Abstracts

The use of self-incompatibility in the production of hybrid eucalyptus seed by 'Aracruz Celulose' in Brazil

Tatiana Góes Junghans¹, Ingrid Peters-Robinson², Fernando L. Bertolucci³

and Acelino Couto Alfenas^1,4

¹ Núcleo de Biotecnologia Aplicada à Agropecuária (BIOAGRO), Universidade Federal de Viçosa (UFV), 36571-000 Viçosa, MG, Brasil. Send correspondence to A.C.A.

²State University of New York, Albany, NY, USA.

³Aracruz Celulose S.A., Aracruz, ES, Brasil.

⁴Departamento de Fitopatologia, UFV, Viçosa, MG, Brasil.

ABSTRACT

Self-incompatibility found in a Eucalyptus grandis clone was used to promote interspecific hybridization between E. grandis and E. urophylla clones. The interspecific hybridization achieved in an open-pollinated commercial seed orchard planted in Espírito Santo, Brazil was evaluated by the multi-locus outcrossing rate (t_m) of the seed producing clone, E. grandis. The percentage of outcrossed seeds reached 95.9%. The outcrossing rate of individual trees was quite variable, but was always above 70.0%. Wright's coefficient was negative (F = -0.30) revealing an excess of heterozygous genotypes in the progeny. Genetic parameters confirmed the high degree of hybridization expected in this orchard. The 800-m wide protection belt of native tropical forest that surrounds this orchard has significantly reduced pollen contamination, in comparison to a 400-m wide belt used in another local seed orchard.

INTRODUCTION

Eucalyptus have a mixed mating system and relatively high outcrossing rates, which are maintained by protandry and various levels of self-incompatibility, reinforced by selection against selfed progeny during the life cycle (Phillips and Brown, 1977; Potts et al., 1987). Self-incompatibility, manifested at different rates, has been frequently observed in Eucalyptus grandis (Hodgson, 1976; Potts and Savva, 1988). Beardsell et al. (1993) suggested that this could be the main mechanism controlling the outcrossing rate in this species. Therefore, selection of self-incompatible plus-trees combined with their cloning and utilization as female parent in crosses represent a promising strategy to reduce the percentage of seeds originated from selfing in open-pollinated seed production.

E. grandis, E. urophylla and the hybrid between these two species are important raw-material for the cellulose and paper industry in Brazil (Bertolucci et al., 1995). Orchards designed for the production of hybrid seeds through open pollination were installed by 'Aracruz Celulose' in the 1970's. In spite of being operationally very desirable, open pollination has the disadvantage of including a certain amount of inbred crosses. The most pronounced inbreeding results from selfing because, in one single step, half of the alleles become fixed in homozygosity, quickly reducing the effect of heterosis and manifesting recessive deleterious characters in the progeny. Consequently, inbreeding can significantly reduce the productivity of plantations that use such seeds (Eldridge and Griffin, 1983; Yeh et al., 1983; Bush and Smouse, 1991; Hardner and Potts, 1995).

We used isozymes to verify the hybridization rate in a seed orchard planted by 'Aracruz Celulose', in which self-incompatibility was used as a tool to enhance the production of E. grandis x E. urophylla hybrid seeds. Isozyme electrophoresis is a fast and efficient technique for estimating outcrossing rates in plants. It offers several co-dominant markers that can be easily scored at the seed, seedling or adult tree stages (Brown and Allard, 1970; Brown, 1989). This technique has been extensively used to investigate mating system parameters in eucalyptus populations (Brown et al., 1975; Phillips and Brown, 1977; Yeh et al., 1983; Fripp et al., 1987; Moran et al., 1989; Campinhos, 1992).

MATERIAL AND METHODS

The seed orchard was examined in the State of Espírito Santo, Brazil. Trees were spaced 4 m x 4 m in a total area of 8 hectares (282 m x 282 m). This orchard was composed of a single clone of E. grandis, identified as highly self-incompatible, and 22 clones of E. urophylla, the pollen donors. These clones were randomized and equally represented within the orchard (Ikemori and Campinhos, 1983). The 22 pollen parents were planted in March 1976. The E. grandis clone, from which seeds were harvested, was planted in August 1978. For sampling purposes, the orchard was sub-divided into 16 sectors in which seeds from the E. grandis clone were harvested in 1994. Twenty-nine trees were sampled. Seedlings were grown in a tree nursery at the Federal University of Viçosa in Minas Gerais. Young leaves from the adult parental clones in the orchard and from 15 seedlings at the age of two months were harvested for electrophoresis. The seedlings were randomly chosen within each family (i.e., within the progeny of each E. grandis tree). Leaf tissue was crushed in mortars with chilled extraction buffer (0.034 M dibasic sodium phosphate, 0.2 M sucrose, 2.56% PVP-40, 5.7 mM ascorbic acid, 5.8 mM di-ethyl-carbamic acid, 2.6 mM sodium bisulfite, 2.5 mM sodium borate, 0.2% b-mercaptoethanol, 1.0% polyethylene glycol-6000) following procedures described in Alfenas et al. (1991) and Campinhos et al. (1996).

The following enzyme systems were found to be polymorphic in preliminary assays: 6-phosphogluconate dehydrogenase (6PGDH), phosphoglucose isomerase (PGI) and isocitrate dehydrogenase (IDH). Electrophoresis was conducted in 13% starch gels and morpholine-citric buffer system (gel buffer: 0.04 M citric acid titrated to pH 6.1 with N-(3-aminopropyl)-morpholine and diluted 1:20 for use; electrode buffer: 0.04 M citric acid titrated to pH 7.1; 6.6 volts/cm for 3 h) (Clayton and Treliak, 1972, listed in Alfenas et al., 1991).

Multilocus (t_m) and average single locus (t_s) outcrossing rates, and the fixation index (F) were estimated by MLT (Ritland, 1990), a FORTRAN program developed by Ritland and Jain (1981). In this orchard, outcrossing represents hybridization.

RESULTS AND DISCUSSION

Enzymes of 6PGDH, IDH and PGI systems are known to be dimeric in eucalyptus. Their electrophoretic patterns and inheritance were examined by other authors in several eucalypt species (Moran and Bell, 1983; Passador, 1994; Nóbrega, 1995). Zymograms of parental clones and of the progeny are shown in Figure 1. Alleles at each locus were designated alphabetically in order of mobility, allele a being the most anodal.

Two loci of 6PGDH were examined. Alleles a, b and c of locus 6Pgdh-1 were found in the parental clones, with Rf values 0.25, 0.23 and 0.17, respectively (Figure 1A). Allele c (Rf = 0.17) characterizes the E. grandis clone. The band corresponding to allele c overlapped with the position of the monomorphic 6Pgdh-2 locus on the gel. A fourth allele, d (Rf = 0.15), not present among the parental genotypes, showed that crosses to foreign pollen occurred in spite of a protection belt of 800 m of native tropical forest that surrounds the whole orchard. The band corresponding to allele d was positioned slightly under the 6Pgdh-2 activity zone in the gel.

IDH bands were concentrated in a single activity zone, representing one locus with three alleles (Rf values: 0.13, 0.10 and 0.80) (Figure 1B). The E. grandis genotype at this locus was bc. The three alleles of Idh were present among the E. urophylla clones. Enzyme PGI, represented by two loci, was variable at the locus Pgi-2 in parents and progeny. Three alleles were identified at this locus in the progeny, with Rf values 0.15, 0.13 and 0.1 (Figure 1C). The E. grandis clone was homozygous aa, and the male parents were identified as aa and ab. Allele c, present only in the progeny, also indicates pollen contamination. The minimum contamination rate in the orchard in 1994 was 2.8%, as estimated by the frequency of progeny genotypes showing allele c of Pgi-2 or d allele of 6Pgdh-1. This amount does not include possible contamination by pollen carrying alleles in common with those present in the 23 clones. The 800-m wide protection belt of native tropical forest that surrounded this orchard significantly reduced pollen contamination, in comparison to a 400-m wide belt used in another local E. grandis x E. urophylla seed-orchard (Campinhos et al., 1996).

Progeny with genotypes carrying the allele a or d at 6Pgdh-1, allele a at Idh, or alleles b or c at Pgi-2 can be directly recognized as resulting from cross-fertilization. In addition, most of the progeny could be classified as originating from selfing or outcrossing based on their multilocus genotypes. Allelic frequencies observed at 6Pgdh-1, Idh and Pgi-2 in the parental clones, progeny, and pollen, as estimated from family genotypic arrays, are listed in Table I. The less frequent allele at 6Pgdh-1 (d) was combined with the next less frequent allele (c) at this locus to constitute a single synthetic allele, because the MLT program assumes three alleles per locus (Ritland, 1983).

Estimated allele frequencies in pollen, based on genotypic frequencies observed in the progeny, differed from allele frequencies in parental clones (Table I). This result is expected, because the E. grandis clone should not contribute to the pollen that originated the examined seed, although it was the source of all female gametes. Progeny allele frequencies approached the average between pollen and ovule frequencies listed in Table I, as expected under complete outcrossing.

Average outcrossing rate at the three loci (t_s), multilocus outcrossing rate (t_m) and fixation index (F) are listed in Table II. The single-locus outcrossing rate at the monomorphic Pgi-2 locus was 0.957. It was estimated by t = H/p, where H is the observed heterozygote frequency in the progeny and p is the joint frequency of non-maternal alleles in the pollen. The multilocus outcrossing rate in the orchard was t_m = 0.96, higher than the mean value reported for Eucalyptus species in nature (t_m = 0.78; Moran and Bell, 1983). The difference between t_s and t_m was very small, suggesting that allele frequencies in the pollen pool were relatively uniform in space and time during the flowering season.

Values of t_m, estimated at a family level, ranged from 0.7 to 2.0 (Table III). The value of t, by definition, ranges from zero to 1.0. Values above 1.0 indicate that the distribution of genotypes within family arrays did not allow the convergence of t estimates to a single value, when applying the MLT algorithm (Ritland and Jain, 1981). This effect is generally the result of sampling error violating one of the assumptions of the mixed mating model. Estimates between 1.0 and 2.0 may be obtained, for example, when preferential mating occurs within the population, violating the assumptions of random mating in the alogamous fraction of the crosses (Ritland and El-Kassaby, 1985). Being self-incompatible, the E. grandis clone crossed preferentially, or almost exclusively, with the E. urophylla clones in the orchard. Another fraction of the cross-fertilizations can be attributed to contaminant pollen. According to field observations (Bertolucci, F.L., Aracrus Celulose, unpublished results), the E. grandis clone was not completely self-incompatible when inflorescences were experimentally isolated under paper bags. Therefore, a small amount of self-fertilization may have taken place as well. Family outcrossing rates above 1.0 should not be considered by their nominal value, but interpreted as complete outcrossing. Families exhibiting t_m= 2.00 were indeed constituted entirely by seedlings with recognized hybrid multilocus genotypes.

When taken together, the parameters F, t_m and t_s give a clearer picture of how geneflow proceeded in the orchard. The value of F measures the correlation between alleles at a locus (Wright, 1969). This index also measures the deviation of the observed frequency of heterozygous genotypes from its expected proportion under random mating (Wright, 1978). The negative value of F (F = -0.30) indicates an excess of heterozygous genotypes in the examined progeny. This "negative inbreeding" is due to negative assortative mating producing hybrid progeny, as a consequence of the self-incompatibility of the E. grandis clone. Similar results were obtained in another orchard involving the same combination of eucalyptus species and analytical procedures (Campinhos et al., 1996).

Comparing the effectiveness of pollen sterility (Campinhos et al., 1996) versus self-incompatibility in promoting cross-fertilization in eucalyptus orchards (t_m = 0.702 and t_m = 0.959, respectively), we concluded that reduced pollen production per se does not adequately predict the crossing behavior of the tree. Plus-trees should be routinely tested for self-incompatibility. Preferential fertilization resulting from strong self-incompatibility combined with natural selection favoring survival and growth of heterozygotes among seedlings and seedling selection practiced in tree-nurseries should have the strongest impact on the quality of hybrids produced by open pollination in eucalyptus breeding.

ACKNOWLEDGMENTS

The authors are grateful to FINEP and Aracruz Celulose/SIF for the financial support of this work.

RESUMO

Verificamos a eficiência do uso da auto-incompatibilidade encontrada em um clone de Eucalyptus grandis para aumentar a hibridação interespecífica entre E. grandis e E. urophylla em um pomar clonal para produção de semente comercial, instalado no Espírito Santo, Brasil. O nível da hibridação interespecífica alcançada nesse pomar foi avaliado por meio da taxa de fecundação cruzada, t_m, do clone de E. grandis produtor de semente. A taxa (t_m) foi estimada usando-se vários locos de isoenzimas simultaneamente. A percentagem de sementes híbridas alcançou 95,9%. A taxa de fecundação cruzada das árvores individuais foi variável, mas sempre superior a 70%. O coeficiente de Wright foi negativo (F = -0,30), indicando haver um excesso de genótipos heterozygotos na progênie. Os parâmetros genéticos estimados confirmaram o alto nível de hibridação esperado nesse pomar. A barreira de proteção, constituída por uma faixa de 800 metros de floresta nativa, reduziu significativamente a contaminação do pomar por pólen de outros povoamentos de eucalipto, em comparação à faixa de 400 metros de largura utilizada em outro pomar na mesma região.

(Received November 25, 1996)

Publication Dates

History