PDF Download
PubReader
Export Citation
Email the Author
Share Facebook
Share Twitter
Cited by 0 Article
Metrics (View:199)
1. INTRODUCTION
The decline in the potential of Indonesia’s natural forests to fulfill timber needs has led to the development of a plantation resource to help meet this demand. Promoting the use of native species is essential to meet the timber demand, as these species are expected to be well-suited and adapted to local biotic and abiotic conditions (Thomas et al., 2014). It is also necessary to reduce the dependence of the wood industry, especially on members of the Dipterocarpaceae (Augustina et al., 2020). Subsequently, nyawai (Ficus variegata Blume belongs to the Moraceae family) is a native species that has prospects for supplying plywood, face veneer (Hendromono and Komsatun, 2008), packaging boxes, and materials for light construction (Sumarni et al., 2009). It is a fast-growing tree species that has attracted the interest of private companies since 2003 (Effendi and Mindawati, 2015) and is being developed as an ‘alternative’ plant for plantations (Susanti and Halwany, 2017). In Indonesia, the natural distribution of this species includes Kalimantan (Haryjanto and Hadiyan, 2014; Wahyuningtyas et al., 2022), Java (Nurtjahjaningsih et al., 2019), Sumatra (Nur’aini et al., 2013; Rofifah et al., 2021), Sulawesi (Pitopang, 2012), West Lombok, Papua (van Heist et al., 2010).
Implementing a selective breeding program is an effective way to increase plantation productivity and improve the genetic quality of trees (Nelson, 2023). Achieving effective selective breeding requires a thorough understanding of the heritability of growth and wood quality traits. Heritability is a statistical measure of the proportion of phenotypic variance explained by genetic variance (Waszak and Cieślik, 2016). Traits with high heritability can be targeted for improvement through selective breeding because they are more likely to be passed on to future generations. The correlations between traits is also important for plant breeding programs in order to create an understanding how one trait influences another through shared genetic factors (de Oliveira et al., 2015). However, in-depth knowledge in these areas is currently lacking for F. variegata.
Measures of the heritability of height and diameter (D) of F. variegata in progeny tests established in 2012 showed that the genetic control of these traits was weak in the early growth phase (Haryjanto et al., 2014). Accurate measurements of tree growth and wood quality at the time of harvesting are crucial from breeding and selection perspective. These measurements hold a high value for predicting the breeding values of tested genotypes. Wood density is regarded as the best single predictor of wood quality (Fundova et al., 2018). Many breeding programs prioritize it as an initial trait for improvement due to its correlations with stiffness, strength, and shrinkage (Hong et al., 2014). The specific gravity of F. variegata is 0.27, ranging from 0.20 to 0.43 (Sumarni et al., 2009), indicating that the wood lacks strength and improving wood quality is necessary to make the species more attractive as a timber species. Wood quality is the main concern when using fast-growing plantations, especially when producing solid wood (Ghani and Lee, 2021). Therefore, improving these properties will expand the utilisation of wood more widely (Laksono et al., 2023).
In breeding programs, the preferred approach for rapidly assessing wood properties in standing trees involves the use of non-destructive methods (Schimleck et al., 2019). The Pilodyn, originally developed to test for wood decay, is often used to assess wood density (Fukatsu et al., 2011). Meanwhile, stress-wave velocity (SWV) which measures the speed of sound waves in standing trees, has been used to measure the speed of sound waves in standing trees and has received considerable attention (Wang et al., 2001). Many studies have used SWV on a standing tree as a stiffness selection criterion (Fukatsu et al., 2015; Matheson et al., 2008) and the potential for prediction of modulus of elasticity (Ishiguri et al., 2008; Van Duong et al., 2022; Wang et al., 2001).
This study has two objectives: First, to investigate the inheritance of traits related to growth characteristics and wood quality. Second, to understand how these traits relate to each other, both genotypic and phenotypic correlations. The acquired information will be examined in the context of formulating suitable selection strategies for F. variegata breeding programs.
2. MATERIALS and METHODS
The study site was in Bantul, Yogyakarta Province, Indonesia (latitude 07°57′30″–07°57′54″S, longitude 110°26′07″–110°26′29″ E, altitude of 75–145 m above sea level). The slope of the site varied between 5°–30°, and the soil was identified as an Oxisol. The climate according to the Köppen classification is Af (tropical rainforest climate) to Am (tropical monsoon climate), or type C (medium wet) of the climate classification by Schmidt and Ferguson (1951), with a mean annual rainfall of 1,502 mm (Bappeda Kabupaten Bantul, 2011).
In this study, two progeny trials involving the first generation from 17 and 19 open-pollinated F. variegata trees were conducted in a half-sib trial established in December 2012. The mother trees were selected from two regions, and the trial was divided into two sublines situated 300 m apart. The experimental design of the progeny test was a randomized complete block design. Subsequently, the provenances of West Nusa Tenggara (WNT) represented 17 families, 5 non-contiguous tree plots in six replicates and Cilacap-Pangandaran (C-P) represented 19 families, 4 non-contiguous tree plots in six replicates. The planting space was 5 × 5 m.
D was measured at 1.3 m above ground, tree height (H) was measured from the ground to the tip of the tree, and stem volume (V) was measured using the equation of Qirom and Supriyadi (2013) for 10-year-old trees:
Where:
V: stem volume (m3), D: diameter at 1.3 m above ground (cm), H: tree height (m), a: –9.22846, b: 1.7456, c: 0.9759.
Stem SWV was measured using a commercially available handheld stress-wave timer (FAKOPP Microsecond Timer, FAKOPP Enterprise) according to previously described methods (Hidayati et al., 2013b; Ishiguri et al., 2007). The start and stop sensors were positioned at 150 and 50 cm from ground level, respectively. Using a small hammer, the start sensor was struck to generate a stress wave. When the stress-wave was received at the stop sensor, the travel time between the two sensors was recorded. Ten measurements of stress-wave propagation time were taken for each tree, and the mean value was calculated. The formula for the calculation of the stress-wave propagation time per unit length was described by Wang etal. (2000, 2004).
Wood density measured indirectly by Pilodyn penetration (P) was carried out using Pilodyn tester (strength of spring, 6 J; D of pin, 2.5 mm, Proceq) at 1.3 m above ground level at three positions for each tree (Hidayati et al., 2013b, 2019; Ishiguri et al., 2008; Wu et al., 2010) and the mean value was calculated. The bark at the measurement points was removed prior to measurement.
Analyses of variance (ANOVA) were carried out to evaluate the differences in D, H, V, P, and SWV among the tested families based on an individual-tree following the linear model:
Where Yijk is the individual tree observation, μ: the overall mean, Bi: fixed effect of the i-th block, Fj: random effect of the j-th family associated with the average genetic effects of the open-pollinated families, εijk: the within-plot error.
Family heritability (h2f) was calculated using the following formula of Zobel and Talbert (1984):
Where σ2f: the family component of variance, σ2w: the within-plot error component of variance, TR: the harmonic mean number of trees per plot by replication.
The coefficient of additive genetic variation (CVA) is a mean-standardized index of the evolvability of a trait (Hill, 2010; Houle, 1992). It was described by Haapanen (2001):
Where CVA: coefficient of additive genetic variation, σ2A: additive genetic variance. The additive genetic variance on a single-site basis was estimated by multiplying the among-family component of variance by four (σ2A = 4σ2f; White et al., 2007; Zobel and Talbert, 1984); χ: population mean for the trait.
Genetic correlation between pairs of traits (rGxy) was calculated following Zobel and Talbert (1984):
Where σf(xy) is the additive genetic covariance component between traits, and σ2f(x) and σ2f(y) are the additive variance components for traits x and y respectively.
The phenotypic correlation between pairs of traits (rP) was calculated as given by Falconer (1989):
Where covP is the covariance of the two traits, and σPx and σPy are the SD for traits x and y respectively.
3. RESULTS and DISCUSSION
The mean, range of values, SEM for D, H, V, SWV, and P of the 17 WNT families and the 19 C-P families are shown in Table 1. The mean growth traits (D, H, V) of WNT families were higher than those of C-P families. Subsequently, P was smaller in WNT than C-P, 29.62 mm, and 33.57 mm, respectively. SWV of WNTs was smaller than that of C-P, 2.03 km/s and 2.08 km/s, respectively. Significant variations (1% level) were observed for all measured characteristics among WNT families. Additionally, differences in D, V, and SWV within the C-P families were found to be significant at the 5% level, as shown in Table 2.
The estimated CVA, variance components, and family heritability estimates are shown in Table 3. The estimated CVA depending on the analyzed traits, and growth traits was higher than wood quality. In the WNT, CVA ranged from 16.47% to 46.61% for the growth traits (D, H, and V) and it was 4.68% (SWV) and 6.20% (P); in C-P, CVA ranged from 5.16% to 24.02% for the growth traits (D, H, and V) and it was 2.58% (SWV) and 2.54% (P).
In the WNT, the estimated contribution of the family component of variance (σ2f) to the total phenotypic variance ranged from 15.80% to 21.47% for the growth traits (D, H, and V). For wood quality, it was 8.95% (P) and 10.43% (SWV). In C-P, the estimated contribution of the family component of variance (σ2f) to the total phenotypic variance ranged from 0.46% to 3.95% for the growth traits (D, H, and V). For wood quality, it was 2.18% (P) and 3.04% (SWV).
In the WNT, family heritability (h2f) for the growth traits (D, H, and V) ranged from 0.81 to 0.86, and for wood quality, it was 0.69 (P) and 0.73 (SWV); in C-P, family heritability for growth traits ranged from 0.26 (H) to 0.49 (V) and for wood quality, it was 0.32 (P) and 0.39 (SWV).
In both WNT and C-P families, highly positive and significant phenotypic correlations were observed for the growth traits (D, H, and V). The correlations ranged between 0.811–0.969 for WNT families and 0.631–0.922 for C-P families, as shown in Table 4. Subsequently, no phenotypic correlations were observed between growth traits (D, H) and wood quality (SWV and P). Significant negative phenotypic correlations were observed between SWV and P in both families, –0.281 (WNT) and –0.125 (C-P), respectively.
Strong estimated positive genetic correlations were observed between D and H (rG = 0.918), and (rG = 0.765) in WNT and C-P families, respectively. Between other growth traits (D and V) and (H and V) they were low (0.047) and moderate (0.502), respectively. Estimated negative genetic correlations among growth characteristics (D, H, V) and P in the WNT families and C-P families were –0.179 to –0.731 and –0.375 to –0.664, respectively.
Estimated negative genetic correlations observed between D and P were strong (rG = –0.731) and moderate (rG = –0.375), and between P and SWV were moderate (rG = –0.460) and strong (rG = –0.786) in WNT families and C-P families, respectively.
In this study, the growth of F. variegata was comparatively lower than that observed in other sites. For instance, in Riam Kiwa (South Kalimantan), F. variegata planted there exhibited a mean annual increment (MAI) in D of 2.10–2.36 cm/yr (Fitriani, 2011), whereas in this study the MAI was 1.74 cm/yr and 1.63 cm/yr for the WNT and C-P populations, respectively. This difference may be attributed to the tropical climate in Riam Kiwa (South Kalimantan) characterized by high humidity and consistent rainfall throughout the year (Yasin et al., 2020). F. variegata, being a tropical tree, thrives in seasonal wet tropical rainforests, often found along watercourses (Spencer et al., 1996).
The SWV values of F. variegata (2.03 km/s and 2.08 km/s for WNT and C-P families, respectively) may be relatively lower than other tropical tree species. Subsequently, the SWV values were at 2.83 km/s for Gmelina arborea 5-year-old (Hidayati et al., 2017) and 3.08 km/s for 13-year-old Falcataria moluccana (Ishiguri et al., 2007). This indicates that G. arborea and F. moluccana are stiffer than F. variegata. This difference may be due to differences in the anatomical characteristics and basic density of these species.
P serves as an indicator of wood density; higher values correspond to lower wood density, and vice versa (Kha et al., 2012). The P value was higher than other tropical tree species. Specifically, the P value was at 25.9 mm for 4.5-year-old in white jabon (Neolamarckia cadamba; Chaerani et al., 2019), 18.7–27.6 mm for 12-year-old Tectona grandis (Hidayati et al., 2013b), 15.7–18.9 mm for 6-year-old Acacia mangium (Hidayati et al., 2019). This indicates that F. variegata has a lower wood density compared to these species.
In this study, a comparison of the growth and wood quality traits between the two populations (WNT and C-P) reveals the following: the WNT families exhibit superior growth and wood density compared to the C-P families. Wood density, as indicated by P, is lower for WNT families than for C-P families. Lower penetration shows denser and stronger wood and is negatively correlated with P (Wei and Borralho, 1997; Wu et al., 2010). Conversely, for wood stiffness, SWV in the WNT families is slightly smaller than the C-P families (2.03 and 2.08, respectively). Although higher values of SWV show greater stiffness (Ishiguri et al., 2007; Wang et al., 2001), it does not always correlate with smaller P as observed in T. grandis (Hidayati et al., 2013a; Seta et al., 2021), A. mangium (Hidayati et al., 2019). In summary, it can be concluded that growth characteristics and wood density are superior in WNT compared to the C-P family. However the wood stiffness are similar for WNT to CP.
With the exception of height (H) and P in the C-P families, all measured traits between the two populations (WNT and C-P) were statistically significantly different. These significant differences in characteristics indicate that these traits are influenced by genetic factors rather than the environment. Similar results were reported in other studies, where significant differences were observed in growth characteristics (D, H) and wood quality (SWV and P) among 65 families of A. mangium (Hidayati et al., 2019), as well as in D, H, and P among 105 families of white jabon (Chaerani et al., 2019). The significant differences observed among the families in these traits suggest the possibility of selecting superior families to support breeding programs.
The CVA is recommended for reporting the results of studies on adaptive variation, as it provides a better guide than heritability to the long-term evolvability of a trait (Houle, 1992). This CVA serves as a valuable indicator for assessing the long-term evolution of a trait and its potential for adaptation under changing environmental conditions (Houle, 1992). This reflects the potential for a given trait to respond to selection pressures in a given population, the larger the coefficient, the greater the evolvability of the trait (Cheung, 2020). In this study, CVA in growth traits (D, H, V) was higher than for wood quality (SWV and P) in WNT and C-P families (Table 3). This suggests that growth traits could be hypothesized as being more related to species adaptation. Additionally, the CVA of V was higher than that of H or D in both populations. This result may reflect the fact that volume is a function of both H and D. Specifically, the CVA of D was higher than that of height in both populations, at 19.84% and 16.47% for WNT, and 9.58% and 5.16% for C-P, respectively. This finding contrasts with a previous study on Pinus elliottii, where the CVA for height (H) was higher at 5.06% compared to the CVA for D at 2.99% (Lai et al., 2017). Therefore, this study suggests that there is a greater scope for selection of D among families compared to height.
Family heritability (h2f) was higher in WNT families compared to C-P families for all traits (Table 3). Additionally, the proportion of the family component variance (σ2f) to the total phenotypic variance was consistently greater in WNT families for specific traits. For instance, in the D trait, the contribution of WNT families to the family component variance was 15.80%, whereas for C-P families, it was 3.95% (Table 3). The genetic variation in WNT families resulted in higher heritability values, indicating that the variation in these traits was primarily due to genetic differences among families (Lai et al., 2017; Ranjan and Gautam, 2020). This suggests a greater additive genetic effect passed on to offspring from WNT families compared to C-P families, emphasizing the significance of genetic variation in breeding programs (Mihai et al., 2014). The observed high heritability implies that selective breeding is more feasible for the considered traits.
Heritability and genetic correlations are important factors influencing the effectiveness of breeding programs, particularly in forest trees. A notable and crucial adverse genetic correlation is observed between tree productivity and wood density, commonly characterized by a strong negative relationship (Klápště et al., 2022). Meanwhile, heritability and genetic correlation are important genetic parameters for the effectiveness of indirect selection. When the genetic correlation between traits is high but heritability is low, indirect selection tends to be less effective in producing offspring with the desired parent traits. Low heritability means that the genetic influence on the traits is not strongly transmitted to the next generation, diminishing the impact of indirect selection. High family heritability was found in the WNT population for the D and P at 0.81 and 0.69, respectively (Table 3), and the strong genetic correlation of D and P rG = –0.731 (Table 4). This indicates that indirect selection of the D trait in the WNT population will improve wood quality.
In this study, family heritability (h2f) for the growth characteristics was higher than for wood quality in the WNT families, and except for H, also for the C-P families. In the WNT families, the family heritability of growth traits (D, H, V) ranged from 0.81 to 0.86, while the family heritability of wood quality traits (SWV, P) ranged from 0.69 to 0.73. Meanwhile, the family heritability in the C-P families ranged from 0.46 to 0.49 and ranged from 0.32 to 0.39 for growth (D, V) and wood quality (SWV, P) respectively, and other studies confirm this findings. The family heritability for height and P were 0.69 and 0.46, respectively in white jabon (Chaerani et al., 2019); in A. mangium, family heritability for height and P were 0.37 and 0.25 respectively (Hidayati et al., 2019). This indicates that growth traits are passed on to offspring more than wood quality traits.
The study showed that high heritability can also be obtained by using a limited number of families, for example, 17 families for WNT and 19 for C-P. The result was similar to in Quercus serrata with a family heritability of 0.80 with 17 families (Kang et al., 2007). However, it is important to note that high heritability does not necessarily mean that the trait is entirely determined by genetic or that it cannot be influenced by environmental factors. Subsequently, heritability estimates are specific to the population being studied and the range of environments and genetic variations present in that population. Traits with high heritability in one population may exhibit lower heritability in another population with different genetic and environmental factors.
Phenotypic and genetic correlations play important roles in tree improvement programs. Subsequently, phenotypic correlations help identify relationships between observable traits, while genetic correlations inform breeders about how they can prioritize the selection of certain traits, knowing that improvement in one trait may have a positive or negative effect on others. By considering both phenotypic and genetic correlations, breeders can formulate comprehensive strategies aimed at simultaneously enhancing multiple traits. This method allows for the development of forest plantations with desired outcomes, taking into account the interplay and relationships between different traits in the breeding process.
This study showed highly significant phenotypic correlations among growth traits (D, H, V) in both WNT families (0.811 to 0.969) and C-P families (0.631 to 0.922; Table 4), consistent with prior results observed in T. grandis (Hidayati et al., 2013a). Positive correlations between D and H were similarly reported in A. mangium (Hidayati et al., 2019), and Eucalyptus urophylla (Wei and Borralho, 1997), suggesting a strong interdependence between growth characteristics, with their values exhibiting predictable simultaneous changes.
There was no phenotype correlation between wood traits (SWV and P) and growth traits (D and height; Table 4). Similarly, no significant phenotype correlation was observed between growth characteristics (D and H) and SWV in teak (Hidayati et al., 2013b). This finding aligns with previous results on A. mangium (Hidayati et al., 2019), but in contrast to G. arborea, which exhibits a positive significant phenotype correlation between D and SWV (Hidayati et al., 2017). In this study, wood traits appear to be independent of growth traits, suggesting that families with superior characteristics for both growth and wood traits can be selected.
Positive genetic correlation between growth traits (D, H, V) ranged from 0.180 to 0.918 in the WNT families and ranged from 0.047 to 0.765 in the C-P families (Table 4). This indicates that these traits tend to be influenced by similar genetic factors. In other words, individuals with favorable genes for one of these traits are likely to also have favorable genes for the others. These results are consistent with other studies, such as those conducted on Populus ussuriensis Kom (Jin et al., 2019), P. elliottii (Lai et al., 2017) and Eucalyptus pellita (Fadwati et al., 2023). These traits are interconnected at the genetic level, and improving one trait can lead to improvements in the others.
P is closely related to the wood density at the outer part of the stem (Wessels et al., 2011; Wu et al., 2010). The estimated negative genetic correlations between P and D were strong (rG = –0.731) in the WNT families and moderate (rG = –0.375) in the C-P families (Table 4). Meanwhile, the genetic correlation of SWV with growth traits (D and H) of both ENT and C-P was both positive and negative, but the correlation was weak (rG = –0.194–0.144; Table 4). The strong and negative correlation observed between P and D implies that as D increases, there is a corresponding decrease in P. This association indicates that increasing D is linked to low P, which, in turn, is indicative of high wood density. This finding is consistent with previous study on white jabon (Chaerani et al., 2019), and E. urophylla (Kien et al., 2008). Some studies have shown an unfavourable relationship between the two traits found in Norway spruce (Nguyen et al., 2022), Catalpa bungei (Xiao et al., 2021), T. grandis (Hidayati et al., 2013b), and E. urophylla (Wei and Borralho, 1997), making it difficult to improve growth and wood quality at the same time. This study shows that the correlations is favourable for breeding both traits (growth and wood quality) simultaneously.
The genetic correlation between P and SWV was moderate (rG = –0.460) to strong (rG = –0.786) negative in the WNT and C-P families, respectively. In other words, selecting trees for increased P (which may be associated with wood density) could lead to a decrease in SWV and vice versa. Subsequently, higher SWV corresponded to lower P values, indicative of denser and stronger wood. This finding is consistent with previous studies on E. urophylla (Wei and Borralho, 1997), and Larix kaempferi (Fukatsu et al., 2015). The result indicates that P was a reliable indirect measure of wood density in F. variegata.
4. CONCLUSIONS
The genetic study on the growth and wood characteristics of F. variegata in Yogyakarta Province, Indonesia aimed to investigate the inheritance of traits related to growth characteristics and wood quality; and to understand how these traits relate to each other, both genotypic and phenotypic correlations. The results obtained are summarized as follows:
-
Significant differences were observed between all measured characteristics in WNT families, except for height, and in the C-P families, except for height and P, indicating that these characteristics were influenced by genetic factors.
-
The family heritability of all traits was much higher in the WNT than in C-P families, demonstrating (1) much higher variability of all traits in the WNT was due to genetic differences among families; (2) greater additive effect passed on to offspring from WNT families than C-P families.
-
Positive genetic correlations between growth traits (D, H, V) were found in the WNT and C-P families. The observed moderate to strong negative genetic correlation between D and P, as well as between P and SWV, suggested that using D as a selection criterion could be a suitable breeding strategy for F. variegata. This method allows for the simultaneous genetic improvement of both growth and wood quality traits in the species.
