1ARC-Roodeplaat Vegetable and Ornamental Plant Institute (Western Cape), Private Bag X1 Elsenburg 7607, South Africa; 2Department Plant Pathology, University of Stellenbosch, Private Bag X1, Matieland, Stellenbosch 7602, South Africa; 3ARC-Roodeplaat Vegetable and Ornamental Plant Institute, Private Bag X293, Pretoria 0001, South Africa; 4ARC-Agrimetrics, Infruitec, Private Bag X5013, Stellenbosch 7599, South Africa

Background and objectives
Late blight of potatoes (Solanum tuberosum L) and tomatoes (Lycopersicum esculentum Mill) is caused by Phytophthora infestans. With the emergence of both mating types and new genotypes of the pathogen in Europe, the USA and other countries, the pathogen that caused the Great Irish famine still remains a threat to potato and tomato production today [1]. The emergence of strains with resistance to phenylamide fungicides, the only systemic fungicide with curative activity for late blight control, further contributed to the resurgence of late blight as a major, economically important disease of potatoes and tomatoes [1,2]. The source of resistant strains has frequently been related to recent migrations of new genotypes from Mexico to other parts of the world [1]. South Africa experienced severe late blight epidemics during the 1995 to 1997 growing season, especially on potatoes. Epidemics were difficult to control and losses occurred. A study was undertaken to establish whether the reported correlation between disease severity, genotype and phenylamide resistance contributed to the severe late blight epidemics in South Africa.

Materials and methods
A country-wide late blight survey was conducted, and 218 single lesion isolates were evaluated for phenylamide sensitivity and mating type. A subset of 90 isolates was analysed for glucose phosphate isomerase (Gpi) allozyme banding patterns, using polyacrylamide electrophoresis and iso-electric focussing. Mating type composition was determined by pairing isolates with known A1 and A2 mating strains on 10% clarified V8 medium. Phenylamide sensitivity was assessed using the in vitro leaf disk method [2]. Five concentrations of metalaxyl, viz. 0(control), 0.1, 1, 10 and 100 ;mg a.i./ml were used. A regression analysis was done for each isolate and the EC50 (effective concentration resulting in 50% growth reduction) values were calculated. A cluster analysis was done on these values to group data.

Results and conclusions
P. infestans isolates were obtained from potatoes sampled at 51 different localities, and 26 isolates from tomatoes sampled at nine localities. All isolates were of the A1 mating type. Allozyme analysis revealed that all isolates tested had the 86/100 banding pattern indicating that the 'old populations' that were found outside Mexico prior to the mid 1970s, are present in South Africa. Cluster analysis of the calculated EC50 values, showed that there were five different sensitivity groups, namely (1) a sensitive group with EC50 0-2.5 ;mg a.i./ml, (2) sensitive intermediate group with EC50 values between 2.6-7 ;mg a.i./ml, (3) intermediate group with EC50 between 7.1-15 ;mg a.i./ml, (4) intermediate resistant group with EC50 between 15.1-30 ;mg a.i./ml and (5) resistant group with EC50 values more than 30 ;mg a.i./ml. Based on the calculated EC50 values there appears to be a distinct split at the 30 ;g a.i./ml concentration differentiating resistant from sensitive isolates. This suggests that 30 ;g a.i./ml is the discriminatory concentration of South African P. infestans isolates. Since only the old population appears to occur in South Africa, the increased frequency and severity of late blight can not be attributed to the migration of new genotypes, but rather seems to be correlated to phenylamide resistance and conducive weather patterns. In view of the present situation strict control measures must be taken to prevent entry of the A2 mating strain and new populations of P. infestans into South Africa.

1. Fry WE, Goodwin SB, 1997. Plant Disease 81, 1349-1357.
2. Staub TH, Sozzi FJ, Gisi U, 1992. EPPO Bulletin 22, 306-309.