1Tree Pathology Cooperative Programme, Forestry and Agriculture Biotechnology Institute (FABI), University of Pretoria, Pretoria, 0002, South Africa; 2Department of Plant Pathology, Iowa State University, Ames, IA 50011, USA

Background and objectives
Armillaria is a cosmopolitan root pathogen throughout temperate and most of the tropical regions of the world. Armillaria species kill a wide range of native and planted conifer and hardwood trees as well as shrubs in gardens and parks. Dead and dying oak (Quercus) and other trees showing typical symptoms of Armillaria root rot have recently been found in the historical Company Gardens, which were established 350 years ago by early settlers in Cape Town to provide fresh products for Dutch East India Company ships. The aim of this study was to identify and characterize the Armillaria species responsible for the disease.

Materials and methods
Isolates of Armillaria were obtained using standard methods. For molecular comparisons, DNA was extracted and used as template for PCR amplification of a portion of the intergenic spacer (IGS) region of the rRNA operon with primers P-1 and 0-1. The amplified IGS products were digested with the restriction endonuclease AiuI to determine differences in RFLPS between the Cape Town isolates as well as between these isolates and known RFLPs of the Northern Hemisphere species of Armillaria. Phylogenetic comparisons were made by sequencing the IGS region and comparing it with known sequences for the Northern Hemisphere Armillaria species. The Cape Town isolates were conclusively identified with the aid of diploid-haploid matings with haploid Armillaria tester strains. The number of Armillaria clones in the Company Gardens was determined using diploid-diploid crossings.

Results and conclusions
The AluI RLFP profiles of the Armillaria isolates from Cape Town were similar to each other. This indicates that they belong to one species. The RFLP profiles of these isolates were similar to those of A. mellea from Europe and Western North America [1]. We therefore believe that the Cape Town isolates represent A. mellea. This was further confirmed by the sexually compatible reaction observed in diploid-haploid matings between the Cape Town isolates and haploid A. mellea tester strains from France. Diploid-haploid matings between the Cape Town isolates and A. mellea from North America gave incompatible reactions. It was thus also expected that Cape Town and European isolates of A. mellea would be more closely related to each other than to A. mellea from North America. Phylogenetic trees constructed from IGS sequence data for isolates from Cape Town and those from Eastern and Western North America, Europe, Japan and South Korea grouped the South African isolates with A. mellea from Europe. This showed that the A. mellea from South Africa is most closely related to A. mellea isolates from Europe. Diploid-diploid crossings revealed the presence of only one clone of A. mellea in the Company Gardens. The age of this clone was estimated to be at least 280 years based on the size of the infection centre and an estimate of 0.8 ;m per year in linear growth rate. We believe that A. mellea was introduced into Cape Town by the early Dutch settlers. This would be consistent with historical records showing that potted citrus trees were planted on the same site where oak trees are currently infected with A. mellea.

1. Harrington TC, Wingfield BD, 1995. Mycologia 87, 280-288.