simulated root freezing in the nursery: effects on the growth and physiology of containerized boreal...

Simulated root freezing in the nursery: effects onthe growth and physiology of containerized borealconifer seedlings after outplanting

Daniel Dumais, Carole Coursolle, Francine J. Bigras, and Hank A. Margolis

Abstract: The effects of induced root freezing injury on 2+0 white spruce (Picea glauca(Moench) Voss), black spruce(Picea mariana(Mill.) BSP), and jack pine (Pinus banksianaLamb.) seedlings were studied. Hardened seedlings wereexposed to freezing during the fall and cold stored until planting. Seedlings were planted in spring on two field siteswith different soil moisture levels (wet or dry). Seedling morphology and physiology were measured periodically dur-ing the first growing season, and mortality was evaluated at the end of the season. With the exception of June mea-surements on the wet site, where daytime water potential fell as low as –2.0 MPa, root damage did not seriously affectshoot water potential. Generally, stomatal conductance decreased with increasing root damage. Net photosynthesis onboth sites decreased between 22 and 39% with increasing root damage. Root damage did not affect the ratio ofintercellular to ambient CO2 concentration. As well, reductions in the nitrogen concentration of current-year foliagewith increasing root damage were observed, suggesting that the observed reductions in net photosynthesis were causedby nonstomatal factors. Root growth was greater on the wet site than on the dry site, particularly between August andOctober, when mean soil minimum temperatures were lower on the dry site. On both sites, aerial dry mass was onlyslightly affected by root damage in July and August, but the effect of damage became more pronounced in October onthe wet site. Black spruce and white spruce seedling mortality began being affected when approximately 50% of theroot systems were damaged, while jack pine mortality was affected starting at 40% damage. Root damage levels of50% caused 2.0 and 1.5 cm reductions in annual height increment of white spruce and black spruce, respectively, and40% damage caused a reduction of 1.0 cm in annual height increment of jack pine.

Résumé: Les effets des dommages racinaires causés par le gel ont été étudiés chez des semis 2+0 d’épinette blanche(Picea glauca(Moench) Voss), d’épinette noire (Picea mariana(Mill.) BSP) et de pin gris (Pinus banksianaLamb.).Des semis endurcis ont été exposés au gel durant l’automne avant d’être entreposés en chambre froide. Au printempssuivant, ces semis ont été plantés sur deux sites ayant des teneurs en eau du sol distinctes (mouilleux et sec). La mor-phologie et la physiologie des semis ont été mesurées périodiquement durant la première saison de croissance. La mor-talité a été évaluée à la fin de la saison. À l’exception des mesures prises sur le site mouilleux en juin, où des valeursde –2,0 MPa ont été observées, les dommages racinaires n’ont pas sérieusement affecté le potentiel hydrique du xy-lème. Généralement, la conductance stomatique a diminué avec l’augmentation des dommages. Des diminutions de 22à 39% de la photosynthèse nette ont été observées sur les deux sites avec l’augmentation des dommages. Les domma-ges aux racines n’ont pas affecté le ratio de la concentration intercellulaire en CO2 sur la concentration ambiante enCO2, mais des réductions de la teneur en azote du feuillage de l’année courante ont été observées, indiquant que lesdiminutions de la photosynthèse nette ont été causées par des facteurs non-stomatiques. La croissance racinaire a étéplus élevée sur le site mouilleux, particulièrement entre août et octobre alors que les températures minimales moyennesdu sol ont été plus basses sur le site sec. Sur les deux sites, l’effet des dommages racinaires sur la biomasse aérienne aété léger en juillet et en août, mais s’est accentué en octobre sur le site mouilleux. Pour l’épinette noire et l’épinetteblanche, la mortalité a été perceptible lorsque environ 50% du système racinaire était endommagé. Pour le pin gris, lamortalité a été perceptible à partir de 40% de dommages. Des dommages racinaires de 50% ont causé des réductionsde croissance en hauteur de 2,0 et 1,5 cm respectivement pour l’épinette blanche et l’épinette noire. Pour le pin gris,des dommages racinaires de 40% ont provoqué une réduction de croissance en hauteur de 1,0 cm.

Dumais et al.615

IntroductionRoot frost damage can at times be a major problem during

containerized tree seedling production in northern climates

(Lindström 1987), because root systems are considerablyless frost resistant than aboveground tissues (Mityga andLanphear 1971; Sakai and Larcher 1987). In Quebec, the

Can. J. For. Res.32: 605–615 (2002) DOI: 10.1139/X02-005 © 2002 NRC Canada

605

Received 8 June 2001. Accepted 14 December 2001. Published on the NRC Research Press Web site at http://cjfr.nrc.ca on23 March 2002.

D. Dumais, C. Coursolle, and H.A. Margolis.Centre de recherche en biologie forestière, Faculté de Foresterie et de Géomatique,Université Laval, Sainte-Foy, QC G1K 7P4, Canada.F.J. Bigras.1 Natural Resources Canada, Canadian Forest Service, Laurentian Forestry Centre, 1055 du P.E.P.S., P.O. Box 3800,Sainte-Foy, QC G1V 4C7, Canada.1Corresponding author (e-mail: [email protected]).

I:\cjfr\cjfr32\cjfr-04\X02-005.vpWednesday, March 20, 2002 3:50:36 PM

Color profile: Generic CMYK printer profileComposite Default screen

probability of root frost damage is increased, because 87%of the seedlings produced for reforestation are grown in con-tainers (Gagnon et al. 2000) and seedlings are overwinteredoutside. One million seedlings were destroyed by root frostin Quebec forest nurseries in 1994 (Hall 1996). Root sys-tems are particularly vulnerable during autumn when tem-peratures can fall below the hardiness level of the seedlingsand later when there is not enough snow cover to protect theroot system.

In controlled environment experiments, root frost damagehas been found to induce a decrease in shoot and root waterpotential of 1+0 black spruce (Picea mariana(Mill.) BSP)(Bigras and Calmé 1994; Bigras 1997) and jack pine seed-lings (Pinus banksianaLamb.) (Bigras and Margolis 1997).Lindström (1986) reported a decrease in stomatal conduc-tance because of water stress in root-damaged 1+0 Scotspine (Pinus sylvestrisL.) and Norway spruce (Picea abies(L.) Karst.) seedlings, while Langerud et al. (1991) andLangerud and Sandvik (1991) reported decreases in photo-synthesis and transpiration for 1+0 Norway spruce seedlingswith root damage caused by immersion in boiling water.Furthermore, a decrease in aerial growth and survival fol-lowing root frost damage has been noted for 1+0 Norwayspruce (Lindström 1986) and black spruce seedlings (Bigras1997), and for 2+0 white spruce (Picea glauca(Moench)Voss) and jack pine seedlings (Coursolle et al. 2000).Coursolle et al. (2002) reported that root frost damage ef-fects on white spruce, black spruce, and jack pine seedlinggrowth varied with substrate moisture content during agreenhouse experiment. Furthermore, after observing lowmortality in root-damaged 2+0 black spruce seedlingsplanted during a rainy summer, Bigras (1998) hypothesizedthat the soil water regime may have a significant influenceon the performance of planted seedlings.

Excessive root damage is a criterion for rejecting coniferseedlings produced for reforestation, thus causing significantfinancial losses for nursery producers. On the other hand, ac-ceptance of root-damaged seedlings can be risky, since theirperformance might decrease after planting (Ritchie 1990).Considering the importance of root frost damage to refores-tation efforts in northern climates, the objectives of thisstudy were (i) to evaluate the impact of simulated autumnroot frost damage on water relations, gas exchange, mineralnutrition, and shoot and root growth of 2+0 white spruce,black spruce, and jack pine seedlings planted on sites classi-fied as either wet or dry during the first growing season and(ii ) to identify the critical thresholds of root damage beyondwhich significant negative effects on seedling growth wouldbe observed. White spruce, black spruce, and jack pine ac-count for more than 94% of the seedlings planted in Quebec(ministère des Ressources naturelles du Québec 1998).

Materials and methods

Plant materialWhite spruce (WS; seed origin: 46°30′N, 73°15′W), black

spruce (BS; 47°05′N, 72°30′W), and jack pine (JP; 46°30′N,72°24′W) seedlings were grown in production tunnels at aforest tree nursery (Reboisement Mauricie Inc., Saint-Étienne-des-Grès, Que., 46°26′N, 72°47′W, altitude 61 m) inIPL® 45-110 containers (45 cells per container, 110 cm3;IPL®, Saint-Damien, Que.) on a peat moss – vermiculite

substrate (4:1, v/v), following standard practices for seedlingproduction in Quebec (ministère de l’Énergie et desRessources 1990). After two growing seasons, mean shootheights were 23, 22, and 19 cm and mean stem diameterswere 3.0, 2.3, and 2.7 mm for WS, BS, and JP, respectively.In early October 1997, seedlings were transferred to theLaurentian Forestry Centre (Canadian Forest Service,Sainte-Foy, Que., 47°N, 71°W) and placed outdoors. On Oc-tober 27, seedlings were placed in a dark cold room at 2°Cfor 23 days until they were exposed to artificial frosts.

Artificial frost treatmentA preliminary test was conducted to determine the freez-

ing temperatures that would produce five incremental rootdamage classes in the approximate range of 20–100% andreferred to as damage classes 2–6, with the control beingclass 1. Using the method described by Coursolle et al.(2000), seedlings were removed from containers and placedin individual plastic tubes (Super “Stubby” Model, 4 ×13.5 cm, 115 cm3, Ray Leach, Cone-Tainer Nursery, Canby,Oreg.). The aerial portions of the seedlings were placed inheated insulated boxes to protect them from frost. Theseboxes were placed in a programmable cold room (MIC6000, Partlow Corp., New Hartford, N.Y.) at 2°C for 1 h.The temperature was lowered at a rate of 2.5°C·h–1 with 1 hplateaus between each 2.5°C drop in temperature. Seedlingswere sampled at the end of the 1 h plateau at six predeter-mined temperatures ranging between –10 and –35°C. Rootsubstrate temperature was monitored using thermocouplesattached to a data logger (CR21X, Campbell Scientific, Lo-gan, Utah). Control seedlings were placed in a dark coldroom at 2°C for the duration of the frost treatment.

Following the preliminary test, root damage level was as-sessed by washing the root system and separating live rootsfrom damaged roots using a stereomicroscope. Damagedroots were characterized by a brownish discoloration of thecambium and (or) by a mushy cortex. Assessment of rootdamage level was calculated using the dry mass of damagedroots as a percentage of total root dry mass. Temperaturesselected for damage classes 2 to 6 were as follows: –10.0,−12.5, –15.0, –20.0, and –22.5°C for WS, –12.5, –15.0,−17.5, –20.0, and –22.5°C for BS, and –10.0, –12.5, –15.0,−17.5, and –20.0°C for JP.

On November 18, seedlings were exposed to an artificialfrost 19 days after the preliminary test, using the method de-scribed above and temperatures determined from the prelim-inary test. Following the frost application, sampled seedlingswere placed in a dark cold room at 2°C for thawing. After48 h, seedlings designated for planting were placed in card-board boxes and stored in a dark cold room at 0°C for190 days until they were transported to the planting sites. Inaddition, five seedlings per species per damage class wereplaced in a growth chamber (20°C day : 15°C night; 80%humidity; 16 h light : 8 h dark photoperiod) for 15 days andthen evaluated separately for root damage using the previ-ously mentioned method, and these results are presented aspreplanting root damage.

Planting sitesSeedlings were planted on May 27–28, 1998, with 1-m

spacing, on wet and dry sites at the Forêt Montmorency(Université Laval’s forest research station located 60 km

© 2002 NRC Canada

606 Can. J. For. Res. Vol. 32, 2002



north of Québec) (47°N, 71°W, altitude 750 m). The wet sitesoil (sub-hygric) has a sandy loam texture with imperfectdrainage (Ontario Institute of Pedology 1985) and had 50%(v/v) water content as determined by time domainreflectometry measurements in June (Moisture Point, MP-917 model, Environmental Sensors, Victoria, B.C.). The drysite soil (xeric) has a gravelly sandy texture with good drain-age and had 30% (v/v) water content in June.

During the summer of 1998, soil water content was mea-sured gravimetrically on 23 July, 4 and 20 August and ex-pressed on a volume basis. Air temperatures at 20 cm abovesoil level and soil temperatures at 7 cm below soil level weremonitored at the two sites using CR-10 data loggers (Camp-bell Scientific Corp., Edmonton, Alta.). Rainfall was mea-sured at the Forêt Montmorency weather station locatednearby.

Physiology

Xylem water potentialThe predawn shoot water potential (Ψpd) was measured

between 00:00 and 3:00 at the end of June, July, and Augustusing a pressure chamber (Model 600, PMS InstrumentsCo., Corvallis, Oreg.). Daytime shoot water potential (Ψd)was measured between 11:00 and 15:00 at the end of June,July, and August on the same day and on the same seedlingsused for gas-exchange measurements.

Gas-exchange measurementsNet photosynthesis (An), stomatal conductance (gs), and the

ratio of intercellular to ambient CO2 concentration (Ci/Ca)were measured at the end of June, July, and August using a0.25-L chamber attached to a portable photosynthesis system(Model Li-6200, LI-COR Inc., Lincoln, Nebr.). The mea-surements were taken on the 3–4 cm length of shoot directlybelow the terminal bud of the main stem or, on the previousyear’s terminal shoot only, if the new shoot was not longenough. Measurements were taken at 1000µmol·m–2·s–1

photosynthetically active radiation using a lamp with light-emitting diodes (670 nm, Quantum Devices Inc., Barneveld,Wis.) specially adapted to fit the LI-COR 0.25-L chamberlid. Net photosynthesis and stomatal conductance measure-ments were expressed on a surface-area basis. Flow was ad-justed for each seedling so that relative humidity was keptconstant (40–65%) during the measurement and reflected theenvironmental conditions for the day of sampling. The sur-face area (SA) was calculated (SA = 4.00(VL)0.5 for WS andBS and 4.59(VL)0.5 for JP) using volume displacement (V)(Brand 1987) and needle length (L) (WinRHIZOTM, RégentInstruments Inc., Québec, Que.). Net photosynthesis wasalso calculated on a dry-mass basis so that it could be re-lated to nutrient concentrations, which are expressed on adry-mass basis. For logistical reasons, seedlings from dam-age classes 2 and 6 were not included in the gas-exchangemeasurements.

Nutrient concentrationsNitrogen, phosphorus, and potassium concentrations of

previous-year and current-year foliage were assessed at theend of July and in mid-October for seedlings in damageclasses 1, 3, 4, and 5. The foliage was digested followingprocedures described in Parkinson and Allen (1975). Nitro-

gen concentration was evaluated using the Kjeldahl method(Bremner and Mulvaney 1982) and phosphorus and potas-sium concentrations by atomic emission spectrometry(Perkin-Elmer Plasma Model 40, Perkin-Elmer Corp.,Norwalk, Conn.).

Seedling growth and mortalityInitial seedling height and stem diameter were measured

before planting. Live root dry mass (LIRDM) and aerial drymass (ADM) were assessed at the end of June, July, and Au-gust and in mid-October. The measurements in June, July,and August were made using the same seedlings as for thegas-exchange measurements. At the end of the growing sea-son, in October, annual height increment (AHI) and finalstem diameter (SD) were measured, and seedling mortalitywas evaluated. Seedlings were classified as dead if they haddry, yellow foliage.

Experimental design and statistical analysisThe experimental design was randomized independently

at each site, although the basic experimental design was thesame. At each site, three complete repetitions had 18 plotseach arranged in a 3 × 6rectangle (i.e., six plots per row),giving a total of 54 plots. In each of the three rows, eachspecies was assigned to two of the six plots, and each dam-age class was assigned to one of the six plots, giving twodamage classes per species for each row. Thus, all 18 combi-nations of the 3 species × 6 damage classes were present inthe repetition. This is known as an incomplete block design,where the incomplete blocks are the rows of six plots withineach complete repetition (Cochran and Cox 1957). The pur-pose of having the smaller incomplete blocks within thelarger repetitions is to improve the precision of comparisonsbetween the means of the species and the damage classes.There were 12 seedlings per plot for a total of 648 seedlingsper site (3 complete repetitions × 3 species × 6 damageclasses × 12 seedlings).

Three seedlings per plot, one per sampling date, wereused to measure daytime water potential, net photosynthesis,stomatal conductance, and the ratio of intercellular to ambi-ent CO2 concentration. For these variables, the design thusbecame a split-plot, incomplete-block design with the threedates as the subplots, i.e., one seedling per subplot. Threeother seedlings per plot, again one per sampling date, wereused to measure predrawn water potential, yielding a split-plot design for this response variable as well. Annual heightincrement was measured on six seedlings per plot in mid-October. Thus, the design for this response variable was thebasic incomplete block but with six subsamples per plot.Nutrient concentrations were measured on the same seed-lings as daytime water potential at the end of July and onone of the six seedlings used to measure annual shootgrowth in mid-October. The design for this response variablewas a split plot with the two dates as the subplots. Live rootdry mass and aerial dry mass were evaluated on the sameseedlings as daytime water potential on the three summersampling dates (June, July, and August), and on the sameseedlings as nutrient concentrations in mid-October. Thus,there were four sampling dates in the subplots for this re-sponse variable. The five remaining seedlings were simplycoded as dead or alive at the end of the growing season.

© 2002 NRC Canada

Dumais et al. 607



The statistical model used for the analysis of variance(ANOVA) was

Response variable = constant + R + B[R] + S

+ C + C × S +between-plot error + D

+ S × D + C × D + S × C × D

+ within-plot error

where R is repetition, B is block, S is species, C is damageclass, and D is date. The terms for date and its interactionswith species and damage class disappear if the response wasonly measured once, and the “within plot error” becomes an“among seedling within plot error” if the response was mea-sured on more than one seedling on any one measurementoccasion. The square brackets indicate that the block (B)was nested within the repetition. The effects of repetitionsand blocks, the between-plot error, and the within-plot errorwere considered random, while those of species, damageclass, date, and their interactions were considered fixed.Species effects were partitioned into orthogonal contrasts be-cause they are discrete factors, while damage class and dateeffects were partitioned into orthogonal polynomial contrastsbecause they are continuous factors. The contrasts for the in-teractions between species and damage class were parti-tioned by multiplying the coefficients of the respectiveorthogonal and orthogonal polynomial matrices. Responsecurves in figures were generated using the significant poly-nomial contrast for each variable.

The statistical model was fit with the MIXED procedureof SAS (Littell et al. 1996). Initial seedling height and diam-eter were used as covariates for annual height increment anddiameter. Dead seedlings were excluded from the statisticalanalyses, except for live root dry mass, a variable directly af-fected by root frost. To satisfy ANOVA assumptions, asquare root transformation was used for aerial dry mass anda natural logarithmic transformation for predawn water po-tential. Only significant effects (p ≤ 0.05) are illustrated. Nostatistical analysis was conducted on seedling mortality be-cause of the small number of seedlings sampled for eachdamage class and species combination (n = 15). Only datarelated to root damage effects are shown, since this was themain objective of the study.

ResultsPreplanting root damage

The frost treatments resulted in a gradient of root damage.However, root damage levels for each class differed with re-spect to species because of their different inherent frost tol-erance levels and acclimation rates. Preplanting root damagefor classes 1–6 were 3, 27, 37, 48, 59, and 65% for whitespruce; 1, 38, 54, 63, 69, and 77% for black spruce; and 5,42, 55, 70, 80, and 85% for jack pine.

Meteorological dataSoil water content (v/v) was 37.7, 56.1, and 54.8% for the

wet site and 28.6, 38.1, and 37.7% for the dry site onJuly 23, August 4, and August 20, respectively. Mean maxi-mum seasonal air and soil temperatures were 2.6 and 3.9°Chigher on the dry site compared with the wet site (Table 1).On the other hand, mean minimum seasonal temperatureswere 3.3 and 0.2°C lower on the dry site compared with thewet site (Table 1). No freezing temperatures at 20 cm abovesoil level were recorded between June 3 and September 19for the wet site, and between June 10 and August 14 for thedry site (107 and 64 days without frost, respectively). Rain-fall, as measured at the nearby weather station, totaled636 mm.

Wet site

PhysiologyPredawn water potential decreased between classes 1 and

4 in June and between classes 3 and 6 in July and August(Fig. 1a, Table 2,p = 0.0165). Predawn water potential wasmore negative in June (mean –0.60 MPa) than on the otherdates (mean –0.28 MPa). Daytime water potential decreasedfrom classes 3 to 6 in June and showed little change in Julyand August (Fig. 1b, Table 2,p = 0.0117). The most nega-tive daytime water potential occurred in June (–2.0 MPacompared with –0.78 MPa for July and August).

Stomatal conductance decreased with increasing rootdamage (Fig. 1c, Table 2,p = 0.0144). Net photosynthesisdecreased by 18, 24, and 33% relative to the control forclasses 3, 4, and 5, respectively, (Fig. 1d, Table 2, p =0.0482). Root damage did not affect the ratio of intercellularto ambient CO2 concentration of seedlings and averaged0.79 (Table 2,p ≥ 0.0607, results not shown).

The nitrogen concentration of previous-year foliage wasnot affected by root damage (Table 3,p ≥ 0.0801, results notshown). However, the current-year nitrogen concentration

© 2002 NRC Canada

608 Can. J. For. Res. Vol. 32, 2002

Wet site Dry site

June July Aug. Sept. Season June July Aug. Sept. Season

Air temperature (°C) (20 cm above soil surface)Maximum 22.1 24.9 25.1 17.0 22.4 26.9 26.9 27.4 18.4 25.0Minimum 7.8 8.8 7.5 4.4 7.2 6.0 4.3 3.6 0.6 3.9Mean 15.0 16.9 16.3 10.7 14.8 16.5 15.6 15.5 9.5 14.5Soil temperature (°C) (7 cm below soil surface)Maximum 11.8 14.4 14.4 11.4 13.1 15.5 17.7 20.6 15.3 17.0Minimum 7.6 10.1 10.1 8.4 9.1 11.1 10.9 8.2 4.9 8.9Mean 9.7 12.3 12.3 9.9 11.1 13.3 14.3 14.4 10.1 13.0

Table 1. Mean monthly maximum, minimum, and mean air and soil temperatures for the wet and drysites, as well as for the overall 1998 growing season.



decreased with increasing root damage (Table 3,p = 0.0311,Fig. 1e). Root damage did not affect the phosphorus orpotassium concentrations of either the previous-year orcurrent-year foliage (results not shown).

Seedling growth and mortalityLive root dry mass decreased with increasing damage in

June, July, and August and to an even greater extent in Octo-ber (Fig. 2a, Table 4,p = 0.0499). The difference in live root

© 2002 NRC Canada

Dumais et al. 609

WS

BS

JP

June

July

Aug.

1 2 3 4 5 6

(f)

(g)

(h)

(i)

(k)

JP

BS

WS

(j)

(a)

Ψpd

(MP

a)

Ψd

(MP

a)

gs

2-2

-1(m

olH

Om

s)

An 2

-1-1

(µm

olC

Og

s)

-0.8

-0.6

-0.4

-0.2

0.0

JuneJuly

Aug.

-2.0

-1.5

-1.0

-0.5

0.0

June

July

Aug.

0.02

0.04

0.06

0.08

N-1

(m

gg

)

0.03

0.04

0.05

0.06

0.07

14

16

18

20

12

(b)

(c)

(d)

(e)

1 2 3 4 5 6

gs

2-2

-1(m

olH

Om

s)

0.02

0.04

0.06

0.08

Wet site Dry site

··

··

··

·

Root damage class

Fig. 1. Relationship between root frost damage, induced during the fall on hardened seedlings, and physiological variables measuredthe following growing season, after planting in the spring, on the wet and dry sites. (a and f) Predawn water potential (Ψpd). Maximumstandard error (MSE) was not available because variable was transformed. Values are the means of nine seedlings for the wet site andof 27 seedlings for the dry site. (b and g) Daytime water potential (Ψd), MSE was 0.2. Values are the means of nine seedlings. (c, h,and i) Stomatal conductance (gs). MSE was 0.006 for the wet site and 0.008 for the dry site. Values are the means of 27 seedlings forthe wet site and nine seedlings for the dry site. (d and j) Net photosynthesis (An). MSE was 0.007 for the wet site and 0.004 for thedry site. Values are the means of 27 seedlings. (e and k) Nitrogen concentration (N) of current-year foliage. MSE was 2 and 1 for thewet and dry sites, respectively. Values are the means of 18 seedlings. Only significant effects (p ≤ 0.05) are shown. Preplanting rootdamage for classes 1–6, respectively, were 3, 27, 37, 48, 59, and 65% for white spruce (WS); 1, 38, 54, 63, 69, and 77% for blackspruce (BS); and 5, 42, 55, 70, 80, and 85% for jack pine (JP).



© 2002 NRC Canada

610 Can. J. For. Res. Vol. 32, 2002

Wa

ter

po

ten

tial

Ga

se

xch

an

ge

So

urc

eo

fva

ria

tiona

dfn

bd

fdb

We

tsi

teΨ

pdc

We

tsi

teΨ

dcD

rysi

teΨ

pd

Dry

site

Ψd

dfn

dfd

We

tsi

teg s

dW

et

site

And

We

tsi

teC

i/Cad

Dry

site

g s

Dry

site

An

Dry

site

Ci/C

a

Fix

ed

eff

ect

s(p

>F

)S

pe

cie

s(S

)2

93

.3≤0

.00

01

≤0.0

00

10

.34

44

≤0.0

00

12

54

.3≤0

.00

01

0.0

28

10

.77

50

0.0

16

2≤0

.00

01

0.1

39

6W

Svs

.B

S(1

)9

3.3

0.8

84

70

.08

11

0.3

21

10

.21

25

(1)

54

.40

.31

22

0.0

40

30

.50

51

0.1

65

00

.00

27

0.3

92

8JP

vs.

spru

ces

(SP

)(1

)9

3.2

≤0.0

00

1≤0

.00

01

0.2

88

9≤0

.00

01

(1)

54

.3≤0

.00

01

0.0

62

10

.78

06

0.0

13

90

.00

02

0.0

66

5D

am

ag

ecl

ass

(C)

59

3.1

0.2

40

10

.00

04

0.2

01

80

.82

18

35

4.2

0.0

61

40

.22

52

0.9

75

40

.03

95

≤0.0

00

10

.48

05

Clin

ea

r(1

)9

3.1

0.0

26

10

.00

03

0.0

26

60

.53

36

(1)

54

.60

.01

44

0.0

48

20

.73

10

0.0

36

6≤0

.00

01

0.8

55

2C

qu

ad

ratic

(1)

93

.10

.19

34

0.1

33

90

.33

66

0.5

61

1(1

)5

4.1

0.6

46

90

.89

87

0.7

29

20

.68

98

0.9

49

70

.13

10

S×

C1

09

3.1

0.7

88

60

.75

87

0.7

71

90

.11

15

65

4.1

0.4

91

90

.96

69

0.4

10

60

.06

45

0.5

13

70

.89

71

(WS

vs.

BS

)×

Clin

ea

r(1

)9

3.1

0.6

58

00

.57

17

0.7

96

10

.27

28

(1)

54

.00

.31

93

0.8

38

80

.98

16

0.0

45

10

.25

99

0.9

51

8(W

Svs

.B

S)

×C

qu

ad

ratic

(1)

93

.00

.28

68

0.7

84

10

.51

90

0.8

60

3(1

)5

4.4

0.3

03

00

.51

34

0.3

48

30

.64

36

0.5

64

50

.99

27

(JP

vs.

SP

)×

Cline

ar

(1)

93

.40

.79

53

0.2

69

10

.62

60

0.0

23

5(1

)5

4.2

0.3

67

10

.54

54

0.3

13

50

.08

44

0.1

25

30

.99

53

(JP

vs.

SP

)×

Cq

ua

dra

tic(1

)9

3.3

0.4

76

60

.24

00

0.7

78

10

.68

96

(1)

53

.90

.80

66

0.5

61

80

.06

07

0.4

39

80

.93

76

0.5

55

5D

ate

(D)

29

3.1

≤0.0

00

1≤0

.00

01

0.0

00

3≤0

.00

01

25

4.0

0.3

48

40

.87

29

0.0

00

10

.00

86

0.0

01

60

.17

28

Dlin

ea

r(1

)9

3.1

≤0.0

00

1≤0

.00

01

0.8

33

4≤0

.00

01

(1)

53

.90

.22

85

0.7

36

40

.01

48

0.0

99

50

.00

04

0.6

86

4D

qu

ad

ratic

(1)

93

.10

.01

34

≤0.0

00

1≤0

.00

01

0.3

83

1(1

)5

4.0

0.4

10

30

.68

58

0.0

00

20

.00

78

0.8

08

00

.06

85

S×

D4

93

.10

.39

09

≤0.0

00

10

.25

19

0.0

00

34

54

.0≤0

.00

01

0.7

05

00

.00

37

≤0.0

00

10

.52

23

0.0

74

8(W

Svs

.B

S)

×D

line

ar

(1)

93

.00

.07

82

0.0

00

90

.38

76

0.8

94

4(1

)5

3.9

0.0

00

20

.33

22

0.5

38

7 ≤0.0

00

10

.45

57

0.0

13

4(J

Pvs

.S

P)

×D

line

ar

(1)

93

.20

.62

93

≤0.0

00

10

.07

70

0.4

39

2(1

)5

4.0

≤0.0

00

10

.33

75

0.0

00

20

.03

21

0.3

60

50

.32

91

(WS

vs.

BS

)×

Dq

ua

dra

tic(1

)9

3.3

0.3

62

4≤0

.00

01

0.4

74

40

.03

18

(1)

54

.0≤0

.00

01

0.9

10

10

.65

53

0.0

00

50

.26

00

0.0

83

0(J

Pvs

.S

P)

×D

qu

ad

ratic

(1)

93

.10

.99

83

0.0

00

40

.46

51 ≤0

.00

01

(1)

54

.10

.00

03

0.8

61

70

.73

04

0.0

18

90

.55

89

0.9

25

8C

×D

10

93

.00

.23

20

0.1

73

60

.82

41

0.7

13

76

54

.00

.76

68

0.5

94

40

.95

73

0.2

20

00

.78

55

0.5

33

0C

line

ar×

Dlin

ea

r(1

)9

3.1

0.6

38

00

.20

82

0.6

40

70

.34

83

(1)

54

.10

.67

17

0.7

49

10

.57

91

0.1

74

20

.72

17

0.4

86

1C

qu

ad

ratic

×D

line

ar

(1)

93

.00

.01

65

0.0

11

70

.71

63

0.5

33

2(1

)5

4.1

0.4

34

40

.66

26

0.6

59

60

.32

10

0.5

64

30

.30

21

Clin

ear

×D

qu

ad

ratic

(1)

93

.00

.20

35

0.1

01

10

.14

38

0.6

22

3(1

)5

3.9

0.1

81

50

.51

67

0.3

93

30

.03

29

0.2

28

80

.21

43

Cq

ua

dra

tic×

Dq

ua

dra

tic(1

)9

3.0

0.3

05

10

.79

05

0.6

34

30

.84

66

(1)

54

.00

.61

60

0.4

09

20

.60

36

0.3

00

10

.95

85

0.3

93

2S

×C

×D

20

93

.10

.48

83

0.2

72

00

.46

78

0.3

27

01

25

4.0

0.3

90

30

.98

49

0.8

86

10

.42

81

0.1

92

10

.82

92

Ra

nd

om

eff

ect

s(p

>|Z

|)R

ep

etit

ion

(R)

0.4

40

50

.47

86

na

en

a0

.44

57

0.5

76

20

.54

52

na

na

na

Blo

ck(B

)[R

]n

an

a0

.95

17

0.6

89

8n

an

a0

.99

40

na

na

0.3

22

6S

×C

×B

[R]

na

na

na

0.89

26

na

0.1

92

8n

an

an

an

aR

esi

du

al

≤0.0

00

1≤0

.00

01

≤0.0

00

1≤0

.00

01

≤0.0

00

1≤0

.00

01

≤0.0

00

1≤0

.00

01

≤0.0

00

1≤0

.00

01

a WS

,w

hite

spru

ce;

BS

,bl

ack

spru

ce;

JP,

jack

pine

.b df

n,de

gree

sof

free

dom

ofnu

mer

ator

;df

d,de

gree

sof

free

dom

ofde

nom

inat

or;

dfd

ofdr

ysi

teΨ

pdva

ried

from

15.5

to73

.0; Ψ

ddf

dva

ried

from

82.0

to82

.8fo

rth

ew

etsi

tean

dfr

om14

.5to

53.0

for

the

dry

site

;df

dof

dry

siteg

sw

as45

.0;A

ndf

dva

ried

from

16.6

to36

.9fo

rth

ew

etsi

tean

dw

as45

.0fo

rth

edr

ysi

te;

Ci/C

adf

dva

ried

from

15.7

to53

.3fo

rth

ew

etsi

tean

dfr

om36

.1to

46.1

for

the

dry

site

.df

nva

lues

inpa

rent

hese

sre

fer

toco

ntra

sts.

c Ψpd

,pr

edaw

nw

ater

pote

ntia

l; Ψd

,da

ytim

ew

ater

pote

ntia

l.d A

n,

net

phot

osyn

thes

is;g s

,st

omat

alco

nduc

tanc

e;Ci/C

a,ra

tioof

inte

rcel

lula

rto

ambi

ent

CO 2

conc

entr

atio

n.e na

,p>

|Z|

was

not

avai

labl

ebe

caus

eof

anes

timat

eeq

ual

to0.

Tabl

e2.

An

aly

ses

of

vari

an

cea

nd

ass

oci

ate

dp

rob

ab

ilitie

sfo

rw

ate

r-p

ote

ntia

la

nd

gas-

exch

an

ge

me

asu

rem

en

ts.



dry mass between classes 1 and 6 increased from 0.5 g inJune to 1.2 g in October.

Aerial dry mass decreased slightly with increasing dam-age in July and August and much more significantly in Oc-tober (Fig. 2b, Table 4,p = 0.0009). The difference in aerialdry mass between classes 1 and 6 increased from 0.41 g inJuly to 2.95 g in October. The decrease in annual height in-crement with increasing root damage was greater for JP thanfor the two spruces (Fig. 2c, Table 4,p = 0.0254). The mostsevere root damage caused annual height increment reduc-tions of 20% for BS, 50% for WS, and 42% for JP. Rootdamage reduced shoot diameter by as much as 12% forclass 6 (Fig. 2d, Table 4, p = 0.0069) compared with thecontrol.

BS and WS mortality only started being affected by rootdamage at classes 5 and 6, respectively, (Table 5), withclass 5 and 6 mortality being nearly equal for BS. On theother hand, JP mortality started at class 4 and increased withincreasing root damage (Table 5).

Dry site

PhysiologyPredawn water potential decreased slightly with increas-

ing root damage (Fig. 1f, Table 2,p = 0.0266). Root damagehad little effect on WS and BS daytime water potential,while JP pine daytime water potential decreased fromclasses 1 to 6 (Fig. 1g, Table 2,p = 0.0235).

Stomatal conductance decreased with increasing rootdamage in June and August (Fig. 1h, Table 2,p = 0.0329).Stomatal conductance of WS decreased from classes 1 to 6,whereas stomatal conductance of BS and JP showed littlevariation (Fig. 1i, Table 2,p = 0.0451). Net photosynthesisdecreased by 22, 26, and 39% relative to the control forclasses 3, 4, and 5 (Fig. 1j, Table 2,p ≤ 0.0001). Root dam-age did not affect the ratio of intercellular to ambient CO2concentration of seedlings and averaged 0.76 (Table 2,p ≥0.1310, results not shown).

The nitrogen concentration of previous-year foliage wasnot significantly affected by root damage (Table 3,p ≥0.0908, results not shown), whereas current-year nitrogenconcentration decreased with increasing root damage(Fig. 1k, Table 3,p = 0.0265). Root damage did not affectthe phosphorus or potassium concentrations of either theprevious-year or current-year foliage (results not shown).

Seedling growth and mortalityGenerally, live root dry mass decreased as root damage in-

creased for all dates (Fig. 2e, Table 4,p = 0.0133). The dif-ference between the control and class 6 was 0.5 g in June,0.5 g in July, 0.5 g in August, and 0.6 g in October.

Aerial dry mass exhibited only very slight decreases withincreasing root damage in July, August, and October(Fig. 2f, Table 4,p = 0.0048). The difference between thecontrol and class 6 was 0.8, 1.0, and 0.5 g in July, August,and October, respectively. Annual height increment fromclasses 2 to 6 decreased by 14, 24, 33, 43, and 52% relative

© 2002 NRC Canada

Dumais et al. 611

Previous-year foliage Current-year foliage

Source of variationa dfnb dfdb Wet site Dry site Wet site Dry site

Fixed effects (p >F)Species (S) 2 21.8 0.0040 ≤0.0001 0.0012 ≤0.0001WS vs. BS (1) 21.6 0.0030 0.0001 0.0029 ≤0.0001JP vs. spruces (SP) (1) 22.0 0.0945 0.0006 0.0111 0.0080Damage class (C) 3 21.7 0.0801 0.5670 0.1297 0.1402C linear (1) 21.5 0.0850 0.5993 0.0331 0.0265C quadratic (1) 22.2 0.3417 0.4772 0.7705 0.6046S × C 6 21.7 0.4644 0.1313 0.9499 0.7119(WS vs. BS) × C linear (1) 21.8 0.4910 0.2939 0.7535 0.2075(WS vs. BS) × C quadratic (1) 21.7 0.1362 0.7767 0.8418 0.9974(JP vs. SP) × C linear (1) 21.2 0.4779 0.0908 0.6385 0.2450(JP vs. SP) × C quadratic (1) 22.5 0.5932 0.3729 0.8061 0.6001Date (D) 1 22.8 ≤0.0001 ≤0.0001 ≤0.0001 ≤0.0001S × D 2 22.8 0.0071 0.0566 0.4338 0.3106C × D 3 22.8 0.1329 0.6200 0.1450 0.3903S × C × D 6 22.7 0.6295 0.3354 0.2620 0.7677

Random effects (p >|Z|)Repetition (R) 0.4031 0.6197 0.5082 nac

Block (B) [R] na 0.4098 na 0.9002S × C × B [R] 0.1405 0.6041 0.0681 naResidual 0.0009 0.0030 0.0087 0.0053aWS, white spruce; BS, black spruce; JP, jack pine.bdfn, degrees of freedom of numerator; dfd, degrees of freedom of denominator; dry site previous-year

foliage dfd varied from 16.1 to 24.1; current-year foliage dfd varied from 16.0 to 18.6 for the wet site andfrom 18.1 to 21.7 for the dry site. dfn values in parentheses refer to contrasts.

cna, p >|Z| not available because of an estimate equal to 0.

Table 3. Analyses of variance and associated probabilities of nitrogen concentration for previous-year and current-year foliage.



to the control (Fig. 2g, Table 4,p ≤ 0.0001). Root damagereduced shoot diameter by as much as 21% for class 6(Fig. 2h, Table 4,p ≤ 0.0001).

Seedling mortality of BS started at class 4 and increasedwith increasing damage thereafter (Table 5). Root damagecaused WS seedling mortality only at damage class 6 and JPseedling mortality increased with increasing root damage.

Discussion

With the exception of June measurements on the wet site,where predawn water potential values fell to –0.7 MPa, rootdamage did not seriously affect predawn water potential(Figs. 1a and 1f). This implies that in most cases the seed-lings were able to reabsorb water at night. Damaged rootsprobably had little effect on water absorption, since the re-maining suberized roots, which are generally more frost tol-erant than fine roots (Lindström and Mattsson 1989), could

still take up a significant amount of water and solutes(Kramer 1983, p. 134). The fact that the planting sites werein an area with abundant summer precipitation may explainthis result. It is also possible that the remaining healthy rootsincreased their water absorption capacity or regeneratedenough new roots to avoid negative effects on predawn wa-ter potential. Except for June, daytime water potential alsoshowed little variation with respect to root damage.Coursolle et al. (2002) also reported that WS, BS, and JPdaytime water potential did not vary with frost-induced rootdamage. However, a reduction in water potential of 1+0 BSand JP seedlings following root frost was observed byBigras and Margolis (1997) and by Bigras (1997).

Generally, net photosynthesis and stomatal conductancedecreased with increasing root damage (Figs. 1c, 1d, and1h–1j). Langerud et al. (1991) reported that 1+0 Norwayspruce seedlings with root systems immersed in boiling wa-ter showed a reduction in photosynthesis and transpiration.

© 2002 NRC Canada

612 Can. J. For. Res. Vol. 32, 2002

1 2 3 4 5 6

LIR

DM

(g)

AD

M

(g)

0.0

2.0

4.0

6.0

0.0

0.4

0.8

1.2

1.6

2.0

(b)

(a)

(c)

3.0

3.5

4.0

4.5

5.0

0

5

10

15

20

(d)

AH

I

(cm

)

SD

(mm

)

JuneJuly

Aug.

Oct.

JuneJuly

Aug.

Oct.

JP

WS

BS

Wet site

(f)

(e)

(g)

(h)

1 2 3 4 5 6

June July

Aug.Oct.

June

Oct.

July

Aug.

Dry site

Root damage class

Fig. 2. Relationship between root frost damage, induced during the fall on hardened seedlings, and growth variables measured the fol-lowing growing season, after planting in the spring, on the wet and dry sites. (a and e) Live root dry mass (LIRDM). Maximum stan-dard error (MSE) was 0.2. Values are the means of three seedlings. (b and f) Aerial dry mass (ADM). MSE was not available becausethe variable was transformed. Values are the means of three seedlings. (c and g) Annual height increment (AHI). MSE was 2 for thewet site and 1 for the dry site. Values are the means of 18 seedlings for the wet site and 54 seedlings for the dry site. (d and h) Stemdiameter (SD). MSE was 0.3 and 0.2 for the wet and dry sites, respectively. Values are the means of 54 seedlings. Only significant ef-fects (p ≤ 0.05) are illustrated. Preplanting root damage for classes 1–6, respectively, were 3, 27, 37, 48, 59, and 65% for white spruce(WS); 1, 38, 54, 63, 69, and 77% for black spruce (BS); 5, 42, 55, 70, 80, and 85% for jack pine (JP).



Lindström (1986) demonstrated that decreases in stomatalconductance associated with water stress were induced byroot damage in 1+0 Norway spruce seedlings. Several stud-ies with conifer species have established a link between wa-ter potential, stomatal conductance, and photosynthesis(Seiler and Johnson 1985; Grossnickle and Blake 1986;Seiler and Cazell 1990; Dang et al. 1997a). A decrease inwater potential may reduce photosynthesis by inducingstomatal closure and a reduction in intercellular CO2 con-centration (Teskey et al. 1995) or by having a direct impacton mesophyll conductance. On the other hand, the ratio ofintercellular to ambient CO2 concentration was not reducedin this study, and root damage had little effect on either pre-dawn or daytime water potential. Therefore, the observed re-

ductions in net photosynthesis were probably caused bynonstomatal factors, such as lower mesophyll conductanceand reduced CO2 fixation by the mesophyll.

The reduced ability to fix CO2 in root-damaged seedlingsmay be the result of the reduced nitrogen levels observed forseedlings with damaged root systems (Figs. 1e and 1k).Many authors have described a positive correlation betweennet photosynthesis and foliage nitrogen concentration (e.g.,Mitchell and Hinckley 1993; Reich et al. 1995; Dang et al.1997b; Schoettle and Smith 1999). Reductions in foliar ni-trogen concentration caused by root frost damage are proba-bly related to decreases in root surface area and (or) to thereduction in root system exploration after injury, whichwould have led to reduced nitrogen absorption. Unlike nitro-

© 2002 NRC Canada

Dumais et al. 613

Wet site Dry site

Source of variationa dfnb dfdb LIRDM c ADM c AHI c SDc LIRDM ADM AHI SD

Fixed effects (p >F)Covariated 1 —e — — nsf ≤0.0001 — — ns ≤0.0001Species (S) 2 28.0 0.1183 0.3553≤0.0001 0.0003 0.0001 ≤0.0001 0.0003 0.0025WS vs. BS (1) 28.0 0.5269 0.5316 ≤0.0001 ≤0.0001 0.0381 ≤0.0001 0.0005 0.0015JP vs. spruces (SP) (1) 28.0 0.0495 0.2029≤0.0001 0.6381 ≤0.0001 0.0004 0.0327 0.0696Damage class (C) 5 28.0 ≤0.0001 0.0045 0.0017 0.0700 ≤0.0001 0.0926 0.0011 0.0006C linear (1) 28.0 ≤0.0001 ≤0.0001 ≤0.0001 0.0069 ≤0.0001 0.0096 ≤0.0001 ≤0.0001C quadratic (1) 28.0 0.8086 0.8441 0.7130 0.6356 0.3284 0.1191 0.1697 0.1794S × C 10 28.0 0.7537 0.3495 0.1979 0.7033 0.4873 0.4410 0.3065 0.5365(WS vs. BS) × C linear (1) 28.0 0.9126 0.2199 0.6485 0.2396 0.1896 0.8613 0.5170 0.4936(WS vs. BS) × C quadratic (1) 28.0 0.6040 0.8583 0.6189 0.5374 0.9864 0.9932 0.1923 0.6421(JP vs. SP) × C linear (1) 28.0 0.2427 0.1921 0.0254 0.4910 0.1462 0.1347 0.0722 0.0947(JP vs. SP) × C quadratic (1) 28.0 0.5573 0.0542 0.1605 0.2267 0.7320 0.4642 0.7146 0.8409Date (D) 3 104.0 ≤0.0001 ≤0.0001 — — ≤0.0001 0.0012 — —D linear (1) 104.0 ≤0.0001 ≤0.0001 — — ≤0.0001 0.0003 — —D quadratic (1) 104.0 0.0417 0.9107 — — 0.6051 0.1954 — —S × D 6 104.0 0.7001 0.3048 — — 0.1261 0.4176 — —(WS vs. BS) × D linear (1) 104.0 0.0993 0.0289 — — 0.3623 0.3792 — —(WS vs. BS) × D quadratic (1) 104.0 0.7057 0.9951 — — 0.0816 0.5805 — —(JP vs. SP) × D linear (1) 104.0 0.4810 0.5449 — — 0.0285 0.0504 — —(JP vs. SP) × D quadratic (1) 104.0 0.7126 0.8325 — — 0.3409 0.3799 — —C × D 15 104.0 0.7852 0.3349 — — 0.0328 0.1079 — —C linear × D linear (1) 104.0 0.0499 0.0009 — — 0.7919 0.2248 — —C quadratic × D linear (1) 104.0 0.2020 0.7006 — — 0.6891 0.3376 — —C linear × D quadratic (1) 104.0 0.2412 0.4792 — — 0.5376 0.0048 — —C quadratic × D quadratic (1) 104.0 0.8267 0.5456 — — 0.1218 0.2360 — —C cubic × D quadratic (1) 104.0 0.6138 0.9691 — — 0.0133 0.7476 — —S × C × D 30 104.0 0.6233 0.9501 — — 0.6182 0.4576 — —

Random effects (p >|Z|)Repetition (R) — — 0.2041 0.4576 0.6056 0.4069 0.1951 nag 0.7522 0.5381Block (B) [R] — — 0.4032 0.8535 na 0.6591 na 0.3554 0.2750 0.5779S × C × B [R] — — 0.2564 0.3256 0.0807 0.0174 0.1057 0.5116 0.0616 0.0471Residual — — ≤0.0001 ≤0.0001 ≤0.0001 ≤0.0001 ≤0.0001 ≤0.0001 ≤0.0001 ≤0.0001aWS, white spruce; BS, black spruce; JP, jack pine.bdfn, degrees of freedom of numerator; dfd, degrees of freedom of denominator; dry site LIRDM dfd varied from 28.5 to 105.0; ADM dfd varied from

21.5 to 106.0 for the wet site and from 29.1 to 108.0 for the dry site. AHI dfd varied from 28.0 to 34.4 for the wet site and from 22.2 to 44.7 for the drysite; SD dfd varied from 26.1 to 50.0 for the wet site and from 21.5 to 49.2 for the dry site. dfn values in parentheses refer to contrasts.

cLIRDM, live root dry mass; ADM, aerial dry mass; AHI, annual height increment; SD, stem diameter.dThe covariate was initial seedling height and stem diameter (before planting out).eNot pertinent.fns, nonsignificant, the effect of the initial height covariate was removed from the model.gna, p >|Z| not available because of an estimate equal to 0.

Table 4. Analyses of variance and associated probabilities for growth variables.



gen, phosphorus and potassium concentrations were not sig-nificantly affected by root damage for any of our three treespecies. These results can be explained by the fact that thefluxes of phosphorus and potassium from the soil to theplant are generally much lower than they are for nitrogen(Munson et al. 1993, 1995).

Seedling growth decreased with increasing root damage(Figs. 2a–2h). Soil moisture seemed to only slightly influ-ence the impact of root damage on aerial growth. The largestgrowth reductions were at classes 5 and 6 on both sites, sug-gesting that extensive root damage was necessary to slowdown height and diameter growth. Bigras and Margolis(1997) reported that seedling height and diameter growth of1+0 JP seedlings grown in growth chambers were reducedfollowing 50% destruction of the root system. Coursolle etal. (2002) reported that root damage levels of 69, 72, and84% for WS, BS, and JP, respectively, were required tocause approximately a 50% reduction in the aerial dry massof greenhouse-grown seedlings. Bigras (1998) reported that3 years after outplanting, height growth of 2+0 WS seed-lings was reduced by 11% and diameter growth by 25%when 40% of roots survived an artificial frost.

When comparing root and shoot growth measurements(LIRDM, ADM, AHI, SD) taken during the study, it is ap-parent that new root and shoot growth occurred mainly be-tween August and October on the wet site (Figs. 2a and 2b).During the same period, root growth on the dry site was lessthan on the wet site (Fig. 2e) and did not show the same

growth spurt as on the wet site (Fig. 2f). The greater seed-ling growth on the wet site can probably be partly attributedto the higher soil water content (49.5% compared with34.8% for the dry site). Day and MacGillivray (1975) alsofound that WS seedling rhizogenesis decreased with soilmoisture content. The higher root growth on the wet site canalso be related to its higher mean minimal soil temperaturesin August (10.1°C compared with 8.2°C) and September(8.4°C compared with 4.9°C). Lyr and Hoffmann (1967) ob-served thatPicea abiesroot growth increased with increas-ing soil temperatures between 0° and 26°C. Finally, lowerseasonal mean minimum air temperatures on the dry site(3.9°C compared with 7.2°C for the wet site) as well as theshorter frost-free period (64 days compared with 107 days)probably also hastened growth cessation on the dry site sothat there was not enough time for the undamaged seedlingsto develop more growth relative to their damaged counter-parts (Johnson and Havis 1977; Smit-Spinks et al. 1985;Sakai and Larcher 1987).

Our results suggest that approximately 50% of spruce rootsystems need to be damaged before mortality is affected.This corresponds to annual height increment reductions of2.0 cm for WS and 1.5 cm for BS. On the other hand, JPmortality starts being affected when approximately 40% ofthe root system is damaged, which is associated with a 1-cmreduction in annual height increment. Coursolle et al. (2002)reported that 60–80% of WS, BS, and JP seedling root sys-tems needed to be damaged before seedling survival in thegreenhouse was affected. Finally, Bigras (1998) observed amortality rate of 17% after outplanting when 40% of the liv-ing roots of the 2+0 BS seedlings survived a frost.

It should be made clear that growth reductions measuredin this study are highly dependent on the planting siteswhere the study took place, the environmental conditionsthat prevailed during the study, and the length of the study.In fact, root damage effects on dry site seedling growth seemto have been muted by lower temperatures and a shorterfrost-free period. It is quite possible that growth reductionswould have become more evident during the subsequentgrowing seasons. Also, because of the magnitude of thestudy, no site repetitions were included, which limits theconclusions that can be drawn with respect to the soil mois-ture effects. Finally, only 15 seedlings per treatment combi-nation were used to evaluate seedling mortality, which mayreduce precision. Therefore, caution should be used whenextrapolating the results of this study to other situations.

Acknowledgements

We thank Luce Ouellet, Yves Dubuc, David Gaumont-Guay, and Marie Coyea for assistance; Michèle Bernier-Cardou for statistical advice; and Mohammed Lamhamedifor reviewing an earlier version of the manuscript. We alsothank Reboisement Mauricie Inc. for providing the seedlingsand the ministère des Ressources naturelles du Québec,Fonds pour la formation des chercheurs et l’aide à la recher-che du Québec, and the Natural Sciences and EngineeringResearch Council of Canada for financial support.

© 2002 NRC Canada

614 Can. J. For. Res. Vol. 32, 2002

Seedlingmortality(n = 15)

SpeciesRoot damageclass

Preplantingroot damage(%)

Wetsite

Drysite

Black spruce 1 1 0 02 38 0 03 54 0 04 63 0 15 69 5 46 77 4 7

White spruce 1 3 0 02 27 0 03 37 0 04 48 0 05 59 0 06 65 3 4

Jack pine 1 5 1 22 42 0 33 55 1 44 70 3 55 80 5 126 85 10 13

Note: Seedlings were exposed to an artificial frost in the fall of 1997so as to induce different root damage levels (classes).

Table 5. Wet and dry site mortality of outplanted black spruce,white spruce, and jack pine seedlings as observed in October1998.



References

Bigras, F.J. 1997. Root cold tolerance of black spruce seedlings:viability tests in relation to survival and regrowth. Tree Physiol.17: 311–318.

Bigras, F.J. 1998. Field performance of containerized black spruceseedlings with root systems damaged by freezing or pruning.New For.15: 1–9.

Bigras, F.J., and Calmé, S. 1994. Viability tests for estimating rootcold tolerance of black spruce seedlings. Can. J. For. Res.24:1039–1048.

Bigras, F.J., and Margolis, H.A. 1997. Shoot and root sensitivity ofcontainerized black spruce, white spruce and jack pine seedlingsto late fall freezing. New For.13: 29–49.

Brand, D.G. 1987. Estimating the surface area of spruce and pinefoliage from displaced volume and length. Can. J. For. Res.17:1305–1308.

Bremner, J.M., and Mulvaney, C.S. 1982. Nitrogen—total.InMethods of soil analysis, Part 2. Chemical and microbiologicalproperties.Edited byA.L. Page, R.H. Millet, and D.R. Keeney.Agron. Monogr. 9. pp. 595–624.

Cochran, W.G., and Cox, G.M. 1957. Experimental design. JohnWiley & Sons, New York.

Coursolle, C., Bigras, F.J., and Margolis, H.A. 2000. Assessmentof root freezing damage of two-year-old white spruce, blackspruce and jack pine seedlings. Scand. J. For. Res.15: 343–353.

Coursolle, C., Bigras, F.J., and Margolis, H.A. 2002. Effects ofroot freezing on the physiology and growth ofPicea glauca,Picea mariana, and Pinus banksianaseedlings under differentsoil moisture regimes. Scand J. For. Res. In press.

Dang, Q.-L., Margolis, H.A., Coyea, M.R., Sy, M., and Collatz,G.J. 1997a. Regulation of branch-level gas exchange of borealtrees: roles of shoot water potential and vapor pressure differ-ence. Tree Physiol.17: 521–535.

Dang, Q.-L., Margolis, H.A., Sy, M., Coyea, M.R., Collatz, G.J.,and Walthall, C.L. 1997b. Profiles of photosynthetically activeradiation, nitrogen and photosynthetic capacity in the boreal for-est: implications for scaling from leaf to canopy. J. Geophys.Res.102: 28 845 – 28 859.

Day, R.J., and MacGillivray, G.R. 1975. Root regeneration of fall-lifted white spruce nursery stock in relation to soil moisturecontent. For. Chron.51: 196–199.

Gagnon, J., Gingras, B.-M., Girard, D., Lamhamedi, M., Renaud,M., Richard, S., Robert, N., and Veilleux, L. 2000. Recherche etdéveloppement en production de plants. Gouvernement du Qué-bec, ministère des Ressources naturelles, Forêt Québec, Direc-tion de la recherche forestière, Québec. Rapp. Interne 453.

Grossnickle, S.C., and Blake, T.J. 1986. Environmental and physi-ological control of needle conductance for bare-root blackspruce, white spruce, and jack pine seedlings on boreal cutoversites. Can. J. Bot.64: 943–949.

Hall, J.P. 1996. Insects and diseases of trees in Canada 1994. Natu-ral Resources Canada, Canadian Forest Service, Ottawa, Ont.

Johnson, J.R., and Havis, J.R. 1977. Photoperiod and temperatureeffects on root cold acclimation. J. Am. Soc. Hortic. Sci.102:306–308.

Kramer, P.J. 1983. Water relations of plants. Academic Press, NewYork.

Langerud, B.R., and Sandvik, M. 1991. Root damage toPiceaabies seedlings: effects on root growth capacity, shoot growthcapacity and transpiration. Scand. J. For. Res.6: 387–394.

Langerud, B.R., Puttonen, P., and Troeng, E. 1991. Viability ofPicea abiesseedlings with damaged roots and shoots. Scand. J.For. Res.6: 59–72.

Lindström, A. 1986. Freezing temperature in the root zone—effects on growth of containerizedPinus sylvestrisand Piceaabiesseedlings. Scand. J. For. Res.1: 371–377.

Lindström, A. 1987. Winter storage and root hardiness ofcontainerized conifer seedlings. Doctoral thesis, Department ofForest Yield Research, Swedish University of Agricultural Sci-ences, Garpenberg.

Lindström, A., and Mattsson, A. 1989. Equipment for freezingroots and its use to test cold resistance of young and matureroots of Picea abiesseedlings. Scand. J. For. Res.4: 59–66.

Littell, R.C., Milliken, G.A., Stroup, W.W., and Wolfinger, R.D. 1996.SAS® system for mixed models. SAS Institute Inc., Cary, N.C.

Lyr, H., and Hoffmann, G. 1967. Growth rates and growth period-icity of tree roots. Int. Rev. For. Res.2: 181–236.

Ministère de l’Énergie et des Ressources. 1990. Guide d’évaluationde la qualité des plants de reboisement produits en récipients.Gouvernement du Québec, Québec.

Ministère des Ressources naturelles du Québec. 1998. Performancedes plantations établies par le ministère des Ressourcesnaturelles dans les forêts publiques du Québec, de 1986 à 1995.Ministère des Ressources naturelles, Forêt Québec, Direction del’assistance technique, Québec.

Mitchell, A.K., and Hinckley, T.M. 1993. Effects of foliar nitrogenconcentration on photosynthesis and water use efficiency inDouglas-fir. Tree Physiol.12: 403–410.

Mityga, H.G., and Lanphear, F.O. 1971. Factors influencing thecold hardiness ofTaxus cuspidataroots. J. Am. Soc. Hortic. Sci.96: 83–87.

Munson, A.D., Margolis, H.A., and Brand, D.G. 1993. Intensivesilvicultural treatment: impacts on soil fertility and planted coni-fer response. Soil Sci. Soc. Am. J.57: 246–255.

Munson, A.D., Margolis, H.A., and Brand, D.G. 1995. Seasonalnutrient dynamics in white pine and white spruce in response toenvironmental manipulation. Tree Physiol.15: 141–149.

Ontario Institute of Pedology. 1985. Field manual for describingsoils. 3rd ed. Ontario Institute of Pedology, Guelph, Ont.

Parkinson, J.A., and Allen, S.E. 1975. A wet oxidation proceduresuitable for the determination of nitrogen and mineral nutrientsin biological material. Comm. Soil Sci. Plant Anal.6: 1–11.

Reich, P.B., Kloeppel, B.D., Ellsworth, D.S., and Walters, M.B.1995. Different photosynthesis–nitrogen relations in deciduoushardwood and evergreen coniferous tree species. Oecologia,104: 24–30.

Ritchie, G.A. 1990. A rapid method for detecting cold injury in co-nifer seedling root systems. Can. J. For. Res.20: 26–30.

Sakai, A., and Larcher, W. 1987. Frost survival of plants. Re-sponses and adaptation to freezing stress. Ecol. Stud. 62.

Schoettle, A.W., and Smith, W.K. 1999. Interrelationships amonglight, photosynthesis and nitrogen in the crown of maturePinuscontorta ssp.latifolia. Tree Physiol.19: 13–22.

Seiler, J.R., and Cazell, B.H. 1990. Influence of water stress on thephysiology and growth of red spruce seedlings. Tree Physiol.6:69–77.

Seiler, J.R., and Johnson, J.D. 1985. Photosynthesis and transpira-tion of loblolly pine seedlings as influenced by moisture-stressconditioning. For. Sci.31: 742–749.

Smit-Spinks, B., Swanson, B.T., and Markhart, A.H., III. 1985.The effect of photoperiod and thermoperiod on cold acclimationand growth ofPinus sylvestris. Can. J. For. Res.15: 453–460.

Teskey, R.O., Sheriff, D.W., Hollinger, D.Y., and Thomas, R.B.1995. External and internal factors regulating photosynthesis.InResource physiology of conifers: acquisition, allocation, and uti-lization. Edited by W.K. Smith and T.M. Hinckley. AcademicPress, New York. pp. 105–140.

© 2002 NRC Canada

Dumais et al. 615



simulated root freezing in the nursery: effects on the growth and physiology of containerized boreal...

Documents