A study of strain variation in erwinia salicis in relation to the epidemiology of watermark disease

Abstract

Erwinia salicis (Day) Chester is the causal agent of watermark disease in several species of Salix L. (willow) particulary in Salix alba var. caerulea (cricket-bat willow). Virtually nothing is known about the epidemiology of watermark disease and the infection route, infection process, transmission and survival of E. salicis is not known. The work presented here investigates the epidemiology of the disease by distinguishing between isolates of E. salicis and then examining their geographical distribution. The first step in this study was to reliably identify the isola- tes without a pathogenicity test. This was done by comparing the cul- tural, biochemical and serological characteristics of authentic iso- lates of E. salicis whose pathogenicity has been confirmed, with those of the suspect cultures. The most important tests were ELISA and API20E profilingi described in Chapter 2. Chapter 3 included the carbon source utilisation assay, bacteri- ophage sensitivity assay and bacteriocin assays. Three bacteriophages specific for E. salicis were isolated from natural sources and a total of 5 bacteriophages were used in a bacteriophage typing scheme. Bacte- riophage sensitivity assays and galactose utilisation tests demonst- rated the differentiation between Dutch and English isolates of the bacterium, but did not reveal sufficient variation for use in the epidemiological study. Serotyping was employed in Chapter 4 and two antisera were raised against two isolates of E. salicis. Only two groups, based on presence or heat-stable '0' antigen, were identified with one of the antisera, indicating the serological homogenicity of E. salicis. Chapter 5 describes the development and optimization of multilo- cus enzyme electrophoresis typing for E. salicis. A total of seven enzymes were chosen for the typing study. Chapter 6 describes the application of multilocus enzyme elect- rophoresis typing for the differentiation of E. salicis isolates. This method indicated extensive variation between the isolates tested and 23 ETs could be identified in a population of 78 isolates of E. sali- cis. It was concluded from this study that in Wiltshire, in a severely diseased plantation, the disease did not spread from tree and infec- tion has presumably arisen from infected propagating material.