Early Growth Performance and Genetic Variation of Araucaria cunninghamii in Bondowoso, East Java Province, Indonesia

Araucaria cunninghamii is a coniferous tree that has high commercial value for tropical and sub-tropical regions and is used as an alternative raw material for the pulp and paper industry. In order to get a good yield from this species, a progeny test was established in Bondowoso, East Java, and the variation in growth was investigated. This trial included 6 populations, 80 families, and 4 replications of 4-tree row plots spaced 4 m x 2 m. A range of percent survival of 98.81%–99.41%, with an average of 99.20%, was reported for adaptability, showing no significant variations between populations or between families. Significant differences existed between populations and between families in terms of growth in height, diameter, and volume. The best diameter was from the Manokwari population (4.93 cm), while the best height and volume characteristics were from the Jayapura population (5.32 m; 0.0051 m3). Height, diameter, and volume characters have moderate to high heritabilities. The genetic and phenotypic relationships between height and diameter, height and volume, and diameter and volume features were all high and positive. With the family selected at a 30% proportion, the predicted genetic gain for height, diameter, and volume were 3.14%, 5.79%, and 0.00073%, respectively.


Introduction
One of the conifer species in the Araucariaceae family is A. cunninghamii Aiton ex D. Don.It natively occurs in Papua New Guinea, Papua Indonesia, and Queensland-Australia's natural forests [1].According to [2], A. cunninghamii naturally occurs in Papua in the areas of Sorong, Manokwari, Fak-Fak, Serui, Nabire, Wamena, and Jayapura.This species that exists in the tropical wild forests of Papua have not been managed and utilised by breeders and managers of plantation forest in Indonesia.This is because many Indonesian plantation forest managers and tree breeders are unaware of the potential, prospects, and economic significance of this species.According to [3], this wood can be used for a wide range of softwood applications, including plywood, veneer, panels, flooring, and woodworking.In addition, the sap has a high economic value.
A. cunninghamii should be honoured and maintained as an alternative raw material for the pulp and paper industry [4].Currently, the main supplier to the pulp industry is Acacia sp, whose productivity is decreasing due to root rot caused by Ganoderma sp [5,6].Due to the massive attack of the disease in Acacia sp plantations, plantation managers need to look for alternative plant species to replace Acacia sp [7].The type of A. cunninghamii as an alternative species is not widely known in the community, so it is necessary to explore more information including in an effort to obtain superior seeds to support this type of planting programme.A series of plant breeding activities need to be carried out, one of which is the development of offspring test plots, where progeny test plots are a form of plant breeding population that can later be selected gradually to produce superior progeny test and seeds.To perform selection appropriately, effectively and efficiently, it is necessary to obtain information on genetic parameters based on data from existing descent test plots.The success of plant selection depends on how much genetic variability exists in the genetic material to be selected [8].In tree breeding programmes, the result of genetic gain is the acquisition of superior seeds from selected seed sources, one of which is a seedling orchard [9].This study aims to determine the adaptability, growth variation, estimated heritability value, genetic correlation, phenotypic and genetic gain at the age of four years in the progeny test of A. cunninghamii.

Methods
The study was conducted in a first generation (F-1) A. cunninghamii progeny test plot in Bondowoso, East Java Province.The research forest is administratively located in Wringinanom village, Sukosari sub-district, Bondowoso district.The progeny test site has climate type B [10] with an average rainfall of 2,400 mm/year.The soil type is andosol.The site is flat with an average slope of 0-10%, located at an altitude of 800 m above sea level [11].The genetic material used to construct the progeny test came from 6 populations consisting of five Papuan natural populations (Fak-fak, Sorong, Serui, Wamena and Manokwari) totalling 60 families and one population from Queensland (CSIRO) totalling 20 families.The progeny test was designed following a randomised complete block design with 6 populations, 80 families, 4 blocks, 4 trees per plot with a spacing of 4m x 2m (Figure 1).The population data used in this study are presented in Table I.The measured tree height is the total tree height, while the stem diameter is measured at chest height (1.3 m).Data on tree height and stem diameter were used to calculate tree volume using the following formula [12]: where: V= volume of tree (m 3 ) D= stem diameter (cm) T= height (m) F= shape figures = 0.486 [13] Analysis of variance for plant adaptability was conducted using data on the average percentage of living plants per plot with the following linear model [14]: where: μ= the overall mean Bi= the effect of the i th replicate Pj= the effect of the j th populations Fk(Pj)= the effect of the k th family nested in the j th populations Eijk= the residual error with a mean of zero Yijk= the plot means of the k th family, j th populations, i th replicate For plant growth traits (height, stem diameter and tree volume), analysis of variance was conducted using individual data.The linear model used for the analysis was as follows: where: μ= the overall mean Bi= the effect of the i th replicate Pj= the effect of the j th populations Fk(Pj)= the effect of the k th family nested in the j th populations Bi*Fk(Pj)= the interaction effect of i th replicate and k th family nested in the j th populations Eijk= the residual error with a mean of zero Yijk= the plot means of the k th family, j th populations, i th replicate Calculation of variance components was obtained using mixed model analysis.If the treatment in the results of each analysis of variance showed significant differences, then further tests with Duncan Multiple Range Test (DMRT) at the 1% or 5% test level were carried out to determine differences in each treatment.The variance component of the family in the population was used to estimate individual (h 2 i) and family (h 2 f) heritability values with the formula according to [14]: where: h 2 = individual tree heritability h 2 = individual tree heritability V 2 f = variance between families-within-populations groups V 2 = variance between families and replications V 2 = variance error B = the mean of number of replications N = harmonic mean number of individuals per plot The family variance component (V 2 ) was assumed to be 1/3 of the additive genetic variance (V 2 A), as seeds were collected from naturally pollinated parent trees in natural forests where some seeds may be the result of kin mating.Estimated genetic (rg) and phenotypic (rp) correlations were calculated based on the following formula [14]. where: Covf (X.Y) = covariance of the two traits at the family level σ 2 f(x) = family-level variance components of trait (x) σ 2 f(y) = family-level variance components of trait (y) r g = genetic correlations where: Covf (X.Y) = phenotypic covariance of the two traits at the family level σ 2 f(x) = phenotypic variance components of trait (x) σ 2 f(y) = phenotypic variance components of trait (y) r p = phenotypic correlations The genetic gain (ΔG) from selection tested offspring was predicted by the formula according to [15]: ΔG = h 2 .i.V where: i = selection intensity [16] (V p) = standard deviation of phenotypes of a trait (h 2 i) = individual heritability of a trait

Results and discussion
One important piece of information from progeny testing is adaptability.The adaptability of plants in their new environment can be seen from the level of percent survival of plant [17].The results of observations of four-year-old A. cunninghamii plants have a variation in percent survival in the range of 98.81%-99.41%with an average of 99.20%.The best population survival rate is from the Fak-fak population, while the lowest population survival rate is from Manokwari (Table 2).The result of variance analysis showed that family did not significantly affect the level of live per cent in the population (p<0.05).Research elsewhere on the adaptability and percent survival of plants from ten provenances of A. cunninghamii reported by [18] that of the 240 individuals planted, 78% survived at the age of 29 years after planting, these results indicate that the plant has good adaptation and percent survival to its new location, considering this species is an exotic species.
The difference in percent survival rate between populations indicates that the origin of the population (Table 2) shows different adaptation traits as argued by [9] that variation of a trait in a tree species can occur between geographical areas and environmental factors related to soil type and fertility, maintenance intensity and climatic factors, especially rainfall, have a stronger influence on growth than climatic factors.Furthermore, it was reported by [15,19] that differences in the growing environment of population origin can be the main driver in the process of differences in genetic makeup due to local adaptation, where differences in genetic makeup will affect plant appearance on certain characters.
The results of measurements of height, diameter and volume in the four-year-old A. cunninghamii of progeny test in Bondowoso showed variations.Variation for height, diameter and volume traits with an average range of (4.98 m-5.32 m, 5.18 m; 4.72 cm-4.93 cm, 4.86 cm; 0.0045 m 3 -0.0051m 3 , 0.0047 m 3 .).The best rankings for height traits came from Jayapura, Serui, Wamena, Queensland, Manokwari and Fak-fak populations, for diameter traits from Manokwari, Jayapura, Wamena, Serui, Fak-fak and Queensland populations and for volume from Jayapura, Serui, Manokwari, Fak-fak and Queensland populations (Table 2).Plant growth in this study was faster when compared to the growth of Araucaria from other populations and in different places, such as Sao Paulo (Brazil) which showed that at the age of twenty-one years the average height was 6.77 m and diameter 9.95 cm, with an estimated tree volume of 0.011 m 3 [20].Meanwhile, another conifer (Pinus pinaster) in five-year-old test plants in southwestern France showed an average height of 3.23 m and a diameter of 4.7 cm [21].This provides information that up to the age of four years the growth of height and diameter in the progeny test of A. cunninghamii remains variable and is able to adapt well to the new environment in Bondowoso.To determine differences between treatments, the DMRT test was conducted on variations in height, diameter and volume growth between populations presented in Table 2.
The results of the DMRT test on the characteristics of height, diameter and volume showed a significant effect, thus at that age range the growth of plant height tends to be more dynamic than the growth of stem diameter (Table 2).At the age of four years, the best growth in height and diameter was sequentially from populations (Jayapura, Serui and Wamena); (Manokwari, Jayapura and Wamena).Variation in the growth of each population for height and diameter traits continued to increase with plant age, indicating that there was variation in growth between populations and within populations [22].The effect of population on the genetic variation of height and diameter traits showed very significant differences, high variation in height traits and significantly different among populations was also found in descent test studies conducted by [23] for A. angustifolia in Brazil, [24] for P. caribaea var.bahamensis in Brazil, [25] for Pinus radiata in New Zealand.This inter-population variation can be utilised to improve crop productivity.Increasing crop productivity can be done through breeding programmes through family selection and selection within families will increase genetic gain [9].At four years of age, the populations tested showed significant differences in the growth traits of height and diameter tested (Table 4).These results are an indication that height and diameter growth are strongly influenced by genetic factors, and provide an opportunity for selection, in order to improve genetic gain.High variation between families and populations, and families nested within populations is very favourable in the selection process [26] furthermore that there is no selection result without variation so that analysis of variance in an evaluation of test plant breeds is very important to do because it can determine how much variability between families tested.The differences among the families and populations tested indicate that within each individual tree there are variations between populations and between individuals [9].Early knowledge of variation in plant height and diameter traits is very important for plant breeding plots.Studies on conifers have reported that in general the most efficient age for selection is between 5-10 years for plant cycles between 25-50 years [27,28].Therefore, the results of the progeny test in Bondowoso provide preliminary information for selection and subsequent conversion to seedling orchards.
Estimated heritability values are important to know because it is a very decisive factor in the success of selection and tree breeding programmes.To determine the selection method and estimate the value of genetic gain from the right selection results, information on heritability values is needed which is the proportion of genetic variance to the total variance of the measured trait [15].At the age of four years, the estimated heritability values of individual and family heritabilities of height, diameter and volume traits in four-year-old A. cunninghamii progeny test plots are presented in Table 5.The estimated individual heritability values for height, diameter and volume traits were 0.22; 0.58 and 0.48, respectively, while the family heritability values for height, diameter and volume traits were 0.41; 0.79 and 0.73, respectively (Table 5).According to [29], the individual heritability of height and volume traits is moderate, while for diameter and volume traits it is high.The moderate to high criteria indicate that the genetic variation of A. cunninghamii is relatively high.The existence of genetic variation in the of progeny test of A. cunninghamii is likely due to the genetic material used to build the of progeny test plot comes from populations that are quite far apart (Table 1).The results of the estimated individual heritability values include moderate criteria in the study of A. cunninghamii conducted by [3] at various ages for height traits aged 5, 10, 13, 16 years (h 2 =0.17; 0.23; 0.30; 0.13) and diameter traits aged 10, 13, 16 years (h 2 =0.18; 0.16; 0.17).Similarly, the results of research on other conifer species (Pinus sylvertris) aged 10 years, the individual heritability value for tree height was 0.18 and stem diameter was 0.26 [30].The estimated value of individual heritability with moderate to high criteria reflects that the estimated value of the genetic variance component is a high proportion of the phenotypic variance, so that the contribution of genetic factors is quite high in controlling growth.
Genetic correlation is very important in tree breeding programmes, especially to improve two different traits based on the application of selection on one trait, with the hope of indirectly improving the other trait [9].The genetic correlation coefficient illustrates how much the relationship between traits is genetically close.In this study, the genetic correlation and phenotypic correlation between height and diameter were positive and high at 0.79 and 0.68, respectively.The high genetic and phenotypic correlation values indicate that an increase in plant height will be followed by an increase in stem diameter or vice versa with a degree of relationship of 0.79% and 68%, respectively.Genetic correlation and phenotypic correlation between height and volume also showed positive and high values of 0.99 and 0.87, respectively.Likewise, the genetic correlation and phenotypic correlation between diameter and volume also showed positive and high values of 0.98 and 0.91, respectively (Table 6).These results illustrate that the height trait has a strong positive influence on the diameter trait.From these results, it can be said that if selection is to be carried out, then selection is sufficiently based on one trait, namely the height trait, because by only prioritising the height trait, the selection can be made based on the diameter trait.Some results of research on eight-year-old A. cunninghamii of progeny test in Queensland reported that the genetic correlation between tree height and trunk diameter was 0.86 [28], of progeny test research on five-year-old Pinus banksiana, the genetic correlation between tree height and trunk diameter was 0.78 [31].In most tree species, the two traits always show a high genetic correlation, so that in volume estimation or calculation of other parameters using these two traits, it is sufficient to use only one of the traits.For the measurement of old or very tall trees, this information is especially valuable because the accuracy of height measurements has declined, so the diameter trait is widely used for volume estimation or other traits.This will make selection more efficient, both in terms of cost and time.However, the A. cunninghamii progeny test is still relatively young so the performance of genetic parameters is still unstable, so it is still necessary to study the correlation in the next few years when the plant is relatively older and the genetic gain is as expected.Estimated genetic gain is a response to selection, which is carried out to improve a trait in order to obtain an increase in yield from one generation to the next.The estimated genetic gain is determined by the heritability value, the amount of selection intensity (i) and the phenotypic standard deviation which is the square root of phenotypic variation [9].How much genetic gain can be achieved is closely related to the heritability value of each trait.A high estimated heritability value will result in high genetic gain and vice versa if the heritability value is low, the genetic gain is also relatively low [14].In this study, simulations to estimate genetic gain using proportional selection of the best families at 30%, 40% and 50% with selection intensities of 1.15, 0.93 and 0.84 [16].The estimated genetic gain for height, diameter and volume traits using 30% proportional best family selection was 3.14%, 5.79% and 0.00073% (Table 7).Estimates of genetic gain from the simulation of family selection conducted showed that the percentage of height traits is greater when compared to the trait of diameter (Table 7), from these results it appears that with the value of the genetic variance component and the greater heritability of a trait has a chance of genetic gain that can be obtained will be greater [32].Given that the age of the plant is still relatively very young, namely four years, where plant growth is still strongly influenced by environmental factors, the possibility of changes in the coefficient of genetic variance components and heritability when the age of the plant increases is still very possible to change [33], thus the opportunity for greater genetic gain can still be expected.

Conclusions
At the age of four years among the tested populations showed significant differences in the growth traits of height and diameter tested, this result is an indication that the growth of height and diameter is strongly influenced by genetic factors, and provides an opportunity to be able to make selection, in order to improve the genetic gain.Estimated heritability values with moderate to high criteria reflect that the estimated value of the genetic variance component is a high proportion of the phenotypic variance, so that the contribution of genetic factors is quite high in controlling growth.A combination of inter-family selection and within-family selection will be effective in improving the growth of this species.Improving growth environment conditions with regular plant maintenance should be done to optimise plant growth, stabilise heritability, so that genetic gain can be optimised.

Table 1 .
Seed sources information of progeny test of A. cunninghamii in Bondowoso, East Java

Table 2 .
Average, range, standard deviation and DMRT test results for survival, height, diameter and volume in 4-year-old A. cunninghamii progeny test

Table 3 .
Analysis of variance of survival in the 4-year-old A. cunninghamii progeny test

Table 4 .
Analysis of variance of height, diameter and volume in A. cunninghamii progeny test 4 years old in Bondowoso, East Java

Table 5 .
Estimated values of individual heritability and family heritability of height, diameter and volume in test plots of 4-year-old A. cunninghamii progeny test i Family heritability (h 2 )

Table 6 .
Estimated values of genetic correlation (rg) and phenotypic correlation (rp) among test plant traits 4-year-old progeny test of A. cunninghamii in Bondowoso, East Java

Table 7 .
Estimated genetic gain of growth in the progeny test of A. cunninghamii at 4 years of age in Bondowoso, East Java