Tree Genetics & Genomes

, Volume 8, Issue 5, pp 1041–1050 | Cite as

Genotypic variation in wood properties and growth traits of triploid hybrid clones of Populus tomentosa at three clonal trials

Original Paper

Abstract

To better understand the genetic control of growth traits (tree height, dbh, and stem volume) and wood properties (basic wood density and fiber length) in triploid hybrid clones of Populus tomentosa, genetic relationships among selected wood properties with growth traits were examined in 5-year-old clonal field trials located in Yanzhou, Gaotang, and Xiangfen, northern China. In total, 180 trees from 10 clones were sampled from the three sites. The site had a moderate effect on basic wood density (BWD), stem wood dry weight (DWT), and tree growth and had a highly significant effect on fiber length (FL) (P < 0.001). Clonal effects were also significant (P < 0.05) for all studied traits (except for diameter at breast height (DBH) and stem volume (SV)). Clone × site interaction was significant for all the studied traits except for FL. The estimated repeatability of clonal means for FL (0.91) was higher than for BWD (0.71), DWT (0.62), tree height (0.62), DBH (0.61), and SV (0.55). Intersite genetic correlation estimates indicated that wood properties were more stable than growth traits. Phenotypic correlation estimates between SV and BWD ranged from −0.29 to −0.10.

Keywords

Triploid clones Wood properties Clonal repeatability Genotypic correlation Genetic gains Populus tomentosa 

Introduction

The breeding objective can be defined as the combination of characteristics or traits that the breeder hopes to improve, usually expressed as a linear combination of the properties of economic importance (Wei and Borralho 1999). Triploid breeding programs for Populus tomentosa have in general emphasized improvements in tree growth, trunk formation, adaptability, and disease resistance in China (Zhu 2006; Zhang et al. 2005). However, little has been done to improve wood density, fiber traits, or other wood properties in triploid breeding programs.

In northern China, triploid hybrids of P. tomentosa are the most important commercial trees within the pulp and paper industry, as well as the sawn timber industry (Zhu et al. 1995; Zhu et al. 1998). Due to the high growth rate, the plantations of genetically improved triploid hybrid poplar clones could reach sawlog size in 8 years with an annual yield of 10–20 m3/yr/ha (Zhu 2006). In fact, potentially high yields and short-rotation cultivation make triploid hybrid poplar clones the preferred tree “species” in most provinces. To obtain high-yield fiber production, triploid hybrid poplar clones are widely planted in Hebei, Henan, Shandong, and Shanxi provinces.

Wood properties related to processing, product yields, and end-use performance are economically important in pulp and paper production (Pu et al. 2002) and include fiber attributes linked to product–performance indices, such as felting coefficient where higher values are associated with better paper resistance (Igartua et al. 2003). Just like pulp yield, wood density is also an important pulpwood parameter and should be an integral component of tree breeding programs (Schimleck et al. 1999). Wood density is considered as one of the most important wood properties, which has a major impact on the freight costs, chipping properties, pulp yield per unit mass of wood, and paper quality (Schimleck et al. 1999; Pliura et al. 2007). However, little is known about the genetics of wood density in triploid hybrid clones of P. tomentosa in China. Previous studies reported that a few of triploid hybrid clonal trials established in China have been evaluated for wood density (Xing and Zhang 2000) and fiber traits (Pu et al. 2002). However, most of these studies were either performed on a single site or based on a small number of samples or clones, whereas the present study used 10 clones on each of the three sites. For tree improvement purpose, estimation of quantitative parameters requires that more genotypes than those used in the previous studies be tested and done at different locations. In addition, it is important to study wood quality traits before incorporating them into tree breeding programs to maximize the economic gains.

In this study, we examined wood properties and growth of nine triploid clones and one diploid clone of hybrid poplar (P. tomentosa) in northern China. The objective of our study was fourfold, to (1) evaluate inheritance of wood properties and growth traits, (2) examine genetic relationship between wood properties and growth traits, (3) compare different scenarios of selection, and (4) test the feasibility of incorporating wood quality traits into triploid hybrid poplar breeding programs. These findings will be used to develop appropriate selection strategies for triploid breeding programs of white poplars in northern China.

Materials and methods

Materials

The materials used in this study were based on three triploid hybrid poplar clonal trials established with cuttings by Beijing Forestry University in the spring of 2004 on agricultural soil with typical conditions of site fertility for P. tomentosa. The planted cuttings were 20–25 cm in length and 1.5–2.0 cm in diameter. There were three replicates in the nursery. The trials were conducted in Yanzhou (lat. 33°10′E, long. 116°49′N, Shandong), Gaotang (lat. 36°51′E, long. 116°04′N, Shandong), and Xiangfen (lat. 35°50′E, long. 111°21′N, Shanxi) in northern China. The trials consisted of nine triploid hybrid clones and one diploid clone on each site (Table 1). The nine triploid clones had been preselected among the triploids for promising growth (Zhu 2006). The diploid clone M1319 was a superior tree of P. tomentosa with high growth rate in northern China. Each clonal trial was established in a randomized block design, with three block replicates each (240 trees per plot giving altogether 720 trees per clone per trial). On each site, clones were planted in rectangular plots containing 240 trees (4 × 60 trees) per plot with 2 × 3-m spacing. No thinning was applied during the trial period until the time of measurements.
Table 1

Identity and origin of the hybrid clones

No.

Clone identity

Parents

Level of ploidy

Sex

1

B301

(P. tomentosa × Populus bolleana) × P. tomentosa

Triploid

2

B302

(P. tomentosa × P. bolleana) × P. tomentosa

Triploid

3

B303

(P. tomentosa × P. bolleana) × P. tomentosa

Triploid

4

B304

(P. tomentosa × P. bolleana) × P. tomentosa

Triploid

5

B305

(P. tomentosa × P. bolleana) × P. tomentosa

Triploid

6

B306

(P. tomentosa × P. bolleana) × P. tomentosa

Triploid

7

B312

(P. tomentosa × P. bolleana) × P. tomentosa

Triploid

8

B330

(Populus alba × Populus glandulosa) × P. tomentosa

Triploid

9

B331

(P. alba × P. glandulosa) × P. tomentosa

Triploid

10

M1319

P. tomentosa × P. tomentosa

Diploid

Measurement of tree growth and evaluation of wood properties

In the spring of 2009, 180 tree samples were randomly collected from the three sites. On each site, six ramets (the six trees) per clone were chosen fully at random for sampling. Total tree height (H) and diameter at breast height (DBH) of all sampled trees were measured at 5 years of age by the triploid poplar breeding program of Beijing Forestry University. Stem volume (SV) was calculated for each sample based on a volume function of DBH and H used by Chen (1989). For wood quality evaluation, all the tree samples were transported to Beijing Forestry University. A 10-cm-thick disk was taken from each tree at breast height for laboratory measurements.

For each disk, the basic wood density (BWD) was determined, on a small rectangular pith-to-bark wood specimen with dimensions of 20 mm vertically × 20 mm radially, using the maximum moisture content method described by Smith (1954). BWD multiplied by SV was calculated for each tree as an index of stem wood dry weight (DWT) (Zhang and Morgenstern 1995).

For intra-ring analysis of fiber properties, matchstick-sized wood specimens (each representing two annual ring pairs) were chipped away from the stem disks and then macerated in a boiling 1:1 (v/v) mixture of acetic acid and hydrogen peroxide. Thereafter, fiber length (FL) was measured using the L&W Fiber Tester (AB Lorentzen & Wettre, Kista, Sweden), which is based on image analysis. In fiber length measurements, a highly diluted suspension flows between two glass plates, which are close to each other and thus limits the possibility of the fibers moving in one direction but allows them to move freely in the other two directions. As a result, the two-dimensional images permit the measurement of fiber length and deformations separately. The use of the L&W Fiber Tester makes it possible to observe a large number of fibers for each sample in a few minutes (i.e., up to tens of thousands of fibers).

Statistical analysis

In the main statistical analysis and the calculation of genetic statistics, the diploid clone (M1319) was not included. Therefore, only the nine triploid clones were involved in the main analyses of variance and estimates of genetic parameters. Analyses of variance were done using the UNIVARIATE procedure of the SPSS software (SPSS for Windows, version 13, SPSS, Chicago, IL). Variation among ramets of the sample clones was analyzed by analysis of variance, using a linear model (Eq. 1) within site:
$$ {X_{{ik}}} = \mu + {C_i} + {\varepsilon_{{ik}}} $$
(1)
where Xik is an observation on the kth ramet from the ith clone, μ is the general mean, Ci is the effect due to the ith clone, and εik is random error. In linear model (Eq. 1), the clone × replicate effects were not considered because the six trees per clone per site were chosen at random and lack of recording the blocks for the sample trees.
Repeatability of clone means within a site was calculated using the following formula (Eq. 2):
$$ R_{\text{c}}^2 = \frac{{\hat{\sigma }_{\text{c}}^2}}{{\hat{\sigma }_{\text{c}}^2 + \frac{{\hat{\sigma }_{\text{e}}^2}}{k}}} $$
(2)
Individual-tree clonal repeatabilities (Rb2) were estimated as
$$ R_b^2 = {{{\hat{\sigma }_{\text{e}}^2}} \left/ {{\left( {\hat{\sigma }_{\text{c}}^2 + \hat{\sigma }_{\text{e}}^2} \right)}} \right.} $$
(3)
where k is the mean of sampled trees per clone within a site, \( \hat{\sigma }_{\text{c}}^2 \) is the estimated variance of clone, and \( \hat{\sigma }_{\text{e}}^2 \) is the estimated variance among ramets within clones.
The linear model (Eq. 4) was used for joint analyses of the three sites together (Zhang et al. 2003):
$$ {X_{{ijk}}} = \mu + {C_i} + {L_j} + {C_i}{L_j} + {\varepsilon_{{ijk}}} $$
(4)
where Xijk is an observation on the kth ramet from the ith clone in the jth location, μ is the overall mean, Ci is the effect due to the ith clone, Lj is the effect due to the jth location, CiLj is the interaction between the ith clone and jth location, and εijk is random error. All terms were considered random, except for location which was considered as a fixed effect.
Repeatability of clone means was calculated according to the following formula:
$$ R_{\text{c}}^2 = \frac{{\hat{\sigma }_{\text{c}}^2}}{{\frac{{k2\hat{\sigma }_{\text{c}}^2}}{{k2}} + \frac{{k1\hat{\sigma }_{{L \times C}}^2}}{{k2}} + \frac{{\hat{\sigma }_{\text{e}}^2}}{{k2}}}} $$
(5)
The individual-tree clonal repeatability across sites was estimated as
$$ R_b^2 = {{{\hat{\sigma }_e^2}} \left/ {{\left( {\hat{\sigma }_c^2 + \hat{\sigma }_{{L \times C}}^2 + \hat{\sigma }_e^2} \right)}} \right.} $$
(6)
where k1 is the coefficient associated with the variance due to the clone × location interaction term (\( \hat{\sigma }_{{L \times C}}^2 \)) and k2 is the coefficient associated with the variance due to clonal variation (\( \hat{\sigma }_{\text{c}}^2 \)). Approximate standard errors (SE) for repeatability estimates were calculated based on the following formula (Becker 1992):
$$ SE\left( {R_C^2} \right) = \sqrt {{\frac{{2(1 - R_{\text{C}}^2)^2[1 + (k{2} - 1)R_{\text{C}}^2]^2}}{{k2(k2 - 1)(N - 1)}}}} $$
(7)
where N is the number of clones tested.
The intertrait clonal correlations on each site were estimated as follows (Becker 1984):
$$ {r_{{{\text{A}}({\text{X}},{\text{Y}})}}} = \frac{{{{\hat{\sigma }}_{{{\text{C}}({\text{x}},{\text{y}})}}}}}{{\sqrt {{\hat{\sigma }_{{{\text{C}}({\text{x}})}}^2\hat{\sigma }_{{{\text{C}}({\text{y}})}}^2}} }} $$
(8)
where rA(X,Y) is the intertrait clonal correlation between the traits x and y, \( {\hat{\sigma }_{{{\text{C}}({\text{x}},{\text{y}})}}} \) is the estimated clonal covariance component between x and y, \( \hat{\sigma }_{{{\text{C}}({\text{x}})}}^2 \) is the estimated clonal variance components for the trait x, and \( \hat{\sigma }_{{{\text{C}}({\text{y}})}}^2 \) is the estimated clonal variance components for the trait y. The corresponding clonal variance components were estimated from data collected from the same individual-tree data, using cross products analogously to using mean squares for estimating variance components.
The approximate standard errors of genetic correlation estimates were computed using the following equation (Falconer 1981):
$$ {\text{SE}} = \frac{{1 - r^2}}{{\sqrt {2} }}\sqrt {{\frac{{{{\hat{\sigma }}_{{(R_{\text{x}}^{{2}})}}}{{\hat{\sigma }}_{{(R_{\text{y}}^{{2}})}}}}}{{R_{\text{x}}^2R_{\text{y}}^2}}}} $$
(9)
where r is the genetic correlation estimate, Rx2 is the repeatability estimate of the trait x, Ry2 is the repeatability estimate of the trait y, \( {\hat{\sigma }_{{(R_{\text{X}}^2)}}} \) is the estimated standard error of Rx2, and \( {\hat{\sigma }_{{(R_{\text{Y}}^2)}}} \) is estimated standard error of Ry2. Significance of between-trait genetic correlations was tested approximately using Eq. 7 as the basis for t tests.
The estimates of genetic correlations between the same traits assessed in different sites were calculated to evaluate the contribution of each pair of treatments to the total clone × environment. These type B genotypic correlations (Burdon 1977) were estimated based on measurements of sample ramets from the same clones planted on different sites using the following formula:
$$ {r_{{{\text{B}}\left( {{\text{X}},{\text{Y}}} \right)}}} = \frac{{{r_{{{\text{p}}({\text{x}}1,{\text{y}}2)}}}}}{{{R_{{{\text{C}}({\text{x}}1)}}}{R_{{{\text{C}}({\text{y}}2)}}}}} $$
(10)
where rp(x1,y2) is the phenotypic correlation coefficient between the clone means estimated between x measured in site 1 and y measured in site 2 and RC(x1) and RC(y2) are the square roots of their clonal mean repeatability, estimated for x and y at site 1 and 2, respectively.

The phenotypic correlations of clonal means for each pair of traits were calculated and tested for significance using the SPSS PROC CORR software.

In this study, the selection intensity is assumed to equal 1.271, which corresponds to the selection of three clones out of 14 or four clones out of 18 (Pliura et al. 2007). The expected gain in trait y △Gy was predicted from the correlated response to clonal selection in trait x using the following formula (Falconer 1981):
$$ \Delta {\text{Gy}} = {i_{\text{x}}}\sqrt {{{R_{\text{x}}}}} {\sigma_{\text{y}}}{r_{\text{xy}}} $$
(11)
where ix (ix = 1.271) is the intensity of selection, Rx is the repeatability of clonal means for trait x, σy is the clonal standard deviation for trait y, and rxy is the genetic correlation between trait x and trait y (rxy ≤ 1).

Results

Basic statistics and variation within and among sites

The mean values, range of variation, and coefficients of phenotypic variation of all studied traits in each of the three clonal trials are presented in Table 2. Trees from the Yanzhou site had the lowest BWD. The highest BWD was observed at the Gaotang site. The difference of BWD between the lowest and highest means was 6.9 % (Table 2). Growth was fastest at Gaotang and slowest at Yanzhou. The longest fibers were observed at Gaotang, and the shortest, at Yanzhou. Among all sites, the largest SV and DWT of trees were found at Gaotang. However, phenotypic variation of these composite traits was highest at the less productive Yanzhou and Xiangfen sites. Joint analysis of all three trials showed significant site effects for all traits (Table 3). BWD and FL except for DWT and SV displayed a small phenotypic variation (coefficient of variation (CV) = 4.9–6.7 %), which was much lower than H variation and DBH variation.
Table 2

Mean values, ranges of variation, and coefficients of phenotypic variation (CVp%) of clonal means of wood properties and growth traits at the three sites

Site

Traits

Triploid clones

Diploid clone (M1319)

Mean ± SE

Range (min–max)

CVp%

Mean ± SE

Range (min–max)

Yanzhou

BWD (kg/m3)

324.3 ± 9.8

286.1–418.0

6.7

335.7 ± 7.9

310.0–370.0

FL (mm)

0.77 ± 0.04

0.68–0.84

5.3

0.67 ± 0.01

0.65–0.69

DWT (kg)

15.7 ± 3.7

9.6–24.5

23.8

9.7 ± 3.0

5.1–13.1

H (m)

12.5 ± 1.2

10.6–15.4

9.7

9.9 ± 0.5

9.2–10.6

DBH (cm)

10.7 ± 1.0

8.5–13.3

9.4

9.0 ± 1.2

7.1–10.4

SV (m3)

0.0482 ± 0.0110

0.0262–0.0704

22.9

0.0287 ± 0.0082

0.0164–0.0395

Gaotang

BWD (kg/m3)

346.8 ± 8.7

302.2–388.7

5.6

381.5 ± 16.1

369.3–414.5

FL (mm)

0.82 ± 0.04

0.73–0.90

4.9

0.69 ± 0.03

0.64–0.74

DWT (kg)

21.5 ± 6.8

7.4–42.8

31.6

14.8 ± 3.2

10.9–20.3

H (m)

13.6 ± 1.2

9.4–16.0

8.9

12.0 ± 0.6

11.1–12.9

DBH (cm)

11.6 ± 1.5

7.8–14.9

12.9

9.7 ± 0.8

8.6–10.8

SV (m3)

0.0618 ± 0.0184

0.0236–0.1139

29.8

0.0386 ± 0.0073

0.0281–0.0489

Xiangfen

BWD (kg/m3)

326.3 ± 7.7

294.1–362.8

5.4

341.5 ± 5.4

330.7–347.7

FL (mm)

0.80 ± 0.04

0.73–0.90

5.0

0.67 ± 0.02

0.64–0.71

DWT (kg)

19.8 ± 5.2

9.8–35.5

26.4

13.4 ± 1.9

10.0–15.9

H (m)

13.2 ± 1.1

11.4–15.8

8.4

11.5 ± 0.4

10.8–12.1

DBH (cm)

11.7 ± 1.2

9.0–14.6

10.3

10.0 ± 0.7

8.8–10.7

SV (m3)

0.0608 ± 0.0156

0.0315–0.1071

25.7

0.0394 ± 0.0057

0.0288–0.0464

BWD basic wood density, FL fiber length, DWT stem wood dry weight, H tree height, DBH diameter at breast height, SV stem volume

Table 3

Summary of results of analyses of variance and estimated repeatabilities (±standard errors) for tree growth and wood properties at the three sites combined

Trait

P value

Rb2

Rc2

Percent

Clones

Sites

Sites × clones

BWD

0.018

0.006

0.015

0.19 ± 0.09

0.71 ± 0.11

61.1

FL

0.000

0.000

0.100

0.45 ± 0.14

0.91 ± 0.04

91.4

DWT

0.043

0.001

0.010

0.17 ± 0.09

0.63 ± 0.13

51.5

H

0.048

0.004

0.008

0.14 ± 0.08

0.62 ± 0.13

49.6

DBH

0.051

0.005

0.010

0.14 ± 0.08

0.61 ± 0.13

49.6

SV

0.083

0.005

0.002

0.12 ± 0.07

0.55 ± 0.13

39.9

Degrees of freedom are 2 for site, 8 for clone, 16 for clone × interaction, and 135 for error

BWD basic wood density, FL fiber length, DWT stem wood dry weight, H tree height, DBH diameter at breast height, SV stem volume, Rb2 individual-tree clonal repeatability, Rc2 repeatability of clonal means, Percent estimated ratio of clonal variance to the sum of clonal plus clone × site variances

Clonal variation and repeatability

Results of analysis of variance for the three sites combined are summarized in Table 3. Significant differences in both wood properties and growth traits were found among the clones (except for DBH and SV) (Table 3). For all traits (except for FL), variance due to error (namely differences among ramets within a clone within a site) accounted for most of the variation in these traits, ranging from 69.2 to 72.6 % of the total variation. Most of the variance in FL (44.7 %), however, was due to the clone. Therefore, FL in the triploid hybrid poplar clones had the highest estimated repeatability (0.91) and estimated individual-tree clonal repeatability (0.45) (Table 3).

Estimates of repeatability at clone-mean and individual-tree levels for tree growth and wood properties at individual sites are presented in Table 4. No significant difference in DWT, H, DBH, and SV was found among the clones at Yanzhou. Therefore, we did not estimate the repeatabilities of DWT, H, DBH, and SV for the four traits. Estimated clonal repeatability for BWD ranged from 0.62 to 0.90 and from 0.79 to 0.90 for FL. Estimated clonal repeatability of growth traits varied from 0.61 to 0.83. However, these differences in repeatability estimates were often insignificant. Repeatability of DWT ranged from 0.74 to 0.75, which was as much as that of SV (0.70 to 0.82).
Table 4

Analysis of variance results and estimated repeatabilities for tree growth and wood properties at each site

Trait

Yanzhou

Gaotang

Xiangfen

P value

Rb2

Rc2

P value

Rb2

Rc2

P value

Rb2

Rc2

BWD

0.018

0.21

0.62 ± 0.14

0.003

0.30

0.72 ± 0.12

0.000

0.59

0.90 ± 0.05

FL

0.000

0.58

0.90 ± 0.05

0.000

0.49

0.85 ± 0.07

0.000

0.38

0.79 ± 0.09

DWT

0.726

0.001

0.34

0.75 ± 0.11

0.002

0.32

0.74 ± 0.11

H

0.389

0.000

0.45

0.83 ± 0.08

0.000

0.39

0.79 ± 0.09

DBH

0.516

0.000

0.44

0.82 ± 0.08

0.020

0.21

0.61 ± 0.14

SV

0.786

0.000

0.43

0.82 ± 0.08

0.005

0.29

0.70 ± 0.12

The degree of freedom for clone is 8 and 45 for error

BWD basic wood density, FL fiber length, DWT stem wood dry weight, H tree height, DBH diameter at breast height, SV stem volume, Rb2 individual-tree clonal repeatability, Rc2 repeatability of clonal means.

Clone × site interaction

In this study, significant interaction of clone × site was observed for the studies of all wood properties (except for FL) and tree growth traits (Table 3). SV had a higher ratio of clone × ite interaction variance to the sum of clonal and clone × site interaction variances (60.1 %) than other traits.

Intersite genotypic correlations are presented in Table 5. Some intersite genotypic correlations between the same traits with pairs of Yanzhou-Gaotang and Yanzhou-Xiangfen were not estimated because no significant difference in DWT, H, DBH, and SV was found among the clones at Yanzhou. Most of the intersite (B type) genotypic correlations between the same traits with pairs of sites were either moderate or strong (Table 5). Higher intersite genotypic correlations were observed in wood properties (except for BWD) than growth traits. Among all studied traits, FL showed the highest intersite genotypic correlations (0.90–0.98), followed by correlations in BWD (0.46–0.71). However, compared with wood properties, growth traits had generally lower intersite genotypic correlations (0.51–0.69). The lowest intersite genotypic correlations observed in SV could be due to the lower estimated ratio of clonal variance to the sum of clonal and clone × site variances (Table 3).
Table 5

Estimated intersite genetic correlations (with approximate standard errors in brackets) for the wood properties and growth traits

Trait

Genotypic correlations between pairs of sites

Yanzhou-Gaotang

Yanzhou-Xiangfen

Gaotang-Xiangfen

BWD

0.65(0.08)

0.46(0.06)

0.71(0.04)

FL

0.91*(0.008)

0.90*(0.01)

0.98*(0.003)

WDT

0.55(0.07)

H

0.69(0.04)

DBH

0.57(0.07)

SV

0.51(0.07)

BWD basic wood density, FL fiber length, DWT stem wood dry weight, H tree height, DBH diameter at breast height, SV stem volume

*P < 0.05 (significant correlations)

Genotypic correlations between traits, genetic gain, and correlated genetic response

Estimated genetic correlations and phenotypic correlations between wood properties and growth traits are presented in Table 6. Most of the genetic correlations between wood properties and growth traits were not estimated at Yanzhou due to the absence of repeatabilities of DWT, H, DBH, and SV. Those between BWD and H, and DBH and SV were not significantly correlated with both phenotypically and genetically. However, a weak and negative estimated correlation between BWD and growth traits existed. This suggests that selection for growth traits might lead to a minor reduction in BWD. A positive or significantly positive estimated genotypic correlation was found between BWD and FL at each site. A positive estimated genotypic correlation between FL and growth traits was also observed at each site. There are some strong positive autocorrelations among SV, H, and DBH because SV was derived from H and DBH. Similarly, DWT, SV, and BWD also had some autocorrelations as DWT was derived from BWD and SV.
Table 6

Estimated intertrait genetic correlations (above the diagonal) and phenotypic correlations (below the diagonal) between wood properties and growth traits across the three sites

Trait

BWD

FL

DWT

H

DBH

SV

Yanzhou

 BWD

 

0.63

 FL

0.47

 

 DWT

0.53

0.58

 

 H

−0.05

0.10

0.18

 

 DBH

−0.15

0.26

0.67*

0.24

 

 SV

−0.10

0.29

0.79*

0.74*

0.90**

 

Xiangfen

 BWD

 

0.93*

0.54

−0.03

−0.05

−0.15

 FL

0.78*

 

0.75

0.52

0.66

0.56

 DWT

0.44

0.57

 

1.22**

1.45**

1.35**

 H

−0.02

0.41

0.93**

 

1.37**

1.31**

 DBH

−0.04

0.46

0.98**

0.95**

 

1.51**

 SV

−0.12

0.42

0.97**

0.98**

0.99**

 

Gaotang

 BWD

 

0.99*

0.27

−0.23

−0.27

−0.38

 FL

0.77*

 

0.56

0.65

0.32

0.35

 DWT

0.20

0.45

 

1.00**

1.21**

1.24**

 H

−0.16

0.55

0.80**

 

0.82*

0.93*

 DBH

−0.21

0.27

0.95**

0.68*

 

1.20**

 SV

−0.29

0.29

0.97**

0.77*

0.98**

 

BWD basic wood density, FL fiber length, DWT stem wood dry weight, H tree height, DBH diameter at breast height, SV stem volume

*P < 0.05(significant correlations)

**P < 0.01

Predicted genetic gains, assuming observed repeatabilities and correlations from direct clonal selection and correlated genetic response in BWD and FL and growth traits, with different selection criteria used, are presented for individual sites in Table 7. Most of predicted genetic gains were not estimated at Yanzhou because of the absence of repeatabilities of DWT, H, DBH, and SV. The predicted gains in BWD were comparable at Gaotang and Xiangfen (4.9 and 5.2 %, respectively), whereas the Yanzhou site had a slightly lower gain (4.3 %). Selection for DBH resulted in the same gains in DWT as direct selection for DWT but at the same time leading to a slight reduction in BWD (Table 7). Selection for BWD led to the lowest gains in DWT. However, the negative effect on stem volume was larger. Predicted gains from direct selection in DWT were as high as the selection for SV. However, they still resulted in an increase of BWD (1.0–2.1 %).
Table 7

Expected response (\( \Delta G/ \overline X \, \times \,100 \)) in wood properties and growth traits at each site when different selection criteria are used

Selection criterion

Response (%)

BWD

FL

DWT

H

DBH

SV

Yanzhou

 BWD

4.3

1.9

 FL

2.4

5.0

Gaotang

 BWD

4.9

3.0

4.0

−0.5

−0.4

−1.6

 FL

4.2

4.1

9.9

3.9

2.7

6.7

 DWT

1.0

1.8

20.7

5.3

9.0

21.3

 H

−0.1

2.3

17.4

7.0

6.7

17.6

 DBH

−0.2

1.1

20.8

4.7

9.8

22.3

 SV

−1.3

1.2

21.3

5.3

9.6

22.7

Xiangfen

 BWD

5.2

2.7

8.1

−1.0

−1.5

−3.6

 FL

3.8

3.2

9.8

2.4

2.2

6.7

 DWT

2.1

1.8

16.6

5.2

4.5

15.2

 H

−0.8

1.3

16.0

5.8

4.6

15.5

 DBH

−1.0

1.3

14.5

4.8

4.7

13.9

 SV

−1.2

1.4

15.8

5.3

4.5

15.4

BWD basic wood density, FL fiber length, DWT stem wood dry weight, H tree height, DBH diameter at breast height, SV stem volume

Discussion

Variation among sites

Site effects reflect the reaction of trees to the combined effects of edaphic and climatic conditions. Even if present experiments were not designed to separate these various effects, some conclusions could be drawn. The slower growth rate at Yanzhou was probably due to poor drainage (because of soil compaction) and wet weather during the early period of the trial. The highest BWD was not observed for the least productive Yanzhou site, as expected, but at the intermediate Gaotang site. Site effects were relatively low for BWD, indicating that trees had less phenotypic plasticity for this trait. Previous studies had reported significant site effects for wood density of poplar hybrids (Nepveu et al. 1986; Song et al. 2000; Xing and Zhang 2000; Zhang et al. 2003). Differences in tree spacing among their trials did not correlate with mean site wood density, unlike other studies on poplar hybrids (Murphey et al. 1979). The BWD of clone B305 (Fig. 1) was higher at the most productive Gaotang site than the least productive Yanzhou site. This tendency would correspond to expectations that plants grow at optimum conditions with higher wood density (Farmer 1970), as wood density was positively correlated with cell wall thickness, which in turn is related to plant nutritional status (Larson 1964). However, the other hybrids had lower wood density at the most productive Gaotang site than the least productive Yanzhou site.
Fig. 1

Mean values of basic wood density of triploid hybrid poplar clones at the three sites

Clonal variation and repeatability

Clonal effects in the joint analysis for all wood properties and growth traits were significant (except for DBH and SV) (P < 0.05). The higher significance of clonal effects was observed at Gaotang and Xiangfen sites (Table 4), indicating that these two sites resulted in larger genetic variation and smaller environment variation than a less productive site. Previous studies (Yanchuk et al. 1983; Yu et al. 2001; Zhang et al. 2003, Li et al. 2005, Seyed et al. 2011) also reported a significant clonal effect in wood density and growth traits of poplars or their hybrids.

From tree breeding and wood utilization perspectives, the variation in density among triploid hybrids was substantial, with a phenotypic CV of 5.4–6.7 %, depending on the site. The BWD values of triploid poplar hybrids, varied from 286.1–418.0 kg/m3, were similar to those reported previously for clones of Populus euramericana (Beaudoin et al. 1992; Hermandez et al. 1998), Populus balsamifera (Ivkovich 1996), Populus deltoides (Olson et al. 1985), and Populus nigra and P. euramericana (Nepveu et al. 1978). However, the present BWD values of triploid clones were somewhat lower than those reported for P. deltoides (Posey et al. 1969), Populus tremuloides Michx. (Yanchuk et al. 1983), and Populus tremula (Kärki 2001). Higher mean wood density values also had been reported for hybrids involving Populus trichocarpa and Populus maximowiczii (Murphey et al. 1979), for hybrid poplars from the AigeirosTacamahaca sections (Phelps et al. 1982) and for different poplar hybrids and P. deltoides (Zhang et al. 2003). These differences might be due to the methods of estimation of wood density, the parentage of hybrids involved in the studies (e.g., P. balsamifera), clonal variation within hybrids, differences in environmental conditions, differences in cambial age of sampled and measured rings, and differences in sampling height. With respect to the radial variation in wood density of poplars, Hermandez et al. (1998) found that the wood density of P. euramericana decreased slightly from the pith to one-third of the diameter and then increased outwards. Trembling aspen wood exhibited a similar pattern of radial variation (Yanchuk et al. 1983).

The mean FL of individual clones ranged from 0.68 to 0.84 mm at Yanzhou, from 0.73 to 0.90 mm at Gaotang, and from 0.73 to 0.90 mm at Xiangfen. Phelps et al. (1982) reported that FL of 4-year-old P. deltoides hybrids ranged from 0.58 to 0.70 mm. Geyer et al. (2000) found that average FL for 4-year-old poplar hybrid clones was 0.84 mm. Arithmetic average FL of hybrid aspen clones measured by Kajaani FS-200 FL Analyzer ranged from 0.43 to 0.70 mm (Yu et al. 2001). Zhang et al. (2008) reported that FL for 5-year-old Populus × euramericana cv. “74/76” ranged from 0.80 to 1.40 mm at DBH. Obviously, the FL of these hybrid poplar clones measured by the Fiber Quality Analyzer was much shorter than that of the other hybrids. The lower FL might be partly due to the fact that the Fiber Quality Analyzer underestimated FL because of the inclusion of curled and broken fibers and vessel elements (Robertson et al. 1999).

Variation among clones for DWT was high, with CV values of 26.4–31.6 %. However, this resulted mostly from larger variations in SV. DWT of clone B302 was the largest among the three sites (Fig. 2). Thus, clone B302 could be considered as the best producer of fiber biomass at sites of different productivity levels.
Fig. 2

Mean values of stem wood dry weight of triploid hybrid poplar clones at the three sites

Clone × site interaction

As poplar breeding programs dealt with developing clones suitable for different environments, genotype × site interaction might have a significant impact on accuracy of breeding values, thus reducing genetic gain. With incongruence between test locations and deployment zones, G × E interaction, if improperly accounted for, could result in bias of estimates and thus a decrease of genetic gain. Significant clone × site interaction for BWD was observed (Table 3). This finding coincided with previous reports on the presence of significant clone × site interaction for wood density of poplar hybrid clones (Farmer 1970; Zhang et al. 2003). The variance component of the clone × site interaction term for BWD in this study (12.0 %) was much higher than that of similar poplar hybrids at age 3 (6.9 %) (Zhang et al. 2003). In other studies on poplars, significant clone × site interaction for wood density was not found (Farmer and Wilcox 1968; Randall and Cooper 1973; Nepveu et al. 1986). Clone × site interaction observed in the current study might have a certain limit because of the wide array of site differences. For growth traits, clone × site interaction in the present study, while modest, might be sufficient to justify detailed clonal testing in order to achieve optimal clonal deployment.

The type B genetic correlation value between Gaotang and Xiangfen for DWT (Table 6) indicated that true G × E interaction was present and that difference between these two sites contributed the most to the clone × site interaction. However, the estimated type B genotypic correlation among sites for FL was mostly positive (Table 6), which indicated that the wood properties of the clones were rather stable across sites. This observation corresponded to the conclusions of Zobel and Jett (1995) that relative performance of genotypes across environments for wood properties was rather stable.

Genotypic correlation between traits

The reports from previous studies on the relationship between growth and wood density were often conflicting. Significant negative genetic or genotypic correlations had been found in many studies involving poplars and poplar hybrids (Yanchuk et al. 1984; Olson et al. 1985; Hermandez et al. 1998; Beaudoin et al. 1992; Pliura et al. 2007). However, genetic relationship between fast growth and wood density had not been observed in other studies (Farmer 1970; Zhang et al. 2003). Based on the analysis of publications on correlations between growth and wood density in many species of distinct wood categories, it was concluded that environmental influence appeared to be one of the reasons why different results had been obtained for the same species (Zhang and Morgenstern 1995). Variations in strength and significance of negative genotypic correlations between most growth traits and wood density at different sites of the present study indicated that under different environments, the expression of genes might be different, and the same trait could be controlled by different gene sets which would result in the alteration of trait values and the interrelations among them. This was supported by recent findings by Wu et al. (2002) that different quantitative trait loci (QTL) for volume growth of poplars were present in different environments and that expression of some QTL depended on the environment.

Earlier studies of different tree species suggested that H had a less negative association with BWD than did DBH (Vargas-Hernandez and Adams 1991; Zhang and Morgenstern 1995; Hermandez et al. 1998; Stener and Hedenerg 2003). In this study, the negative correlation between BWD and H tended to be weaker than that between BWD and DBH at each site (Table 5). Therefore, selection for tree height as opposed to tree diameter might have a smaller negative effect on wood quality.

In general, an increase in tree growth was associated with slightly greater FL. Therefore, genetic improvement of hybrid poplar growth would probably associate with an increase of FL. Numerous studies reported that FL increased with faster growth (Kennedy 1957; Posey et al. 1969; Yanchuk et al. 1984). Xie et al. (1996) found that the genetic correlations between FL and growth traits were positive and statistically significant in black poplar. In this study, positive genotypic correlations were observed between FL and growth traits across the three sites. The result suggested that triploid breeding of poplar increased not only the tree growth but also the FL.

As DWT was derived from SV and BWD, positive correlations between BWD and DWT were larger than between SV and DWT (Table 5). Therefore, DWT should be preferable to primary traits (BWD and SV) for use as a selection index in tree breeding programs.

Implications for tree improvement

The quantity and quality of wood and fiber properties of trees affect the suitability of genotypes as a raw material for pulp and paper and for mechanical wood processing (Zobel and van Buijtenen 1989). This is because relatively small changes in the material properties of wood affect the profitability of processes and the properties of final products. Faced with this concern, tree breeders have realized that wood quantity and quality should not be treated as independent factors. This study shows that wood density has a slightly lower magnitude of genotypic variation than tree height, which is the primary trait used in most tree breeding programs. This suggests that breeding for wood density can be less efficient than breeding for height growth. On the other hand, higher genetic gains may be achievable in wood properties than in tree height as clonal repeatabilities of some of the wood properties are higher than those for height in the entire experiment (Table 3). Growth traits and wood density should be combined to select suitable triploid clones to increase product value.

Due to the presence of a clone × site interaction for DWT and a clone × site interaction for BWD, correspondence among clonal variance components is much lower over the entire experiment than each site taken individually. Thus the expected gain for these traits may be significantly lower with the entire experiment. This calls for selecting clones separately for each site. According to P values, clone × site interactions, though modest, may be sufficient to justify detailed clonal testing in order to achieve optimal clonal deployment.

Our estimates of genetic correlations and prediction of genetic gains are tentative because of the nature and size of the sample. The genetic gains are expected from different criteria at Yanzhou site without repeatability for DWT and growth traits. The expected genetic gains can be achieved in a range of traits at Gaotang and Xiangfen. The choice of selection traits, while suggested by Table 7, remains tentative.

Negative genetic correlations between growth traits and wood density indicate that selection for faster growth can diminish wood density if such factor is not considered in tree breeding. On the other hand, the inclusion of wood density into tree breeding programs can lead to a partial loss of growth. Nevertheless, the use of this criterion can ultimately benefit both types of end uses (e.g., solid-wood products and fiber-based products), as product value will rise with increasing wood density. Simultaneous improvement of growth and wood properties of triploid poplars is impractical when strong negative correlations exist between these traits. Given the apparently weak negative genotypic correlations and the clone × site interactions, there are solutions to improve both growth traits and wood properties of triploid clones: breeding for specialized subpopulations or breeding for a general population using index selection or another scheme of multitrait selection.

One way to incorporate wood density as a selection criterion into tree breeding programs for a general population is to use wood density as a component of a multitrait selection index with economic weights. Another way is to use dry weight as a selection trait, which combines wood volume and density in the most natural and economically feasible way.

Notes

Acknowledgments

This study was mainly supported by the Fundamental Research Funds for Central Universities (YX2010-14) and the Forestry Public Benefit Research Foundation (20100400900-3). The authors would like to thank Dr. Jicheng Pei from Tianjin University of Science and Technology for his assistance with fiber properties measurements. The authors would also like to thank Professor Xiangning Jiang from Beijing Forestry University for his editing.

References

  1. Beaudoin M, Hermandez RE, Koubaa A, Poliquin J (1992) Interclonal, intraclonal and within-tree variation in wood density of poplar hybrid clones. Wood Fiber Sci 24:147–153Google Scholar
  2. Becker WA (1984) Manual of quantitative genetics. Academic Enterprises, Pullman, p 190Google Scholar
  3. Becker WA (1992) Manual of quantitative genetics. Academic Enterprises, Pullman, p 160Google Scholar
  4. Burdon RD (1977) Genetic correlation as a concept for studying genotype × environment interaction in forest tree breeding. Silvae Genet 26:168–175Google Scholar
  5. Chen ZS (1989) The establishment of binomial stock volume tables of poplar. J For Res 2(1):78–83Google Scholar
  6. Falconer DS (1981) Introduction to quantitative genetics, 3rd edn. Longman, London, pp 285–286Google Scholar
  7. Farmer RE Jr (1970) Variation and inheritance of eastern cottonwood growth and properties under two soil moisture regimes. Silvae Genet 19:5–8Google Scholar
  8. Farmer RE Jr, Wilcox JR (1968) Preliminary testing of eastern cottonwood clones. Theor Appl Genet 38:197–201CrossRefGoogle Scholar
  9. Geyer WAG, Dewyke J, Walawender WP (2000) Biomass and gasification properties of young Populus clones. Wood Fiber Sci 32:375–384Google Scholar
  10. Hermandez RE, Koubaa A, Beaudoin M, Fortin Y (1998) Selected mechanical properties of fast-growing poplar hybrid clones. Wood Fiber Sci 30:138–147Google Scholar
  11. Igartua DV, Monteoliva SE, Monterubbianesil MG, Villegas MS (2003) Basic density and fiber length at breast height of Eucalyptus globulus ssp. globulus for parameter prediction of the whole tree. IAWA J 24(2):73–184Google Scholar
  12. Ivkovich M (1996) Genetic variation of wood properties in balsam poplar (Populus balsamifera L.). Silvae Genet 45:119–124Google Scholar
  13. Kärki T (2001) Variation of wood density and shrinkage in European aspen (Populus tremula). Holz als Roh und Werkstoff 59:79–84CrossRefGoogle Scholar
  14. Kennedy RW (1957) Fiber length of fast- and slow-grown black cottonwood. For Chron 33:46–50Google Scholar
  15. Larson PR (1964) Some indirect effects of environment on wood formation. In: Zimmerman MH (ed) Formation of wood in forest trees. Academic Press, New York, pp 345–365Google Scholar
  16. Li S, Jiang Y, Wang G, Wang W, Qiao Y (2005) Joint selection for growth and wood properties in Aigeiros clones. Sci Silvae Sinicae 41(2):53–58Google Scholar
  17. Murphey WK, Bowersox TW, Blankenhorn PR (1979) Selected wood properties of young Populus hybrids. Wood Sci 11:263–267Google Scholar
  18. Nepveu G, Keller R, Tessier du Cross E (1978) Sélection juvenile pour la qualité du bois chez certains peupliers noirs. Ann Sci Forest 35:69–92CrossRefGoogle Scholar
  19. Nepveu G, Barneoud C, Polge H, Aubert M (1986) Variabilité clonale des contraintes de croissance et de quelques autres propriétés du bois dans le genre Populus. In: Fiabilite de l’appreciation de la Qualité du Bios a l’aide de carottes de Sondage, Annales de echerchers Sylvicoles. AFOCEL, France, pp 337–357Google Scholar
  20. Olson JR, Jourdain CR, Rousseau RJ (1985) Selection for cellulose content, specific gravity and volume in young Populus deltoides clones. Can J For Res 15:393–396CrossRefGoogle Scholar
  21. Phelps JE, Isebrands JG, Jowett D (1982) Raw material quality of short-term, intensively cultured Populus clones I. A comparison of stem and branch properties at three spacing. IAWA Bull. n.s. 3:193-200Google Scholar
  22. Pliura A, Zhang SY, MacKay J, Bousquet J (2007) Genotypic variation in wood density and growth traits of poplar hybrids at four clonal trails. For Ecol Manag 238:92–106CrossRefGoogle Scholar
  23. Posey CE, Bridgwater FE, Buxton JA (1969) Natural variation in specific gravity, fiber length, and growth rate of eastern cottonwood in the southern Great Plains. Tappi 52:1508–1511Google Scholar
  24. Pu J, Song J, Yao C (2002) Variations in fiber forms of triploid clones of Populus tomentosa. J Beijing For Univ 24(2):62–66Google Scholar
  25. Randall WK, Cooper D (1973) Predicted genotypic gains from cottonwood clonal tests. Silvae Genet 22:165–167Google Scholar
  26. Robertson G, Olson J, Allen P, Chan B, Seth R (1999) Measurement of fiber length, coarseness, and shape with the fiber quality analyzer. TAPPI 82:93–98Google Scholar
  27. Schimleck LR, Michell AJ, Raymond CA, Muneri A (1999) Estimation of basic density of Eucalyptus globulus using near-infrared spectroscopy. Can J For Res 29(2):194–202CrossRefGoogle Scholar
  28. Seyed Mehdi Safavi, Mohsen Farshadfar, denial Kahrizi, Seyed Afshin Safavi. (2011) Genetic variability of some morphological traits in poplar clones. Am J Sci Res 13:113-117Google Scholar
  29. Smith DM (1954) Maximum moisture content method for determining specific gravity of small wood samples. For Pro Lab Rep, No. 2014Google Scholar
  30. Song W, Zhang Z, Xu J (2000) Study on inheritance and variation of wood basic density of Populus tomentosa Carr. clones. SCI SILVAE SINICAE 36(sp. 1):125–130Google Scholar
  31. Stener LG, Hedenerg Ö (2003) Genetic parameters of wood, fiber, stem quality and growth traits in a clone test with Betula pendula. Scand J For Res 18:103–110CrossRefGoogle Scholar
  32. Vargas-Hernandez J, Adams WT (1991) Genetic variation of wood density components in young coastal Douglas-fir: implications for tree breeding. Can J For Res 21:1801–1807CrossRefGoogle Scholar
  33. Wei X, Borralho NMG (1999) Objectives and selection criteria for pulp production of Eucalyptus urophylla plantations in south east China. For Genet 6(3):181–190Google Scholar
  34. Wu R, Ma C-X, Zhu J, Casella G (2002) Mapping epigenetic quantitative trait loci (QTL) altering a developmental trajectory. Genome 45:28–33PubMedCrossRefGoogle Scholar
  35. Xie HF, Yu ZK, Chen DL, Guo DC, Song FX (1996) Selection and genetic analysis for wood properties in pulpwood varieties of black poplar. J Shandong For Sci 1:1–5Google Scholar
  36. Xing X, Zhang Z (2000) Genetic variation in wood density of triploid clones of Populus tomentosa. J Beijing For Univ 22(6):16–20Google Scholar
  37. Yanchuk AD, Dancik BP, Micko MM (1983) Intraclonal variation in wood density of trembling aspen in Alberta. Wood Fiber Sci 15:387–394Google Scholar
  38. Yanchuk AD, Dancik BP, Micko MM (1984) Variation and heritability of wood density and fiber length of trembling aspen in Alberta. Canada Silvae Genet 33:11–16Google Scholar
  39. Yu Q, Pulkkinen P, Rautio M, Haapanen M, Alen R, Stener LG, Beuker E, Tigerstedt PMA (2001) Genetic control of wood physiochemical properties, growth and phenology in hybrid aspen clones. Can J For Res 31:1348–1356CrossRefGoogle Scholar
  40. Zhang S, Morgenstern EK (1995) Genetic variation and inheritance of wood density in black spruce (Picea mariana) and its relationship with growth: implications for tree breeding. Wood Sci Technol 30:63–75CrossRefGoogle Scholar
  41. Zhang SY, Yu Q, Chauret G, Koubaa A (2003) Selection for both growth and wood properties in hybrid poplar clones. For Sci 49:901–908Google Scholar
  42. Zhang Q, Zhang ZY, Lin SZ, Lin YZ (2005) Resistance of transgenic hybrid triploids in Populus tomentosa Carr. against 3 species of lepidopterans following two winter dormancies conferred by high level expression of cowpea trypsin inhibitor gene. Silvae Genetica 54(3):108–116Google Scholar
  43. Zhang L, Zhao RJ, Fei BH (2008) Morphological analysis of the fibers of Populus × euramericana cv.“74/76”. China Pulp Paper 27(5):28–31Google Scholar
  44. Zhu Z (2006) Genetic improvement of Populus tomentosa. Chinese Forestry Press, BeijingGoogle Scholar
  45. Zhu Z, Lin H, Kang X (1995) Studies on allotriploid breeding of Populus tomentosa B301 clones. SCI SILVAE SINICAE 31(6):499–505Google Scholar
  46. Zhu Z, Kang X, Zhang Z (1998) Studies on selection of natural triploids of Populus tomentosa. SCI SILVAE SINICAE 34(4):22–31Google Scholar
  47. Zobel BJ, Jett JB (1995) Genetics of wood production. Springer-Verlag, Berlin, p 337CrossRefGoogle Scholar
  48. Zobel BJ, van Buijtenen JP (1989) Wood variation, its causes and control. Springer-Verlag, Berlin, p 363CrossRefGoogle Scholar

Copyright information

© Springer-Verlag 2012

Authors and Affiliations

  1. 1.National Engineering Laboratory for Tree Breeding, Key laboratory of Genetics and Breeding in Forest Trees and Ornamental Plants, Ministry of EducationBeijing Forestry UniversityBeijingPeople’s Republic of China

Personalised recommendations