3.1. Within- and Among-Site Variation
The mean values of all studied growth traits for 7-year-old clones of hybrid white poplar in each trial are presented in
Table 3. On average, the greatest DBH, H, and V were found at Fengfeng nursery, and the lowest were observed at Guanxian nursery, with 21.80%, 23%, and 75.25% differences, respectively. Clearly, growth was fastest at Fengfeng and slowest at Guanxian. The ANOVA for the three sites in
Table 4 revealed a significant effect of growth site on the growth characteristics studied.
Table 3 and
Table 4 show that Fengfeng differed significantly (α = 0.05) in all growth traits from the other sites.
Table 3.
Mean growth traits of four clones from three sites.
Table 3.
Mean growth traits of four clones from three sites.
Trait and Geographical Origin (Site) | Mean ± SE | Mean | Significance |
---|
Yiyang-1 | Yiyang-2 | Yiyang-3 | LM50 |
---|
Fengfeng Nursery |
DBH (cm) | 20.4 ± 0.1 a | 20.9 ± 0.1 a | 16.5 ± 0.2 c | 18.3 ± 0.3 b | 19.0 | ** |
H (m) | 15.1 ± 0.1 ab | 14.4 ± 0.2 b | 15.5 ± 0.4 a | 14.9 ± 0.2 ab | 15.0 | * |
V (m3) | 0.2043 ± 0.0296 a | 0.2058 ± 0.0137 a | 0.1374 ± 0.0336 c | 0.1633 ± 0.0854 b | 0.1777 | ** |
Guanxian Nursery |
DBH (cm) | 14.9 ± 0.2 b | 17.7 ± 0.1 a | 15.1 ± 0.1 b | 14.7 ± 0.1 b | 15.6 | ** |
H (m) | 11.9 ± 0.1 b | 11.9 ± 0.1 b | 13.2 ± 0.1 a | 12.1 ± 0.2 b | 12.2 | * |
V (m3) | 0.0892 ± 0.0230 b | 0.1303 ± 0.0080 a | 0.0983 ± 0.0194 b | 0.0878 ± 0.0211 b | 0.1014 | ** |
Weixian Nursery |
DBH (cm) | 18.9 ± 0.2 a | 15.5 ± 0.2 b | 13.7 ± 0.2 c | 17.7 ± 0.2 a | 16.5 | ** |
H (m) | 14.9 ± 0.2 a | 14.5 ± 0.1 a | 15.0 ± 0.1 a | 15.6 ± 0.0 a | 15.0 | NS |
V (m3) | 0.1758 ± 0.0408 a | 0.1161 ± 0.0251 b | 0.0931 ± 0.0119 c | 0.1605 ± 0.0309 a | 0.1364 | ** |
Table 4.
Genetic stability of four clones at three sites.
Table 4.
Genetic stability of four clones at three sites.
Clones | ΔV (%) | CV (%) | Ci | p < 0.01 |
---|
Yiyang-1 | 14.0 | 35.66 | 0.0179 | A |
Yiyang-2 | 9.8 | 31.87 | 0.0122 | B |
LM50 | 0.0 | 33.58 | –0.0013 | C |
Yiyang-3 | –20.1 | 24.34 | –0.0289 | D |
Mean | – | 31.36 | 0.0000 | |
The significant site effect was consistent with previous reports [
5,
15,
16,
21,
22,
23]. Site effects reflect the reaction of trees to the combined effects of edaphic and climatic conditions [
2,
16,
21]. With the exception of cultivation and the management of irrigation and drainage, variation in growth rate in this study is attributable in part to geography, soil quality, annual precipitation, etc. Individuals from Fengfeng nursery show characteristics superior to those from Weixian or Guanxian: all clones planted at Fengfeng exceeded the productivity of the other sites in terms of stem volume. Fengfeng has high annual precipitation compared to the other two sites, and more fertile soil than Guanxian, which may explain these differences in wood quantity and quality.
3.2. Clonal Variation in Growth Traits
The variation in growth traits at each site differed in significance (with the exception of H at Weixian) among the four hybrid clones (
Table 3). At α = 0.05, the V ranking for the hybrid clones at each site varied.
For the Fengfeng nursery, Yiyang-1 and Yiyang-2 had greater V values than did LM50 or Yiyang-3, with Yiyang-3 having the lowest value. Specifically, the V of Yiyang-1 and Yiyang-2 were greater than that of LM50 (the control) by 25.1%, 26.0%, and −15.9%, respectively. However, we found no significant differences between Yiyang-1 and Yiyang-2. For the Guanxian nursery, Yiyang-1 had a V similar to that of LM50, and the percentage of excesses V in Yiyang-2 and Yiyang-3 reached 48.4% and 12.0%, respectively. Yiyang-2 had the greatest V and differed significantly from Yiyang-1, LM50, and Yiyang-3, respectively, whereas no significant differences were observed among the three clones. For the Weixian nursery, the V of Yiyang-1 was similar to that of LM50 and greater than those of Yiyang-2 and Yiyang-3, with the lowest V observed for Yiyang-3.
Yiyang-3 displayed notably lower growth performance than the other three clones, except at Guanxian nursery where it was similar to that of LM50 (
Table 3;
Figure 1). However, Yiyang-1 and Yiyang-2 exhibited faster growth than Yiyang-3 and LM50 at Fengfeng and Guanxian. Although significant differences between Yiyang-1 and Yiayng-2 were minor or inconsistent, Yiyang-1 had the greatest total mean volume and the largest percentage-of-volume increment
vs. LM50 across all three sites (
Table 4). Among all clones, Yiyang-1 was ranked first, followed by Yiyang-2, LM50, and Yiyang-3. In this case, Yiyang-1 grew faster than LM50 and the other two new hybrid clones.
Figure 1.
Individual clone mean as a percentage of the LM50 clone mean (dashed line) at each site.
Figure 1.
Individual clone mean as a percentage of the LM50 clone mean (dashed line) at each site.
The site × clone interactions of all studied properties were significant (
Table 5), which demonstrated that the clones performed differently among the three locations. Overall, these differences demonstrate that the clonal effect on wood growth traits was considerable, suggesting that the properties of the hybrids were mainly correlated with those of their parents. Pairwise comparisons of the samples from each clone at each site showed no statistically significant differences (α = 0.05), indicating potential genetic stability for each clone.
Table 5 focuses on the genetic stability and growth adaptability of the four clones. V and clonal effect values (Ci) reflect yield levels. Genetic stability refers to the stability of traits controlled by genetics in different environments. The coefficient of variation (CV) in the site × clone interaction was applied for this test. The comprehensive ranking of the hybrid poplar clones was Yiyang-1 > Yiyang-2 > LM50 > Yiyang-3, with high significance (
p < 0.01). Theoretically, the smaller the CV value, the more stable the growth will be. We found a negative correlation between the CV and clonal growth stability. All clones exhibited strong stability, with CV values ranging from 24.34% to 35.66%. The CV was greatest for Yiyang-1 and lowest for Yiyang-3, inferring that Yiyang-1 exhibits better growth under better cultivation practices, that Yiyang-2 has no special cultivation requirements, and that Yiyang-3 is adapted to relatively poor environmental conditions.
Table 5.
Summary of the ANOVA results for tree growth at the three sites combined.
Table 5.
Summary of the ANOVA results for tree growth at the three sites combined.
| DBH (cm) | H (m) | V (m3) |
---|
| DF | MS | F | P | MS | F | P | MS | F | P |
Sites | 2 | 50.80 | 1636.05 | < 0.001 | 40.54 | 1356.78 | < 0.001 | 0.023 | 1960.45 | < 0.001 |
Clones | 3 | 23.88 | 769.20 | < 0.001 | 1.85 | 61.76 | < 0.001 | 0.005 | 439.55 | < 0.001 |
Site × Clone | 6 | 11.27 | 363.06 | < 0.001 | 0.61 | 20.36 | < 0.001 | 0.003 | 280.25 | < 0.001 |
Error | 36 | 0.03 | | | 0.03 | | | | 1.191 × 10−5 | |
3.3. Clonal Variation in Wood Properties
BWD is an indicator of the degree of wood compactness, and is positively associated with dry air density, mechanical strength, and hardness. The mean and standard error and pairwise comparisons of several physical properties at breast height from four clones at Fengfeng are presented in
Table 6. The overall mean BWD of the four hybrid clones was 0.324 g cm
−3. Yiyang-3 displayed the greatest BWD of the hybrids, but BWD did not differ significantly among clones. Notably, the differences in fiber traits among the four clones were statistically significant. The mean FLs ranged from 872.90 to 1062.61 µm for all hybrids (
Figure 2). Yiyang-3 had the lowest average FL, at 17.85%, 17.60%, and 17.58% shorter than those of LM50, Yiyang-1, and Yiyang-2, respectively. Yiyang-1 and Yiyang-2 both had slightly shorter FLs, but were statistically similar to those of LM50. The mean FWs ranged from 23.13 to 27.69 µm for all hybrids (
Figure 2). Yiyang-2 had the largest average FW, while Yiyang-3 had the smallest. Yiyang-2 displayed markedly higher FW than the other clones, whereas Yiyang-1 did not differ significantly from Yiyang-3. LM50 displayed 6.9% and 9.2% higher FW than Yiyang-1 and Yiyang-3, respectively. The mean FL:FW of the four hybrids ranged from 37.74 to 44.83 (
Figure 2). Yiyang-1 had the greatest average FL:FW, while Yiyang-3 had the lowest. The FL:FW for LM50 was 10.0% and 11.6% greater than those of Yiyang-2 and Yiyang-3, respectively, while that for Yiyang-1 was 6.5% greater than that for LM50. The greatest difference in FL:FW occurred between Yiyang-1 and Yiyang-3, but there was no significant difference between Yiyang-2 and Yiyang-3. Variation in VDS, crystallinity, and MFA, which are related to wood mechanical properties, was not significant among the four clones.
Based on our results, the fiber traits (FL and FW) of Yiyang-1, Yiyang-2, and Yiyang-3 are consistent with the ranges reported for several FL and FW values in various species or hybrids obtained from the literature [
5]. Additionally, the FL values of Yiyang-1 and Yiyang-2 were > 1000 μm, which corresponds to an intermediate wood FL of 0.9–1.6 mm according to the International Wood Anatomical Society [
24]. Fiber properties have a major effect on the quality of pulp and paper products. Greater fiber lengths and narrower fiber widths (with a length to width ratio of > 30) are preferred for pulp and paper production, and result in higher-density paper sheets, smoother paper, more uniform formation,
etc. [
25,
26]. In this study, Yiyang-1 and LM50 both showed similarly high fiber lengths but different fiber widths than Yiyang-2, and the FL: FW ratio of Yiyang-1 was the greatest among clones. With respect to fiber traits, Yiyang-1 would be particularly promising for pulp and paper manufacturing applications, followed by LM50, Yiyang-2, and Yiyang-3. The cellulose contents of Yiyang-2, Yiyang-1, LM50, and Yiyang-3 were 49.52%, 49.14%, 49.07%, and 48.36%, respectively (
Figure 2). The holocellulose and lignin contents ranged from 75.81% to 77.35% and from 19.72% to 21.06%, respectively (
Figure 2). Yiyang-2 had the highest average cellulose content and Yiyang-3 had the lowest; Yiyang-3 was markedly lower than Yiyang-2, Yiyang-1, and LM50, while there were no significant differences among Yiyang-2, Yiyang-1, and LM50. The mean holocellulose content ranged from 75.81% to 77.35%. Yiyang-1 was 0.8%, 1.4%, and 2.0% greater than LM50, Yiyang-2, and Yiyang-3, respectively. Mean lignin content varied in a range from 19.72% to 21.06%. Similar results among different poplar clones were reported previously [
27,
28]: Yiyang-1 was the lowest, 2.8%, 3.7%, and 6.4% lower than LM50, Yiyang-2, and Yiyang-3, respectively. However, there was a small but significant difference among LM50, Yiyang-2, and Yiyang-3.
Table 6.
Mean values and standard error of the wood quality at the Fengfeng nursery.
Table 6.
Mean values and standard error of the wood quality at the Fengfeng nursery.
Clone Identity | BWD (g cm−3) | FL (µm) | FW (µm) | FL:FW | VDS (%) | MFA | C (%) | Water (%) | Cellulose (%) | Holocellulose (%) | Lignin (%) |
---|
LM50 | 0.326 ± 0.005 | 1062.61 ± 8.06 | 25.26 ± 0.75 | 42.10 ± 1.44 | 6.75 ± 1.02 | 31.00 ± 0.49 | 44.30 ± 1.02 | 90.4 ± 0.20 | 49.07 ± 0.15 | 76.7 ± 0.52 | 20.28 ± 0.16 |
Yiyang-1 | 0.318 ± 0.006 | 1059.34 ± 28.25 | 23.63 ± 0.25 | 44.83 ± 1.07 | 8.00 ± 3.98 | 31.13 ± 0.29 | 45.32 ± 0.75 | 91.26 ± 0.54 | 49.14 ± 0.52 | 77.35 ± 0.50 | 19.72 ± 0.67 |
Yiyang-2 | 0.325 ± 0.005 | 1059.10 ± 11.70 | 27.69 ± 0.59 | 38.26 ± 1.07 | 12.80 ± 1.88 | 31.38 ± 0.54 | 44.18 ± 1.57 | 92.72 ± 0.24 | 49.52 ± 0.25 | 76.29 ± 0.28 | 20.48 ± 0.41 |
Yiyang-3 | 0.326 ± 0.010 | 872.90 ±9.71 | 23.13 ± 0.30 | 37.74 ± 0.35 | 10.02 ± 2.48 | 31.21 ± 0.35 | 44.20 ± 1.12 | 91.92 ± 0.79 | 48.36 ± 0.06 | 75.81 ± 0.27 | 21.06 ± 0.80 |
Mean | 0.324 | 1013.49 | 24.83 | 40.73 | 9.39 | 31.18 | 44.50 | 91.58 | 49.02 | 76.54 | 20.39 |
F-value | 1.076 | 77.264 | 66.407 | 27.376 | 3.009 | 0.554 | 0.868 | 10.541 | 15.547 | 7.383 | 3.957 |
p-value | 0.427 | <0.001 | <0.001 | 0.001 | 0.116 | 0.602 | 0.467 | 0.008 | 0.003 | 0.019 | 0.072 |
Figure 2.
Boxplots of fiber traits and wood chemical components of four hybrids. (maximum-minimum values; the top and bottom of each box indicate the 75th and 25th percentiles, respectively. FL refers to fiber length; FW refers to fiber width.).
Figure 2.
Boxplots of fiber traits and wood chemical components of four hybrids. (maximum-minimum values; the top and bottom of each box indicate the 75th and 25th percentiles, respectively. FL refers to fiber length; FW refers to fiber width.).
Chemical components have an important effect on pulping and papermaking. Lignin, a complex phenolic heteropolymer, represents an important fitness trait in tree adaptation to the environment. However, lignin content has adverse effects on the pulping process when lignin removal is accompanied by cellulose degradation [
1,
29]. According to Hart
et al. [
1], transgenic poplar trees with lowered lignin content showed as much as 15% improvement in their efficiency at converting cellulose to corresponding monomers. Therefore, higher pulp yield and quality are related to higher cellulose and lower lignin contents [
30]. Our results showed that Yiyang-3 was markedly inferior in terms of wood chemical components than the other three clones studied, but that the differences among Yiyang-1, Yiyang-2, and LM50 were small. Nevertheless, minor changes will certainly affect the processing operations of large-scale pulp mills. In this case, with its slightly higher holocellulose and lower lignin content relative to Yiyang-2, Yiyang-1 would be superior in wood performance.
The differences in wood quantity and quality among the clones indicate heterosis and genetic control of properties (
Table 3,
Table 4 ,
Table 5 and
Table 6). The hybrid poplar clones Yiyang-1, Yiyang-2, and Yiyang-3, selected as quality varieties, have a common female parent (
P. tomentosa ×
P. bolleana) and a different male parent.
Populus tomentosa “Truncata” was selected as the male parent of clones Yiyang-1 and Yiyang-2, while
P. tomentosa “LM50” was selected as the male parent of clone Yiyang-3.
Populus tomentosa ×
P. bolleana,
P. tomentosa “Truncata”, and
P. tomentosa “LM50” were all excellent varieties, offering a straight stem, fast growth, strong resistance, and good wood properties.
Populus tomentosa “Truncata” is a mutant variety of
P. tomentosa Carr. (
P. tomentosa ×
P. bolleana) hybridized by the Chinese Academy of Forestry in the 1950s as an improved variety.
Populus tomentosa “LM50”, bred in Shandong Province, has been recognized as a superior clone. Numerous studies have shown both significant and stable heterosis in F1 hybrids of
Populus spp. However, the genetic cause is not yet fully understood [
31,
32].
There is also some doubt as to whether breast height measurements provide an accurate projection of the economic and commercial value of trees [
33]. Several investigations have attempted to determine a representative sampling height, and some authors concluded that sampling at heights other than breast height provided better predictions of the BWD and FL of the entire tree [
34,
35]. Therefore, additional research is needed to better evaluate performance in terms of tree applications and production.