Wood mechanical properties and discoloured heartwood proportion in sugar maple and yellow birch grown in New Brunswick

Duchesne, I., Vincent, M., Wang, X., Ung, C. H., and Swift, D. E. (2016). "Wood mechanical properties and discoloured heartwood proportion in sugar maple and yellow birch grown in New Brunswick," BioRes. 11(1), 2007-2019.

Abstract

Rising interest in using wood in non-residential multi-story building structures opens up new opportunities for utilising low-grade hardwoods. The primary objective of this study was to evaluate the geographic variation in modulus of elasticity (MOE) and modulus of rupture (MOR) of sugar maple and yellow birch wood in relation to stand and tree characteristics for two regions in New Brunswick, Canada. Mixed effects statistical models were developed to test the effects of stand, tree, and wood sample variables. A second objective was to examine geographic variation in heartwood discolouration in relation to stand and tree characteristics. Between-tree differences (trees nested within sites) accounted for 44% and 35% of the total variation in yellow birch (MOE and MOR, respectively) and for 69% and 60% of total variation in sugar maple. The fixed effects explained only a very small part for the variation in MOE and MOR in the sugar maple data (10% for MOE and 5% for MOR). For sugar maple, mechanical properties (MOE and MOR) at 50% of the radius were considerably lower than those close to the bark, but this radial variation was not noteworthy for yellow birch. Discoloured heartwood proportion had no significant effect on wood mechanical properties.

Download PDF

Full Article

Wood Mechanical Properties and Discoloured Heartwood Proportion in Sugar Maple and Yellow Birch Grown in New Brunswick

Isabelle Duchesne,^a,* Manon Vincent,^b Xiaodong (Alice) Wang,^c, Chhun-Huor Ung,^a and D. Edwin Swift ^d

Keywords: Low-grade hardwoods; End-use suitability; MOE; MOR; Sugar maple; Yellow birch; Statistical models; Geographic variation

Contact information: a: Natural Resources Canada, Canadian Wood Fibre Centre, P.O. Box 10380, Québec, QC, G1V 4C7, Canada; b: Ministère des Forêts, de la Faune et des Parcs du Québec, 5700, 4e Avenue Ouest, Québec QC, G1H 6R1, Canada; c: Wood Technology and Engineering, Luleå University of Technology, Forskargatan 1, SE-931 87 Skellefteå, Sweden (Former postdoc at Canadian Wood Fibre Centre, Québec, Canada); d: Natural Resources Canada, Canadian Wood Fibre Centre, P.O. Box 4000, Fredericton, NB, E3B 5P7, Canada; *Corresponding author: isabelle.duchesne@canada.ca

INTRODUCTION

Hardwood sawmills generate direct economic outcomes similar to the softwood industry, but at a smaller production scale (Trudelle et al. 2009). Over the past decade, the eastern Canadian hardwood lumber industry has been less competitive because of the limited availability of high-quality hardwoods combined with a particularly difficult economic situation (FPInnovations 2014). The decline in the overall quality of northern hardwood forests has been attributed to repeated selective harvest (high grading) practices of the past centuries.

In eastern Canada and northeastern United States, sugar maple (Acer saccharum Marsh.) and yellow birch (Betula alleghaniensis Britt.) are among the most important commercial hardwood species and have been typically used for the manufacturing of furniture, cabinetry, millwork, and flooring (Mullins and McKnight 1981). These appearance-based products generally require high-quality clear wood cuttings with a uniform colour that are free of visual defects on one or two board surfaces, depending on specific end-uses (Pearson 2009; McDonald 2010). In degraded stands, wood discolouration can reduce grade and volume yields and cause major value loss (Ohman 1968). However, although the availability of high-quality hardwoods has declined, the high density of these species makes them suitable for a range of structural applications. Indeed, in recent years, the interest in using hardwoods in structural applications has increased (Dill-Langer and Aicher 2014; Gong et al. 2015). Efforts have been made to add value to hardwood blocks by using them in engineered structural wood products (Verreault 2000). Hardwoods have been used for the production of structural plywood or glued-laminated materials, such as truck bedding (Sellers et al. 1988) and laminated wood railway ties (Gong et al. 2013). When stands contain small-diameter trees of marginal value for traditional uses, the lower-grade hardwood resources can be used in composite products, such as fibreboard, particleboard, and flakeboard, where product quality is not a direct function of stem quality (Sellers et al. 1988). The emergence of multi-story wood-frame buildings in Canada and European countries may also provide an opportunity to explore the potential use of lower-grade hardwoods in new structural applications such as in high-performance timber columns (Aicher et al. 2014; Candelario 2015). To this end, it is important to increase our knowledge of hardwood properties to ensure uniform or predictable quality of products from our forests. Compared with softwoods, hardwoods have received little attention in terms of characterising its wood fibre attributes such as stiffness and strength. Among the few studies available, Jessome (2000) and Kretschmann (2010) characterised the strength properties of sugar maple and yellow birch from eastern Canada and northeastern United States, respectively. However, mechanical property studies on hardwoods provide often limited or no information about stand growth conditions. Consequently, there is still a gap in our understanding of the relationships between wood fibre characteristics and stand growth conditions (Duchesne and Letarte 2013).

The objective of this study was to evaluate the geographic variation in modulus of elasticity (MOE) and modulus of rupture (MOR), and discoloured heartwood proportion (HW) of sugar maple and yellow birch wood in relation to stand and tree characteristics for two ecoregions of the province of New Brunswick, Canada. Mixed effects statistical models are developed in this study to test the effects of stand, tree, and wood sample variables on hardwood MOE and MOR.

EXPERIMENTAL

Stand and Tree Measurements

In the autumn of 2009, eight sites were selected in two ecoregions in New Brunswick (NB): the Central Uplands near Fredericton, and the Northern Uplands northeast of Edmundston (Table 1, Zelazny et al. 1989). Located in the northern hardwoods of the Acadian Forest Region (Rowe 1972), the study sites consisted primarily of various proportions of sugar maple (Acer saccharum Marsh.), yellow birch, red maple (Acer rubrum L.), and American beech (Fagus grandifolia Ehrh.), which are very typical of the hardwood resource in New Brunswick. All stands were unevenly aged and mature and had had some partial harvesting (most recent cuts occurred 3 to 15 years before tree harvest). Because eastern North America has a long history of silvicultural activities, selective harvesting has been practiced over the last 100 years (Swift et al. 2013). At each site, two 11.28-m radius circular plots (400 m²) were laid, in which the dominant tree height and the diameter at breast height (DBH) of each live merchantable tree larger than 9 cm were measured. Basal area (G) was averaged from the data of the two plots (Table 1). Thereafter, five sample trees were randomly selected per temporary plot, for a total of 10 trees per site. The 80 trees were harvested for analysis of lumber product volume and grade recoveries (Duchesne et al. 2012). However, for the modelling of mechanical properties, 18 trees could not be used because of internal defects (e.g., checks, decay, and knots) and/or incomplete stand- or tree-level data, for a total number of 62 trees (Table 2). Sample tree DBH ranged between 24 and 46 cm for both species (Table 3). The measured tree attributes were stem diameter at 1.30 m (DBH_cm), total height (H_tot_m) and crown variables, i.e., crown width (WidCr_dm), live crown length (lenCr_m), and crown area (Crown_area_m2). Tree age (Tree age) was also estimated based on the ring count on a disk cut at a stump height of approximately 15 cm above ground (no age correction).

Table 1. Study Site Descriptive Data

Table 2. Number of Trees and Small Clear Specimens by Site and Species; Specimen Location: A) Near the bark; C) 50% of radius

Table 3. Stand and Tree Characteristics

¹Stdev: Standard deviation

Small Clear Specimen Preparation and Testing in Static Bending

A 30-cm bolt was cut after the first sawlog of each sample tree. Because trees were bucked to maximise lumber recovery, the height at which bolts were cut varied between 1.8 m and 7.0 m (3.7 m on average), depending on tree quality. Expressed as the relative tree height, the bolts were extracted on average at 18% of the total tree height (range: 9% to 36%).

Heartwood diameter (HW_Diam_mm and HW_rel_Diam), defined as brown-reddish discoloured wood that is not white sapwood, was measured on the cross section of each bolt (average of north-south and east-west diameters). For each bolt, a slice including the pith was first sawn, from which two 10 mm x 10 mm x 190 mm small clear specimens were extracted: one near the bark in the sapwood (A), the other at 50% of the radius (C). These slices were sawn in the east-west direction, unless a major defect occurred. The 62 sample trees produced a total of 122 defect-free small clear specimens that could be successfully tested (Table 2).

For each specimen tested, the following data were recorded: location (Spec_Loc, A or C), number of rings per specimen (Nb_Ring), mean ring width (MeanRingW_mm), and sample height within the tree (Sample_H). Thereafter, specimens were placed in a conditioning room (20 °C, 65% relative humidity) until they reached an equilibrium moisture content of 12%.

MOE and MOR tests were performed at FPInnovations with an MTS ReNew Upgrade universal testing machine (MTS Headquarters, Eden Prairie, Minnesota, U.S.A.) following the ASTM-D-143-94 (ASTM 2007) standard test method for small clear specimens. Specimens were placed with growth rings horizontal and tested at 12% moisture content using a span of 140 mm. Basic density (oven-dry wood weight/green volume) of each specimen was measured according to ASTM-D-2395-07 (ASTM 2009). Table 4 is a summary of specimen characteristics.

Table 4. Small Clear Wood Specimen Characteristics and Average Mechanical Properties

¹ Measured on the transversal section of each 30-cm bolt (before cutting the two specimens)

² Measured on each specimen

Model Development for MOE and MOR

The dataset had a hierarchical structure, implying interdependence of observations. Specifically, MOE and MOR measurements were nested within trees, which were nested within sites. Mixed linear models were thus used to investigate variations within as well as among trees and sites (Brown and Prescott 2006). Random site and tree effects were included in the models to allow parameter estimates to vary around the population mean at the level of each grouping factor. Normality of variables was verified graphically and data transformation (centering) applied when needed.

Using the variables listed in Table 5, mixed models were developed in successive steps. First, three groups of models were built to described MOE and MOR variations with 1) tree, 2) stand, and 3) sample and wood attribute characteristics, then global models were built, which accounts for the joint effects of the tree and stand, tree and sample, tree, sample and stand characteristics on MOE and MOR variations. Interaction terms between Spec_Loc X Sample_H_m, Nb_ring_Spec X Sample_H_m, HW_Diam_mm X Spec_Loc, and H_tot_m X Sample_H_m were assumed to incorporate the effect of juvenile wood on mechanical properties. All a priori multilevel linear models were then compared to identify the main factors related to MOE and MOR variations in both yellow birch and sugar maple. Model selection was performed using the AICcmodavg package in R (Mazerolle 2012). This led to uncertainties regarding the selection of the best model to be assessed using a model averaging technique (also referred to as “multimodel inference”). The package computes the weighted estimates of the predictions for a given predictor variable across all models. The weighting of parameter estimates is given by the model probabilities, which are derived from Akaike’s weights (Mazerolle 2006). Normality of residuals was verified graphically, and the multicollinearity between data and the distribution of residuals vs. predicted values was verified using a variance inflation factor (VIF).

Table 5. List of Explanatory Variables Tested for Modelling MOE and MOR

RESULTS AND DISCUSSION

Hardwood Mechanical Property Models

Several candidate models were developed and compared for each species. For yellow birch MOR, five models appeared equivalent among all the models tested (Akaike weight of first models lower than 0.90, and delta AICc (_i)< 2, Mazerolle 2006, data not shown). Three models included sample and tree variables, and two more complex models included interactions between Sample_H_m and other variables. The best model for predicting birch MOE included only sample and tree variables. For maple MOE and MOR, a dozen models appeared equivalent. All these models included only sample and tree variables for MOE, sample variables alone, and sample and tree variables for MOR. Using models averaging and the 95% confidence intervals (CI), it appears that only Nb_ring_Spec (CI= 261.84, 707.5), Tree_age (CI= -39.24, -4.07), and LenCr_m (CI= -756.38, -138.74) showed strong evidence of having a significant influence (effect ≠0) in yellow birch MOE variations. In birch MOR models, Nb_ring_Spec (CI= 1.06, 5.19) and Tree_age (CI= -0.39, -0.08) had a significant effect on MOR variations. Spec_Loc (CI= -1900.1, -745.61) was the only sample variable having a significant influence on sugar maple MOE while Spec_Loc (CI=-10.83, -1.77) and Nb_ring_Spec (CI= 0.27, 2.82) had a significant effect on sugar maple MOR variations. The fixed effects parameters estimates, standard deviation (Stdev) and summary statistics of final models are presented (Table 6). Thus, the final equations for models are:

Sugar maple MOE_ijk = α₀ + a_i + b_j(i) + α₄Spec_Loc + e_ijk (1)

Sugar maple MOR_ijk = α₀ + a_i + b_j(i) + α₂Nb_ring_Spec + α₄Spec_Loc + e_ijk (2)

Yellow birch MOE_ijk = α₀ + a_i + b_j(i) + α₁Nb_ring_Spec + α₂Tree_age +α₃LenCr_m + e_ijk(3)

Yellow birch MOR_ijk = α₀ + a_i + b_j(i) + α₁Nb_ring_Spec + α₂Tree_age + e_ijk (4)

where α₀, α₁, α₂ and α₃ are the fixed effects parameters. The random elements a, b, and e are assumed to be independent and normally distributed. a_i denotes the site random effect b_j(i)the tree nested in site random effect and e_ijk the within group error.

Table 6. Parameters Estimates, Standard Errors (SE), Variance Component (Stdev), and Summary Statistics of Final Models

Results showed that between-site differences represented a negligible portion of the total variation for yellow birch mechanical properties in this study, while it accounted for about 20% of the total variation for sugar maple. However, between-tree differences (tree nested within sites) accounted for 44% and 35% of the total variation in yellow birch (MOE and MOR respectively) and for 69% and 60% of total variation in sugar maple. The fixed effects explained only a very small part for the variation in MOE and MOR in the sugar maple data (10% for MOE and 5% for MOR). The HW_Diam_mm effect and its interaction with Spec_Loc were tested, and these were not significant either in sugar maple or in yellow birch. Yellow birch age and crown length significantly affected MOE and MOR, but this was not the case for maple. Yellow birch MOE and MOR appeared significantly lower in older trees and birch MOE was lower in trees with long live crowns (Fig. 1). Older birch trees may lack vigour (senescence) and grow wood with lower mechanical properties. Jelonek et al. (2015) studied the effect of tree senescence on the properties of wood tissues in Scots pine (Pinus sylvestris L.). They found a marked decline in basic density and MOE after the tree age exceeded approximately 75 to 80 years, indicating a dynamic ageing process expressed in the gradual deterioration of wood tissue properties. In the North Shore region of Québec, Torquato et al.(2014) found that MOE and MOR in black spruce were higher in samples from stands of regular structure (mean tree age: 101 years) compared with that of very old stands of irregular structure (mean tree age: 165 years). It can be hypothesized that wood formation in maple and birch may follow a senescence pattern similar to that of softwoods with tree age. No information on the effects of tree senescence on hardwood properties was found, perhaps because the primary use of hardwoods is for appearance-based products.

On the other hand, long crowns of dominant trees may also confer lower mechanical properties when growth rates are excessive in relation to the normal growth of the species. Thus, it seems that we observe two interacting, complex mechanisms regulating birch wood formation and properties: one related to tree age and the other to growth rate.

Fig. 1. Predicted MOE and MOR variations with tree age and crown length in yellow birch. Lines are the predicted values using fixed effects parameters only. Average crown length (13 m) was chosen within the equation to show fitted MOE and MOR variations with age, whereas average tree age (112 years old) was chosen within the equation to show fitted MOE and MOR variations with crown length.

Geographic Variation in Mechanical Properties

Small clear specimen average MOE and MOR values are shown in Table 4. Compared with data in the literature (Jessome 2000), the sugar maple average value was 24% lower for MOE (14100 MPa vs. 10684 MPa in this study) but similar for MOR (115.0 vs. 113.2 MPa). For yellow birch, MOE was 22% lower (14100 MPa vs. 10954 MPa in this study) and similar for MOR (106.0 vs. 106.5 MPa in this study). A simple Pearson correlation analysis indicated a strong positive link between MOE and MOR for both species (r = 0.87 for sugar maple and 0.85 for yellow birch).

MOE and MOR variation between sites appeared larger for sugar maple, while birch MOE and MOR appear relatively homogeneous between sites (Fig. 2). These results should be regarded with caution because of the very limited number of sample trees analysed at each site, especially for birch (Table 2). As shown in the previous modelling section, site variables had no effect on MOE and MOR. There was no statistically significant difference in MOE and MOR between the ecoregions of Northern and Central Uplands for both species, suggesting that growth conditions within the Acadian Forest region were comparable, or did not vary to the point of inducing notable changes in MOE and MOR. The relationship between mechanical properties and specimen basic density was similar for the two species: MOE stayed more or less constant throughout the range of wood densities measured while MOR tended to slightly increase with increasing density (but the trend was not statistically significant).

The effect of specimen location on MOE and MOR is shown in Fig. 3. For maple, MOE from clearwood located at 50% of the radius (C) was 9961 MPa, and it increased to 11407 MPa near the bark (A) (+14.5%), while for MOR it increased from 110.0 MPa at position C to 116.4 MPa at position A, (+5.8%). For yellow birch, MOE increased from 10471.8 MPa (C) to 11436.0 MPa (A) (+9.2%), and MOR from 104.9 MPa (C) to 108.1 MPa (A) (+3.1%), but this variation, which is related to specimen location, was not statistically significant. For maple, specimens near the bark (A) were stiffer than those at position C at all sites. For birch, four out of six sites had stiffer wood near the bark (A). The two sites that showed the contrary also had the smallest number of samples of all birch sites: P2 (2 trees) and P7 (1 tree). This suggests that birch may follow a similar radial trend as maple, but a more extensive sampling would be needed to verify this.

Geographic Variation in Discoloured Heartwood

In this study, heartwood refers to a darker brown-reddish discolouration of traumatic origin (also called red heartwood) developed in sugar maple and yellow birch wood as a result of tree injuries and invasion by microorganisms (Shigo 1967; Hallaksela and Niemistö 1998; Drouin et al. 2009). These hardwoods do not develop a regular, genetically-programmed “true”-coloured heartwood as in other species (e.g., oaks). Traumatic heartwood discolouration has no effect on wood mechanical properties (Shmulsky and Jones 2011). Similarly, no effect on the mechanical properties of wood as a result of a change from sapwood to regular, “true”-colored heartwood has been found in most species in the United States (USDA 1966), and for white and red oaks in Europe (Merela and Cufar 2013).

Discoloured heartwood proportion varied greatly from site to site and was larger in birch compared with maple (27.6% vs. 20.8%, Fig. 4). It tended to increase with tree age only for birch (Fig. 5). In the province of Québec, red heartwood proportions of 36.4% and 36.8% were reported for sugar maple and yellow birch, respectively (Havreljuk et al. 2013). The proportion of discoloured wood had no significant effect on small clear mechanical properties, which concurs with the literature (Shmulsky and Jones 2011).

Fig. 2. Box-plot of MOE and MOR variations in relation to geographic sites. The New Brunswick Ecoregions of Central Uplands and Northern Uplands are represented by study sites P1 to P4 and P5 to P8, respectively.

Fig. 3. Box-plot of MOE and MOR variations with specimen location (A: near the bark, C: 50% of the radius). Bold black lines show median of the sample. Thin, dotted lines show the average MOE and MOR for each location.

Fig. 4. Heartwood proportion (%) in relation to sites for sugar maple and yellow birch in New Brunswick’s ecoregions of Central (P1 to P4) and Northern (P5 to P8) Uplands

Fig. 5. Heartwood proportion (%) of sugar maple and yellow birch in relation to tree age

CONCLUSIONS

In this study, there was no significant effect of ecoregion on hardwood MOE and MOR. In the models developed, a large amount of MOE and MOR variation was explained by random effects, which means that the explanatory variables tested poorly explained the response variables, especially for sugar maple where fixed effects explained only 10% and 5% of MOE and MOR variation, respectively. These results were due to between-site variation that was more important in sugar maple models compared with yellow birch models.
Clear wood specimens located at 50% of the radius showed a MOE and MOR significantly lower than those located close to the bark for maple, but not for birch. MOR in maple was also slightly affected by the number of rings by specimen (growth rate indicator). In birch, the number of rings per specimen as well as tree age significantly affected MOE and MOR, both of which decreased with tree age. Crown length negatively affected birch MOE.
A positive relationship was observed between tree age and heartwood proportion, only for birch. Heartwood proportion did not have any significant effect on MOE and MOR.
The indication that crown length and age affected internal wood attributes in yellow birch opens up the possibility for forest managers to positively manipulate tree growth conditions to obtain specific internal wood characteristics.

ACKNOWLEDGEMENTS

This study is part of the Eastern Hardwood Research Initiative launched in 2008 by FPInnovations to assist in the maximisation of the hardwood value chain: from forest to end product. This initiative includes the provinces of New Brunswick, Nova Scotia, Quebec, and Ontario, and receives federal government support under the Transformative Technologies Program. The authors are grateful to the New Brunswick Department of Natural Resources for providing initial study site information. A special thanks to Roger Gagné, Francis Tanguay, Luc Bédard, Ghislain Veilleux, Gavin Comeau, Riley DeMerchant, and Brian Williams for collecting field data, and to Pamela Cheers for an editorial review.

REFERENCES CITED

Aicher, S., Christian, Z., and Dill-Langer, G. (2014). “Hardwood glulams – Emerging timber products of superior mechanical properties,” in: Proceedings of the World Conference on Timber Engineering (WCTE 2014), Quebec City, Canada, pg. 10.

ASTM D143-94 (2007). “Standard test methods for small clear specimens of timber,” ASTM International, West Conshohocken, Pennsylvania.

ASTM D2395-07 (2009). “Standard test methods for specific gravity of wood and wood-based materials,” pg. 298–306. ASTM International, West Conshohocken, PA.

Brown, H., and Prescott, R. (2006). Applied Mixed Models in Medicine, 2^nd Ed., John Wiley & Sons Ltd., Chichester, England.

Candelario, J. D. (2015). High performance of timber columns (HPTC) – Literature review, FPInnovations Report No. 301009344, pg. 8.

Dill-Langer, G., and Aicher, S. (2014). “Glulam composed of glued laminated veneer lumber made of beech wood: superior performance in compression loading,” in: Materials and Joints in Timber Structures – Recent Developments of Technology, Aicher, S., Reinhardt, H.-W., and Garrecht, H. (eds.), RILEM Bookseries, Vol. 9, pp. 603-613, ISSN 2211-0844, RILEM, 2014.

Drouin, M., Beauregard, R., and Duchesne, I. (2009). “Variability of wood color in paper birch in Québec,” Wood Fiber Sci. 41(4), 333-345.

Duchesne, I., Ung, C. H., Swift, E. Ferland-Raymond, B., and Giroux, Y. (2012). Hardwood Initiative: Project 16 – Development of a Product Matrix by Tree Grade and DBH Class for Sugar Maple and Yellow Birch, Canadian Wood Fibre Centre, FPInnovations Report.

Duchesne, I., and Letarte, M. (2013). “Les relations entre la sylviculture et les propriétés du bois,” Chapitre dans Ministère des Ressources Naturelles, Le Guide Sylvicole du Québec, Tome 2, Les Concepts et l’Application de la Sylviculture, Les Publications du Québec [“Relations between forestry and wood properties,” Chapter in the Ministry of Natural Resources, The Quebec Silvicultural Guide, Vol. 2, The Concepts and Application of Forestry, Published in Québec].

FPInnovations (2014). (http://hardwoodinitiative.fpinnovations.ca/en-index.php).

Gong, M., Delahunty, S. Chui, Y. H. and Li, L. (2013). “Use of low grade hardwoods for fabricating laminated railway ties,” Constr. Build. Mater. 41, 73-78. DOI: 10.1016/j.conbuildmat.2012.11.114

Gong, M., Tu, D., Li, L., and Chui, Y. H. (2015). “Planar shear properties of hardwood cross layer in hybrid cross laminated timber,” in: Proc. 5^th International Scientific Conference on Hardwood Processing (ISCHP2015), Sept.15-17, Québec, Canada.

Hallaksela, A. M., and Niemistö, P. (1998). “Stem discoloration of planted silver birch,” Scand. J. For. Res. 13(1-4), 169-176. DOI: 10.1080/02827589809382973

Havreljuk, F., Achim, A., and Pothier, D. (2013). “Regional variation in the proportion of red heartwood in sugar maple and yellow birch,” Can. J. For. Res 43, 278-287. DOI:10.1139/cjfr-2012-0479

Jelonek, T., Tomczak, A., Jakubowski, M., and Grzywinski, W. (2015). The effect of tree senescence on selected properties of wood tissue in Scots pine (Pinus sylvestris L.). Ann. WULS – SGGW, For. Wood Technol. 92, 153-160.

Jessome, A. P. (2000). Strength and Related Properties of Woods Grown in Canada, Report, Forintek Canada Corp.

Kretschmann, D. E. (2010). “Mechanical properties of wood,” in: Wood Handbook, General Technical Report FPL-GTR-190, U.S. Department of Agriculture, Forest Service, Forest Products Laboratory, Madison, WI.

McDonald, J. (2010). Développement de nouveaux procédés et technologies pour l’industrie des bois feuillus – Établissement de liens entre les besoins des utilisateurs et la qualité des sciages requis, FPInnovations Report 201002820, pg. 31.

Mazerolle, M. J. (2006). “Improving data analysis in herpetology: Using Akaike’s information criterion (AIC) to assess the strength of biological hypotheses,” Amphib-Reptilia 27(2), 169-180. DOI: 10.1163/156853806777239922

Mazerolle, M. J. (2012). Package ‘AICcmodavg’. (http://www.icesi.edu.co/CRAN/web/packages/AICcmodavg/AICcmodavg.pdf).

Merela, M., and Cufar, K. (2013). “Mechanical properties of sapwood versus heartwood, in three different oak species,” Drvna Industrija 64(4), 323-334.

Mullins, E. J., and McKnight, T. S. (eds.) (1981). Canadian woods: Their Properties and Uses, Environment Canada, Canadian Forestry Service, Headquarters, Ottawa, 3rd Edition. University of Toronto Press, Toronto, Ontario, pg. 390.

Ohman, J. H. (1968). “Decay and discoloration of sugar maple,” U.S. Department of Agriculture Forest Pest Leaflet 110. Forest Service, pg. 6. http://www.fs.usda.gov/Internet/FSE_DOCUMENTS/fsbdev2_043098.pdf

Pearson, N. (2009). “Identifying the needs of hardwood users in secondary and tertiary processing,” FPInnovations-Forintek Division Report 680-6183, pg. 36.

Rowe, J. S. (1972). Forest Regions of Canada, Publication No. 1300, Natural Resources Canada, Canadian Forest Service, Ottawa, Ontario, Canada.

Sellers Jr., T., McSween, J. R., and Nearn, W. T. (1988). Gluing of Eastern Hardwoods: A Review (General Technical Report SO-71), US Department of Agriculture, Forest Service, New Orleans, LA.

Shigo, A. (1967). “Successions of organisms in discoloration and decay of wood,” in: International Review of Forestry Research, Vol. 2, Romberger, J. A., and Mikola, P. (eds.), Academic Press, New York, NY, pg. 237-299.

Shmulsky, R., and Jones, P. D. (2011). Forest Products and Wood Science: An Introduction, 6^th ed., Wiley-Blackwell, Ames, IA.

Swift, D. E., Duchesne, I. Ung, C. H., Wang, X., and Gagné, R. (2013). Impact of Partial Harvesting on Stand Dynamics and Tree Grades for Northern Hardwood Trees of the Acadian Forest Region, Information Report FI-X-009E, NRCan, CFS-CWFC.

Torquato, L. P., Auty, D., Hernàndez, R., Duchesne, I., Pothier, D., and Achim, A. (2014). “Black spruce trees from fire-origin stands have higher wood mechanical properties than those from older, irregular stands,” Can. J. For. Res. 44(2), 118-127.

Trudelle, M., Gélinas, N., and Beauregard, R. (2009). “Estimation des retombées économiques directes engendrées par le réseau de création de valeur de la filière bois de feuillus durs au Québec,” [“Estimated direct economic benefits generated by the network to create value in the wood industry of hardwood in Quebec,”] For. Chron. 85(4), 538-547.

USDA (1966). “Differences between heartwood and sapwood,” U.S. Forest Service Research Note FPL-0147 (http://www.fpl.fs.fed.us/documnts/fplrn/fplrn147.pdf), accessed July 22, 2015.

Verreault, C. (2000). Utilisation des Coeurs de Bois Franc dans des Bois d’Ingénierie à Usage Structural [Using heartwood of Hard Wood in Structural Engineered Wood], Project Report No. 1949, Forintek Canada Corp, Québec, QC, Canada.

Zelazny, V. F., Ng, T. T. M., Hayter, M. G., Bowling, C. L., and Bewick, D. A. (1989). Field Guide to Forest Site Classification in New Brunswick: Napadogan – Tobique Site Region,New Brunswick Dept. of Natural Resources, Fredericton, NB, Canada.

Article submitted: August 25, 2015; Peer review completed: December 1, 2015; Revised version received and accepted: December 15, 2015; Published: January 13, 2016.

DOI: 10.15376/biores.11.1.2007-2019