Where Did E. maimaiga Presently in North America Come From?

This question has been asked during virtually every presentation I’ve made about this fungus. E. maimaiga occurs in Japan, Korea, north-eastern China and the Russian Far East. Based on molecular studies, we now know that the fungal strain present in North America originated from Japan (Nielsen et al ., 2005).

Could this fungus have been successfully introduced by Speare and Colley in 1910–1911? This does not seem likely because the fungal strains established in North America today are quite different from isolates of E. maimaiga collected near Tokyo. Also, it seems unlikely that this fungus would not have been found infecting gypsy moth larvae between 1911 and 1989, especially because appropriate weather conditions for epizootics occurred several times during this timespan (Weseloh, 1998) and massive field surveys of gypsy moth pathogens were undertaken. Could this fungus have been successfully introduced in 1985–1986? Once again, molecular studies showed that the isolate that was released was more different from the isolate recovered in 1989 than would have been expected, given that changes in the released isolate would have to have occurred between release of the fungus in 1985–1986 and isolation of strains during and after the 1989 epizootics. At present, the best-accepted explanation for the origin of E. maimaiga present today in North America is that an accidental introduction must have occurred at some time since 1971 (Weseloh, 1998).

Studies of Effects of E. maimaiga on Non-target Lepidoptera

The gypsy moth and its natural enemies live in naturally occurring forests as well as urban forests. When this virulent pathogen was discovered in North America in 1989 and people began discussing redistribution and potential use for inundative or inoculative augmentation, questions about environmental safety were posed: Would E. maimaiga have unexpected negative impacts on the native fauna? We conducted studies to address this question from 1992 to 2001. We quickly learned that E. maimaiga could only potentially affect lepidopteran larvae that are present in the spring when gypsy moth larvae are also present. First, larvae of representatives of the many species of forest Lepidoptera present in the forest during spring at the same time as gypsy moth larvae were collected and challenged with the fungus in the laboratory. Under these optimal conditions E. maimaiga could infect about one-third of the 78 species challenged, but infection was only consistently high among the three species of tussock moths (Lymantriidae) tested plus one laboratory colony of a hawk moth (Sphingidae) (Hajek et al ., 1995a). We were surprised by these results because in the forest we virtually never saw cadavers of any species other than gypsy moth hanging on tree trunks during epizootics. Therefore, next we collected native lepidopteran larvae in forests in Virginia, Michigan and New York during spring. While some of the gypsy moth larvae collected were always found to be infected, only two of 1511 larvae belonging to 52 species of non-targets from moderate-density gypsy moth sites in Virginia were infected and both of these individuals belonged to relatively common species (Hajek et al ., 1996). These results clearly showed that the range of hosts that this natural enemy can infect in the laboratory (the physiological host range) was not equivalent to the range of species infected in the field (the ecological host range).

Meanwhile, during other studies we had found high levels of infection among gypsy moth larvae caged on the soil. We had not investigated the potential for infections in non-target Lepidoptera at the soil surface. Late-instar gypsy moth larvae often spend the daytime hours resting in the leaf litter, and our studies have shown that this behaviour, which is highly unusual for most lepidopteran larvae, results in high levels of infection (Hajek, 2001). Studies of lepidopteran larvae in the leaf litter yielded extremely low levels of infection except among gypsy moth larvae (Hajek et al ., 2000).

The only non-target Lepidoptera consistently infected in the laboratory were lymantriids, which belong to the same family as gypsy moth. No lymantriids occurred in the leaf litter during our studies and few had been collected during our foliage studies. Therefore, we needed to focus our efforts on larvae of this family of endemics, which are often less common. Over 5 years, all lymantriids present in 18 plots in the forested mountains of Virginia and West Virginia were reared to detect infections (Hajek et al ., 2004). Among the seven species of native lymantriids collected, only three were ever found to be infected by E. maimaiga and never at 50%, although infection among gypsy moth larvae was high. In summary, our extensive studies have demonstrated that this obligate pathogen is highly host specific, although limited levels of infection among a few species of sympatric lepidopteran larvae are occasionally possible.

Use of E. maimaiga for Control

The major uses of E. maimaiga for control have been introductions of small amounts of field-collected resting spores to areas along the edge of spread of gypsy moth. All indications suggest that the fungus will spread into these areas eventually but at present there are no data documenting how long this would take. When gypsy moth first colonizes a new area, population dynamics are chaotic along the leading edge of spread, and during this time land managers and the public are eagerly searching for means for controlling this pest. Cadaver- or soilborne resting spores have been field collected and released in at least ten states. 2 In addition, cadavers bearing resting spores were collected in North America and released in Bulgaria, Siberia and possibly Romania (Hajek et al ., 2005).

Results from limited studies suggest that this fungus could be useful when applied in areas where it already exists to augment pre-existing levels of fungal inoculum and cause earlier initiation of fungal epizootics. However, to use E. maimaiga in this way, there would need to be a source of fungal inoculum. E. maimaiga is an obligate pathogen and is not easy to grow in the laboratory. Mass production of E. maimaiga is not feasible at present, although there are yearly requests for material to release in areas recently colonized by gypsy moth. The best stage for distribution is the resting spore, which is constitutively dormant after maturation. Although progress toward mass production of resting spores has been slow, these spores can now be produced in vitro ; dormancy can be prevented and methods for storage have been developed. Time will tell whether methods for mass production of this fungus will be developed in the future to enable vastly improved availability.

Notes

1 Classical biological control shipments today are maintained in quarantine before field release, in part to make sure that only the species of interest is released.

2 Permits must be obtained from state and federal agencies before soil can be moved.