Bernadette Nanayakkara, Frederic Lagane, Pat Hodgkiss, Mike Dibley, Simeon Smaill, Mark Riddell, Jonathan Harrington and Dave Cown

Effects of induced drought and tilting on biomass allocation, wood properties, compression wood formation and chemical composition of young Pinus radiata genotypes (clones)

De Gruyter | Published online: November 22, 2013


Eight genotypes (clones) of Pinus radiata were subjected to drought and stem inclination to assess genotype response to common stressors. While drought stress reduced diameter growth, height growth and total biomass accumulation, root to shoot (R/S) ratios were unaffected. Drought-stressed plants had significantly lower average acoustic velocity, but longitudinal shrinkage (LS) and density were not different from those of the control plants. Radial diameter growth and R/S ratios were unaffected by tilting. Inclined stems had significantly lower acoustic velocity, and significantly higher LS and density than control stems. Acoustic velocity had a strong negative correlation with LS (r2=0.79). Compression wood (CW) content was much higher in tilted plants, compared to control and drought treatment plants. The CW of tilted trees had different chemistry than that of the CW of drought and control plants. Genotypes differed significantly in the amount of CW formed as a response to tilting, demonstrating that the formation and extent of CW is genetically influenced. Mechanical perturbation in conjunction with acoustic methods for assessing stiffness would be a useful approach for early-age selection of genotypes less prone to form CW.


Wood formation is a complex process initiated in the tree crown and the vascular cambium and the resulting effects on wood properties are controlled by genetics and environmental factors (Larson 1994). Environmental stresses have adverse effects on plant growth and productivity, and are predicted to become more severe and widespread in decades to come (Osakabe et al. 2012). Understanding the cambial response to environment stresses is crucial in order to learn about their cumulative effects on the wood quality characteristics.

Many studies have examined the relationship between tree water status and xylem development in Pinus radiata D. Don (Shepherd 1964; Sheriff and Whitehead 1984; Downes 1990; Snowdon and Waring 1991; Stone et al. 2012) and other species (Rossi et al. 2009; Jyske et al. 2010; Christensen-Dalsgaard and Ennos 2012). Decreased frequency of cambial cell division, increased cell-wall thickness and decreased lumen diameter (latewood-like cells) generally occur in conifers in response to water deficit. Drought has often been associated with change in wood properties such as increased density (Christensen-Dalsgaard and Ennos 2012) or an increased density peak within earlywood (EW) (Rozenberg et al. 2002; Grabner et al. 2006). Ochroma pyramidale, Betula pendula and Acacia karro seedling stems subjected to drought had stiffer and stronger stems on average than well-watered individuals (Christensen-Dalsgaard and Ennos 2012). Eucalyptus nitens subjected to severe drought had the smallest micro fibrill angle (MFA) and when water stress was released, the MFA began to increase (Wimmer et al. 2002). Most studies on the effects of drought on biomass reported changes in biomass allocation pattern between the roots and the shoots (López et al. 2009; Schall et al. 2012; Espinoza et al. 2013).

The genetic component of a tree’s physiological response to water stress has been investigated (Monclus et al. 2006; Meier and Leuschner 2008; Rowell et al. 2009). Genetic variation was observed in growth but not in some of the drought responses such as δ13C (carbon isotope composition) and the R/S ratio.

Young plants growing in an open or wind-exposed environment, produce large amounts of randomly distributed compression wood (CW), in the process of righting themselves (Bamber 2001). In mature trees, wood quality is deteriorated by higher amounts of CW due to high longitudinal shrinkage (LS) and warp, which, in turn, can be related partly to higher cellulose MFA in CW (Harris 1977; Timell 1986b; Donaldson and Turner 2001). CW is chemically different from normal wood (NW) in that it has more galactan and lignin, and less cellulose and galactoglucomannan (Timell 1986a). As the severity of CW increases, the amount of lignin and galactan increases accordingly; hence, these act as chemical indicators of CW severity (Nanayakkara et al. 2005).

The heritability of CW has been reported to be significantly high ranging from 0.3 to 0.9 (Shelbourne et al. 1969; Apiolaza et al. 2011b). Genotypes showed differences in their propensity to form CW at age 12 (Burdon 1975). Significant differences between P. radiata seedling families were found with more susceptible families exhibiting greater initial magnitudes of stem lean (Downes and Turvey 1993; Downes et al. 1994).

The question remains whether genotypes differ in their response to mechanical stress, and whether it is possible to assess genotypes less prone to CW formation at very early ages. Sierra-De-Grado et al. (2008) induced the formation of CW in one-year-old provenances of P. pinaster and studied the biomechanical response to this stimulus. A similar approach was applied (Telewski and Jaffe 1986; Nakada 2007) for selection purposes at early ages. The rationale for this selection method is that genetic differences in a young plant’s ability to control stem form will continue throughout the life of the tree. This concept should not be confused with growing young plants at a lean as a means of obtaining NW free of CW for wood quality screening purposes (Apiolaza et al. 2011a,b).

The objective of the study was to investigate the response of different genotypes of young P. radiata trees to drought and mechanical stresses. The responses of growth, biomass, wood density, acoustic velocity, LS, CW severity, and chemical composition were measured as a function of drought and stem inclination. Also, the bacterial enzyme 1-aminocyclopropane-1-carboxylate (ACC) deaminase, which is a precursor to the plant stress hormone ethylene, was assessed to understand the soil bacterial response to drought. Such genetic differences, if present, could be valuable in early clonal selection programmes to differentiate between superior genotypes that are more drought resistant or less prone to CW formation.

Materials and methods

Experimental design

A two-way factorial design experiment (genotype×stress) was established at Scion nursery, Rotorua, New Zealand (Latitude, 38.2° and Longitude, 176.3°). Eight P. radiata genotypes were planted in 20 l planter bags with standard potting mix and drip irrigated. The planter bags were placed outside at 0.5 m intervals on permeable polythene matting in an area sheltered from wind. Five genotypes, 28, 31, 38, 41, and 45 (Forest Genetics Ltd), were improved production clones and were propagated from 6-month-old cuttings, while genotypes 708, 738, and 794 (Forest Genetics Ltd, Rotorua, New Zealand) were unimproved clones propagated by embryogenesis. The physiological age difference between these two groups of clones was approximately 6 months. Physiological ageing is the phenomenon whereby trees propagated from cuttings may exhibit some of the physiological characteristics of the older trees from which they were propagated.

Plants were grown in their planter bags without any artificial stress for 3 months prior to applying treatments to one-year-old plants. The trees were then grown under stress treatments from June 2010 to June 2011. The experiment consisted of nine plots, with three replicate plots for each of three treatments, control, tilt, and drought. Each of the replicate plots contained three individuals (or ramets) of each of the eight genotypes. In total, 216 plants included 8 genotypes×3 ramets×3 treatments×3 replicates for each treatment.


Drought stress was applied by covering the soil in each bag with a water-proof membrane skirting and restricting the watering to obtain a 50% field capacity of the potting mix by a controlled sprinkler system. Field capacity is the maximum amount of water that a soil or rock can hold, as by capillary action, before the water drains away. At regular intervals, pots were weighed and field capacity was calculated to check that the irrigation was performing as expected (Figure 1). Control and mechanical treatment watering regimes were targeted at near 100% field capacity. Mechanical stress was applied by tying the seedling stems to a stake at 5, 15, and 30 cm to maintain the lean at the lower 5–30 cm part of the stem, and then tilting the planter bag to a 30° lean. Treatments were applied from June 2010 to June 2011, and the plants were harvested during July–August 2011.

Figure 1 Variation in field capacity for drought and control pots monitored over a few months in 2011.

Figure 1

Variation in field capacity for drought and control pots monitored over a few months in 2011.

Growth and biomass measurements

The overall mean height and root collar diameter at the beginning and at the end of stress application is given in Table 1. Relative height (Δheight=Xt-Xo) and diameter increments (Δdiameter=Xt-Xo) were selected as parameters, where Xo represents the initial value (June 2010) and Xt the value after the treatment (June 2011). At harvest, stems were separated into the following components: stems, stem needles, branches, and coarse roots (>2 mm). Dry weight of the plant components was determined after drying at 70°C to a constant weight. At this stage, branch needles were stripped off from branches and weighed separately.

Table 1

Mean stem diameter at the base, and mean height before and after stress application.

Treatment June 2010 June 2011
Diam. (mm) Height (mm) Diam. (mm) Height (mm)
Control 6.2 33 24.8 109
Drought 6.0 32 22.2 91
Tilting 6.2 33 24.8 99

Assessment of ACC deaminase activity

The activity of the bacterial enzyme ACC deaminase, which is a precursor to the plant stress hormone ethylene, was assessed as an indicator of drought stress on the soil microbial community. Based on variation in growth measurements, a slow growing genotype (31) and a fast growing genotype (41) were selected for an examination of bacterial ACC deaminase activity. Soil samples (from 0 to 10 cm depth) were collected from the planter bags and ACC deaminase activity was assessed by a modified method of Smaill et al. (2010), in which 5 ml of 20 mM ACC in 0.1 M Tris buffer (pH 8.5) is the substrate for enzyme activity. Enzyme activity (α-ketobutyrate production) was determined by absorbance at 540 nm with a Fluostar Optima (BMG Labtech, Offenburg, Germany).

Compression wood (CW)

The amount of visible moderate and severe CW was assessed from red, green and blue (RGB) images made of freshly cut and wet cross ends scanned on a flatbed scanner. Images were taken at 5, 15, 20, and 30 cm up the stem from the base and an average CW content was calculated. CW areas were visually identified and manually estimated as a percentage of the whole stem cross sectional area with the GIMP 2.8 image analysis program (

Wetting the surface enhanced the colour contrast between NW and severe CW, but mild CW cannot be reliably identified visually (Donaldson et al. 1999). Owing to the large number of samples involved, microscopic examination was only performed from a subset of samples to confirm the visual identification.

Acoustic velocity

Sections (free of branches) of live plant stems (between 15 and 30 cm up the stem) were selected and longitudinal wave time-of-flight (ToF) acoustic velocity (Vs) was determined based on the method of Emms et al. (2012) with accelerometers attached at 10 cm apart along the selected section. The same 10 cm stem section was then cut from the stem and bark was removed, and the stem section was equilibrated to 12% moisture content (MC) in a climate chamber (60%, 25°C). The acoustic velocity was measured by the longitudinal-wave acoustic resonance (Vdx) technique as described by Emms et al. (2013). Modulus of elasticity (MoE) or stiffness of dry xylem was then calculated: MoE=ρν2, where ρ is dry density of the stem section.

Longitudinal shrinkage (LS) and density

This was performed on the 10 cm stem piece used for resonance measurements. LS was measured by means of a custom-designed jig attached with a micrometer. Two spherical-headed map pins were inserted in line on the opposing end faces of each specimen in the fresh green condition and secured by epoxy putty. The pinheads formed the reference points for length measurements. The spherical heads of the pins rested precisely on the tubular ends of the jig, one of which corresponded to the hollow cylindrical end of the stem of the micrometer. The displacement of the micrometer was recorded. Shrinkage was measured from green to oven dry (55°C) condition. LS was calculated: LS=100(Lg-Lod)/Lg, where, Lg=length of green stem, Lod=length after oven drying. Air-dry wood density was assessed by obtaining the weight and volume (by water immersion) of the stem sections at 12% MC.


To assess the impact of stress on cell-wall chemical composition, CW and NW areas were sampled separately and analysed. Chemical analysis was only performed on genotypes 28, 31, 38, 41, and 45, and for a particular genotype and a wood type, ground wood prepared from three individual ramets was pooled to form a single sample. A stem piece taken at 5–15 cm from the base was sliced into thin discs with a band saw and visually identified CW and NW areas were carefully isolated by using a chisel. All wood samples were air-dried, ground in a Wiley mill to pass a 20-mesh screen, then extracted with dichloromethane overnight in a Soxhlet extractor and reground to pass a 60-mesh screen. The total lignin content was determined in duplicate as the sum of the Klason lignin plus acid-soluble lignin following standard methods (Tappi standard T222 om-88 1988; Tappi standard UM-250 1991) scaled down to analyse 0.25 g wood. Monomeric sugars in the filtrates from Klason lignin determination were analysed by ion chromatography (Petterson and Schwandt 1991) and the results were expressed as anhydro sugar units.

Data analysis

The statistical analysis was done by means of the mixed procedure of SAS statistical analysis package (SAS version 9.2; Firstly, the plot means were calculated for each of the nine plots. Statistical analysis was carried out on the basis of these plot means as the experimental units. A latin-square approach with stress treatment as the mainplot factor was applied with genotype as the subplot factor, and an interaction factor (treatment×genotype). Significant differences were identified with the Tukey-Kramer test for multiple comparisons among least square means. A probability level of 0.05 was assumed to represent a significant difference.

Results and discussion

Growth and biomass

Diameter and height growth of drought-stressed plants (Tdrought) were significantly lower than those of the control plants (Tcontrol), whereas in case of tilt treatment (Ttilting) these parameters were not significantly different from those of the Tcontrol (Table 2). The ratio “height to diameter growth” for treated materials was essentially the same as that of the Tcontrol. As expected, the decreased water availability reduced the annual height and radial increment, in agreement with previous reports on P. radiata plants (Downes 1990). During drought stress, cambium enters a semi-dormant state, which inhibits cell division and stem growth (Shepherd 1964). There were no significant differences in growth between leaned and straight young P. radiata plants (Apiolaza et al. 2011a). Similarly, there was no significant difference in growth between staked and the unstaked plants (Downes et al. 1993). There were significant differences in the height and diameter increment across genotypes (P<0.0001; Table 3). Clone 41 was the fastest growing among all genotypes while 708 was the slowest in terms of diameter growth and 45 was the slowest in terms of height. The treatment×genotype interaction was significant for height increment (P=0.032) but not for the diameter increment.

Table 2

Least square means of morphological features and biomass allocation.

Treatment ΔHt (cm) ΔDiam. (mm) Ratio ΔHt/ΔDia Stem (g) Foliage (g) Branches (g) Foliage (g) Total massa (g) Roots (g) Ratio root/shoot
Control 76.4b 18.7a 4.1a 102.5bc 34.7a 37.3b 74.0a 247.6a 152.0a 0.6194a
Drought 58.8a 16.3b 3.6a 75.7a 32.8a 30.9a 79.8a 217.0a 129.4b 0.6188a
Tilting 65.9ab 18.6a 3.6a 101.3b 35.3a 39.0bc 94.6a 270.2ab 156.64a 0.5856a
SE 3.3 0.5 0.1 5.2 1.0 1.8 10.0 10.5 4.6 0.0279
F-value 7.15 7.16 4.98 8.58 1.63 6.40 1.13 6.43 10.24 0.91
P-value 0.0258 0.0258 0.0531 0.0174 0.2721 0.0325 0.3829 0.0322 0.0116 0.4502

aAbove ground.

Means in the same column followed by different letters indicate significant differences between different treatment means (Tukey-Kramer’s SD, P=0.05). SE, standard error of treatment means.

Table 3

Morphological features and biomass allocation by genotype.

Genotype ΔHt (cm) ΔDiam. (mm) Ratio ΔHt/ΔDia Stem (g) Stem Foliage (g) Branches (g) Branch Foliage (g) Total Massa (g) Roots (g) Ratio root/shoot
28 62.3 18.7 3.34 110.4 37.3 48.5 96.7 293.8 175.9 0.5852
31 65.3 16.9 3.88 84.0 31.8 42.6 90.4 250.5 166.9 0.6913
38 70.4 18.2 3.86 123.3 42.4 41.6 92.6 300.0 170.1 0.5675
41 77.1 19.2 4.05 129.0 36.4 40.1 78.5 284.4 146.1 0.5152
45 60.5 17.2 3.53 87.1 32.1 41.2 125.1 285.2 187.5 0.6747
708 62.0 16.6 3.78 54.4 28.9 16.6 40.1 138.7 95.0 0.7017
738 71.5 17.7 4.09 67.9 31.9 25.2 68.5 193.5 101.5 0.5401
794 65.0 18.4 3.62 87.9 32.6 33.5 73.2 225.4 126.7 0.5687
SE 2.7 0.52 0.1 4.8 1.6 2.3 8.5 11.7 6.2 0.0279
F-value 7.60 4.06 5.45 42.8 7.0 23.3 13.5 29.0 34.4 6.96
P-value <0.0001 0.0017 0.0002 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001

aAbove ground.

SE, standard error of genotype means.

Across all genotypes, above ground and below ground biomass accumulation was less (13 and 15%, respectively) in the Tdrought compared to the Tcontrol (Table 2). An average root to shoot (R/S) ratio of 0.60 was observed across all treatments. In most studies focusing on drought stress, the R/S ratio was changed (López et al. 2009; Schall et al. 2012; Espinoza et al. 2013). However, four-year-old Fagus sylvatica L. (European beech) plants (Meier and Leuschner 2008) and Norway spruce plants (Schall et al. 2012) subjected to summer drought did not show any changes in R/S ratios.

In the Ttilting, total above ground biomass accumulation was higher than that of Tcontrol due to the fact that this treatment had more branches and branch foliage than the Tcontrol. The interpretation is that inhibiting the movement of the stem in response to external forces fosters the apical control over the stem formation and impedes branch production. A reduction in branch number in the staked trees over the unstaked plants has been observed previously (Downes et al. 1993).

It is interesting to compare the differences in response to tilting and flexing. While our results showed that Ttilting does not affect growth and R/S ratios significantly, wind or other mechanical perturbations such as flexing had a direct effect of increasing radial growth, reducing stem elongation, and decreasing total biomass above ground (Telewski and Jaffe 1986; Stokes et al. 1997; Telewski and Pruyn 1998; Kern et al. 2005).

Significant clonal differences were seen in the R/S ratio (Table 3). Genotype 708 had the highest R/S ratio of 0.70 and clone 41 the lowest of 0.51. The treatment×genotype interaction was significant for stem (P=0.032), total above-ground biomass (P=0.020) and roots (P=0.012). Plants grown in planter bags showed significantly higher R/S ratios (0.70–0.51) compared to plants grown in nursery beds, which had R/S ratios around 0.2 (Beets et al. 2007). Similar ratios have been reported with young Sitka spruce plants grown in pots to determine the effect of flexing and nutrition, which had R/S ratios ranging from 0.62 to 0.78 (Stokes et al. 1997).

ACC deaminase activity

ACC deaminase activity was significantly greater in Tdrought than in Tcontrol, (Table 4), showing that the bacterial community responds to the drought stress of the plants. The stressed plants exude ACC and reduce ethylene production (Glick et al. 1998; Hontzeas et al. 2006). Enzyme activity did not vary with genotype and the interaction term was not significant (P=0.180). Accordingly, genotypes 31 and 41 did not differ in their ability to influence ACC deaminase activity in the soil bacterial community. A tree genotype capable of influencing soil bacteria to increase the activity of this enzyme will likely have increased tolerance to drought stress.

Table 4

Effect of drought and genotype on ACC deaminase activity as determined by α-ketobutyrate production.

Treatment α-ketobutyrate [μg (g soil)-1 day-1]
Control 107 (22)b
Drought 241 (30)a
Genotype 31 165 (45)
Genotype 41 184 (25)

SE values are given in parentheses. Means in the same column followed by different letters indicate significant (P=0.05) differences.

Wood density, shrinkage and resonance velocity (Vdx)

In-stem (Vs) and dry xylem resonance velocity (Vdx) in drought-stressed stems were significantly lower than those in Tcontrol; however, LS and density were not different from those of the Tcontrol (Table 5). Probably, drought-stressed plants formed only a mild CW (see section on CW severity), hence wood properties such as LS were not affected seriously. Drought response may vary between species and higher density may be one of the responses to drought (Jyske et al. 2010; Christensen-Dalsgaard and Ennos 2012). P. radiata plants subjected to intermittent drought and irrigation had average densities lower than the well-watered trees (Drew et al. 2011). The densities of EW and LW are different because of the differences in their lumen diameter and cell-wall thicknesses. Therefore, the average density resulting from Tdrought might be dependent on the circumstances of drought stress, which influences the LW production.

Table 5

Least square means of wood properties.

Treatment CW (%) Acoustic properties MoEa (GPa) Shrinkage (gr. to o.d.)b (%) Density (kg m-3)
Vs (m s-1) Vdx (m s-1)
Control 7 1172a 2895a 9.3a 0.751a 437a
Drought 5 1084b 2711b 8.4ab 0.804a 442a
Tilting 31 1094b 2658b 7.9b 1.255b 465b
SE 0.90 25.6 34.8 0.2 0.050 5.1
F-value 260 5.10 11.75 11.33 31.33 8.1
P-value <0.0001 0.0299 0.0084 0.0092 0.0007 0.0195

aModulus of elasticity for dry xylem.

bGreen to oven dry.

Vs, in stem acoustic velocity; Vdx, resonance velocity of dry xylem; CW, compression wood; SE, standard error; F-value, value calculated; Pr, probability calculated from F-value. Means in the same column followed by different letters indicate significant differences between different treatments (Tukey-Kramer’s SD, P=0.05).

The stems of Ttilting, had a lower Vdx (or assumed lower stiffness), and higher LS and density compared to the Tcontrol (all statements significant at 0.05 level). There was nearly two-fold difference in LS values between the Ttilting and the Tcontrol. The formation of elevated amounts of severe CW in the Ttilting resulted in stems of lower stiffness, and, therefore, of lower acoustic velocity. Both the reduced stiffness and higher LS are due to higher MFA in the S2 layer of CW (Yin et al. 2011). These results are in agreement with measured wood properties of 18-month-old seedling CW formed by tilting wherein shrinkage increased and velocity decreased with increasing MFA (Apiolaza et al. 2011a,b).

Significant (P<0.001) genotypic differences were observed concerning Vdx or stiffness. Clone 31 had the highest Vdx out of five production clones tested while clone 708 had the lowest Vdx out of all clones and out of three non-production clones (Table 6). The large differences in Vdx between genotypes might be due to extremes in the MFA in juvenile wood (JW). In JW a strong negative correlation between MFA and acoustic velocity has been reported (Cown et al. 1999; Lachenbruch et al. 2010; Hasegawa et al. 2011). A very high MFA in 708 would result in a high LS and a low Vdx. A better knowledge about the MFA may be the key to understanding these effects better.

Table 6

Mean values of measured wood properties by genotype.

Genotype CW

Acoustic properties MoEa


(gr. to o.d.)b (%)

(kg m-3)

(m s-1)

(m s-1)
28 13 1428 2679 7.8 0.956 434
31 18 1782 3306 12.0 0.558 492
38 13 1458 2860 9.0 0.836 477
41 11 1494 2953 9.6 0.627 456
45 14 1413 2710 8.1 0.892 448
708 12 1190 2359 6.2 1.620 441
738 8 1227 2618 7.5 0.988 421
794 24 1359 2652 7.7 1.090 415
SE 1.45 20.1 36.3 0.23 0.079 7.53
F-value 11.8 17.35 65.5 68.3 13.1 12.5
P-value <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001

aModulus of elasticity for dry xylem.

bGreen to oven dry.

Vs, in stem acoustic velocity; Vdx, resonance velocity of dry xylem; CW, compression wood; SE, standard error.

LS and density showed significant differences between genotypes. Clone 708 had the highest LS and clone 31 the highest density out of all genotypes studied, resulting in a significant treatment×genotype interaction for shrinkage (P=0.028) and density (P=0.006).

While Vdx (and derived stiffness) was positively correlated with in-stem acoustic velocity (Vs) (r2=0.88), it was negatively correlated with LS (r2=0.79) (Figure 2). Clone 31, with the highest velocity (Vdx), had the lowest LS, and clone 708, with the lowest velocity, had the highest LS. In 18-month-old P. radiata plants, both green and Vdx data were highly negatively correlated with LS and volumetric dimensional shrinkage (Apiolaza et al. 2011b). Acoustic measurements coupled with mechanical perturbation have a high potential as an early screening tool for dimensional stability, as suggested in the literature (Lindstrom et al. 2002, 2005; Apiolaza et al. 2011b)

Figure 2 Relationship between LS and dry xylem resonance velocity.

Figure 2

Relationship between LS and dry xylem resonance velocity.

Compression wood (CW)

CW was present in all treatments, substantially smaller amounts in both the Tcontrol (0–13%) and Tdrought (0–8%) compared to Ttilting (40–60%), which was significantly different from Tcontrol and Tdrought (P=<0.0001) (Table 5). CW in vertical stems results from corrections to a previous leaning or from small, less obvious asymmetric loadings within the stem. Occurrence of 13–17% CW in rocked and straight plants has been reported (Apiolaza et al. 2011a), apparently caused by the stem movement. Several studies have suggested that the formation of CW aids stems to remain straight, and can be prevalent even in straight trees (Warensjö and Rune 2004). There was no consistent pattern between percentage CW at different heights up the stem although, in general, more CW was detected in the lower part of the plant stem. One clear observation was the high variability of the trees’ response to Ttilting. This observation is in agreement with data of Lachenbruch et al. (2010), where CW in inclined five-year-old P. radiata varied from 0 to 59% and where higher values were found at the tree bases.

Genotypes differed significantly in the amount of CW formed (P=<0.0001) (Table 6, Figure 3). Clone 794 produced large amounts of CW under Ttilting. The interaction between genotype and CW formed in response to Ttilting was significant (P=0.004). The genotypes tested here differed significantly in their response to Ttilting. Hence, warp propensity could be screened at the seedling stage by evaluating the propensity of CW formation through mechanical perturbation. Three seven-year-old P. radiata clones subjected to the same 45° lean had differing amounts of visual CW (Lindstrom et al. 2005). Telewski and Jaffe (1981) observed that six-month-old P. taeda plants of different genetic background respond in slightly different ways to identical mechanical perturbations under similar environmental conditions.

Figure 3 The average amount of CW formed by each genotype in response to the stress treatments. Values shown are treatment×genotype means±standard errors.

Figure 3

The average amount of CW formed by each genotype in response to the stress treatments. Values shown are treatment×genotype means±standard errors.

The LS was positively correlated with the amount of CW (r2=0.46) (Figure 4). However, this correlation improved (r2=0.77) when genotype 708 was excluded. Genotype 708 had the highest LS of all genotypes tested but it does not appear to be driven by the amount of CW formed. Generally genotypes that produced higher amounts of CW showed higher LS, e.g., clone 794. According to Öhman (1999), the amount of CW, which is relatively easy to measure, could be a rough indicator for LS. LS is highly correlated with MFA (Xu et al. 2009), and the CW MFA is higher than that of NW. Core wood or JW is known to have very high MFA (>40°), and in such cases the correlation between MFA and LS is no longer valid.

Figure 4 Relationship between LS and CW content.

Figure 4

Relationship between LS and CW content.

Chemical composition and CW severity

Drought and tilting stresses did not alter the chemical composition of NW beyond the usual between-tree variation (Kibblewhite et al. 2010); however, CW chemistry was significantly affected (Table 7). CW in the Tcontrol and Tdrought had lignin and galactose contents of 31–33% and 5–8%, respectively, while in Ttilting, lignin and galactose contents were in the range 37–40% and 9–12%, respectively (Figure 5). There was a linear relationship between galactose and Klason lignin content showing three clusters for NW, mild CW, and severe CW (Figure 5). Changes in CW severity are associated with increased lignin and galactose contents, and decreased glucose content (Nanayakkara et al. 2009; Kibblewhite et al. 2010). Based on the results in the present paper it is obvious that, Tcontrol and Tdrought produce mild to moderate CW, while Ttilting leads to severe CW (Nanayakkara et al. 2009; Brennan et al. 2012).

Table 7

Least square means of chemical composition of normal wood (NW) and compression wood (CW) (%).

Klason lignin Arabinose Galactose Glucose Xylose Mannose
Control 27.94a 31.71a 2.32a 1.94a 2.40a 5.31a 43.45a 40.01a 8.68a 7.20a 10.47a 9.15a
Drought 28.20a 31.77a 2.70b 2.24b 2.92b 7.04b 45.25b 40.45a 9.82b 7.61a 10.62a 9.15a
Tilting 28.34a 36.76b 2.35a 1.68c 2.92b 10.01c 42.26a 34.87b 8.70a 6.30b 10.87a 7.75b
SE 0.24 0.24 0.02 0.03 0.06 0.35 0.35 0.49 0.07 0.21 0.12 0.17
F-value 0.75 154.07 84.85 65.15 28.08 50.16 18.69 42.28 90.53 10.42 2.99 29.01
P-value 0.5103 <0.0001 <0.0001 <0.0001 0.0009 0.0002 0.0026 0.0003 <0.0001 0.0112 0.1255 0.0008

Means in the same column followed by different letters indicate significant differences between treatments (Tukey’s SD, P=0.05). SE, standard error.

Figure 5 Relationship between lignin content and galactose content.

Figure 5

Relationship between lignin content and galactose content.

In CW, arabinose and xylose showed significant variation among genotypes. In NW, all wood chemical components, except lignin, revealed significant variation among genotypes (Table 8). The interaction between treatment and genotype was significant in the case of CW lignin content and NW xylose content. According to our results, carbohydrate composition of NW is genetically controlled. Donaldson et al. (1997) investigated the clonal variability of 16-year-old P. radiata and found significant variation among clones in all wood chemical components except for glucose. The authors emphasised the necessity for sampling of CW and NW separately for an objective comparison between genotypes.

Table 8

Genotype means of chemical composition of normal wood (NW) and compression wood (CW) (%).

Klason lignin Arabinose Galactose Glucose Xylose Mannose
28 27.77a 33.23a 2.49b 1.92ab 2.56a 7.24a 43.51ab 38.21a 9.35b 6.99ab 10.46a 8.44a
31 28.11a 33.70a 2.30a 1.76a 2.81a 7.50a 44.72b 39.23a 8.81a 7.03ab 10.62ab 8.78a
38 28.37a 32.76a 2.33a 1.94b 2.83a 6.78a 43.73ab 39.25a 9.28b 7.28ab 10.68ab 9.12a
41 28.63a 33.93a 2.53cb 2.00cb 2.66a 7.54a 42.52a 37.37a 9.28b 7.33b 11.08b 8.90a
45 27.92a 33.45a 2.63c 2.14c 2.86a 8.22a 43.79ab 38.14a 8.61a 6.53a 10.41a 8.44a
SE 0.24 0.31 0.03 0.04 0.07 0.45 0.45 0.56 0.08 0.20 0.12 0.22
F-value 2.61 2.26 22.50 11.03 3.16 1.39 3.03 2.22 14.23 3.06 5.45 1.93
P-value 0.0605 0.0957 <0.0001 <0.0001 0.0321 0.2715 0.0371 0.0994 <0.0001 0.0378 0.0029 0.1405

Means in the same column followed by different letters indicate significant differences between genotypes (Tukey-Kramer’s SD, P=0.05). SE, standard error.


Different genotypes displayed significant differences in growth but no differences in biomass accumulation under drought stress. Apparently, drought response in P. radiata plants is not under genetic control. Drought stress does not affect wood properties significantly at an early age, but this statement still needs confirmation. Drought not only reduces P. radiata growth, it also leads to the formation of a mild type of CW. In contrast, genotypes responded differently to tilting stress, and it was interpreted that the formation and extent of CW is genetically controlled. This variability can be detected in young trees by means of acoustic tools, which could be beneficial in clonal selection programmes to differentiate between superior genotypes at an early age of around three years. Testing young plants in clonal selection programmes has many advantages. For example, small scale genetic tests at closer spacing, shorter time span for a screening cycle, young stems are easier to handle and measure, large number of individuals can be tested, and a better control of environment conditions can be achieved by growing of plants in containers.


The authors wish to thank Scion nursery staff for setting up and maintaining the trial, Christine Dodunski, Rodrigo Osorio and John Lee for assistance with field work, Grant Emms for his valuable advice on acoustic measurements, Mark Kimberley for advice on experimental design, and Mike Watt for reviewing the manuscript. This study was supported by the Future Forest Research Ltd.


Apiolaza, L.A., Butterfield, B.G., Chauhan, S.S., Walker, J.C.F. (2011a) Characterization of mechanically perturbed young stems: can it be used for wood quality screening? Ann. Forest Sci. 68:407–414. Search in Google Scholar

Apiolaza, L.A., Chauhan, S.S., Walker, J.C.F. (2011b) Genetic control of very early compression and opposite wood in Pinus radiata and its implications for selection. Tree Genet. Genom. 7:563–571. Search in Google Scholar

Bamber, R.K. (2001) A general theory for the growth stresses in reaction wood: How trees stay upright. IAWA J. 22:205–212. Search in Google Scholar

Beets, P.N., Pearce, S.H., Oliver, G.R., Clinton, P.W. (2007) Root/shoot ratios for deriving below-ground biomass of Pinus radiata clones. New Zealand J. Forest Sci. 37:267–288. Search in Google Scholar

Brennan, M., McLean, J.P., Altaner, C.M., Ralph, J., Harris, P.J. (2012) Cellulose microfibril angles and cell-wall polymers in different wood types of Pinus radiata. Cellulose 19:1385–1404. Search in Google Scholar

Burdon, R.D. (1975) Compression wood in Pinus radiata clones on four different sites. New Zealand J. Forest Sci. 5:152–164. Search in Google Scholar

Christensen-Dalsgaard, K.K., Ennos, A.R. (2012) Effects of drought acclimation on the mechanical properties of Ochroma pyramidale, Betula pendula and Acacia karroo tree seedling stems. Forestry 85:215–223. Search in Google Scholar

Cown, D.J., Herbert, J., Ball, R.D. (1999) Modelling Pinus radiata lumber characteristics. New Zealand J. Forest Sci. 29: 203–213. Search in Google Scholar

Donaldson, L.A., Turner, J.C.P. (2001) The influence of compression wood and microfibril angle on the occurrence of distortion in window frames made from radiata pine (Pinus radiata). Holz Roh. Werkst. 59:163–168. Search in Google Scholar

Donaldson, L.A., Croucher, J.M., Uprichard, J.M. (1997) Clonal varation of wood chemistry in radiata pine (Pinus radiata D. Don) wood. Holzforschung 51:537–542. Search in Google Scholar

Donaldson, L.A., Singh, A.P., Yoshinaga, A., Takabe, K. (1999) Lignin distribution in mild compression wood of Pinus radiata. Can. J. Botany 77:41–50. Search in Google Scholar

Downes, G. (1990) Does water-stress lead to formation of traumatic tissue and tracheid collapse in poorly lignified Pinus radiata. Forest Ecol. Manage. 30:139–145. Search in Google Scholar

Downes, G.M., Turvey, N.D. (1993) Relationships between stem structure and bending strength in Pinus radiata seedlings. Trees 7:86–91. Search in Google Scholar

Downes, G.M., Beckers, E.P.J., Turvey, N.D., Porada, H. (1993) Strength and structure of stems from fast grown Pinus radiata. Trees 7:131–136. Search in Google Scholar

Downes, G.M., Moore, G.A., Turvey, N.D. (1994) Variations in response to induced stem bending in seedlings of Pinus radiata. Trees Struct. Funct. 8:151–159. Search in Google Scholar

Drew, D.M., Downes, G.M., Evans, R. (2011) Short-term growth responses and associated wood density fluctuations in variously irrigated Eucalyptus globulus. Trees Struct. Funct. 25:153–161. Search in Google Scholar

Emms, G., Nanayakkara, B., Harrington, J. (2012) A novel technique for non-damaging measurement of sound speed in seedlings. Eur. J. For. Res. 131:1449–1459. Search in Google Scholar

Emms, G., Nanayakkara, B., Harrington, J. (2013) Application of long-wave time-of-flight sound speed measurements to Pinus radiata seedlings. Can. J. For. Res. 43:750–756. Search in Google Scholar

Espinoza, S.E., Magni, C.R., Martinez, V.A., Ivkovic, M. (2013) The effect of water availability on plastic responses and biomass allocation in early growth traits of Pinus radiata D. Don. Forest Systems 22:3–14. Search in Google Scholar

Glick, B.R., Penrose, D.M., Li, J.P. (1998) A model for the lowering of plant ethylene concentrations by plant growth-promoting bacteria. J. Theor. Biol. 190:63–68. Search in Google Scholar

Grabner, M., Cherubini, P., Rozenberg, P., Hannrup, B. (2006) Summer drought and low earlywood density induce intra-annual radial cracks in conifers. Scand. J. For. Res. 21:151–157. Search in Google Scholar

Harris, J.M. (1977) Shrinkage and density of radiata pine compression wood in relation to its anatomy and mode of formation. New Zealand J. Forest Sci. 7:91–106. Search in Google Scholar

Hasegawa, M., Takata, M., Matsumura, J., Oda K. (2011) Effect of wood properties on within-tree variation in ultrasonic wave velocity in softwood. Ultrasonics 51:296–302. Search in Google Scholar

Hontzeas, N., Hontzeas, C.E., Glick, B.R. (2006) Reaction mechanisms of the bacterial enzyme 1-aminocyclopropane-1-carboxylate deaminase. Biotechnol. Adv. 24:420–426. Search in Google Scholar

Jyske, T., Holtta, T., Makinen, H., Nojd, P., Lumme, I., Spiecker, H. (2010) The effect of artificially induced drought on radial increment and wood properties of Norway spruce. Tree Physiol. 30:103–115. Search in Google Scholar

Kern, K.A., Ewers, F.W., Telewski, F.W., Koehler, L. (2005) Mechanical perturbation affects conductivity, mechanical properties and aboveground biomass of hybrid poplars. Tree Physiol. 25:1243–1251. Search in Google Scholar

Kibblewhite, R.P., Suckling, I.D., Evans, R., Grace, J.C., Riddell, M.J.C. (2010) Lignin and carbohydrate variation with earlywood, latewood, and compression wood content of bent and straight ramets of a radiata pine clone. Holzforschung 64:101–109. Search in Google Scholar

Lachenbruch, B., Johnson, G.R., Downes, G.M., Evans, R. (2010) Relationships of density, microfibril angle, and sound velocity with stiffness and strength in mature wood of Douglas-fir. Can. J. For. Res. 40:55–64. Search in Google Scholar

Larson, P.R. The Vascular Cambium – Development and Structure. Springer-Verlag, Berlin, Heidelberg, 1994. Search in Google Scholar

Lindstrom, H., Harris, P., Nakada, R. (2002) Methods for measuring stiffness of young trees. Holz Roh Werkst. 60:165–174. Search in Google Scholar

Lindstrom, H., Evans, R., Reale, M. (2005) Implications of selecting tree clones with high modulus of elasticity. New Zealand J. Forest Sci. 35:50–71. Search in Google Scholar

López, R., Rodríguez-Calcerrada, J., Gil L. (2009) Physiological and morphological response to water deficit in seedlings of five provenances of Pinus canariensis: Potential to detect variation in drought-tolerance. Trees Struct. Funct. 23:509–519. Search in Google Scholar

Meier, I.C., Leuschner, C. (2008) Genotypic variation and phenotypic plasticity in the drought response of fine roots of European beech. Tree Physiol. 28:297–309. Search in Google Scholar

Monclus, R., Dreyer, E., Villar, M., Delmotte, F.M., Delay, D., Petit, J.M., Barbaroux, C., Thiec, D., Brechet, C., Brignolas, F. (2006) Impact of drought on productivity and water use efficiency in 29 genotypes of Populus deltoides x Populus nigra. New Phytologist 169:765–777. Search in Google Scholar

Nakada, R. (2007) Within-tree variation of wood characteristics in confiers and the anatomical characteristics specific to very young trees. In: The Compromised Wood Workshop 2007. Eds. Entwistle, K.M., Harris, P. and Walker, J. University of Canterbury, Christchurch, New Zealand. pp. 51–67. Search in Google Scholar

Nanayakkara, B., Manley-Harris, M., Suckling, I.D., Donaldson, L.A. Chemical characterisation of compression wood in Pinus radiata. Paper presented at the Proceedings from the 59th Appita Annual Conference, Auckland, New Zealand, 2005. Search in Google Scholar

Nanayakkara, B., Manley-Harris, M., Suckling, I.D., Donaldson, L.A. (2009) Quantitative chemical indicators to assess the gradation of compression wood. Holzforschung 63:431–439. Search in Google Scholar

Öhman, M. (1999) Correspondence between manually estimated compression wood in Norway spruce and the warp of the sawn timber. Holz als Roh-und Werkstoff 57:391–396. Search in Google Scholar

Osakabe, Y., Kawaoka, A., Nishikubo, N., Osakabe, K. (2012) Responses to environmental stresses in woody plants: key to survive and longevity. J. Plant Res. 125:1–10. Search in Google Scholar

Petterson, R.C., Schwandt, V.H. (1991) Wood sugar analysis by anion chromatography. J. Wood Chem. Technol. 11:495–501. Search in Google Scholar

Rossi, S., Simard, S., Rathgeber, C.B.K., Deslauriers, A., De Zan, C. (2009) Effects of a 20-day-long dry period on cambial and apical meristem growth in Abies balsamea seedlings. Trees Struct. Funct. 23:85–93. Search in Google Scholar

Rowell, D.M., Ades, P.K., Tausz, M., Arndt, S.K., Adams, M.A. (2009) Lack of genetic variation in tree ring delta13C suggests a uniform, stomatally-driven response to drought stress across Pinus radiata genotypes. Tree Physiol. 29:191–198. Search in Google Scholar

Rozenberg, P., Van Loo, J., Hannrup, B., Grabner, M. (2002) Clonal variation of wood density record of cambium reaction to water deficit in Picea abies (L.) Karst. Ann. Forest Sci. 59:533–540. Search in Google Scholar

Schall, P., Loedige, C., Beck, M., Ammer, C. (2012) Biomass allocation to roots and shoots is more sensitive to shade and drought in European beech than in Norway spruce seedlings. Forest Ecol. Manage. 266:246–253. Search in Google Scholar

Shelbourne, C.J., Zobel, B.J., Stonecypher, R.W. (1969) The inheritance of compression wood and its genetic and phenotypic correlations with six other traits in 5-year-old loblolly pine. Silvae Genet. 18:43–47. Search in Google Scholar

Shepherd, K.R. (1964) Some obeservations on the effect of drought on the growth of Pinus radiata D.Don. Aust. For. Res. 28:7–22. Search in Google Scholar

Sheriff, D.W., Whitehead, D. (1984) Photosynthesis and wood strucutre in Pinus radiata D. Don during dehydration and immediately after rewatering. Plant Cell Environ. 7:53–62. Search in Google Scholar

Sierra-De-Grado, R., Pando, V., Martinez-Zurimendi, P., Penalvo, A., Bascones, E., Moulia, B. (2008) Biomechanical differences in the stem straightening process among Pinus pinaster provenances. A new approach for early selection of stem straightness. Tree Physiol. 28:835–846. Search in Google Scholar

Smaill, S.J., Leckie, A.C., Clinton, P.W., Hickson, A.C. (2010) Plantation management induces long-term alterations to bacterial phytohormone production and activity in bulk soil. Appl. Soil Ecology 45:310–314. Search in Google Scholar

Snowdon, P., Waring, H.D. (1991) Effects of irrigation and artificial drought on the growth and health of Pinus radiata near Canberra, A.C.T. Australian Forestry 54:174–186. Search in Google Scholar

Stokes, A., Nicoll, B.C., Coutts, M.P., Fitter, A.H. (1997) Responses of young Sitka spruce clones to mechanical perturbation and nutrition: effects on biomass allocation, root development, and resistance to bending. Can. J. For. Res. 27: 1049–1057. Search in Google Scholar

Stone, C., Penman, T., Turner, R. (2012) Managing drought-induced mortality in Pinus radiata plantations under climate change conditions: A local approach using digital camera data. Forest Ecol. Manage. 265:94–101. Search in Google Scholar

Telewski, F.W., Jaffe, M.J. (1981) Thigomorphogenesis: changes in the morphology and chemical composition induced by mechanical perturbation in 6-month-old Pinus teada seedlings. Can. J. For. Res. 11:380–387. Search in Google Scholar

Telewski, F.W., Jaffe, M.J. (1986) Thigmomorphogenesis: anatomical, morphological and mechanical analysis of genetically different sibs of Pinus taeda in response to mechanical perturbation. Physiol. Plant. 66:219–226. Search in Google Scholar

Telewski, F.W., Pruyn, M.L. (1998) Thigmomorphogenesis: a dose response to flexing in Ulmus americana seedlings. Tree Physiol. 18:65–68. Search in Google Scholar

Timell, T.E. (1986a) Chemical properties of compression wood. In: Compression Wood in Gymnosperms Volume 1. Springer-Verlag, Berlin. pp. 289–408. Search in Google Scholar

Timell, T.E. (1986b) Compression wood in lumber, plywood and board manufacturing. In: Compression Wood in Gymnosperms Volume 3. Springer-Verlag, Berlin. pp 1799–1830. Search in Google Scholar

Warensjö, M., Rune, W. (2004) Stem straightness and compression wood in a 22-year-old stand of container-grown Scots pine trees. Silva Fenn. 38:143–153. Search in Google Scholar

Wimmer, R., Downes, G.M., Evans, R. (2002) Temporal variation of microfibril angle in Eucalyptus nitens grown in different irrigation regimes. Tree Physiol. 22:449–457. Search in Google Scholar

Xu, P., Liu, H.W., Evans, R., Donaldson, L.A. (2009) Longitudinal shrinkage behaviour of compression wood in radiata pine. Wood Sci. Technol. 43:423–439. Search in Google Scholar

Yin, Y.F., Bian, M.M., Song, K.L., Xiao, F.M., Jiang, X.M. (2011) Influence of microfibril angle on within-tree variations in the mechanical properties of chinese fir (Cunninghamia lanceolata). IAWA J. 32:431–442. Search in Google Scholar

Received: 2013-3-27
Accepted: 2013-10-25
Published Online: 2013-11-22
Published in Print: 2014-5-1

©2014 by Walter de Gruyter Berlin/Boston