Physical and Mechanical Properties of Fast Growing Polyploid Acacia Hybrids (A. auriculiformis × A. mangium) from Vietnam

Acacia plants are globally important resources in the wood industry, but particularly in Southeast Asian countries. In the present study, we compared the physical and mechanical properties of polyploid Acacia (3x and 4x) clones with those of diploid (2x) clones grown in Vietnam. We randomly selected 29 trees aged 3.8 years from different taxa for investigation. BV10 and BV16 clones represented the diploid controls; X101 and X102 were the triploid clones; and AA-4x, AM-4x, and AH-4x represented neo-tetraploid families of Acacia auriculiformis, Acacia mangium, and their hybrid clones. The following metrics were measured in each plant: stem height levels, basic density, air-dry equilibrium moisture content, modulus of rupture (MOR), modulus of elasticity (MOE), compression strength, and Young’s modulus. We found that the equilibrium moisture content significantly differed among clones, and basic density varied from pith-to-bark and in an axial direction. In addition, the basic density of AA-4x was significantly higher than that of the control clones. Furthermore, the MOR of AM-4x was considerably lower than the control clones, whereas the MOE of X101 was significantly higher than the control values. The compression strength of AM-4x was significantly lower than that of the control clones, but AH-4x had a significantly higher Young’s modulus. Our results suggest that polyploid Acacia hybrids have the potential to be alternative species for providing wood with improved properties to the forestry sector of Vietnam. Furthermore, the significant differences among the clones indicate that opportunities exist for selection and the improvement of wood quality via selective breeding for specific properties.


Introduction
Plants of the genus Acacia are an important resource in global furniture manufacturing and are vital to the wood industry of Southeast Asian countries [1][2][3][4][5][6][7][8][9][10][11]. Improvements in technology and tree breeding techniques have increased the interest in Acacia products [12]. An Acacia hybrid (an interspecific hybrid of Acacia mangium and Acacia auriculiformis) was first recognized in Malaysia in 1972 [10]; since then, this hybrid has rapidly become an important tree in the industrial output of many countries, including Vietnam.
Many studies on this Acacia hybrid, aimed at providing information for end-use applications, have shown that it is fast growing, has a medium strength, and can be utilized in many ways. For example,  Figure 1 illustrates the samples collected in this study. Logs that were 0.2-1.3 m and 1.5-3 m in length were taken from each tree stem after felling, to extract samples for measurement of mechanical properties. 5-cm disks were also taken to test physical properties at various height levels (0.2-13.5 m). Disk edges were coated with wax to prevent decay and other environmental alterations. DBH (1.3 m) was determined as the mean of two cross diameters. After felling, total height was also measured.

Air-Dry Moisture Content and Basic Density
Pith-to-bark strips (diameter × 30 × 20 mm) were cut from each disk after air drying. The strips were then divided into small pieces at a distance of 1 cm from pith-to-bark with a razor blade. Each piece was weighed at the air-dry stage and then dried at 103°C ± 3°C to determine dry weight. Equilibrium air-dry moisture content was determined at room temperature (around 25°C) and humidity (25-35% relative humidity (RH)). Basic density was measured with an electronic densimeter (EW-300SG, Alpha Mirage, Osaka, Japan). Samples for basic density measurements were soaked in distilled water for at least 24 h to ensure full hydration. They were then immersed and suspended using a metal shield inside the water bath of the densimeter to determine wood volume by water displacement. Samples were then oven-dried at 103 °C for around 48 h until they reached a constant mass; they were then re-weighted. Basic density was calculated as oven-dry mass divided by wood volume.
The change in basic density with the diameter growth of each disk is described as follows:

Air-Dry Moisture Content and Basic Density
Pith-to-bark strips (diameter × 30 × 20 mm) were cut from each disk after air drying. The strips were then divided into small pieces at a distance of 1 cm from pith-to-bark with a razor blade. Each piece was weighed at the air-dry stage and then dried at 103 • C ± 3 • C to determine dry weight. Equilibrium air-dry moisture content was determined at room temperature (around 25 • C) and humidity (25-35% relative humidity (RH)). Basic density was measured with an electronic densimeter (EW-300SG, Alpha Mirage, Osaka, Japan). Samples for basic density measurements were soaked in distilled water for at least 24 h to ensure full hydration. They were then immersed and suspended using a metal shield inside the water bath of the densimeter to determine wood volume by water displacement. Samples were then oven-dried at 103 • C for around 48 h until they reached a constant mass; they were then re-weighted. Basic density was calculated as oven-dry mass divided by wood volume.
The change in basic density with the diameter growth of each disk is described as follows: . . BDn are basic densities for disk diameters of 1, 2, . . . n cm, respectively; BD 1 , BD 2 , . . . BD n are basic densities for samples at position 1, 2, . . . n cm).

Mechanical Testing
Specimens for static bending (20 × 20 × 320 mm, radial × tangential × axial) and compression (20 × 20 × 30 mm, radial × tangential × axial) were cut from a 20-mm-thick board. The specimens were conditioned to a constant mass at 20 • C ± 2 • C and a RH of 65% ± 5%. They were then maintained in this condition until required for testing. The average moisture content of the test samples at this stage was 12%. The modulus of rupture (MOR) and modulus of elasticity (MOE) were determined by a three-point bending test according to ISO 13061-3:2014 and ISO 13061-4:2014 standards [31,32].
After these tests were completed, samples were taken from undamaged portions to determine density for estimating the relationship between bending and density.
Compression parallel to the grain was performed in a 100 kN universal testing machine (AG-I, Shimazdu, Japan). The displacement was measured using strain gage (FLAB 511, TML, Japan) with a gage factor of 2.1% ± 1%. Compression strength and Young's modulus were determined according to ISO 13061-17:2017 and ISO 130614:2014 standards [32,33].

Statistical Analysis
The statistical analysis involved a completely randomized design. ANOVA was used to test for differences in the outcomes of experiments with different numbers of samples. Averages were compared using Tukey's test. The Mann-Whitney U test was also used to test for differences between clone in terms of their physical and mechanical properties. The significance level in all tests was p < 0.05.

Moisture Content and Basic Density
The equilibrium air-dry moisture content of seven clones is described in Figure 2. As shown in Figure 2, X102 had the highest EMC value of the clones, whereas AA-4x had the lowest value. As presented in Table 2, the mean moisture contents of the seven clones were significantly different. Moreover, the data could be divided into two groups: the 4x genotypes in one group and the other clones in a second group. Because wood is always exposed to varying climatic conditions, defining the equilibrium moisture content (EMC) is important for the effective use of wood. Mechanical properties and shrinkage changes, which are probably the most important issues in the end-use applications of solid wood, are positively correlated with the EMC of wood [34][35][36]. The wood properties are related to heartwood and sap wood ratio, where the heartwood absorbs less water than sapwood [37][38][39][40][41][42]. This is same tendency with changing in EMC, which is affected by heartwood to sapwood ratio [43] as well as wood density [41,42]. The relationship between basic density and EMC in Figure 3 pointed out that basic density had a weakly correlation with EMC in this study. Thus, the ratio of heartwood-sapwood may explain the difference in EMC between clones. In previous research, triploid Acacia had a higher proportion of heartwood than diploid Acacia when the plants were of equal age (3.8 years) and from similar locations [30]. Thus, the EMC of clones may be higher when the ratio of heartwood increases. This observation was consistent with the findings of a previous study, which showed that the EMC for the heartwood of Pinus radiata was higher than that of its sapwood at all humidity and temperature conditions tested [43]. In further research, it will be necessary to investigate the heartwood ratio of tetraploid clones, in order to better explain the relationship between EMC and the heartwood ratio.        Figure 4 shows the basic densities of the seven clones. The clone AA-4x had the highest basic density, while AM-4x had the lowest. This result contrasts with those for EMC, where AA-4x had the lowest rather than highest value. This trend is consistent with previous studies, in which EMC was reported to decrease as basic density increased [34,35]. As shown in Table 2, the basic densities of AA-4x and AM-4x were significantly different from those of BV10 and BV16; however, the basic densities of X101, X102, and AH-4x did not differ from those of BV10 and BV16. The wood density of Acacia hybrids was previously reported by PV et al. [30] They showed that the basic density of diploid and triploid Acacia hybrids was 0.36-0.49 g/cm 3 , and their mean values for diploid and triploid hybrids in southern Vietnam are consistent with our results. They also showed that Acacia hybrids from the same clone at the same age planted in southern Vietnam had similar basic densities. In another study, Ismail and Farawahida [7] indicated that the density of 6-year-old Acacia hybrids was 0.47 g/cm 3 . Likewise, Acacia hybrids at 7-8 years old are reported to have 0.43-0.49 g/cm 3 densities [15,17]. In other studies, the densities of Acacia hybrids at 8 and 9-12 years old was 0.61-0.69 and 0.58 g/cm 3 , respectively, whereas A. mangium density was 0.52 g/cm 3 Figure 4 shows the basic densities of the seven clones. The clone AA-4x had the highest basic density, while AM-4x had the lowest. This result contrasts with those for EMC, where AA-4x had the lowest rather than highest value. This trend is consistent with previous studies, in which EMC was reported to decrease as basic density increased [34,35]. As shown in Table 2, the basic densities of AA-4x and AM-4x were significantly different from those of BV10 and BV16; however, the basic densities of X101, X102, and AH-4x did not differ from those of BV10 and BV16. The wood density of Acacia hybrids was previously reported by PV et al. [30] They showed that the basic density of diploid and triploid Acacia hybrids was 0.36-0.49 g/cm 3 , and their mean values for diploid and triploid hybrids in southern Vietnam are consistent with our results. They also showed that Acacia hybrids from the same clone at the same age planted in southern Vietnam had similar basic densities. In another study, Ismail and Farawahida [7] indicated that the density of 6-year-old Acacia hybrids was 0.47 g/cm 3 . Likewise, Acacia hybrids at 7-8 years old are reported to have 0.43-0.49 g/cm 3 densities [15,17]. In other studies, the densities of Acacia hybrids at 8 and 9-12 years old was 0.61-0.69 and 0.58 g/cm 3 , respectively, whereas A. mangium density was 0.52 g/cm 3 and A. auriculiformis densities were 0.54 g/cm 3 at 5.5 years and 0.69 g/cm 3 at 11 years [44][45][46]. Previous research showed that a direct comparison of species is difficult because the wood density of Acacia species varies greatly and depends on site conditions and tree age [46]. For our study, the most interesting observation was that the basic density of AA-4x was significantly higher than that of the 2x clone.    The mean DBHs and tree heights are shown in Table 1. As shown, X101 was the tallest tree and AM-4x had the largest DBH. In addition, X101 and AM-4x had higher stem volumes than those of other clones. Furthermore, X101 and X102 showed higher growth than BV10 and BV16. Nonetheless, faster growth does not affect wood density, because a relationship between wood density and growth The mean DBHs and tree heights are shown in Table 1. As shown, X101 was the tallest tree and AM-4x had the largest DBH. In addition, X101 and AM-4x had higher stem volumes than those of other clones. Furthermore, X101 and X102 showed higher growth than BV10 and BV16. Nonetheless, faster growth does not affect wood density, because a relationship between wood density and growth rate has not been found [47,48]. This trend is applicable to Acacia hybrids, as shown in a similar study of diploid and triploid hybrids [30].

MC (%) BD (g/cm 3 ) MOR (MPa) MOE (GPa) σ (N/mm 2 ) E (GPa)
The radial variation of basic density at each stem height is shown in Figure 5. The basic density increased from pith to the bark in the seven clones, which is consistent with the findings of Walker [49], Kim et al. [18], and Machado et al. [50], who reported that the specific gravity of Acacia increased from the pith to the outer region near the bark. Variance in this condition may be affected by variability inside the trees and other factors, such as climate conditions, during tree growth [50]. The axial variation of the basic density is presented in Figure 6. The basic density was highest at the stump, and it tended to initially decline moving up the tree, before increasing again toward the top. Similarly, recent studies have indicated that density decreases above breast height to about the middle of the tree before increasing toward the top [18,51].   Figure 7 shows the results for MOR in the different clones. The X101 clone had the highest mean MOR at 78.8 MPa, whereas the AM-4x had the lowest value at 56.2 MPa. As shown in Table 2, there was a significant difference between the MOR of X101 and AM-4x and those of the control clones BV10 and BV16.

Bending Test
The MOE of the seven clones is shown in Figure 8. As shown, X101 had the highest MOE at 9.7 GPa, whereas AM-4x had the lowest MOE at 7.2 GPa. Table 2 indicates that the difference between the MOE of X101 and that of the control clones was significant. Taken together, the results of the bending test suggest that the MOR and MOE of clone X101 were significantly higher than that of the control clones. However, the bending properties of AM-4x were considerably lower than those of the controls.
The MOR and MOE values from the static bending test in the present study are lower than those reported by Rokeya et al. [16], Jusoh et al. [15], and Sharma et al. [17] This inconsistency may be explained by variations in age, genotype, or location condition in our study and the others [46,47]. Figures 9 and 10 show the relationships of density with MOR and MOE, respectively. Contrary to expectations, we did not find a significant correlation between bending properties and density. This finding differs from previous studies, which have suggested that density is strongly correlated with bending properties [45,47]. These differences may be explained in part by the relationship between strength and anatomical properties. For example, Nakada et al. [52] and Zhu et al. [53] reported that strength may be related to the microfibril angle of wood fibers to which wood density is less sensitive [52,53]. The reasoning here is that density increases while microfibril angle decreases with age; this impacts mechanical tests and results in weak correlations when density alone is measured [54][55][56]. Grain angle has also been found to affect the correlation between density and mechanical properties [57].  Figure 7 shows the results for MOR in the different clones. The X101 clone had the highest mean MOR at 78.8 MPa, whereas the AM-4x had the lowest value at 56.2 MPa. As shown in Table 2, there was a significant difference between the MOR of X101 and AM-4x and those of the control clones BV10 and BV16.  The MOE of the seven clones is shown in Figure 8. As shown, X101 had the highest MOE at 9.7 GPa, whereas AM-4x had the lowest MOE at 7.2 GPa. Table 2 indicates that the difference between the MOE of X101 and that of the control clones was significant. Taken together, the results of the bending test suggest that the MOR and MOE of clone X101 were significantly higher than that of the control clones. However, the bending properties of AM-4x were considerably lower than those of the controls.    The MOR and MOE values from the static bending test in the present study are lower than those reported by Rokeya et al. [16], Jusoh et al. [15], and Sharma et al. [17] This inconsistency may be explained by variations in age, genotype, or location condition in our study and the others [46,47]. Figures 9 and 10 show the relationships of density with MOR and MOE, respectively. Contrary to expectations, we did not find a significant correlation between bending properties and density. This finding differs from previous studies, which have suggested that density is strongly correlated with bending properties [45,47]. These differences may be explained in part by the relationship between strength and anatomical properties. For example, Nakada et al. [52] and Zhu et al. [53] reported that strength may be related to the microfibril angle of wood fibers to which wood density is less sensitive [52,53]. The reasoning here is that density increases while microfibril angle decreases with age; this impacts mechanical tests and results in weak correlations when density alone is measured [54][55][56]. Grain angle has also been found to affect the correlation between density and mechanical properties [57].

Compression Test
Ultimate stress and Young's modulus in the compression test are shown in Figures 11 and 12, respectively. For both compression properties, AH-4x had the highest values, while AM-4x had the lowest. There was a significant difference in the stress value between AM-4x and the control clones. Likewise, the Young's modulus of AM-4x was significantly lower than that of the diploid group, whereas that of AH-4x was considerably higher than the values of the control group (Table 2). This result suggests that other triploid and the tetraploid clones have similar compression strengths to the control. By comparison, the stress values observed in the present study were not different from those of 8-year-old Acacia hybrids [17], but lower than those of Acacia hybrids at 6-7 years [7,15].

Compression Test
Ultimate stress and Young's modulus in the compression test are shown in Figures 11 and 12, respectively. For both compression properties, AH-4x had the highest values, while AM-4x had the lowest. There was a significant difference in the stress value between AM-4x and the control clones. Likewise, the Young's modulus of AM-4x was significantly lower than that of the diploid group, whereas that of AH-4x was considerably higher than the values of the control group (Table 2). This result suggests that other triploid and the tetraploid clones have similar compression strengths to the control. By comparison, the stress values observed in the present study were not different from those of 8-year-old Acacia hybrids [17], but lower than those of Acacia hybrids at 6-7 years [7,15].

Compression Test
Ultimate stress and Young's modulus in the compression test are shown in Figures 11 and 12, respectively. For both compression properties, AH-4x had the highest values, while AM-4x had the lowest. There was a significant difference in the stress value between AM-4x and the control clones. Likewise, the Young's modulus of AM-4x was significantly lower than that of the diploid group, whereas that of AH-4x was considerably higher than the values of the control group (Table 2). This result suggests that other triploid and the tetraploid clones have similar compression strengths to the control. By comparison, the stress values observed in the present study were not different from those of 8-year-old Acacia hybrids [17], but lower than those of Acacia hybrids at 6-7 years [7,15].  In the Acacia breeding program in Vietnam, there has been a focus on increasing stem volume and growth rate [3,8,21,44,58]. Improving wood quality, however, should also be an important consideration in new breeding generations. Furthermore, developing improved clones of the Acacia hybrid with increased properties of strength would be useful, since the hybrid suffers related problems in some extreme conditions in Vietnam [44]. Previous research suggested that the hybrid clones from Vietnam had lower densities and mechanical properties than A. auriculiformis, or were intermediate between their parents' characteristics [14,18]. However, according to our data, the triploid clone X101 had a similar basic density, similar compression test values, and higher bending test properties than those of diploid clones. Thus, this clone could be considered as a choice for improved Acacia wood quality.
It should be noted that some variation in physical and mechanical properties was unexplained by the measured factors. For instance, the effect of different site conditions and anatomical characteristics on the clone will require additional attention in further research.

Conclusions
Considering wood density, strength, and compression properties will be important for the improvement of the polyploid Acacia breeding program. Various clones had various advantages and disadvantages in this study. For example, AA-4x had significantly higher wood density than the other clones, while AM-4x had greater stem volume than the control clones. In addition, X101 had higher stem volume and bending properties (MOR and MOE) than the other clones, and similar wood density to the control. In general, triploid and tetraploid Acacia hybrids have the potential to be alternative species to supply Acacia wood as valuable hardwood timber. Moreover, X101 could potentially be used for selection to improve and increase the production of high-quality Acacia wood. In the Acacia breeding program in Vietnam, there has been a focus on increasing stem volume and growth rate [3,8,21,44,58]. Improving wood quality, however, should also be an important consideration in new breeding generations. Furthermore, developing improved clones of the Acacia hybrid with increased properties of strength would be useful, since the hybrid suffers related problems in some extreme conditions in Vietnam [44]. Previous research suggested that the hybrid clones from Vietnam had lower densities and mechanical properties than A. auriculiformis, or were intermediate between their parents' characteristics [14,18]. However, according to our data, the triploid clone X101 had a similar basic density, similar compression test values, and higher bending test properties than those of diploid clones. Thus, this clone could be considered as a choice for improved Acacia wood quality.
It should be noted that some variation in physical and mechanical properties was unexplained by the measured factors. For instance, the effect of different site conditions and anatomical characteristics on the clone will require additional attention in further research.

Conclusions
Considering wood density, strength, and compression properties will be important for the improvement of the polyploid Acacia breeding program. Various clones had various advantages and disadvantages in this study. For example, AA-4x had significantly higher wood density than the other clones, while AM-4x had greater stem volume than the control clones. In addition, X101 had higher stem volume and bending properties (MOR and MOE) than the other clones, and similar wood density to the control. In general, triploid and tetraploid Acacia hybrids have the potential to be alternative species to supply Acacia wood as valuable hardwood timber. Moreover, X101 could potentially be used for selection to improve and increase the production of high-quality Acacia wood.