Use of remote sensing and geographic information systems to study the epidemiology of oak wilt in Dallas, Texas

Abstract

Color infrared (CIR) aerial photography and a geographic information system (GIS/ArcView) were used to categorized forty-four oak wilt centers in a 50 square mile area to study why disease incidence, caused by Ceratocystis fagacearum, has increased in Dallas, Texas. Twenty-four C. fagacearum isolates were obtained from the disease centers and were characterized for growth, fungicide sensitivity, sporulation and mating type (A &B) on two culture media. Mating types were found in equal proportions within the study area, (A=11, B=13). Distinct clustering patterns existed for isolates for each mating type, growth, and sporulation. Two clusters were found to include both mating types suggesting sexual reproduction of the fungus is occurring within the study area. Slope, aspect, soil phase and elevation at the forty four disease centers were measured to determine whether environmental variables influence the occurrence of oak wilt. Results from principal component analysis of the environmental variables were not significant and uninformative in explaining the occurrence of disease in the area. Chi-square and geostatistical analyses revealed a distinct spatial pattern of the occurrence and spread of oak wilt in the east-west direction. Statistically more disease centers (p[]0.002) occurred on streets in the east-west direction. Variograms for disease centers and the phenotype characteristics of mating type revealed similarity in the east-west direction for a distance of 1500m. Variograms for medium growth rate revealed spatial continuity of centers for distances of 3000m in the east-west direction. Variograms for low spore production revealed spatial continuity of centers for distances of 3000m in the northwest-southeast direction. Variograms for high spore production revealed spatial continuity of centers for distances of 2000m in the east-west direction.