POPULATION AND POPULATION CYCLES: Established gypsy moth populations remain low for varying periods of time, sometimes permanently. During this phase, predators including especially native mice (Peromyscus spp.) among other vertebrates and invertebrates, as well as native and introduced parasitoids (Diptera, Hymenoptera) exert some degree of control (e.g. Doane and McManus, 1981, Nichols, 1980, FEIS, 1995, Weseloh, 1985, Weseloh et al., 1983) and numerous other publications). Introduced predatory beetles (Calosoma) tend to have more impact at higher densities. Despite predators and parasitoids, populations in vulnerable forest types can increase to a point where natural enemies no longer exert effective control. Populations then build up within about three years to outbreak levels. In peak years of severe outbreaks in oak dominated forests 100% defoliation of all favored to moderately resistant trees often occurs. White oaks and other highly favored species may incur substantial defoliation the year before the general outbreak. Severe defoliation generally occurs for one or two seasons followed by a crash. Occasionally populations will fail to collapse for longer periods, and moderate to severe defoliation may continue to occur locally after generalized outbreaks in neighboring areas have collapsed. Such persistence is (or at least was) most likely in areas recently invaded by the gypsy moth, but since about the late 1990s has been much less frequent than previously due to the fungus Entomophaga maimaiga (Richard Reardon, USFS, pers. comm.). Collapse is most likely after a season with near 100% defoliation of oaks.

An unusual and overlooked situation can persist for years in coastal plain southern New Jersey (especially in Cumberland County 1980s to about 1996 when Entomophaga ended the situation) and may recur farther south locally if that fungus is slow to establish. Key ingredients appear to be low density (5-20%) of canopy oaks and a lot of sweetgum. Early instars concentrate on the scattered oaks, which are defoliated in late May. Older larvae then disperse to sweetgums--which were unacceptable to early instars. The sweetgums may or may not incur moderate to heavy defoliation depending on their density and proximity to oaks. Larvae forced off oaks find abundant food, disperse sufficiently that they are no longer at high density, and tend to produce normal egg masses. Red maple and blueberry are also readily available alternate foodplants but sweetgum seems preferred. The percent stand defoliation remains low but individual oaks are defoliated repeatedly. Because of the greater number (often six or more) of heavy (60-100%) defoliations per decade these oaks on mesic to hydric soils had higher mortality in the 1980s and 1990s than oaks in some xeric sites with more normal crashes. A few sweetgums and hickories were also killed. The Nature Conservancy's Eldora Preserve is an example of this damage pattern and most mature oaks were killed in the 1980s. New Jersey's only significant stand of Quercus nigra south of Dividing Creek lost almost all mature trees of that species, but hundreds (possibly >1000) of saplings and pole-sized water oaks persist as of November 2002.

Heavy defoliation may occur in somewhat interrupted areas of several hundred thousand acres during the worst seasons and thousand acre outbreaks are not unusual. In June 1981 most oak and mixed forests from southern Maine to coastal Connecticut were heavily defoliated. Occasionally in sprayed areas (mainly older reports, e.g. Nichols, 1980) populations can rebound to outbreak levels in three years, but generally they remain low for five or more years once they crash. Prior to the emergence of Entomophaga in 1989 the general rule was an outbreak every six to 12 years in New England etc. In the southern New Jersey Pinelands region many oak-pine forests (especially in Burlington, Cape May and Cumberland Counties) were defoliated several times when the gypsy moth first invaded in the late 1970s-1980s, often with substantial mortality to subcanopy or even canopy oaks. Other similar oak-pine forests in Salem, Ocean, and Atlantic Counties have (as of 2002) never had an outbreak, in many cases without any control efforts. The 1995 FEIS and other sources report similar observations elsewhere. Failure of expected outbreaks to materialize was a problem in the Sample et al. (1996) field studies. Outbreaks often start in highly favored stressed sites such as ridgetops. Some old reports suggested they sometimes started at the edge of developed areas, perhaps due to increased shelter for pupae and reduced predator (Peromyscus) densities (e.g. due to housecats and habitat changes).

Outbreaks are not fully synchronized, so that there are almost always some areas of heavy defoliation in any given season at least along the leading edge of spread. There are years with none over vast areas behind the leading edge. Likewise, even in the worst years (1981 to date) there are always areas with no noticeable defoliation. In New England, eastern New York and northeastern Pennsylvania, outbreaks usually collapse after one to three seasons. Following collapse, it may be difficult to find any stage of the gypsy moth for a few years. This fact has hampered studies of low-level gypsy moth populations so there are some gaps in knowledge of their population dynamics. Since 1989 (see below) frequency of outbreaks in the Northeast has declined markedly and much of southern New England has been outbreak free (as of 2002) since 1981 or 1982.

NATURAL CONTROLS. Outbreak collapse usually involves death of an overwhelming majority of gypsy moth larvae due to some combination of gypsy moth neuclear polyhedrosis virus (NPV), or since 1989 the fungus Entomophaga maimaiga, or starvation. Both pathogens are introduced, although there is some uncertainty of the exact origin of the current fungus, and are generally the most important natural controls in outbreak conditions. The fungus can also provide excellent control in pre-outbreak conditions and has prevented outbreaks since 1989 in large areas of New England. It is now also an important mortality agent in low density populations. Mortality from parasitism can become very high or may remain low in outbreaking populations. It seems to be the consensus that the important cumulative impact of parasitoids and predators is to slow the rate of increase in low-level populations and thus to lengthen the period between outbreaks, more than actually ending or preventing outbreaks.

Several mostly non-native parasitoids utilize gypsy moth larvae and pupae. Some useful references for identification and basic information include Hoy (####), Nichols (1980) and Simons et al. (1979). Nichols regarded nine species as of some importance in Pennsylvania. None of the parasitic wasps sting humans. Weseloh (1985) and Weseloh et al. (1983) and some of the references therein are among the important studies of parasitoid impacts including in low level larval populations. At least two parasitoids are often noticeable in many states. A tiny introduced wasp, Ooencyrtus kuvanae, attacks the egg masses and has five or six generations per year, from about July to December. Due to its small size, it can only destroy eggs near the surface of the masses. Nichols reports 20 to 50% parasitism of eggs. They kill the highest percentage of eggs in the small egg masses laid in outbreak years. These wasps commonly find virtually 100% of egg masses in an area. Another small wasp Cotesia (formerly Apanteles) melanoscelus kills gypsy moth larvae in about the third instar. The larva dies near or attached to the small white cocoon made by the wasp larva. This tiny wasp is apparently a specialist on Lymantriidae (gypsy moth, satin moth, native Orgyia) (Wieber et al., 2003). Most gypsy moth larvae observed by the author in parts of Hamden, Connecticut in 1982 (the year after a massive outbreak) and about half at Eldora, Cape May County, New Jersey in the early and mid 1990s (a suboutbreak period) were killed by this parasitoid. In general though its impact is limited by a complex of hyperparasitoids ((Wieber, et al., 2003). Parasites of eggs and early instars are generally not considered to have a major impact on gypsy moth populations, especially at high densities, but probably do help slow increase of low density populations.

Compsilura concinnata is an introduced tachinid fly whose larvae parasitize gypsy moth caterpillars and hundreds of other species including a few sawflies. Sometimes it reaches high levels in outbreaking gypsy moth populations, but it usually does not greatly impact them, probably because it is a multiple brooded generalist that quickly becomes limited by lack of alternate hosts which it may deplete (see Boettner et al. 2000). Its long-term impact on native summer feeding Lepidoptera has apparently been drastic reductions and state level extirpations of once widespread (Farquhar, 1934) summer species in New England. It does not appear to have greatly impacted any native spring feeders. As far as known, other gypsy moth biocontrols are not seriously impacting native species.