Growth and wood properties of a 38-year-old Populus simonii × P. nigra plantation established with different densities in semi-arid areas of northeastern China

Ammar Khalil Mohamed Ahmed · Zhixiang Fu · Changjun Ding ·Luping Jiang · Xidong Han · Aiguo Yang · Yinghui Ma · Xiyang Zhao

Abstract To characterize the effects of plantation densities on the growth characteristics (diameter at breast level, tree height and volume) and the common wood properties,38-year-old Populus simonii × P. nigra clones planted with four levels of spacing (2 m × 2 m, 3 m × 3 m, 4 m × 4 m,and 5 m × 5 m) in a semi-arid area in northeastern China were examined. The results of ANOVA showed significant differences (P ＜0.01) for all the investigated growth traits and wood properties under different plantation densities,except for the chemical composition of wood. The repeatability and phenotypic variations of all the traits varied from 0.34 to 0.99 and from 13.45 to 59.65%, respectively. Except for wood density, which was significantly negatively correlated with the growth traits, a positive correlation was observed between the growth traits and all the other wood mechanical properties. However, most of the correlations between the growth traits and the chemical composition of the wood were not significant. The path analysis for the wood mechanical characteristics and the growth in the prediction of volume were significant and ranged from 0.18 to 0.72 for wood density and diameter at breast height, respectively,while those for the chemical composition of wood ranged from 0.001 to 0.336, which showed a low impact on the volume. The highest stand volume (610 m3) per hectare was observed with the 2 m × 2 m spacing, which consequently provided a high total price and income, while a high individual volume growth per tree was observed with the 5 m × 5 m spacing. The results suggested that for the poplar trees younger than 40 years in a semi-arid area in China,2 m × 2 m spacing is suitable for obtaining a high volume per hectare, while 5 m × 5 m spacing is best for obtaining a high individual volume per tree.

Keywords Fiber length · Hemicellulose content · Lignin content · Planting spacing · Poplar · Wood density

Introduction

Poplar is an important high-yielding tree species with short-rotation coppices for wood production, biomass and ecological purposes (Manzone et al. 2014). Populus simonii × P. nigra (P. xiaohei) is an improved clone that was selected in the 1970s among hybrid clones of P.simonii and P. nigra (Zhang et al. 2016; Yu et al. 2017) and has been well-suited for afforestation and commercial forestry in northern China. P. xiaohei is a short-rotation tree species with excellent wood characteristics and high resistance to different environmental stresses (Wei et al.2012; Zhang et al. 2016). P. xiaohei is an important source of pulp, fiber and energy (Ares 2002). Therefore, this species is considered to be the most important species in economical forests and has attracted attention from commercial and scientific researchers in different countries throughout the world who are interested in improving the silvicultural breeding of poplar species (San et al. 2015).Plantation density is one of the most common silvicultural practices, which affects the growth and quality of forests (Fujimoto and Koga 2010) and is an important factor for improving silvicultural breeding (Hebert et al.2016). Different plantation densities impact wood strength characteristics and density (Guler et al. 2015). In high plantation densities, the individual tree volume growth decreases, and tree mortality increases through excessive intraspecific competition (Akers et al. 2013). Low plantation densities can increase individual tree volume, therefore increase the resistance properties (Guler et al. 2015).The wood mechanical properties are key factors in predicting the suitability of a product for end use and therefore its value (Cassidy et al. 2013). Therefore, the improvement of wood mechanical properties, such as stiffness and density, is becoming an important issue for breeding programs(Hannrup et al. 2000; Fries and Ericsson 2006). Moreover,the breeding processes for fiber and wood properties are related to the economic benefits and yield quality (Yin et al. 2017). For example, tree height (H) and diameter(DBH) directly affect the volume, and the some properties of wood characteristics, such as fiber width (FW) and fiber length (FL); the wood chemical composition (cellulose content, CC and lignin content, LC), have direct impacts on the timber quality and final products (Bos and Caligari 2008; Yin et al. 2017). To evaluate tree growth and mortality, the breeders must keep in mind many aspects,especially the growth characteristics, wood properties and stress resistance (Liu et al. 2015). In addition, the most suitable approach is to combine all of the factors under specific environmental conditions, such as arid and semiarid areas (Yin et al. 2017).

The adaptation of plants and regeneration in arid and semi-arid areas is very difficult, owing to the climate conditions (Joubert et al. 2013). Therefore, trees enhance different mechanisms to ensure persistence in a semi-arid environment. Tree populations that are located in a limited rainfall area face greater mortality and weaker regenerations compared with tree populations that are located in optimal environment conditions due to drought and the stress environment. The variety among tree populations in terms of the growth rate, regeneration and mortality in response to environmental conditions manifests in plant growth and morphology (O’connor and Goodall 2017).Therefore, in semi-arid regions, it is important to keep plants healthy and stable by understanding and applying suitable silvicultural practices (Li et al. 2016).

Although there have been many studies on growth and different wood traits in different tree species, the adaptive planting density, the relationship of growth and the wood characters, such as mechanical and chemical properties of trees in semi-arid areas, have been insufficiently studied or ignored. In this research, 38-year-old P. xiaohei clones were studied, and their growth traits, chemical and mechanical properties were investigated in four levels of initial plantation spacing (PD) (2 m × 2 m, 3 m × 3 m,4 m × 4 m and 5 m × 5 m). The main objectives of the study were (1) to select the most appropriate planting density for poplars in semi-arid areas in northeast China,(2) to evaluate the genetic variation parameters among the different densities, and (3) to identify the most important factors that affect volumes per individual tree.

Materials and methods

Site conditions

The experimental site was located at Wujiazhan County(44°96′N, 125°68′E, and 189 m.a.s.l.), Fuyu City, Jilin Province, which belongs to a semi-arid area in northeastern China. The precipitation, frost-free season and annual average temperature were 300-400 mm, 135-140 d and 4.5 °C, respectively.

Materials and experimental design

The species (P. xiaohei) used in this research was the result of a cross between P. simonii and P. nigra which, in the 1970s, was selected as an improved poplar clone in northeastern China. The experiment was a randomized complete block design, which included five blocks and four levels of spacing (2 m × 2 m, 3 m × 3 m, 4 m × 4 m and 5 m × 5 m), which were randomly arranged in 1980,with 270 plants (30 × 9) in each block.

Growth trait measurement

In September, 2017, when the plantation was 38 years old,30 trees were randomly selected from each block (150 trees per planting density). The diameter at breast height and tree height (H) of each selected tree were measured by a laser vertex and DBH meter tap, respectively. The tree volume(V) was calculated according to Gama et al. (2010).

Wood property measurement

From each planting density (6 trees from each block), 30 wood cores were collected from south to north at breast level (1.3 m) by using an increment borer and were then taken to the laboratory to investigate the wood properties.The wood mechanical properties (FL, WD and FW) and chemical compositions of wood (LC, HEC and HOC) were measured following the method of Yin et al. (2017) and Mu et al. (2009), respectively.

Statistical analysis

All data were analyzed using SPSS 20.0 software (SPSS Inc.,Chicago, IL, USA). Effects of planting density on growth traits and on the wood properties were analyzed. For variance of (ANOVA) the data were analyzed according to Yin et al.(2017). The coefficient of phenotypic variation (PCV),repeatability (R) and phenotypic correlations rA(xy) were calculated using the formula of Hai et al.(2008), Liang et al.(2018) and Bi et al.(2000), respectively. In addition, the path analysis of direct and indirect effects were calculated according to Couto et al. (2017). Statistically, SPSS cannot run with nonnumeric values; therefore, different plantation densities (PD) were defined as follows:

Plantation densities 2 × 2 = 4

Plantation densities 3 × 3 = 9

Plantation densities 4 × 4 = 16

Plantation densities 5 × 5 = 25.

Results

Growth traits and wood properties

There are highly significant differences (P ＞0.01) for all of the investigated traits, except for the wood chemical composition among different planting densities (Table 1). The average values of H, V, DBH, FL, WD, FW, HEC, LC, and HOC for the P. xiaohei in different plantation spacings are shown in Table 2. The average values of H, DBH and V increased while the WD decreased with the lower plantation spacing. The 5 m × 5 m plantation spacing showed the highest average values of growth traits (H: 23.2 m, DBH:36.4 cm, and V: 0.947 m3) and wood traits (FL: 1.881 mm,and FW: 0.017 mm). For WD, the 2 m × 2 m plantation spacing resulted in the highest value, and this value was higher than the lowest values by 0.642 g cm-3.

Phenotypic coefficient and repeatability

The phenotypic coefficient and repeatability of H, V, DBH,FL, WD, FW, HEC, LC, and HOC are shown in Table 3.The PCV values ranged from 13.5 to 59.6%. For the growth traits, V showed the highest PCV values (59.6%),which were higher than that of H (18.9%) and DBH(22.6%) by 40.704% and 37.019%, respectively. Among the wood properties, WD showed the highest PCV value(36.494%), which was higher than the PCV values of FL and FW by 11.169% and 9.665%, respectively. The chemical wood composition showed the lowest PCV values, which ranged from 13.5 to 18.8% for HEC and LC,respectively. The repeatability of all the investigated traits varied from 0.34 to 0.99, for which LC and DBH obtained the lowest and highest values, respectively.

Correlation analysis

The correlations of the different traits under different plantation densities are shown in Table 4. The correlation coefficient of all investigated traits varied from 0.059 (HEC with WD) to 0.981 (V with DBH). The correlation coefficients among growth traits (H, DBH and V) ranged from 0.698(DBH with H) to 0.981 (V with DBH), all of which showed a significant positive correlation. For the wood properties, WD showed no significant correlation with any of the other woodproperties (- 0.245 to 0.272); FL showed a significant positive correlation with FW, HOC, HEC and LC; HOC,HEC and LC showed significant positive correlations with each other. Except for WD, which showed a significantly negative correlation, and FL and FW, which were significantly positively correlated with the growth traits, there were no other significant correlations among the growth traits and wood properties. Furthermore, PD showed a significant positive correlation with traits H, DBH, V, and FL, and a significantly negative correlation with WD, HOC and LC.

Table 1 ANOVA analyses of different growth traits and wood properties of P. xiaohei under different planting densities

LC (%)HEC (%)HOC (%)21.700 ± 1.252a 0.244 ± 0.021 a 1.527 ± 0.475 b 1.556 ± 0.216 a 0.006 ± 0.041 a 48.6 ± 0.024 8.0 ± 0.005 40.7 ± 0.019 18.100 ± 3.872a 22.600 ± 1.955 a 0.296 ± 0.083 a 1.035 ± 0.227 a 1.417 ± 0.194 a 0.006 ± 0.036 a 47.6 ± 0.037 7.8 ± 0.003 39.8 ± 0.040 21.500 ± 1.780b 28.200 ± 1.229 b 0.537 ± 0.076 b 0.900 ± 0.160 a 1.381 ± 0.119 a 0.005 ± 0.037 a 43.4 ± 0.052 8.3 ± 0.011 35.1 ± 0.059 23.200 ± 1.751b 36.400 ± 2.591 c 0.947 ± 0.156 c 0.885 ± 0.276 a 1.881 ± 0.490 b 0.017 ± 0.049 b 45.8 ± 0.023 8.5 ± 0.019 37.3 ± 0.036 H tree height; DBH diameter at breast height; V tree volume; WD wood density; FL fiber length; FW fiber width; HOC holocellulose content; HEC hemicellulose content; LC lignin content.Table 2 Average and standard deviation of different growth traits and wood properties of P. xiaohei under different planting densities FW (mm)FL (mm)-3)WD (g cm 3)V (m DBH (cm)16.600 ± 1.83a Significant different (P ＜0.05) of multiple comparison results were showed by different alphabet (a, b, c)H (m)Plantation density(m × m)2 × 2 3 × 3 4 × 4 5 × 5

Path coefficient analysis

The path coefficient analyses of all the investigated traits are shown in Fig. 1. In the prediction of V, the regression weights and direct impact for the growth traits and wood properties ranged from - 0.178 to 0.724; for the growth traits, the DBH showed the highest effect on V. H also showed a high effect on V, the value of which was 0.612. The effect of WD, FL, FW, HEC, HOC and LC on V ranged from- 0.178 (WD) to 0.286 (HEC). H and DBH showed a negative direct impact on the wood density, - 0.242 and- 0.160, respectively. H, DBH and LC showed a positive direct impact on FL, the values of which ranged from 0.331 to 0.473. The lowest path coefficient was shown between FW and WD, the value of which was only 0.001.

Prices and total stand volume in trade

The prices and total stand volume per hectare under different plantation spacings are shown in Table 5. The highest V (610 m3) per hectare was observed in a 2 m × 2 m plantation, which was higher than the lowest V per hectare (3 m × 3 m, 328.9 m3) by 281.1 m3. The individual volume per tree was substantially increased with the decrease in plantation density, which ranged from 0.244 to 0.947 m3. On the other hand, considering the prices and income potential, the 2 m × 2 m plantation density showed the highest price for total V per hectare($67,100, at a rate of $110 per m3), which was higher than the lowest (3 m × 3 m plantation) price by $30,921.

Discussion

Growth traits and wood properties

In this study, ANOVA revealed significant differences(P ＞0.01) between the growth and wood characteristics;except for the chemical composition of wood. These results indicate that different plantation densities affect the growth and wood traits more than the chemical composition of wood; thus, the plantation density can be used to selectwood properties for different uses. This finding is in accordance with the analysis of a prior study (Hebert et al.2016), which reported that different plantation densities had significant effects on the growth traits and wood traits in Pinus banksiana, but contrasted with a study by Zanuncio et al. (2013), who found significant differences in the chemical composition of eucalyptus wood under different plantation densities. This difference might be explained by changes in some important factors, such as the tree species and the environment. This allows for the determination of whether a positive growth response to different plantation densities would affect tree growth or wood properties.

Table 3 Repeatability and coefficient of variation of growth traits and wood properties of P. xiaohei under different planting densities

The averages of the tree height, volume and diameter at the breast were greatly increased with the decrease in plantation densities. This indicates that different plantation densities deeply affect poplar growth. This result can be useful in selecting suitable plantation densities for poplar growth in semi-arid regions according to cultivation and breeding program goals. These findings are in agreement with prior research (Akhtar et al. 2008; Prasad et al. 2011;Miah et al. 2014), which recorded that the average growth traits of Leucaena leucocephala, Eucalyptus camaldulensis and Avicennia officinalis were significantly greater in low plantation densities. For the wood traits, understanding wood quality is important for meeting different wood product demands from the industry (Gardiner and Moore 2013). Wood density is directly related to mechanical and physical wood properties, such as strength and hardness and is also closely related with the production of pulp and fiber properties (Downes et al. 2006). Wood density can be regarded as a key factor of the internal wood structure measurement, and it is mainly used as a proxy measure for other wood characteristics, especially mechanical and functional wood traits (Barotto et al. 2018). In this study,the wood density decreased from 1.527 to 0.885 g cm-3with the decrease in plantation densities, which indicates that the wood density has a negative relationship with the plantation density of P. xiaohei in semi-arid areas. The findings are the same as that of the study that was made on Hevea brasiliensis (Naji et al. 2012), which reported that at a lower planting density (wide spacing), the wood density decreased slightly. For papermaking, the toughness and strength of paper are greatly related to the fiber properties,especially fiber length and fiber width (Li 2013). In this study, the average values of fiber width and fiber length increased with the decrease in plantation density, which indicated that wide spacing can promote the toughness and strength of wood quality, which is suitable for paper making. The results suggest that wood properties of the poplar are highly affected by plantation density, which could meet the diverse needs of various wood products.Lignin, holocellulose, cellulose, and hemicellulose are considered to be four of the most important chemical compositions of wood, which have an effect onpapermaking (Stickler 2008). Greater paper strength is related to high hemicellulose in the material (Wang et al.2010). In the present study, the averages of the chemical composition of wood varied slightly among the different plantation densities, which indicate that in semi-arid areas,the chemical composition of wood in poplars is not significantly affected by plantation density. The average values of the holocellulose content are higher than that of the LC, which indicates that P. xiaohei has advantageous characteristics for papermaking. This result is in agreement with the results of a study on white paper (Ma et al. 2015).

Table 4 Correlation coefficients between growth traits, wood properties and planting densities of P. xiaohei

Fig. 1 Path analysis diagram to estimate the impact of the different traits on volume of P.xiaohei (solid lines with a single-headed arrow show the non-significant direct effects,while dashed lines with a singleheaded arrow show the significant effects. A doubleheaded arrow shows the correlations between the two variables). Note: H, DBH, V,WD, FL, FW, LC, HEC and HOC indicate the tree height,diameter at breast height,volume, wood density, fiber length, fiber width, lignin content, hemicellulose content and holocellulose content,respectively

Table 5 Prices and total stand volume per hectare under different planting densities

Repeatability and coefficients of variation

In the research field of tree breeding programs, genetic variation is considered to be one of the most important parameters for selection, and repeatability is the parameter that is used to estimate the repeatable portion of the variation (Mwase et al. 2008). Our results observed that the phenotypic coefficient of variations in the diameter at breast height, volume and tree height ranged from 18.95 to 59.65%, which were higher than those that were observed in another study on poplar clones (Si et al. 2016). Higher phenotypic variations (PCV) values indicated that there was a high environmental effect on growth traits and that selection based on different plantation densities was feasible. Furthermore, the repeatability of the growth traits in the present study ranged from 0.90 to 0.99, which were higher than that in a study on Populus tomentosa (Zhang et al. 2013). The logic of this higher repeatability might be that different environments and ages led to differences in the repeatability levels. For the wood properties, PCV values ranged from 25.33 to 36.49%, with a repeatability higher than 0.90, which was higher than that observed by Yu et al. (2014). This finding indicates that selecting a suitable plantation density based on growth and wood traits is convenient. For the chemical composition of wood, PCV and repeatability showed lower values than those that were observed in growth and wood traits, which reveals the stability of chemical properties compared with the growth traits among different plantation densities.

Correlation coefficients

Correlation coefficients describe the degrees of correlation between different traits and characters (Lee et al. 2002) and are important for breeding strategies (Goncalves et al.2005). In the present study, the phenotypic correlation coefficients among height, volume and diameter at breast height ranged from 0.70 to 0.98, which showed a high positive correlation. The results were similar to the results of the research on the Norway spruce (Gerendiain et al.2008), which suggested that selecting suitable plantation density could be beneficial for improving growth traits.Theoretically, short rotation species are related to lower wood density. In this study, the values of wood density decreased with the increase in planting density, but the single tree growth traits showed an opposite tendency.Wood density showed a significant negative correlation with the height, diameter at breast height and volume,which was the same result as findings from previous studies. For example, in a study of the Douglas fir, negative genetic correlations were found to exist between the wood density and diameter at breast height and height (Johnson and Gartner 2006). Additionally, in Guazuma crinite, lower planting densities were obtained with larger trees, but the selection of fast-growing materials could be made at an early growth stage without significantly reducing the wood density (Weber and Montes 2008). For the chemical composition of wood (holocellulose, hemicellulose and lignin contents), the results showed a weak correlation with the growth traits, which was in agreement with studies on Larix olgensis (Yin et al. 2017) and Pinus massoniana (Ji et al. 2007). This result indicates that the growth characteristics and the chemical composition of wood may be genetically independent, which is beneficial for a combined selection. For the correlations between plant density and other investigated traits, the results were interesting and reasonable. There was a significantly positive correlation between PD and H, DBH and V, and, except for FL, all of the wood properties were negatively correlated with PD, as was shown by the coefficients to be significant. The results were similar to the findings of research of Verburg and Van(2003), which indicated that the adaptive planting density is a very important parameter for tree growth and for the development of wood properties.

Path coefficients

A path coefficient analysis is one of the most important tools for determining the relationships between different parameters and for revealing whether the influence is directly reflected within the plant or requires other pathways to produce an effect (Yasin and Singh 2010; Tsegaye et al. 2012). Poplar is the most important commercial tree species, and, as such, growth traits should be primarily considered in the breeding process for high-fiber varieties.In this study, the DBH and H showed a higher direct impact on the V, and the DBH had the highest value of the characteristics investigated. The results were the same as that in a study on the path analysis of poplar clones (Wang et al.2011), which indicated that the H and DBH were the key traits affecting the volume growth. When the growth traits satisfy the needed requirements, the wood properties of the poplar should be considered. The WD, LC, and HOC were important traits for the wood properties that were evaluated(Chen et al. 2016). In this research, the WD and HOC showed a negative direct impact, and the LC showed a positive impact on the volume, but the values were low.The results were similar to the research on Picea mariana(Zhang et al. 2007), which found a negative relationship between the wood density and growth traits and suggested that the relationship of wood density and growth traits may vary to some extent with the genotype and environment(Zhang et al. 2007). The chemical composition of the wood showed a weak impact on other wood and growth traits,which may be because the chemical composition of wood affects the mechanical properties, such as the hardness of the cell, strength and stiffness, more than the wood traits(FL and FW). This is in agreement with a previous study by Cruz et al. (2018), who reported that the main chemical components of wood, such as the contents of holocellulose,hemicellulose and lignin, contribute to and affect the mechanical and physical properties of wood to different degrees.

Economic benefits

An important step of the breeding program is estimating the economic benefits to quantify whether investments in tree breeding can be justified. The selection of a suitable plantation density is a key silvicultural decision that is predominantly driven by financial considerations and the performance of end-use products (Zhang et al. 2002; Watt et al. 2017). In this study, 2 m × 2 m plantation spacing resulted in the highest stand volume (610 m3) per hectare and, consequently, showed the highest price ($67,100).However, 5 m × 5 m plantation spacing showed the highest individual volume per tree, but with lower total stand volume and prices than 2 m × 2 m plantation spacing. This indicates that in semi-arid areas in northeastern China, with trees younger than 38 years old, the number of trees also limits the total stand volume per hectare and,consequently, limits the total income. This result is in accordance with that of previous research (Baldwin et al.2000; Cardoso et al. 2013), which reported that, at the stand level, the stand volume of Pinus species increases with higher plantation densities. In addition, numerous studies have already shown that higher plantation densities can decrease individual volume growth and increase the total stand volume (Barrongafford et al. 2003; Will et al.2010; Akers et al. 2013). By considering only prices and income, a plantation with a high number of trees per hectare could be favored if the objective is to maximize the total income without considering the individual volume growth at an early age.

Conclusions

The initial planting density has a clear effect on tree growth. Therefore, understanding the relationship between planting density and tree growth is very important for breeding strategies (Akhtar et al. 2008). Control of the tree stand density and the selection of tree stock with improved traits are two of the main processes by which a breeding program can influence the growth and wood properties of trees in semi-arid areas. In this study, the effects of the plantation density on growth traits and the wood properties of P. xiaohei in semi-arid areas were examined. Significant variation and high repeatability were observed among different traits, except for the chemical composition of wood, and PCV values for the different traits exceeded 10%. Furthermore, a path coefficient analysis revealed that all growth and wood traits significantly affected the volume, while the diameter at breast height had a high direct impact on the volume. For economic gain, 2 m × 2 m plantation spacing showed the highest economic values by hectare, and 5 m × 5 m spacing showed the highest individual volume and economic values compared with other plantation spacing. In conclusion, in semi-arid areas,depending upon the performance and value of the end-use products, the 5 m × 5 m plantation spacing of P. xiaohei could be prescribed for high individual volume growth and 2 m × 2 m plantation spacing for a high stand volume per hectare.