Host plant resistance and epidemiology of Fusarium ear rot in maize.

Abstract

This thesis examines fusarium ear rot epidemiology in maize (Zea mays L.) and host plant resistance to Fusarium species causing ear rot, in eastern Canada, through three major groups of experiments. A three-year disease survey that determines the causal agents of maize ear rot in Ontario; a three-year study on the stability of different genotypes to two different inoculation techniques across different environments; and a two-year study on fungal infection from an inoculum source using methods to detect infection spread. A survey of maize ear rot was conducted in eastern and western Ontario from 1991-1993 in order to develop a predictive model of ear rot incidence. Field data were collected on disease incidence, insect and/or bird damage, ear tip infection (silk), rainfall, and temperature. Isolations were made in the laboratory from samples showing evidence of infection, and the five prevalent Fusarium species (F. graminearum, F. moniliforme, F. proliferatum, F. sporotrichioides, and F. subglutinans) were enumerated. The profile of Fusarium species isolated varied considerably from year to year, and in a certain proportion between sites. Fusarium subglutinans was the most frequently isolated species and was associated with low rainfall in July. Fusarium moniliforme and F. proliferatum occurred mostly following the hot, dry summer of 1991. Fusarium graminearum presence increased significantly with higher precipitation in July. Presence of F. sporotrichioides was not associated directly with the incidence of ear rot. Insect damage and infection through the ear tip were the most important mode of fungal entry into maize ears. In the second major field study undertaken in east-central Ontario and south-western Quebec, the interaction between genotypes and environments was described using a Joint Regression Model and an Additive Main effects and Multiplicative Interaction effects (AMMI) model. The AMMI model was able to capture 68% of the interaction across both inoculation methods, compared to 84% for the Joint regression model. The number of cool and wet days (relative humidity over 80%) observed in August was more critical for silk channel inoculation infections, and the number of wet days was more critical to achieve severe ear rot with the kernel-stab inoculation technique. A disease spread study showed that the Power Law model is a suitable model to predict fusarium ear rot spread over a distance below five metres from a small inoculum source of F. graminearum (strain sc DAOM 180378). (Abstract shortened by UMI.)