The population dynamics of a “leading edge” (= at the edge of the expanding gypsy moth invasion) gypsy moth, Lymantria dispar (L.), population was monitored for 3 years (1995–97), with emphasis on the interactions of the gypsy moth nuclear polyhedrosis virus (LdNPV) and the fungus Entomophaga maimaiga Humber, Shimazu, & Soper. Gypsy moth populations in the woodlots varied from very sparse to high (potentially defoliating) levels. LdNPV was strongly density dependent, being confirmed only from the higher populated woodlots. In contrast, the fungus was similarly active in both sparse and highly-populated woodlots. In 1995, the fungal epizootic developed late in the season, with most larvae succumbing during stadia 5–6 and producing mainly resting spores (azygospores). Estimated mortality due to fungus averaged 68% in high-density plots and 85% in low-density plots. LdNPV mortality occurred in a two-wave epizootic, although second-wave LdNPV mortality was undoubtedly reduced because of the reduction of late-season larvae due to fungus activity. Estimated mortality due to LdNPV averaged 14% in highly-populated plots and 1% in low-population plots. In 1996, high levels of fungal-induced mortality occurred earlier in the gypsy moth season than in the previous year. Most gypsy moth larvae in 1996 died in a mid-season wave of fungal-induced mortality, with necropsied cadavers containing only conidia. This resulted in relatively few larvae surviving to late instars. At this time, a second wave of fungus-induced mortality occurred, with over half of the necropsied cadavers containing resting spores. The depletion of the gypsy moth populations by the fungus in 1995 resulted in a greatly reduced first wave of LdNPV in all plots in 1996, and perhaps due to the early appearance of the fungus in 1996, LdNPV was nearly absent from late-season larvae collected from all plots. In 1997, gypsy moth populations were uniformly low, and no dead larvae were found in any of the plots.
