ESTIMATES OF GENETIC GAINS FOR GROWTH TRAITS IN YOUNG PLANTS OF Eucalyptus urophylla

The objective of this study was the selection of superior genotypes for growth traits, correlating them to initial height growth in the fi eld, at age eight months. A random block design was used in the nursery, with eight clones, three replicates and four plants per plot. And a random block design was also used in the fi eld, with eight clones, four replicate blocks and nine plants per plot. Data being analyzed in the nursery at age 120 days included: height of fi eld seedling, at age eight months (Hc), height of nursery seedling (Hm), root collar diameter (Dc), shoot diameter (Db), shoot dry matter (PMSPA), root dry matter (PMSR), total dry matter (PMST), ratio of shoot dry matter to root dry matter (PMSPA/PMSR), Dickson quality index of root collar diameter (IQD-Dc), Dickson quality index of shoot diameter (IQD-Db). Analyses of variance showed that signifi cant genetic differences exist among clones for all traits and, given the high heritability values found, the estimated genetic gains were generically very high. As for predicted indirect genetic gain, selection in nursery seedlings for Dc, PMSPA/PMSR, IQD-Dc provided the highest values of indirect gain in fi eld seedling height.


INTRODUCTION
Species of genus Eucalyptus play an important economic, social and environmental role for Brazil.Up until 2007, the estimated area reforested with Eucalyptus was 3,771,867 ha (ABRAF, 2008), expected to produce enough timber to supply mainly the Brazilian industrial sector, in particularly cellulose pulp production and charcoal-fuelled steel metallurgy.
The successful use of Eucalyptus species is largely due to the following aspects: i) dozens of species are easily adaptable and grow rapidly under various Brazilian ecological conditions, ii) these species are the most commonly used in large-scale production of short-fi ber pulp and wood charcoal, iii) they have the best potential for use in large-scale clonal silviculture, helping obtain a more homogeneous raw material, in amounts and with the quality to best suit each specifi c industrial use, iv) the considerable inter-and intra-specifi c variability of genus Eucalyptus, expressed by its diversifi ed wood properties, allows its use in a wide range of technological processes.
Given the expansion of reforested areas in Brazil and the increasing demand for timber by the Brazilian forest-based industry, more and more forest planting is expected to be carried out in all seasons of the year.

Godoy, T. G. & Rosado, S. C. da S.
And this will require the use of seedlings with superior physiological quality so as to increase tolerance to environmental stresses, including stress when planting is conducted other than in the rainy season or when unexpected rainfall shortage looms during the rainy season.
One of the major concerns in silviculture is taking healthy, vigorous seedlings to the fi eld, as they will have to survive, develop and grow under biotic and abiotic stress conditions once planted.It is thus crucial to conduct genetic improvement programs seeking selection of material that will, in the nursery stage, present seedlings with quality traits and thus help achieve the desired success in fi eld conditions.
The objective of this work was to select superior genotypes for growth and quality traits of seedlings from clonal E. urophylla, and to evaluate the effi ciency of seedling selection in the nursery in relation to their growth response in the fi eld, at age eight months.

MATERIAL AND METHODS
In this experiment clonal seedlings were used, as produced by the minicutting technique (vegetative propagation in the nursery of Carvovale Ltd.).The forest nursery owned by Carvovale Ltd. is located in the municipality of Taiobeiras, in Vale do Jequitinhonha region, at coordinates 42º14'10" west longitude and 15º48'30" south latitude.According to Koppen classifi cation, the predominating local climate is dry, continental, with average annual precipitation less than 1,000 mm and average maximum temperatures around 34ºC (COMPANHIA DE PESQUISA DE RECURSOS MINERAIS -CPRM, 2004).
Seedlings were produced in 50 cm 3 tubes and kept in plastic trays with 96 cells 80 cm above ground level.The substrate used was composed of 40% vermiculite, 30% scorched rice husk, 10% coconut husk fi ber, 20% slash pine bark and needles, and 5.5 pH.The management practices adopted were similar to those used in commercial scale production by the providing company, with fertilization being done via irrigating water in the initial, growth and hardening off phases, and dosages being applied per cubic meter of substrate.
In the nursery, eight clones were assessed using a random block design with three replicates and four plants per plot, to a total of 96 seedlings, at age 120 days.In the fi eld, a random block design was also used consisting of eight clones with four replicates and nine plants per plot, to a total of 288 plants arranged with 3 x 2 m spacings, at age eight months.
Seedling height data were collected in the fi eld, at age eight months.In the nursery, the following data were collected, at age 120 days: seedling height (Hm), root collar diameter (Dc), shoot diameter (Db).Shoot dry matter (PMSPA) and root dry matter (PMSR) were obtained in a laboratory.The following data were also obtained: total dry matter (PMST), ratio of shoot dry matter to root dry matter (PMSPA/PMSR), Dickson quality index of root collar diameter (IQD-Dc), Dickson quality index of shoot diameter (IQD-Db).
In order to measure seedling height in the fi eld, a measuring tape marked in cm was used.For shoot height in the nursery, a ruler marked in cm was used.For measurements of root collar diameter and shoot diameter, a caliper marked in mm was used.The root portion of each seedling was then carefully washed, trying to keep all roots intact.The vegetal materials were crosscut at root collar level, separated into shoot and root portions, placed in labeled paper bags and taken to a forced air oven at 70ºC, to a constant weight.They were then weighed using an analytical scale accurate to 0.01 g to determine shoot dry matter (PMSPA) and root dry matter (PMSR).
The Dickson quality index of root collar diameter (IQD-Dc) and Dickson quality index of shoot diameter (IQD-Db) were calculated according to Dickson et al. (1960).
Nursery and fi eld data were subjected to analysis of variance, following procedures described in program GENES (CRUZ, 2001).The statistical model of choice was: where: Y ij is the mean value observed in the i th clone of the j th block; μ is the grand mean; g i is the effect of the i th clone (fi xed effect); b j is the effect of the j th block (random effect); ε ji is the experimental error.
The analysis of variance summary and expected mean square values for phenotypic, genotypic and environmental parameters are provided in Table 1.
Estimates of variance components, values of coeffi cient of genotypic determination (h 2 ) and coeffi cient of variation were obtained according to expressions described by Cruz and Carneiro (2003).Gains were predicted by selecting one clone in four being assessed (25% selection), normalizing selection intensity at 1.400 (COTTERILL; DEAN, 1990).

RESULTS AND DISCUSSION
From Table 2 data, a significant difference is observed between clones, by the F test (P ≤ 0.01), for all growth traits.This indicates the possibility of clone selection for improvement of growth traits in the nursery, and height growth in the fi eld after planting.
Table 3 provides predicted genetic gains to be attained in all growth traits from direct selection of nursery seedlings, as well as indirect gains in initial fi eld growth, at eight months after planting.
In forest improvement programs, correlations are used to assist selection involving two traits simultaneously, in other words, selection in one may have refl ections, positive or negative, on another.Where refl ections are positive, gains are directly obtained in the selected trait and indirect gains are obtained in the correlated trait.However, in order for this to happen, it is necessary that the genotypic correlation (r g ) among the relevant traits be relatively high.
From the above correlation, a genetic material may be obtained capable of simultaneously gathering various favorable attributes (CRUZ; REGAZZI, 2001).On the other hand, where a trait is negatively correlated to another, one should be cautious that selection in one trait does not cause undesirable changes in another.
Considering a selection intensity of one clone in four, that is, 25%, the growth trait showing the smallest direct gain (GD%) was nursery seedling height (Hm), with a GD of 10.0%.The highest estimate of expected GD (39.7%) was found for the Dickson coeffi cient index using shoot diameter (IQD-Db).
As for estimates of expected indirect genetic gains GI (%) in fi eld height growth (Hc), the most effi cient seedling growth trait, at age 120 days, was root collar diameter (Dc), with a GI of 10.6%, revealing indirect selection efficiency (ESI) of 84.4% (Table 3).Other important growth traits included PMSPA/PMST and IQD-Dc, with ESI values of 71.0% and 65.0% respectively.These ESI values indicate that PMSPA/PMSR, IQD-Dc and Dc constitute the main growth traits indicative of seedling quality in this clonal population.
According to Falconer (1987), indirect selection constitutes an important strategy to improve low heritability traits and/or traits considered difficult to measure.
In this specifi c study, these higher ESI values occurred due to the fact that the heritability of initial height growth in the fi eld showed a heritability value relatively lower than growth traits of most effi cient seedlings and/or due to higher values of genotypic correlation.From Table 3 data it was noted that the highest genotypic correlation coeffi cient (r g = 0.8) found was between Hc and Dc, followed by r g = 0.7 and rg = 0.6 for PMSPA/PMSR and IQD-Dc respectively.
Considering the three most important traits only regarding indirect gains expected for initial growth of fi eld seedlings, it was noted that two clones, namely ITA-063 and ITA-069, showed the greatest height values in the fi eld eight months after planting, 2.8m and 2.7m respectively (Table 4).Table 4 data shows that these clones also showed the highest values of root collar diameter, 3.23 mm and 2.69mm respectively.Considering other correlations, it was noted that PMSPA/PMSR traits of clone ITA-069 showed the highest value (2.543).For clone ITA-063 the same ratio was 2.083.For IQD-Dc, clone ITA-063 showed the highest value, 0.213.

CONCLUSIONS
-Signifi cant differences exist among clones, for all traits being studied.
-The genetic control of traits is relatively high.
-A high correlation was found between initial fi eld height and root collar diameter (Dc), ratio of shoot dry matter to root dry matter (PMSPA/PMSR) and Dickson quality index of root collar diameter (IQD-Dc) of nursery seedlings.
Estimates of genetic gains for growth traits ...
between blocks; c: number of clones; r: number of blocks.

Table 1 -
Analysis of variance summary and expected mean square values, for phenotypic, genotypic and environmental parameters.

Table 2 -
Analysis of variance and estimate of population parameters for growth and morphological traits of nursery seedlings and for height growth of young fi eld plants, at age eight months, of E. urophylla clones.Análise de variância e estimativa dos parâmetros populacionais para caracteres morfológicos e de crescimento de mudas no viveiro e de crescimento em altura de plantas jovens no campo, aos oito meses de idade, de clones de E. urophylla.
e : variation index.

Table 3 -
Estimated direct gain (GD), genetic correlation (r g ), estimated indirect gain (GI) and indirect selection effi ciency (ESI) for growth traits of seedlings and height of young fi eld plants, at age eight months, of E. urophylla clones.Previsão de ganho direto (GD), correlação genética (r g ), previsão de ganho indireto (GI) e efi ciência de seleção indireta (ESI) para caracteres de crescimento de mudas e altura de plantas jovens no campo, aos oito meses de idade, de clones de E. urophylla.Dc: root collar diameter; Db: shoot diameter; PMSPA: shoot dry matter; PMSR: root dry matter; PMST: total dry matter; PMSPA/ PMSR: ratio of shoot dry matter to root dry matter; IQD-Dc: Dickson quality index using root collar diameter and IQD-Db: Dickson quality index of shoot diameter, Hm: height of nursery seedling.Estimates of genetic gains for growth traits ...