Biometric analysis of cassava clones

• Autor(a) principal: Almeida, Ísis Fernanda de
• Data de Publicação: 2019
• Outros Autores: Jesus, Adriana Madeira Santos, Almeida, Ramon Vinicius de, Leite, Bianca Stefáni Arantes, Prado, Mayara Cardoso do
• Tipo de documento: Artigo
• Idioma: eng
• Título da fonte: Revista de Ciências Agrárias (Belém. Online)
• Texto Completo: https://ajaes.ufra.edu.br/index.php/ajaes/article/view/2950
• topic: Mixed models, genotypic effects, analysis of deviance, path analysis; clones de mandioca

In the last ten years cassava roots represented the fourth most produced commodity in Brazil. Given its commercial importance, higher yields are constantly sought in breeding programs. This study was aimed at conducting a biometric analysis of cassava clones based on the estimation/prediction of genetic parameters and correlated genetic gain using mixed models and path analysis, respectively. Forty-eight clones were evaluated in a randomized block design with two replicates. The experiment was carried out in northern Minas Gerais in 2010. The agronomic characteristics evaluated were plant height (PH), fresh weight of aerial parts (FWAP), fresh root weight (FRW), fresh weight of commercial roots (FWCR), root length (RL), and root diameter (RD). These traits were evaluated at six and twelve months after planting. All traits examined were significantly affected by genotype. FWAP, RL and RD changed according the time of harvesting and RL was superior at six months. Accuracy was highest for PH (0.90) and lowest for FRW and FWCR (0.64). UFLA 42 was the most commercially productive. The trait RL exhibited the highest gain via correlated response to FWCR at twelve months after planting. At six months after planting, no traits were suitable for indirect selection. The traits PH and FWAP had little relevance as secondary components in path analysis.