Pathogenicity and virulence of Entomophaga maimaiga Humber, Shimazu, Soper and Hajek (Entomophthorales: Entomophthoraceae), a fungal pathogen of gypsy moth, Lymantria dispar L. (Lepidoptera: Lymantriidae)

• Main Author: Thomas, Stephen R
• Language: ENG
• Published: ScholarWorks@UMass Amherst 2005
• Subjects: Entomology|Microbiology
• Online Access: https://scholarworks.umass.edu/dissertations/AAI3193947

Invertebrate pathologists have multiple definitions for the terms pathogenicity and virulence, and these definitions vary across disciplines that focus on host/pathogen interactions. We surveyed various literatures and found most define pathogenicity as the broader term, which incorporates virulence. Virulence is seen as the severity of disease manifestation that can only be measured in infected individuals. The effectiveness of pathogens as biological control agents is thought to be related to their virulence. E. maimaiga, a fungal pathogen of gypsy moths, has had variable success as a biological control agent in North America. I compared the pathogenicity and virulence of American isolates of E. maimaiga to Japanese isolates with injection and dipping bioassays and found no consistent difference in LD50 and LT50 between the isolates from each country, when tested on a laboratory colony. I also found contrary to some reports, the Michigan isolate was the fastest killing and had the lowest LD50 of any isolate that I tested. I then compared the virulence of a Japanese and a North American isolate of E. maimaiga on gypsy moths from Japan, and North America as well as the standard laboratory strain (New Jersey strain) from North America. I found that North American gypsy moths were the most susceptible for both isolates in terms of median number dying and time till death of infected larvae. We also found that the fungal isolates produced more spores when they infected hosts on which they were naturally found. When I examined the within host dynamics of E. maimaiga, I found hyphal bodies seen as early as 4 days after infection which corresponded with the within host time needed to survive the premature death of the host. After four rounds of selection for increased growth within the host, I saw a decline in replication rate of the pathogen; however, both median number killed and time to kill increased as measured with injection bioassays. These results were also compared to theoretical predictions made by Ebert and Weisser's (1997) model for optimal virulence.