The effect of planting spacing on Pinus patula stem straightness, microfibril angle and wood density

The effect of planting spacing on Pinus patula stem straightness, microfibril angle and wood density Abstract Improved growth rates and shorter rotation ages have caused a reduction in the stiffness of structural lumber from South African-grown pine plantations. Microfibril angle (MFA) and wood density are known to be two wood properties that influence wood stiffness. Therefore, the objective of this study was to determine the effect of planting spacing of Pinus patula trees, on the MFA and wood density, as well as stem straightness. A total of 171 trees from four spacing treatments (403, 1097, 1808 and 2981 stems ha−1) from an 18-year old experimental P. patula plantation located in Mpumalanga, South Africa, were analysed for wood density, MFA, and ring width. A sub-sample of 81 trees was scanned for tree form using a terrestrial laser scanner. A non-linear mixed-effects model using a power function was developed to model MFA and wood density as a function of ring number and ring width. Planting spacing had a highly significant effect on stem straightness with the most widely spaced trees having the worst mean stem straightness. However, the stem straightness did not increase consistently with increasing stems ha−1. The dynamic modulus of elasticity of standing P. patula trees increased greatly with closer spacing – more so than any other species reported in literature. The mixed model showed that, after accounting for differences due to ring number and ring width, spacing treatment had a significant effect on both the initial MFA and its rate of change with age. For wood density, this remaining effect of spacing treatment was only displayed in its radial rate of change. Based on these results, it seems as if planting spacing has great potential as a management intervention to improve the mechanical wood properties and in certain cases also the stem straightness of South African-grown P. patula at final harvest. Introduction Of all sawn wood produced and sold in South Africa, ~75 per cent is regarded as structural lumber (Crickmay and Associates, 2015), making it the single most important product category for local sawmills. The most important tree resource for these lumber processors is Pinus patula, which accounts for 52.2 per cent of the total softwood area in South Africa (DAFF, 2014). Pinus patula is also widely planted in other African and South American countries with an estimated worldwide total of one million hectares planted with this species in 1994 (Wright, 1994). A critical issue for P. patula structural lumber producers is that a large portion of their end products must conform to the minimum mechanical requirements for structural lumber. This has become more difficult in South Africa in recent years as changes to the plantation resources resulted in reduced mechanical properties of lumber (Burdzik, 2004; Dowse and Wessels, 2013). As improvements in forest management and genetic material have increased growth rates, the harvesting age of South African-grown pine trees, mainly Pinus patula, elliottii, taeda and radiata, managed for saw-log production, has been reduced considerably from ~28 years in 1983 to ~23 years in 2003 (Crickmay et al., 2005). Since then, South African studies have shown a significant reduction in important mechanical properties of lumber, particularly the mean stiffness (modulus of elasticity, MOE) of visually graded lumber (Burdzik, 2004; Wessels et al., 2011; Dowse and Wessels, 2013). Dowse and Wessels (2013) and Wessels et al. (2014) reported the mean MOE of lumber, processed from 16 to 20 year-old P. patula stands, to be ~25 per cent less than required for the lowest and most produced structural grade in South Africa. Globally, reduced mechanical properties of fast growing trees has also become a growing concern as studies from other countries, using different species, have accordingly reported significant proportions of non-compliant structural products harvested at younger ages (Cown, 1992; Kretschmann and Bendtsen, 1992; Biblis and Brinker, 1993; Biblis, 2006). In light of these reports, the South African sawmilling industry needs to address the low MOE of P. patula and other softwood resources to continue the processing thereof into acceptable structural products. The structure of wood cell walls largely determines the mechanical properties such as the MOE of wood (Barnett and Bonham, 2004; Tsoumis, 2009). Microfibril angle (MFA), the orientation of cellulose microfibrils in the secondary cell wall with respect to the longitudinal axis of tracheid cells, and wood density have been shown to be the two most influential properties for Pinus radiata (Cown et al., 1999; Evans and Ilic, 2001; Downes et al., 2002; Xu and Walker, 2004) and P. patula (Wessels et al., 2015a) wood stiffness. However, a poor relationship between MOE and wood density of P. radiata corewood has been reported in some studies (Burdon et al., 2001; Lasserre et al., 2009; Watt et al., 2010). Research by Cown et al. (1999) noted that wood density in P. radiata does become more influential with increasing cambial age. Some authors argued that MFA is the only property to account for large variations in radial MOE trends in fast grown softwoods with wood density acting only as a supporting property (Cave and Walker, 1994; Walker and Butterfield, 1996). In contrast, research relating the average MFA to the stiffness of full-sized lumber, instead of small clear specimens, has shown wood density to have a similar influence on lumber MOE (Wessels et al., 2015a), and in some cases even more so than MFA (Downes et al., 2002; Cown et al., 2004; Vikram et al., 2011). Previous studies on several softwood species showed that planting spacing might influence the mechanical properties and volume recovery. Closely spaced plantations display positive effects on wood stiffness in P. radiata (Lasserre et al., 2005, 2008, 2009; Waghorn et al., 2007a, b; Moore et al., 2015; Wessels and Froneman, 2015) and in other species (Wang and Ko, 1998; Chuang and Wang, 2001; Ishiguri et al., 2005; Roth et al., 2007; Clark III et al., 2008; Moore et al., 2009; Amateis et al., 2013; Rais et al., 2014). The increase in stiffness with increasing stems ha−1 has frequently been attributed to the increase in the height/diameter ratio (slenderness). Based on Euler’s buckling theory, tall, slender trees in competitive environments will require wood that is higher in stiffness in order to resist buckling due to their increasing self-weight (Spatz and Bruechert, 2000; Watt et al., 2006a, b; Waghorn and Watt, 2013a; Merlo et al., 2014; Wessels et al., 2015b). Planting spacing can also influence the straightness of trees (Macdonald and Hubert, 2002), which has economic consequences for log processors. Both the yield and quality of lumber is greatly affected by crooked stems (Cown et al., 1984; Monserud et al., 2003; Ivković et al., 2007; Lachenbruch et al., 2010) and some studies suggest value losses in the sawmill process of roughly 10 per cent due to poor stem straightness (Carino et al., 2006). Leaning stems and those with excessive sweep are known to cause compression wood (Timell, 1986; Krause and Plourde, 2008), which has been shown to reduce wood stiffness (Lindström et al., 2004; Sonderegger et al., 2008). The effect of planting spacing on stem straightness is, however, not always consistent. Trees in stands planted with narrow spacing have been shown to display better stem straightness, improving volume recovery (Malinauskas, 2003; Tong and Zhang, 2005; Belley et al., 2013; Froneman, 2014; Smith et al., 2014) although spacing effects on stem straightness were sometimes less clear (Egbäck et al., 2012; Liziniewicz et al., 2012). On the other hand, trees grown under suppressed conditions in closely spaced plantations may also display poor stem straightness (Theron and Bredenkamp, 2004). The main objective of this study was to determine the effect that planting spacing has on the important properties of MFA and wood density of P. patula trees. At the same time, we also wanted to establish the effect of planting spacing on stem straightness. The results of the study would be useful in formulating future forest management regimes for P. patula grown in South Africa. To the authors’ best knowledge, this is the first study measuring the effect of planting spacing of P. patula on MFA and stem straightness. Materials and methods Experimental layout This study was conducted using an 18 year-old P. patula spacing experiment located in the Mpumalanga escarpment on the Montrose plantation near the town of Barberton, South Africa (25.9037° S, 30.8729° E). This area has a mean annual rainfall of ~850 mm and mean midday temperatures of ~17°C (Barberton aviation weather station, Code: FABR, 25.7175° S, 30.9750° E, 681 m ASL). The trees in this study were pruned at 5, 7 and 9 years after planting to 2, 3.5 and 5.5 m respectively. The experiment followed a randomized complete block design consisting of four planting spacing levels of 403, 1097, 1808 and 2981 stems ha−1, each replicated in two blocks. Each sampling plot had been planted with 49 seedlings in a 7 × 7 tree layout (variable area plots) but only the centre 25 trees were included in the study reported here as the outer trees were considered buffer rows. Out of a possible 200 trees, only 171 were still available for analysis due to mortality, indicated by the survival percentages of treatments in Table 1. Table 1 Plot data and sample sizes Planting spacing (stems ha−1) Mean DBH1 (cm) Mean height1 (m) Survival (%) Total stem volume (m3 ha−1) Sample size1 403 32.7 (34.1, 34.1) 23.3 (23.9, 24.9) 96 321 48 (22, 10) 1097 23.8 (22.8, 25.3) 21.5 (20.9, 22.8) 96 428 48 (24, 10) 1808 19.9 (21.5, 21.7) 20.6 (22.2, 21.2) 84 451 42 (19, 10) 2981 16.9 (17.3, 17.8) 20.3 (20.1, 21.4) 66 417 33 (16, 10) Planting spacing (stems ha−1) Mean DBH1 (cm) Mean height1 (m) Survival (%) Total stem volume (m3 ha−1) Sample size1 403 32.7 (34.1, 34.1) 23.3 (23.9, 24.9) 96 321 48 (22, 10) 1097 23.8 (22.8, 25.3) 21.5 (20.9, 22.8) 96 428 48 (24, 10) 1808 19.9 (21.5, 21.7) 20.6 (22.2, 21.2) 84 451 42 (19, 10) 2981 16.9 (17.3, 17.8) 20.3 (20.1, 21.4) 66 417 33 (16, 10) 1The first and second values in parenthesis indicate values of the sub-sampled trees for stem straightness (TLS measurements) and the removal of increment cores respectively. Table 1 Plot data and sample sizes Planting spacing (stems ha−1) Mean DBH1 (cm) Mean height1 (m) Survival (%) Total stem volume (m3 ha−1) Sample size1 403 32.7 (34.1, 34.1) 23.3 (23.9, 24.9) 96 321 48 (22, 10) 1097 23.8 (22.8, 25.3) 21.5 (20.9, 22.8) 96 428 48 (24, 10) 1808 19.9 (21.5, 21.7) 20.6 (22.2, 21.2) 84 451 42 (19, 10) 2981 16.9 (17.3, 17.8) 20.3 (20.1, 21.4) 66 417 33 (16, 10) Planting spacing (stems ha−1) Mean DBH1 (cm) Mean height1 (m) Survival (%) Total stem volume (m3 ha−1) Sample size1 403 32.7 (34.1, 34.1) 23.3 (23.9, 24.9) 96 321 48 (22, 10) 1097 23.8 (22.8, 25.3) 21.5 (20.9, 22.8) 96 428 48 (24, 10) 1808 19.9 (21.5, 21.7) 20.6 (22.2, 21.2) 84 451 42 (19, 10) 2981 16.9 (17.3, 17.8) 20.3 (20.1, 21.4) 66 417 33 (16, 10) 1The first and second values in parenthesis indicate values of the sub-sampled trees for stem straightness (TLS measurements) and the removal of increment cores respectively. Measurements The stem deviation (from perfect straightness) was measured up to 6 m from tree base using data from a terrestrial laser scanning (TLS) system for one randomly chosen replication block. A Trimble FX phase shift scanner (Trimble Inc.) with angular resolution of 8 sec was used, which results in a sample step of 4 mm at a distance of 20 m. The scan setup used a minimum of four scans per plot. Of the 85 trees available in the chosen replication, a total of 81 were reconstructed from TLS scans and analysed for stem straightness. The other four trees were excluded from the analysis due to either limited scans, causing insufficient points in the point cloud for those trees, or forking below 6 m. The variation of these tree dimensions (manually measured on the 81 trees) from the full sample is indicated in Table 1. We defined stem straightness as the maximum deviation from the stem’s centreline perpendicular to a straight line (chord) joining the two centre-points of the stem at the base and 6 m (Figure 1). The perpendicular deviations were derived through vector equations using three-dimensional coordinates of tree stems which were provided by the 3D Forest software package, version 0.31 (www.3dforest.eu). Stem straightness was then calculated from the set of perpendicular distances for each tree by selecting the maximum. Figure 1 View largeDownload slide Illustration of the measurement of stem straightness. Figure 1 View largeDownload slide Illustration of the measurement of stem straightness. Basal area and the relative stand density (RD), according to Curtis (1982), was calculated for each spacing treatment. The diameter at breast height (DBH) and standing tree height were manually recorded for all trees (Table 1). The slenderness of trees was taken as the ratio of tree height to DBH. The dynamic MOE of standing trees was calculated from stress wave velocities at breast height obtained using the Fakopp Treesonic instrument (Fakopp Enterprise Bt.; Divos, 2010). From the wood density (ρ) – assumed constant at 1000 kg m−3 (Wielinga et al., 2009) – and the stress wave velocity (V), the MOE was then estimated from the following: MOE=ρV2 (1) The probe generally penetrated ~20 mm into the wood and thus effectively only recorded outerwood MOE and was not hindered by bark. Increment cores were taken at breast height (1.3 m) from the northern side of 10 randomly chosen trees per spacing treatment – 40 trees in total. Water in the increment cores was replaced by ethanol in three stages before the cores were dried to equilibrium moisture content. The MFA, wood density and ring widths of each sample were measured using the CSIRO Silviscan 3 apparatus (Evans, 1999) in Melbourne, Australia, at a radial resolution of 2 mm for MFA and 0.025 mm for wood density. Ring widths were defined by the distance between the maxima of wood density of successive rings in the radial wood density profile. In this study, the majority of annual rings had no latewood (LW) according to both interpretations of Mork’s definition of LW cells (Denne, 1989) – which showed that this definition could not be used in our study. A wood density threshold of 500 kg m−3 was then chosen as a definition of LW percentage. This was based on values from literature (Koubaa et al., 2002) and overlaying wood density profiles with images of increment core samples; P. patula typically has distinctly visible darker bands of LW zones. The growth of individual trees varied and therefore the width of the first year rings depended on when the height of a specific tree reached 1.3 m (which was the sampling height) – resulting in widely varying ring widths for the first year ring. Due to varying growth rates some trees only reached a height of 1.3 m after several growth seasons and therefore sometimes had fewer than 15 year rings at breast height. Cores also contained mostly earlywood for the last annual ring as trees were sampled just before winter. Therefore, in the statistical analysis only the 2nd to 13th annual rings were considered. The mean width of rings, from pith to bark, and for each spacing treatment, were also overlaid with a cant sawing pattern and a 4 mm sawing kerf to simulate which annual rings will be present in a given board position. The mean wood properties for simulated board positions were then calculated from the rings demarcated by the sawing pattern. It would have been preferable to destructively sample trees and measure the MOE of the lumber from these trees, but at the time of this study no P. patula spacing experiments were available for destructive testing and therefore the focus of this study was rather on the basic properties of MFA and wood density as well as stem straightness. Statistical analysis The R system for statistical computing (R Core Team, 2016) and Statistica (Dell Inc, 2016) were used for data analysis. Pearson correlations between the various individual tree dimensional variables and average wood properties were computed. The effect of planting spacing on DBH, tree height, stem straightness and dynamic MOE was tested using one-way analysis of variance (ANOVA) where Tukey’s LSD post hoc tests were subsequently performed. Two non-linear mixed-effects models (Pinheiro and Bates, 2000), fitted with the R package ‘nlme’ (Pinheiro et al., 2016), were developed to examine the effects of planting spacing on the pith-to-bark variation in wood density and MFA. The first model was the power function presented by Moore et al. (2015): Yijk=(αi+aij)RNijk(βi+bij)+εijk (2) where Yijk, RNijk and εijk are the response variable (MFA or wood density), the ring number, and the residual error of the kth annual ring in the jth tree in the ith spacing treatment, respectively. The parameters αi and βi correspond to the initial value (ring 1) and the radial rate of change in the response variable, respectively, which could vary for the ith spacing treatment. The aij and bij terms are the random effects for the jth tree in the ith spacing treatment. Considering that planting spacing heavily affects ring width, an additional model incorporating ring width (cf. Auty et al., 2013) was also developed to test if planting spacing still had any effect on the estimated parameters after accounting for differences in ring number and ring width: Yijk=(α0,i+aij+α1,iRWijk)RNijk(β0,i+β1,iRWijk)+εijk (3) All parameters were thus adjusted for RWijk, the ring width of the kth annual ring in the jth tree in the ith spacing treatment, by the α1,i and β1,i parameters which could also vary with the ith spacing treatment. Because only rings 2–13 were considered, the 2nd annual ring was designated as ring number 1 for this analysis. Likelihood ratio tests and Akaike’s information criterion, AIC (Akaike, 1974), were used to evaluate the significance of including each term in both models – random effects, fixed effects and the effect of spacing treatment. Heteroscedasticity was modelled as a power function of ring number (Auty et al., 2013) while the random effect parameters were considered to account for correlations among residuals (Moore et al., 2015). Subsequently, annual ring widths, modelled as an exponential function of cambial age (parameters not shown), were used to predict the radial profiles of wood properties given by equation (3), for each spacing treatment (Auty et al., 2017). Results Stem straightness Planting spacing had a highly significant effect (P < 0.001) on the average stem straightness (Figure 2A and Table 2). However, the trend across spacing treatments was inconsistent. The most widely spaced treatment (403 stems ha−1) had the least straight stems with a mean deviation of 83 mm. Stem straightness for trees from the 1097 stems ha−1 treatment was significantly (P < 0.001) greater than the 403 stems ha−1 treatment and displayed the lowest mean stem deviation of 41 mm. There were no significant differences in the mean stem straightness between the two most closely spaced treatments. Figure 2 View largeDownload slide Means and 95 per cent confidence intervals for stem straightness (A), DBH (B), height (C) and slenderness ratio (D) for each spacing treatment. Different letters denote significant differences at P < 0.05. Figure 2 View largeDownload slide Means and 95 per cent confidence intervals for stem straightness (A), DBH (B), height (C) and slenderness ratio (D) for each spacing treatment. Different letters denote significant differences at P < 0.05. Table 2 Stand-level characteristics for each spacing treatment Planting spacing (stems ha−1) Stem straightness (mm)1 Slenderness MOE (MPa)1 Basal area (m2 ha−1) Relative stand density 403 83.1 (18.8) 0.73 10 150 (620) 33.4 5.8 1097 40.7 (7.3) 0.92 12 739 (624) 48.2 9.8 1808 70.4 (10.0) 1.06 14 607 (1202) 50.0 11.1 2981 55.5 (15.0) 1.23 15 044 (893) 46.3 11.1 Planting spacing (stems ha−1) Stem straightness (mm)1 Slenderness MOE (MPa)1 Basal area (m2 ha−1) Relative stand density 403 83.1 (18.8) 0.73 10 150 (620) 33.4 5.8 1097 40.7 (7.3) 0.92 12 739 (624) 48.2 9.8 1808 70.4 (10.0) 1.06 14 607 (1202) 50.0 11.1 2981 55.5 (15.0) 1.23 15 044 (893) 46.3 11.1 1Values in parenthesis are ± SE. Table 2 Stand-level characteristics for each spacing treatment Planting spacing (stems ha−1) Stem straightness (mm)1 Slenderness MOE (MPa)1 Basal area (m2 ha−1) Relative stand density 403 83.1 (18.8) 0.73 10 150 (620) 33.4 5.8 1097 40.7 (7.3) 0.92 12 739 (624) 48.2 9.8 1808 70.4 (10.0) 1.06 14 607 (1202) 50.0 11.1 2981 55.5 (15.0) 1.23 15 044 (893) 46.3 11.1 Planting spacing (stems ha−1) Stem straightness (mm)1 Slenderness MOE (MPa)1 Basal area (m2 ha−1) Relative stand density 403 83.1 (18.8) 0.73 10 150 (620) 33.4 5.8 1097 40.7 (7.3) 0.92 12 739 (624) 48.2 9.8 1808 70.4 (10.0) 1.06 14 607 (1202) 50.0 11.1 2981 55.5 (15.0) 1.23 15 044 (893) 46.3 11.1 1Values in parenthesis are ± SE. DBH, height, slenderness and site occupancy Planting spacing had a highly significant effect (P < 0.001) on DBH as shown in Figure 2B. As expected, the mean DBH decreased with increasing stems ha−1 – in total, DBH reduced by 48 per cent from the widest to closest spacing (Table 1). Planting spacing had a highly significant effect (P < 0.001) on mean tree height – tree height decreased by 8 per cent from 403 stems ha−1 to 1097 stems ha−1. There was a further non-significant decrease in mean tree height of only 6 per cent between 1097 stems ha−1 and 2981 stems ha−1. The mean slenderness for each spacing treatment can be seen in Table 2 and Figure 2D. The effect of planting spacing on slenderness was highly significant (P < 0.001), increasing by 68 per cent from 403 to 2981 stems ha−1. Both RD and basal area followed an increasing trend from 403 to 1808 stems ha−1, above which RD remained constant while basal area then decreased (Table 2). The most notable increases were between 403 and 1097 stems ha−1. Dynamic MOE Spacing treatment had a highly significant effect (P < 0.001) on the dynamic MOE (Table 2). The mean MOE increased by 48 per cent from 403 to 2981 stems ha−1 – a mean rate of increase (Δ MOE/Δ planting spacing) of 1.9 MPa ha stems−1 (Figure 3). Differences in means were the greatest between 403 and 1097 stems ha−1 and thereafter, displaying smaller differences between the more closely spaced treatments showing an asymptotic type response. Figure 3 View largeDownload slide Means and 95 per cent confidence intervals for MOE for each spacing treatment. Different letters denote significant differences at P < 0.05. Figure 3 View largeDownload slide Means and 95 per cent confidence intervals for MOE for each spacing treatment. Different letters denote significant differences at P < 0.05. Microfibril angle The mean MFA per annual ring across all spacing treatments decreased from 31° to 7° between the 2nd and the 13th year rings (Figure 4A). As expected, MFA displayed a clear decreasing trend with increasing ring number (cambial age). For a given annual ring, the overall trend in MFA was a decreasing angle from 403 stems ha−1 to the closer spacing treatments. The mean MFA for the 403 stems ha−1 treatment decreased to about the 11th annual ring before it reached a constant level of ~12°, while in the more closely spaced treatments, MFA rapidly decreased up to the seventh and eighth annual ring before stabilizing. As a result, the mean MFA decreased by 5°, on average, for the first nine rings from 403 to 1097 stems ha−1. There was also a 5° decrease from 1097 to 1808 stems ha−1 near the pith (the second to fourth annual ring) for equivalent annual rings while the MFA values for treatments 1808 and 2981 stems ha−1 were similar at all annual rings. The model given by equation (2) (parameters not shown, AIC = 2415), showed that only αi was significantly influenced by spacing (P < 0.001). The αi term was significantly greater for 403 and 1097 stems ha−1 compared with the other treatments. Figure 4 View largeDownload slide Variation in microfibril angle (A), wood density (B), latewood percentage (C) and ring width (D) at different spacing treatments and rings from pith. Vertical bars denote 95 per cent confidence intervals. Figure 4 View largeDownload slide Variation in microfibril angle (A), wood density (B), latewood percentage (C) and ring width (D) at different spacing treatments and rings from pith. Vertical bars denote 95 per cent confidence intervals. When ring width was included, spacing treatment had a significant (P < 0.001) effect on all parameters of the model given by equation (3), which had a considerably better fit to the data (lower AIC value of 2244) (Table 3). Modelled MFA was greater for the 403 and 1097 stems ha−1 treatments compared with the other two treatments up to ring 6 (Figure 5A). The rate of decline was clearly lower for the 403 stems ha−1 treatment compared with the other treatments, although this gradient was not mediated through neither β1,i or β0,i (non-significant) but determined by ring width through the α1,i term. Modelled MFA was negatively influenced by ring width due to the significant α1,i term for the 1097 stems ha−1 treatment and the β1,i term for the 1808 stems ha−1 treatment (Table 3). The α0,i and β0,i parameters differed significantly, even after ring width had been taken into account. The α0,i parameter for the 1097 stems ha−1 was significantly greater than for the other treatments, while its β0,i parameter was the only significantly different value relative to 403 stems ha−1. Table 3 Parameter estimates, standard errors, P-values and standard deviations for the random effect estimates of equation (3) Parameters MFA Wood density Estimate SE t-Value P-value Estimate SE t-Value P-value α0 (Int./Fixed) 13.965 2.812 4.966 <0.001 431.803 13.556 31.854 <0.001    α0,1097 24.495 3.616 6.774 <0.001    α0,1808 −9.844 3.063 −3.213 0.001    α0,2981 −4.650 3.475 −1.338 0.182 α1 (Int./Fixed) 0.713 0.119 6.014 <0.001 −2.164 0.624 −3.467 <0.001   α1,1097 −0.992 0.144 −6.895 <0.001   α1,1808 0.502 0.124 4.030 <0.001   α1,2981 0.541 0.162 3.330 <0.001 β0 (Int.) −0.086 0.068 −1.270 0.205 0.023 0.015 1.526 0.128   β0,1097 −0.715 0.073 −9.807 <0.001 0.077 0.012 6.357 <0.001   β0,1808 0.145 0.083 1.739 0.083 0.115 0.013 9.003 <0.001   β0,2981 −0.060 0.100 −0.603 0.547 0.115 0.013 9.099 <0.001 β1 (Int.) 0.001 0.003 0.316 0.753 0.001 0.001 0.446 0.656   β1,1097 0.030 0.004 8.021 <0.001 −0.006 0.002 −3.031 0.003   β1,1808 −0.014 0.004 −3.199 0.002 −0.010 0.002 −4.429 <0.001   β1,2981 −0.001 0.006 −0.069 0.945 −0.012 0.002 −5.433 <0.001 Random parameters Std. Dev. Std. Dev. a0,i - tree 3.098 37.584 εij - residual 0.661 24.486 Parameters MFA Wood density Estimate SE t-Value P-value Estimate SE t-Value P-value α0 (Int./Fixed) 13.965 2.812 4.966 <0.001 431.803 13.556 31.854 <0.001    α0,1097 24.495 3.616 6.774 <0.001    α0,1808 −9.844 3.063 −3.213 0.001    α0,2981 −4.650 3.475 −1.338 0.182 α1 (Int./Fixed) 0.713 0.119 6.014 <0.001 −2.164 0.624 −3.467 <0.001   α1,1097 −0.992 0.144 −6.895 <0.001   α1,1808 0.502 0.124 4.030 <0.001   α1,2981 0.541 0.162 3.330 <0.001 β0 (Int.) −0.086 0.068 −1.270 0.205 0.023 0.015 1.526 0.128   β0,1097 −0.715 0.073 −9.807 <0.001 0.077 0.012 6.357 <0.001   β0,1808 0.145 0.083 1.739 0.083 0.115 0.013 9.003 <0.001   β0,2981 −0.060 0.100 −0.603 0.547 0.115 0.013 9.099 <0.001 β1 (Int.) 0.001 0.003 0.316 0.753 0.001 0.001 0.446 0.656   β1,1097 0.030 0.004 8.021 <0.001 −0.006 0.002 −3.031 0.003   β1,1808 −0.014 0.004 −3.199 0.002 −0.010 0.002 −4.429 <0.001   β1,2981 −0.001 0.006 −0.069 0.945 −0.012 0.002 −5.433 <0.001 Random parameters Std. Dev. Std. Dev. a0,i - tree 3.098 37.584 εij - residual 0.661 24.486 Estimates for the fixed parameters show their values and intercept (Int., i.e. 403 stems ha–1), and values for the other treatments relative to the intercept (i.e. the change in estimate from 403 stems ha−1). The α0 and α1 parameters for wood density are the only fixed parameters (single value for all treatments). Table 3 Parameter estimates, standard errors, P-values and standard deviations for the random effect estimates of equation (3) Parameters MFA Wood density Estimate SE t-Value P-value Estimate SE t-Value P-value α0 (Int./Fixed) 13.965 2.812 4.966 <0.001 431.803 13.556 31.854 <0.001    α0,1097 24.495 3.616 6.774 <0.001    α0,1808 −9.844 3.063 −3.213 0.001    α0,2981 −4.650 3.475 −1.338 0.182 α1 (Int./Fixed) 0.713 0.119 6.014 <0.001 −2.164 0.624 −3.467 <0.001   α1,1097 −0.992 0.144 −6.895 <0.001   α1,1808 0.502 0.124 4.030 <0.001   α1,2981 0.541 0.162 3.330 <0.001 β0 (Int.) −0.086 0.068 −1.270 0.205 0.023 0.015 1.526 0.128   β0,1097 −0.715 0.073 −9.807 <0.001 0.077 0.012 6.357 <0.001   β0,1808 0.145 0.083 1.739 0.083 0.115 0.013 9.003 <0.001   β0,2981 −0.060 0.100 −0.603 0.547 0.115 0.013 9.099 <0.001 β1 (Int.) 0.001 0.003 0.316 0.753 0.001 0.001 0.446 0.656   β1,1097 0.030 0.004 8.021 <0.001 −0.006 0.002 −3.031 0.003   β1,1808 −0.014 0.004 −3.199 0.002 −0.010 0.002 −4.429 <0.001   β1,2981 −0.001 0.006 −0.069 0.945 −0.012 0.002 −5.433 <0.001 Random parameters Std. Dev. Std. Dev. a0,i - tree 3.098 37.584 εij - residual 0.661 24.486 Parameters MFA Wood density Estimate SE t-Value P-value Estimate SE t-Value P-value α0 (Int./Fixed) 13.965 2.812 4.966 <0.001 431.803 13.556 31.854 <0.001    α0,1097 24.495 3.616 6.774 <0.001    α0,1808 −9.844 3.063 −3.213 0.001    α0,2981 −4.650 3.475 −1.338 0.182 α1 (Int./Fixed) 0.713 0.119 6.014 <0.001 −2.164 0.624 −3.467 <0.001   α1,1097 −0.992 0.144 −6.895 <0.001   α1,1808 0.502 0.124 4.030 <0.001   α1,2981 0.541 0.162 3.330 <0.001 β0 (Int.) −0.086 0.068 −1.270 0.205 0.023 0.015 1.526 0.128   β0,1097 −0.715 0.073 −9.807 <0.001 0.077 0.012 6.357 <0.001   β0,1808 0.145 0.083 1.739 0.083 0.115 0.013 9.003 <0.001   β0,2981 −0.060 0.100 −0.603 0.547 0.115 0.013 9.099 <0.001 β1 (Int.) 0.001 0.003 0.316 0.753 0.001 0.001 0.446 0.656   β1,1097 0.030 0.004 8.021 <0.001 −0.006 0.002 −3.031 0.003   β1,1808 −0.014 0.004 −3.199 0.002 −0.010 0.002 −4.429 <0.001   β1,2981 −0.001 0.006 −0.069 0.945 −0.012 0.002 −5.433 <0.001 Random parameters Std. Dev. Std. Dev. a0,i - tree 3.098 37.584 εij - residual 0.661 24.486 Estimates for the fixed parameters show their values and intercept (Int., i.e. 403 stems ha–1), and values for the other treatments relative to the intercept (i.e. the change in estimate from 403 stems ha−1). The α0 and α1 parameters for wood density are the only fixed parameters (single value for all treatments). Figure 5 View largeDownload slide Variation in predicted microfibril angle (A) and wood density (B) from equation (3) at different spacing treatments and rings from pith. Figure 5 View largeDownload slide Variation in predicted microfibril angle (A) and wood density (B) from equation (3) at different spacing treatments and rings from pith. Wood density Wood density varied from ~370 kg m−3 close to the pith to ~600 kg m−3 at ring 10 (Figure 4B). The general trend was an increase in wood density with increasing cambial age up until the 10th annual ring after which it then began to decline. No gradient change was observed in ring width after the 10th annual ring but a similar observation was displayed in the latewood percentage (Figure 4C). There were no clear differences in the wood density between spacing treatments within the first six annual rings, after which the general trend for equivalent annual rings was an increase in wood density with increasing stems ha−1. These differences were most pronounced between 403 stems ha−1 and the other treatments. The model parameters (equation (2)) (AIC = 4941) were significantly different between wide and closer spacing treatments, both αi and βi, showing that the variation of wood density was affected by planting spacing. The model for wood density given by equation (3) (AIC = 4867) (Table 3, Figure 5B) showed that spacing treatment did not significantly influence the initial density of wood ( α0,i). Ring width, however, did not emerge as being a significant contributor to differences in initial wood density between different spacing treatments ( α1,i). Both components of the radial rate of change in wood density ( β0,i and β1,i) was significantly influenced by spacing treatment. This was evident as ring width differed between treatments especially near the pith, while wood density displayed no clear differences in the same region. The rate parameter β0,i , for a given ring width, increased significantly with increased stems ha−1 while ring width also had an increasingly negative influence on modelled wood density in closer spacings ( β1,i, Table 3). Accordingly, the predicted wood density between treatments were similar near the pith, but the incline rate clearly increased from 403 stems ha−1 to 1808 and 2981 stems ha−1 (Figure 5B). The modelled wood density curves for 1808 and 2981 stems ha−1 were nearly completely overlapping (Figure 5B). Relationship between properties Correlations between measured properties were reported in Table 4. Slenderness displayed insignificant correlations with wood density, LW percentage and tree height (despite being a function of tree diameter and tree height) but was strongly related to DBH (r = 0.86). Although slenderness also correlated significantly to MOE, a weak relationship was still displayed (r = 0.42 or r2 = 0.18). MFA was the property with the highest Pearson correlation with MOE (r = −0.66) which weakened somewhat when only considering the outer 20 mm of increment cores. Wood density and LW percentage both correlated significantly with ring width, with an especially high correlation coefficient between wood density and LW percentage (r = 0.89). Table 4 Pearson correlation coefficients between the mean tree variables and wood properties measured or calculated from all 40 increment cores Variable DBH Height Slenderness Ring width Density LW % MFA MOE1 DBH 1.00 0.65 −0.86 0.80 −0.26 −0.19 0.49 −0.51 Height 1.00 −0.24 0.58 0.01 0.05 0.30 −0.19 Slenderness 1.00 −0.63 0.28 0.23 −0.39 0.42 Ring width 1.00 −0.43 −0.36 0.48 −0.42 Density 1.00 0.95 −0.11 0.20 (0.24) LW % 1.00 −0.03 0.05 MFA 1.00 −0.66 (−0.43) MOE 1.00 Variable DBH Height Slenderness Ring width Density LW % MFA MOE1 DBH 1.00 0.65 −0.86 0.80 −0.26 −0.19 0.49 −0.51 Height 1.00 −0.24 0.58 0.01 0.05 0.30 −0.19 Slenderness 1.00 −0.63 0.28 0.23 −0.39 0.42 Ring width 1.00 −0.43 −0.36 0.48 −0.42 Density 1.00 0.95 −0.11 0.20 (0.24) LW % 1.00 −0.03 0.05 MFA 1.00 −0.66 (−0.43) MOE 1.00 1Values in parentheses indicate correlations between MOE and the mean wood properties for the last 20 mm of increment cores; bold correlations are significant at P < 0.05. Table 4 Pearson correlation coefficients between the mean tree variables and wood properties measured or calculated from all 40 increment cores Variable DBH Height Slenderness Ring width Density LW % MFA MOE1 DBH 1.00 0.65 −0.86 0.80 −0.26 −0.19 0.49 −0.51 Height 1.00 −0.24 0.58 0.01 0.05 0.30 −0.19 Slenderness 1.00 −0.63 0.28 0.23 −0.39 0.42 Ring width 1.00 −0.43 −0.36 0.48 −0.42 Density 1.00 0.95 −0.11 0.20 (0.24) LW % 1.00 −0.03 0.05 MFA 1.00 −0.66 (−0.43) MOE 1.00 Variable DBH Height Slenderness Ring width Density LW % MFA MOE1 DBH 1.00 0.65 −0.86 0.80 −0.26 −0.19 0.49 −0.51 Height 1.00 −0.24 0.58 0.01 0.05 0.30 −0.19 Slenderness 1.00 −0.63 0.28 0.23 −0.39 0.42 Ring width 1.00 −0.43 −0.36 0.48 −0.42 Density 1.00 0.95 −0.11 0.20 (0.24) LW % 1.00 −0.03 0.05 MFA 1.00 −0.66 (−0.43) MOE 1.00 1Values in parentheses indicate correlations between MOE and the mean wood properties for the last 20 mm of increment cores; bold correlations are significant at P < 0.05. Simulated lumber properties The pith boards of the 2981 stems ha−1 treatment had a mean MFA of 19.3° compared with 30.3° for 403 stems ha−1 – a 57 per cent increase (Figure 6). Wood density showed a similar trend of increasing with board position and stems ha−1 except for the wood density of the pith boards for 403 stems ha−1, which was quite high due to the high wood density values of the first three rings from the pith for this treatment. The 403 stems ha−1 treatment had trees with greater diameters and displayed the capacity to produce additional boards, but in terms of MFA and wood density, its best board (second from the pith), was still worse than the first boards next to the pith of 2981 stems ha−1. Figure 6 View largeDownload slide The mean ring widths for 403 stems ha−1 (A), 1097 stems ha−1 (B), 1808 stems ha−1 (C) and 2981 stems ha−1 (D), overlaid with a 40 × 120 mm cant sawing strategy drawn to scale including a saw kerf of 4 mm. Maximum and minimum rings are indicated for each board position with their mean MFA and wood density. Figure 6 View largeDownload slide The mean ring widths for 403 stems ha−1 (A), 1097 stems ha−1 (B), 1808 stems ha−1 (C) and 2981 stems ha−1 (D), overlaid with a 40 × 120 mm cant sawing strategy drawn to scale including a saw kerf of 4 mm. Maximum and minimum rings are indicated for each board position with their mean MFA and wood density. Discussion It was expected that the closely spaced treatments would result in straighter trees as some studies have found that decreased planting spacing and competition for sunlight and other growth resources from evenly planted neighbouring trees can help direct growth straight upwards (Woods et al., 1992; Smith et al., 2014). This was partly supported by the finding in our study that the 403 stems ha−1 treatment had the least straight stems. However, an inconsistent trend points to some opposing effects. Part of the reason for this might be that certain soil nutrient deficiencies has been shown to result in stem deformity (Birk, 1991; Turvey et al., 1992). Severe competition between densely planted trees might therefore result in nutrient deficiencies for individual trees. The high RD of both 1808 and 2981 stems ha−1 is close to the defined level of imminent mortality of 12 for South African pines (Kotze and du Toit, 2012) indicating severe competition which could possibly have led to nutrient deficiencies for some trees. We hypothesize that another reason for decreases in stem straightness at closer spacing might be the high mortality rate (Table 1), which can cause uneven openings in the canopy. Poor stem straightness might result from the tree positioning itself towards openings in the canopy. The low survival rates in the two most closely spaced treatments (Table 1) supports this hypothesis. Another factor that should be considered in future studies, but which did not play a role in this study, is the influence of thinning. Treatments with higher stand densities did not always result in straighter stems, but when thinning is performed, they provide the opportunity to remove more trees with poor stem straightness earlier in the life of the trees – thereby removing competition for better shaped trees to grow faster (Macdonald and Hubert, 2002). Planting spacing should thus be considered in combination with possible thinning regimes when evaluating tree form at final harvest. Stem straightness has a big influence on sawmill volume recovery and processing efficiency (Carino et al., 2006; Hamner et al., 2007; Yerbury and Cooper, 2017) and future work should focus on a better understanding of the influence of growing space, manipulated through both planting spacing and thinning, on stem straightness of P. patula. In terms of wood properties, the overall effect of growing space was much clearer. The strong positive influence of planting spacing on the dynamic MOE (Figure 3) was higher than that found in studies on any other softwood species. In studies by Waghorn et al. (2007a; b) and Lasserre et al. (2005) on P. radiata, Roth et al. (2007) on Pinus taeda, and Froneman (2014) on Pinus elliottii, the mean dynamic MOE gradients varied between 0.8 and 1.2 MPa ha stem−1 increases in stems ha−1, which was roughly 50 per cent lower than found in this study (1.8 MPa ha stem−1). However, the range of planting spacing in our study was greater than that of the other studies with the exception of Froneman (2014). At the individual tree level, the poor relationship between MOE and slenderness was contrary to the results of various other studies that found a comparatively good linear relationship. Results by Watt et al. (2006a, 2009), Waghorn et al. (2007b) and Lasserre et al. (2009) on P. radiata found slenderness to explain 49–71 per cent of the variation in tree dynamic MOE. Interestingly, Lasserre et al. (2008) found stem slenderness to seemingly account for variation in tree MOE as a previously significant effect of planting spacing became non-significant once adjustments were made for differences in slenderness (adding slenderness as a covariate), suggesting tree slenderness to be the main mechanism through which closely spaced trees improve wood stiffness. Contrastingly, Roth et al. (2007) found the effect of planting spacing on the MOE of young P. taeda to remain significant after adjustments for stem slenderness, with the authors suggesting environmental and genetic factors to control the outerwood dynamic MOE through mechanisms other than stem form. Similarly, it seems as if the increased MOE with increasing stems ha−1 from this study was not only mediated through tree slenderness. MFA differences observed in our study were similar to results of previous studies displaying reduced MFA with closer spacing at either establishment (Lasserre et al., 2009; Watt et al., 2011) or after thinnings (Moore et al., 2015; Auty et al., 2017), although our study covered a wider range of tree spacing. In general, our study found a positive influence of growth rate on MFA, similar to various other studies (Lindström et al., 1998; Sarén et al., 2004; Auty et al., 2013). Furthermore, a residual effect of planting spacing on the variation of wood properties over and above cambial age and ring width was also evident. This highlights the limitations associated with the common use of ring width (growth rate) as a proxy for spacing since it may not fully capture cause and effect (Zobel, 1992; Auty et al., 2017). It remains unclear why a tree would exhibit lower MFA when planted more densely and growth (or ring width) is reduced, if not due mainly to slenderness and growth rate differences. Other explanations for the apparent differences in wood properties with tree spacing may include the ratio of live crown length to tree height (Kuprevicius et al., 2013). The lower MFA near the pith with increasing stems ha−1 in this study (Figures 4A and 5A), is an important finding. This inner region of the stem forms part of the boards which usually display the worst MOE in logs of P. patula (Wessels et al., 2014), and other species (Xu and Walker, 2004; Vikram et al., 2011; Moore et al., 2012, 2013; Rais et al., 2014; Wessels and Froneman, 2015). The generally high MFA values near the pith is part of the reason for the lower stiffness of pith boards, as saplings with small diameters require more flexibility to prevent fracture of the stem when subjected to wind loading (Barnett and Bonham, 2004; Burgert, 2006). This is one theory that could also possibly explain why MFA generally improved with increasing stems ha−1. Closely planted trees should result in less wind exposure (Green et al., 1995) which has been shown to decrease taper while increasing wood stiffness (Gardiner et al., 1997; Spatz and Bruechert, 2000; Bascuñán et al., 2006; Brüchert and Gardiner, 2006). The decrease of 5° in MFA, on average, from 403 to 1097 stems ha−1 and from 1097 to 1808 stems ha−1 (only for some rings) could potentially be of significant value to structural lumber producers. For P. radiata, a 5° improvement in corewood has been suggested to be enough to represent an increase in wood stiffness of up to 50 per cent (Walker and Butterfield, 1996). In a study by Wessels et al. (2015a) it was shown that the mean wood density, MFA and ring width, calculated from year rings could be used to successfully predict the stiffness of boards. Ring width is an important property since it affects the geometry of sawing and subsequently the individual lumber properties. Therefore, although the improved MFA from 1097 to 1808 stems ha−1 was restricted to only the first few rings, these growth rings will occupy a significant proportion of volume within pith boards as the growth rate is typically greatest at the pith (Figure 6). Furthermore, in trees harvested at a young age, pith boards constitute a large percentage of the total recovered product. Rings closer to the bark, with better mechanical properties, are unfortunately less prevalent as they are mostly chipped away in the sawmilling process. It must be noted that for some increment cores pith eccentricity and compression wood may have had a negative influence on the accuracy of some ring width and wood property measurements respectively. It was noted that the earlier the apparent onset of competition commenced in a stand of trees, the sooner the MFA began to stabilize. The point where maturewood begins is often defined as the radial position where wood properties have stabilized – usually after a transition zone (Cown, 1992; Zobel and Van Buijtenen, 2012). When only considering MFA, it seems as if this transition then occurs earlier for plantations with increased stems ha−1. On the other hand, it was also apparent that the radial MFA and wood density gradients decreased with decreasing stems ha−1 (Figure 4A and B), which has also previously been reported (Malan et al., 1997; Moore et al., 2015). In terms of wood quality, a low MFA gradient is usually considered a positive trait that will guard against uneven shrinkage which leads to warping of lumber products (Huang et al., 2003; Malan, 2010). An interesting result was the decline in wood density after the 10th annual ring (Figure 4B). Distinct shifts in wood density profiles has previously been reported (Moore et al., 2015). However, this decline was consistent for all the spacing treatments and was only mirrored in LW percentage, suggesting that it was probably a function of growing conditions not related to spacing (Cregg et al., 1988; Filipescu et al., 2014). Conclusions In summary, the narrow spacing treatments (1808 and 2981 stems ha−1) gave three distinct advantages in terms of wood properties compared with the wide spacing treatments (403 stems ha−1): First, the absolute mean MFA values of the more closely spaced treatments were significantly lower for rings close to the pith. Second, based on MFA, the juvenile core seems to be restricted to the first seven or eight year rings from the pith, whereas for the 403 stems ha−1 treatment, the juvenile core transition only started at rings 10 or 11. Third, due to suppressed growth, the centre boards of narrow spacings will contain more mature rings than that of the 403 stems ha−1. Combined, this resulted in improved MFA properties at similar board positions for closer spacings which, excluding the pith boards, was also the case for wood density. Comprehensive sawing and lumber testing studies will be required to evaluate the effect of planting spacing on final product properties. After accounting for differences due to ring number and ring age, spacing treatment had a significant effect on both the initial MFA and its rate of change with age. For wood density, this remaining effect of spacing treatment was only displayed in its radial rate of change. The results of this study showed that increased stems ha−1 has the potential to improve the underlying wood properties controlling lumber stiffness. It might be possible to reap the benefits of closely spaced plantations to control wood properties during juvenile growth and then thin a stand to inhibit mortality and potentially improve the average stem form. The sample size in this study was limited and so additional studies are required to reinforce these results. Future work should also include destructive sampling of trees and processing into lumber to evaluate the effect of planting spacing on the actual final product. Stem straightness at the final harvest could possibly also be improved using narrow spacing and thinning but results were not conclusive. More work is required to understand the effect of planting spacing on stem straightness as well as the possible effect of thinning. Acknowledgements We appreciate the help of Wilmour Hendrikse, Hannes Vosloo, Phillip Fischer, Prof. Martin Kidd and Dr Robert Evans with field work, laboratory measurements, data preparation and statistical analysis. Thanks to Prof. Alexis Achim and two anonymous reviewers for very good suggestions to improve earlier versions of the manuscript. Funding The following organizations are gratefully acknowledged for funding this project: Sawmilling SA, Forestry SA, Sappi, York Timbers, Hans Merensky Foundation, and the SA government’s THRIP programme. Conflict of interest statement None declared. References Akaike , H. 1974 A new look at the statistical model identification . IEEE Trans. Automat. Contr. 19 , 716 – 723 . Google Scholar CrossRef Search ADS Amateis , R.L. , Burkhart , H.E. and Jeong , G.Y. 2013 Modulus of elasticity declines with decreasing planting density for loblolly pine (Pinus taeda) plantations . Ann. For. Sci. 70 , 743 – 750 . Google Scholar CrossRef Search ADS Auty , D. , Gardiner , B.A. , Achim , A. , Moore , J.R. and Cameron , A.D. 2013 Models for predicting microfibril angle variation in Scots pine . Ann. For. Sci. 70 , 209 – 218 . Google Scholar CrossRef Search ADS Auty , D. , Moore , J. , Achim , A. , Lyon , A. , Mochan , S. and Gardiner , B. 2017 Effects of early respacing on the density and micro fibril angle of Sitka spruce wood . Forestry , 1 – 13 . https://doi.org/10.1093/forestry/cpx004 Barnett , J.R. and Bonham , V.A. 2004 Cellulose microfibril angle in the cell wall of wood fibres . Biol. Rev. Camb. Philos. Soc. 79 , 461 – 472 . Google Scholar CrossRef Search ADS PubMed Bascuñán , A. , Moore , J.R. and Walker , J.C.F. 2006 Variations in the dynamic modulus of elasticity with proximity to the stand edge in radiata pine stands on the Canterbury Plains, New Zealand . N. Z. J. For. 51 , 4 – 8 . Belley , D. , Beaudoin , M. , Duchesne , I. , Vallerand , S. , Tong , Q.J. and Swift , D.E. 2013 Assessment of white spruce and jack pine stem curvature from a nelder spacing experiment . Wood fiber Sci. 45 , 237 – 249 . Biblis , E. 2006 Flexural properties and compliance to visual grade requirements of 2 by 4 and 2 by 6 loblolly pine lumber obtained from a 19-year-old plantation . For. Prod. J. 56 , 71 – 73 . Biblis , E. and Brinker , R. 1993 Effect of stand age on flexural properties and grade compliance of lumber from loblolly pine plantation timber . For. Prod. J. 43 , 23 . Birk , E.M. 1991 Stem and branch form of 20-year-old radiata pine in relation to previous land use . Aust. For. 54 , 30 – 39 . Google Scholar CrossRef Search ADS Brüchert , F. and Gardiner , B. 2006 The effect of wind exposure on the tree aerial architecture and biomechanics of Sitka spruce (Picea sitchensis, Pinaceae) . Am. J. Bot. 93 , 1512 – 1521 . Google Scholar CrossRef Search ADS PubMed Burdon , R.D. , Britton , R.A.J. and Walford , G.B. 2001 Wood stiffness and bending strength in relation to density in four native provenances of Pinus radiata . N. Z. J. For. Sci. 31 , 130 – 146 . Burdzik , W.M.G. 2004 Grade verification of SA pine—bending, modulus of rupture, modulus of elasticity, tension and compression . S. Afr. For. J. 202 , 21 – 27 . Burgert , I. 2006 Exploring the micromechanical design of plant cell walls . Am. J. Bot. 93 , 1391 – 1401 . Google Scholar CrossRef Search ADS PubMed Carino , H.F. , Arano , K.G. , Blanche , C.A. and Aries Patawaran , M. 2006 Impact of curve sawing on southern pine dimension lumber manufacturing. Part 1. Lumber volume and value yields . For. Prod. J. 56 , 61 – 68 . Cave , I. and Walker , J. 1994 Stiffness of wood in fast-grown plantation softwoods: the influence of microfibril angle . For. Prod. J. 44 , 43 . Chuang , S. and Wang , S. 2001 Evaluation of standing tree quality of Japanese cedar grown with different spacing using stress-wave and ultrasonic-wave methods . J. wood Sci. 47 , 245 – 253 . Google Scholar CrossRef Search ADS Clark , A. , III , Jordan , L. , Schimleck , L. and Daniels , R.F. 2008 Effect of initial planting spacing on wood properties of unthinned loblolly pine at age 21 . For. Prod. J. 58 , 78 – 83 . Cown , D.J. 1992 Corewood (juvenile wood) in Pinus radiata – should we be concerned? N. Z. J. For. Sci. 22 , 87 – 95 . Cown , D.J. , Ball , R.D. and Riddell , M.J.C. 2004 10 Pinus radiata clones: distribution and influence on product performance . N. Z. J. For. Sci. 34 , 293 – 315 . Cown , D.J. , Hebert , J. and Ball , R. 1999 Modelling Radiata pine lumber characteristics. Part 1: mechanical properties of small clears . N. Z. J. For. Sci. 29 , 203 – 213 . Cown , D.J. , McConchie , D.L. and Treloar , C. 1984 Timber recovery from pruned Pinus radiata butt logs at Mangatu: effect of log sweep . N. Z. J. For. Sci. 14 , 109 – 123 . Cregg , B.M. , Dougherty , P.M. and Hennessey , T.C. 1988 Growth and wood quality of young loblolly pine trees in relation to stand density and climatic factors . Can. J. For. Res. 18 , 851 – 858 . Google Scholar CrossRef Search ADS Crickmay , D.G. , Le Brasseur , J. , Stubbings , J.A. and Daugherty , A.E. 2005 Study of the supply and demand of industrial roundwood in South Africa. Copy obtainable from mandy@crickmay.co.za. Crickmay and Associates 2015 South African Lumber Index for February 2015. Curtis , R.O. 1982 A simple index of stand density for Douglas-fir . For. Sci. 28 , 92 – 94 . DAFF 2014 Report on commercial timber resources and primary roundwood processing in South Africa. Dell Inc 2016 STATISTICA (data analysis software system). Denne , M. 1989 Definition of latewood according to Mork (1928) . IAWA J 10 , 59 – 62 . Google Scholar CrossRef Search ADS Divos , F. 2010 Acoustic tools for seedling, tree and log selection. In: The Future of Quality Control for Wood & Wood Products, 4–7th May 2010, Edinburgh The Final Conference of COST Action E53. Downes , G.M. , Nyakuengama , J.G. , Evans , R. , Northway , R. , Blakemore , P. , Dickson , R.L. , et al. 2002 Relationship between wood density, microfibril angle and stiffness in thinned and fertilized Pinus radiata . IAWA J 23 , 253 – 265 . Google Scholar CrossRef Search ADS Dowse , G.P. and Wessels , C.B. 2013 The structural grading of young South African grown Pinus patula sawn timber . South. For. J. For. Sci. 75 , 37 – 41 . Egbäck , S. , Liziniewicz , M. , Högberg , K. , Ekö , P. and Nilsson , U. 2012 Influence of progeny and initial stand density on growth and quality traits of 21 year old half-sib Scots pine (Pinus sylvestris L.) . For. Ecol. Manage. 286 , 1 – 7 . Google Scholar CrossRef Search ADS Evans , R. 1999 A variance approach to the X-ray diffractometric estimation of microfibril angle in wood . Appita J. 52 , 283 – 289 . Evans , R. and Ilic , J. 2001 Rapid prediction of wood stiffness from microfibril angle and density . For. Prod. J. 51 , 53 – 57 . Filipescu , C.N. , Lowell , E.C. , Koppenaal , R. and Mitchell , A.K. 2014 Modeling regional and climatic variation of wood density and ring width in intensively managed Douglas-fir . Can. J. For. Res. 44 , 220 – 229 . Google Scholar CrossRef Search ADS Froneman , G.M. 2014 The effect of initial planting density on selected physical and mechanical properties of South African grown Pinus elliottii and Pinus radiata. University of Stellenbosch. Gardiner , B. a , Stacey , G.R. , Belcher , R.E. and Wood , C.J. 1997 Field and wind tunnel assessments of the implications of respacing and thinning for tree stability . Forestry 70 , 233 – 252 . Google Scholar CrossRef Search ADS Green , S.R. , Grace , J. and Hutchings , N.J. 1995 Observations of turbulent air flow in three stands of widely spaced Sitka spruce . Agric. For. Meteorol. 1923 , 205 – 225 . Google Scholar CrossRef Search ADS Hamner , P. , White , M. and Araman , P. 2007 The frequency and level of sweep in mixed hardwood saw logs in the eastern United States . For. Prod. J. 57 , 23 . Huang , C.-L. , Lindström , H. , Nakada , R. and Ralston , J. 2003 Cell wall structure and wood properties determined by acoustics – a selective review . Holz als Roh- und Werkst 61 , 321 – 335 . Google Scholar CrossRef Search ADS Ishiguri , F. , Kasai , S. , Yokota , S. , Iizuka , K. and Yoshizawa , N. 2005 Wood quality of sugi (Cryptomeria japonica) grown at four initial spacings . IAWA J. 26 , 375 – 386 . Google Scholar CrossRef Search ADS Ivković , M. , Wu , H.X. , Spencer , D.J. , Mcrae , T.A. , Wu , H.X. , Spencer , D.J. , et al. 2007 Modelling the effects of stem sweep, branch size and wood stiffness of radiata pine on structural timber production . Aust. For. 70 , 173 – 184 . Google Scholar CrossRef Search ADS Kotze , H. and du Toit , B. 2012 Silviculture of industrial pine plantations ins South Africa. In South African Forestry Handbook . 5th edn . Bredenkamp B.V. and Upfold S. (eds). The Southern African Institute of Forestry , pp. 123 – 140 . Koubaa , A. , Zhang , S.Y.T. and Makni , S. 2002 Defining the transition from earlywood to latewood in black spruce based on intra-ring wood density profiles from X-ray densitometry . Ann. For. Sci. 59 , 511 – 518 . Google Scholar CrossRef Search ADS Krause , C. and Plourde , P. 2008 Stem deformation in young plantations of black spruce (Picea mariana (Mill.) B.S.P.) and jack pine (Pinus banksiana Lamb.) in the boreal forest of Quebec, Canada . For. Ecol. Manage. 255 , 2213 – 2224 . Google Scholar CrossRef Search ADS Kretschmann , D. and Bendtsen , B. 1992 Ultimate tensile stress and modulus of elasticity of fast-grown plantation loblolly pine lumber . Wood Fiber Sci. 24 , 189 – 203 . Kuprevicius , A. , Auty , D. , Achim , A. and Caspersen , J.P. 2013 Quantifying the influence of live crown ratio on the mechanical properties of clear wood . Forestry 86 , 361 – 369 . Google Scholar CrossRef Search ADS Lachenbruch , B. , Droppelmann , F. , Balocchi , C. , Peredo , M. and Perez , E. 2010 Stem form and compression wood formation in young Pinus radiata trees . Can. J. For. Res. 40 , 26 – 36 . Google Scholar CrossRef Search ADS Lasserre , J. , Mason , E. and Watt , M. 2005 The effects of genotype and spacing on Pinus radiata corewood stiffness in an 11-year old experiment . For. Ecol. Manage. 205 , 375 – 383 . Google Scholar CrossRef Search ADS Lasserre , J.P. , Mason , E.G. and Watt , M.S. 2008 Influence of the main and interactive effects of site, stand stocking and clone on Pinus radiata D. Don corewood modulus of elasticity . For. Ecol. Manage. 255 , 3455 – 3459 . Google Scholar CrossRef Search ADS Lasserre , J.P. , Mason , E.G. , Watt , M.S. and Moore , J.R. 2009 Influence of initial planting spacing and genotype on microfibril angle, wood density, fibre properties and modulus of elasticity in Pinus radiata D. Don corewood . For. Ecol. Manage. 258 , 1924 – 1931 . Google Scholar CrossRef Search ADS Lindström , B.H. , Evans , J.W. and Verrill , S.P. 1998 Influence of cambial age and growth conditions on microfibril angle in young Norway Spruce (Picea abies [L.] Karst.) . Holzforschung 52 , 573 – 581 . Google Scholar CrossRef Search ADS Lindström , H. , Harris , P. , Sorensson , C.T. and Evans , R. 2004 Stiffness and wood variation of 3-year old Pinus radiata clones . Wood Sci. Technol. 38 , 579 – 597 . Google Scholar CrossRef Search ADS Liziniewicz , M. , Ekö , P.M. and Agestam , E. 2012 Effect of spacing on 23-year-old lodgepole pine (Pinus contorta Dougl. var. latifolia) in southern Sweden . Scand. J. For. Res. 27 , 361 – 371 . Google Scholar CrossRef Search ADS Macdonald , E. and Hubert , J. 2002 A review of the effects of silviculture on timber quality of Sitka spruce . Forestry 75 , 107 – 138 . Google Scholar CrossRef Search ADS Malan , F.S. 2010 Corewood in South African pine: necessity and opportunities for improvement . South. For. J. For. Sci. 72 , 99 – 105 . Malan , F.S. , Retief , R.J. and Male , J.R. 1997 The influence of planting espacement on the wood density and pulping properties of Pinus patula . S. Afr. For. J. 180 , 23 – 32 . Malinauskas , A. 2003 Influence of initial density and planning design on the quality of Butt Logs in Scots Pine (Pinus sylvestris L.) plantations . Balt. For. 9 , 10 – 15 . Merlo , E. , Alvarez-Gonzalez , J.G. , Santaclara , O. and Riesco , G. 2014 Modelling modulus of elasticity of Pinus pinaster Ait. in northwestern Spain with standing tree acoustic measurements, tree, stand and site variables . For. Syst. 23 , 153 – 166 . Monserud , R.A. , Parry , D.L. and Todoroki , C.L. 2003 Recovery from simulated sawn logs with sweep recovery from simulated sawn logs with sweep . N. Z. J. For. Sci. 34 , 190 – 205 . Moore , J. , Achim , A. , Lyon , A. , Mochan , S. and Gardiner , B. 2009 Effects of early re-spacing on the physical and mechanical properties of Sitka spruce structural timber . For. Ecol. Manage. 258 , 1174 – 1180 . Google Scholar CrossRef Search ADS Moore , J.R. , Cown , D.J. , McKinley , R.B. and Sabatia , C.O. 2015 Effects of stand density and seedlot on three wood properties of young radiata pine grown at a dry-land site in New Zealand . N. Z. J. For. Sci 45 , 15 . Google Scholar CrossRef Search ADS Moore , J.R. , Lyon , A.J. and Lehneke , S. 2012 Effects of rotation length on the grade recovery and wood properties of Sitka spruce structural timber grown in Great Britain . Ann. For. Sci. 69 , 353 – 362 . Google Scholar CrossRef Search ADS Moore , J.R. , Lyon , A.J. , Searles , G.J. , Lehneke , S.A. and Ridley-Ellis , D.J. 2013 Within- and between-stand variation in selected properties of Sitka spruce sawn timber in the UK: implications for segregation and grade recovery . Ann. For. Sci. 70 , 403 – 415 . Google Scholar CrossRef Search ADS Pinheiro , J.C. and Bates , D.M. 2000 Mixed-Effects Models in S and S-PLUS . Springer . Google Scholar CrossRef Search ADS Pinheiro , J. , Bates , D. , DebRoy , S. and Sarkar , D. , R Core Team 2016 NLME: Linear and nonlinear mixed effects models. R Packag. version 3.1–128 Version 3., 1–336. R Core Team 2016 R: A Language and Environment for Statistical Computing . R Foundation for Statistical Computing . Rais , A. , Poschenrieder , W. , Pretzsch , H. and Van De Kuilen , J.W.G. 2014 Influence of initial plant density on sawn timber properties for Douglas-fir (Pseudotsuga menziesii (Mirb.) Franco) . Ann. For. Sci. 71 , 617 – 626 . Google Scholar CrossRef Search ADS Roth , B.E. , Li , X. , Huber , D.A. and Peter , G.F. 2007 Effects of management intensity, genetics and planting density on wood stiffness in a plantation of juvenile loblolly pine in the southeastern USA . For. Ecol. Manage. 246 , 155 – 162 . Google Scholar CrossRef Search ADS Sarén , M.-P. , Serimaa , R. , Andersson , S. , Saranpää , P. , Keckes , J. and Fratzl , P. 2004 Effect of growth rate on mean microfibril angle and cross-sectional shape of tracheids of Norway spruce . Trees 18 , 354 – 362 . Google Scholar CrossRef Search ADS Smith , R.G.B. , Rowell , D. , Porada , H. and Bush , D. 2014 Pinus pinaster and Pinus radiata survival, growth and form on 500–800 mm rainfall sites in southern NSW . Aust. For. 77 , 105 – 113 . Google Scholar CrossRef Search ADS Sonderegger , W. , Mandallaz , Æ.D. and Niemz , Æ.P. 2008 An investigation of the influence of selected factors on the properties of spruce wood . Wood Sci. Technol. 42 , 281 – 298 . Google Scholar CrossRef Search ADS Spatz , H. and Bruechert , F. 2000 Basic biomechanics of self-supporting plants: wind loads and gravitational loads on a Norway spruce tree . For. Ecol. Manage. 135 , 33 – 44 . Google Scholar CrossRef Search ADS Theron , K. and Bredenkamp , B.V. 2004 Stand density and stocking in plantations. In Encyclopedia of Forest Sciences . Burley J. , et al. (eds). Elsevier , pp. 829 – 836 . Google Scholar CrossRef Search ADS Timell , T. 1986 Compression wood in gymnosperms. Tong , Q.J. and Zhang , S.Y. 2005 Impact of initial spacing and precommercial thinning on jack pine tree growth and stem quality . For. Chron. 81 , 418 – 428 . Google Scholar CrossRef Search ADS Tsoumis , G. 2009 Science and technology of wood: Strructure, properties, Utilization. Turvey , N.D. , Carlyle , C. and Down , G.M. 1992 Effects of micronutrients on the growth form of two families of Pinus radiata (D. Don) seedlings . Plant Soil 139 , 59 – 65 . Google Scholar CrossRef Search ADS Vikram , V. , Cherry , M.L. , Briggs , D. , Cress , D.W. , Evans , R. and Howe , G.T. 2011 Stiffness of Douglas-fir lumber: effects of wood properties and genetics . Can. J. For. Res. 41 , 1160 – 1173 . Google Scholar CrossRef Search ADS Waghorn , M.J. , Mason , E.G. and Watt , M.S. 2007 a Influence of initial stand density and genotype on longitudinal variation in modulus of elasticity for 17-year-old Pinus radiata . For. Ecol. Manage. 252 , 67 – 72 . Google Scholar CrossRef Search ADS Waghorn , M.J. and Watt , M.S. 2013 a Stand variation in Pinus radiata and its relationship with allometric scaling and critical buckling height . Ann. Bot. 111 , 675 – 680 . Google Scholar CrossRef Search ADS PubMed Waghorn , M.J. , Watt , M.S. and Mason , E.G. 2007 b Influence of tree morphology, genetics, and initial stand density on outerwood modulus of elasticity of 17-year-old Pinus radiata . For. Ecol. Manage. 244 , 86 – 92 . Google Scholar CrossRef Search ADS Walker , J.C.F. and Butterfield , B.G. 1996 The importance of microfibril angle for the processing industries . New Zeal. For. 40 , 34 – 40 . Wang , S.-Y. and Ko , C.-Y. 1998 Dynamic modulus of elasticity and bending properties of large beams of Taiwan-grown Japanese cedar from different plantation spacing sites . J. Wood Sci. 44 , 62 – 68 . Google Scholar CrossRef Search ADS Watt , M.S. , Clinton , P.C. , Parfitt , R.L. , Ross , C. and Coker , G. 2009 Modelling the influence of site and weed competition on juvenile modulus of elasticity in Pinus radiata across broad environmental gradients . For. Ecol. Manage. 258 , 1479 – 1488 . Google Scholar CrossRef Search ADS Watt , M.S. , D’Ath , R. , Leckie , A.C. , Clinton , P.W. , Coker , G. , Davis , M.R. , et al. 2006 b Modelling the influence of stand structural, edaphic and climatic influences on juvenile Pinus radiata fibre length . For. Ecol. Manage. 254 , 166 – 177 . Google Scholar CrossRef Search ADS Watt , M.S. , Moore , J.R. , Facon , J.P. , Downes , G.M. , Clinton , P.W. , Coker , G. , et al. 2006 a Modelling environmental variation in young’s modulus for Pinus radiata and implications for determination of critical buckling height . Ann. Bot. 98 , 765 – 775 . Google Scholar CrossRef Search ADS PubMed Watt , M.S. , Sorensson , C. , Cown , D.J. , Dungey , H.S. and Evans , R. 2010 Determining the main and interactive effect of age and clone on wood density, microfibril angle, and modulus of elasticity for Pinus radiata . Can. J. For. Res. Can. Rech. For. 40 , 1550 – 1557 . Google Scholar CrossRef Search ADS Watt , M.S. , Zoric , B. , Kimberley , M.O. and Harrington , J. 2011 Influence of stocking on radial and longitudinal variation in modulus of elasticity, microfibril angle, and density in a 24-year-old Pinus radiata thinning trial . Can. J. For. Res. 41 , 1422 – 1431 . Google Scholar CrossRef Search ADS Wessels , C.B. , Dowse , G.P. and Smit , H.C. 2011 The flexural properties of young Pinus elliottii × Pinus caribaea var. hondurensis timber from the Southern Cape and their prediction from acoustic measurements . South. For. J. For. Sci. 73 , 137 – 147 . Wessels , B.C. and Froneman , G.M. 2015 The potential for improving the stiffness of young South African grown Pinus radiata lumber by using high planting densities . Pro Lingo 11 , 58 – 64 . Wessels , C.B. , Malan , F.S. , Kidd , M. and Rypstra , T. 2015 a The variation of microfibril angle in South African grown Pinus patula and its influence on the stiffness of structural lumber . South. For. J. For. Sci. 77 , 213 – 219 . Wessels , C.B. , Malan , F.S. , Nel , D.G. and Rypstra , T. 2014 Variation in strength, stiffness and related wood properties in young South African-grown Pinus patula . South. For. J. For. Sci. 76 , 37 – 46 . Wessels , C.B. , Malan , F.S. , Seifert , T. , Louw , J.H. and Rypstra , T. 2015 b The prediction of the flexural lumber properties from standing South African-grown Pinus patula trees . Eur. J. For. Res. 134 , 1 – 18 . Google Scholar CrossRef Search ADS Wielinga , B. , Raymond , C.A. , James , R. and Matheson , A.C. 2009 Effect of green density values on Pinus radiata stiffness estimation using a stress-wave technique . N. Z. J. For. Sci. 39 , 71 – 79 . Woods , P.V. , Nambiar , E.K.S. and Smethurst , P.J. 1992 Effect of annual weeds on water and nitrogen availability to Pinus radiata trees in a young plantation . For. Ecol. Manage. 48 , 145 – 163 . Google Scholar CrossRef Search ADS Wright , J.A. 1994 Utilization of Pinus patula: an annotated bibliography. Oxford Forestry Institute Occasional Paper no. 45. Xu , P. and Walker , J. 2004 Stiffness gradients in radiata pine trees . Wood Sci. Technol. 38 , 1 – 9 . Google Scholar CrossRef Search ADS Yerbury , M.D. and Cooper , R.J. 2017 Curve sawing spruce sawlogs containing sweep can reduce drying distortion when compared with conventional sawing . Forestry 83 , 443 – 450 . Google Scholar CrossRef Search ADS Zobel , B. 1992 Silviculture effects on wood properties . IPEF Int. , 2 , 31 – 38 . Zobel , B. and Van Buijtenen , J. 2012 Wood Variation: Its Causes and Control . Springer Science & Business Media . © Institute of Chartered Foresters, 2018. All rights reserved. For Permissions, please e-mail: journals.permissions@oup.com. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Forestry: An International Journal Of Forest Research Oxford University Press

The effect of planting spacing on Pinus patula stem straightness, microfibril angle and wood density

Loading next page...
 
/lp/ou_press/the-effect-of-planting-spacing-on-pinus-patula-stem-straightness-S7R4TzL73e
Publisher
Oxford University Press
Copyright
© Institute of Chartered Foresters, 2018. All rights reserved. For Permissions, please e-mail: journals.permissions@oup.com.
ISSN
0015-752X
eISSN
1464-3626
D.O.I.
10.1093/forestry/cpy005
Publisher site
See Article on Publisher Site

Abstract

Abstract Improved growth rates and shorter rotation ages have caused a reduction in the stiffness of structural lumber from South African-grown pine plantations. Microfibril angle (MFA) and wood density are known to be two wood properties that influence wood stiffness. Therefore, the objective of this study was to determine the effect of planting spacing of Pinus patula trees, on the MFA and wood density, as well as stem straightness. A total of 171 trees from four spacing treatments (403, 1097, 1808 and 2981 stems ha−1) from an 18-year old experimental P. patula plantation located in Mpumalanga, South Africa, were analysed for wood density, MFA, and ring width. A sub-sample of 81 trees was scanned for tree form using a terrestrial laser scanner. A non-linear mixed-effects model using a power function was developed to model MFA and wood density as a function of ring number and ring width. Planting spacing had a highly significant effect on stem straightness with the most widely spaced trees having the worst mean stem straightness. However, the stem straightness did not increase consistently with increasing stems ha−1. The dynamic modulus of elasticity of standing P. patula trees increased greatly with closer spacing – more so than any other species reported in literature. The mixed model showed that, after accounting for differences due to ring number and ring width, spacing treatment had a significant effect on both the initial MFA and its rate of change with age. For wood density, this remaining effect of spacing treatment was only displayed in its radial rate of change. Based on these results, it seems as if planting spacing has great potential as a management intervention to improve the mechanical wood properties and in certain cases also the stem straightness of South African-grown P. patula at final harvest. Introduction Of all sawn wood produced and sold in South Africa, ~75 per cent is regarded as structural lumber (Crickmay and Associates, 2015), making it the single most important product category for local sawmills. The most important tree resource for these lumber processors is Pinus patula, which accounts for 52.2 per cent of the total softwood area in South Africa (DAFF, 2014). Pinus patula is also widely planted in other African and South American countries with an estimated worldwide total of one million hectares planted with this species in 1994 (Wright, 1994). A critical issue for P. patula structural lumber producers is that a large portion of their end products must conform to the minimum mechanical requirements for structural lumber. This has become more difficult in South Africa in recent years as changes to the plantation resources resulted in reduced mechanical properties of lumber (Burdzik, 2004; Dowse and Wessels, 2013). As improvements in forest management and genetic material have increased growth rates, the harvesting age of South African-grown pine trees, mainly Pinus patula, elliottii, taeda and radiata, managed for saw-log production, has been reduced considerably from ~28 years in 1983 to ~23 years in 2003 (Crickmay et al., 2005). Since then, South African studies have shown a significant reduction in important mechanical properties of lumber, particularly the mean stiffness (modulus of elasticity, MOE) of visually graded lumber (Burdzik, 2004; Wessels et al., 2011; Dowse and Wessels, 2013). Dowse and Wessels (2013) and Wessels et al. (2014) reported the mean MOE of lumber, processed from 16 to 20 year-old P. patula stands, to be ~25 per cent less than required for the lowest and most produced structural grade in South Africa. Globally, reduced mechanical properties of fast growing trees has also become a growing concern as studies from other countries, using different species, have accordingly reported significant proportions of non-compliant structural products harvested at younger ages (Cown, 1992; Kretschmann and Bendtsen, 1992; Biblis and Brinker, 1993; Biblis, 2006). In light of these reports, the South African sawmilling industry needs to address the low MOE of P. patula and other softwood resources to continue the processing thereof into acceptable structural products. The structure of wood cell walls largely determines the mechanical properties such as the MOE of wood (Barnett and Bonham, 2004; Tsoumis, 2009). Microfibril angle (MFA), the orientation of cellulose microfibrils in the secondary cell wall with respect to the longitudinal axis of tracheid cells, and wood density have been shown to be the two most influential properties for Pinus radiata (Cown et al., 1999; Evans and Ilic, 2001; Downes et al., 2002; Xu and Walker, 2004) and P. patula (Wessels et al., 2015a) wood stiffness. However, a poor relationship between MOE and wood density of P. radiata corewood has been reported in some studies (Burdon et al., 2001; Lasserre et al., 2009; Watt et al., 2010). Research by Cown et al. (1999) noted that wood density in P. radiata does become more influential with increasing cambial age. Some authors argued that MFA is the only property to account for large variations in radial MOE trends in fast grown softwoods with wood density acting only as a supporting property (Cave and Walker, 1994; Walker and Butterfield, 1996). In contrast, research relating the average MFA to the stiffness of full-sized lumber, instead of small clear specimens, has shown wood density to have a similar influence on lumber MOE (Wessels et al., 2015a), and in some cases even more so than MFA (Downes et al., 2002; Cown et al., 2004; Vikram et al., 2011). Previous studies on several softwood species showed that planting spacing might influence the mechanical properties and volume recovery. Closely spaced plantations display positive effects on wood stiffness in P. radiata (Lasserre et al., 2005, 2008, 2009; Waghorn et al., 2007a, b; Moore et al., 2015; Wessels and Froneman, 2015) and in other species (Wang and Ko, 1998; Chuang and Wang, 2001; Ishiguri et al., 2005; Roth et al., 2007; Clark III et al., 2008; Moore et al., 2009; Amateis et al., 2013; Rais et al., 2014). The increase in stiffness with increasing stems ha−1 has frequently been attributed to the increase in the height/diameter ratio (slenderness). Based on Euler’s buckling theory, tall, slender trees in competitive environments will require wood that is higher in stiffness in order to resist buckling due to their increasing self-weight (Spatz and Bruechert, 2000; Watt et al., 2006a, b; Waghorn and Watt, 2013a; Merlo et al., 2014; Wessels et al., 2015b). Planting spacing can also influence the straightness of trees (Macdonald and Hubert, 2002), which has economic consequences for log processors. Both the yield and quality of lumber is greatly affected by crooked stems (Cown et al., 1984; Monserud et al., 2003; Ivković et al., 2007; Lachenbruch et al., 2010) and some studies suggest value losses in the sawmill process of roughly 10 per cent due to poor stem straightness (Carino et al., 2006). Leaning stems and those with excessive sweep are known to cause compression wood (Timell, 1986; Krause and Plourde, 2008), which has been shown to reduce wood stiffness (Lindström et al., 2004; Sonderegger et al., 2008). The effect of planting spacing on stem straightness is, however, not always consistent. Trees in stands planted with narrow spacing have been shown to display better stem straightness, improving volume recovery (Malinauskas, 2003; Tong and Zhang, 2005; Belley et al., 2013; Froneman, 2014; Smith et al., 2014) although spacing effects on stem straightness were sometimes less clear (Egbäck et al., 2012; Liziniewicz et al., 2012). On the other hand, trees grown under suppressed conditions in closely spaced plantations may also display poor stem straightness (Theron and Bredenkamp, 2004). The main objective of this study was to determine the effect that planting spacing has on the important properties of MFA and wood density of P. patula trees. At the same time, we also wanted to establish the effect of planting spacing on stem straightness. The results of the study would be useful in formulating future forest management regimes for P. patula grown in South Africa. To the authors’ best knowledge, this is the first study measuring the effect of planting spacing of P. patula on MFA and stem straightness. Materials and methods Experimental layout This study was conducted using an 18 year-old P. patula spacing experiment located in the Mpumalanga escarpment on the Montrose plantation near the town of Barberton, South Africa (25.9037° S, 30.8729° E). This area has a mean annual rainfall of ~850 mm and mean midday temperatures of ~17°C (Barberton aviation weather station, Code: FABR, 25.7175° S, 30.9750° E, 681 m ASL). The trees in this study were pruned at 5, 7 and 9 years after planting to 2, 3.5 and 5.5 m respectively. The experiment followed a randomized complete block design consisting of four planting spacing levels of 403, 1097, 1808 and 2981 stems ha−1, each replicated in two blocks. Each sampling plot had been planted with 49 seedlings in a 7 × 7 tree layout (variable area plots) but only the centre 25 trees were included in the study reported here as the outer trees were considered buffer rows. Out of a possible 200 trees, only 171 were still available for analysis due to mortality, indicated by the survival percentages of treatments in Table 1. Table 1 Plot data and sample sizes Planting spacing (stems ha−1) Mean DBH1 (cm) Mean height1 (m) Survival (%) Total stem volume (m3 ha−1) Sample size1 403 32.7 (34.1, 34.1) 23.3 (23.9, 24.9) 96 321 48 (22, 10) 1097 23.8 (22.8, 25.3) 21.5 (20.9, 22.8) 96 428 48 (24, 10) 1808 19.9 (21.5, 21.7) 20.6 (22.2, 21.2) 84 451 42 (19, 10) 2981 16.9 (17.3, 17.8) 20.3 (20.1, 21.4) 66 417 33 (16, 10) Planting spacing (stems ha−1) Mean DBH1 (cm) Mean height1 (m) Survival (%) Total stem volume (m3 ha−1) Sample size1 403 32.7 (34.1, 34.1) 23.3 (23.9, 24.9) 96 321 48 (22, 10) 1097 23.8 (22.8, 25.3) 21.5 (20.9, 22.8) 96 428 48 (24, 10) 1808 19.9 (21.5, 21.7) 20.6 (22.2, 21.2) 84 451 42 (19, 10) 2981 16.9 (17.3, 17.8) 20.3 (20.1, 21.4) 66 417 33 (16, 10) 1The first and second values in parenthesis indicate values of the sub-sampled trees for stem straightness (TLS measurements) and the removal of increment cores respectively. Table 1 Plot data and sample sizes Planting spacing (stems ha−1) Mean DBH1 (cm) Mean height1 (m) Survival (%) Total stem volume (m3 ha−1) Sample size1 403 32.7 (34.1, 34.1) 23.3 (23.9, 24.9) 96 321 48 (22, 10) 1097 23.8 (22.8, 25.3) 21.5 (20.9, 22.8) 96 428 48 (24, 10) 1808 19.9 (21.5, 21.7) 20.6 (22.2, 21.2) 84 451 42 (19, 10) 2981 16.9 (17.3, 17.8) 20.3 (20.1, 21.4) 66 417 33 (16, 10) Planting spacing (stems ha−1) Mean DBH1 (cm) Mean height1 (m) Survival (%) Total stem volume (m3 ha−1) Sample size1 403 32.7 (34.1, 34.1) 23.3 (23.9, 24.9) 96 321 48 (22, 10) 1097 23.8 (22.8, 25.3) 21.5 (20.9, 22.8) 96 428 48 (24, 10) 1808 19.9 (21.5, 21.7) 20.6 (22.2, 21.2) 84 451 42 (19, 10) 2981 16.9 (17.3, 17.8) 20.3 (20.1, 21.4) 66 417 33 (16, 10) 1The first and second values in parenthesis indicate values of the sub-sampled trees for stem straightness (TLS measurements) and the removal of increment cores respectively. Measurements The stem deviation (from perfect straightness) was measured up to 6 m from tree base using data from a terrestrial laser scanning (TLS) system for one randomly chosen replication block. A Trimble FX phase shift scanner (Trimble Inc.) with angular resolution of 8 sec was used, which results in a sample step of 4 mm at a distance of 20 m. The scan setup used a minimum of four scans per plot. Of the 85 trees available in the chosen replication, a total of 81 were reconstructed from TLS scans and analysed for stem straightness. The other four trees were excluded from the analysis due to either limited scans, causing insufficient points in the point cloud for those trees, or forking below 6 m. The variation of these tree dimensions (manually measured on the 81 trees) from the full sample is indicated in Table 1. We defined stem straightness as the maximum deviation from the stem’s centreline perpendicular to a straight line (chord) joining the two centre-points of the stem at the base and 6 m (Figure 1). The perpendicular deviations were derived through vector equations using three-dimensional coordinates of tree stems which were provided by the 3D Forest software package, version 0.31 (www.3dforest.eu). Stem straightness was then calculated from the set of perpendicular distances for each tree by selecting the maximum. Figure 1 View largeDownload slide Illustration of the measurement of stem straightness. Figure 1 View largeDownload slide Illustration of the measurement of stem straightness. Basal area and the relative stand density (RD), according to Curtis (1982), was calculated for each spacing treatment. The diameter at breast height (DBH) and standing tree height were manually recorded for all trees (Table 1). The slenderness of trees was taken as the ratio of tree height to DBH. The dynamic MOE of standing trees was calculated from stress wave velocities at breast height obtained using the Fakopp Treesonic instrument (Fakopp Enterprise Bt.; Divos, 2010). From the wood density (ρ) – assumed constant at 1000 kg m−3 (Wielinga et al., 2009) – and the stress wave velocity (V), the MOE was then estimated from the following: MOE=ρV2 (1) The probe generally penetrated ~20 mm into the wood and thus effectively only recorded outerwood MOE and was not hindered by bark. Increment cores were taken at breast height (1.3 m) from the northern side of 10 randomly chosen trees per spacing treatment – 40 trees in total. Water in the increment cores was replaced by ethanol in three stages before the cores were dried to equilibrium moisture content. The MFA, wood density and ring widths of each sample were measured using the CSIRO Silviscan 3 apparatus (Evans, 1999) in Melbourne, Australia, at a radial resolution of 2 mm for MFA and 0.025 mm for wood density. Ring widths were defined by the distance between the maxima of wood density of successive rings in the radial wood density profile. In this study, the majority of annual rings had no latewood (LW) according to both interpretations of Mork’s definition of LW cells (Denne, 1989) – which showed that this definition could not be used in our study. A wood density threshold of 500 kg m−3 was then chosen as a definition of LW percentage. This was based on values from literature (Koubaa et al., 2002) and overlaying wood density profiles with images of increment core samples; P. patula typically has distinctly visible darker bands of LW zones. The growth of individual trees varied and therefore the width of the first year rings depended on when the height of a specific tree reached 1.3 m (which was the sampling height) – resulting in widely varying ring widths for the first year ring. Due to varying growth rates some trees only reached a height of 1.3 m after several growth seasons and therefore sometimes had fewer than 15 year rings at breast height. Cores also contained mostly earlywood for the last annual ring as trees were sampled just before winter. Therefore, in the statistical analysis only the 2nd to 13th annual rings were considered. The mean width of rings, from pith to bark, and for each spacing treatment, were also overlaid with a cant sawing pattern and a 4 mm sawing kerf to simulate which annual rings will be present in a given board position. The mean wood properties for simulated board positions were then calculated from the rings demarcated by the sawing pattern. It would have been preferable to destructively sample trees and measure the MOE of the lumber from these trees, but at the time of this study no P. patula spacing experiments were available for destructive testing and therefore the focus of this study was rather on the basic properties of MFA and wood density as well as stem straightness. Statistical analysis The R system for statistical computing (R Core Team, 2016) and Statistica (Dell Inc, 2016) were used for data analysis. Pearson correlations between the various individual tree dimensional variables and average wood properties were computed. The effect of planting spacing on DBH, tree height, stem straightness and dynamic MOE was tested using one-way analysis of variance (ANOVA) where Tukey’s LSD post hoc tests were subsequently performed. Two non-linear mixed-effects models (Pinheiro and Bates, 2000), fitted with the R package ‘nlme’ (Pinheiro et al., 2016), were developed to examine the effects of planting spacing on the pith-to-bark variation in wood density and MFA. The first model was the power function presented by Moore et al. (2015): Yijk=(αi+aij)RNijk(βi+bij)+εijk (2) where Yijk, RNijk and εijk are the response variable (MFA or wood density), the ring number, and the residual error of the kth annual ring in the jth tree in the ith spacing treatment, respectively. The parameters αi and βi correspond to the initial value (ring 1) and the radial rate of change in the response variable, respectively, which could vary for the ith spacing treatment. The aij and bij terms are the random effects for the jth tree in the ith spacing treatment. Considering that planting spacing heavily affects ring width, an additional model incorporating ring width (cf. Auty et al., 2013) was also developed to test if planting spacing still had any effect on the estimated parameters after accounting for differences in ring number and ring width: Yijk=(α0,i+aij+α1,iRWijk)RNijk(β0,i+β1,iRWijk)+εijk (3) All parameters were thus adjusted for RWijk, the ring width of the kth annual ring in the jth tree in the ith spacing treatment, by the α1,i and β1,i parameters which could also vary with the ith spacing treatment. Because only rings 2–13 were considered, the 2nd annual ring was designated as ring number 1 for this analysis. Likelihood ratio tests and Akaike’s information criterion, AIC (Akaike, 1974), were used to evaluate the significance of including each term in both models – random effects, fixed effects and the effect of spacing treatment. Heteroscedasticity was modelled as a power function of ring number (Auty et al., 2013) while the random effect parameters were considered to account for correlations among residuals (Moore et al., 2015). Subsequently, annual ring widths, modelled as an exponential function of cambial age (parameters not shown), were used to predict the radial profiles of wood properties given by equation (3), for each spacing treatment (Auty et al., 2017). Results Stem straightness Planting spacing had a highly significant effect (P < 0.001) on the average stem straightness (Figure 2A and Table 2). However, the trend across spacing treatments was inconsistent. The most widely spaced treatment (403 stems ha−1) had the least straight stems with a mean deviation of 83 mm. Stem straightness for trees from the 1097 stems ha−1 treatment was significantly (P < 0.001) greater than the 403 stems ha−1 treatment and displayed the lowest mean stem deviation of 41 mm. There were no significant differences in the mean stem straightness between the two most closely spaced treatments. Figure 2 View largeDownload slide Means and 95 per cent confidence intervals for stem straightness (A), DBH (B), height (C) and slenderness ratio (D) for each spacing treatment. Different letters denote significant differences at P < 0.05. Figure 2 View largeDownload slide Means and 95 per cent confidence intervals for stem straightness (A), DBH (B), height (C) and slenderness ratio (D) for each spacing treatment. Different letters denote significant differences at P < 0.05. Table 2 Stand-level characteristics for each spacing treatment Planting spacing (stems ha−1) Stem straightness (mm)1 Slenderness MOE (MPa)1 Basal area (m2 ha−1) Relative stand density 403 83.1 (18.8) 0.73 10 150 (620) 33.4 5.8 1097 40.7 (7.3) 0.92 12 739 (624) 48.2 9.8 1808 70.4 (10.0) 1.06 14 607 (1202) 50.0 11.1 2981 55.5 (15.0) 1.23 15 044 (893) 46.3 11.1 Planting spacing (stems ha−1) Stem straightness (mm)1 Slenderness MOE (MPa)1 Basal area (m2 ha−1) Relative stand density 403 83.1 (18.8) 0.73 10 150 (620) 33.4 5.8 1097 40.7 (7.3) 0.92 12 739 (624) 48.2 9.8 1808 70.4 (10.0) 1.06 14 607 (1202) 50.0 11.1 2981 55.5 (15.0) 1.23 15 044 (893) 46.3 11.1 1Values in parenthesis are ± SE. Table 2 Stand-level characteristics for each spacing treatment Planting spacing (stems ha−1) Stem straightness (mm)1 Slenderness MOE (MPa)1 Basal area (m2 ha−1) Relative stand density 403 83.1 (18.8) 0.73 10 150 (620) 33.4 5.8 1097 40.7 (7.3) 0.92 12 739 (624) 48.2 9.8 1808 70.4 (10.0) 1.06 14 607 (1202) 50.0 11.1 2981 55.5 (15.0) 1.23 15 044 (893) 46.3 11.1 Planting spacing (stems ha−1) Stem straightness (mm)1 Slenderness MOE (MPa)1 Basal area (m2 ha−1) Relative stand density 403 83.1 (18.8) 0.73 10 150 (620) 33.4 5.8 1097 40.7 (7.3) 0.92 12 739 (624) 48.2 9.8 1808 70.4 (10.0) 1.06 14 607 (1202) 50.0 11.1 2981 55.5 (15.0) 1.23 15 044 (893) 46.3 11.1 1Values in parenthesis are ± SE. DBH, height, slenderness and site occupancy Planting spacing had a highly significant effect (P < 0.001) on DBH as shown in Figure 2B. As expected, the mean DBH decreased with increasing stems ha−1 – in total, DBH reduced by 48 per cent from the widest to closest spacing (Table 1). Planting spacing had a highly significant effect (P < 0.001) on mean tree height – tree height decreased by 8 per cent from 403 stems ha−1 to 1097 stems ha−1. There was a further non-significant decrease in mean tree height of only 6 per cent between 1097 stems ha−1 and 2981 stems ha−1. The mean slenderness for each spacing treatment can be seen in Table 2 and Figure 2D. The effect of planting spacing on slenderness was highly significant (P < 0.001), increasing by 68 per cent from 403 to 2981 stems ha−1. Both RD and basal area followed an increasing trend from 403 to 1808 stems ha−1, above which RD remained constant while basal area then decreased (Table 2). The most notable increases were between 403 and 1097 stems ha−1. Dynamic MOE Spacing treatment had a highly significant effect (P < 0.001) on the dynamic MOE (Table 2). The mean MOE increased by 48 per cent from 403 to 2981 stems ha−1 – a mean rate of increase (Δ MOE/Δ planting spacing) of 1.9 MPa ha stems−1 (Figure 3). Differences in means were the greatest between 403 and 1097 stems ha−1 and thereafter, displaying smaller differences between the more closely spaced treatments showing an asymptotic type response. Figure 3 View largeDownload slide Means and 95 per cent confidence intervals for MOE for each spacing treatment. Different letters denote significant differences at P < 0.05. Figure 3 View largeDownload slide Means and 95 per cent confidence intervals for MOE for each spacing treatment. Different letters denote significant differences at P < 0.05. Microfibril angle The mean MFA per annual ring across all spacing treatments decreased from 31° to 7° between the 2nd and the 13th year rings (Figure 4A). As expected, MFA displayed a clear decreasing trend with increasing ring number (cambial age). For a given annual ring, the overall trend in MFA was a decreasing angle from 403 stems ha−1 to the closer spacing treatments. The mean MFA for the 403 stems ha−1 treatment decreased to about the 11th annual ring before it reached a constant level of ~12°, while in the more closely spaced treatments, MFA rapidly decreased up to the seventh and eighth annual ring before stabilizing. As a result, the mean MFA decreased by 5°, on average, for the first nine rings from 403 to 1097 stems ha−1. There was also a 5° decrease from 1097 to 1808 stems ha−1 near the pith (the second to fourth annual ring) for equivalent annual rings while the MFA values for treatments 1808 and 2981 stems ha−1 were similar at all annual rings. The model given by equation (2) (parameters not shown, AIC = 2415), showed that only αi was significantly influenced by spacing (P < 0.001). The αi term was significantly greater for 403 and 1097 stems ha−1 compared with the other treatments. Figure 4 View largeDownload slide Variation in microfibril angle (A), wood density (B), latewood percentage (C) and ring width (D) at different spacing treatments and rings from pith. Vertical bars denote 95 per cent confidence intervals. Figure 4 View largeDownload slide Variation in microfibril angle (A), wood density (B), latewood percentage (C) and ring width (D) at different spacing treatments and rings from pith. Vertical bars denote 95 per cent confidence intervals. When ring width was included, spacing treatment had a significant (P < 0.001) effect on all parameters of the model given by equation (3), which had a considerably better fit to the data (lower AIC value of 2244) (Table 3). Modelled MFA was greater for the 403 and 1097 stems ha−1 treatments compared with the other two treatments up to ring 6 (Figure 5A). The rate of decline was clearly lower for the 403 stems ha−1 treatment compared with the other treatments, although this gradient was not mediated through neither β1,i or β0,i (non-significant) but determined by ring width through the α1,i term. Modelled MFA was negatively influenced by ring width due to the significant α1,i term for the 1097 stems ha−1 treatment and the β1,i term for the 1808 stems ha−1 treatment (Table 3). The α0,i and β0,i parameters differed significantly, even after ring width had been taken into account. The α0,i parameter for the 1097 stems ha−1 was significantly greater than for the other treatments, while its β0,i parameter was the only significantly different value relative to 403 stems ha−1. Table 3 Parameter estimates, standard errors, P-values and standard deviations for the random effect estimates of equation (3) Parameters MFA Wood density Estimate SE t-Value P-value Estimate SE t-Value P-value α0 (Int./Fixed) 13.965 2.812 4.966 <0.001 431.803 13.556 31.854 <0.001    α0,1097 24.495 3.616 6.774 <0.001    α0,1808 −9.844 3.063 −3.213 0.001    α0,2981 −4.650 3.475 −1.338 0.182 α1 (Int./Fixed) 0.713 0.119 6.014 <0.001 −2.164 0.624 −3.467 <0.001   α1,1097 −0.992 0.144 −6.895 <0.001   α1,1808 0.502 0.124 4.030 <0.001   α1,2981 0.541 0.162 3.330 <0.001 β0 (Int.) −0.086 0.068 −1.270 0.205 0.023 0.015 1.526 0.128   β0,1097 −0.715 0.073 −9.807 <0.001 0.077 0.012 6.357 <0.001   β0,1808 0.145 0.083 1.739 0.083 0.115 0.013 9.003 <0.001   β0,2981 −0.060 0.100 −0.603 0.547 0.115 0.013 9.099 <0.001 β1 (Int.) 0.001 0.003 0.316 0.753 0.001 0.001 0.446 0.656   β1,1097 0.030 0.004 8.021 <0.001 −0.006 0.002 −3.031 0.003   β1,1808 −0.014 0.004 −3.199 0.002 −0.010 0.002 −4.429 <0.001   β1,2981 −0.001 0.006 −0.069 0.945 −0.012 0.002 −5.433 <0.001 Random parameters Std. Dev. Std. Dev. a0,i - tree 3.098 37.584 εij - residual 0.661 24.486 Parameters MFA Wood density Estimate SE t-Value P-value Estimate SE t-Value P-value α0 (Int./Fixed) 13.965 2.812 4.966 <0.001 431.803 13.556 31.854 <0.001    α0,1097 24.495 3.616 6.774 <0.001    α0,1808 −9.844 3.063 −3.213 0.001    α0,2981 −4.650 3.475 −1.338 0.182 α1 (Int./Fixed) 0.713 0.119 6.014 <0.001 −2.164 0.624 −3.467 <0.001   α1,1097 −0.992 0.144 −6.895 <0.001   α1,1808 0.502 0.124 4.030 <0.001   α1,2981 0.541 0.162 3.330 <0.001 β0 (Int.) −0.086 0.068 −1.270 0.205 0.023 0.015 1.526 0.128   β0,1097 −0.715 0.073 −9.807 <0.001 0.077 0.012 6.357 <0.001   β0,1808 0.145 0.083 1.739 0.083 0.115 0.013 9.003 <0.001   β0,2981 −0.060 0.100 −0.603 0.547 0.115 0.013 9.099 <0.001 β1 (Int.) 0.001 0.003 0.316 0.753 0.001 0.001 0.446 0.656   β1,1097 0.030 0.004 8.021 <0.001 −0.006 0.002 −3.031 0.003   β1,1808 −0.014 0.004 −3.199 0.002 −0.010 0.002 −4.429 <0.001   β1,2981 −0.001 0.006 −0.069 0.945 −0.012 0.002 −5.433 <0.001 Random parameters Std. Dev. Std. Dev. a0,i - tree 3.098 37.584 εij - residual 0.661 24.486 Estimates for the fixed parameters show their values and intercept (Int., i.e. 403 stems ha–1), and values for the other treatments relative to the intercept (i.e. the change in estimate from 403 stems ha−1). The α0 and α1 parameters for wood density are the only fixed parameters (single value for all treatments). Table 3 Parameter estimates, standard errors, P-values and standard deviations for the random effect estimates of equation (3) Parameters MFA Wood density Estimate SE t-Value P-value Estimate SE t-Value P-value α0 (Int./Fixed) 13.965 2.812 4.966 <0.001 431.803 13.556 31.854 <0.001    α0,1097 24.495 3.616 6.774 <0.001    α0,1808 −9.844 3.063 −3.213 0.001    α0,2981 −4.650 3.475 −1.338 0.182 α1 (Int./Fixed) 0.713 0.119 6.014 <0.001 −2.164 0.624 −3.467 <0.001   α1,1097 −0.992 0.144 −6.895 <0.001   α1,1808 0.502 0.124 4.030 <0.001   α1,2981 0.541 0.162 3.330 <0.001 β0 (Int.) −0.086 0.068 −1.270 0.205 0.023 0.015 1.526 0.128   β0,1097 −0.715 0.073 −9.807 <0.001 0.077 0.012 6.357 <0.001   β0,1808 0.145 0.083 1.739 0.083 0.115 0.013 9.003 <0.001   β0,2981 −0.060 0.100 −0.603 0.547 0.115 0.013 9.099 <0.001 β1 (Int.) 0.001 0.003 0.316 0.753 0.001 0.001 0.446 0.656   β1,1097 0.030 0.004 8.021 <0.001 −0.006 0.002 −3.031 0.003   β1,1808 −0.014 0.004 −3.199 0.002 −0.010 0.002 −4.429 <0.001   β1,2981 −0.001 0.006 −0.069 0.945 −0.012 0.002 −5.433 <0.001 Random parameters Std. Dev. Std. Dev. a0,i - tree 3.098 37.584 εij - residual 0.661 24.486 Parameters MFA Wood density Estimate SE t-Value P-value Estimate SE t-Value P-value α0 (Int./Fixed) 13.965 2.812 4.966 <0.001 431.803 13.556 31.854 <0.001    α0,1097 24.495 3.616 6.774 <0.001    α0,1808 −9.844 3.063 −3.213 0.001    α0,2981 −4.650 3.475 −1.338 0.182 α1 (Int./Fixed) 0.713 0.119 6.014 <0.001 −2.164 0.624 −3.467 <0.001   α1,1097 −0.992 0.144 −6.895 <0.001   α1,1808 0.502 0.124 4.030 <0.001   α1,2981 0.541 0.162 3.330 <0.001 β0 (Int.) −0.086 0.068 −1.270 0.205 0.023 0.015 1.526 0.128   β0,1097 −0.715 0.073 −9.807 <0.001 0.077 0.012 6.357 <0.001   β0,1808 0.145 0.083 1.739 0.083 0.115 0.013 9.003 <0.001   β0,2981 −0.060 0.100 −0.603 0.547 0.115 0.013 9.099 <0.001 β1 (Int.) 0.001 0.003 0.316 0.753 0.001 0.001 0.446 0.656   β1,1097 0.030 0.004 8.021 <0.001 −0.006 0.002 −3.031 0.003   β1,1808 −0.014 0.004 −3.199 0.002 −0.010 0.002 −4.429 <0.001   β1,2981 −0.001 0.006 −0.069 0.945 −0.012 0.002 −5.433 <0.001 Random parameters Std. Dev. Std. Dev. a0,i - tree 3.098 37.584 εij - residual 0.661 24.486 Estimates for the fixed parameters show their values and intercept (Int., i.e. 403 stems ha–1), and values for the other treatments relative to the intercept (i.e. the change in estimate from 403 stems ha−1). The α0 and α1 parameters for wood density are the only fixed parameters (single value for all treatments). Figure 5 View largeDownload slide Variation in predicted microfibril angle (A) and wood density (B) from equation (3) at different spacing treatments and rings from pith. Figure 5 View largeDownload slide Variation in predicted microfibril angle (A) and wood density (B) from equation (3) at different spacing treatments and rings from pith. Wood density Wood density varied from ~370 kg m−3 close to the pith to ~600 kg m−3 at ring 10 (Figure 4B). The general trend was an increase in wood density with increasing cambial age up until the 10th annual ring after which it then began to decline. No gradient change was observed in ring width after the 10th annual ring but a similar observation was displayed in the latewood percentage (Figure 4C). There were no clear differences in the wood density between spacing treatments within the first six annual rings, after which the general trend for equivalent annual rings was an increase in wood density with increasing stems ha−1. These differences were most pronounced between 403 stems ha−1 and the other treatments. The model parameters (equation (2)) (AIC = 4941) were significantly different between wide and closer spacing treatments, both αi and βi, showing that the variation of wood density was affected by planting spacing. The model for wood density given by equation (3) (AIC = 4867) (Table 3, Figure 5B) showed that spacing treatment did not significantly influence the initial density of wood ( α0,i). Ring width, however, did not emerge as being a significant contributor to differences in initial wood density between different spacing treatments ( α1,i). Both components of the radial rate of change in wood density ( β0,i and β1,i) was significantly influenced by spacing treatment. This was evident as ring width differed between treatments especially near the pith, while wood density displayed no clear differences in the same region. The rate parameter β0,i , for a given ring width, increased significantly with increased stems ha−1 while ring width also had an increasingly negative influence on modelled wood density in closer spacings ( β1,i, Table 3). Accordingly, the predicted wood density between treatments were similar near the pith, but the incline rate clearly increased from 403 stems ha−1 to 1808 and 2981 stems ha−1 (Figure 5B). The modelled wood density curves for 1808 and 2981 stems ha−1 were nearly completely overlapping (Figure 5B). Relationship between properties Correlations between measured properties were reported in Table 4. Slenderness displayed insignificant correlations with wood density, LW percentage and tree height (despite being a function of tree diameter and tree height) but was strongly related to DBH (r = 0.86). Although slenderness also correlated significantly to MOE, a weak relationship was still displayed (r = 0.42 or r2 = 0.18). MFA was the property with the highest Pearson correlation with MOE (r = −0.66) which weakened somewhat when only considering the outer 20 mm of increment cores. Wood density and LW percentage both correlated significantly with ring width, with an especially high correlation coefficient between wood density and LW percentage (r = 0.89). Table 4 Pearson correlation coefficients between the mean tree variables and wood properties measured or calculated from all 40 increment cores Variable DBH Height Slenderness Ring width Density LW % MFA MOE1 DBH 1.00 0.65 −0.86 0.80 −0.26 −0.19 0.49 −0.51 Height 1.00 −0.24 0.58 0.01 0.05 0.30 −0.19 Slenderness 1.00 −0.63 0.28 0.23 −0.39 0.42 Ring width 1.00 −0.43 −0.36 0.48 −0.42 Density 1.00 0.95 −0.11 0.20 (0.24) LW % 1.00 −0.03 0.05 MFA 1.00 −0.66 (−0.43) MOE 1.00 Variable DBH Height Slenderness Ring width Density LW % MFA MOE1 DBH 1.00 0.65 −0.86 0.80 −0.26 −0.19 0.49 −0.51 Height 1.00 −0.24 0.58 0.01 0.05 0.30 −0.19 Slenderness 1.00 −0.63 0.28 0.23 −0.39 0.42 Ring width 1.00 −0.43 −0.36 0.48 −0.42 Density 1.00 0.95 −0.11 0.20 (0.24) LW % 1.00 −0.03 0.05 MFA 1.00 −0.66 (−0.43) MOE 1.00 1Values in parentheses indicate correlations between MOE and the mean wood properties for the last 20 mm of increment cores; bold correlations are significant at P < 0.05. Table 4 Pearson correlation coefficients between the mean tree variables and wood properties measured or calculated from all 40 increment cores Variable DBH Height Slenderness Ring width Density LW % MFA MOE1 DBH 1.00 0.65 −0.86 0.80 −0.26 −0.19 0.49 −0.51 Height 1.00 −0.24 0.58 0.01 0.05 0.30 −0.19 Slenderness 1.00 −0.63 0.28 0.23 −0.39 0.42 Ring width 1.00 −0.43 −0.36 0.48 −0.42 Density 1.00 0.95 −0.11 0.20 (0.24) LW % 1.00 −0.03 0.05 MFA 1.00 −0.66 (−0.43) MOE 1.00 Variable DBH Height Slenderness Ring width Density LW % MFA MOE1 DBH 1.00 0.65 −0.86 0.80 −0.26 −0.19 0.49 −0.51 Height 1.00 −0.24 0.58 0.01 0.05 0.30 −0.19 Slenderness 1.00 −0.63 0.28 0.23 −0.39 0.42 Ring width 1.00 −0.43 −0.36 0.48 −0.42 Density 1.00 0.95 −0.11 0.20 (0.24) LW % 1.00 −0.03 0.05 MFA 1.00 −0.66 (−0.43) MOE 1.00 1Values in parentheses indicate correlations between MOE and the mean wood properties for the last 20 mm of increment cores; bold correlations are significant at P < 0.05. Simulated lumber properties The pith boards of the 2981 stems ha−1 treatment had a mean MFA of 19.3° compared with 30.3° for 403 stems ha−1 – a 57 per cent increase (Figure 6). Wood density showed a similar trend of increasing with board position and stems ha−1 except for the wood density of the pith boards for 403 stems ha−1, which was quite high due to the high wood density values of the first three rings from the pith for this treatment. The 403 stems ha−1 treatment had trees with greater diameters and displayed the capacity to produce additional boards, but in terms of MFA and wood density, its best board (second from the pith), was still worse than the first boards next to the pith of 2981 stems ha−1. Figure 6 View largeDownload slide The mean ring widths for 403 stems ha−1 (A), 1097 stems ha−1 (B), 1808 stems ha−1 (C) and 2981 stems ha−1 (D), overlaid with a 40 × 120 mm cant sawing strategy drawn to scale including a saw kerf of 4 mm. Maximum and minimum rings are indicated for each board position with their mean MFA and wood density. Figure 6 View largeDownload slide The mean ring widths for 403 stems ha−1 (A), 1097 stems ha−1 (B), 1808 stems ha−1 (C) and 2981 stems ha−1 (D), overlaid with a 40 × 120 mm cant sawing strategy drawn to scale including a saw kerf of 4 mm. Maximum and minimum rings are indicated for each board position with their mean MFA and wood density. Discussion It was expected that the closely spaced treatments would result in straighter trees as some studies have found that decreased planting spacing and competition for sunlight and other growth resources from evenly planted neighbouring trees can help direct growth straight upwards (Woods et al., 1992; Smith et al., 2014). This was partly supported by the finding in our study that the 403 stems ha−1 treatment had the least straight stems. However, an inconsistent trend points to some opposing effects. Part of the reason for this might be that certain soil nutrient deficiencies has been shown to result in stem deformity (Birk, 1991; Turvey et al., 1992). Severe competition between densely planted trees might therefore result in nutrient deficiencies for individual trees. The high RD of both 1808 and 2981 stems ha−1 is close to the defined level of imminent mortality of 12 for South African pines (Kotze and du Toit, 2012) indicating severe competition which could possibly have led to nutrient deficiencies for some trees. We hypothesize that another reason for decreases in stem straightness at closer spacing might be the high mortality rate (Table 1), which can cause uneven openings in the canopy. Poor stem straightness might result from the tree positioning itself towards openings in the canopy. The low survival rates in the two most closely spaced treatments (Table 1) supports this hypothesis. Another factor that should be considered in future studies, but which did not play a role in this study, is the influence of thinning. Treatments with higher stand densities did not always result in straighter stems, but when thinning is performed, they provide the opportunity to remove more trees with poor stem straightness earlier in the life of the trees – thereby removing competition for better shaped trees to grow faster (Macdonald and Hubert, 2002). Planting spacing should thus be considered in combination with possible thinning regimes when evaluating tree form at final harvest. Stem straightness has a big influence on sawmill volume recovery and processing efficiency (Carino et al., 2006; Hamner et al., 2007; Yerbury and Cooper, 2017) and future work should focus on a better understanding of the influence of growing space, manipulated through both planting spacing and thinning, on stem straightness of P. patula. In terms of wood properties, the overall effect of growing space was much clearer. The strong positive influence of planting spacing on the dynamic MOE (Figure 3) was higher than that found in studies on any other softwood species. In studies by Waghorn et al. (2007a; b) and Lasserre et al. (2005) on P. radiata, Roth et al. (2007) on Pinus taeda, and Froneman (2014) on Pinus elliottii, the mean dynamic MOE gradients varied between 0.8 and 1.2 MPa ha stem−1 increases in stems ha−1, which was roughly 50 per cent lower than found in this study (1.8 MPa ha stem−1). However, the range of planting spacing in our study was greater than that of the other studies with the exception of Froneman (2014). At the individual tree level, the poor relationship between MOE and slenderness was contrary to the results of various other studies that found a comparatively good linear relationship. Results by Watt et al. (2006a, 2009), Waghorn et al. (2007b) and Lasserre et al. (2009) on P. radiata found slenderness to explain 49–71 per cent of the variation in tree dynamic MOE. Interestingly, Lasserre et al. (2008) found stem slenderness to seemingly account for variation in tree MOE as a previously significant effect of planting spacing became non-significant once adjustments were made for differences in slenderness (adding slenderness as a covariate), suggesting tree slenderness to be the main mechanism through which closely spaced trees improve wood stiffness. Contrastingly, Roth et al. (2007) found the effect of planting spacing on the MOE of young P. taeda to remain significant after adjustments for stem slenderness, with the authors suggesting environmental and genetic factors to control the outerwood dynamic MOE through mechanisms other than stem form. Similarly, it seems as if the increased MOE with increasing stems ha−1 from this study was not only mediated through tree slenderness. MFA differences observed in our study were similar to results of previous studies displaying reduced MFA with closer spacing at either establishment (Lasserre et al., 2009; Watt et al., 2011) or after thinnings (Moore et al., 2015; Auty et al., 2017), although our study covered a wider range of tree spacing. In general, our study found a positive influence of growth rate on MFA, similar to various other studies (Lindström et al., 1998; Sarén et al., 2004; Auty et al., 2013). Furthermore, a residual effect of planting spacing on the variation of wood properties over and above cambial age and ring width was also evident. This highlights the limitations associated with the common use of ring width (growth rate) as a proxy for spacing since it may not fully capture cause and effect (Zobel, 1992; Auty et al., 2017). It remains unclear why a tree would exhibit lower MFA when planted more densely and growth (or ring width) is reduced, if not due mainly to slenderness and growth rate differences. Other explanations for the apparent differences in wood properties with tree spacing may include the ratio of live crown length to tree height (Kuprevicius et al., 2013). The lower MFA near the pith with increasing stems ha−1 in this study (Figures 4A and 5A), is an important finding. This inner region of the stem forms part of the boards which usually display the worst MOE in logs of P. patula (Wessels et al., 2014), and other species (Xu and Walker, 2004; Vikram et al., 2011; Moore et al., 2012, 2013; Rais et al., 2014; Wessels and Froneman, 2015). The generally high MFA values near the pith is part of the reason for the lower stiffness of pith boards, as saplings with small diameters require more flexibility to prevent fracture of the stem when subjected to wind loading (Barnett and Bonham, 2004; Burgert, 2006). This is one theory that could also possibly explain why MFA generally improved with increasing stems ha−1. Closely planted trees should result in less wind exposure (Green et al., 1995) which has been shown to decrease taper while increasing wood stiffness (Gardiner et al., 1997; Spatz and Bruechert, 2000; Bascuñán et al., 2006; Brüchert and Gardiner, 2006). The decrease of 5° in MFA, on average, from 403 to 1097 stems ha−1 and from 1097 to 1808 stems ha−1 (only for some rings) could potentially be of significant value to structural lumber producers. For P. radiata, a 5° improvement in corewood has been suggested to be enough to represent an increase in wood stiffness of up to 50 per cent (Walker and Butterfield, 1996). In a study by Wessels et al. (2015a) it was shown that the mean wood density, MFA and ring width, calculated from year rings could be used to successfully predict the stiffness of boards. Ring width is an important property since it affects the geometry of sawing and subsequently the individual lumber properties. Therefore, although the improved MFA from 1097 to 1808 stems ha−1 was restricted to only the first few rings, these growth rings will occupy a significant proportion of volume within pith boards as the growth rate is typically greatest at the pith (Figure 6). Furthermore, in trees harvested at a young age, pith boards constitute a large percentage of the total recovered product. Rings closer to the bark, with better mechanical properties, are unfortunately less prevalent as they are mostly chipped away in the sawmilling process. It must be noted that for some increment cores pith eccentricity and compression wood may have had a negative influence on the accuracy of some ring width and wood property measurements respectively. It was noted that the earlier the apparent onset of competition commenced in a stand of trees, the sooner the MFA began to stabilize. The point where maturewood begins is often defined as the radial position where wood properties have stabilized – usually after a transition zone (Cown, 1992; Zobel and Van Buijtenen, 2012). When only considering MFA, it seems as if this transition then occurs earlier for plantations with increased stems ha−1. On the other hand, it was also apparent that the radial MFA and wood density gradients decreased with decreasing stems ha−1 (Figure 4A and B), which has also previously been reported (Malan et al., 1997; Moore et al., 2015). In terms of wood quality, a low MFA gradient is usually considered a positive trait that will guard against uneven shrinkage which leads to warping of lumber products (Huang et al., 2003; Malan, 2010). An interesting result was the decline in wood density after the 10th annual ring (Figure 4B). Distinct shifts in wood density profiles has previously been reported (Moore et al., 2015). However, this decline was consistent for all the spacing treatments and was only mirrored in LW percentage, suggesting that it was probably a function of growing conditions not related to spacing (Cregg et al., 1988; Filipescu et al., 2014). Conclusions In summary, the narrow spacing treatments (1808 and 2981 stems ha−1) gave three distinct advantages in terms of wood properties compared with the wide spacing treatments (403 stems ha−1): First, the absolute mean MFA values of the more closely spaced treatments were significantly lower for rings close to the pith. Second, based on MFA, the juvenile core seems to be restricted to the first seven or eight year rings from the pith, whereas for the 403 stems ha−1 treatment, the juvenile core transition only started at rings 10 or 11. Third, due to suppressed growth, the centre boards of narrow spacings will contain more mature rings than that of the 403 stems ha−1. Combined, this resulted in improved MFA properties at similar board positions for closer spacings which, excluding the pith boards, was also the case for wood density. Comprehensive sawing and lumber testing studies will be required to evaluate the effect of planting spacing on final product properties. After accounting for differences due to ring number and ring age, spacing treatment had a significant effect on both the initial MFA and its rate of change with age. For wood density, this remaining effect of spacing treatment was only displayed in its radial rate of change. The results of this study showed that increased stems ha−1 has the potential to improve the underlying wood properties controlling lumber stiffness. It might be possible to reap the benefits of closely spaced plantations to control wood properties during juvenile growth and then thin a stand to inhibit mortality and potentially improve the average stem form. The sample size in this study was limited and so additional studies are required to reinforce these results. Future work should also include destructive sampling of trees and processing into lumber to evaluate the effect of planting spacing on the actual final product. Stem straightness at the final harvest could possibly also be improved using narrow spacing and thinning but results were not conclusive. More work is required to understand the effect of planting spacing on stem straightness as well as the possible effect of thinning. Acknowledgements We appreciate the help of Wilmour Hendrikse, Hannes Vosloo, Phillip Fischer, Prof. Martin Kidd and Dr Robert Evans with field work, laboratory measurements, data preparation and statistical analysis. Thanks to Prof. Alexis Achim and two anonymous reviewers for very good suggestions to improve earlier versions of the manuscript. Funding The following organizations are gratefully acknowledged for funding this project: Sawmilling SA, Forestry SA, Sappi, York Timbers, Hans Merensky Foundation, and the SA government’s THRIP programme. Conflict of interest statement None declared. References Akaike , H. 1974 A new look at the statistical model identification . IEEE Trans. Automat. Contr. 19 , 716 – 723 . Google Scholar CrossRef Search ADS Amateis , R.L. , Burkhart , H.E. and Jeong , G.Y. 2013 Modulus of elasticity declines with decreasing planting density for loblolly pine (Pinus taeda) plantations . Ann. For. Sci. 70 , 743 – 750 . Google Scholar CrossRef Search ADS Auty , D. , Gardiner , B.A. , Achim , A. , Moore , J.R. and Cameron , A.D. 2013 Models for predicting microfibril angle variation in Scots pine . Ann. For. Sci. 70 , 209 – 218 . Google Scholar CrossRef Search ADS Auty , D. , Moore , J. , Achim , A. , Lyon , A. , Mochan , S. and Gardiner , B. 2017 Effects of early respacing on the density and micro fibril angle of Sitka spruce wood . Forestry , 1 – 13 . https://doi.org/10.1093/forestry/cpx004 Barnett , J.R. and Bonham , V.A. 2004 Cellulose microfibril angle in the cell wall of wood fibres . Biol. Rev. Camb. Philos. Soc. 79 , 461 – 472 . Google Scholar CrossRef Search ADS PubMed Bascuñán , A. , Moore , J.R. and Walker , J.C.F. 2006 Variations in the dynamic modulus of elasticity with proximity to the stand edge in radiata pine stands on the Canterbury Plains, New Zealand . N. Z. J. For. 51 , 4 – 8 . Belley , D. , Beaudoin , M. , Duchesne , I. , Vallerand , S. , Tong , Q.J. and Swift , D.E. 2013 Assessment of white spruce and jack pine stem curvature from a nelder spacing experiment . Wood fiber Sci. 45 , 237 – 249 . Biblis , E. 2006 Flexural properties and compliance to visual grade requirements of 2 by 4 and 2 by 6 loblolly pine lumber obtained from a 19-year-old plantation . For. Prod. J. 56 , 71 – 73 . Biblis , E. and Brinker , R. 1993 Effect of stand age on flexural properties and grade compliance of lumber from loblolly pine plantation timber . For. Prod. J. 43 , 23 . Birk , E.M. 1991 Stem and branch form of 20-year-old radiata pine in relation to previous land use . Aust. For. 54 , 30 – 39 . Google Scholar CrossRef Search ADS Brüchert , F. and Gardiner , B. 2006 The effect of wind exposure on the tree aerial architecture and biomechanics of Sitka spruce (Picea sitchensis, Pinaceae) . Am. J. Bot. 93 , 1512 – 1521 . Google Scholar CrossRef Search ADS PubMed Burdon , R.D. , Britton , R.A.J. and Walford , G.B. 2001 Wood stiffness and bending strength in relation to density in four native provenances of Pinus radiata . N. Z. J. For. Sci. 31 , 130 – 146 . Burdzik , W.M.G. 2004 Grade verification of SA pine—bending, modulus of rupture, modulus of elasticity, tension and compression . S. Afr. For. J. 202 , 21 – 27 . Burgert , I. 2006 Exploring the micromechanical design of plant cell walls . Am. J. Bot. 93 , 1391 – 1401 . Google Scholar CrossRef Search ADS PubMed Carino , H.F. , Arano , K.G. , Blanche , C.A. and Aries Patawaran , M. 2006 Impact of curve sawing on southern pine dimension lumber manufacturing. Part 1. Lumber volume and value yields . For. Prod. J. 56 , 61 – 68 . Cave , I. and Walker , J. 1994 Stiffness of wood in fast-grown plantation softwoods: the influence of microfibril angle . For. Prod. J. 44 , 43 . Chuang , S. and Wang , S. 2001 Evaluation of standing tree quality of Japanese cedar grown with different spacing using stress-wave and ultrasonic-wave methods . J. wood Sci. 47 , 245 – 253 . Google Scholar CrossRef Search ADS Clark , A. , III , Jordan , L. , Schimleck , L. and Daniels , R.F. 2008 Effect of initial planting spacing on wood properties of unthinned loblolly pine at age 21 . For. Prod. J. 58 , 78 – 83 . Cown , D.J. 1992 Corewood (juvenile wood) in Pinus radiata – should we be concerned? N. Z. J. For. Sci. 22 , 87 – 95 . Cown , D.J. , Ball , R.D. and Riddell , M.J.C. 2004 10 Pinus radiata clones: distribution and influence on product performance . N. Z. J. For. Sci. 34 , 293 – 315 . Cown , D.J. , Hebert , J. and Ball , R. 1999 Modelling Radiata pine lumber characteristics. Part 1: mechanical properties of small clears . N. Z. J. For. Sci. 29 , 203 – 213 . Cown , D.J. , McConchie , D.L. and Treloar , C. 1984 Timber recovery from pruned Pinus radiata butt logs at Mangatu: effect of log sweep . N. Z. J. For. Sci. 14 , 109 – 123 . Cregg , B.M. , Dougherty , P.M. and Hennessey , T.C. 1988 Growth and wood quality of young loblolly pine trees in relation to stand density and climatic factors . Can. J. For. Res. 18 , 851 – 858 . Google Scholar CrossRef Search ADS Crickmay , D.G. , Le Brasseur , J. , Stubbings , J.A. and Daugherty , A.E. 2005 Study of the supply and demand of industrial roundwood in South Africa. Copy obtainable from mandy@crickmay.co.za. Crickmay and Associates 2015 South African Lumber Index for February 2015. Curtis , R.O. 1982 A simple index of stand density for Douglas-fir . For. Sci. 28 , 92 – 94 . DAFF 2014 Report on commercial timber resources and primary roundwood processing in South Africa. Dell Inc 2016 STATISTICA (data analysis software system). Denne , M. 1989 Definition of latewood according to Mork (1928) . IAWA J 10 , 59 – 62 . Google Scholar CrossRef Search ADS Divos , F. 2010 Acoustic tools for seedling, tree and log selection. In: The Future of Quality Control for Wood & Wood Products, 4–7th May 2010, Edinburgh The Final Conference of COST Action E53. Downes , G.M. , Nyakuengama , J.G. , Evans , R. , Northway , R. , Blakemore , P. , Dickson , R.L. , et al. 2002 Relationship between wood density, microfibril angle and stiffness in thinned and fertilized Pinus radiata . IAWA J 23 , 253 – 265 . Google Scholar CrossRef Search ADS Dowse , G.P. and Wessels , C.B. 2013 The structural grading of young South African grown Pinus patula sawn timber . South. For. J. For. Sci. 75 , 37 – 41 . Egbäck , S. , Liziniewicz , M. , Högberg , K. , Ekö , P. and Nilsson , U. 2012 Influence of progeny and initial stand density on growth and quality traits of 21 year old half-sib Scots pine (Pinus sylvestris L.) . For. Ecol. Manage. 286 , 1 – 7 . Google Scholar CrossRef Search ADS Evans , R. 1999 A variance approach to the X-ray diffractometric estimation of microfibril angle in wood . Appita J. 52 , 283 – 289 . Evans , R. and Ilic , J. 2001 Rapid prediction of wood stiffness from microfibril angle and density . For. Prod. J. 51 , 53 – 57 . Filipescu , C.N. , Lowell , E.C. , Koppenaal , R. and Mitchell , A.K. 2014 Modeling regional and climatic variation of wood density and ring width in intensively managed Douglas-fir . Can. J. For. Res. 44 , 220 – 229 . Google Scholar CrossRef Search ADS Froneman , G.M. 2014 The effect of initial planting density on selected physical and mechanical properties of South African grown Pinus elliottii and Pinus radiata. University of Stellenbosch. Gardiner , B. a , Stacey , G.R. , Belcher , R.E. and Wood , C.J. 1997 Field and wind tunnel assessments of the implications of respacing and thinning for tree stability . Forestry 70 , 233 – 252 . Google Scholar CrossRef Search ADS Green , S.R. , Grace , J. and Hutchings , N.J. 1995 Observations of turbulent air flow in three stands of widely spaced Sitka spruce . Agric. For. Meteorol. 1923 , 205 – 225 . Google Scholar CrossRef Search ADS Hamner , P. , White , M. and Araman , P. 2007 The frequency and level of sweep in mixed hardwood saw logs in the eastern United States . For. Prod. J. 57 , 23 . Huang , C.-L. , Lindström , H. , Nakada , R. and Ralston , J. 2003 Cell wall structure and wood properties determined by acoustics – a selective review . Holz als Roh- und Werkst 61 , 321 – 335 . Google Scholar CrossRef Search ADS Ishiguri , F. , Kasai , S. , Yokota , S. , Iizuka , K. and Yoshizawa , N. 2005 Wood quality of sugi (Cryptomeria japonica) grown at four initial spacings . IAWA J. 26 , 375 – 386 . Google Scholar CrossRef Search ADS Ivković , M. , Wu , H.X. , Spencer , D.J. , Mcrae , T.A. , Wu , H.X. , Spencer , D.J. , et al. 2007 Modelling the effects of stem sweep, branch size and wood stiffness of radiata pine on structural timber production . Aust. For. 70 , 173 – 184 . Google Scholar CrossRef Search ADS Kotze , H. and du Toit , B. 2012 Silviculture of industrial pine plantations ins South Africa. In South African Forestry Handbook . 5th edn . Bredenkamp B.V. and Upfold S. (eds). The Southern African Institute of Forestry , pp. 123 – 140 . Koubaa , A. , Zhang , S.Y.T. and Makni , S. 2002 Defining the transition from earlywood to latewood in black spruce based on intra-ring wood density profiles from X-ray densitometry . Ann. For. Sci. 59 , 511 – 518 . Google Scholar CrossRef Search ADS Krause , C. and Plourde , P. 2008 Stem deformation in young plantations of black spruce (Picea mariana (Mill.) B.S.P.) and jack pine (Pinus banksiana Lamb.) in the boreal forest of Quebec, Canada . For. Ecol. Manage. 255 , 2213 – 2224 . Google Scholar CrossRef Search ADS Kretschmann , D. and Bendtsen , B. 1992 Ultimate tensile stress and modulus of elasticity of fast-grown plantation loblolly pine lumber . Wood Fiber Sci. 24 , 189 – 203 . Kuprevicius , A. , Auty , D. , Achim , A. and Caspersen , J.P. 2013 Quantifying the influence of live crown ratio on the mechanical properties of clear wood . Forestry 86 , 361 – 369 . Google Scholar CrossRef Search ADS Lachenbruch , B. , Droppelmann , F. , Balocchi , C. , Peredo , M. and Perez , E. 2010 Stem form and compression wood formation in young Pinus radiata trees . Can. J. For. Res. 40 , 26 – 36 . Google Scholar CrossRef Search ADS Lasserre , J. , Mason , E. and Watt , M. 2005 The effects of genotype and spacing on Pinus radiata corewood stiffness in an 11-year old experiment . For. Ecol. Manage. 205 , 375 – 383 . Google Scholar CrossRef Search ADS Lasserre , J.P. , Mason , E.G. and Watt , M.S. 2008 Influence of the main and interactive effects of site, stand stocking and clone on Pinus radiata D. Don corewood modulus of elasticity . For. Ecol. Manage. 255 , 3455 – 3459 . Google Scholar CrossRef Search ADS Lasserre , J.P. , Mason , E.G. , Watt , M.S. and Moore , J.R. 2009 Influence of initial planting spacing and genotype on microfibril angle, wood density, fibre properties and modulus of elasticity in Pinus radiata D. Don corewood . For. Ecol. Manage. 258 , 1924 – 1931 . Google Scholar CrossRef Search ADS Lindström , B.H. , Evans , J.W. and Verrill , S.P. 1998 Influence of cambial age and growth conditions on microfibril angle in young Norway Spruce (Picea abies [L.] Karst.) . Holzforschung 52 , 573 – 581 . Google Scholar CrossRef Search ADS Lindström , H. , Harris , P. , Sorensson , C.T. and Evans , R. 2004 Stiffness and wood variation of 3-year old Pinus radiata clones . Wood Sci. Technol. 38 , 579 – 597 . Google Scholar CrossRef Search ADS Liziniewicz , M. , Ekö , P.M. and Agestam , E. 2012 Effect of spacing on 23-year-old lodgepole pine (Pinus contorta Dougl. var. latifolia) in southern Sweden . Scand. J. For. Res. 27 , 361 – 371 . Google Scholar CrossRef Search ADS Macdonald , E. and Hubert , J. 2002 A review of the effects of silviculture on timber quality of Sitka spruce . Forestry 75 , 107 – 138 . Google Scholar CrossRef Search ADS Malan , F.S. 2010 Corewood in South African pine: necessity and opportunities for improvement . South. For. J. For. Sci. 72 , 99 – 105 . Malan , F.S. , Retief , R.J. and Male , J.R. 1997 The influence of planting espacement on the wood density and pulping properties of Pinus patula . S. Afr. For. J. 180 , 23 – 32 . Malinauskas , A. 2003 Influence of initial density and planning design on the quality of Butt Logs in Scots Pine (Pinus sylvestris L.) plantations . Balt. For. 9 , 10 – 15 . Merlo , E. , Alvarez-Gonzalez , J.G. , Santaclara , O. and Riesco , G. 2014 Modelling modulus of elasticity of Pinus pinaster Ait. in northwestern Spain with standing tree acoustic measurements, tree, stand and site variables . For. Syst. 23 , 153 – 166 . Monserud , R.A. , Parry , D.L. and Todoroki , C.L. 2003 Recovery from simulated sawn logs with sweep recovery from simulated sawn logs with sweep . N. Z. J. For. Sci. 34 , 190 – 205 . Moore , J. , Achim , A. , Lyon , A. , Mochan , S. and Gardiner , B. 2009 Effects of early re-spacing on the physical and mechanical properties of Sitka spruce structural timber . For. Ecol. Manage. 258 , 1174 – 1180 . Google Scholar CrossRef Search ADS Moore , J.R. , Cown , D.J. , McKinley , R.B. and Sabatia , C.O. 2015 Effects of stand density and seedlot on three wood properties of young radiata pine grown at a dry-land site in New Zealand . N. Z. J. For. Sci 45 , 15 . Google Scholar CrossRef Search ADS Moore , J.R. , Lyon , A.J. and Lehneke , S. 2012 Effects of rotation length on the grade recovery and wood properties of Sitka spruce structural timber grown in Great Britain . Ann. For. Sci. 69 , 353 – 362 . Google Scholar CrossRef Search ADS Moore , J.R. , Lyon , A.J. , Searles , G.J. , Lehneke , S.A. and Ridley-Ellis , D.J. 2013 Within- and between-stand variation in selected properties of Sitka spruce sawn timber in the UK: implications for segregation and grade recovery . Ann. For. Sci. 70 , 403 – 415 . Google Scholar CrossRef Search ADS Pinheiro , J.C. and Bates , D.M. 2000 Mixed-Effects Models in S and S-PLUS . Springer . Google Scholar CrossRef Search ADS Pinheiro , J. , Bates , D. , DebRoy , S. and Sarkar , D. , R Core Team 2016 NLME: Linear and nonlinear mixed effects models. R Packag. version 3.1–128 Version 3., 1–336. R Core Team 2016 R: A Language and Environment for Statistical Computing . R Foundation for Statistical Computing . Rais , A. , Poschenrieder , W. , Pretzsch , H. and Van De Kuilen , J.W.G. 2014 Influence of initial plant density on sawn timber properties for Douglas-fir (Pseudotsuga menziesii (Mirb.) Franco) . Ann. For. Sci. 71 , 617 – 626 . Google Scholar CrossRef Search ADS Roth , B.E. , Li , X. , Huber , D.A. and Peter , G.F. 2007 Effects of management intensity, genetics and planting density on wood stiffness in a plantation of juvenile loblolly pine in the southeastern USA . For. Ecol. Manage. 246 , 155 – 162 . Google Scholar CrossRef Search ADS Sarén , M.-P. , Serimaa , R. , Andersson , S. , Saranpää , P. , Keckes , J. and Fratzl , P. 2004 Effect of growth rate on mean microfibril angle and cross-sectional shape of tracheids of Norway spruce . Trees 18 , 354 – 362 . Google Scholar CrossRef Search ADS Smith , R.G.B. , Rowell , D. , Porada , H. and Bush , D. 2014 Pinus pinaster and Pinus radiata survival, growth and form on 500–800 mm rainfall sites in southern NSW . Aust. For. 77 , 105 – 113 . Google Scholar CrossRef Search ADS Sonderegger , W. , Mandallaz , Æ.D. and Niemz , Æ.P. 2008 An investigation of the influence of selected factors on the properties of spruce wood . Wood Sci. Technol. 42 , 281 – 298 . Google Scholar CrossRef Search ADS Spatz , H. and Bruechert , F. 2000 Basic biomechanics of self-supporting plants: wind loads and gravitational loads on a Norway spruce tree . For. Ecol. Manage. 135 , 33 – 44 . Google Scholar CrossRef Search ADS Theron , K. and Bredenkamp , B.V. 2004 Stand density and stocking in plantations. In Encyclopedia of Forest Sciences . Burley J. , et al. (eds). Elsevier , pp. 829 – 836 . Google Scholar CrossRef Search ADS Timell , T. 1986 Compression wood in gymnosperms. Tong , Q.J. and Zhang , S.Y. 2005 Impact of initial spacing and precommercial thinning on jack pine tree growth and stem quality . For. Chron. 81 , 418 – 428 . Google Scholar CrossRef Search ADS Tsoumis , G. 2009 Science and technology of wood: Strructure, properties, Utilization. Turvey , N.D. , Carlyle , C. and Down , G.M. 1992 Effects of micronutrients on the growth form of two families of Pinus radiata (D. Don) seedlings . Plant Soil 139 , 59 – 65 . Google Scholar CrossRef Search ADS Vikram , V. , Cherry , M.L. , Briggs , D. , Cress , D.W. , Evans , R. and Howe , G.T. 2011 Stiffness of Douglas-fir lumber: effects of wood properties and genetics . Can. J. For. Res. 41 , 1160 – 1173 . Google Scholar CrossRef Search ADS Waghorn , M.J. , Mason , E.G. and Watt , M.S. 2007 a Influence of initial stand density and genotype on longitudinal variation in modulus of elasticity for 17-year-old Pinus radiata . For. Ecol. Manage. 252 , 67 – 72 . Google Scholar CrossRef Search ADS Waghorn , M.J. and Watt , M.S. 2013 a Stand variation in Pinus radiata and its relationship with allometric scaling and critical buckling height . Ann. Bot. 111 , 675 – 680 . Google Scholar CrossRef Search ADS PubMed Waghorn , M.J. , Watt , M.S. and Mason , E.G. 2007 b Influence of tree morphology, genetics, and initial stand density on outerwood modulus of elasticity of 17-year-old Pinus radiata . For. Ecol. Manage. 244 , 86 – 92 . Google Scholar CrossRef Search ADS Walker , J.C.F. and Butterfield , B.G. 1996 The importance of microfibril angle for the processing industries . New Zeal. For. 40 , 34 – 40 . Wang , S.-Y. and Ko , C.-Y. 1998 Dynamic modulus of elasticity and bending properties of large beams of Taiwan-grown Japanese cedar from different plantation spacing sites . J. Wood Sci. 44 , 62 – 68 . Google Scholar CrossRef Search ADS Watt , M.S. , Clinton , P.C. , Parfitt , R.L. , Ross , C. and Coker , G. 2009 Modelling the influence of site and weed competition on juvenile modulus of elasticity in Pinus radiata across broad environmental gradients . For. Ecol. Manage. 258 , 1479 – 1488 . Google Scholar CrossRef Search ADS Watt , M.S. , D’Ath , R. , Leckie , A.C. , Clinton , P.W. , Coker , G. , Davis , M.R. , et al. 2006 b Modelling the influence of stand structural, edaphic and climatic influences on juvenile Pinus radiata fibre length . For. Ecol. Manage. 254 , 166 – 177 . Google Scholar CrossRef Search ADS Watt , M.S. , Moore , J.R. , Facon , J.P. , Downes , G.M. , Clinton , P.W. , Coker , G. , et al. 2006 a Modelling environmental variation in young’s modulus for Pinus radiata and implications for determination of critical buckling height . Ann. Bot. 98 , 765 – 775 . Google Scholar CrossRef Search ADS PubMed Watt , M.S. , Sorensson , C. , Cown , D.J. , Dungey , H.S. and Evans , R. 2010 Determining the main and interactive effect of age and clone on wood density, microfibril angle, and modulus of elasticity for Pinus radiata . Can. J. For. Res. Can. Rech. For. 40 , 1550 – 1557 . Google Scholar CrossRef Search ADS Watt , M.S. , Zoric , B. , Kimberley , M.O. and Harrington , J. 2011 Influence of stocking on radial and longitudinal variation in modulus of elasticity, microfibril angle, and density in a 24-year-old Pinus radiata thinning trial . Can. J. For. Res. 41 , 1422 – 1431 . Google Scholar CrossRef Search ADS Wessels , C.B. , Dowse , G.P. and Smit , H.C. 2011 The flexural properties of young Pinus elliottii × Pinus caribaea var. hondurensis timber from the Southern Cape and their prediction from acoustic measurements . South. For. J. For. Sci. 73 , 137 – 147 . Wessels , B.C. and Froneman , G.M. 2015 The potential for improving the stiffness of young South African grown Pinus radiata lumber by using high planting densities . Pro Lingo 11 , 58 – 64 . Wessels , C.B. , Malan , F.S. , Kidd , M. and Rypstra , T. 2015 a The variation of microfibril angle in South African grown Pinus patula and its influence on the stiffness of structural lumber . South. For. J. For. Sci. 77 , 213 – 219 . Wessels , C.B. , Malan , F.S. , Nel , D.G. and Rypstra , T. 2014 Variation in strength, stiffness and related wood properties in young South African-grown Pinus patula . South. For. J. For. Sci. 76 , 37 – 46 . Wessels , C.B. , Malan , F.S. , Seifert , T. , Louw , J.H. and Rypstra , T. 2015 b The prediction of the flexural lumber properties from standing South African-grown Pinus patula trees . Eur. J. For. Res. 134 , 1 – 18 . Google Scholar CrossRef Search ADS Wielinga , B. , Raymond , C.A. , James , R. and Matheson , A.C. 2009 Effect of green density values on Pinus radiata stiffness estimation using a stress-wave technique . N. Z. J. For. Sci. 39 , 71 – 79 . Woods , P.V. , Nambiar , E.K.S. and Smethurst , P.J. 1992 Effect of annual weeds on water and nitrogen availability to Pinus radiata trees in a young plantation . For. Ecol. Manage. 48 , 145 – 163 . Google Scholar CrossRef Search ADS Wright , J.A. 1994 Utilization of Pinus patula: an annotated bibliography. Oxford Forestry Institute Occasional Paper no. 45. Xu , P. and Walker , J. 2004 Stiffness gradients in radiata pine trees . Wood Sci. Technol. 38 , 1 – 9 . Google Scholar CrossRef Search ADS Yerbury , M.D. and Cooper , R.J. 2017 Curve sawing spruce sawlogs containing sweep can reduce drying distortion when compared with conventional sawing . Forestry 83 , 443 – 450 . Google Scholar CrossRef Search ADS Zobel , B. 1992 Silviculture effects on wood properties . IPEF Int. , 2 , 31 – 38 . Zobel , B. and Van Buijtenen , J. 2012 Wood Variation: Its Causes and Control . Springer Science & Business Media . © Institute of Chartered Foresters, 2018. All rights reserved. For Permissions, please e-mail: journals.permissions@oup.com. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)

Journal

Forestry: An International Journal Of Forest ResearchOxford University Press

Published: Mar 2, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off