1. INTRODUCTION
Eucalyptus pellita F. Muell is one of the fast-growing species widely distributed in Indonesia, Papua New Guinea, and Queensland-Australia. It is prioritized for industrial forest plantations and has great potential as an alternative species to replace Acacia mangium which is currently experiencing a decline due to root rot disease in the tropics (Lee, 1993). This species has high adaptability and grows fast with single trunk, straight stem, free of high branches, as well as resistant to pests and diseases (Harwood et al., 1997). Diseases rendered A. mangium unviable in south and central Sumatra, and the industry responded by changing the species to E. pellita (Nambiar et al., 2018). This species has also been widely selected for the development of forest plantations in Indonesia and Southeast Asia (Lukmandaru et al., 2016). Plantation forest development in Indonesia has reached 10.8 million hectares (Hadi et al., 2019). Currently, E. pellita had replaced A. mangium in about 465,000 ha in Sumatra and 225,500 ha in Kalimantan, Indonesia (Hardiyanto et al., 2021). Several forest plantations are located in East Kalimantan and there is limited information on the results of research (growth characteristics and wood properties) on E. pellita in East Kalimantan. In contrast, the wood pulp industry still appears to be a top government priority of Indonesia to the expected increased supplies from industrial forest plantations.
Forest plantations have an important role in meeting the wood needs of the forest industry as they meet about a third of the world's timber demand (Barua et al., 2014). Eucalyptus plantation dominates the tropics and sub-tropics with proportions of 42% and 26% (Binkley et al., 2017), as well as the hardwood plantations (Ghani and Lee, 2021). The high productivity increase in plantation forests is currently dominated by clonal forestry. The clones were selected because they have advantages, namely high productivity, applicable as hybrids (Griffin, 2014), and can also maximize genetic gain (White et al., 2007). However, to supply many improved productions clones, a large clonal plantation needs to be supported by a program of breeding and clonal development and testing (Kha et al., 2012). Identified the critical issues that would help to increase the wood productivity by practicing an integrated approach to applications research and comparable environment (Nambiar et al., 2018). This will provide clones adapted to the range of target planting environments, as well as yield improvement in performance and wood properties (Burdon and Aimers-Halliday, 2006).
To meet the demand for E. pellita wood in Indonesia, both for carpentry such as log timber, plywood and for other materials such as wood pulp, the strategy used by forest plantations is not sufficient only focused on developing clones to enhance plantation productivity and product uniformity. In this context, the determining of basic density is crucial in determining the best strategies for clonal breeding and testing and in predicting genetic parameter from developing the best clones (Osorio et al., 2001). Basic density is important in the value of timber and pulpwood also other material. It is highly correlated with major strength characteristics and pulp and paper properties (Wu et al., 2011b). Therefore, to increasing the growth rate of E. pellita, improving wood quality has become an urgent goal for genetic improvement.
The increasing volume and higher basic density at harvest are desirable traits for all stakeholders, especially in plantation forests. The process of clonal forestry starts with the creation of clone-based plantation forests which are carried out by selection, propagation, evaluation, and management (Wendling et al., 2014). Plantation companies must provide clones with a broad genetic base that can adapt to different environments. Genetic responses can be influenced by the environment which might cause a decrease in genetic advantage when selected simultaneously (dos Santos et al., 2016). Ideally, clonal forestry should be selected for stable performance across a variety of environments (combination sites) or no interaction between genetic by environment (G × E). The interaction between the genotype and environment can lead the selection clone of most productive genotype (clone) for planting at specific site because the most productive ones at one site may not be the most productive at another site. There are numerous strategies of minimizing interaction effects of G × E that become unstable clone in performance for example, by identifying genotypes exhibiting specific adaptation to specific environments (Cruz and Castoldi, 1991) or the stratification of a heterogeneous area into more homogeneous parts, referred to as breeding zones, where the genotype selection is performed separately in each zone or site (Santos et al., 2015). Error in selecting species at the designated site is usually the main factor that causes low growth of Eucalyptus sp. in North Sumatra (Latifah et al., 2014). Therefore, an understanding of environmental effects is very important in the selection of genetic material (Oliveira et al., 2018).
Growth and basic density of wood is a genetic control that can support improvement (Hung et al., 2014) as an effort to increase the productivity of E. pellita plantations in Indonesia. To determine performance, it is necessary to estimate genetic parameters such as heritability, expected gain, and genetic correlation (Hodge and Dvorak, 2012). The selection of genetic performance in the most accurate clone test is carried out at half-cycle age. Many studies have been carried out on Eucalyptus, both on its growth performance and its wood properties (Augustina et al., 2020; Harrand et al., 2009; Iswanto et al., 2021; Kien et al., 2010; Lukmandaru et al., 2016; Luo et al., 2012; Schulz et al., 2021). However, in the case of E. pellita in early growth in Indonesia, there are few reports in the literature regarding genetic parameters for growth and wood quality of clonal material. This information is very important for the next plantation in Indonesia to increase productivity and wood quality of the E. pellita. Therefore, this study aims to determine growth characteristics (height, diameter, and volume of the tree) and basic density and its genetic parameters (heritability, genetic gain and genetic correlation) on 3.5-year-old E. pellita planted in two different sites in Kutai Kartanegara, East Kalimantan, Indonesia. The results are expected to provide information for ongoing breeding, clonal selection, and deployment strategies of E. pellita clonal forestry for industrial forest plantation companies to support short cycle forestry with high productivity in Indonesia.
2. MATERIALS and METHODS
The clonal test was conducted at two locations, namely sites 1 and 2 located in Kutai Kartanegara Regency, East Kalimantan Province, Indonesia. The first clone is located at site 1 at WGS 1984/ UTM Zone 50M latitude coordinate: 9913706 (North) and longitude: 475419 (East) with an area of 2.5 ha and planting was carried out in May 2018. Meanwhile, the second clone test location is located at WGS 1984/ UTM Zone 50M latitude coordinates: 9937153 (North) and Longitude: 441947 (East) with an area of 2.5 ha and planting was conducted in May 2018, for detail in Table 1. The climate type of both sites was based on the Schmidt-Ferguson classification, while rainfall was ranged from 1,600–2,000 mm per year.
About 3.5-year-old E. pellita trees were used for the analysis of growth characteristics such as diameter, height, and volume of trees, as well as basic density. Both clone test sites were designed using a completely randomized block design. The number tested in both trials was 50 clones, divided into 5 blocks, and 5 trees/plot (Fig. 1). Tree diameter of 1.3 m from ground level was measured using a diameter band (phi band), while tree height was measured using a vertex hypsometer. Tree height and diameter were measured for all trees. For the measurement of basic density, samples were taken from 5 trees per clone using a Resistograph (1 sample tree/block), samples were selected randomly from each block. The Resistograph was assessed as a means of quantifying basic density in individual standing plantation trees (Downes et al., 2018). This instrument drives a 3 mm diameter needle through a tree at a set forward speed (feed speed) and rotation rate (rpm), while also measuring the resistance to turning. A 400 mm long trace can be taken from a single tree in less than 20 seconds, with tests conservatively showing that over 40 trees/hour were sampled (Downes et al., 2018). The trace represents a profile of resistance every 0.1 mm and indicates the radial variation in density (Gao et al., 2012). The results of resistograph measurement are traces or commonly referred to Resi traces. Then, the Resi traces were transferred to the resistograph tools Pro software and exported as text files. To get the basic density value, a model must be built in using the software. The Resistograph value as the needles travels through the disc was initially accounted for using a linear friction correction (Downes et al., 2018). Furthermore, the Resistograph predicted basic density for each sampling interval at every 0.1 mm was defined from the linear relationship between laboratory-measured density and mean drilling resistance amplitude for segment (slope and intercept). The mean predicted basic density value was calculated for each segment (Gendvilas et al., 2021). A company has developed a resistograph model for quantifying the basic density. This model was then used for all Eucalypt in the material breeding program. Previous study showed that the relationship between mean Resistograph value and core basic density was strong in Eucalyptus (Downes et al., 2018).
The stem volume (VOL) was calculated using the following formula by West (2009):
where: VOL is tree stem volume, HT is tree total height in m, DBH is tree diameter at breast height in cm.
Analysis of variance (ANOVA) was applied to evaluate differences in the growth traits and basic density. ANOVA was performed using individual clone data based on analysis of one location (single site) or a combination of two location (combination site/multiple site). In the combined ANOVA for determine the interaction between clone and environment, environmental variation parameters tested included: site 1 and site 2. The model used as follows by Leksono and Kurinobu (2005):
For each location:
where: Where Yijk is the phenotype observation of the kth individual tree in the ith block and the jth clone, μ is the overall mean, Bi is the block effect of the ith, Fj is the clone effect of the jth and Eijk is the error effect of the ijkth.
For the combination of locations:
where: Yijkl is the phenotypic observation of the lth individual tree in the ith location, the jth block and the kth clone, μ is the overall mean, Li is the location effect of the ith, Bj(L)i is the block effect of the jth in the ith location, Fk is the clone effect of the kth, FLki is the interaction effect of the clone kth with the ith location, Eijkl is the error effect of the ijklth. Based on the measurement carried out, data analysis was conducted using ANOVA. It was obtained through the general linear model procedure using plot mean data, while genetic variance and covariance were estimated using proc varcomp type 1 and REML (Residual Maximum Likelihood or Restricted Maximum Likelihood) in the SAS On Demand for Academics software. Average clone heritability on each location (A) and combination location (B), were determined by expressions by White and Hodge (1989); Zobel and Talbert (1984):
(A) Each location:
(B) Combination location:
where: H2 = heritability in a broad sense; σ2k = component variance of clone; σ2kb = component of the clone-block variance; σ2kl = component of the clone-location interaction variance; σ2kb(l) = component of in-location clone-block interaction variance; σ2e = component of error variance; L = number of locations; B = number of blocks per location; and N = number of ramets/tree per plot.
Expected gain (Pg) was calculated using the formula used by Luechanimitchit et al. (2017):
where: i = intensity of selection; P = SD of the phenotype; H2 = heritability and x̅ is the phenotypic mean of each trait. Genetic correlation results from calculating the covariance with the SD for two different traits. The genetic correlation was assessed using the formula proposed by Zobel and Talbert (1984):
where: rG = genetic correlation; σk (xy) = covariance component for properties x and y; σ2k(x) = component of clone variance for trait x; σ2k(y) = component of clone variance for trait y.
3. RESULTS and DISCUSSION
Clonal variations for growth parameters including diameter, height, and volume of trees, as well as basic density were evaluated in this study. Table 2 shows the mean values of diameter, height, and volume of trees, where the mean values in each site were 6.4–13.8 cm, 11.2–20.9 m, and 0.02–0.13 m3 in site 1; 6.0–14.6 cm, 7.9–20.5 m, and 0.01–0.15 m3 in site 2; as well as 7.0–13.6 cm, 10.2–19.8 m, and 0.02–0.13 m3, in the combination site, respectively. Ramadan et al. (2018) reported that tree height, diameter and tree volume of 4-year-old E. pellita clone planted in East Kalimantan were 18.6–18.8 m, 11.7–12.5 cm, and 0.082–0.098 m3, respectively. Tree height and diameter of 9-year-old E. pellita planted in North Sumatra, Indonesia were 16.7 m and 16.8 cm, respectively (Prasetyo et al., 2017). In Brazil, tree height and diameter of 8-year-old E. hybrid were 17.64 m and 14.43 cm (Oliveira et al., 2018). Tree height, diameter and volume of 3- and 4-year-old E. pellita planted in Central Java, Indonesia were 10.9 and 12.75 m; 7.79–8.91 cm, and 0.0296–0.0456 m3 (Kartikaningtyas et al., 2020). The tree height, diameter, and tree volume of 7-year-old E. hybrid planted in China were 17.93–18.82 m, 11.98–12.60 cm, and 0.093–0.106 m3, respectively (Wu et al., 2011a). In Brazil, growth characteristics of 4.5-year-old Eucalyptus grandis were 18.82–22.32 m and 15.59–19.87 for tree heigh and stem diameter (Harrand et al., 2009). In the present study, the average value of growth characteristics was lower comparing to the previous results (Kartikaningtyas et al., 2020; Oliveira et al., 2018; Prasetyo et al., 2017; Ramadan et al., 2018; Wu et al., 2011a). These differences are due to the differences of genetic materials, ages, site, and environment. Development of the clones is sensitive to the site condition and environment (Ramadan et al., 2018). Leslie et al. (2012) state that increasing of the tree growth is varies depend on the environment (soil, climate, and biotic factor) and genetic factor.
Table 2 also shows the mean values of basic density, where the mean values in each site were 409–547 kg/m3 in 1, 400–562 kg/m3 in site 2; and 407–555 kg/m3 in the combination site, respectively. Ramadan et al. (2018) reported that basic density of 4-year-old E. pellita planted in East Kalimantan was 450–465.2 kg/m3. Basic density of 9-year-old E. pellita planted in North Sumatra, Indonesia was 400–450 kg/m3 (Prasetyo et al., 2017). The basic density of 7-year-old E. hybrid planted in China were 404–427 kg/m3 (Wu et al., 2011a). Basic density of the present study was lower comparing result by Ramadan et al. (2018). However, the result is higher comparing results by Prasetyo et al. (2017) and Wu et al. (2011a) even though their samples were older. The differences of the basic density could be due to the different genetic material, sites, age, etc. Further research is needed to clarify the effect of tree age on the basic density in E. pellita.
The results of ANOVA are shown in Table 3, and significant differences were observed in all measured characteristics, among the 50 clones either on individual site or on combination sites, and between two sites. Similar results were found by Kartikaningtyas et al. (2020) which was tree diameter, height and volume showed significant differences between 15 clones of 3.8-year-old E. pellita, planted in Wonogiri, Central Java, Indonesia. Dhillon and Singh (2010) who discovered significant differences for tree diameter and heigth among the progenies of about 23 trees of E. tereticornis at the age of 3.5 years old from different sites in Punjab at Regional Research Station, India. Similarly, significant differences in growth characteristics and basic density among 19 clones were reported in E. hybrid at the age of 4.2-year-old in southern China (Wu et al., 2011b), as well as E. pellita at 4 years old in East Kalimantan, Indonesia (Ramadan et al., 2018). Hidayati et al. (2019) also found significant differences for growth traits (tree diameter dan height) and wood properties between 65 families of A. mangium at the age of 6-year-old in Indonesia. Moreover, this result is in line with Prasetyo et al. (2017) who reported that growth characteristics (tree diameter, height, dan volume) and basic density of 9-year-old E. pellita planted in North Sumatra, Indonesia were significantly different. In the present study, the results are also consistent with previous studies (Dhillon and Singh, 2010; Hidayati et al., 2019; Kartikaningtyas et al., 2020; Prasetyo et al., 2017; Ramadan et al., 2018; Wu et al., 2011b). Differences in the growth characteristics and basic density between clones is effect of genetic factor. It is provided a potential to increase wood productivity and quality of the E. pellita (Kartikaningtyas et al., 2020).
The differences in growth characteristics (height and diameter) and basic density were found between two sites (Table 3). This is due to the differences in the topography of the location, where site 1 has a rather steep topography while site 2 has a slightly slope (flatter than site 1; Table 1). Soil texture is also thought to have an effect on clone performance where site 1 is with clay stone and site 2 with sandy stone. Hardiyanto et al. (2021) stated that the soil influences productivity on tree growth. In general, site 2 show the better result of diameter and the basic density. It could be related to the relatively flat topography and soil type of sandy stone in site 2. However, the tree volume is the same for both of site (Table 2). Similar result of the growth characteristics between two site was also found by other researchers in another Eucalypt (Harrand et al., 2009; Wu et al., 2011a).
Table 3 shows the results of the interaction variation between clones and locations (G × E). Significant differences were found in all growth characteristics (diameter, height, and volume of trees) and basic density (Table 3). This indicates that there is genetic instability in all traits where the ranking of the best clones at site 1 is not necessarily the best clone at site 2. Khasa et al. (1995) reported that significant site × provenance interactions were found in Acacia auriculiformis and A. mangium. A. auriculiformis in Vietnam was also observed to have significant interactions between site × clone, hence, clonal development and different clones should be used for different sites (Hai et al., 2008). According to Luo et al. (2012), the selection of clones suitable for their environment will increase maximum results even though in practice it requires a significant investment. The significant genotype by environment interaction might be attributable to factors such as environmental effects, narrow genetic expression changes (Bouvet et al., 2005), and the diversity of environments examined (Hardner et al., 2010). In the present study, result of G × E interaction was similar to the previous studies (Hai et al., 2008; Khasa et al., 1995). This indicates that G × E effects are practically important for selection growth parameters and basic density. Consequently, different clones should be used on different sites and not in combination for the next breeding strategies for E. pellita. To manage G × E interactions, the best clones can be selected for specific sites to maximize deployment gains (Libby and Rauter, 1984). Ramburan et al. (2011) reported clay is generally considered an important site factor causing variation in cultivar responses in the interaction genotype by environment. However, in other species Lee et al. (2022) reported that the growth of Pinus densiflora conditions were found to have a significant effect in the relationship between the annual ring growth and climate also in Korean Pine was significant different between microfibil angle with climate (Kim et al., 2020).
Table 4 shows the estimates of heritability for growth characteristics and basic density of E. pellita clones, where the heritability values of tree diameter, height, volume and basic density in each site were 0.83, 0.86, 0.87, and 0.91 in site 1; and 0.90, 0.87, 0.90, and 0.93 in site 2, respectively. Ramadan et al. (2018) reported the heritability of tree diameter, height, volume and basic density to be 0.53, 0.66, 0.63, and 0.89, respectively for the 4-year-old E. pellita clones planted in East Kalimantan, Indonesia. Furthermore, heritability was reportedly 0.82, 0.82, and 0.95, for tree diameter, height, volume and basic density respectively in 3.5-year-old Eucalyptus camaldulensis planted in Ba Vi, North Vietnam (Kien et al., 2010). Wu et al. (2011b) also mentioned that the heritability among 19 clones in E. hybrid at the age of 4.2 years old in Shankou, Southern China was 0.86, 0.73, 0.83, and 0.74 for tree diameter, height, volume and basic density. Meanwhile, the estimated values for 8-year-old Eucalyptus urophylla × E. tereticornis clones planted in southern China were 0.91 for tree diameter and heigth, as well as 0.87 for basic density (Yang et al., 2018). The heritability results of growth and basic density were higher than those in previous studies (Kien et al., 2010; Ramadan et al., 2018; Wu et al., 2011b), hence, there is a potential for improving growth characteristics and basic density in E. pellita, through the breeding program. Additionally, the growth characteristics heritability in this study was lower than previous results from Yang et al. (2018) at the age of 8-year-old, suggesting their ability to increase with tree age.
The estimated heritability of basic density in this study is higher than that of growth characteristics. This result indicate that basic density exhibits genetic control and is one of the preferred traits for wood quality improvement. This result is in line with Pliura et al. (2007) who reported that the value of the basic density heritability was higher than growth characteristics. Therefore, basic density has strong genetic control compared to environmental factors. Heritability is a parameter that describes the number of parental traits passed on to the offspring and is very important because it is closely related to expected gain and tree breeding strategies to obtain genetic enhancement (Zobel and Talbert, 1984). The basic density, which has been shown to have high heritability, affects timber strength, machinability and hardness (Xiao et al., 2021). Generally, density in eucalypts has been reported to be under strong genetic control with individual heritabilities ranging between 0.4 and 0.84 (Borralho et al., 1992). Based on the results, E. pellita has good opportunities for genetic improvement of the examined traits to increase wood productivity and quality.
The expected genetic gain value which is the response to the selection was calculated based on the estimated value of heritability. Table 5 shows the expected gain for growth characteristics and basic density of E. pellita clones, where the values (%) of tree diameter, height, volume and basic density in each site by choosing the best 1–5 clones were 25%, 24%, 70%, and 11% in site 1; and 31%, 29%, 80%, and 13% in site 2, respectively. Meanwhile, in E. urophylla × E. grandis clones at 5 years old in Congo, the expected gain was 26% direct selection for height and 62% for volume by selecting the best two clones (Makouanzi et al., 2017). Ramadan et al. (2018) also reported that the expected gain of E. pellita at 4 years old in East Kalimantan, Indonesia was 31.8% selection for volume and 7.4% for basic density by selecting 1–7 best clones.
The genetic gain of measured parameters is in the range of the previous results (Makouanzi et al., 2017). Proper use of clones will increase the expected gain by 20%–25% compared to the use of seed origin even when carried out in a breeding population of the same location and silvicultural technique (Rezende et al., 2014). The high expected genetic gain in clonal forestry might be due to the large proportion of additive genetic variation, ability to capture all non-additive genetic variations and exploit genetic as well as environmental interactions (White et al., 2007). However, the expected gain of basic density in this study is lower compared to that of the growth characteristics. Similar results were also found in the 5-year-old E. camaldulensis clones where in the growth characteristics increased by 22%–32% which is relatively small compared to basic density (Kien et al., 2010). Despite of this difference, the basic density character displays strong genetic control as demonstrated by the high heritability value in Table 4. This indicates that parental selection can improve wood quality.
The next genetic parameter examined was the correlation between traits. It was assessed to determine the relationship between one trait and another. Genetic correlation has an essential role in tree breeding programs, especially to develop two different characters or characteristics based on the application of selection on one character (Zobel and Talbert, 1984). The value of genetic correlation (rG) between height and diameter showed a close and positive relationship, ranging from 0.93 to 0.96 (Table 6). A similar result was also found in a 5-year-old E. camaldulensis clone planted in Vietnam with a genetic correlation value of 0.85–0.90 (Kien et al., 2010) and a 9-year-old E. urophylla clone with a value of 0.87 (Kien et al., 2009). High genetic correlation between height and diameter of 0.94 was also found in a 4-year-old E. globulus clone in Tasmania (Silva et al., 2013). Furthermore, the 4.2-year-old E. hybrid in China had a genetic correlation between height and diameter of 0.90–0.91 (Wu et al., 2011b). The results are also in line with Harrand et al. (2009), which found that E. grandis at the age of 4.5 years and 8.5 years had a genetic correlation of > 0.90 in the growth characteristics of height, diameter and volume of trees. Meanwhile, the results are higher than those obtained by Ramadan et al. (2018) in a 4-year-old E. pellita which ranged from 0.71 to 0.79 in East Kalimantan, Indonesia.
The genetic correlation between height and basic density was 0.37–0.47; while for diameter and basic density, it was 0.23–0.33 as described in Table 6. Therefore, the relationship between growth characteristics namely height and diameter, as well as basic density is moderate ranging from 0.23–0.47. At site 1, the correlation between growth characteristics and basic density was 0.23–0.37, while site 2 showed the highest genetic correlation range, of 0.33–0.47 and 0.31–0.44 in the combination of sites. Basic density is positively correlated with many of the mechanical properties of solid wood. Then, to improve these properties, should identify clones which give the highest wood density on a particular site type if performance of the clone is unstable. According to Kien et al. (2010), the genetic correlation between tree height and basic density of 5 years old clones of E. camaldulensis was 0.01–0.21 in Vietnam. Ramadan et al. (2018) also reported that the genetic correlation between height and basic density in the E. pellita planted in East Kalimantan, Indonesia was 0.04–0.23. Furthermore, a negative weak genetic correlation (–0.03) between diameter and basic density was reported by Yang et al. (2018) for 8-year-old E. uropyhlla × E. tereticornis clones planted in China. There was no correlation between the diameter and basic density of E. globulus (Apiolaza et al., 2005). Wu et al. (2011b) also stated that the genetic correlation between height and basic density was –0.12, while that of diameter and basic density was 0.28 in E. urophylla at the age 5.9-year-old planted in Guangdong, China. Moreover, Hidayati et al. (2019), found highly positive correlations between the growth characteristics, but no significant correlations were established with wood properties such as pilodyn penetration and stress wave velocity of A. mangium at the age 6-year-old in Indonesia. The growth characteristics, including tree height, diameter, and volume, had a strong correlation with a value > 0.90. These results illustrate that in making the selection, only one trait, namely diameter, can be used. This is because by prioritizing the actual diameter, improvements will occur indirectly to other properties. However, correlations between all growth characteristics and basic density were moderate and it is potentially good results for clone selection. This results genetic correlations among growth characteristics (height, diameter, volume) and basic density showed positive moderate to highly significant value. It is interesting to note that none of the character had shown negative correlation with each other, thus suggested that these characters may be used to the advantage of the breeder for bringing improvement in these characteristics simultaneously.
4. CONCLUSIONS
A total of 50 clones of E. pellita from two different sites in Kutai Kartanegara, East Kalimantan, Indonesia was used in this study. Growth characteristics namely diameter, height, and volume of trees, as well as basic density, were significantly different among clones and sites. The results showed that there was an interaction between genetics and the environment which further will require separate clonal test locations and deployment populations for growth traits and basic density. Therefore, different clones should be used on various sites for the next breeding strategies for E. pellita. The heritability value of growth characteristics (tree diameter, height and volume) and basic density were high in each site, 0.83, 0.86, 0.87, and 0.91 in site 1; and 0.90, 0.87, 0.90, and 0.93 in site 2, respectively. It is suggesting that these characteristics are genetically controlled. This implies that E. pellita has great potential for genetic improvement to increase wood productivity and quality. The value of the expected genetic gain for growth characteristics was higher than that of basic density (tree diameter, height, volume and basic density in each site by choosing the best 1–5 clones were 25%, 24%, 70%, and 11% in site 1; and 31%, 29%, 80%, and 13% in site 2, respectively). Furthermore, the genetic correlation between growth characteristics showed a strong relationship ranging from 0.93 to 0.96, while that of growth characteristics and basic density was positive and significantly moderate ranging from 0.23 to 0.47. The results of the responses also indicate that using growth characteristics and basic density as selection traits together will be very beneficial to increase the productivity and wood quality of E. pellita clones.
