Variation of lumber properties in genetically improved full-sib families of Douglas-fir in British Columbia, Canada

Variation of lumber properties in genetically improved full-sib families of Douglas-fir in... Abstract Tree breeding to increase forest productivity and resilience is an active area of research. Many studies have examined wood traits of interest to lumber manufacturing, such as wood density, knottiness and microfibril angle, due to the generally negative correlation between rate of growth and wood quality. Relatively little is known, however, about the variation in structural parameters of lumber, i.e. modulus of elasticity (MoE) and modulus of rupture (MoR), in genetically improved trees. In this study we evaluate physico-mechanical properties of lumber from 12 full-sib families in a first-generation Douglas-fir progeny trial. Trees were harvested at age 33 and milled into boards that were tested for MoE and MoR, with specific gravity (SG) and acoustic velocity (AV) also measured. Results indicate that families with lower growth tend to perform better for MoE and MoR, although there are certain families that exhibit higher growth and better MoE and MoR. Boards with less juvenile wood had higher MoE and MoR indicating the significance of board orientation due to sawing pattern; this suggests that radial variation of wood properties is an important factor for genetically improved families. Overall, AV was a better predictor than SG for both MoE and MoR, indicating the potential of using AV in future tree breeding, as well as for product segregation. Findings from this study provide evidence to further develop breeding programmes of Douglas-fir in order to optimize wood production, product quality and ultimately value recovery. Introduction Forest tree breeding has been an active area of research in recent decades, driven by the need to enhance forest productivity to meet increasing demands for timber and other ecosystem services (Hubert and Lee, 2005). Projected global change of climate and the resulting increased impact of pathogens and insects further underscore the additional need to breed tree species for increased resilience to damaging biotic and abiotic factors (Gray et al., 2016; Steffenrem et al., 2016). While the initial objectives of tree breeding aimed to increase the productivity of genetically improved planting stock, the focus has widened to include wood properties in more recent times, given the generally observed trade-off between wood quality and volume production (McLean et al., 2016). Going forward, it is expected that tree breeding will continue to pay attention to wood quality due to the increased use of intensive management regimes and shorter rotations in order to maintain the economic viability and sustainability of forest-related activities. Defined as ‘a measure of the aptness of wood for a particular end use’ (Briggs and Smith, 1986), wood quality emphasizes different groups of wood properties for different types of forest products. For example, strength and stiffness of lumber relies on physico-mechanical properties, such as modulus of elasticity (MoE) and modulus of rupture (MoR), that are closely related to wood specific gravity (SG) and microfibril angle (MFA). Therefore, tree breeders generally tailor breeding programmes based on signals from intended markets and final utilization in conjunction with site characteristics and management regimes (Burdon, 2010). Douglas-fir (Pseudotsuga menziesii) is a major commercial species in British Columbia and the US Pacific Northwest (PNW), owing to its wood properties, fast growth and feasible management regimes. Douglas-fir is grown primarily for solid wood products, which is reflected by current price of sawlogs in local markets ($80–100 per m3), while pulpwood ($30 per m3) is generally considered a by-product of sawlog production. Therefore, breeding programmes of Douglas-fir have traditionally emphasized volume production in connection with characteristics relevant to lumber manufacturing (Howe et al., 2006). Stem volume, branch diameter and wood density have been identified as traits of interest for Douglas-fir selection programmes in the PNW (Aubry et al., 1998), in line with traits of importance identified for exotic plantations of Douglas-fir grown in New Zealand (Whiteside et al., 1976). Multi-trait breeding objectives have been examined for various other species grown primarily for lumber: loblolly pine (Pinus taeda) – growth, stem form, branching and sawtimber potential (Cumbie et al., 2012); radiata pine (Pinus radiata) – growth, stem straightness, branching, wood density and stiffness (Evison and Apiolaza, 2015); Norway spruce (Picea abies) – wood density, stiffness and strength (Chen et al., 2015); Eucalyptus nitens – wood density, stiffness and board checking (Blackburn et al., 2010). The overall take on multi-trait breeding is that complex trade-offs (i.e. negative genetic correlations) between traits make it challenging to optimize selection programmes. It is notable that traits of importance may also differ based on the actual economic model, for example growth potential was identified as the most important trait for plantation growers, while stiffness was the most important trait for sawmills for radiata pine in Australia (Ivkovic et al., 2006). The systematic breeding and testing of coastal Douglas-fir in British Columbia has been initiated in 1976. Progeny tests were established in eight series, on 11 sites every year, for a total of 88 test sites. Currently, advanced generation breeding and testing is carried out by complimentary testing (Stoehr et al., 2008). Over the last 5 years (2012–2016) 15.3 million Douglas-fir trees have been planted annually in BC, with seedlings produced from orchard seed ranging in genetic gain (volume gain in percentage at rotation age of 60) between 15 and 20 per cent (Woods, 2016). On an average site index SI50 of 30 m at a planting density of 1111 trees/ha and rotation age of 60, a 15 per cent genetic gain seedlot can yield 590 m3/ha merchantable volume compared to 512 m3/ha for non-improved trees, based on TASS-TIPSY projections (BC MFLNRO, 2017). It is notable that due to negative genetic correlations between wood density and growth rate in Douglas-fir (El-Kassaby et al., 2011), caution must be exercised so that wood quality is not jeopardized by selecting too aggressively for improved growth. The maturation of the first generation field tests, such as progeny trials, provides a valuable opportunity to revisit and evaluate the impact of genetic selection on growth and wood properties. Our overall goal focused on examining the variation of lumber properties in genetically improved full-sib families of Douglas-fir in a first generation progeny trial in British Columbia, Canada. Evidence will contribute to further refining of breeding programmes to optimize wood production and product quality. Specifically we are asking the following questions: (1) what is the effect of growth gain on wood properties of full-sib families selected for growth potential? and (2) how could correlations between lumber properties be used to improve breeding programmes of Douglas-fir? Material and methods The study site is located on the eastern side of Vancouver Island, south of Nanaimo, British Columbia (49°05′ N, 123°52′ W), in the Coastal Douglas-fir biogeoclimatic zone, moist maritime subzone, characterized by warm, dry summers and mild, wet winters (Green and Klinka, 1994). The test site is located on a flat terrain in a river bed in the close vicinity of the Nanaimo River; thus, the site does not experience summer moisture deficit. This progeny test is part of Series 4, which tested 9 non-overlapping 6-tree partial diallels that created 135 full-sib families from 54 parents. The progeny test was planted at 3 × 3 m spacing in February 1979 in a randomized complete block design (each block containing all 4-tree row plots of each family). The trial did not contain wildstand controls; however, the parent trees were not selected for wood quality traits, therefore they could be considered a random sample of the population. Selections were made in 1989 based on height at age 12 and finalized after Pilodyn estimates of wood density were obtained at age 13. The goal was to select fast-growing parents while holding wood density (estimated by Pilodyn) within 5 per cent of the test population level. A total of 31 forward selections were made at the time on this site. In 1994 all test trees were pruned to a height of 3 m to improve site access and roughly 25 per cent of the surviving trees were removed by cutting the smallest diameter tree in each 4-tree row plot. For the current study we selected 12 full-sib families, four families for each of three genetic gain levels: low-level genetic gain (<2 per cent), mid-level (4.5 per cent) and top-level (11.2 per cent). Gain was expressed as the mid-parent value of volume gain (in percentage above wild stand controls) and calculated as follows: Gainvolumeatrotation=(BV♀+BV♂)/2∗RA (1) where BV♀ and BV♂ are breeding values for female and male parents respectively in volume at age 12 and projected to rotation by including Lambeth’s (1980) juvenile mature correlation; RA is rotation age 60. Lambeth’s correlation is calculated as follows (Xie and Yanchuk, 2003): Lambeth’scorrelation=1.02+0.308ln(selectionage)/RA (2) The low-level genetic gain families could be considered controls as they were not selected for growth and wood quality. Four trees were randomly selected in each of the 12 families for a total sample of 48 trees (Table 1). A wide range of biometric and non-destructive measurements were collected on the standing trees and on the cut stems. Following tree falling, each merchantable stem was sectioned into logs, with the lowest 3 m log used for lumber processing. Table 1 Summary of sample tree characteristics (mean and range) grouped by family level of genetic gain (low, mid and top gain families) Low (n = 16) Mid (n = 16) Top (n = 16) dbh (cm) 30.8 (22.9–42.8) 32.8 (23.0–42.3) 33.5 (22.1–42.6) Stem height (m) 25.9 (22.4–28.8) 27.1 (23.1–32.2) 27.6 (22.4–32.1) Height to live crown (m) 13.1 (2.4–21.5) 14.5 (4.7–18.1) 14.9 (4.1–19.9) Low (n = 16) Mid (n = 16) Top (n = 16) dbh (cm) 30.8 (22.9–42.8) 32.8 (23.0–42.3) 33.5 (22.1–42.6) Stem height (m) 25.9 (22.4–28.8) 27.1 (23.1–32.2) 27.6 (22.4–32.1) Height to live crown (m) 13.1 (2.4–21.5) 14.5 (4.7–18.1) 14.9 (4.1–19.9) Note: n is number of sample trees. Table 1 Summary of sample tree characteristics (mean and range) grouped by family level of genetic gain (low, mid and top gain families) Low (n = 16) Mid (n = 16) Top (n = 16) dbh (cm) 30.8 (22.9–42.8) 32.8 (23.0–42.3) 33.5 (22.1–42.6) Stem height (m) 25.9 (22.4–28.8) 27.1 (23.1–32.2) 27.6 (22.4–32.1) Height to live crown (m) 13.1 (2.4–21.5) 14.5 (4.7–18.1) 14.9 (4.1–19.9) Low (n = 16) Mid (n = 16) Top (n = 16) dbh (cm) 30.8 (22.9–42.8) 32.8 (23.0–42.3) 33.5 (22.1–42.6) Stem height (m) 25.9 (22.4–28.8) 27.1 (23.1–32.2) 27.6 (22.4–32.1) Height to live crown (m) 13.1 (2.4–21.5) 14.5 (4.7–18.1) 14.9 (4.1–19.9) Note: n is number of sample trees. Logs were transported off-site and the 3 m logs were milled into dimension lumber using a portable sawmill. Boards were stacked and air dried in the shade until reaching an equilibrium moisture level of 15 per cent, based on measurements with a 2-pin moisture metre. After trimming the length to 2.44 m (8 ft) and planing, all boards were assessed by a professional grader and graded visually using the National Lumber Grading Authority (NLGA) rules for dimension structural lumber. A subsample of two 2 × 4 boards, one edge-grain and one flat-grain, were selected for each sample tree, for a total of 96 sample boards. The edge-grain boards contained areas closer to the pith with a higher content of juvenile wood than the flat-grain boards that contained generally rings closer to the bark. The boards were subsequently transported to FPInnovations in Vancouver, British Columbia for standardized lumber testing. Acoustic velocity was measured for all boards using the HITMAN HM200 tool (Fibergen, New Zealand). The lumber was tested for edgewise modulus of elasticity (MoE) and modulus of rupture (MoR) under third point bending, based on ASTM D4761-13 (ASTM, 2013). The edgewise bending test used a 59.5 inch (151 cm) span, for a 17:1 span-to-depth ratio. The load was applied at a constant displacement rate of 3.78 inch (6.3 cm) per minute. The bending stiffness test was performed three consecutive times for each board to seat it in the machine supports. On the third loading, the sample board was taken to failure. The edgewise-bending MoE and MoR were calculated using the board dimension taken at mid-span at the time of the test. The specific gravity (SG) was determined for each board using a full cross-section block ~1.5 inches (3.8 cm) long that was taken near the point of failure. Statistical analyses were done using the R statistical programming environment (R Core Team, 2016) to examine the impact of genetic selection on lumber properties. The effect of growth rate, board orientation and family was tested using the following mixed-effects model: yijk=μ+G+Oi+fj+εijk (3) where i is the board orientation (edge-grain and flat-grain); j is the full-sib family j = 1,…,12; k represents the board (two boards per sample tree); yijk is the response variable of interest (specific gravity, acoustic velocity, modulus of elasticity, modulus of rupture); μ is the overall (fixed) mean; G is the fixed effect (continuous variable) of genetic gain; Oi is the fixed effect of the board orientation; fj is the random effect of the j family; and εijk is the residual variation. BLUPs (best linear unbiased predictors) that show departure from the population average for models of wood properties were obtained to evaluate the relative performance of selected full-sib families. The relationships between MoE or MoR and specific gravity (SG) and acoustic velocity (AV) were examined using the following mixed-effects model: yij=μ+d1SGj+d2AVj+fi+εij (4) where i is the full-sib family i = 1,…,12; j is the board (two boards per sample tree, four sample trees per family); yij is the response variable of interest (MoE or MoR); SGj and AVj are specific gravity and acoustic velocity of board j; d1 and d2 are dummy variables that allow the inclusion of either SG or AV or both; fi is the random effect of the i family; and εij is the residual variation. Random effects were assumed to be independent, normally distributed and with 0 centred mean. Results The distributions of lumber parameters indicated differences by board orientation, with inferior properties for edge-grain boards compared with flat-grain boards (Figure 1). The distribution of visual grades for the 2 × 4 boards did not seem to be affected by the level of genetic gain (Table 2). Lumber properties for the 96 tested boards averaged 0.48 for SG, 4.7 km/s for AV, 10.2 GPa for MoE, and 60.1 MPa for MoR (Table 3). There was a declining trend for these wood properties with increased growth expressed as genetic gain (Table 4). Parameter estimates show that for a 10 per cent genetic gain there was a decline of 16 per cent in SG, 12 per cent in AV, 11 per cent in MoE and 13 per cent in MoR, respectively. Figure 1 View largeDownload slide Distributions of modulus of elasticity (MoE) and modulus of rupture (MoR) for tested 2 × 4 boards by board type (n is number of sample boards per category). Figure 1 View largeDownload slide Distributions of modulus of elasticity (MoE) and modulus of rupture (MoR) for tested 2 × 4 boards by board type (n is number of sample boards per category). Table 2 Distribution of sample 2 × 4 boards by visual grade and family level of genetic gain Grade Low Mid Top Total Structural (S) 9 7 13 29 #1 13 14 12 39 #2 10 6 7 23 #3 – 5 – 5 Grade Low Mid Top Total Structural (S) 9 7 13 29 #1 13 14 12 39 #2 10 6 7 23 #3 – 5 – 5 Note: Structural (S) is the highest quality grade, followed by #1, #2 and #3. View Large Table 2 Distribution of sample 2 × 4 boards by visual grade and family level of genetic gain Grade Low Mid Top Total Structural (S) 9 7 13 29 #1 13 14 12 39 #2 10 6 7 23 #3 – 5 – 5 Grade Low Mid Top Total Structural (S) 9 7 13 29 #1 13 14 12 39 #2 10 6 7 23 #3 – 5 – 5 Note: Structural (S) is the highest quality grade, followed by #1, #2 and #3. View Large Table 3 Summary of lumber characteristics (n = 96) Average Minimum Maximum Reference a b Specific gravity SG 0.48 0.38 0.57 0.48 Acoustic velocity AV (km/s) 4.7 3.5 5.3 Modulus of elasticity MoE (GPa) 10.2 3.6 14.9 10.8 11.5 Modulus of rupture MoR (MPa) 60.1 32.7 77.5 55.1 42.1 Average Minimum Maximum Reference a b Specific gravity SG 0.48 0.38 0.57 0.48 Acoustic velocity AV (km/s) 4.7 3.5 5.3 Modulus of elasticity MoE (GPa) 10.2 3.6 14.9 10.8 11.5 Modulus of rupture MoR (MPa) 60.1 32.7 77.5 55.1 42.1 Note: Reference values from (a) Evans and Green (1988), and (b) Dahlen et al. (2012); n is number of sample boards. View Large Table 3 Summary of lumber characteristics (n = 96) Average Minimum Maximum Reference a b Specific gravity SG 0.48 0.38 0.57 0.48 Acoustic velocity AV (km/s) 4.7 3.5 5.3 Modulus of elasticity MoE (GPa) 10.2 3.6 14.9 10.8 11.5 Modulus of rupture MoR (MPa) 60.1 32.7 77.5 55.1 42.1 Average Minimum Maximum Reference a b Specific gravity SG 0.48 0.38 0.57 0.48 Acoustic velocity AV (km/s) 4.7 3.5 5.3 Modulus of elasticity MoE (GPa) 10.2 3.6 14.9 10.8 11.5 Modulus of rupture MoR (MPa) 60.1 32.7 77.5 55.1 42.1 Note: Reference values from (a) Evans and Green (1988), and (b) Dahlen et al. (2012); n is number of sample boards. View Large Table 4 Parameter estimates and fit statistics for lumber properties: specific gravity (SG), acoustic velocity (AV), modulus of elasticity (MoE) and modulus of rupture (MoR) as a function of genetic gain and board orientation Parameter SG AV MoE MoR Intercept 0.5007 4.7625 9.4203 57.2250 Genetic gain –0.0082 –0.0659 –0.1063 –0.7266 Orientation 0.0244 0.3247 2.1462 10.7812 Residual σ 0.0299 0.3239 2.2201 7.7476 R2 0.72 0.64 0.39 0.52 Parameter SG AV MoE MoR Intercept 0.5007 4.7625 9.4203 57.2250 Genetic gain –0.0082 –0.0659 –0.1063 –0.7266 Orientation 0.0244 0.3247 2.1462 10.7812 Residual σ 0.0299 0.3239 2.2201 7.7476 R2 0.72 0.64 0.39 0.52 Note: parameter estimates are based on equation (3); significance level α = 0.05; R2 values are calculated based on the Pearson correlation between observed and predicted values for fixed effects only. View Large Table 4 Parameter estimates and fit statistics for lumber properties: specific gravity (SG), acoustic velocity (AV), modulus of elasticity (MoE) and modulus of rupture (MoR) as a function of genetic gain and board orientation Parameter SG AV MoE MoR Intercept 0.5007 4.7625 9.4203 57.2250 Genetic gain –0.0082 –0.0659 –0.1063 –0.7266 Orientation 0.0244 0.3247 2.1462 10.7812 Residual σ 0.0299 0.3239 2.2201 7.7476 R2 0.72 0.64 0.39 0.52 Parameter SG AV MoE MoR Intercept 0.5007 4.7625 9.4203 57.2250 Genetic gain –0.0082 –0.0659 –0.1063 –0.7266 Orientation 0.0244 0.3247 2.1462 10.7812 Residual σ 0.0299 0.3239 2.2201 7.7476 R2 0.72 0.64 0.39 0.52 Note: parameter estimates are based on equation (3); significance level α = 0.05; R2 values are calculated based on the Pearson correlation between observed and predicted values for fixed effects only. View Large Variation of observed lumber properties illustrated that families with lower levels of growth tended to have better wood properties (Table 5). These results identified several families that were good performers, such as families 141 and 137, as well as families that showed inferior results (families 46 and 39). Table 5 BLUPs (best linear unbiased predictors) for the 12 families for modulus of elasticity (MoE) and modulus of rupture (MoR) Low gain families Mid gain families Top gain families 4 76 111 141 12 46 67 137 33 39 112 140 MoE 0.334 0.876 –0.002 1.063 –1.293 –1.702 –0.714 0.201 0.426 –0.207 0.289 0.313 MoR 0.399 2.247 3.020 3.252 –5.836 –10.962 –0.293 3.129 3.151 –1.640 2.153 1.377 Low gain families Mid gain families Top gain families 4 76 111 141 12 46 67 137 33 39 112 140 MoE 0.334 0.876 –0.002 1.063 –1.293 –1.702 –0.714 0.201 0.426 –0.207 0.289 0.313 MoR 0.399 2.247 3.020 3.252 –5.836 –10.962 –0.293 3.129 3.151 –1.640 2.153 1.377 Note: BLUPs indicate departures from the population average based on equation (3). Table 5 BLUPs (best linear unbiased predictors) for the 12 families for modulus of elasticity (MoE) and modulus of rupture (MoR) Low gain families Mid gain families Top gain families 4 76 111 141 12 46 67 137 33 39 112 140 MoE 0.334 0.876 –0.002 1.063 –1.293 –1.702 –0.714 0.201 0.426 –0.207 0.289 0.313 MoR 0.399 2.247 3.020 3.252 –5.836 –10.962 –0.293 3.129 3.151 –1.640 2.153 1.377 Low gain families Mid gain families Top gain families 4 76 111 141 12 46 67 137 33 39 112 140 MoE 0.334 0.876 –0.002 1.063 –1.293 –1.702 –0.714 0.201 0.426 –0.207 0.289 0.313 MoR 0.399 2.247 3.020 3.252 –5.836 –10.962 –0.293 3.129 3.151 –1.640 2.153 1.377 Note: BLUPs indicate departures from the population average based on equation (3). Predictive models of MoE showed that AV was a better predictor than SG (Figure 2). The MoE model based on AV explained 47 per cent of the observed variation. Moreover, when both AV and SG were used to predict MoE, the parameter estimate for SG became non-significant (Table 6). Similarly, AV was a better predictor than SG for MoR, and combining these two predictors significantly improved predictive performance by explaining 87 per cent of the variation in MoR. Figure 2 View largeDownload slide Relationships of modulus of elasticity (MoE) and modulus of rupture (MoR) with specific gravity (SG) and acoustic velocity (AV) for tested lumber boards (n = 96). Figure 2 View largeDownload slide Relationships of modulus of elasticity (MoE) and modulus of rupture (MoR) with specific gravity (SG) and acoustic velocity (AV) for tested lumber boards (n = 96). Table 6 Parameter estimates and fit statistics for lumber models of modulus of elasticity (MoE) and modulus of rupture (MoR) as a function of lumber specific gravity (SG) and acoustic velocity (AV) Parameter MoE (SG) MoE (AV) MoE (SG, AV) MoR (SG) MoR (AV) MoR (SG, AV) β0 –1.7691 (ns) –10.659 –11.052 –24.948 –46.896 –64.358 β1 (SG) 25.076 1.628 (ns) 117.456 71.978 β2 (AV) 4.466 4.383 22.860 19.217 Residual σ 2.344 1.889 1.897 7.075 4.081 3.493 R2 0.116 0.474 0.469 0.427 0.811 0.872 Parameter MoE (SG) MoE (AV) MoE (SG, AV) MoR (SG) MoR (AV) MoR (SG, AV) β0 –1.7691 (ns) –10.659 –11.052 –24.948 –46.896 –64.358 β1 (SG) 25.076 1.628 (ns) 117.456 71.978 β2 (AV) 4.466 4.383 22.860 19.217 Residual σ 2.344 1.889 1.897 7.075 4.081 3.493 R2 0.116 0.474 0.469 0.427 0.811 0.872 Note: parameter estimates are based on equation (4); significance level α = 0.05, unless otherwise noted (ns); R2 values are calculated based on the Pearson correlation between observed and predicted values for fixed effects only. Dummy variables in equation (4) allowed for the inclusion of either SG (1,0), AV (0,1) or both SG and AV (1,1). View Large Table 6 Parameter estimates and fit statistics for lumber models of modulus of elasticity (MoE) and modulus of rupture (MoR) as a function of lumber specific gravity (SG) and acoustic velocity (AV) Parameter MoE (SG) MoE (AV) MoE (SG, AV) MoR (SG) MoR (AV) MoR (SG, AV) β0 –1.7691 (ns) –10.659 –11.052 –24.948 –46.896 –64.358 β1 (SG) 25.076 1.628 (ns) 117.456 71.978 β2 (AV) 4.466 4.383 22.860 19.217 Residual σ 2.344 1.889 1.897 7.075 4.081 3.493 R2 0.116 0.474 0.469 0.427 0.811 0.872 Parameter MoE (SG) MoE (AV) MoE (SG, AV) MoR (SG) MoR (AV) MoR (SG, AV) β0 –1.7691 (ns) –10.659 –11.052 –24.948 –46.896 –64.358 β1 (SG) 25.076 1.628 (ns) 117.456 71.978 β2 (AV) 4.466 4.383 22.860 19.217 Residual σ 2.344 1.889 1.897 7.075 4.081 3.493 R2 0.116 0.474 0.469 0.427 0.811 0.872 Note: parameter estimates are based on equation (4); significance level α = 0.05, unless otherwise noted (ns); R2 values are calculated based on the Pearson correlation between observed and predicted values for fixed effects only. Dummy variables in equation (4) allowed for the inclusion of either SG (1,0), AV (0,1) or both SG and AV (1,1). View Large Discussion In this study, we examined the variation of lumber properties for 12 full-sib coastal Douglas-fir families in a first-generation progeny trial in British Columbia, Canada. The observed height growth differences between the low gain families and mid and high gain families were 4.6 and 6.6 per cent (Table 1). This is very close to the expected differences as we expected height growth gains to be ~3.6 and 7.7 per cent given the Lambeth’s age–age correlation (r = 0.7 for selection age 12 and forecast to age 35). Tree breeding has the potential to increase volume growth in the range of 10–25 per cent (Jansson et al., 2017), and by doing so to improve the profitability of forest operations and shorten commercial rotations. Historically, the evaluation of lumber properties has relied on small clear specimens owing to the ease of obtaining and testing, despite generally mediocre correlations with lumber properties (Butler et al., 2016). Whenever feasible, it is preferable to rely on testing actual lumber specimens. Our results indicate that higher growth may lead to lower wood properties. In addition, radial variation showed to be an important factor as indicated by lower values for wood properties of edge-grain boards compared with flat-grain boards. On one hand, this points to technological advances that have the potential to customize sawing to optimize and differentiate higher-quality products; on the other hand, with edge-grain boards containing a larger proportion of juvenile wood, it emphasizes possible issues arising from shorter rotations as an indirect result of selecting fast growing planting stock. Given that wood properties of Douglas-fir tend to improve with the transition from juvenile into mature wood (Filipescu et al., 2014), silvicultural practices may have the ability to overcome such pitfalls by planting higher-density stands, followed by thinning to accelerate diameter growth once crowns have lifted (Auty et al., 2016). This strategy would be particularly well suited to Douglas-fir that has the ability to grow vigorously at advanced ages (Curtis, 1992). Planting rapidly growing selected stock may lead to concerns over shorter rotations and by extension harvesting of younger trees containing higher proportions of juvenile wood. Trees in our study were harvested at age 33, relatively younger than the typical commercial rotations (40–60 years) for coastal Douglas-fir in British Columbia. However, our results were encouraging as averages of SG (0.48), MoE (10.2 GPa) and MoR (60.1 MPa) were comparable with reference values for Douglas-fir (Evans and Green, 1988; Dahlen et al., 2012). While the initial planting density (1100 trees per ha) of this progeny trial was similar to operational planting densities, there were several factors that were specific to our study. Trees were pruned at age 15; therefore the influence of knots on lumber properties was limited, but in the absence of pruning, stand density management may be useful by maintaining denser stands in the initial stages of development (Tong et al., 2013; Lowell et al., 2014). The site was located in a river bed, thus it was not characterized by summer moisture deficit that typically limits the formation of denser and stronger latewood (Kantavichai et al., 2010), although silviculture regimes may alleviate moisture deficits by thinning (Diaconu et al., 2015). Moreover, as an additional limitation, it should be noted that lumber in our study was obtained from the lower 3 m log, and stiffness typically declines with log position in the stem (Todoroki et al., 2012). Larger samples covering multiple sites will increase the inference in future studies. Testing of 2 × 4 boards in our study indicated that families of lower genetic gain tend to produce lumber with better physico-mechanical properties, while certain families exhibited both superior growth and wood traits. The existence of families with good growth and wood quality is encouraging and findings are in agreement with previous studies (i.e. Beaulieu et al., 2006) that found a strong negative genetic correlation between stem volume and lumber stiffness, but an absence of significant correlation at phenotypic level for white spruce (Picea glauca), which would allow for selective propagation without compromising growth potential. Similarly, the existence of selected genotypes with superior growth in response to intensive silviculture and no subsequent reduction in stiffness has also been observed in loblolly pine (Roth et al., 2007). Testing of lumber properties to evaluate genetically improved families is possible, as shown by findings of our study; however, this can be challenging as lumber traits are expressed only in later stages of development and lumber testing can be costly and time consuming. In this regard, breeders attempt to reduce the length of selection cycles by utilizing quick and inexpensive non-destructive testing at early ages (Chauhan et al., 2013) or by means of genomic selection (Isik, 2014). Nevertheless, studies of lumber properties from genetically improved families are necessary to provide validation data and baseline information. Increasing productivity and shortening rotations are some of the immediate obvious benefits in using genetically selected planting stock. In addition, anecdotal evidence suggests that rapid initial height growth may also alleviate needs for vegetation management, which would reduce overall regeneration costs. Families in the top-level of genetic gain tested in our study had, on average, an 11 per cent gain, while current operational planting typically uses seed with 15–20 per cent gain; therefore some caution is warranted when generalizing lumber testing results from our study. Furthermore, recent studies have documented that stem defects tend to be more frequent for Douglas-fir grown on productive sites closer to coastal areas in the PNW (Magalska and Howe, 2014), highlighting the need to include stem defects as a secondary selection criterion in breeding programmes. Climate change has the additional potential to influence SG in Douglas-fir (Stoehr et al., 2009) and growth through phenological patterns in the PNW (Ford et al., 2016), as well as increasing the impact of pathogens, such as Swiss needle cast (Phaeocryptopus gaeumannii), that influences growth (Maguire et al., 2011) and indirectly wood quality (Johnson et al., 2005). Models of MoE in our study showed that AV was the main explanatory variable, while SG played only a minor role. ‘Our results also indicated a stronger relationship between MoR and AV (a surrogate for MFA) than between MoE and AV, contrary to what has been reported in previous studies (e.g. McLean et al., 2016; Butler et al., 2017)’.This could be explained by SG having already been included in the criteria of selection of families for growth potential; at the same time, it is possible that the relative young age of trees in our study may have been a factor. Both potential explanations, i.e. the use of SG as criterion of selection and the relative young age of trees in our study, may reduce the overall variability compared with lumber from older trees in natural populations. A better understanding of underlying factors will require further research. In this regard, given shifting structural and functional requirements of trees with age (Lachenbruch et al., 2011), the relative importance of density and AV as determinants of stiffness may vary over time. Both density and MFA were important for predicting dynamic MoE of Pinus patula boards (Wessels et al., 2015) and of small clears of Scots pine (Auty et al., 2016), while density was more important than MFA for stiffness of Douglas-fir mature wood (Lachenbruch et al., 2010). Nevertheless, AV measured on boards of radiata pine was strongly correlated with stiffness (Downes et al., 2002), while the importance of density as a selection criterion may have been overemphasized in previous selection programmes (Apiolaza, 2009). Therefore, AV may play an increasing role both in selection of Douglas-fir for stiffness (Cherry et al., 2008; Vikram et al., 2011), as well as product segregation (Matheson et al., 2002; Wang et al., 2007). In addition, our models of MoR showed a more balanced contribution of both SG and AV when used as predictors, in agreement with other studies that have shown SG has a higher correlation to MoR than to MoE in juvenile wood of radiata pine (Ivkovic et al., 2009). In conclusion, the evaluation of lumber characteristics from genetically selected trees offers a valuable opportunity to assess the impact of early selection decisions, as well as improve breeding programmes going forward. Findings from our case study indicate that families with lower growth tend to perform better in regards to physico-mechanical characteristics of lumber; however there are some families that exhibit both higher growth and good lumber properties. Acoustic velocity shows promising potential for future tree breeding, as well as for product segregation, given that acoustic velocity has been a better predictor than specific gravity for models of modulus of elasticity and modulus of rupture. Evidence from this work will inform development of policies supporting sustainable forest management decisions to optimize wood production and product quality. Acknowledgements Authors express their appreciation to numerous people and organizations that made this study possible. The Nanaimo Park District facilitated site access and tree sampling. Field support was provided by Keith Bird, Lisa Hayton, Tom Bown, Kristina Beckmann, and Ross Koppenaal. Ciprian Pirvu facilitated lumber testing at FPInnovations Vancouver. Funding Funding was provided by the Land Based Investment Strategy (Government of British Columbia) and the Forest Innovation Program (Government of Canada). Conflict of interest statement None declared. References Apiolaza , L.A. 2009 Very early selection for solid wood quality: screening for early winners . Ann. For. Sci. 66 ( 601 ), 1 – 10 . ASTM American Society for Testing and Materials . 2013 . Standard test methods for mechanical properties of lumber and wood-base structural material, D4761-13. Aubry , C.A. , Adams , W.T. and Fahey , T.D. 1998 Determination of relative economic weights for multitrait selection in coastal Douglas-fir . Can. J. For. Res. 28 , 1164 – 1170 . Google Scholar CrossRef Search ADS Auty , D. , Achim , A. , Macdonald , E. , Cameron , A.D. and Gardiner , B.A. 2016 Models for predicting clearwood mechanical properties of Scots pine . For. Sci. 62 , 403 – 413 . BC MFLNRO . 2017 . Growth and Yield Modelling. TIPSY. Ministry of Forests, Lands and Natural Resource Operations. https://www.for.gov.bc.ca/hts/growth/tipsy/tipsy_description.html (last accessed May 25, 2017) Beaulieu , J. , Zhang , S.Y. , Yu , Q. and Rainville , A. 2006 Comparison between genetic and environmental influences on lumber bending properties in young white spruce . Wood Fiber Sci. 38 , 553 – 564 . Blackburn , D. , Hamilton , M. , Harwood , C. , Innes , T. , Potts , B. and Williams , D. 2010 Stiffness and checking of Eucalyptus nitens sawn boards: genetic variation and potential for genetic improvement . Tree Gen. Gen. 6 , 757 – 765 . Google Scholar CrossRef Search ADS Briggs , D.G. and Smith , W.R. 1986 Effects of silvicultural practices on wood properties – a review. In Douglas-fir: Stand Management for the Future . Oliver C.D. , Hanley D.P. and Johnson J.A. (eds). College of Forest Resources, University of Washington , pp. 108 – 117 . Contribution No. 55. Burdon , R.D. 2010 Wood properties and genetic improvement of radiata pine . N. Z. J. For. 55 , 22 – 27 . Butler , M.A. , Dahlen , J. , Antony , F. , Kane , M. , Eberhardt , T.L. , Jin , H. , et al. 2016 Relationships between loblolly pine small clear specimens and dimension lumber tested in static bending . Wood Fiber Sci. 48 , 81 – 95 . Butler , M.A. , Dahlen , J. , Eberhardt , T.L. , Montes , C. , Antony , F. and Daniels , R.F. 2017 Acoustic evaluation of loblolly pine tree- and lumber-length logs allows for segregation of lumber modulus of elasticity, not for modulus of rupture . Ann. For. Sci. 74 , 20 . Google Scholar CrossRef Search ADS Chauhan , S.S. , Sharma , M. , Thomas , J. , Apiolaza , L.A. , Collings , D.A. and Walker , J.C.F. 2013 Methods for the very early selection of Pinus radiata D. Don. for solid wood products . Ann. For. Sci. 70 , 439 – 449 . Google Scholar CrossRef Search ADS Chen , Z.Q. , Karlsson , B. , Lundquist , S.O. , Gil. , M.R.G. , Olsson , L. and Wu , H.X. 2015 Estimating solid wood properties using Pilodyn and acoustic velocity on standing trees of Norway spruce . Ann. For. Sci. 72 , 499 – 508 . Google Scholar CrossRef Search ADS Cherry , M.L. , Vikram , V. , Briggs , D. , Cress , D.W. and Howe , G.T. 2008 Genetic variation in direct and indirect measures of wood stiffness in coastal Douglas-fir . Can. J. For. Res. 38 , 2476 – 2486 . Google Scholar CrossRef Search ADS Cumbie , W.P. , Isik , F. and McKeand , S.E. 2012 Genetic improvement of sawtimber potential in loblolly pine . For. Sci. 58 , 168 – 177 . Curtis , R.O. 1992 A new look at an old question – Douglas-fir culmination age . West. J. Appl. For. 7 , 97 – 99 . Dahlen , J. , Jones , P.D. , Deale , R.D. and Shmulsky , R. 2012 Bending strength and stiffness of in-grade Douglas-fir and southern pine No. 2 2 × 4 lumber . Can. J. For. Res. 42 , 858 – 867 . Google Scholar CrossRef Search ADS Diaconu , D. , Kahle , H.P. and Spiecker , H. 2015 Tree- and stand-level thinning effects on growth of European beech (Fagus sylvatica L.) on a northeast- and a southwest-facing slope in southwest Germany . Forests 6 , 3256 – 3277 . Google Scholar CrossRef Search ADS 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 El-Kassaby , Y.A. , Mansfield , S. , Iisk , F. and Stoehr , M. 2011 In situ wood quality assessment in Douglas-fir . Tree Genet. Gen. 7 , 553 – 561 . Google Scholar CrossRef Search ADS Evans , J.W. and Green , D.W. 1988 Mechanical Properties of Visually Graded Dimension Lumber: Vol. 2 Douglas Fir-Larch . National Technical Information Service , 476 . Publication number PB-88-159-397. Evison , D.C. and Apiolaza , L.A. 2015 Incorporating economic weights into radiata pine breeding selection decisions . Can. J. For. Res. 45 , 135 – 140 . Google Scholar CrossRef Search ADS 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 Ford , K.R. , Harrington , C.A. , Bansal , S. , Gould , P.J. and St. Clair , J.B. 2016 . Will changes in phenology track climate change? A study of growth initiation timing in coast Douglas-fir . Glob Change Biol 22 : 3712 – 3723 . Gray , L.K. , Rweyongeza , D. , Hamann , A. , John , S. and Thomas , B.R. 2016 Developing management strategies for tree improvement programs under climate change: Insights gained from long-term field trials with lodgepole pine . For. Ecol. Manage. 377 , 128 – 138 . Google Scholar CrossRef Search ADS Green , R.N. and Klinka , K. 1994 A field guide to site identification for the Vancouver Forest Region. In Land Management Handbook 28 . Province of BC, Ministry of Forests . Howe , G.T. , Jayawickrama , K. , Cherry , M. , Johnson , G.R. and Wheeler. , N.C. 2006 Breeding Douglas-fir . Plant Breed. Rev. 27 , 245 – 353 . Hubert , J. and Lee , S. 2005 A review of the relative roles of silviculture and tree breeding in tree improvement: the example of Sitka spruce in Britain and possible lessons for hardwood breeding . Forestry 78 , 109 – 120 . Google Scholar CrossRef Search ADS Isik , F. 2014 Genomic selection in forest tree breeding: the concept and an outlook to the future . New For. 45 , 379 – 401 . Google Scholar CrossRef Search ADS Ivkovic , M. , Gapare , W.J. , Abarquez , A. , Ilic , J. , Powell , M.B. and Wu , H.X. 2009 Prediction of wood stiffness, strength, and shrinkage in juvenile wood of radiata pine . Wood Sci. Technol. 43 , 237 – 257 . Google Scholar CrossRef Search ADS Ivkovic , M. , Wu , H.X. , McRae , T.A. and Powell , M.B. 2006 Developing breeding objectives for radiata pine structural wood production. I. Bioeconomic model and economic weights . Can. J. For. Res. 36 , 2920 – 2931 . Google Scholar CrossRef Search ADS Jansson , G. , Hansen , J.K. , Haapanen , M. , Kvaalen , H. and Steffenrem , A. 2017 The genetic and economic gains from forest tree breeding programmes in Scandinavia and Finland . Scand. J. For. Res . 32 , 273 – 286 . Google Scholar CrossRef Search ADS Johnson , G.R. , Grotta , A.T. , Gartner , B.L. and Downes , G. 2005 Impact of the foliar pathogen Swiss needle cast on wood quality of Douglas-fir . Can. J. For. Res. 35 , 331 – 339 . Google Scholar CrossRef Search ADS Kantavichai , R. , Briggs , D. and Turnblom , E. 2010 Modeling effects of soil, climate, and silviculture on growth ring specific gravity of Douglas-fir on a drought-prone site in western Washington . For. Ecol. Manage. 259 , 1085 – 1092 . Google Scholar CrossRef Search ADS Lachenbruch , B. , Johnson , G.R. , Downes , G.M. and Evans , R. 2010 Relationships of density, microfibril angle, and sound velocity with stiffness and strength in mature wood of Douglas-fir . Can. J. For. Res. 40 , 55 – 64 . Google Scholar CrossRef Search ADS Lachenbruch , B. , Moore , J.R. and Evans , R. 2011 Radial variation in wood structure and function in woody plants, and hypotheses for its occurrence. In Age- and Size-related Changes in Tree Structure and Function . Meinzer F.C. , Lachenbruch B. and Dawson T.E. (eds). Springer , pp. 121 – 164 . Google Scholar CrossRef Search ADS Lambeth , C.C. 1980 Juvenile-mature correlations in Pinaceae and implications for early selection . For. Sci. 26 , 571 – 580 . Lowell , E.C. , Maguire , D.A. , Briggs , D.G. , Turnblom , E.C. , Jayawickrama , K.J.S. and Bryce , J. 2014 Effects of silviculture and genetics on branch/knot attributes of coastal Pacific Northwest Douglas-fir and implications for wood quality – a synthesis . Forests 5 , 1717 – 1736 . Google Scholar CrossRef Search ADS Magalska , L. and Howe , G.T. 2014 Genetic and environmental control of Douglas-fir stem defects . For. Ecol. Manage. 318 , 228 – 238 . Google Scholar CrossRef Search ADS Maguire , D.A. , Mainwaring , D.B. and Kanaskie , A. 2011 Ten-year growth and mortality in young Douglas-fir stands experiencing a range in Swiss needle cast severity . Can. J. For. Res. 41 , 2064 – 2076 . Google Scholar CrossRef Search ADS Matheson , A.C. , Dickson , R.L. , Spencer , D.J. , Joe , B. and Ilic , J. 2002 Acoustic segregation of Pinus radiata logs according to stiffness . Ann. For. Sci. 59 , 471 – 477 . Google Scholar CrossRef Search ADS McLean , J.P. , Moore , J.R. , Gardiner , B.A. , Lee , S.J. , Mochan , S.J. and Jarvis , M.C. 2016 Variation of radial wood properties from genetically improved Sitka spruce growing in the UK . Forestry 89 , 109 – 116 . Google Scholar CrossRef Search ADS R Development Core Team 2016 R: A Language and Environment for Statistical Computing . R Foundation for Statistical Computing , Vienna, Austria. Available at: https://www.R-project.org/. 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 Steffenrem , A. , Solheim , H. and Skroppa , T. 2016 Genetic parameters for wood quality traits and resistance to the pathogens Heterobasidion parviporum and Endoconidiophora polonica in a Norway spruce breeding population . Eur. J. For. Res. 135 , 815 – 825 . Google Scholar CrossRef Search ADS Stoehr , M.U. , Ukrainetz , N.K. , Hayton , L.K. and Yanchuk , A.D. 2009 Current and future trends in juvenile wood density for coastal Douglas-fir . Can. J. For. Res. 39 , 1415 – 1419 . Google Scholar CrossRef Search ADS Stoehr , M. , Yanchuk , A. , Xie , C.-Y. and Sanchez , L. 2008 Gain and diversity in advanced generation coastal Douglas-fir selections for seed production populations . Tree Genet. Gen. 4 , 193 – 200 . Google Scholar CrossRef Search ADS Todoroki , C.L. , Lowell , E.C. , Dykstra , D.P. and Briggs , D.G. 2012 Maps and models of density and stiffness within individual Douglas-fir trees . N. Z. J. For. Sci. 42 , 1 – 13 . Tong , Q.J. , Duchesne , I. , Belley , D. , Beaudoin , M. and Swift , E. 2013 Characterization of knots in plantation white spruce . Wood Fibre Sci. 45 , 84 – 97 . 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 Wang , X. , Carter , P. , Ross , R.J. and Brashaw , B.K. 2007 Acoustic assessment of wood quality of raw forest materials – a path to increased profitability . For. Prod. J. 57 , 6 – 14 . Wessels , C.B. , Malan , F.S. , Kidd , M. and Rypstra , T. 2015 The variation of microfibril angle in South African grown Pinus patula and its influence on the stiffness of structural lumber . South. For. 77 , 213 – 219 . Whiteside , I.D. , Wilcox , M.D. and Tustin , J.R. 1976 New Zealand Douglas fir timber quality in relation to silviculture . N. Z. J. For. 22 , 24 – 44 . Woods , J.H. 2016 . Forest Genetics Council of BC Business Plan 2016/17. Available at: http://www.fgcouncil.bc.ca/ Xie , C.-Y. and Yanchuk , A.D. 2003 Breeding values of parental trees, genetic worth of seed orchard seedlots and yields of improved stocks in British Columbia . W. J. Appl. For. 18 ( 2 ), 88 – 100 . © 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

Variation of lumber properties in genetically improved full-sib families of Douglas-fir in British Columbia, Canada

Loading next page...
 
/lp/ou_press/variation-of-lumber-properties-in-genetically-improved-full-sib-OSQ9SoF0OC
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/cpy011
Publisher site
See Article on Publisher Site

Abstract

Abstract Tree breeding to increase forest productivity and resilience is an active area of research. Many studies have examined wood traits of interest to lumber manufacturing, such as wood density, knottiness and microfibril angle, due to the generally negative correlation between rate of growth and wood quality. Relatively little is known, however, about the variation in structural parameters of lumber, i.e. modulus of elasticity (MoE) and modulus of rupture (MoR), in genetically improved trees. In this study we evaluate physico-mechanical properties of lumber from 12 full-sib families in a first-generation Douglas-fir progeny trial. Trees were harvested at age 33 and milled into boards that were tested for MoE and MoR, with specific gravity (SG) and acoustic velocity (AV) also measured. Results indicate that families with lower growth tend to perform better for MoE and MoR, although there are certain families that exhibit higher growth and better MoE and MoR. Boards with less juvenile wood had higher MoE and MoR indicating the significance of board orientation due to sawing pattern; this suggests that radial variation of wood properties is an important factor for genetically improved families. Overall, AV was a better predictor than SG for both MoE and MoR, indicating the potential of using AV in future tree breeding, as well as for product segregation. Findings from this study provide evidence to further develop breeding programmes of Douglas-fir in order to optimize wood production, product quality and ultimately value recovery. Introduction Forest tree breeding has been an active area of research in recent decades, driven by the need to enhance forest productivity to meet increasing demands for timber and other ecosystem services (Hubert and Lee, 2005). Projected global change of climate and the resulting increased impact of pathogens and insects further underscore the additional need to breed tree species for increased resilience to damaging biotic and abiotic factors (Gray et al., 2016; Steffenrem et al., 2016). While the initial objectives of tree breeding aimed to increase the productivity of genetically improved planting stock, the focus has widened to include wood properties in more recent times, given the generally observed trade-off between wood quality and volume production (McLean et al., 2016). Going forward, it is expected that tree breeding will continue to pay attention to wood quality due to the increased use of intensive management regimes and shorter rotations in order to maintain the economic viability and sustainability of forest-related activities. Defined as ‘a measure of the aptness of wood for a particular end use’ (Briggs and Smith, 1986), wood quality emphasizes different groups of wood properties for different types of forest products. For example, strength and stiffness of lumber relies on physico-mechanical properties, such as modulus of elasticity (MoE) and modulus of rupture (MoR), that are closely related to wood specific gravity (SG) and microfibril angle (MFA). Therefore, tree breeders generally tailor breeding programmes based on signals from intended markets and final utilization in conjunction with site characteristics and management regimes (Burdon, 2010). Douglas-fir (Pseudotsuga menziesii) is a major commercial species in British Columbia and the US Pacific Northwest (PNW), owing to its wood properties, fast growth and feasible management regimes. Douglas-fir is grown primarily for solid wood products, which is reflected by current price of sawlogs in local markets ($80–100 per m3), while pulpwood ($30 per m3) is generally considered a by-product of sawlog production. Therefore, breeding programmes of Douglas-fir have traditionally emphasized volume production in connection with characteristics relevant to lumber manufacturing (Howe et al., 2006). Stem volume, branch diameter and wood density have been identified as traits of interest for Douglas-fir selection programmes in the PNW (Aubry et al., 1998), in line with traits of importance identified for exotic plantations of Douglas-fir grown in New Zealand (Whiteside et al., 1976). Multi-trait breeding objectives have been examined for various other species grown primarily for lumber: loblolly pine (Pinus taeda) – growth, stem form, branching and sawtimber potential (Cumbie et al., 2012); radiata pine (Pinus radiata) – growth, stem straightness, branching, wood density and stiffness (Evison and Apiolaza, 2015); Norway spruce (Picea abies) – wood density, stiffness and strength (Chen et al., 2015); Eucalyptus nitens – wood density, stiffness and board checking (Blackburn et al., 2010). The overall take on multi-trait breeding is that complex trade-offs (i.e. negative genetic correlations) between traits make it challenging to optimize selection programmes. It is notable that traits of importance may also differ based on the actual economic model, for example growth potential was identified as the most important trait for plantation growers, while stiffness was the most important trait for sawmills for radiata pine in Australia (Ivkovic et al., 2006). The systematic breeding and testing of coastal Douglas-fir in British Columbia has been initiated in 1976. Progeny tests were established in eight series, on 11 sites every year, for a total of 88 test sites. Currently, advanced generation breeding and testing is carried out by complimentary testing (Stoehr et al., 2008). Over the last 5 years (2012–2016) 15.3 million Douglas-fir trees have been planted annually in BC, with seedlings produced from orchard seed ranging in genetic gain (volume gain in percentage at rotation age of 60) between 15 and 20 per cent (Woods, 2016). On an average site index SI50 of 30 m at a planting density of 1111 trees/ha and rotation age of 60, a 15 per cent genetic gain seedlot can yield 590 m3/ha merchantable volume compared to 512 m3/ha for non-improved trees, based on TASS-TIPSY projections (BC MFLNRO, 2017). It is notable that due to negative genetic correlations between wood density and growth rate in Douglas-fir (El-Kassaby et al., 2011), caution must be exercised so that wood quality is not jeopardized by selecting too aggressively for improved growth. The maturation of the first generation field tests, such as progeny trials, provides a valuable opportunity to revisit and evaluate the impact of genetic selection on growth and wood properties. Our overall goal focused on examining the variation of lumber properties in genetically improved full-sib families of Douglas-fir in a first generation progeny trial in British Columbia, Canada. Evidence will contribute to further refining of breeding programmes to optimize wood production and product quality. Specifically we are asking the following questions: (1) what is the effect of growth gain on wood properties of full-sib families selected for growth potential? and (2) how could correlations between lumber properties be used to improve breeding programmes of Douglas-fir? Material and methods The study site is located on the eastern side of Vancouver Island, south of Nanaimo, British Columbia (49°05′ N, 123°52′ W), in the Coastal Douglas-fir biogeoclimatic zone, moist maritime subzone, characterized by warm, dry summers and mild, wet winters (Green and Klinka, 1994). The test site is located on a flat terrain in a river bed in the close vicinity of the Nanaimo River; thus, the site does not experience summer moisture deficit. This progeny test is part of Series 4, which tested 9 non-overlapping 6-tree partial diallels that created 135 full-sib families from 54 parents. The progeny test was planted at 3 × 3 m spacing in February 1979 in a randomized complete block design (each block containing all 4-tree row plots of each family). The trial did not contain wildstand controls; however, the parent trees were not selected for wood quality traits, therefore they could be considered a random sample of the population. Selections were made in 1989 based on height at age 12 and finalized after Pilodyn estimates of wood density were obtained at age 13. The goal was to select fast-growing parents while holding wood density (estimated by Pilodyn) within 5 per cent of the test population level. A total of 31 forward selections were made at the time on this site. In 1994 all test trees were pruned to a height of 3 m to improve site access and roughly 25 per cent of the surviving trees were removed by cutting the smallest diameter tree in each 4-tree row plot. For the current study we selected 12 full-sib families, four families for each of three genetic gain levels: low-level genetic gain (<2 per cent), mid-level (4.5 per cent) and top-level (11.2 per cent). Gain was expressed as the mid-parent value of volume gain (in percentage above wild stand controls) and calculated as follows: Gainvolumeatrotation=(BV♀+BV♂)/2∗RA (1) where BV♀ and BV♂ are breeding values for female and male parents respectively in volume at age 12 and projected to rotation by including Lambeth’s (1980) juvenile mature correlation; RA is rotation age 60. Lambeth’s correlation is calculated as follows (Xie and Yanchuk, 2003): Lambeth’scorrelation=1.02+0.308ln(selectionage)/RA (2) The low-level genetic gain families could be considered controls as they were not selected for growth and wood quality. Four trees were randomly selected in each of the 12 families for a total sample of 48 trees (Table 1). A wide range of biometric and non-destructive measurements were collected on the standing trees and on the cut stems. Following tree falling, each merchantable stem was sectioned into logs, with the lowest 3 m log used for lumber processing. Table 1 Summary of sample tree characteristics (mean and range) grouped by family level of genetic gain (low, mid and top gain families) Low (n = 16) Mid (n = 16) Top (n = 16) dbh (cm) 30.8 (22.9–42.8) 32.8 (23.0–42.3) 33.5 (22.1–42.6) Stem height (m) 25.9 (22.4–28.8) 27.1 (23.1–32.2) 27.6 (22.4–32.1) Height to live crown (m) 13.1 (2.4–21.5) 14.5 (4.7–18.1) 14.9 (4.1–19.9) Low (n = 16) Mid (n = 16) Top (n = 16) dbh (cm) 30.8 (22.9–42.8) 32.8 (23.0–42.3) 33.5 (22.1–42.6) Stem height (m) 25.9 (22.4–28.8) 27.1 (23.1–32.2) 27.6 (22.4–32.1) Height to live crown (m) 13.1 (2.4–21.5) 14.5 (4.7–18.1) 14.9 (4.1–19.9) Note: n is number of sample trees. Table 1 Summary of sample tree characteristics (mean and range) grouped by family level of genetic gain (low, mid and top gain families) Low (n = 16) Mid (n = 16) Top (n = 16) dbh (cm) 30.8 (22.9–42.8) 32.8 (23.0–42.3) 33.5 (22.1–42.6) Stem height (m) 25.9 (22.4–28.8) 27.1 (23.1–32.2) 27.6 (22.4–32.1) Height to live crown (m) 13.1 (2.4–21.5) 14.5 (4.7–18.1) 14.9 (4.1–19.9) Low (n = 16) Mid (n = 16) Top (n = 16) dbh (cm) 30.8 (22.9–42.8) 32.8 (23.0–42.3) 33.5 (22.1–42.6) Stem height (m) 25.9 (22.4–28.8) 27.1 (23.1–32.2) 27.6 (22.4–32.1) Height to live crown (m) 13.1 (2.4–21.5) 14.5 (4.7–18.1) 14.9 (4.1–19.9) Note: n is number of sample trees. Logs were transported off-site and the 3 m logs were milled into dimension lumber using a portable sawmill. Boards were stacked and air dried in the shade until reaching an equilibrium moisture level of 15 per cent, based on measurements with a 2-pin moisture metre. After trimming the length to 2.44 m (8 ft) and planing, all boards were assessed by a professional grader and graded visually using the National Lumber Grading Authority (NLGA) rules for dimension structural lumber. A subsample of two 2 × 4 boards, one edge-grain and one flat-grain, were selected for each sample tree, for a total of 96 sample boards. The edge-grain boards contained areas closer to the pith with a higher content of juvenile wood than the flat-grain boards that contained generally rings closer to the bark. The boards were subsequently transported to FPInnovations in Vancouver, British Columbia for standardized lumber testing. Acoustic velocity was measured for all boards using the HITMAN HM200 tool (Fibergen, New Zealand). The lumber was tested for edgewise modulus of elasticity (MoE) and modulus of rupture (MoR) under third point bending, based on ASTM D4761-13 (ASTM, 2013). The edgewise bending test used a 59.5 inch (151 cm) span, for a 17:1 span-to-depth ratio. The load was applied at a constant displacement rate of 3.78 inch (6.3 cm) per minute. The bending stiffness test was performed three consecutive times for each board to seat it in the machine supports. On the third loading, the sample board was taken to failure. The edgewise-bending MoE and MoR were calculated using the board dimension taken at mid-span at the time of the test. The specific gravity (SG) was determined for each board using a full cross-section block ~1.5 inches (3.8 cm) long that was taken near the point of failure. Statistical analyses were done using the R statistical programming environment (R Core Team, 2016) to examine the impact of genetic selection on lumber properties. The effect of growth rate, board orientation and family was tested using the following mixed-effects model: yijk=μ+G+Oi+fj+εijk (3) where i is the board orientation (edge-grain and flat-grain); j is the full-sib family j = 1,…,12; k represents the board (two boards per sample tree); yijk is the response variable of interest (specific gravity, acoustic velocity, modulus of elasticity, modulus of rupture); μ is the overall (fixed) mean; G is the fixed effect (continuous variable) of genetic gain; Oi is the fixed effect of the board orientation; fj is the random effect of the j family; and εijk is the residual variation. BLUPs (best linear unbiased predictors) that show departure from the population average for models of wood properties were obtained to evaluate the relative performance of selected full-sib families. The relationships between MoE or MoR and specific gravity (SG) and acoustic velocity (AV) were examined using the following mixed-effects model: yij=μ+d1SGj+d2AVj+fi+εij (4) where i is the full-sib family i = 1,…,12; j is the board (two boards per sample tree, four sample trees per family); yij is the response variable of interest (MoE or MoR); SGj and AVj are specific gravity and acoustic velocity of board j; d1 and d2 are dummy variables that allow the inclusion of either SG or AV or both; fi is the random effect of the i family; and εij is the residual variation. Random effects were assumed to be independent, normally distributed and with 0 centred mean. Results The distributions of lumber parameters indicated differences by board orientation, with inferior properties for edge-grain boards compared with flat-grain boards (Figure 1). The distribution of visual grades for the 2 × 4 boards did not seem to be affected by the level of genetic gain (Table 2). Lumber properties for the 96 tested boards averaged 0.48 for SG, 4.7 km/s for AV, 10.2 GPa for MoE, and 60.1 MPa for MoR (Table 3). There was a declining trend for these wood properties with increased growth expressed as genetic gain (Table 4). Parameter estimates show that for a 10 per cent genetic gain there was a decline of 16 per cent in SG, 12 per cent in AV, 11 per cent in MoE and 13 per cent in MoR, respectively. Figure 1 View largeDownload slide Distributions of modulus of elasticity (MoE) and modulus of rupture (MoR) for tested 2 × 4 boards by board type (n is number of sample boards per category). Figure 1 View largeDownload slide Distributions of modulus of elasticity (MoE) and modulus of rupture (MoR) for tested 2 × 4 boards by board type (n is number of sample boards per category). Table 2 Distribution of sample 2 × 4 boards by visual grade and family level of genetic gain Grade Low Mid Top Total Structural (S) 9 7 13 29 #1 13 14 12 39 #2 10 6 7 23 #3 – 5 – 5 Grade Low Mid Top Total Structural (S) 9 7 13 29 #1 13 14 12 39 #2 10 6 7 23 #3 – 5 – 5 Note: Structural (S) is the highest quality grade, followed by #1, #2 and #3. View Large Table 2 Distribution of sample 2 × 4 boards by visual grade and family level of genetic gain Grade Low Mid Top Total Structural (S) 9 7 13 29 #1 13 14 12 39 #2 10 6 7 23 #3 – 5 – 5 Grade Low Mid Top Total Structural (S) 9 7 13 29 #1 13 14 12 39 #2 10 6 7 23 #3 – 5 – 5 Note: Structural (S) is the highest quality grade, followed by #1, #2 and #3. View Large Table 3 Summary of lumber characteristics (n = 96) Average Minimum Maximum Reference a b Specific gravity SG 0.48 0.38 0.57 0.48 Acoustic velocity AV (km/s) 4.7 3.5 5.3 Modulus of elasticity MoE (GPa) 10.2 3.6 14.9 10.8 11.5 Modulus of rupture MoR (MPa) 60.1 32.7 77.5 55.1 42.1 Average Minimum Maximum Reference a b Specific gravity SG 0.48 0.38 0.57 0.48 Acoustic velocity AV (km/s) 4.7 3.5 5.3 Modulus of elasticity MoE (GPa) 10.2 3.6 14.9 10.8 11.5 Modulus of rupture MoR (MPa) 60.1 32.7 77.5 55.1 42.1 Note: Reference values from (a) Evans and Green (1988), and (b) Dahlen et al. (2012); n is number of sample boards. View Large Table 3 Summary of lumber characteristics (n = 96) Average Minimum Maximum Reference a b Specific gravity SG 0.48 0.38 0.57 0.48 Acoustic velocity AV (km/s) 4.7 3.5 5.3 Modulus of elasticity MoE (GPa) 10.2 3.6 14.9 10.8 11.5 Modulus of rupture MoR (MPa) 60.1 32.7 77.5 55.1 42.1 Average Minimum Maximum Reference a b Specific gravity SG 0.48 0.38 0.57 0.48 Acoustic velocity AV (km/s) 4.7 3.5 5.3 Modulus of elasticity MoE (GPa) 10.2 3.6 14.9 10.8 11.5 Modulus of rupture MoR (MPa) 60.1 32.7 77.5 55.1 42.1 Note: Reference values from (a) Evans and Green (1988), and (b) Dahlen et al. (2012); n is number of sample boards. View Large Table 4 Parameter estimates and fit statistics for lumber properties: specific gravity (SG), acoustic velocity (AV), modulus of elasticity (MoE) and modulus of rupture (MoR) as a function of genetic gain and board orientation Parameter SG AV MoE MoR Intercept 0.5007 4.7625 9.4203 57.2250 Genetic gain –0.0082 –0.0659 –0.1063 –0.7266 Orientation 0.0244 0.3247 2.1462 10.7812 Residual σ 0.0299 0.3239 2.2201 7.7476 R2 0.72 0.64 0.39 0.52 Parameter SG AV MoE MoR Intercept 0.5007 4.7625 9.4203 57.2250 Genetic gain –0.0082 –0.0659 –0.1063 –0.7266 Orientation 0.0244 0.3247 2.1462 10.7812 Residual σ 0.0299 0.3239 2.2201 7.7476 R2 0.72 0.64 0.39 0.52 Note: parameter estimates are based on equation (3); significance level α = 0.05; R2 values are calculated based on the Pearson correlation between observed and predicted values for fixed effects only. View Large Table 4 Parameter estimates and fit statistics for lumber properties: specific gravity (SG), acoustic velocity (AV), modulus of elasticity (MoE) and modulus of rupture (MoR) as a function of genetic gain and board orientation Parameter SG AV MoE MoR Intercept 0.5007 4.7625 9.4203 57.2250 Genetic gain –0.0082 –0.0659 –0.1063 –0.7266 Orientation 0.0244 0.3247 2.1462 10.7812 Residual σ 0.0299 0.3239 2.2201 7.7476 R2 0.72 0.64 0.39 0.52 Parameter SG AV MoE MoR Intercept 0.5007 4.7625 9.4203 57.2250 Genetic gain –0.0082 –0.0659 –0.1063 –0.7266 Orientation 0.0244 0.3247 2.1462 10.7812 Residual σ 0.0299 0.3239 2.2201 7.7476 R2 0.72 0.64 0.39 0.52 Note: parameter estimates are based on equation (3); significance level α = 0.05; R2 values are calculated based on the Pearson correlation between observed and predicted values for fixed effects only. View Large Variation of observed lumber properties illustrated that families with lower levels of growth tended to have better wood properties (Table 5). These results identified several families that were good performers, such as families 141 and 137, as well as families that showed inferior results (families 46 and 39). Table 5 BLUPs (best linear unbiased predictors) for the 12 families for modulus of elasticity (MoE) and modulus of rupture (MoR) Low gain families Mid gain families Top gain families 4 76 111 141 12 46 67 137 33 39 112 140 MoE 0.334 0.876 –0.002 1.063 –1.293 –1.702 –0.714 0.201 0.426 –0.207 0.289 0.313 MoR 0.399 2.247 3.020 3.252 –5.836 –10.962 –0.293 3.129 3.151 –1.640 2.153 1.377 Low gain families Mid gain families Top gain families 4 76 111 141 12 46 67 137 33 39 112 140 MoE 0.334 0.876 –0.002 1.063 –1.293 –1.702 –0.714 0.201 0.426 –0.207 0.289 0.313 MoR 0.399 2.247 3.020 3.252 –5.836 –10.962 –0.293 3.129 3.151 –1.640 2.153 1.377 Note: BLUPs indicate departures from the population average based on equation (3). Table 5 BLUPs (best linear unbiased predictors) for the 12 families for modulus of elasticity (MoE) and modulus of rupture (MoR) Low gain families Mid gain families Top gain families 4 76 111 141 12 46 67 137 33 39 112 140 MoE 0.334 0.876 –0.002 1.063 –1.293 –1.702 –0.714 0.201 0.426 –0.207 0.289 0.313 MoR 0.399 2.247 3.020 3.252 –5.836 –10.962 –0.293 3.129 3.151 –1.640 2.153 1.377 Low gain families Mid gain families Top gain families 4 76 111 141 12 46 67 137 33 39 112 140 MoE 0.334 0.876 –0.002 1.063 –1.293 –1.702 –0.714 0.201 0.426 –0.207 0.289 0.313 MoR 0.399 2.247 3.020 3.252 –5.836 –10.962 –0.293 3.129 3.151 –1.640 2.153 1.377 Note: BLUPs indicate departures from the population average based on equation (3). Predictive models of MoE showed that AV was a better predictor than SG (Figure 2). The MoE model based on AV explained 47 per cent of the observed variation. Moreover, when both AV and SG were used to predict MoE, the parameter estimate for SG became non-significant (Table 6). Similarly, AV was a better predictor than SG for MoR, and combining these two predictors significantly improved predictive performance by explaining 87 per cent of the variation in MoR. Figure 2 View largeDownload slide Relationships of modulus of elasticity (MoE) and modulus of rupture (MoR) with specific gravity (SG) and acoustic velocity (AV) for tested lumber boards (n = 96). Figure 2 View largeDownload slide Relationships of modulus of elasticity (MoE) and modulus of rupture (MoR) with specific gravity (SG) and acoustic velocity (AV) for tested lumber boards (n = 96). Table 6 Parameter estimates and fit statistics for lumber models of modulus of elasticity (MoE) and modulus of rupture (MoR) as a function of lumber specific gravity (SG) and acoustic velocity (AV) Parameter MoE (SG) MoE (AV) MoE (SG, AV) MoR (SG) MoR (AV) MoR (SG, AV) β0 –1.7691 (ns) –10.659 –11.052 –24.948 –46.896 –64.358 β1 (SG) 25.076 1.628 (ns) 117.456 71.978 β2 (AV) 4.466 4.383 22.860 19.217 Residual σ 2.344 1.889 1.897 7.075 4.081 3.493 R2 0.116 0.474 0.469 0.427 0.811 0.872 Parameter MoE (SG) MoE (AV) MoE (SG, AV) MoR (SG) MoR (AV) MoR (SG, AV) β0 –1.7691 (ns) –10.659 –11.052 –24.948 –46.896 –64.358 β1 (SG) 25.076 1.628 (ns) 117.456 71.978 β2 (AV) 4.466 4.383 22.860 19.217 Residual σ 2.344 1.889 1.897 7.075 4.081 3.493 R2 0.116 0.474 0.469 0.427 0.811 0.872 Note: parameter estimates are based on equation (4); significance level α = 0.05, unless otherwise noted (ns); R2 values are calculated based on the Pearson correlation between observed and predicted values for fixed effects only. Dummy variables in equation (4) allowed for the inclusion of either SG (1,0), AV (0,1) or both SG and AV (1,1). View Large Table 6 Parameter estimates and fit statistics for lumber models of modulus of elasticity (MoE) and modulus of rupture (MoR) as a function of lumber specific gravity (SG) and acoustic velocity (AV) Parameter MoE (SG) MoE (AV) MoE (SG, AV) MoR (SG) MoR (AV) MoR (SG, AV) β0 –1.7691 (ns) –10.659 –11.052 –24.948 –46.896 –64.358 β1 (SG) 25.076 1.628 (ns) 117.456 71.978 β2 (AV) 4.466 4.383 22.860 19.217 Residual σ 2.344 1.889 1.897 7.075 4.081 3.493 R2 0.116 0.474 0.469 0.427 0.811 0.872 Parameter MoE (SG) MoE (AV) MoE (SG, AV) MoR (SG) MoR (AV) MoR (SG, AV) β0 –1.7691 (ns) –10.659 –11.052 –24.948 –46.896 –64.358 β1 (SG) 25.076 1.628 (ns) 117.456 71.978 β2 (AV) 4.466 4.383 22.860 19.217 Residual σ 2.344 1.889 1.897 7.075 4.081 3.493 R2 0.116 0.474 0.469 0.427 0.811 0.872 Note: parameter estimates are based on equation (4); significance level α = 0.05, unless otherwise noted (ns); R2 values are calculated based on the Pearson correlation between observed and predicted values for fixed effects only. Dummy variables in equation (4) allowed for the inclusion of either SG (1,0), AV (0,1) or both SG and AV (1,1). View Large Discussion In this study, we examined the variation of lumber properties for 12 full-sib coastal Douglas-fir families in a first-generation progeny trial in British Columbia, Canada. The observed height growth differences between the low gain families and mid and high gain families were 4.6 and 6.6 per cent (Table 1). This is very close to the expected differences as we expected height growth gains to be ~3.6 and 7.7 per cent given the Lambeth’s age–age correlation (r = 0.7 for selection age 12 and forecast to age 35). Tree breeding has the potential to increase volume growth in the range of 10–25 per cent (Jansson et al., 2017), and by doing so to improve the profitability of forest operations and shorten commercial rotations. Historically, the evaluation of lumber properties has relied on small clear specimens owing to the ease of obtaining and testing, despite generally mediocre correlations with lumber properties (Butler et al., 2016). Whenever feasible, it is preferable to rely on testing actual lumber specimens. Our results indicate that higher growth may lead to lower wood properties. In addition, radial variation showed to be an important factor as indicated by lower values for wood properties of edge-grain boards compared with flat-grain boards. On one hand, this points to technological advances that have the potential to customize sawing to optimize and differentiate higher-quality products; on the other hand, with edge-grain boards containing a larger proportion of juvenile wood, it emphasizes possible issues arising from shorter rotations as an indirect result of selecting fast growing planting stock. Given that wood properties of Douglas-fir tend to improve with the transition from juvenile into mature wood (Filipescu et al., 2014), silvicultural practices may have the ability to overcome such pitfalls by planting higher-density stands, followed by thinning to accelerate diameter growth once crowns have lifted (Auty et al., 2016). This strategy would be particularly well suited to Douglas-fir that has the ability to grow vigorously at advanced ages (Curtis, 1992). Planting rapidly growing selected stock may lead to concerns over shorter rotations and by extension harvesting of younger trees containing higher proportions of juvenile wood. Trees in our study were harvested at age 33, relatively younger than the typical commercial rotations (40–60 years) for coastal Douglas-fir in British Columbia. However, our results were encouraging as averages of SG (0.48), MoE (10.2 GPa) and MoR (60.1 MPa) were comparable with reference values for Douglas-fir (Evans and Green, 1988; Dahlen et al., 2012). While the initial planting density (1100 trees per ha) of this progeny trial was similar to operational planting densities, there were several factors that were specific to our study. Trees were pruned at age 15; therefore the influence of knots on lumber properties was limited, but in the absence of pruning, stand density management may be useful by maintaining denser stands in the initial stages of development (Tong et al., 2013; Lowell et al., 2014). The site was located in a river bed, thus it was not characterized by summer moisture deficit that typically limits the formation of denser and stronger latewood (Kantavichai et al., 2010), although silviculture regimes may alleviate moisture deficits by thinning (Diaconu et al., 2015). Moreover, as an additional limitation, it should be noted that lumber in our study was obtained from the lower 3 m log, and stiffness typically declines with log position in the stem (Todoroki et al., 2012). Larger samples covering multiple sites will increase the inference in future studies. Testing of 2 × 4 boards in our study indicated that families of lower genetic gain tend to produce lumber with better physico-mechanical properties, while certain families exhibited both superior growth and wood traits. The existence of families with good growth and wood quality is encouraging and findings are in agreement with previous studies (i.e. Beaulieu et al., 2006) that found a strong negative genetic correlation between stem volume and lumber stiffness, but an absence of significant correlation at phenotypic level for white spruce (Picea glauca), which would allow for selective propagation without compromising growth potential. Similarly, the existence of selected genotypes with superior growth in response to intensive silviculture and no subsequent reduction in stiffness has also been observed in loblolly pine (Roth et al., 2007). Testing of lumber properties to evaluate genetically improved families is possible, as shown by findings of our study; however, this can be challenging as lumber traits are expressed only in later stages of development and lumber testing can be costly and time consuming. In this regard, breeders attempt to reduce the length of selection cycles by utilizing quick and inexpensive non-destructive testing at early ages (Chauhan et al., 2013) or by means of genomic selection (Isik, 2014). Nevertheless, studies of lumber properties from genetically improved families are necessary to provide validation data and baseline information. Increasing productivity and shortening rotations are some of the immediate obvious benefits in using genetically selected planting stock. In addition, anecdotal evidence suggests that rapid initial height growth may also alleviate needs for vegetation management, which would reduce overall regeneration costs. Families in the top-level of genetic gain tested in our study had, on average, an 11 per cent gain, while current operational planting typically uses seed with 15–20 per cent gain; therefore some caution is warranted when generalizing lumber testing results from our study. Furthermore, recent studies have documented that stem defects tend to be more frequent for Douglas-fir grown on productive sites closer to coastal areas in the PNW (Magalska and Howe, 2014), highlighting the need to include stem defects as a secondary selection criterion in breeding programmes. Climate change has the additional potential to influence SG in Douglas-fir (Stoehr et al., 2009) and growth through phenological patterns in the PNW (Ford et al., 2016), as well as increasing the impact of pathogens, such as Swiss needle cast (Phaeocryptopus gaeumannii), that influences growth (Maguire et al., 2011) and indirectly wood quality (Johnson et al., 2005). Models of MoE in our study showed that AV was the main explanatory variable, while SG played only a minor role. ‘Our results also indicated a stronger relationship between MoR and AV (a surrogate for MFA) than between MoE and AV, contrary to what has been reported in previous studies (e.g. McLean et al., 2016; Butler et al., 2017)’.This could be explained by SG having already been included in the criteria of selection of families for growth potential; at the same time, it is possible that the relative young age of trees in our study may have been a factor. Both potential explanations, i.e. the use of SG as criterion of selection and the relative young age of trees in our study, may reduce the overall variability compared with lumber from older trees in natural populations. A better understanding of underlying factors will require further research. In this regard, given shifting structural and functional requirements of trees with age (Lachenbruch et al., 2011), the relative importance of density and AV as determinants of stiffness may vary over time. Both density and MFA were important for predicting dynamic MoE of Pinus patula boards (Wessels et al., 2015) and of small clears of Scots pine (Auty et al., 2016), while density was more important than MFA for stiffness of Douglas-fir mature wood (Lachenbruch et al., 2010). Nevertheless, AV measured on boards of radiata pine was strongly correlated with stiffness (Downes et al., 2002), while the importance of density as a selection criterion may have been overemphasized in previous selection programmes (Apiolaza, 2009). Therefore, AV may play an increasing role both in selection of Douglas-fir for stiffness (Cherry et al., 2008; Vikram et al., 2011), as well as product segregation (Matheson et al., 2002; Wang et al., 2007). In addition, our models of MoR showed a more balanced contribution of both SG and AV when used as predictors, in agreement with other studies that have shown SG has a higher correlation to MoR than to MoE in juvenile wood of radiata pine (Ivkovic et al., 2009). In conclusion, the evaluation of lumber characteristics from genetically selected trees offers a valuable opportunity to assess the impact of early selection decisions, as well as improve breeding programmes going forward. Findings from our case study indicate that families with lower growth tend to perform better in regards to physico-mechanical characteristics of lumber; however there are some families that exhibit both higher growth and good lumber properties. Acoustic velocity shows promising potential for future tree breeding, as well as for product segregation, given that acoustic velocity has been a better predictor than specific gravity for models of modulus of elasticity and modulus of rupture. Evidence from this work will inform development of policies supporting sustainable forest management decisions to optimize wood production and product quality. Acknowledgements Authors express their appreciation to numerous people and organizations that made this study possible. The Nanaimo Park District facilitated site access and tree sampling. Field support was provided by Keith Bird, Lisa Hayton, Tom Bown, Kristina Beckmann, and Ross Koppenaal. Ciprian Pirvu facilitated lumber testing at FPInnovations Vancouver. Funding Funding was provided by the Land Based Investment Strategy (Government of British Columbia) and the Forest Innovation Program (Government of Canada). Conflict of interest statement None declared. References Apiolaza , L.A. 2009 Very early selection for solid wood quality: screening for early winners . Ann. For. Sci. 66 ( 601 ), 1 – 10 . ASTM American Society for Testing and Materials . 2013 . Standard test methods for mechanical properties of lumber and wood-base structural material, D4761-13. Aubry , C.A. , Adams , W.T. and Fahey , T.D. 1998 Determination of relative economic weights for multitrait selection in coastal Douglas-fir . Can. J. For. Res. 28 , 1164 – 1170 . Google Scholar CrossRef Search ADS Auty , D. , Achim , A. , Macdonald , E. , Cameron , A.D. and Gardiner , B.A. 2016 Models for predicting clearwood mechanical properties of Scots pine . For. Sci. 62 , 403 – 413 . BC MFLNRO . 2017 . Growth and Yield Modelling. TIPSY. Ministry of Forests, Lands and Natural Resource Operations. https://www.for.gov.bc.ca/hts/growth/tipsy/tipsy_description.html (last accessed May 25, 2017) Beaulieu , J. , Zhang , S.Y. , Yu , Q. and Rainville , A. 2006 Comparison between genetic and environmental influences on lumber bending properties in young white spruce . Wood Fiber Sci. 38 , 553 – 564 . Blackburn , D. , Hamilton , M. , Harwood , C. , Innes , T. , Potts , B. and Williams , D. 2010 Stiffness and checking of Eucalyptus nitens sawn boards: genetic variation and potential for genetic improvement . Tree Gen. Gen. 6 , 757 – 765 . Google Scholar CrossRef Search ADS Briggs , D.G. and Smith , W.R. 1986 Effects of silvicultural practices on wood properties – a review. In Douglas-fir: Stand Management for the Future . Oliver C.D. , Hanley D.P. and Johnson J.A. (eds). College of Forest Resources, University of Washington , pp. 108 – 117 . Contribution No. 55. Burdon , R.D. 2010 Wood properties and genetic improvement of radiata pine . N. Z. J. For. 55 , 22 – 27 . Butler , M.A. , Dahlen , J. , Antony , F. , Kane , M. , Eberhardt , T.L. , Jin , H. , et al. 2016 Relationships between loblolly pine small clear specimens and dimension lumber tested in static bending . Wood Fiber Sci. 48 , 81 – 95 . Butler , M.A. , Dahlen , J. , Eberhardt , T.L. , Montes , C. , Antony , F. and Daniels , R.F. 2017 Acoustic evaluation of loblolly pine tree- and lumber-length logs allows for segregation of lumber modulus of elasticity, not for modulus of rupture . Ann. For. Sci. 74 , 20 . Google Scholar CrossRef Search ADS Chauhan , S.S. , Sharma , M. , Thomas , J. , Apiolaza , L.A. , Collings , D.A. and Walker , J.C.F. 2013 Methods for the very early selection of Pinus radiata D. Don. for solid wood products . Ann. For. Sci. 70 , 439 – 449 . Google Scholar CrossRef Search ADS Chen , Z.Q. , Karlsson , B. , Lundquist , S.O. , Gil. , M.R.G. , Olsson , L. and Wu , H.X. 2015 Estimating solid wood properties using Pilodyn and acoustic velocity on standing trees of Norway spruce . Ann. For. Sci. 72 , 499 – 508 . Google Scholar CrossRef Search ADS Cherry , M.L. , Vikram , V. , Briggs , D. , Cress , D.W. and Howe , G.T. 2008 Genetic variation in direct and indirect measures of wood stiffness in coastal Douglas-fir . Can. J. For. Res. 38 , 2476 – 2486 . Google Scholar CrossRef Search ADS Cumbie , W.P. , Isik , F. and McKeand , S.E. 2012 Genetic improvement of sawtimber potential in loblolly pine . For. Sci. 58 , 168 – 177 . Curtis , R.O. 1992 A new look at an old question – Douglas-fir culmination age . West. J. Appl. For. 7 , 97 – 99 . Dahlen , J. , Jones , P.D. , Deale , R.D. and Shmulsky , R. 2012 Bending strength and stiffness of in-grade Douglas-fir and southern pine No. 2 2 × 4 lumber . Can. J. For. Res. 42 , 858 – 867 . Google Scholar CrossRef Search ADS Diaconu , D. , Kahle , H.P. and Spiecker , H. 2015 Tree- and stand-level thinning effects on growth of European beech (Fagus sylvatica L.) on a northeast- and a southwest-facing slope in southwest Germany . Forests 6 , 3256 – 3277 . Google Scholar CrossRef Search ADS 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 El-Kassaby , Y.A. , Mansfield , S. , Iisk , F. and Stoehr , M. 2011 In situ wood quality assessment in Douglas-fir . Tree Genet. Gen. 7 , 553 – 561 . Google Scholar CrossRef Search ADS Evans , J.W. and Green , D.W. 1988 Mechanical Properties of Visually Graded Dimension Lumber: Vol. 2 Douglas Fir-Larch . National Technical Information Service , 476 . Publication number PB-88-159-397. Evison , D.C. and Apiolaza , L.A. 2015 Incorporating economic weights into radiata pine breeding selection decisions . Can. J. For. Res. 45 , 135 – 140 . Google Scholar CrossRef Search ADS 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 Ford , K.R. , Harrington , C.A. , Bansal , S. , Gould , P.J. and St. Clair , J.B. 2016 . Will changes in phenology track climate change? A study of growth initiation timing in coast Douglas-fir . Glob Change Biol 22 : 3712 – 3723 . Gray , L.K. , Rweyongeza , D. , Hamann , A. , John , S. and Thomas , B.R. 2016 Developing management strategies for tree improvement programs under climate change: Insights gained from long-term field trials with lodgepole pine . For. Ecol. Manage. 377 , 128 – 138 . Google Scholar CrossRef Search ADS Green , R.N. and Klinka , K. 1994 A field guide to site identification for the Vancouver Forest Region. In Land Management Handbook 28 . Province of BC, Ministry of Forests . Howe , G.T. , Jayawickrama , K. , Cherry , M. , Johnson , G.R. and Wheeler. , N.C. 2006 Breeding Douglas-fir . Plant Breed. Rev. 27 , 245 – 353 . Hubert , J. and Lee , S. 2005 A review of the relative roles of silviculture and tree breeding in tree improvement: the example of Sitka spruce in Britain and possible lessons for hardwood breeding . Forestry 78 , 109 – 120 . Google Scholar CrossRef Search ADS Isik , F. 2014 Genomic selection in forest tree breeding: the concept and an outlook to the future . New For. 45 , 379 – 401 . Google Scholar CrossRef Search ADS Ivkovic , M. , Gapare , W.J. , Abarquez , A. , Ilic , J. , Powell , M.B. and Wu , H.X. 2009 Prediction of wood stiffness, strength, and shrinkage in juvenile wood of radiata pine . Wood Sci. Technol. 43 , 237 – 257 . Google Scholar CrossRef Search ADS Ivkovic , M. , Wu , H.X. , McRae , T.A. and Powell , M.B. 2006 Developing breeding objectives for radiata pine structural wood production. I. Bioeconomic model and economic weights . Can. J. For. Res. 36 , 2920 – 2931 . Google Scholar CrossRef Search ADS Jansson , G. , Hansen , J.K. , Haapanen , M. , Kvaalen , H. and Steffenrem , A. 2017 The genetic and economic gains from forest tree breeding programmes in Scandinavia and Finland . Scand. J. For. Res . 32 , 273 – 286 . Google Scholar CrossRef Search ADS Johnson , G.R. , Grotta , A.T. , Gartner , B.L. and Downes , G. 2005 Impact of the foliar pathogen Swiss needle cast on wood quality of Douglas-fir . Can. J. For. Res. 35 , 331 – 339 . Google Scholar CrossRef Search ADS Kantavichai , R. , Briggs , D. and Turnblom , E. 2010 Modeling effects of soil, climate, and silviculture on growth ring specific gravity of Douglas-fir on a drought-prone site in western Washington . For. Ecol. Manage. 259 , 1085 – 1092 . Google Scholar CrossRef Search ADS Lachenbruch , B. , Johnson , G.R. , Downes , G.M. and Evans , R. 2010 Relationships of density, microfibril angle, and sound velocity with stiffness and strength in mature wood of Douglas-fir . Can. J. For. Res. 40 , 55 – 64 . Google Scholar CrossRef Search ADS Lachenbruch , B. , Moore , J.R. and Evans , R. 2011 Radial variation in wood structure and function in woody plants, and hypotheses for its occurrence. In Age- and Size-related Changes in Tree Structure and Function . Meinzer F.C. , Lachenbruch B. and Dawson T.E. (eds). Springer , pp. 121 – 164 . Google Scholar CrossRef Search ADS Lambeth , C.C. 1980 Juvenile-mature correlations in Pinaceae and implications for early selection . For. Sci. 26 , 571 – 580 . Lowell , E.C. , Maguire , D.A. , Briggs , D.G. , Turnblom , E.C. , Jayawickrama , K.J.S. and Bryce , J. 2014 Effects of silviculture and genetics on branch/knot attributes of coastal Pacific Northwest Douglas-fir and implications for wood quality – a synthesis . Forests 5 , 1717 – 1736 . Google Scholar CrossRef Search ADS Magalska , L. and Howe , G.T. 2014 Genetic and environmental control of Douglas-fir stem defects . For. Ecol. Manage. 318 , 228 – 238 . Google Scholar CrossRef Search ADS Maguire , D.A. , Mainwaring , D.B. and Kanaskie , A. 2011 Ten-year growth and mortality in young Douglas-fir stands experiencing a range in Swiss needle cast severity . Can. J. For. Res. 41 , 2064 – 2076 . Google Scholar CrossRef Search ADS Matheson , A.C. , Dickson , R.L. , Spencer , D.J. , Joe , B. and Ilic , J. 2002 Acoustic segregation of Pinus radiata logs according to stiffness . Ann. For. Sci. 59 , 471 – 477 . Google Scholar CrossRef Search ADS McLean , J.P. , Moore , J.R. , Gardiner , B.A. , Lee , S.J. , Mochan , S.J. and Jarvis , M.C. 2016 Variation of radial wood properties from genetically improved Sitka spruce growing in the UK . Forestry 89 , 109 – 116 . Google Scholar CrossRef Search ADS R Development Core Team 2016 R: A Language and Environment for Statistical Computing . R Foundation for Statistical Computing , Vienna, Austria. Available at: https://www.R-project.org/. 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 Steffenrem , A. , Solheim , H. and Skroppa , T. 2016 Genetic parameters for wood quality traits and resistance to the pathogens Heterobasidion parviporum and Endoconidiophora polonica in a Norway spruce breeding population . Eur. J. For. Res. 135 , 815 – 825 . Google Scholar CrossRef Search ADS Stoehr , M.U. , Ukrainetz , N.K. , Hayton , L.K. and Yanchuk , A.D. 2009 Current and future trends in juvenile wood density for coastal Douglas-fir . Can. J. For. Res. 39 , 1415 – 1419 . Google Scholar CrossRef Search ADS Stoehr , M. , Yanchuk , A. , Xie , C.-Y. and Sanchez , L. 2008 Gain and diversity in advanced generation coastal Douglas-fir selections for seed production populations . Tree Genet. Gen. 4 , 193 – 200 . Google Scholar CrossRef Search ADS Todoroki , C.L. , Lowell , E.C. , Dykstra , D.P. and Briggs , D.G. 2012 Maps and models of density and stiffness within individual Douglas-fir trees . N. Z. J. For. Sci. 42 , 1 – 13 . Tong , Q.J. , Duchesne , I. , Belley , D. , Beaudoin , M. and Swift , E. 2013 Characterization of knots in plantation white spruce . Wood Fibre Sci. 45 , 84 – 97 . 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 Wang , X. , Carter , P. , Ross , R.J. and Brashaw , B.K. 2007 Acoustic assessment of wood quality of raw forest materials – a path to increased profitability . For. Prod. J. 57 , 6 – 14 . Wessels , C.B. , Malan , F.S. , Kidd , M. and Rypstra , T. 2015 The variation of microfibril angle in South African grown Pinus patula and its influence on the stiffness of structural lumber . South. For. 77 , 213 – 219 . Whiteside , I.D. , Wilcox , M.D. and Tustin , J.R. 1976 New Zealand Douglas fir timber quality in relation to silviculture . N. Z. J. For. 22 , 24 – 44 . Woods , J.H. 2016 . Forest Genetics Council of BC Business Plan 2016/17. Available at: http://www.fgcouncil.bc.ca/ Xie , C.-Y. and Yanchuk , A.D. 2003 Breeding values of parental trees, genetic worth of seed orchard seedlots and yields of improved stocks in British Columbia . W. J. Appl. For. 18 ( 2 ), 88 – 100 . © 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 28, 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