Interspecific hybridization and inbreeding effect in seed from a Eucalyptus grandis x E. urophylla clonal orchard in Brazil

Interspecific hybridization and inbreeding effect in seed from a Eucalyptus grandis x E. urophylla clonal orchard in Brazil

Eduardo N. Campinhos¹, Ingrid Peters-Robinson², Fernando L. Bertolucci³ and Acelino C. 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

We used allozyme markers to estimate the amount of natural hybridization between Eucalyptus grandis and E. urophylla in a 7.4-hectare commercial hybrid-seed orchard planted in Espírito Santo, Brazil. This orchard was planted in 1982 using a honeycomb design, with each hexagonal plot containing one E. grandis tree surrounded by six E. urophylla trees. There were 267 replicated hexagonal plots in the orchard. Seeds were harvested from the E. grandis clone only. The multilocus outcrossing rate estimated for the E. grandis clone averaged 70.2%, ranging from 33.0 to 99.0% among individual trees. Contaminant pollination, inferred from progeny genotypes containing alleles not present in the seven parental clones, accounted for 14.4% of the hybrid seed. Contaminant pollen was attributed to neighboring eucalyptus stands isolated from the orchard by a 400-m wide belt of native forest. Inbred and hybrid progenies were identified by their allozyme genotypes and transplanted to the field. Field growth of inbred progeny was 30% lower than that of hybrid plants at two and three years of age.

INTRODUCTION

The genus Eucalyptus contains the most frequently planted forestry species in tropical and sub-tropical countries (Eldridge et al., 1993). Interspecific hybridization in this genus has often been reported for natural and cultivated stands (Pryor and Johnson, 1981; Ashton and Sandiford, 1988). Geographical and temporal reproductive isolation are considered to be the major barriers for gene flow between Eucalyptus species in their native range (Griffin et al., 1988). Reproductive isolation between species of Eucalyptus can easily break down in multispecies plantations established in new environments, as observed in Brazil (Brune and Zobel, 1981) and South Africa (Burgess and Bell, 1983; Burgess et al., 1985). In many cases the F1 generation performance of these natural hybrids has been found to be superior to the original germplasm. However, F2 and subsequent generations generally show poor growth and undesirable canopy shape (Eldridge et al., 1993).

In contrast, controlled interspecific hybridization has been a very successful strategy in eucalypt breeding programs. Eucalyptus grandis x Eucalyptus urophylla hybrids presently support a major portion of the cellulose and paper industry in Brazil (Bertolucci et al., 1995). Fast early growth, good canopy shape and superior wood properties have made E. grandis a major choice for plantations in sub-tropical and warm temperate climates, although this species is rather susceptible to diseases in lowland regions in the humid tropics (Eldridge et al., 1993). In Brazil, E. urophylla is known to present higher resistance to canker disease (Cryphonectria cubensis) than E. grandis. Hybrids between E. grandis and E. urophylla often combine higher resistance to canker and higher wood density. Clonal propagation of such superior hybrids has been a major goal in the eucalypt breeding program conducted by 'Aracruz Celulose' Company in Espírito Santo, Brazil (Brandão et al., 1984).

Eucalyptus species are predominantly cross-fertilized in nature and generally display strong inbreeding depression if experimentally selfed (Eldridge and Griffin, 1983). For this reason, commercial production of hybrid seeds calls for control of both self-fertilization and pollen contamination in the seed orchard. Isozyme electrophoresis provides a fast and relatively inexpensive technique for examining breeding systems in eucalypts (Brown et al., 1975; Phillips and Brown, 1977; Hopper and Moran, 1981; Moran and Bell, 1983; Griffin et al., 1987). The present study reports the level of interspecific hybridization achieved by open pollination in a seed orchard installed by 'Aracruz Celulose'. We also compared the growth rate of hybrid and selfed progenies, identified by means of their isozyme genotypes.

MATERIAL AND METHODS

The seed orchard examined was installed in 1982 in Aracruz (latitude: 19.49^o S; longitude: 40.16^o W), in Espírito Santo, Brazil, by the 'Aracruz Celulose' Company. Canker-free plus-trees selected in provenance tests of E. urophylla germplasm collected in Bessi-Lau, Flores and Timor (Indonesia) were clonally propagated to constitute the pollen donor in this orchard. Clones were obtained from cuttings of basal sprouts that grow after the coppice of the target trees. The single E. grandis clone used as the seed parent in this orchard was selected from a progeny test of germplasm collected in Atherton (Australia) using the same propagation technique. This particular E. grandis plus-tree was chosen because it was identified as partially male-sterile. According to field observations (Bertolucci, F.L., Aracruz Celulose, unpublished results), the flowers of this tree produced poorly developed anthers and seed set was very low when the inflorescences were protected from cross-pollination. The seven clones were replicated in 267 hexagonal plots in the orchard, constituting a modified honey-comb design that covered 7.4 hectares. Trees were spaced 6 m x 6 m in the field. In each hexagonal plot the central E. grandis tree was surrounded by six E. urophylla clones. The orchard was framed by a 400-m wide protection belt consisting of native tropical rain forest (Ikemori and Campinhos, 1983; Martins and Ikemori, 1987).

Parental clones were screened for polymorphic allozyme loci in preliminary essays conducted by Campinhos (1992). Those loci that provided well-defined banding patterns and had known genetic interpretation of zymograms were chosen as genetic markers. Young leaves harvested from the adult trees in the orchard were used for isozyme characterization of parental clones. Leaf samples were packed individually in plastic bags, which were sealed and placed between layers of sawdust and ice in a Styrofoam box. Samples were harvested early in the morning and transported, on the same day, to the Federal University of Viçosa, MG. Enzyme extraction was done the following morning. Replicated sampling and isozyme essays were conducted to confirm the genotypes of the seven parental clones. In 1991, seeds were harvested from 30 E. grandis trees sampled along transects crossing the orchard in three spaced sectors. The seeds were germinated in Gerbox containers with sterilized humid sand. After germination, seedlings were transplanted to 11 x 20 cm-plastic bags containing a 1:1 mixture of sterilized organic soil and sand, and cultivated in a tree-nursery at the University. Enzymes were extracted from young leaves when seedlings were 5 months old. Fifteen vigorous seedlings in each family were chosen for electrophoresis, simulating the selection practiced in commercial nurseries.

For allozyme analysis, approximately one gram of leaf tissue of each plant was homogenized in 3 ml of 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 dithioethylcarbamic acid, 2.6 mM sodium bisulfite, 2.5 mM sodium borate, 0.2% b-mercaptoethanol, 1.0% polyethylene glycol-6000 (Alfenas et al., 1991)) using chilled mortars and pistils. The protein extract was filtered through a piece of tissue paper placed directly upon the macerate. The filtrate was absorbed on Whatman 3MM chromatographic paper wicks (12 x 5 mm) which were stored at -85^oC in microfuge tubes.

Electrophoresis was conducted in 13% starch gels using a morpholine-citric buffer system (0.04 M citric acid titrated to pH 6.1 with N-(3-aminopropyl)-morpholine and diluted 1:20; electrode buffer: same as the gel buffer but not diluted and titrated to pH 7.1) (modified from Clayton and Tretiak, 1972) and a discontinuous Tris-citric lithium borate buffer system (0.065 M Tris, 0.01 M citric acid, pH 8.1; electrode buffer: 0.05 M lithium hydroxide, 0.19 M boric acid, pH 8.6) (Moran and Bell, 1983). Electrophoresis and gel staining procedures followed Alfenas et al. (1991).

Only the progeny of the E. grandis clone was examined by electrophoresis, because this is the seed harvested for commercial purpose in the orchard. Multilocus (t_m), average single-locus (t_s) outcrossing rates, fixation index (F), and gene frequencies in the pollen pool were estimated using the MLT program (Ritland, 1990) developed by Ritland and Jain (1981). The estimated outcrossing rates represent the percentage of hybridization achieved by open pollination. Allele frequencies in ovules and pollen were unequal, given that the parental clones belong to different species. Single-locus outcrossing rates at the loci found homozygous in E. grandis but polymorphic in E. urophylla were calculated as t = H/p, where H is the observed frequency of heterozygotes in the progeny, and p is the estimated joint frequency of the marker alleles in the pollen.

After the allozyme analysis, the 450 seedlings were tagged and transplanted to an experimental field in 'Aracruz Celulose', Espírito Santo, spaced 3 m x 3 m. Plant height and circumference of the stem at breast height (CBH) were measured in the second and third year of field growth.

RESULTS AND DISCUSSION

Three enzyme systems were found to be polymorphic in the parental clones: a-esterase (a-EST), 6-phosphogluconate dehydrogenase (6PGDH) and phosphoglucose isomerase (PGI). Activity and migration rate (Rf) of the enzyme bands were the same whether extracted from selfed or hybrid progeny, suggesting that the respective genes are expressed similarly in the two genetic backgrounds. Zymograms obtained from the parental clones and their progeny are shown in Figure 1.

Figure 1
- Zymogram and genotypes for the enzymes a-esterase (a-EST), 6-phosphogluconate dehydrogenase (6PGDH) and phosphoglucose isomerase (PGI). A, Female parent; B, pollen parent; C, progeny. *Genotypes revealing contaminant pollination.

Several activity zones could be distinguished in the a-EST gels, but only the locus corresponding to the most cathodal staining zone, here called a-Est-1, provided bands of reliable activity and resolution (Figure 1A). Three alleles of this locus (alleles b, c and d) were observed among the parental clones (Rf 0.75, 0.70 and 0.67, respectively). A fourth allele, a (Rf 0.80), was found only in the progeny.

Two loci were recognized in gels stained for 6PGDH. Locus 6Pgdh-1 was polymorphic (Figure 1B), and allele c (Rf 0.17) was unique to the E. grandis clone. The position of its corresponding band overlapped that of the monomorphic 6Pgdh-2 locus. All six E. urophylla clones shared the genotype bb at the locus 6Pgdh-1 (Rf 0.22). Consequently, seedlings showing the heterozygous bc genotype could be immediately classified as outcrossed progeny. A third allele, a (Rf 0.30), was again identified only in the progeny. Two loci were distinguished in the PGI zymograms (Figure 1C), Pgi-1 being monomorphic. The E. grandis clone was found to be homozygous aa at the locus Pgi-2 (Rf 0.47), while E. urophylla clones were either cc or bc. Therefore, all heterozygous progeny at the locus Pgi-2 resulted from outcrossing. A fourth allele d (Rf 0.35) was found in the progeny.

Allele a from both Est-1 and 6Pgdh-1, and allele d from Pgi-2 were attributed to fertilization by foreign pollen. The pollen contamination detected by means of isozymes in the 1991 seed samples did account for 14.2% of the examined seedlings, and 20.2% of the outcrossed progeny. This finding showed that pollination vectors (insects and hummingbirds) were able to cross the 400-m isolation belt that surrounded the orchard. The reported level of gene flow between local eucalyptus stands does not include non-identified contamination by pollen carrying alleles that were also present among the seven parental clones.

The observed allelic frequencies at a-Est , 6Pgdh-1 and Pgi-2 in the parental clones and progeny, as well as the estimated pollen allele frequencies are listed in Table I. Allele frequencies in the pollen pool and parental clones were clearly different. The six E. urophylla clones were not equally represented in the pollen pool. It should be noticed that the E. grandis clone, being at least partially male-sterile, did not contribute much to the pollen that gave rise to the examined progeny. Another point to be noticed in Table I is that unique maternal alleles at 6Pgdh-1 and Pgi-2 were not equally represented in the pollen as they should have been. The difference between their frequencies suggests that part of the progeny classified as resulting from self-pollination based on allozyme genotypes may instead have originated from crosses with undetected contaminant pollen. The three detected contaminant alleles were represented at comparable frequencies in the progeny.

The locus Est-1, heterozygous in the E. grandis clone, allowed the computation of t_m and t_s values as well as allele frequencies in the pollen using MLT (Ritland and Jain, 1981). The MLT program uses no more than three alleles per locus. For this reason, the less frequent allele at Est-1 (d) was combined with the next less frequent allele (c) at this locus to constitute a single synthetic allele (Ritland, 1983). The same procedure was adopted for allele d at Pgi-2. In both cases, the next less frequent allele happened to be absent in the maternal genotype.

The outcrossing rate observed at loci 6-Pgdh-1 and Pgi-2 were, respectively, t = 0.701 and t = 0.702. Single-locus outcrossing estimate at Est-1 was not estimated because the progeny of heterozygotes does not provide adequate information for estimation of t (Brown et al., 1975). The average single-locus (t_s) and the multilocus (t_m) outcrossing rates, and the fixation index (F) are listed in Table II. The amount of hybridization (measured by t_m) was 70.2%, which falls in the range of intraspecific outcrossing rates (0.69 to 0.86) reported by Moran and Bell (1983) for ten Eucalyptus species in nature. The difference between values of t_s and t_m was very small, indicating that the outcrossing estimates obtained independently by the three allozyme loci were comparable. This implies that allele frequencies in the pollen pool during the flowering season were relatively uniform in space and time. Values of t_m estimated at a family level (Table III) varied between 0.33 and 0.99. Analysis of variance did not show differences among the transect's average t_m values either (F ratio = 1.5993; probability = 0.22).

According to Adams and Birkes (1989), the two main reasons for variation in outcrossing rates are the relative availability of selfing versus outcrossing pollen at the time of fertilization and the different abilities of clones to produce viable embryos after selfing, mostly as a function of the number of lethal gene equivalents they possess. The variation in t_m among E. grandis trees in this orchard was due exclusively to environmental fluctuation, because all mother plants had the same genotype. Asynchrony in the onset of flowering among the paternal clones would have been detected as a divergence between t_m and t_s, which was not the case. Finally, fluctuation in outcrossing opportunities may result from the behavior of the pollinators (Hopper and Moran, 1981). The replicas of the E. grandis clone in the orchard can differ randomly in their intensity of flowering and, consequently, receive differential attention from the pollinators. Variation in the outcrossing rate among ramets of a single clone is not an uncommon observation. A comparable within-clone variation in outcrossing rate was observed by Junghans et al. (1998) in another E. grandis x E. urophylla seed orchard installed in the same region by 'Aracruz Celulose'. The orchard studied by these authors has a different field design, but seeds were also harvested from a single E. grandis clone.

The fixation index, F, measures coancestry in the parental population (Wright, 1969). No coancestry is expected to exist in the parental population, because the seed and the pollen-producing clones belong to different species. The negative value obtained, F = -0.300, shows that the observed frequency of heterozygote genotypes found in the progeny is far higher than expected in a population in Hardy-Weinberg equilibrium. This excess of heterozygotes is due to negative assortative mating resulting from the partial male-sterility of the E. grandis clone, combined with the divergence between male and female gene frequencies in the gametic pool. In addition, the seedling selection practiced before electrophoresis must have favored heterozygotes because hybrid progeny tended to be more vigorous.

The growth of hybrid and selfed progenies, measured at two and three years after the seedlings were transplanted to the field, are compared in Table IV. The mean circumference of the stem at breast height (CBH) and the height of the hybrid plants were both significantly larger than in the inbred group (t-test, P < 0.01). The coefficient of variation for both measurements was much higher in the inbred group. The level of variability in this group added to the previously stated concern that some of the plants were misclassified as inbred because of contamination by foreign pollen bearing alleles in common with the E. grandis clone.

Selection against selfed progeny in eucalyptus has been frequently reported (Phillips and Brown, 1977; Eldridge and Griffin, 1983; Potts et al., 1987; Hardner and Potts, 1995) and the apparent advantage of heterozygote genotypes at the intraspecific level in tree species has been investigated by several authors (Moran and Brown, 1980; Moran et al., 1989; Mitton, 1989; Bush and Smouse, 1991). Variations in seedling survival and development must have favored hybrid seedlings in our sampling procedure since the earliest developmental stages. The artificial selection practiced against weak seedlings by most commercial nurseries, as well as any stand thinning operation based on performance in the field, probably eliminated a considerable part of inbred progeny in eucalyptus plantations. Variation in t_m estimates obtained from a single E. grandis clone and the effect of inbreeding on progeny performance, as reported in our study, illustrate how outcrossing rates in open pollinated stands can influence the interpretation of genetic variability observed in progeny tests conducted in plant breeding (Squillage, 1974; Ghai, 1982; Mori, 1993).

ACKNOWLEDGMENTS

We would like to thank FINEP and Aracruz Celulose/SIF for financial support of this research.

RESUMO

Utilizamos aloenzimas como marcadores para estimar o grau de hibridação natural entre Eucalyptus grandis e E. urophilla em um pomar de semente híbrida comercial instalado no Espírito Santo, Brasil. Esse pomar, que compreende 7,4 hectares, foi plantado em 1982 e é constituído por 267 parcelas hexagonais. Em cada parcela a matriz E. grandis, produtora de sementes, está cercada pelos seis clones de E. urophilla, doadores de pólen. A taxa de fecundação cruzada (t_m), estimada simultaneamente por vários locos de aloenzimas, foi da ordem de 70,2%, variando entre 33 e 99% em árvores individuais. Genótipos encontrados na progênie contendo alelos não presentes nos sete clones parentais permitiram detectar a contaminação do pomar por pólen de outra origem. Esses genótipos constituíram 14,4% da progênie híbrida. A origem do pólen contaminante foi atribuída a povoamentos de eucaliptos vizinhos, isolados do pomar por uma faixa de floresta nativa com 400 metros de largura. Os híbridos e as plantas resultantes de autofecundação puderam ser identificados por meio de seus genótipos de aloenzimas. Todas as plantas examinadas foram transplantadas para o campo. Medições feitas no segundo e terceiro ano após o transplante mostraram que o crescimento da progênie de autofecundação foi 30% inferior ao da progênie híbrida.

(Received November 25, 1996)

• Adams, W.T. and Birkes, D.S. (1989). Mating patterns in seed orchards. In: Proceedings of the Twentieth Southern Forest Tree Improvement Conference. June 26-30, Charleston, South Carolina, pp. 75-86.
• Alfenas, A.C., Peters, I., Brune, W. and Passador, G.C. (1991). Eletroforese de Proteínas e Isoenzimas de Fungos e Essęncias Florestais Imprensa Universitária, Universidade Federal de Viçosa, Viçosa.
• Ashton, D.H. and Sandiford, E.M. (1988). Natural hybridization between Eucalyptus regnans F. Muell., and E. macrorhyncha F. Muell. in the Cathedral Range, Victoria. Austr. J. Bot. 36: 1-22.
• Bertolucci, F., Rezende, G. and Penchel, R. (1995). Produçăo e utilizaçăo de híbridos de eucalipto. Silvicultura 51: 12-16.
• Brandăo, L.G., Campinhos, E. and Ikemori, Y.K. (1984). Brazil's new forest soars to success. Pulp Pap. Int. 26: 38-40.
• Brown, A.H.D., Matheson, A.C. and Eldridge, K.G. (1975). Estimation of the mating system for Eucalyptus obliqua L'Herit. using allozyme polymorphism. Austr. J. Bot.23: 931-949.
• Brune, A. and Zobel, B. (1981). Genetic base populations, gene pools and breeding populations for Eucalyptus in Brazil. Silvae Genet. 30: 4-5.
• Burgess, I.P. and Bell, J.C. (1983). Comparative morphology and allozyme frequencies of Eucalyptus grandis Hill ex Maiden and E. saligna Sm. Austr. For. Res. 13: 133-149.
• Burgess, I.P., Bell, J.C. and Van Wyk, G. (1985). The identity of trees currently known as Eucalyptus grandis in the Republic of South Africa based on isozyme frequencies. South African For. J. pp. 24-30.
• Bush, R.M. and Smouse, P.E. (1991). The impact of electrophoretic genotype on life history traits in Pinus taeda Evolution 45: 481-489.
• Campinhos, E.N. (1992). Utilizaçăo de isoenzimas para a determinaçăo do grau de hibridaçăo em um pomar de semente de eucalyptus Monografia, Universidade Federal de Viçosa, Viçosa.
• Clayton, J.W. and Tretiak, D.N. (1972). Amine-citrate buffers for pH control in starch gel electrophoresis. J. Fish. Res. Board Can. 29: 1169-1172.
• Eldridge, K.G. and Griffin, A.R. (1983). Selfing effects in Eucalyptus regnans Silvae Genet. 32: 216-221.
• Eldridge, K., Davidson, C., Hardwood, G. and Van Wyk, G. (1993). Eucalypt Domestication and Breeding Oxford University Press, Oxford.
• Ghai, G.L. (1982). Covariances among relatives in populations under mixed self-fertilization and random mating. Biometrics 38: 87-92.
• Griffin, A.R., Moran, G.F. and Fripp, J. (1987). Preferential outcrossing in Eucalyptus regnans F. Muell. Aust. J. Bot. 35: 465-475.
• Griffin, A.R., Burgess, I.P. and Wolf, L. (1988). Patterns of natural and manipulated hybridisation in the genus Eucalyptus L'Hérit. A review. Aust. J. Bot. 36: 41-66.
• Hardner, C.M. and Potts, B.M. (1995). Inbreeding depression and changes in variation after selfing in Eucalyptus globulus ssp. globulus Silvae Genet. 44: 46-54.
• Hopper, S.D. and Moran,G.F. (1981). Bird pollination and the mating system of Eucalyptus stoatei Austr. J. Bot.29: 625-638.
• Ikemori, Y.K. and Campinhos, E. (1983). Produçăo de sementes de Eucalyptus grandis x Eucalyptus urophylla por polinizaçăo aberta. Resultados preliminares. Silvicultura, 8: 306-308.
• Junghans, T., Peters-Robinson, I., Bertolucci, F.L. and Alfenas, A.C. (1998). The use of self-incompatibility in the production of hybrid eucalypt seed by 'Aracruz Celulose' in Brazil. Genet. Molec. Biol. 21:
• Martins, F.C.G. and Ikemori, Y.K. (1987). Produçăo de híbridos de eucalipto na Aracruz. Trabalho apresentado na Reuniăo sobre Técnicas para Produçăo de Híbridos. ESALQ-USP, Piracicaba.
• Mitton, J.B. (1989). Physiological and demographic variation associated with allozyme variation. In: Isozymes in Plant Biology (Soltis, D.E. and Soltis, P.S., eds.). Discorides Press, Portland, pp. 127-145.
• Moran, G.F. and Bell, J.C. (1983). Eucalyptus. In: Isozymes in Plant Genetics and Breeding Part B (Tanksley, S.D. and Orton, T.J., eds.). Elsevier, Amsterdam, pp. 423-441.
• Moran, G.F. and Brown, A.H.D. (1980). Temporal heterogeneity of outcrossing rates in Alpine Ash (Eucalyptus delegatensis R.T. Bak.). Theor. Appl. Genet. 57: 101-105.
• Moran, G.F., Bell, J.C. and Griffin, A.R. (1989). Reduction in levels of inbreeding in a seed orchard of Eucalyptus regnans F. Muell. compared with natural populations. Silvae Genet. 38: 32-36.
• Mori, E. (1993). Variabilidade Genética isoenzimática em uma populaçăo de Eucalyptus grandis Hill ex Maiden submetida a diferentes intensidades de seleçăo. Doctoral thesis, Escola Superior de Agricultura "Luiz de Queiroz", Piracicaba.
• Phillips, M.A. and Brown, A.H.D. (1977). Mating system and hybridity in Eucalyptus pauciflora Austr. J. Biol. Sci. 30: 337-344.
• Potts, B.M., Potts, W.C. and Cauvin, B. (1987). Inbreeding and interspecific hybridization in Eucalyptus gunnii Silvae Genet. 36: 194-199.
• Pryor, L.D. and Johnson, L.S.A. (1981). Eucalyptus, the universal Australian. In: Ecological Biogeography of Australia (Keast, A., ed.). Junk, The Hague, pp. 499-536.
• Ritland, K. (1983). Estimation of mating systems. In: Isozymes in Plant Genetics and Breeding Part A (Tanksley, S.D. and Orton, T.J., eds.). Elsevier, Amsterdam, pp. 289-302.
• Ritland, K. (1990). A series of FORTRAN computer programs for estimating plant mating systems. J. Hered. 81: 235-237.
• Ritland, K. and Jain, S. (1981). A model for the estimation of outcrossing rate and gene frequencies using n independent loci. Heredity 47: 35-52.
• Squillace, A.E. (1974). Average genetic correlations among offspring from open pollinated forest trees. Silvae Genet. 23: 149-156.
• Wright, S. (1969). Evolution and the Genetics of Populations. Vol. 2. The Theory of Gene Frequencies University of Chicago Press, Chicago.

Publication Dates

• Publication in this collection
  23 Feb 1999
• Date of issue
  Sept 1998

History

• Received
  25 Nov 1996