Meta-analysis of the linear relationship between soybean yield loss and rust severity from uniform fungicide trials.

• Autor(a) principal: PONTE, E. M. del
• Data de Publicação: 2010
• Outros Autores: MAIA, A. H. N., ESKER, P. D., GODOY, C. V.
• Idioma: eng
• Título da fonte: Repositório Institucional da EMBRAPA (Repository Open Access to Scientific Information from EMBRAPA - Alice)
• Texto Completo: http://www.alice.cnptia.embrapa.br/alice/handle/doc/663067
• topic: Soja; Doença de planta; Ferrugem; Soybeans; Rust diseases; Plant diseases and disorders

To quantify yield loss in soybean as affected by rust severity, meta-analysis was used to examine data from uniform fungicide trials (UFTs) conducted across 25 locations and 5 years (2003?2008) in the main soybean production regions of Brazil. Trials (n = 81) were selected based on predefined criteria and examined for the linear relationships between soybean yield loss (%) and rust severity (%) at the R6 growth stage. Slopes of the yield loss × severity relationship were estimated by linear regression. The influence of potential moderator variables on slopes was examined by a random effects model. Forty-one trials were excluded due to the presence of influential observations in the negative linear relationship between yield and severity. There was considerable variation among slope estimates for the increase in yield loss with respect to the change in soybean rust severity (0.31?1.81 p.p./p.p.). A significant and positive relationship between soybean rust severity and yield loss was found. For every increase of 1 p.p. in soybean rust severity, the average increase in yield loss was 0.71 p.p. (95% CI = 0.60?0.82 p.p.). The moderator variable timing of disease onset and severity class accounted for 17 and 16% of the overall and between-trial variability of slopes, respectively. This study provides an approach for quantitative risk assessments of rust-related soybean yield loss, an important step for an economic evaluation of alternative rust management strategies.