Department of Plant Pathology, University of Stellenbosch, Private Bag X1, Matieland 7602 South Africa

Background and objectives
Evidence on the infection pathway, and on latency in the life cycle of Botrytis cinerea on grape, is primarily circumstantial. In California, McClellan and Hewitt [1] found that B. cinerea invades the stigma and style and then becomes latent in necrotic stigma and style tissue at the style end of the berry. At véraison or later the pathogen resumes growth to rot the berry. In France, however, Pezet and Pont [2] found no evidence of this infection pathway and showed that latent infection was predominantly pedicel-associated. De Kock and Holz [3] found no relation between early infection and subsequent disease development or post-harvest decay on table grapes. More information is therefore needed on the infection pathway, and on latency in the life cycle of B. cinerea in planning disease-control strategies. In this study, infection pathways and latency of B. cinerea were determined on naturally infected table grape bunches in the Western Cape province.

Materials and methods
Bunches were obtained from vineyards during the growing season, and from commercial packers after cold storage. Infection was monitored by incubating berry skin segments in petri dishes on Kerssies' B. cinerea-selective medium and by incubating detached berries bearing only the receptacle part of the pedicel, or rachis sections bearing a group of three to seven berries at a high relative humidity in moist chambers. Material used in moist chambers was first treated with paraquat (30 s in 30 ml/l water) or low temperature (1 h at -12°C) in order to negate active host responses in the cuticle and underlying cells and to show up latent infections. The material was kept at 22°C and examined daily for B. cinerea growth.

Results and conclusions
Isolation studies from vineyard-sampled berries showed that berry cheeks were virtually free from B. cinerea infection during all developmental stages. In the case of detached berries bearing the receptacle part of the pedicel, and those on rachis sections, infection was relatively high at pea-size stage. Infection then diminished to low frequencies during later stages. In all berries examined during pea-size to véraison, the pathogen developed first in the receptacle part of the pedicel, from where it colonized the berry. Decay was pedicel-, cheek- and style end-associated with berries sampled later. Overall, 2% of berries were infected; 92.92% of these infections were associated with the pedicel, 7.07% were associated with berry cheeks and 0.01% with the style end. On cold-stored bunches surface-sterilized with sulphur dioxide, percentage berry infection associated with the pedicel was nearly similar to that associated with the berry cheek. None of the berries yielded the pathogen from the style end. However, on unsterile bunches berry infection was predominantly cheek-associated, pedicel-associated infection was at a low level, and a fraction (0.02%) of the berries yielded the pathogen at the style end. These results indicated that latent infection was predominantly pedicel-associated, and confirms the role of the pedicel as an infection pathway as postulated by Pezet and Pont [3]. The study furthermore showed that latency of B. cinerea in grape may be regulated primarily by the pedicel, and confirms the pre-véraison natural resistance [1] of berry cheeks to infection.

1. McClellan WD, Hewitt WB, 1973. Phytopathology 63, 1151-1157.
2. Pezet R, Pont V, 1986. Revue Suisse de Viticulture, Arboriculture et Horticulture 18, 317-322.
3. De Kock PJ, Holz G, 1991, Phytophylactica 23, 73-80.