Analysing Crop-Loss in a Bean Rust Pathosystem. II. Severity - Damage Relationships

Abstract

Field experiments were conducted from 1988 to 1993 at Ambo, Awassa and Debre Zeit, Ethiopia, to determine relationships of damage to crop growth and rust intensity in a bean rust pathosystem. A systemic fungicide (oxycarboxin) was applied at 5 spray frequencies to create differences in rust epidemic. The experiment was carried out in a production situation with low external inputs. Critical point models were examined based on rust severity and yield loss. Multiple regression models were developed for yield and yield loss by attempting to incorporate leaf area index, and rust incidence and severity in three canopy layers at one or more growth stages. Multiple regression models based on leaf area index and rust severity at growth stages R6 (flowering) and R7B (pod formation)  fitted the data with R values of 0.85. Addition of an incidence parameter did not improve the goodness of fit. Single point
models developed at pod formation showed better fit than those developed at or prior to flowering. Models based on crop and disease assessments of the upper canopy layer produced good fit, but the goodness of fit was improved by including data from
the middle canopy. For rust management purposes, crop and disease assessments from before flowering up to seed filling stages can be used. For survey purposes models based on assessments at the pod filling stage will be satisfactory.