EFFICIENCY OF PHOSPHORUS USE IN YOUNG PLANTS OF Eucalyptus urophylla

The objective of this study is to select superior genotypes for effi ciency of phosphorus use in seedlings of Eucalyptus urophylla, correlating them with initial height growth in the fi eld, at age eight months. A completely randomized block design was used in the nursery, consisting of eight clones, three replicates and four plants per plot. The same design was used in the fi eld, consisting of eight clones, four replicate blocks and nine plants per plot. Data being collected in the nursery at age 120 days included: phosphorus concentration in shoot (CPPA); phosphorus concentration in root (CPR); phosphorus content in shoot (CtPPA); phosphorus content in root (CtPR); total phosphorus content in seedling (CtPT); phosphorus use effi ciency in shoot (EUP-PA); phosphorus use effi ciency in root (EUP-R) and total phosphorus use effi ciency (EUP-T). Analyses of variance showed signifi cant genetic differences among clones for all traits and, given the high heritability values found, estimated genetic gains were generically very high. As regards predicted indirect genetic gain, it was noted that selection on nursery seedlings for all EUPs provided the highest values of indirect gain in height of fi eld seedlings.


INTRODUCTION
Species of genus Eucalyptus play an important part in the Brazilian social, economic and environmental scene.Up until 2007, an estimated 3,771,867 ha in area was reforested with Eucalyptus (ASSOCIAÇÃO BRASILEIRA DE PRODUTORES DE FLORESTAS PLANTADAS -ABRAF, 2008), expected to produce enough timber to cater for the Brazilian industrial sector, in particularly cellulose pulp production and charcoalfuelled steel metallurgy.
From a nutritional standpoint, a superior species and/or genotype is one that is capable of developing and producing good yields under unfavorable fertility conditions, and having the ability to absorb the required nutrients in smaller quantities and/or distribute them more effi ciently to various plant components, without affecting productivity (FURLANI et al., 1984).
Species with an enhanced ability to absorb and/or use nutrients would be more desirable.This is so because not only would they make better use of the nutrients being supplied via fertilization, but smaller amounts of nutrients should suffi ce (LUCA, 1997), resulting in greater biomass yield, which is an important characteristic for successful growth and improved productivity where soils are low in P levels (MACHADO, 2000).
Bearing that in mind, the objective of this work is to select superior genotypes for phosphorus uptake effi ciency in clonal seedlings of Eucalyptus urophylla, estimating variance components for phenotypic, genotypic and Godoy, T. G. e Rosado, S. C. da S.
environmental parameters, as well as heritability, direct and indirect genetic gains, correlating them to young plants in the fi eld at age eight months.

MATERIAL AND METHODS
Clonal seedlings were used in this experiment, as produced by the minicutting technique (vegetative propagation) in the forest nursery of Carvovale Ltda.The nursery 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 Köppen classifi cation, the predominant local climate is dry, continental, with average annual precipitation less than 1,000 mm and average maximum temperature around 34ºC (COMPANHIA DE PESQUISA DE RECURSOS MINERAIS -CPRM, 2004).
Seedlings were produced in tubes with a volume capacity of 50 cm 3 and stored in plastic trays containing 96 cells, suspended 80 cm above ground level.The substrate used consisted of 40% vermiculite, 30% scorched rice husk, 10% coconut fi ber, 20% slash pine bark and needles, at a pH of 5.5.
In the nursery, eight clones were assessed using a completely randomized block design, with three replicates and four plants per plot.In the fi eld, the same design was used consisting of eight clones with four replicates and nine plants per plot, at age eight months, arranged at 3 x 2 m spacing.According to Tomé (2002), in order to obtain more precise estimates of treatment means or selection gain, models with a random progeny effect are more suitable.This is so because the effects of progenies are better represented by random effects, since the usual assumption in quantitative genetics is that genetic values related to a genotype result from the sum of small effects of individual contributions from a large number of Mendelian segregating genes for trait manifestation.
At age 120 days, the root portion of each seedling was washed carefully, trying to preserve roots as much as possible.The vegetal materials were placed in labeled paper bags and taken to a forced air oven at 70ºC, to a constant weight.They were then ground using a stainless steel Wiley mill and submitted to the Laboratory of Leaf Analysis of the Departamento de Ciência do Solo of the Universidade Federal de Lavras (UFLA) where, using a nitric-perchloric mixture, P concentrations were determined in the shoot and root portions, by colorimetry (ASSOCIATION OF OFFICIAL ANALYTICAL CHEMISTS -AOAC, 1975).
The following data were collected: height of fi eld seedlings after eight months (Hc); phosphorus concentration in shoot (CPPA); phosphorus concentration in root (CPR); P content in shoot (CtPPA); P content in root (CtPR); total P content in seedling (CtPT); phosphorus use effi ciency in shoot (EUP-PA); phosphorus use effi ciency in root (EUP-R); total phosphorus use effi ciency (EUP-T).Effi ciencies were calculated on the basis of the following methodologies: Siddiqui and Glass (1981) and Swiader et al. (1994).
Data were submitted to analysis of variance, following program procedures in GENES (CRUZ, 2001).The statistical model used was where: Y ij is the mean value observed in the i th genotype of the j th block; μ is grand mean; g i is effect of the i th genotype (random effect); b j is effect of the j th block (random effect); ε ji is experimental error.
The analysis of variance summary and expected mean square values for estimates of phenotypic, genotypic and environmental parameters are provided in Table 1.Estimates of variance components, values of heritability and coeffi cients of variation were derived according to expressions described by Cruz and Carneiro (2003).
Table 3 provides predicted genetic gains to be attained from direct selection for initial height growth in the fi eld, as assessed eight months after planting (Hc), and for P nutritional traits, as well as indirect gains in Hc with selection for nutritional traits CPPA, CPR, CtPPA, CtPR, CtPT, EUP-PA, EUP-R and EUP-T.
In forest improvement programs, correlations are used to assist with selection involving two traits simultaneously, in other words, selection on one trait 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, for this to happen, it is necessary that the genotypic correlation among the traits of interest be relatively high.The negative values of genotypic correlation between Hc and CPPA, CPR and CPT were due to the fact that, in these concentration traits, there may be a dilution of P in the dry matter of seedlings Table 4 provides mean values per clone type for all nutritional traits of interest.Considering only the traits showing better indirect gain responses in initial fi eld height, it is noted that EUP-PA ranged between 10.830 g 2 .mg - (clone ITA-011) and 31.030g 2 .mg - (clone ITA-018); EUP-R ranged between 4.213 g 2 .mg - (clone ITA-011) and 10.797 g 2 .mg - (clone ITA-063), while EUP-T ranged between 7.253 g 2 .mg - (clone ITA-011) and 18.036 g 2 .mg - (clone ITA 018).
Clones ITA-063 and ITA-069 showed the highest mean values of Hc (height of fi eld seedlings), 2.8 and 2.7m respectively.Considering only those traits with higher ESI values, it is noted that said clones are included among clones also showing the highest estimated means for EUP-T, EUP-PA and EUP-R.
In the case of forest stands, the shoot portion being economically more important, P use effi ciency becomes a good indicator of nutrient conversion into biomass, in Eucalyptus clones (FURTINI NETO, 1994).
Results found in this study point to different behavioral patterns in eucalyptus clones regarding nutritional phosphorus use effi ciency.As far as forest activity is concerned, given the wide assortment of species and genotypes used as well as the use of differing fertility soils, this aspect should be given special attention, so as to ensure that the interaction between soil x species and/ or genotypes is done sensibly.
From that correlation, a genetic material may be obtained capable of simultaneously gathering several favorable attributes (CRUZ; REGAZZI, 2001).By the same token, where a trait is negatively correlated to another, one should be cautious that selection on one trait does not cause undesirable changes on other traits.
Table 3 above provides genotypic correlation coeffi cients among the traits of interest.From among nutritional traits, the lowest direct gain (GD%=18.3) was estimated for CPR and the highest (GD%=53.9)was estimated for CtPPA, considering a 25% selection.However, when observing estimates of expected indirect genetic gains (GI%) in fi eld height growth (Hc), it is noted that the most effi cient nutritional trait, at 120 days, was total P use effi ciency (EUP-T), with a GI of 8.8%, revealing an indirect selection effi ciency (ESI) of 70.0%.Other important nutritional traits, showing an ESI above 50%, included EUP-PA and EUP-R, with 64.9% and 57.1% respectively.These ESI values thus suggest that the nutritional trait 'P use effi ciency' can be an indicator of seedling quality in this clonal population.
According to Falconer (1987), indirect selection constitutes an important strategy for improving low heritability traits and/or traits considered diffi cult to measure.
These higher ESI values were due to the heritability of initial height growth in the fi eld (Hc) (h 2 = 80%) being relatively lower than P use effi ciency traits and/or due to the higher genotypic correlation values (Table 3).From Table 3 it is noted that the genotypic correlation coeffi cients (r g ) found between Hc and EUP-T, Hc and EUP-R, and Hc and EUP-PA were 0.6, 0.5 and 0.6 respectively.

CONCLUSIONS
Signifi cant differences were found among clones regarding genotypes, for all traits of interest.
Heritability values were found to be relatively high.
A high correlation was found between initial fi eld height and EUP-T, EUP-PA and EUP-R, for nutritional traits.
Clone ITA063 was found to have the highest CtP value, while clone ITA 069, the lowest.
The most P use effi cient clone was ITA063, while the most ineffi cient (EUP-T, EUP-PA and EUP-R) was ITA 011.

σ
, signifi cant differences are noted among clones, by the F test (p≤0.01),for initial height Effi ciency of phosphorus use in young plants ...

Table 1 -
Analysis of variance summary and expected mean square values for phenotypic, genotypic and environmental parameters.
ˆb σ : variance between blocks; g: number of genotypes; b: number of blocks.

Table 2 -
Analysis of variance summary and estimate of genetic, phenotypic and environmental parameters for nutritional characteristics and height of young fi eld plants, at age eight months, of E. urophylla clones.Resumo da análise de variância e estimativa dos parâmetros genéticos, fenotípicos e ambientais para características nutricionais e altura de plantas jovens no campo, aos oito meses de idade, de clones de E. urophylla.
** P≤0.01; Hc: height of fi eld seedlings, after eight months; CPPA: P concentration in shoot; CPR: P concentration in root; CtPPA: P content in shoot; CtPR: P content in root; CtPT: total P content; EUP-PA: P use effi ciency in shoot; EUP-R: P use effi ciency in root; EUP-T: total P use effi ciency; CV e : experimental coeffi cient of variation (error); CV g : genetic coeffi cient of variation;  2 f : phenotypic variation;  2 e : environmental variation,  2 g : genetic variation; h 2 : coeffi cient of genotypic determination; CV g /CV e : variation index.Godoy, T. G. e Rosado, S. C. da S.