INTERAÇÃO GENÓTIPO-AMBIENTE E GANHOS NA SELEÇÃO EM TESTES DE PROGÊNIES DE Corymbia citriodora HOOK.
Souza, Bruno Marchetti
MetadataShow full item record
Corymbia citriodora is one of the most cultivated hardwood species by small farmers in Brazil, it combines a high growth rate with high wood density. Its timber is used for a wide range of products, such as poles, railway sleepers, posts, scantlings, wharf and construction lumber. Despite its commercial importance, this species does not have a formal breeding program. Genotype-by-environment interaction (GE) is one of the most important elements in the management of a tree breeding program to define breeding zones and to select genotypes that are targeted to specific environmental conditions. The aim of this research was to estimate genetic parameters to understand GE patterns based on C. citriodora progeny tests. The experiment was established using 56 open-pollinated families in three different locations within the Luiz Antônio‟s experimental station in state of São Paulo, Brazil. The three sites differ contrastingly on their soil type (LVM: Red Latosol, QN: Quartzarenic Neosol, LVA: Clay Latosol). The following traits were measured at 30 years of age: total height (HT), diameter at breast height (DBH), stem form and survival. Based on these data the individual stem volume (VOL) was calculated. Statistical analyses and genetic parameters were estimated by REML/BLUP methodology. The MHPRVG (harmonic mean relative performance of genetic values) predicted by BLUP was used to analyze the productivity, stability and adaptability. Based on the genetic values, the genetic gains were estimated considering 4 selection methodologies combined with 10 intensity levels. And finally, an optimal intensity of selection regarding the diversity was calculated for methodology. The survival ranged from 38% (QN site) to 55% (LVC site). The overall mean values for HT and DBH were 26.79 m and 21.5 cm, respectively. The highest growth rates were found in the LVC soil. The GE was not significant to all growth traits. A complex GE was detected only for survival, proving the importance of choosing the right genotypes of the species to specific sites. The joint analysis showed a significant difference between families for DBH, survival and VOL. In the individual analyses, progenies growth traits performances were considerably different only in the QN site. This site also presented high narrow-sense heritability values. The estimated gains were low to moderate peaking 20% on QN site. The selection method that provides superior gains was the family selection. Considering the optimal selection regarding the diversity, the gains ranged from 3.36% to 5.68%, selecting different numbers of individuals. In summary, the families studied here presents potential to obtain considerable genetic gains through selection. However to keep these gains continuous over the next selection cycles it is necessary to incorporate more genetic diversity in order to increase the diversity observed among the progenies.