Comments: While such claims are made, often to promote spraying, it is not true that unchecked gypsy moth infestation will result in deforestation of large areas. Or at least that failed to happen in the first 130 years. It is however quite likely that some trees will be killed during the first outbreak in an area and quite possible others will be in subsequent outbreaks. Mortality might include high quality canopy trees as well as many already stressed or weak individuals. It is also not true that every tree that refoliates a few weeks after an outbreak will recover. Many to most trees that die do so a year or more later often due to secondary agents such as two lined chestnut borer (a beetle) or Armillaria fungi (Dunbar and Stevens, 1975). As gypsy moth invades new areas the best case scenario (increasingly likely now due to Entomophaga) is that no serious defoliation will occur, the worst case probably is roughly 50% mortality to oaks and other highly favored trees when two or more severe defoliations coincide with drought. Such severe damage is much less likely in subsequent outbreaks.

Because of database limitations of the IMPACTS and some other fields, most discussion relating to impacts to native Lepidoptera from gypsy moths and past control programs is placed in this section. Obviously in North America, threats to gypsy moth populations are not of concern, but threats to other organisms due to gypsy moth outbreaks or, more often, control efforts are a major issue. At present the highly persistent (on foliage), highly lethal biocide Diflubenzuron (tradename Dimilinis the main known cause of concern for biodiversity effects. This biocide is highly lethal to at least immatures of any Arthropods (old, broad sense) that eat it, and also as a contact insecticide. The most serious threats and impacts from a biodiversity perspective in the past generally were related to control efforts (including biocontrols), although at least in terms of widespread species, threats from current programs are not comparable to those from practices in the 1950s through 1980s. Current threats which could have consequences persisting more than a few months include tree mortality, impacts to native fauna from gypsy moth outbreaks, impacts to native Lepidoptera from increased generalist parasitoids, and especially eradication of local Lepidoptera or other invertebrate populations by chemical biocides or less likely by BTK. Defoliation itself has many temporary effects to both terrestrial and aquatic environments (see 1995 FEIS) but few or none of these would threaten long term loss of native biota, although short term temperature changes could affect woodland stream fauna for a generation (often a year) or more.

Unlike for example balsam or hemlock woolly adelgids, out of control white tailed deer, or potentially several emerging oak pathogens, gypsy outbreaks may change species composition but will not fundamentally alter the forest or eliminate strata. If tree mortality does occur some species will likely increase and some will likely decrease. On harsh sites, oaks that are killed by gypsy moth are usually replaced by release of understory oaks. On "better" sites there is a greater chance of replacement by other genera such as maples, or on sandy sites by released understory pines. Despite defoliation and tree mortality, mixed forests will remain mixed forests and oak forest or woodland will usually remain oak dominated, as have millions of acres in New England. The forests at Medford and Melrose Massachusetts first experienced gypsy moth outbreaks around 1880 and in 1988 were still primarily oak or oak-hickory dominated on hotter slopes with more mixed hardwoods including some very large oaks in more mesic places (Schweitzer, personal observations). Oaks were reproducing in part because there were almost no deer there in 1988. Witch hazel, pines and hemlocks were common. Some outcrops and places that burned multiple times per decade were dominated by scrub oak and heaths. As Musika (1993) points out tree mortality is not the only effect of gypsy moth outbreaks that can alter stand composition. Herb and shrub species commonly benefit from the increased light. Stand changes are possible if tree mortality occurs and it is possible but undocumented that such could occur even without direct overstory mortality. This topic is complex and impact likely to be very variable with location, climate, and site conditions. Pre-gypsy moth stand composition also will usually not be really "natural". Impacts of stand composition change may not be great for arboreal herbivores (e.g. Lepidoptera, Orthoptera) overall because so many species are polyphagous among forest tree genera, although of course more specialized herbivores (e.g. Catocala moths) will be affected by changes in abundance of foodplant species or (usually) genus.

There was minimal concern for non-target impacts of gypsy moth control efforts as recently as the mid 1980s (e.g. see the near lack of coverage in the FEIS, 1985), especially not regarding biological control efforts. Therefore there was no systematic documentation of impacts at the time. This has changed since about 1990 at least in part due to US Forest Service sponsored research projects. Appendix G of the 1995 FEIS has extensive coverage of non-target risks.

Localized populations of spring feeding Lepidoptera can be completely eradicated by gypsy moth spraying including BTK. There now seems virtually no doubt that the skipper Pyrgus wyandot is extirpated in most of its range and imperiled in the rest due largely or entirely to gypsy moth spraying from about 1957 to 1989. The species was extirpated in New Jersey by about 1960 (records compiled by David Iftner). All of the populations mentioned by Schweitzer (1989) from 1985 and 1986 field work in West Virginia were apparently eradicated by 1990 from gypsy moth spraying as was the classic occurrence in Green Ridge State Forest, Maryland. It is unknown if any have been recolonized. Several conversations with Lepidopterists in western Maryland and West Virginia in the late 1980s suggested eradication of P. wyandot and drastic reduction (at least short-term) of other spring butterflies in sprayed areas to be the norm. Dozens of other Lepidoptera declined or disappeared with that skipper in the late 1950s in northern New Jersey. Some like Papilio cresphontes and Chlosyne nycteis are probably no longer resident in the state although much habitat remains for the former. While search effort has been inadequate and the species probably will be rediscovered at least in Virginia and maybe widely, the grasshopper Appalachia hebardi is now ranked globally historic by NatureServe with no known collections since most of its known range in Pennsylvania was sprayed with DDT or Carbaryl in the late 1950s to 1970s. It is also very unlikely that the regional collapse of Erynnis persius persius in and near New England during and after the 1950s could have been unrelated to massive multi-million acre gypsy moth spraying peaking in 1958. Based on specimens in older collections, that species was clearly of comparable abundance to E. baptisiae before the 1950s and it is now absent at almost all "suitable" sites. The uniquely (for their size) depauperate specialized moth fauna of the Rhode Island pine barrens is probably at least partly explainable by gypsy moth spraying from the 1950s into the 1980s (Schweitzer, pers. obs.).

Some have suggested the decline of the regal fritillary including the near extinction of the eastern subspecies Speyeria idalia idalia was related to gypsy moth spraying. Probably the multi-million acre spray programs in the late 1950s did eradicate some populations in and near New England and the last record in mainland Massachusetts was in 1958. However in most regions populations persisted for at least another ten to 25 years, and eventually died out even in unsprayed areas. Other factors were obviously involved. Compsilura (below) could have played a role but it must be noted that of the five eastern Speyeria, this one is the least associated with forests where that fly is most abundant. Also this fly generally hunts for hosts in trees and shrubs not leaf litter or forest floor herbs. Modest declines of the other three Speyeria in the Northeast appear easily explainable by other factors.

An even greater impact of past gypsy moth control programs was the drastic reduction or large-scale eradication of dozens of formerly common large summer moths. See additional topics for more detail. Almost all of the species were common or at least widespread (Farquhar, 1934) and not localized rarities or habitat specialists. There is disagreement as to how much long-term impact resulted from the spraying of about 12,000,000 acres (Doane and McManus, 1981) of forest with DDT or Carbaryl in the late 1950s versus impacts from the introduced tachinid Compsilura concinnata (Hessel, 1976; Boettner et al., 2000). It is reasonably clear that most of these species collapsed in or about the late 1950s at least in western Massachusetts and from Connecticut southward to central New Jersey. Virtually all Lepidopterists active from northern New Jersey to Massachusetts at the time considered the crashes immediate, not gradual, and blamed aerial spraying at least in part (e.g., see Hessel, 1976; Gochfeld and Burger, 1997; Schweitzer, 1988; also personal communications of this writer with Asher Treat, Charles Remington, Roger Tory Peterson, Joseph Muller, Sidney Hessel and others).

Most likely though an even greater long-term impact was the introduced parasitoid Compsilura concinnata. This fly could pose a serious threat of eradication of localized summer species in parts of western North America particularly in isolated canyons. Boettner et al. (2000) experimentally document extremely high mortality to H. cecropia and others from Compsilura. Early anecdotal reports of declines in Saturniid abundance, but not eradications, date back to soon after the introduction of Compsilura (see Boettner et al., 2000). While Boettner et al. tend to discount the impact of spraying, Compsilura alone seems an unlikely sole cause for immediate (vs. gradual) crashes in the late 1950s since it had been present nearly half a century, but there is little doubt this parasitoid has drastically impeded or completely prevented recoveries. Many of the most widely eradicated species (e.g. Citheronia regalis, Eacles imperialis, Sphinx drupiferarum, most Datana, and undoubtedly others) frequently remain as pupae for two or three years (Schweitzer, unpublished) and so could not be eradicated even by one year of 100% larval mortality over a large area.

Whatever combination of factors were involved, there is no doubt these crashes occurred within a few years of 1958, although it is a common myth that only Saturniidae and Sphingidae were affected. Gregarious Notodontidae such as Datana, Schizura, Clostera were also very scarce or absent in the affected areas in the 1960s, 1970s and at least into the 1980s and in some areas are still. Overall probably the most widely eradicated genera were Datana, Citheronia, Eacles, Sphinx, Lophocampa. The obvious common characteristics of all of these affected species are moderately to extremely large size, exposed (often gregarious) tree and/or shrub feeding caterpillars maturing from late July to October--precisely when large numbers of Compsilura females need native caterpillars. It is less clear why other Notodontidae, Lapara and smerinthine Sphingidae were not also widely eliminated, although some of them apparently were reduced at least in the 1970s. It is noteworthy that Anisota are difficult, but possible, for Compsilura to parasitize (David Wagner, pers. comm., 2001). Perhaps their granular cuticle makes Smerinthinae similarly difficult targets. While other Ceratocampinae (Eacles imperialis, both Citheronia) have disappeared from 99-100% of their former ranges on the New England mainland, two of the three Anisota and the related Dryocampa have not been widely eradicated. A. virginiensis and Dryocampa remain at least moderately common in many places, although they were apparently absent in parts of southern Connecticut in the mid and late 1970s (Schweitzer and Yale samples). Assuming old identifications are correct A. stigma has been eradicated from much of its New England range, becoming confined to Cape Cod region barrens. There however, it appears to be increasing in distribution since 1990 (Michael Nelson, pers. comm., 2001). A. stigma is the only one of the moths discussed in this section for which habitat loss (pine barrens) is a plausible major factor in its decline. The other large moths discussed here routinely utilize a variety of common habitats including dominant forest types, hedgerows, thickets, although a few probably do best in shrublands. Anisota senatoria now has a somewhat reduced range in New England being largely eradicated from the western half, but is still regular eastward and around Albany, New York. A. senatoria is/was itself an outbreak-crash late summer defoliator of oaks. Boettner et al., (2000) found moderate Compsilura parasitism in Hemileuca larvae, although Jennifer Selfridge also working in Massachusetts (email to Dale Schweitzer, September 2002) did not despite simultaneous high levels in A. polyphemus. There is no evidence of major impact to either Hemileuca species though. Although H. maia has disappeared with its habitat in many places, it still persists in all of the substantial pitch pine-scrub oak barren areas in Massachusetts, eastern New York, Rhode Island and probably Maine, but not in New Hampshire. H. lucina, an outbreak species of wet shrublands and powerlines increased drastically in abundance in the late 1970s and early 1980s, and expanded its limited range into Franklin County, Massachusetts, northeastern Connecticut, adjacent Rhode Island, and Vermont (Schweitzer, pers. obs.).

While the precise impacts of massive aerial biocide applications and an out of control biocontrol cannot be deciphered with certainty now, between these two impacts the genera Citheronia, eacles, Datana, tree feeding (but not shrub or herb feeding) Sphinx species, Lophocampa caryae, Manduca jasminearum and others have been eradicated from substantial portions, or all, of their New England range and parts of adjacent New York and Pennsylvania. Attacine Saturniidae and Automeris are greatly reduced. Most, but not Citheronia, have partially recovered (starting mainly in the 1990s) in northern New Jersey and by the early 2000s Datana were scarce but no longer absent in parts of Connecticut. Species in these groups for which suitable habitats are present appeared unaffected in the 1970s and 1980s on Block Island, Nantucket, Martha's Vineyard and extreme outer Cape Cod, areas long considered as poor habitat for Compsilura but also with reduced aerial spraying compared to the mainland, or on Block Island none at all.

It is far from certain whether major Compsilura impacts will continue to spread. Obviously Compsilura will occupy suitable areas of North America where it was intentionally released and at least eastward it will spread. However, established Compsilura does not necessarily mean obliteration of large native summer feeding moths. This fly has long been present in southern New Jersey but at modest levels and with no obvious impacts at the population level to large moths. Saturniidae and Datana are still among the most common moths at lights for example.

Observations of possible Compsilura impacts and south and west of the Poconos region are equivocal. Datana, Eacles, Citheronia and other Saturniidae do not seem to be abnormally scarce now in central and southern Pennsylvania based on collecting by Stephen Johnson and to some extent this author through 2002. Indeed Callosamia promethea seems abnormally abundant in central Pennsylvania with caged females sometimes assembling over 100 males (Johnson) and Citheronia sepulcralis is at least widespread. Nor were such big moths at all scarce in the 2000 Great Smokey Mountains National Park BioBlitz. Except for a lack of Sphinx species Schweitzer noted reasonable to good numbers, especially of Eacles and Datana, in 1988 samples from Prince William Forest Park in Virginia. However Saturniidae, Sphingidae and some others were remarkably scarce in the 1999 USFS samples around Highlands, North Carolina (Adams, 2001). The extent to which Compsilura concinnata threatens North American Lepidoptera is unclear at present. Given its host breadth and the size of the US Lepidoptera fauna, thousands of species are potential casualties and it has apparently drastically reduced or eradicated formerly widespread or even common large summer moths from several states. This suggests it could have greater impacts in places like forested western canyons where equivalent summer moths are much more localized than their eastern counterparts. There is probably nothing preserve managers can do to mitigate Compsilura impacts. Probably of even greater concern are the impacts of Compsilura on native parasitoids of Lepidoptera, especially native Tachinidae.

Besides its lasting impact on native summer Lepidoptera, Compsilura probably should be credited with the near eradication of the introduced brown tailed moth, formerly a pest in and north of Massachusetts, from most of New England. Farquhar (1934) noted it was already declining by then and today its refugia are similar to those for severely impacted native species: extreme outer Cape Cod, various offshore islands and headlands, and far northern New England.

Impacts, if any, from an established biocontrol are generally unavoidable and long term whether such impacts be control of the target species or reduction or eradication of native non-targets. Not all impacts of introduced species on native species are significant at the population level. Hajek (1995) documented that some native Lepidoptera in several families can become infected by Entomophaga maimaiga under extreme laboratory conditions not meant to mimic field exposure. However, Hajek et al. (1996) found only two cases of infection (Malacosoma disstria (Lasicocmpidae), Catocala ilia (Noctuidae)) among 1790 native caterpillars in one large random study with high rates of infection among gypsy moth larvae. These authors also note that E. maimaiga is known only from gypsy moth in its native Japan. In other samples though they did find low to moderate infectivity in native Lymantriidae (Hajek et al., 1996; Hajek et al., 2000; Hajek et al., in press;) especially if the larvae spend time on the ground or in the leaf litter. The highest field incidence for any native species was 36% for Dasychira obliquata during a peak gypsy moth year. In most years no infected native Lymantriidae were recovered (Hajek et al., 1996; Hajek et al., in press). They also found single cases of infection in a Gelechiid (n=84) and in the noctuid Sunira bicolorago (n=20) among 358 caterpillars in samples from in and near forest leaf litter (Hajek et al., 2000). Low to moderate levels of mortality in extreme years should be easily absorbed by populations of common forest moths and probably have less impact than gypsy moth outbreaks. An ability to infect native species at levels too low to threaten their populations could be an extremely useful feature allowing the fungus to better maintain itself when gypsy moth larvae become scarce for long periods. Non-target impacts from Entomophaga maimaiga do not appear to pose any conservation concerns in terms of native Lepidoptera including Lymantriidae. This increasingly successful biocontrol will probably eventually occupy essentially the same range as the gypsy moth at least in humid eastern North America.

Since about 1990 gypsy moth control has taken a more focused Integrated Pest Management (IPM) approach. Potential biocontrols are getting some level of screening for non-target effects, but it is still possible another damaging species could be released and established. Under Cooperative Suppression Programs, spraying does not occur unless gypsy moth densities exceed certain thresholds. Completely unwarranted private spraying still occurs though and unscrupulous or ignorant operators deceive gullible neighborhood associations into annual spraying. Massive indiscriminant aerial spraying such as in the 1950s to early 1970s no longer occurs (FEIS, 1995), although hundreds of thousands of acres were sprayed some years, mostly with Diflubenzuron, in West Virginia and adjacent Maryland in the 1980s. Even by the mid 1970s or early 1980s such massive biocide use had ceased in Connecticut and became more spotty elsewhere in the Northeast where gypsy moth was well established. At current scales of biocide applications, eradication of common or widespread species at least among dispersive insect groups in areas with substantial forest would not be expected, and has not been reported. The potential for local eradications of native fauna is probably much higher for parts of the Midwest where forests are reduced to 10-100 acre scraps in a sea of agriculture. Mortality levels even from BTK could be sufficient to cause eradications in such small habitat islands and recolonization could be difficult even for some common species on such landscapes.

Long term to permanent impacts, if any, of current suppression programs in extensively forested regions should be limited to species occurring in localized colonies in special habitats, such as Pyrgus wyandot. For this reason care needs to be taken not to spray large portions of shale barrens, pitch pine-scrub oak barrens, and other habitats likely to harbor specialized spring feeding Lepidoptera even with BTK. Another species of concern in Appalachian gypsy moth eradication and suppression projects recently has been Phyciodes batesii maconensis. A major global consideration for Speyeria diana in the 1980s was the potential for massive impacts from large-scale gypsy moth spraying to this known BTK-sensitive species, such as occurred (largely using Diflubenzuron) just north of its range in West Virginia in the late 1980s when the species was placed on the USFWS Candidate (C2) list. For a variety of reasons such threats did not materialize since indiscriminant large scale spraying of US Forest Service and National Park Service lands did not occur and does not now seem likely under current IPM based policies.

A shift from chemical biocides to BTK should benefit summer feeding species and insensitive to moderately BTK-sensitive spring species by preventing defoliation and parasitoid buildup. It is distinctly possible that BTK based suppression efforts have been a factor in the partial recovery of Saturniidae, Datana, Lophocampa etc. in places like northern New Jersey during the late 1980s and 1990s. Actual data on this topic would be interesting.

Threats from gypsy moth outbreaks to native species are generally moderate or less and short term (FEIS, 1995. Appendix G). But on closer consideration severe short term to long term impacts are likely to a few Lepidoptera and other herbivores in some situations. Tree mortality and large-scale starvation of native herbivores in outbreak years are probably the direct effects most likely to have long-term impacts, if either occur. If tree mortality occurs there could be potential for increased invasion by understory exotic plants. Large-scale starvation of spring feeding caterpillars such as most Xylenini, Orthosia, the earlier Catocala species, Bistonini, is unlikely in most forest types (but see below) because most spring caterpillars mature before defoliation typically occurs. Most summer tree feeders (Notodontidae, Ceratocampinae, some Limacodidae) reared by Schweitzer have one or both of two life history traits that would prevent outright loss of a population to starvation form a single defoliation. For about 40% of species 5% to 60% of pupae overwinter more than once so there is always a reserve. For about half eclosion is staggered over 30 to 70 days starting near the time of peak defoliation so later larvae would have at least low quality refoliated leaves even in severe outbreaks. A few spring feeders such as Feralia major and Psaphida rolandi also sometimes overwinter two or more times as underground pupae, but most spring feeders overwinter as eggs or adults which can only do so once.

Synchronized early summer species beginning their larval stage at the time of maximum defoliation or during the next three to four weeks (longer if hatchlings cannot use young foliage) are obviously at high risk of starvation when defoliation is severe. Examples could include hatchlings of Hypomecis, Lytrosis, Euchlaena, Hyperaeschra georgica, and Hyperstrotia in June. The potential for starvation is also high for Nadata gibbosa, Actias luna and Antheraea polyphemus both to early larvae present with gypsy moth larvae and to more numerous later larvae hatching during the period trees are leafless.

There are also a few spring feeding species whose late instars develop slowly that are at risk during the defoliation period. Extreme examples are Morrisonia confusa (often on oaks, see also Wood and Butler, 1991), an oak feeding Hydriomena (probably H. pluviata), and Lambdina turbataria". All of these start feeding in April or May but remain as larvae until late August to October in Cumberland County, New Jersey (Schweitzer, pers. obs.). It is unclear how, or if, these species survive severe outbreaks. The four species of the Catocala amica group mature later than most other spring oak specialists, about when or just after defoliation typically occurs in New Jersey and Connecticut and so are at risk. A few hickory-specialized Catocala also linger as larvae (Sargent, 1976; Schweitzer, pers. obs.) and are at high to extreme risk of starvation in severe outbreaks. Such species include C. habilis, C. robinsoni, C. vidua and C. obscura. The BTK assay data in Peacock et al. (1998) imply at least two of these should fare better with a BTK application than in severe defoliation. Buckmoth larvae (Hemileuca maia) cannot complete development before defoliation but if they can reach late penultimate instar (which most can in southern New Jersey) mortality appears low. Somehow many survive and mature quickly when refoliation occurs. They may find some alternate food, perhaps Gaylusaccia leaves, or they may simply survive three weeks or so without food. By direct observation of captive larvae, 8 days of starvation has little, if any, lasting effect on last instar buckmoth larvae.

Scrub oak feeding Lepidoptera and other insects on xeric ridgetops and west slopes are at high to extreme risk of starvation, which could eradicate them if they do not occur in other microhabitats. For example Schweitzer observed on West Rock ridge, New Haven Connecticut that oak defoliation was virtually 100% on May 25, 1981 on the crest and west face but on the lower east slope did not reach that level until June 11. This difference should have a huge impact on survival of spring feeders, as most (sleeved and wild larvae of Xylenini and some Catocala) were observed to mature during that interval. As Gall (1984) observed survival on the west face would have been virtually impossible for many Catocala, and not all could have matured even by June 11. Indeed both of us observed many small Catocala adults there in July 1981. Specimens at Yale and elsewhere indicate two species of scrub oak habitats, Erynnis brizo and Chaetaglaea tremula, occurred on West Rock Ridge in the 1960s. Neither did in 1975-1982 (Schweitzer, pers. obs.) after outbreaks such as in 1971. Both species are larvae in spring. In an even more extreme incident at Hopkinton, Rhode Island on 16 May 1986 Dale Schweitzer observed nearly 100% defoliation of scrub oak buds in a small (<100 acre) sand plain pine barren by first and second instar gypsy moth larvae obviously blown in from adjacent phenologically more advanced oak forest. It is very unlikely any obligate scrub oak feeders survived that spring.

Native Lepidoptera are affected by other aspects of outbreaks besides starvation (useful discussion in Sample et al., 1996) but it is very unlikely these impacts would be long term and as those authors suggest such impacts may be within the range of normal fluctuations. As they document and one can easily observe, even in suboutbreak conditions gypsy moth larvae can far outnumber all native caterpillars combined and this might directly affect the latter. Native Lepidoptera may be impacted by induced increases in tannins or other defensive chemicals (Schultz and Baldwin, 1982) in replacement leaves and the next season. Earlier than normal foliation the next spring (Heichel and Turner, 1976) would affect synchrony of egg hatch and weight at maturity of spring feeders such as Xylenini and Alsophila (Schweitzer, 1979; Schneider, 1980). The main effect of foodplant changes will generally be small size and/or reduced fecundity (at least for Xylenini and Catocala, Schweitzer pers. obs.), which is well known for gypsy moth itself. Undersized, but functional adults are commonly observed after defoliation for summer feeding Notodontidae as well. Fecundity reductions are undoubtedly common in natives as well as gypsy moths and Sample et al. (1996) suggest they might occur even without severe defoliation. However there are no such data supporting or refuting this suggestion for natives.

Indeed general moth collecting, for Catocala and Xylenini collecting in particular, are very often excellent immediately after outbreaks and the 1981 New Haven butterfly count (Gall and Schweitzer, 1982) established a record number of 55 species for an eastern US count that held for over a decade, despite severe defoliation and limited effort by the observers. Sargent documents 1971 as a very good Catocala year in southwestern New England as it was in the Poconos of Pennsylvania (Schweitzer). This was also a year of high gypsy moth damage. Probably factors which favor gypsy moth increases (like warm, dry springs) also favor natives. Native species sometimes crash the year following gypsy moth outbreaks but in some cases (such as the 1982 and 1983 New Haven butterfly counts) crashes could be explained by extreme weather events such as torrential spring rains and floods or may have been due to some effect of the widespread defoliation Data on impacts of gypsy moth outbreaks to native species remain few at this time. Those of Sample et al. (1996) are given mainly at the family level--not an ecologically useful OTU for most macromoths other than Notodontidae and Limacodidae.

Undoubtedly starvation can seriously impact some native Lepidoptera at least when defoliation comes early. However, data are minimal. Lepidoptera that often seem to be reduced following gypsy moth outbreaks include Satyrium species, Nemoria bistriaria and some Acronicta species. The mechanism in the last case may not be nutritional. With 20-20 hindsight though it does seem clear that gypsy moth outbreaks do not generally cause extirpations of widespread forest species in the affected areas. To what extent they threaten localized ridgetop species remains undocumented, but past outbreaks probably do explain some current absences of these natives as noted above. It now seems very likely some of the differences between ridgetop and sand plain pine barrens moth faunas reported by Schweitzer and Rawinski (1987) reflect species losses related to past gypsy moth outbreaks. Since then as more sites are collected for moths, most of the so-called sand plain species have turned up on one or more large ridgetop barren.

In terms of non-target impacts chemical biocides have greatest effect on most native biota of the current management strategies. BTK probably reduces native Lepidoptera more than outbreaks themselves overall, but this could not possibly hold consistently at the species level given the great variation in both BTK sensitivity (including lack of any), variable risks for starvation, and other factors. BTK unlike chemical biocides other than perhaps Mimichas little or no impact beyond Lepidoptera. If benign options like pheromone flakes, Gypchek are available, any of these will probably greatly reduce or completely prevent negative impacts associated with gypsy moth with little or no non-target impacts. See the 1995 FEIS for more discussion on the many short-term impacts of gypsy moth outbreaks on other fauna, especially birds and other vertebrates.