Summary Regeneration patterns of Pinus sylvestris L. juve-niles in central Siberian glades were studied in relation tocold-induced photoinhibition. Spatial distribution of seedlingsin different height classes revealed higher seedling densitiesbeneath the canopy than beyond the canopy, and significantlyhigher densities of seedlings < 50 cm tall on the north side ofthe trees. These patterns coincided with differences in lightconditions. Compared with plants on the north side of canopytrees (north-exposed), photosynthetic photon flux (PPF) re-ceived by plants on the south side of canopy trees (south-ex-posed) was always higher, making south-exposed plants moresusceptible to photoinhibition, especially on cool mornings.Chlorophyll fluorescence data revealed lower photochemicalefficiency and increased non-photochemical quenching ofsmall (20–50 cm in height), south-exposed seedlings fromspring to early autumn, indicating increased excitation pres-sure on photosynthesis. Maximum rate of oxygen evolutionwas less in south-exposed plants than in north-exposed plants.Increased pools of xanthophyll cycle pigments and formationof the photoprotective zeaxanthin provided further evidencefor the higher susceptibility to photoinhibition of south-ex-posed seedlings. A linear mixed model analysis explainedmany of the physiological differences observed in seedlingsaccording to height class and aspect with early morning tem-perature and PPF as predictors. The link between photoinhi-bition and differential distribution of seedlings by height classsuggests that photoinhibition, together with other environmen-tal stresses, decreases the survival of small, south-exposedP. sylvestris seedlings, thereby significantly affecting the re-generation pattern of central Siberian pine glades.

Keywords: environmental stress, excitation pressure, Pinussylvestris, Siberia, xanthophyll cycle.

Introduction

Light can cause photochemical damage to foliage when ab-sorbed in excess of photosynthetic capacity. In the short term,photochemical efficiency may decrease as a result of the pro-tective increase in non-photochemical quenching (qN) of lightenergy (Demmig and Björkman 1987, Osmond 1994, Holt etal. 2004). In the long term, excess light causes sustainedphotoinhibition, mainly by damaging photosystem II (PSII)reaction center proteins (Kyle et al. 1987, Adams et al. 1995a).At low temperatures (Öquist et al. 1987, Huner et al. 1998,Öquist and Huner 2003), when photosynthetic energy flow islimited by the rate of carboxylation and the Calvin cycle, theimbalance between light absorption and energy consumptionincreases the excitation pressure on PSII (Huner et al. 1998).Mechanisms providing protection against photo-oxidativedamage from increased excitation pressure include qN of ex-cess light, mainly attributable to pigments of the xanthophyllcycle and the light-dependent conversion of violaxanthin tozeaxanthin (Verhoeven et al. 1998, Ensminger et al. 2004). Inevergreen boreal conifers, like Pinus sylvestris L., sustainedformation of zeaxanthin is part of the cold-hardening processby which a tree adjusts metabolically to winter conditions.This seasonal down-regulation of physiological activity pro-tects the needles from photo-oxidative damage during coolwinter weather (Ottander et al. 1995, Ensminger et al. 2004).During spring, an increase in temperature induces the reorga-nization of the photosynthetic apparatus of P. sylvestris.

In the boreal zone, the solar angle is generally low in spring,which has important consequences for understory seedlings,which, typically, are then at the most vulnerable stage of theirlife cycle (Koslowski and Pallardy 1997). The generally opennature of boreal canopies and the selective shading effect ofoverstory trees expose seedlings growing on the south side of

Tree Physiology 25, 1139–1150© 2005 Heron Publishing—Victoria, Canada

Regeneration patterns in boreal Scots pine glades linked tocold-induced photoinhibition

MARTIJN SLOT,1–3 CHRISTIAN WIRTH,1,4 JENS SCHUMACHER,1 GODEFRIDUSM. J. MOHREN,2 OLGA SHIBISTOVA,5 JON LLOYD1,6 and INGO ENSMINGER1,7,8

1 Max Planck Institute for Biogeochemistry, Hans-Knöll-Strasse 10, 07745 Jena, Germany2 Forest Ecology and Forest Management Group, Wageningen University, P.O. Box 342, 6700 AH, Wageningen, The Netherlands3 Present address: Resource Ecology Group, Wageningen University, Bornsesteeg 69, 6708 PD Wageningen, The Netherlands4 Department of Ecology and Evolutionary Biology, Princeton University, Princeton, NJ 08544, USA5 Institute of Forests, Russian Academy of Sciences, Siberian Branch, Akademgorodok, 660036 Krasnoyarsk, Russia6 Present address: School of Geography, University of Leeds, Leeds, LS2 9JT, UK7 Corresponding author (iensming@uwo.ca)8 Present address: Department of Biology and The BIOTRON, University of Western Ontario, London, ON N6A 5B7, Canada

Received September 15, 2004; accepted February 18, 2005; published online July 4, 2005

parent trees (south-exposed) to a significantly higher photo-synthetic photon flux (PPF) than that experienced by seedlingson the north side of parent trees (north-exposed). We predictthat the different amounts of sunlight to which the understoryseedlings are exposed combined with the typical subzeromorning temperatures in such climates differentially affectsphotochemistry of north- and south-exposed seedlings.

Although P. sylvestris is generally considered shade intoler-ant (Nikolov and Helmisaari 1992), observations in Siberiashowed a clustering of regeneration on the northern side ofparent trees (Figure 1). Similar patterns have been observed inan alpine tree line ecotone of Picea engelmannii Parry exEngelm. and Abies lassiocarpa (Hook.) Nutt. (Germino et al.2002). These findings suggest that, under extreme conditions(clear skies and low temperatures), photoinhibition is a deter-minant of regeneration patterns. Moreover, severe photoin-hibition can reduce the growth (Lundmark and Hällgren 1987,Farage and Long 1991, Close et al. 2002) and survival of seed-lings in subalpine and boreal climates (Ball et al. 1991, Hollyet al. 1994, Germino and Smith 1999). Therefore, we assessedwhether sustained photoinhibition is ecologically significantin structuring open boreal stands with a crown cover of lessthan 30%, a common feature of the central Siberian landscape.Specifically, we (1) analyzed spatial distribution patterns inthe understory regeneration of these forests and the densitiesof juveniles in different height classes; (2) linked these obser-vations with differences in the photosynthetic characteristicsof Scots pine seedlings; and (3) tested the hypothesis that seed-ling distribution is linked to the alleviation of photoinhibitionby canopy shade.

Materials and methods

Study site

Located at the eastern edge of the west Siberian lowland on theYenisei River, 800 km north of Krasnoyarsk, the study area in-cluded unmanaged Scots pine forests on alluvial sand dunessurrounded by peat bogs (Wirth et al. 1999). Mean annual tem-perature is –3.8 °C and the growing season is limited to173 days with mean air temperatures above 0 °C and 91 dayswith a mean air temperature above 10 °C (Schulze et al. 2002).Mean annual precipitation is 535 mm, of which 70% occurs assummer rainfall. The summer water balance is close to zero(Kelliher et al. 1998) and there is no permafrost.

The study site (66°44′ N, 89°22′ E), a glade, had a low den-sity of canopy trees (hereafter referred to as parent trees) withregeneration concentrated beneath them. The mean height ofthe parent trees was 9.7 ± 1.5 m, with an approximate age of180 years. The density of parent trees (diameter at breastheight > 50 cm) was 130 ha– 1, with a crown cover of 30.3%and a basal area of 7.1 m2 ha– 1. About 60% of the podzolic soilwas covered with lichen (Cladonia and Cladina ssp.) andVaccinium vitis-idaea L. For additional details, see Wirth et al.(1999).

Regeneration patterns

Transects 2 m wide were established around 23 parent trees,extending 5 m both south and north (Figure 2a). Within thetransect were three subareas: outside the crown projection area(CPA); the southern CPA; and the northern CPA. The numberof individuals in each area was recorded in the followingheight classes: 0–5 cm; 5–20 cm; 20–50 cm; 0.5–2.0 m;2–4 m; 4–6 m; and 6–8 m.

Micrometeorology

Micrometeorological data were recorded from March 15 toSeptember 17, 2002, within the CPA of one typical parent treewith understory regeneration (Figure 2b). Air temperature wasmeasured at a height of 1.5 m at a distance of 1.5 m from thetrunk of the parent tree in both the northern and the south-ern CPA (Pt 100, Vaisala Oyj, Helsinki, Finland). IncidentPPF was measured, with photosynthetically active radiation(PAR)-calibrated BPW-21 photodiodes (Siemens AG, Mu-nich, Germany) at eight points along the north–south transectunder the parent tree. All sensors were connected to a data log-ger (Delta-T Devices, Cambridge, U.K.). Data were recordedat 10-s intervals and used to calculate and record 10-min meanvalues.

Parent tree plots

For three parent tree plots, each consisting of a parent tree andthe juveniles within the CPA and including the micrometeor-ology plot, four groups of juveniles were distinguished: indi-viduals 20–50 cm tall on the north and on the south side of thetrunk of the parent tree, and individuals 70–150 cm tall on thenorth and south side (Figure 2c). These groups are hereafterreferred to as SN (small north), SS (small south), LN (largenorth) and LS (large south). Preliminary mapping of the regen-eration patterns suggested that seedlings between 5 and 50 cmtall suffered the highest mortality in high-light environments;

1140 SLOT ET AL.

TREE PHYSIOLOGY VOLUME 25, 2005

Figure 1. Regeneration of Pinus sylvestris under a parent tree, show-ing the typical clustering of seedlings on the north side (right of theparent tree) compared with the smaller number of smaller trees on thesouth side (left of the parent tree). During the morning in spring, whenthe sun angle is low, the parent tree canopy casts shade over the seed-lings on the north side, whereas the seedlings on the south side of theparent tree are exposed to full sunlight.

however, because needles of seedlings from the 5–20-cm sizeclass were too small for repeated sampling, seedlings in the20–50-cm size class were the smallest seedlings studied.

Sampling

Exposed second-year needles were collected from the upperbranches of three plants per size class around solar noon. Asubsample was immediately immersed in liquid nitrogen forlater analysis of pigments, nonstructural carbohydrates andstable isotopes (Foliar 13C) and the remaining 20–30 needleswere taken to the laboratory for physiological measurements.Samples were collected from each of the three plots in spring(May 5, 7 and 9, 2002), early summer (June 5, 6 and 7, 2002),late summer (August 20 and 23, 2002) and early autumn (Sep-tember 25, 2002). On each sampling day, samples of all sizeclasses were taken. Sampling was repeated on the next or fol-lowing days, again from all classes, to ensure that differencesbetween the size classes did not represent day-to-day varia-tions in temperature and PPF.

Experimental plots

To study the effects of shading on photosynthesis under con-trolled conditions, seedlings 20–50 cm tall were either ex-posed to full sunlight or artificially shaded after removal of theparent tree (Figure 2d). Shade was provided by pyramidalframes covered with shade cloth that reduced direct irradianceby 70%. Six seedlings that were previously beneath the can-opy of the removed tree were monitored. Of these, three wereartificially shaded, whereas three were unshaded. Samplingand measurements were performed shortly after snowmelt(May 12, 2002) and 1 month later (June 11, 2002). On both oc-casions, samples were taken from second-year needles aroundsolar noon. Treatment and plot samples were pooled for fur-ther measurements and analysis.

Field measurements

From spring until early summer and from late summer untilearly autumn, chlorophyll fluorescence of PSII was regularlymeasured in the field. Optimum quantum efficiency of PSII,estimated by the ratio of variable to maximum chlorophyll flu-orescence (Fv /Fm = (Fm – F0)/Fm, where F0 = instantaneousfluorescence at open PSII centers), was measured after 20 minof dark adaptation, using dark leaf clips attached to thefiberoptics of a PAM-2000 portable fluorometer (Heinz WalzGmbH, Effeltrich, Germany). On each occasion, three north-erly exposed plants and three southerly exposed plants fromeach of the 20–50- and 70–150-cm size classes were measuredat each of three parent tree plots. For the experimentally ma-nipulated plots, Fv/Fm was measured seven times betweenMay 12 and June 11, 2002.

Laboratory measurements

Photosynthetic activity of P. sylvestris was studied in the labo-ratory based on measurements of chlorophyll fluorescence,photosynthetic oxygen exchange and analysis of photosyn-thetic pigments. The response of photosynthetic oxygen evo-lution to saturating PPF (1200 µmol m– 2 s– 1) (P1200) was mea-sured on needle discs in a leaf chamber (LD2/3, Hansatech,King’s Lynn, U.K.) with a Clark-type electrode. Simulta-neously, chlorophyll fluorescence was recorded with a

TREE PHYSIOLOGY ONLINE at http://heronpublishing.com

PHOTOINHIBITION AND REGENERATION OF SCOTS PINE 1141

Figure 2. Experimental design of the field plots. Twenty-three parenttree plots of the design shown in (a) were analyzed for regenerationdistribution. Meteorology was recorded in one plot (b). Samples formeasurement of photosynthetic capacity were taken from three inde-pendent plots as shown in (c). Three independent plots were used forexperimental shading of nine seedlings after removal of the parenttree (d) (� = shaded seedling). For laboratory measurements, pigmentand carbohydrates analysis samples were pooled by treatment andplot. Field measurements were made on all individuals.

PAM-2000 fluorometer (Ensminger et al. 2004) at 20 °C in anatmosphere containing 5% CO2. The efficiency of open PSIIreaction centers in the light (effective quantum yield) was esti-mated as the ratio of ∆F/Fm′ = (Fm′ – Ft)/Fm′ , where Fm′ =maximal fluorescence at closed PSII centers, and Ft = tran-sient fluorescence at partly closed PSII centers during illumi-nation with actinic light (Schreiber et al. 1994). Non-photo-chemical quenching, a measure of radiationless dissipation ofabsorbed light energy, was calculated from the fluorescencedata as (Fm – Fm′)/(Fm – F0′), where F0′ = minimal fluores-cence at open PSII centers immediately after illumination.Fully relaxed Fm values are required to calculate winter valuesof qN (Adams et al. 1995a, 1995b). During deep, sustainedquenching of excitation energy by a photosystem primed forwinter conditions, Fm measured after brief dark adaptation(e.g., a few hours) does not represent the fully relaxed value.To overcome the problem of underestimation of Fm duringwinter, we extrapolated relaxed (or summer) values assumingthat Fm at its optimum will exceed F0 by a factor of five(Schreiber et al. 1994). We then used this value to estimate qN.We believe this procedure yielded reasonable values of Fm, be-cause we obtained similar Fm values during summer from di-rect measurements and by calculation. The same protocol wasused to measure photosynthetic characteristics of seedlings inthe experimentally manipulated plots before and 1 month aftertreatment.

Photosynthetic pigments

Frozen needle samples from the field were ground in liquid ni-trogen and freeze-dried. Chlorophylls and carotenoids wereextracted from subsamples of the freeze-dried samples accord-ing to Ensminger et al. (2004). Total amounts of chlorophylland carotenoids were determined spectrophotometrically(Helios, ThermoSpectronics, Cambridge, U.K.), as describedby Lichtenthaler (1987). Pigment composition was analyzedby high performance liquid chromatography (HPLC) (Agilent1100, Agilent Technologies, Palo Alto, CA) with a reversedphase C-18 column (Knaur, Berlin, Germany).

Foliar δ13C

The relative abundance of stable isotopes of carbon in com-busted samples was determined as described by Werner andBrand (2001) on subsamples of the freeze-dried foliar samplesused for pigment analysis.

Nonstructural carbohydrates

Soluble sugars were extracted from a subsample of thefreeze-dried samples according to Ögren (1996). The totalpool of soluble sugars (predominantly glucose and fructose)was determined by the anthrone method (Jermyn 1975) in aHelios spectrometer (ThermoSpectronics), with α-D-glucose(Carl Roth GmbH, Karlsruhe, Germany) in 80% ethanol as astandard. Starch was determined by the anthrone method fol-lowing extraction from the residues with perchloric acid(Iivonen et al. 2001). Starch (Merck KGaA, Darmstadt, Ger-many) in 30% perchloric acid was used as a standard.

Statistical analyses

To determine whether the distribution of individuals across thethree transect areas (outside the CPA, northern CPA and south-ern CPA) departed from randomness, the data for all parenttrees were pooled for each height class. Randomness was eval-uated by a χ2 test.

For each season, exposure effects on chlorophyll fluores-cence parameters and pigment composition were analyzed bythe paired t test. Differences between treatments were testedby repeated-measures analysis of variance (ANOVA). We alsoused repeated-measures ANOVA to test for differences be-tween size classes.

Because insufficient material was available for some δ13Cdeterminations, we allowed for missing data with the Re-stricted Maximum Likelihood (REML) method (Gilmour etal. 1995) to analyze seasonal effects on this parameter. Theequation fitted was:

δ µ α βij i j ijS= + + + + e (1)

where δij is foliar δ13C, µ is the overall mean δ13C, αi + βj arefixed effects representing the effects of aspect and seedlingheight, respectively, S is a random seasonal effect that takes ac-count of variation in δ13C with time of year, and eij is the ran-dom (unexplained) error.

Risk of photoinhibition

Indicators for photosynthetic activity and photoinhibitionwere modeled with air temperature, PPF, exposure (aspect)and size class (size) as predictor variables in a linear mixedmodel, treating individual trees as a random factor. Aspect andsize as intrinsic components of the study design were includedin all models. The best predictive model was selected based onthe Akaike Information Criterion (AIC). Statistical signifi-cance of individual predictor variables was assessed by thelikelihood ratio test. All models were estimated with the maxi-mum-likelihood method (Pinheiro and Bates 2000) as imple-mented in the lme-procedure in S-Plus 6.0 (Insightful, Rein-ach, Switzerland). To assess the significance of the selectedpredictor variables we used a stepwise approach: aspect andsize were included in all models. We then added mean air tem-peratures at different times of the day (0600, 0900 and 1200 h).The best model was used in the next step, where we also in-cluded PPF at different times of day. Finally, we tested for in-teractive effects between temperature and PPF.

The micrometeorological parameters obtained from oneplot do not contribute to an explanation of differences betweenplots. They provide, however, a parsimonious and biologicallyreasonable explanation of differences between seasons that areshared by all plots.

Results

Regeneration patterns

Seedlings less than 20 cm tall occurred at significantly higherdensities outside the CPA than inside the CPA, whereas seed-

1142 SLOT ET AL.

TREE PHYSIOLOGY VOLUME 25, 2005

lings 20–50 cm tall were more often found within the CPA andon the north side of the parent tree (Table 1, Figure 3). Seed-lings > 50 cm tall grew almost exclusively within the northernCPA. Random regeneration patterning was rarely observedacross the areas examined. Seedling densities of the two small-est size classes, 0–5 and 5–20-cm-tall seedlings, were similar,indicating a transition probability (Pt) of Pt = 0.8 for seedlings< 5 cm tall reaching the 5–20-cm size class when located out-side the CPA. In contrast to the 5–20-cm size class, seedlingdensities in the 20–50-cm size class outside the CPA were low:the Pt for seedlings in the 5–20-cm size class reaching the nextsize class was only 0.09. However, in the > 50-cm size class,no significant decrease in seedling density was observed, re-sulting in Pt = 1. Estimated transition probabilities were signif-icantly higher beneath the canopy than outside the canopy forseedlings in the 5–20- and the 20–50-cm size classes (Pt =0.33 and 0.93, respectively), but unlike the pattern observed

outside the CPA, a decrease in seedling density persistedthroughout the taller size classes. Seedling densities also dif-fered between southern and northern CPAs. The transitionprobability for seedlings in the 5–20- and the 20–50-cm sizeclasses was significantly greater on the north side of the trees(Pt = 0.39 and 0.95, respectively) than on the south side (0.20and 0.84, respectively), whereas for more mature seedlings,the transition probabilities were similar on the north and southsides of the trees.

Micrometeorology

Within the CPA, there were large differences in the light envi-ronment of the studied trees, with the southern and northernCPAs being clearly differentiated as high-light and low-lighthabitats, respectively (Figure 4). A comparison of diurnal pat-terns showed increased irradiance in the southern CPA early inthe morning, with midday PPF generally twice as high as in

TREE PHYSIOLOGY ONLINE at http://heronpublishing.com

PHOTOINHIBITION AND REGENERATION OF SCOTS PINE 1143

Table 1. Numbers of observed (obs) and expected (exp) seedlings of Pinus sylvestris in different regeneration environments in relation to a parenttree: outside the crown projection area (CPA) versus inside the north side and inside the south side of the CPA. Data were pooled for 23 parenttrees. Areas and percentages are given for all parent trees. Abbreviation: Pt = transition probability.

Size class Outside CPA CPA north side CPA south side Chi-square P202 m2 ≅ 43.9% 132 m2 ≅ 51.2% 126 m2 ≅ 48.8%

obs Pt exp obs Pt exp obs Pt exp

0–5 cm 437 0.80 313 182 1.00 204 93 1.00 195 105.3 < 0.0015–20 cm 348 0.09 289 216 0.39 189 95 0.20 181 56.1 < 0.00120–50 cm 33 0.09 60 85 0.95 39 19 0.84 38 74.3 < 0.0010.5–2 m 3 1.00 44 81 0.38 28 16 0.38 27 138.0 < 0.0012–4 m 2 1.00 17 31 0.65 11 6 0.67 11 50.4 < 0.0014–6 m 3 1.00 12 20 0.45 7 4 0.50 7 27.5 < 0.0016–8 m 5 7 9 4 2 4 6.1 0.047

Figure 3. Spatial distribution of regeneration in relation to parent tree.Seedlings were grouped into seven height classes. The data representthe mean densities recorded in steps of 1 m along 2 × 10-m transectsestablished across the crown projection area (CPA) of 23 parent trees.The circle and rectangle represent the positions of the crown and stemof the parent tree.

Figure 4. Seasonal changes in mean daytime photosynthetic photonflux (PPF) along the north–south transect within the crown projectionarea (CPA) of the parent tree plots. Symbols: � = winter; � = spring;� = summer; and � = autumn. Each value represents the mean of5 days during the respective sampling periods.

the northern CPA (Figure 5). Differences in air temperatureswere typically less than 2 °C between south and north aspectsthroughout the study period (Figure 5). The diurnal patterns ofair temperature in spring (Figure 5) changed rapidly withmorning temperatures between –3 and –5 °C and rising tomore than 5 °C early in the afternoon.

Chlorophyll fluorescence and gas exchange

For both the smaller (20–50 cm) and larger (70–150 cm) seed-lings, minimum Fv /Fm values were observed in spring(May 5–9; Figures 6a and 6b), indicating sustained low func-tional PSII activity in response to winter stress. In early andlate summer (June 5–7 and August 20–23, respectively),Fv /Fm values were highest, ranging between 0.7 and 0.8. Byearly autumn (September 25), Fv /Fm had already decreased,presumably indicating the beginning of the cold-hardeningprocess (Figures 6a and 6b). Repeated-measures ANOVA re-vealed higher Fv /Fm values in north-exposed plants than insouth-exposed plants (P < 0.05), and larger seedlings hadhigher Fv /Fm values than smaller seedlings (P < 0.05).

Photochemical efficiency (∆F/Fm′) in saturating light(1200 µmol m– 2 s– 1) was lowest in spring (May 5–9), highestin summer (August 20–23), and slightly lower in early autumn(September 25) (Figures 6c and 6d). For 20–50-cm-tall seed-lings from the southern CPA, ∆F/Fm′ was consistently lowerthan for similar-sized seedlings from the northern CPA,whereas there was no effect of exposure on ∆F/Fm′ of seed-lings in the 70–150-cm size class (Figure 6d). Radiationlessdissipation of excess light is illustrated by qN in Figures 6eand 6f. Irrespective of season, qN was higher in south-exposed

plants in the 20–50-cm class than in the 70–150-cm class, andwas slightly lower in north-exposed seedlings in the 70–150-cm class than in the 20–50-cm class, except in spring(Figures 6e and 6f ).

Throughout the study, rates of photosynthetic oxygen evolu-tion at saturating light (1200 µmol m– 2 s– 1), P1200, were con-siderably higher in north-exposed seedlings than in south-exposed seedlings in the 20–50-cm size class (Table 2). Incontrast, the difference between north- and south-exposedplants was less pronounced among larger plants, and P1200 didnot differ systematically between LS and LN. Decreased P1200

was observed in seedlings in the 20–50-cm size class in latesummer, particularly for SS individuals, perhaps in response tosummer soil water deficits.

Photosynthetic pigments

Chlorophyll concentrations increased from spring until sum-mer, but the effect of aspect was apparent only in seedlings ofthe smaller size class (Figure 7a). Both the ratio of xanthophyllcycle pigments (V = violaxanthin, A = antheraxanthin and Z =zeaxanthin) per chlorophyll (VAZ Chl – 1) and the de-epoxi-dation state of the xanthophyll cycle pigments (DEPS) werehighest in spring (Figures 7c–f). Within the 20–50-cm sizeclass, north-exposed seedlings had consistently lower concen-trations of VAZ Chl – 1 and lower DEPS compared with south-exposed seedlings (Figures 7c and 7e). For the 70–150-cmsize class, differences between aspects were smaller and lesssystematic for VAZ Chl – 1 and DEPS (Figures 7d and 7f).Most significant was the difference between small and largeseedlings: VAZ Chl – 1 and DEPS were typically five to six

1144 SLOT ET AL.

TREE PHYSIOLOGY VOLUME 25, 2005

Figure 5. Diurnal changes in air tem-perature and photosynthetic photonflux (PPF) within the crown projectionarea (CPA) of the parent tree plots.Black lines = air temperature withinthe northern CPA; gray lines = air tem-perature within the southern CPA; grayarea = PPF within the northern CPA;and white area = PPF within the south-ern CPA. Values are 10-min meansover a typical day.

times higher in small plants than in larger plants throughoutthe summer.

Experimentally manipulated plots

After 1 month of light manipulation, Fv /Fm had increased inseedlings in both treatments (Table 3), and it was significantlyhigher in shaded seedlings than in exposed seedlings. Duringearly spring, ∆F/Fm′ values were low and few differences ex-

isted between shaded and sun-exposed plants. One month afterthe imposition of shading, ∆F/Fm′ had increased—confirmingthe recovery of functional PSII as suggested by the increasedFv /Fm values—and the capacity for photosynthetic electrontransport in the light also increased. However, Fv /Fm was sig-nificantly lower in exposed seedlings than in shaded seedlings,which the higher qN values confirmed. Nevertheless, after1 month, P1200 did not differ significantly between exposed andshaded plants (8.55 ± 1.86 and 10.36 ± 2.08 µmol m– 2 s– 1, re-spectively; Table 4).

Within 1 month following application of the shading treat-ment, chlorophyll concentrations of shaded seedlings weretwice those of unshaded plants, indicating increased capacityfor light absorption (Table 3). In addition, shaded seedlingscontained only about one third the quantity of xanthophyll cy-cle pigments as exposed plants. In contrast, exposed plants ex-ploited the photoprotective function of zeaxanthin: DEPS wasas high as 0.774 ± 0.074 mol mol– 1. Shaded seedlings retainedxanthophyll cycle pigments principally as violaxanthin.

Foliar δ13C composition

Seasonally adjusted foliar δ13C, used to assess variations in theratio of intercellular to ambient CO2 concentrations (Ci /Ca)(Table 4), showed a statistically significant effect for both as-pect and size (P < 0.05) but no interaction between aspect andsize. The less negative δ13C values for south-exposed seedlingssuggest a lower Ci /Ca ratio, and may reflect increased soil wa-ter deficits reducing stomatal conductances; however, differ-ences were small and variations might also be attributable tolight-induced differences in leaf anatomy and internal conduc-tances for CO2 diffusion (Syvertsen et al. 1995).

Nonstructural carbohydrates

In small plants from north and south CPAs, as well as in largeplants from south CPAs, amounts of soluble sugars (17.8–18.4% of DM)) and starch (3.0–3.2% of DM) in needles weresimilar (Table 4). Among seedlings, only needles from 70–150-cm-tall north-exposed plants had significantly greateramounts of soluble sugars (22.0% of DM, one-way ANOVA,P < 0.05; Table 4). The pooled data showed that starch concen-trations from spring to autumn did not change significantlywithin groups. Nevertheless, as indicated by the mean of allgroups, starch concentrations were higher during early sum-mer (5.0 ± 0.8% of DM) compared with spring, late summer

TREE PHYSIOLOGY ONLINE at http://heronpublishing.com

PHOTOINHIBITION AND REGENERATION OF SCOTS PINE 1145

Figure 6. Seasonal changes in photosystem II (PSII) activity of seed-lings (20–50- and 70–150-cm size classes) from south- (open bars)and north-exposed (filled bars) habitats. Optimum quantum yield ofPSII (Fv /Fm) under ambient field conditions (a,b); effective quantumyield of PSII (∆F/Fm′; c,d); and non-photochemical quenching (qN;e,f) measured during exposure to high photosynthetic photon flux(PPF) (1200 µmol m– 2 s– 1) under laboratory conditions. Each valuerepresents the mean of three measurements in each of three parent treeplots ± SD.

Table 2. Seasonal changes in net photosynthetic oxygen evolution at a photosynthetic photon flux (PPF) of 1200 µmol m– 2 s– 1 (P1200) in Pinussylvestris seedlings. Each value represents the mean of three independent measurements ± SD (n = 3); in autumn, pooled samples of three plantsper plot were used.

Size class Exposure P1200 (µmol O2 m– 2 s– 1)

Spring Early summer Late summer Autumn

20–50 cm Southern 5.88 ± 1.99 7.62 ± 5.23 5.24 ± 1.73 12.14Northern 9.49 ± 0.74 12.6 ± 0.43 9.99 ± 3.57 12.43

70–150 cm Southern 7.59 ± 3.59 9.02 ± 2.13 11.9 ± 1.08 14.26Northern 8.3 ± 0.75 15.1 ± 3.85 12.1 ± 1.16 14.96

and autumn (3.5 ± 1.3, 2.8 ± 0.6 and 2.6 ± 0.8% of DM, respec-tively) (one-way ANOVA, Tukey’s test, P < 0.05).

Risk of photoinhibition

A linear mixed-model analysis was used to determine if lowmorning temperatures and a higher PPF accounted for the in-

creased risk of photoinhibition in south-exposed seedlingscompared with north-exposed seedlings. Likelihood ratio testsdirectly assessed the importance of each variable added to themodel (Table 5), and the Akaike Information Criterion (AIC)was used to compare the relative predictive performance of themodels. The nature and number of independent variables thatprovided the best model fit varied between the different indi-cators for photosynthesis and photoinhibition (Table 5).

For Fv /Fm, the highest-ranking model (with the lowest AIC)included aspect, size, air temperature at 0600 h (T06.00), PPFat 0900 h (PPF09.00) and the interactive effect of T06.00 andPPF09.00. Although aspect and size had no significant effect onFv /Fm, we found that increases in T06.00 led to increased Fv /Fm,whereas increases in PPF09.00 led to a decrease. The significantinteraction effect shows that the decrease induced by higherPPF09.00 was less at higher temperatures (Table 5).

The same set of predictor variables explained the degree ofqN (AIC –166.36), but omitting the interaction between T06.00

and PPF09.00 improved the model, as the lower AIC of –168.36indicates.

Aspect, size, T12.00, PPF09.00, and the interaction of T12.00 andPPF09.00 provided the best model for net photosynthesis (AIC204.54). However, the predictive performance of a similarmodel without the interactive effect of T12.00 and PPF09.00 wasslightly better (AIC 202.90).

Low morning temperatures (T06.00) and PPF09.00 were impor-tant parameters for explaining differences in the amount ofxanthophyll cycle pigments. Again, the interactive effects oftemperature and PPF did not improve the model. The bestmodel to explain the formation of zeaxanthin and antheraxan-thin from violaxanthin (DEPS) used T12.00 in addition to aspectand size (AIC = –33.27). However, the AIC value in a morecomplex model that added PPF12.00 provided a similar AICvalue (–31.57), emphasizing the overall importance of tem-perature and light as predictors.

Discussion

Photosynthetic acclimation reflects environmentaldifferences

Differences in photosynthetic characteristics of south- and

1146 SLOT ET AL.

TREE PHYSIOLOGY VOLUME 25, 2005

Figure 7. Seasonal changes in pigment composition of needles fromsouth- (open bars) and north-exposed (filled bars) seedlings (20–50-and 70–150-cm size classes). Total chlorophyll per dry mass (a, b),the pool of xanthophyll cycle pigments per total chlorophyll (VAZChl – 1) (c, d) and the de-epoxidation status of the xanthophyll cyclepigments (DEPS = (0.5A + Z)(VAZ)– 1) (e, f). Each value is the meanof three measurements in each of three parent tree plots ± SD.

Table 3. Effects of artificial shading on photosynthetic parameters of Pinus sylvestris seedlings. Each value represents the mean of three independ-ent measurements ± SD (n = 3); * = significant difference between shade treatments P < 0.05; and ** = P < 0.01(t test). Abbreviations: Fv /Fm =optimum quantum yield of photosystem II; ∆F/Fm′1200 = photochemical efficiency at saturating light (1200 µmol m– 2 s– 1); qN1200 = non-photo-chemical quenching at saturating light; P1200 = rate of oxygen evolution at saturating light (µmol O2 m– 2 s– 1); Chl a + Chl b = total chlorophyll (mggDM

– 1 ); VAZ Chl – 1 xanthophyll cycle pigments per chlorophyll (mmol mol– 1); and DEPS = (0.5A + Z)(VAZ)– 1, de-epoxidation status ofxanthophyll cycle pigments (mol mol–1).

Shading Treatment Fv/Fm ∆F/Fm′1200 qN1200 P1200 Chl a + Chl b VAZ Chl– 1 DEPS

Before Exposed 0.39 ± 0.03 0.08 ± 0.03 0.95 ± 0.01 6.01 ± 2.26 0.98 ± 0.02 113.02 ± 9.61 0.86 ± 0.01Shaded 0.44 ± 0.03 0.11 ± 0.01 0.93 ± 0.02 5.86 ± 1.41 1.07 ± 0.07 97.92 ± 14.85 0.85 ± 0.00

After Exposed 0.71** ± 0.03 0.63 ± 0.11 0.85** ± 0.02 8.55 ± 1.86 1.19* ± 0.02 67.42* ± 13.25 0.77** ± 0.07Shaded 0.81 ± 0.01 0.63 ± 0.02 0.74 ± 0.02 10.36 ± 2.08 2.17 ± 0.10 17.81 ± 0.55 0.00 ± 0.00

north-exposed seedlings (Figure 5) were associated with ef-fects of aspect on seasonal and diurnal light conditions. Sea-sonal changes in photosynthetic physiology, such as therapidly changing chlorophyll fluorescence parameters in

spring (Figure 6), were, however, superimposed on the differ-ences attributable to aspect and size. Nevertheless, compara-tive observations made at several stages throughout thegrowing season allow us to differentiate transitional differ-ences from those attributable to aspect and size.

Optimum quantum yield indicated decreased maximum ef-ficiency of PSII in SS plants that persisted throughout the mea-surement period (Figure 6). This confirms the results of theshading experiment, which revealed a negative relationshipbetween Fv /Fm and exposure to increased PPF, indicating thatexcess PPF decreases the quantum yield of photosynthesis.Similarly, ∆F/Fm′ and P1200 reflected impaired photosyntheticefficiency of the south-exposed unshaded plants comparedwith north-exposed shade plants (Figure 6, Table 2).

Decreased photosynthetic efficiency in SS plants was con-sistent with the greater fraction of excitation energy quenchednon-photochemically, involving the photoprotective proper-ties of the xanthophyll cycle (Figure 7, Table 3). This entailsthe light-dependent conversion of the xanthophyll violaxan-thin into the protective form zeaxanthin (Adams et al. 1995a).Both the total pool of VAZ and the conversion of violaxanthin

TREE PHYSIOLOGY ONLINE at http://heronpublishing.com

PHOTOINHIBITION AND REGENERATION OF SCOTS PINE 1147

Table 4. Differences in δ13C values and carbohydrate concentrationsin needles of Pinus sylvestris seedlings as the means for all samplestaken over the growing season. Data represent mean of n = 9–11 ± SE.For soluble sugars and starch, different letters indicate significant dif-ferences between seedlings at P < 0.05 (one-way ANOVA and LSDpost hoc test). For δ13C values, statistical analysis was performed bythe Restricted Maximum Likelihood method.

Size class Exposure δ13C Soluble sugars Starch(‰) (% DM) (% DM)

20–50 cm South –28.4 ± 0.3 a 18.3 ± 1.4 b 3.2 ± 0.3 aNorth –29.0 ± 0.2 bc 18.4 ± 1.4 b 3.3 ± 1.3 a

70–150 cm South –29.0 ± 0.2 b 17.8 ± 1.5 ab 3.0 ± 0.4 aNorth 29.5 ± 0.1 c 22.0 ± 3.0 bc 3.5 ± 0.5 a

Table 5. The five best models for indicators of photoinhibition in Pinus sylvestris seedlings from natural plots. Abbreviations: Fv/Fm = optimumquantum yield of photosystem II; Pnet = net assimilation rate; qN = non-photochemical quenching; VAZ Chl – 1 = xanthophyll cycle pigments pertotal chlorophyll; DEPS = de-epoxidation status of xanthophyll cycle pigments; AIC = Akaike information criterion; Aspect = South (–), North(+); size = small (–), large (+); T = temperature at respective time of the day; and PPF = photosynthetic photon flux (µmol m– 2 s– 1). Respective co-efficients of the best model are given in brackets below.

Score Indicator AIC Source L ratio P

4 Fv /Fm –32.66 Aspect 0.556 0.45595 –31.24 Aspect, size 0.574 0.44863 –124.94 Aspect, size, T06.00 95.707 < 0.00012 –159.40 Aspect, size, T06.00, PPF09.00 36.452 < 0.00011 –171.45 Aspect, size, T06.00, PPF09.00, T06.00 × PPF09.00 (–0.04, 0.04, 0.02, –0.001, 0.00008) 14.055 0.0002

5 Pnet 216.37 Aspect 8.267 0.00404 214.70 Aspect, size 3.677 0.05523 208.82 Aspect, size, T12.00 7.875 0.00501 202.90 Aspect, size, T12.00, PPF09.00 (0.22, 0.97, 0.41, –0.03) 7.922 0.00492 204.54 Aspect, size, T12.00, PPF09.00, T12.00 × PPF09.00 0.356 0.5505

4 qN –106.89 Aspect 0.783 0.37635 –104.95 Aspect, size 0.066 0.79743 –159.24 Aspect, size, T06.00 56.285 < 0.00011 –168.36 Aspect, size, T06.00, PPF09.00 (0.01, 0.01, –0.01, 0.0003) 11.126 0.00092 –166.52 Aspect, size, T06.00, PPF09.00, T06.00 × PPF09.00 0.160 0.6892

5 VAZ Chl –1 413.71 Aspect 0.303 0.58204 411.89 Aspect, size 3.821 0.05063 360.69 Aspect, size, T06.00 43.772 < 0.00011 359.76 Aspect, size, T06.00, PPF09.00 (14.77, –20.21, –6.27, 0.18) 12.352 0.00042 360.16 Aspect, size, T06.00, PPF09.00, T06.00 × PPF09.00 1.600 0.2063

4 DEPS 21.29 Aspect 0.134 0.24805 14.77 Aspect, size 8.517 0.00351 –33.27 Aspect, size, T12.00 (–0.06, –0.22, –0.08) 50.032 < 0.00012 –31.57 Aspect, size, T12.00, PPF12.00 0.307 0.57973 –29.57 Aspect, size, T12.00, PPF12.00, T12.00 × PPF12.00 < 0.003 0.9559

into zeaxanthin were about 25–30% higher in SS plants than inSN plants (Figure 7). A consistent pattern was also observedbetween sun-exposed and artificially shaded plants (Table 4).Such up-regulation of the pool of VAZ and the constitutive in-crease in DEPS are typical for plants experiencing increasedexcitation pressure as a result of high PPF or conditions unfa-vorable for photosynthesis (Adams et al. 1995a, Ensminger etal. 2001, Müller et al. 2001), and are especially well docu-mented for cold-stressed evergreen conifers (Ottander et al.1995, Ensminger et al. 2004) and eucalypts (Close et al. 2002).Evidently, south-exposed plants acclimated in response to in-creased PSII excitation pressure. The above-mentioned re-sponses are thought to be protective processes to avoid photo-oxidative damage; however, constant exposure to unfavorableconditions also induces severe stress and sustained increasesin PSII excitation pressure. In the long term, cold-inducedchronic photoinhibition likely affects plant fitness as shown inoverwintering Brassica napus L. and evergreen Eucalyptuspauciflora Sieber ex A. Spreng. (Farage and Long 1991, Ballet al. 1997). Because detailed physiological measurements onseedlings less than 20 cm tall could not be included in thisstudy, we cannot provide the biochemical data for plants in thesize class with the highest mortality rates (Table 1).

Effect of morning temperatures

Differences in Fv /Fm of south- and north-exposed seedlingswere largest in spring (14 and 10% lower Fv /Fm in SS and LS

compared with SN and LN, respectively) (Figure 6). In spring,when low morning temperatures increase susceptibility tophotoinhibition (Ball et al. 1991, Blennow and Lindkvist2000), air temperature in the CPA did not differ substantiallybetween north and south areas. However, because the southernCPA was illuminated about 1 h earlier, the temperature, whenneedles first intercepted direct sunlight, was about 2–3 °Clower on the south side than on the north side of the parent tree(Figure 5b). Therefore, growth conditions on the south sidepromote cold-induced photoinhibition, a conclusion sup-ported by results of the mixed-model analysis. For three out offive indicators of photoinhibition (Fv/Fm, qN and VAZ Chl – 1),early morning temperature (T06.00) contributed significantly tothe quality of the model (Table 5). Except for DEPS, the differ-ences in PPF at 0900 h significantly improved the models, sup-porting the hypothesis that early morning cold-induced pho-toinhibition is linked to the observed spatial pattern ofregeneration.

Smaller plants more affected

When reorganization of the photosynthetic apparatus wascompleted in late spring, the north–south difference in photo-synthetic characteristics and pigment composition persisted inthe 20–50-cm plants. Compared with the larger size class, thesmaller plants placed a large demand on the photoprotectiveproperties of the xanthophyll cycle during summer, as indi-cated by a two- to eightfold increase in the amount of DEPSduring early and late summer (Figures 7e and 7f). Within thesmall size class, SS plants always showed higher physiological

indicators of light stress than SN plants. Persistently high zea-xanthin concentrations must have primed the photosystems ofsmall and especially SS plants for increased dissipation of ex-cess light by protective qN throughout the year. Apparentlythere was no such strict acclimation to light-stress for LS

plants. Because photosynthetic characteristics of large (LS andLN) plants were similar once the plants had recovered fromwinter stress, we conclude that larger plants have sufficient re-serves to repair and maintain the photosynthetic machinery inspring.

Does photoinhibition reduce survival?

Our results support the findings of Ball et al. (1991) and Hollyet al. (1994), which showed a relationship between distribu-tion of regeneration and photoinhibition in the evergreens Eu-calyptus pauciflora Sieb. ex Spreng. and E. polyanthemosSchauer., respectively. Asymmetries in the structure of theseedling population and differences in Pt from one size class tothe next (Table 1) might reflect greater mortality and slowergrowth at sites where the seedlings were subject to more fre-quent and more prolonged excess irradiance combined withlower minimum temperatures. Germino et al. (2002) recentlydescribed similar asymmetrical patterns of conifer seedlingdistribution based on a correlation between meteorologicaldata and the frequency of occurrence of young conifers at thealpine tree line. Without actually measuring photosynthesis orphotoinhibition, the authors concluded that seedling survivalwas strongly affected by high sunlight and low temperature.However, Adams et al. (1995a) argue that photoinhibition andthe excess energy dissipation process are unlikely to limitgrowth or establishment of juveniles in the field. Instead, theprotective photoinhibition mechanism indicates the degree ofstress experienced by the plants, assuming that environmentalstress factors limit the utilization of absorbed light energy.

This may be the case in our study because we found a qual-itative relationship between putative mortality and lowergrowth rate, as reflected in asymmetries in the structure of theseedling populations (Table 1) and photoinhibition in smallseedlings (Figure 6, Table 6). Despite the observed chronicsuppression of photosynthetic efficiency, we have no direct ev-idence that photoinhibition kills seedlings, and the analyses ofsoluble carbohydrates and starch contents of the needles of ex-posed and shaded seedlings showed no significant differences(Table 3), contrary to what would be expected if chronicallyphotoinhibited plants starved to death.

An earlier study on growth rates of seedlings (20–50 cmtall) at the same site found that height growth of survivingsouth-exposed seedlings surpassed that of north-exposedplants (C. Wirth, Max-Planck-Institut für Biogeochemie, Jena,Germany, unpublished data). Although these results were notsurprising, considering that P. sylvestris is a shade-intoleranttree (Nikolov and Helmisaari 1992), this contrasts with the re-duced photosynthetic capacity of high PPF-exposed seedlingsdocumented here. We conclude that high PPF-exposed seed-lings have a lower chance of survival than seedlings exposed tolower PPF. But once established, high-PPF-exposed plants

1148 SLOT ET AL.

TREE PHYSIOLOGY VOLUME 25, 2005

have higher growth rates and probably an increased probabil-ity of long term-survival.

Germino et al. (2002) suggested exposure to excess light ex-acerbated the effects of low temperature and water stress, thusinhibiting the establishment of conifer seedlings at the alpinetree line. Similarly, in boreal glades, greater susceptibility tosoil water deficits may interact with chronic photoinhibition toincrease mortality of SS plants. During summer, the soil watercontent in the rooting zone of the sandy soil was as low as0.02 g g– 1 and with no difference between the northern and thesouthern aspects (M. Slot, unpublished data). Probably be-cause smaller plants had not yet developed a deep taproot, theSS plants may have been more prone to desiccation than theirnorthern counterparts (soil water deficit here is reflected by theleast negative δ13C values in SS seedlings) (Table 3). Values ofδ13C in SN, as well as in the larger size class, were significantlymore negative. Reduced susceptibility to soil water deficit inaddition to light stress may also explain why, despite experi-encing a similar light environment to that of SS seedlings, LS

seedlings showed no symptoms of light stress during summer.Whether the interaction between drought stress and chronicphotoinhibition kills small south-exposed plants deserves fur-ther attention.

Herbivory is another important factor that may alter regen-eration patterns. However, herbivory by capercaillie (Tetraourogallus L.) as a potential cause of seedling mortality hasbeen observed in our study area only on recent burns, whereneedle nitrogen concentrations are twice the normal concen-tration (0.017 versus 0.008 g N gDM

– 1 ), and can therefore be ex-cluded. Moose prefer deciduous vegetation and are rarearound our sites because of unregulated hunting.

In conclusion, we demonstrated that the photosynthetic ca-pacity and efficiency of small plants (20–50 cm tall) exposedto high PPFs were reduced compared with plants growing in amore shaded environment. This pattern does not simply reflectdynamic inhibition of photosynthesis (Osmond 1994). In-stead, as revealed by the shading experiment, these patterns re-flect sustained down-regulation of the photosynthetic appara-tus as a result of the acclimation of photosynthesis to the pre-vailing environmental conditions. The higher mortality ofsouth-exposed plants compared with north-exposed plantsconfirms that the distribution of P. sylvestris regeneration inthe study area is linked to the alleviation of photoinhibition bycanopy shelter. However, the mechanism underlying the rela-tionship between photoinhibition and tree mortality may alsoinvolve drought stress interacting with photoinhibition. If so,both factors may be important in determining regenerationsuccess and the structure of the boreal Siberian pine forest. Be-cause we did not observe permanent north–south differencesin the photosynthetic performance within the larger size class,we conclude that once juveniles become established, they areable to cope with the adverse environmental conditions andmortality is no longer caused by high irradiances.

Acknowledgments

We thank Olaf Kolle for supply and advice on the use of micro-meteorological equipment, Galina Zrazhewskaya for organization of

fieldwork, Sascha Dolgushin for assistance in the field and LillianSchmidt for laboratory help.

References

Adams, III, W.W., B. Demmig-Adams, A.S. Verhoeven and D.H.Barker. 1995a. Photoinhibition during winter stress—involvementof sustained xanthophyll cycle-dependent energy-dissipation.Aust. J. Plant Physiol. 22:261–276.

Adams, III, W.W., A. Hoehn and B. Demmig-Adams. 1995b. Chillingtemperatures and the xanthophyll cycle: a comparison of warm-grown and overwintering spinach. Aust. J. Plant Physiol. 22:75–85.

Ball, M.C., V.S. Hodges and G.P. Laughlin. 1991. Cold-inducedphotoinhibition limits regeneration of snow gum at tree-line. Funct.Ecol. 5:663–668.

Ball, M.C., J.J.G. Egerton, R. Leuning, R.B. Cunningham andP. Dunne. 1997. Microclimate above grass adversely affects springgrowth of seedling snow gum (Eucalyptus pauciflora). Plant CellEnviron. 20:155–166.

Blennow, K. and L. Lindkvist. 2000. Models of low temperature andhigh irradiance and their application to explaining the risk of seed-ling mortality. For. Ecol. Manage. 135:289–301.

Close, C.D., C.L. Beadle, G.K. Holz and M.J. Hovenden. 2002. Effectof shadecloth tree shelters on cold-induced photoinhibition, foliaranthocyanin and growth of Eucalyptus globulus and E. nitens seed-lings during establishment. Aust. J. Bot. 50:15–20.

Demmig, B. and O. Björkman. 1987. Comparison of the effect of ex-cessive light on chlorophyll fluorescence (77K) and photon yield ofO2 evolution in leaves of higher plants. Planta 171:171–184.

Ensminger, I., M. Xylander, C. Hagen and W. Braune. 2001. Strate-gies providing success in a variable habitat, III. Ecophysiology ofphotosynthesis in Cladophora glomerata L. Plant Cell Environ. 24:769–779.

Ensminger, I., D. Sveshnikov, D.A. Campbell, C. Funk, S. Jansson,J. Lloyd, O. Shibistova and G. Öquist. 2004. Intermittent low tem-peratures constrain spring recovery of photosynthesis in borealScots pine forests. Global Change Biol. 10:1005–1008.

Farage, P.K. and S.P. Long. 1991. The occurrence of photoinhibitionin an over-wintering crop of oilseed rape (Brassica napus L.) andits correlation with changes in crop growth. Planta 185:279–286.

Germino, M.J. and W.K. Smith. 1999. Sky exposure, crown architec-ture and low-temperature photoinhibition in conifer seedlings atalpine treeline. Plant Cell Environ. 22:407–415.

Germino, M.J., W.K. Smith and A.C. Resor. 2002. Conifer seedlingdistribution and survival in an alpine-treeline ecotone. Plant Ecol.162:157–168.

Gilmour, A.R., R. Thompson and B.R. Cullis. 1995. Average infor-mation REML, an efficient algorithm for variance parameter esti-mation in linear mixed models. Biometrics 51:1440–1450.

Holly, C., G.P. Laughlin and M.C. Ball. 1994. Cold-induced photo-inhibition and design of shelters for establishment of Eucalyptus inpasture. Aust. J. Bot. 42:139–147.

Holt, N.E., G.R. Fleming and K.K. Niyogi. 2004. Toward an under-standing of the mechanism of nonphotochemical quenching ingreen plants. Biochemistry 43:8281–8288.

Huner, N.P.A., G. Öquist and F. Sarhan. 1998. Energy balance andacclimation to light and cold. Trends Plant Sci. 3:224–230.

Iivonen, S., R. Rikala and E. Vapaavuori. 2001. Seasonal root growthof Scots pine seedlings in relation to shoot phenology, carbohy-drate status and nutrient supply. Can. J. For. Res. 31:1569–1578.

Jermyn, M.A. 1975. Increasing the sensitivity of the anthrone methodfor carbohydrates. Anal. Biochem. 68:332–335.

TREE PHYSIOLOGY ONLINE at http://heronpublishing.com

PHOTOINHIBITION AND REGENERATION OF SCOTS PINE 1149

Kelliher, F.M., J. Lloyd, A. Arneth et al. 1998. Evaporation from acentral Siberian pine forest. J. Hydrol. 205:279–296.

Koslowski, T.T. and S.G. Pallardy. 1997. Growth control in woodyplants. Academic Press, Harcourt Brace and Company, San Diego,641 p.

Kyle, D.J., C.B. Osmond and C.J. Arntzen. 1987. Photoinhibition.Topics in photosynthesis. Vol. 9. Elsevier Science Publishers, Am-sterdam, 595 p.

Lichtenthaler, H.K. 1987. Chlorophylls and carotenoids: pigments ofphotosynthetic biomembranes. Method Enzymol. 128:350–382.

Lundmark, T. and J.-E. Hällgren. 1987. Effects of frost on shaded andexposed spruce and pine seedlings planted in the field. Can. J. For.Res. 17:1197–1201.

Müller, P., X.-P. Li and K.K. Niyogi. 2001. Non-photochemicalquenching. A response to excess light energy. Plant Physiol. 125:1558–1566.

Nikolov, N. and H. Helmisaari. 1992. Silvics of the circumpolar bo-real forest species. In A Systems Analysis of the Global BorealForest. Eds. H.H. Shugart, R. Leemans and G.B. Bonan. Cam-bridge University Press, Cambridge, pp 13–84.

Ögren, E. 1996. Relationship between temperature, respiratory loss ofsugar and premature dehardening in dormant Scots pine seedlings.Tree Physiol. 17:47–51.

Öquist, G. and N.P.A. Huner. 2003. Photosynthesis in overwinteringevergreen plants. Annu. Rev. Plant Physiol. Mol. Biol. 54:329–355.

Öquist, G., D.H. Greer and E. Ögren. 1987. Light stress at low tem-perature. In Photoinhibition. Eds. D.J. Kyle, C.B. Osmond andC.J. Arntzen. Elsevier Science Publishers, Amsterdam, pp 67–87.

Osmond C.B. 1994. What is photoinhibition? Some insights fromcomparisons of shade and sun plants. In Photoinhibition of Photo-synthesis: From Molecular Mechanisms to the Field. Eds. N.R.Baker and J.R. Bowyer. BIOS Scientific Publishing, Oxford, pp1–24.

Ottander, C.D., D. Campbell and G. Öquist. 1995. Seasonal changesin PSII organisation and pigment composition in Pinus sylvestris.Planta 197:176–183.

Pinheiro, J.C. and D.M. Bates. 2000. Mixed effect models in S andS-plus. Springer, New York, 529 p.

Schreiber, U., W. Bilger and C. Neubauer. 1994. Chlorophyll fluores-cence as a nonintrusive indicator for rapid assessment of in vivophotosynthesis. In Ecological Studies 100: Ecophysiology of Pho-tosynthesis. Eds. E.-D. Schultze and M.M. Caldwell. Springer-Verlag, Heidelberg, pp 49–70.

Schulze, E.-D., N.N. Vygodskaya, N. Tschebakova et al. 2002. TheEurosiberian transect: an introduction to the experimental region.Tellus 54B:421–428.

Syvertsen, J.P., J. Lloyd, C.A. McConchie and G.D. Farquhar. 1995.On the relationship between leaf anatomy and CO2 diffusionthrough the mesophyll of hypostomatous leaves. Plant Cell Envi-ron. 18:149–157.

Verhoeven, A.S., W.W. Adams, III, and B. Demmig-Adams. 1998.Two forms of sustained xanthophyll cycle-dependent energy dissi-pation in overwintering Euonymus kiautschovicus. Plant Cell Envi-ron. 21:893–903.

Werner, R.A. and W.A. Brand. 2001. Referencing strategies and tech-niques in stable isotope ratio analysis. Rapid Commun. MassSpectrom. 15:501–519.

Wirth, C., E.-D. Schulze, W. Schulze et al. 1999. Above-ground bio-mass and structure of pristine Siberian Scots pine forests as con-trolled by competition and fire. Oecologia 121:66–80.

1150 SLOT ET AL.

TREE PHYSIOLOGY VOLUME 25, 2005