Restoration Potential: Except perhaps in Great Britain where it is extirpated, restoration of gypsy moth is obviously not an issue. Restoration of damaged flora and fauna might be. Native fauna that are impacted but not eradicated by either control efforts or gypsy moth itself will usually recover within a few years without intervention unless limiting factors (e.g. Compsilura, loss of foodplant) persist. Native fauna eradicated by control programs may or may not eventually recolonize. For example Pyrgus wyandot has not done so in northern NJ after >40 years and even if this species does not become extinct (as now seems fairly likely) recovery in much of its former range would surely require reintroduction. Isolated populations may have no recovery potential if eradicated. For example, a truly isolated relict bog copper occurrence southward could not be expected to recolonize before the next post-glacial period. More generalized forest species should recolonize within one to three years in most cases. Restoration of Saturniidae, Sphingidae, Datana and others, and native parasitoids, in New England for now seems impossible and will remain so unless Compsilura permanently declines. The recent partial recoveries of many affected species in northern New Jersey (based on 1990s to 2002 records in the collections of Rutgers university, Alnne Barlow and Tony McBride) and very recent observations by Elizabeth Johnson of Eacles near Pottersville does offer some hope that the same might someday happen in New England. No effective control other than lack of host caterpillars is known for Compsilura although possibly some native hyperparasites might be helpful. Release of exotic hyperparasites is not recommended out of concern for already depleted native Tachinidae.
Planting or encouraging affected tree species once the initial invasion has passed could mitigate stand changes in second growth forests. This should be especially feasible where Entomophaga is providing good gypsy moth control and could be considered for restoring populations of state-rare species of oaks in New Jersey for example. However, biodiversity managers should consider that pre-gypsy moth forest composition in many parts of eastern North America was already quite unnatural having been affected by centuries of logging, altered fire regimes, and loss of American chestnut, a former dominant in many stands. The argument is often made that a reduction in oaks is beneficial in reducing the frequency of severe defoliation, and it is also by no means certain that such a reduction makes a forest any less "natural".
Management Requirements: The widely available options include no action, use of a chemical biocide such as Dimilinand use of one or more applications of Bacillus thuringiensis variety kurstaki. All of these have non-target impacts, except non target impacts of no action are minimal if severe defoliation does not occur as a result. Gypchek has no known non-target impacts but is not always available. There are other methods discussed below that do not have non-target impacts but they may not be options. Biodiversity oriented managers must first determine which options are available and decide if any such treatment option will be considered or allowed. This should be done in advance. Decisions and treatments should be made before defoliation is obvious. Defoliation usually is not noticeable until gypsy moth larvae are penultimate or last instar and by then BTK will not be effective. Even Dimilin will not be at all effective with last instar larvae because it cannot kill them until their next molt, which would be at pupation so feeding would be unaffected. Sevin would likely prevent further defoliation and Mimic at least reduce it in such circumstances, but like Dimilin a manager concerned with biodiversity conservation would generally regard the non-target impacts as unacceptable. If defoliation is already occurring a biodiversity oriented manager will usually have no choice but to allow it to run its course and plan monitoring and possible treatment options for next year. See MONITORING section. However in rare situations where serious defoliation is suddenly noticed (usually on west facing ridges) due to extreme number of third or earlier instars BTK will reduce feeding overnight and will kill most of them and might be effective in reducing defoliation depending on how many there were. In such cases risks of starvation to non-target Lepidoptera are probably very high which should be weighed against BTK impacts to them.
If it is determined that gypsy moth control is needed on lands managed for biodiversity, see the rest of this section and also THREATS and IMPACTS sections, for more information on the possible methods some of which have serious non-target impacts. Diflubenzuron (Dimilin or Carbaryl (Sevin are inappropriate except at very small scales (e.g. small campgrounds, high visitations sites, parking areas) if biodiversity conservation is among the management goals. Mimic has not been thoroughly reviewed for this document but would likely have broad impacts within Lepidoptera at least, for an uncertain period of time. BTK should be considered and is appropriate in environmentally sensitive areas that can be reasonably assumed to not have specialized spring feeding Lepidoptera or rare swallowtails. NPV (Gypchek pheromone based mating disruption, trapping of males, sterile male release have very minimal or no negative non-target impacts but may not be available or appropriate--for example if egg mass counts are already very high. USFS staff and sometimes other professionals may be able to offer good up to date information regarding these options.
Chemical biocides have been very widely used to combat gypsy moth and as recently as the 1985 FEIS very minimal consideration was given to non-target species. Millions of acres were sprayed with DDT and/or Carbaryl (see Doane and McManus, 1981) with a peak of about 8,000,000 acres in 1958 sprayed with DDT. However in the 1990s concern for non-target impacts increased greatly and the US Forest Service funded or conducted substantial research on the topic. Diflubenzuron (trade name Dimilin promoted as an "insect growth regulator" thus avoiding terms like pesticide, is the only chemical treatment currently widely used in Cooperative Suppression Programs and there has been a marked shift toward BTK in most states and by the US Forest Service. Mimic has recently had some use against gypsy moth and is also promoted as a "growth regulator". It is said to be specific to caterpillars and kills by inducing a prmature lethal molt. Carbaryl (Sevin is still used in some private operations. It is acutely toxic, especially by ingestion, to most Arthropods and very destructive in aquatic systems and severely impacts pollinators, including native species and honeybees. However it does not persist for months or longer on foliage, leaf litter, bark etc. like Diflubenzuron can, so Carbaryl probably does not impact canopy herbivores as severely. Unlike Diflubenzuron, Carbaryl kills adult as well as immature Arthropods quickly. Carbaryl is not discussed further in this document.
Literature on Diflubenzuron (Dimilin is too extensive to cover fully here (see 1995 FEIS). Diflubenzuron disrupts chitin formation in insect and other Arthropods (broad sense) that produce it by similar processes, such as a great diversity of Insecta and Crustacea. Death is at the next molt. Therefore it cannot kill adult insects. It is potentially lethal to immature Arthropods that ingests it, often at doses of a few parts per million and for some aquatics at doses of a few parts per billion. It also impacts fungi that produce chitin. It is also considered a contact insecticide but most research suggests this is not a major source of non-target mortality in applications aimed at gypsy moth and that ingestion is clearly the major source of mortality to most terrestrial organisms and to aquatic leaf shredders. It can kill eggs upon contact or affect fecundity of exposed adult females of some insects and even at least one Nematode. It is apparently not clear how widespread impacts to eggs are. See 1995 FEIS, Appendix G for an extensive list of affected organisms.
Some published studies, in some cases aimed mainly at assessing immediate impacts to vertebrate food supplies, show only modest impact to moths overall especially the same summer. Most moths in same summer samples come from unexposed overwintered pupae and/or come from habitats other than local forests. Larval sampling in areas treated with Diflubenzuron underestimates mortality (FEIS, 1995), probably greatly, because generally no attempt is made attempt to rear "survivors" to adults. Undoubtedly most would die at subsequent molts, especially considering that they would continue to ingest Diflubenzuron at subsequent feedings. Nevertheless, field studies show major reductions of Lepidoptera in treated areas, which remain significant at least through the second summer. In general mortality to non-target spring feeding Lepidoptera and probably other immature leaf chewing Arthropods should be similar to that of gypsy moth larvae. Mortality might be lower but is substantial to summer feeding caterpillars (FEIS, 1995), and probably few mobile larvae escape lethal exposure even in mid or late summer. Sedentary species occasionally might. Good data on actual mortality levels for summer leaf eaters were not found and the FEIS estimates probably are too low since caterpillars typically move around and will likely eventually eat a lethal dose.
See Appendix G of the 1995 FEIS for discussion of the fate of Diflubenzuron in the environment. Much Diflubenzuron washes off foliage in the first major rainfall, but a a substantial amount adheres to leaf surfaces for weeks to months. Typically 20 to 80% of the original amount applied remains two to three weeks after treatment, but thereafter its decline in the canopy is slow for the rest of the season and 5% to 50% will remain until leaf fall. There is some effective dilution via leaf expansion but not likely enough to reduce mortality to leaf chewers. At least usually potentially lethal doses remain on foliage for the rest of the growing season. Diflubenzuron probably can remain lethal on broadleaf evergreens for more than a year. However on pine needles the FEIS (citing Mutanen et al., 1988) states that by 61 days levels were undetectable on two samples and 10 and 25% of the original on two others. Other studies show traces of Diflubenzuron or its metabolites on foliage or leaf litter were still present at 319 days. Longer persistence has been presented at conferences but is not reported in the FEIS. Citing data from Mary Wimmer, the FEIS states that residues in leaf litter were over 1000 ppb soon after application, dropping due to microbial activity to 15 to 200 ppb just before leaf fall and then rising again with leaf fall. Residues remained stable over winter and declined to 100 to 400 ppb by the end of the second summer. Obviously leaf litter will remain highly toxic to litter feeding Lepidoptera and many other detritivores through at least two seasons (noted in FEIS, Appendix G) and probably to some extent into the third season or beyond. It is also well-known that residues adhered to falling leaves kill leaf shredders in streams for at least several months.
Diflubenzuron breaks down within a few days from microbial action in some soils and eutrophic pond mud but this can vary with microbial action and other factors. Usually half-life is a few days in eutrophic ponds. However significant amounts commonly do remain in the water column for several days to two weeks (FEIS, 1995, Appendix G, page 7-6). One study showed 98% degradation within four weeks and another showed 50% degradation within two days in soil in field situations. A half-life of 3.5 to 7 days may be typical in soil but persistence is longer in soils with low microbial activity or at low temperature. Without microbial action in sterile soils degradation is reported at 6% in four weeks in one study and negligible in one year in another. Diflubenzuron does not biomagnify up the food chain like DDT.
Impact to narrowly endemic cave fauna in Appalachia is a serious conservation concern. Diflubenzuron is not expected to reach groundwater because it rarely penetrates far into soil. Therefore there should not be an impact if a cave is fed solely by ground water. However if there is a surface stream involved lethal levels could easily occur since Diflubenzuron is somewhat soluble in water, is often suspended in the water column, and can readily enter caves adhered to leaf litter and other particulate matter. Short-term exposure to Diflubenzuron causes very high mortality to immatures of crustaceans (including true crabs and horseshoe crabs as well as small fresh water types), and diverse insect orders. The expected impact of a lethal dose (which could be as low as 10 ppb or slightly less) in a cave system would be death at next molt for most to all immatures of aquatic cave Arthropods (broad sense) and possibly sterilization of adult females. Extinction of endemic species would be a very real possibility. Exactly what species are at risk would depend in part whether exposure was from dissolved or suspended Diflubenzuron or via contaminated leaf remains and depends in part on the feeding habitats of the organisms. Potential impacts to endemic Carabidae are particularly hard to assess. High mortality to terrestrial cave Arthropods from Diflubenzuron seems unlikely unless they would actually ingest it but this is uncertain. They might be exposed via predation or scavenging.
In general Lepidoptera larvae, other chewing herbivores, and leaf shredders in streams are the most severely impacted non-target groups from Diflubenzuron. Insects that do not ingest Diflubenzurin, including sucking insects in the canopy, bees, wasps, ants, and many others are usually not reported to be greatly reduced in field studies, although species level data are few. Some studies indicate reductions of spiders. However sucking insects might in fact be impacted since Kim et al. (1992) document major fecundity reductions in the lab. Reductions in canopy and subcanopy Arthropods are substantial enough overall that Whitemore et al. (1993) were able to demonstrate significant reductions in fat accumulation in seven species of Neotropical migrant bird species. Bird species that depend heavily on caterpillars would be most likely to be affected. See the 1995 FEIS and references therein for details.
Studies such as Martinat et al. (1988) that use very broad operational taxonomic units (OTUs) and do not follow immature insects to the adult stage underestimate Diflubenzuron effects, probably severely so for Lepidoptera. See above and also comments in the 1995 FEIS. Such field studies may however provide good data relevant to birds and other predators for whom viability of prey is not directly important. Abagrotis alternata was noted as apparently unaffected by a Diflubezuron treatment by Butler et al. (1997), but that species would be present overwhelmingly as last instars at spray time (e.g. Peacock et al., 1998) and only late instars are easily sampled under burlap bands. There is no known mechanism by which last instars could be affected by Diflubenzuron prior to pupation when most or all treated larvae probably died unobserved underground. No caterpillars of gypsy moth, Lithophane hemina, Orthosia rubescens, "Polia" latex, the four most common species collected under burlap bands, were found post-spray in the treated plots that year (Butler et al., 1997, Table 3). These three native noctuids are collected by this method as late instars, solely as last instars for L. hemina and O. rubescens whose penultimate and younger larvae are green and stay among foliage. L. hemina would be mostly mid instars at spray time (e.g. see data on congenerics, especially L. petulca, in Peacock et al., 1998), O. rubescens would probably be mostly antepenultimates and penultimates (this species runs slightly later than the two congenerics in Peacock et al., 1998), and few or no eggs of "P". latex would have been laid yet (Wood and Butler, 1989; Schweitzer, pers. obs.). The data for these three plus gypsy moth indicate virtual eradication of spring Macrolepidoptera larvae by the last instar, including "P". latex which not yet even present as larvae at spray time. As would be expected, the difference was not as great for caterpillars collected off foliage, which would include various instars of questionable viability which was not assessed. It has not been demonstrated, and is not expected, that substantial numbers of immature leaf chewers, such as caterpillars, katydids or tree crickets, survive to the adult stage after eating Diflubenzuron. Also caterpillars alive on sample dates would likely ingest additional Diflubenzuron. A reasonable assumption is that mortality to native spring and early summer feeding caterpillars and probably other leaf chewers from this biocide will be comparable to that for gypsy moth.
Dale Schweitzer examined a number of moth samples from summer 1989 for Delaware Natural Heritage Program. The samples noted here were from more or less wooded habitat sprayed with Diflubenzuron in May. Nearby fields, thickets and tidal marsh were not sprayed. These were not paired controlled studies, but the interpretation is obvious to anyone who has ever run a blacklight in an eastern forest on a summer night and has a basic understainding of moth phenology. Marsh, thicket, old field, and lawn species were well represented, but there were almost no forest species except for some that came from previous years' pupae. There were no litter feeding Herminiinae (normally an abundant and diverse group) in most of these samples. The total number of tree feeders that could have been larvae in May to July that year was five moths of four species in a blacklight sample on 6 August at Pike Creek "Natural Area". Two of these were Hypsoropha hormos a persimmon feeder and so probably from an insprayed old field or thicket. There were six individuals of three forest species in a 2 September trap sample out of about 33 moths, mainly weedy Pyralidae. Four of the six were Semiothisa granitata, which suggests residues on pines were no longer lethal by about late July, which is consistent with the 1995 FEIS, Appendix G (see above). The other two could easily have come off wild cherries in old fields or thickets but were larvae since June. A July 12 sample contained a mere 47 macromoths at Ted Harvey Wildlife Refuge. Two oak feeding Catocala and one Abagrotis alternata were probably the only individuals that would have been larvae in forest trees that year. One Anavitrinella could have come from the forest but probably came from another habitat. Old field and thicket species such as the Prunus feeding Catocala ultronia (3) and dogbane feeders (3 of two species) seemed normal and there were three of the persimmon feeding H. hormos. Forest tree feeders from previous years' pupae were better represented with 12 individuals of at least four species of Notodontidae in that sample which probably came mostly from oaks or maples. Overall one would conclude a modest reduction (perhaps 70-80%) of moths comparable to some published studies, but these samples strongly suggest forest species were virtually eradicated as larvae at least into July. Identical traps in comparable, except more light-polluted, New Jersey forests a few dozen kilometers to the east generally take 100-600 forest moths per night at those seasons (Schweitzer 1999-2001 data), suggesting about a 95-100% reduction of forest species.
Because of its severe impact to several major non-target groups in forest canopy and aquatic systems, its persistence on foliage, leaf litter and other surfaces, and the availability of more benign alternatives, Diflubenzuron is inappropriate for large-scale use on lands where biodiversity preservation is among the management goals. Natural Heritage Programs should re-rank tracked Lepidoptera, Orthoptera and certain aquatic insect occurrences to historic if most habitat is treated with Diflubenzuron unless the species is known to have multiple overwintering as pupae. Diflubenzuron, Carbaryl, or DDT applications, even decades previously, should also be considered a negative factor in evaluating "natural community" occurrences if community is assumed to include more than flora, since eradication of specialized Arthropods may have occurred. Data on persistence and non-target impacts from the apparently caterpillar-specific Mimichave not been evaluated for this document. Mimicis not now used in Cooperative Suppression Programs. While BTK or chemical spraying will generally prevent defoliation they do not provide long-term control.
Dichlorvos, the organic phosphate neurotoxin in no-pest strips (see FEIS, 1995, Appendix G) is is placed in small pheromone baited traps. Disparlureis a synthetic gypsy moth pheromone often used this way. Grids of such traps are used for detection or quatitative survey work and are not now used for control purposes . Males are attracted and killed inside the trap. Dichlorovos will likely kill other invertebrates that wander inside the traps or outside them if the traps are destroyed and knocked to the ground. It is very unlikely the dosage would be harmful to any organism any meaningful distance from the traps in normal use. Significant non-target effects are not expected and not reported in the FEIS or elsewhere. The only likely objection to this methodology by preserve managers would be on philosophical grounds. An alternative may be traps with a sticky coating (such as Tangelfootinside but these might kill a few small mammals should they fall to the ground.
Gypchek the commercial product of the NPV virus is available through the US Forest Service (contact Richard Reardon, Morgantown WVA office) but only for government (federal, state, county etc.) sponsored suppression or eradication programs. It can be used on private lands as part of suppression efforts. Gypchek does affect some native Lymantriidae at least in the laboratory (R. Reardon, pers. comm., March 2003) but otherwise there are no suspected negative impacts to non-target species. See Reardon and Podgwaite (1992) and Reardon et al. (1996) for details.
Synthetic gypsy moth pheromones can disrupt mating and can give good control or even local eradication when used against low-density populations. While traps are used for monitoring, when used for control pheromone is broadcast on tiny flakes over substantial areas. This technology has no known harmful impacts. For now pheromone based suppression will not be technically feasible for most managers unless as part of a larger USFS sponsored program in the area. Since the gypsy moth has no native close relatives, disruption of other species seems unlikely in North America and notably there have been no reports of males of other species being lured to traps baited with gypsy moth pheromone. Sterile male release (see Reardon and Castro, 1993) is also without negative impacts but is usually not an option.
Various Microsporidia have been investigated as biological controls of gypsy moth larvae. Some are reported to be significant in gypsy moth populations in Europe (McManus et al., 1989) and some have the advantage that they are passed on to offspring via eggs. According to Richard Reardon (pers. comm., 2003) they are more likely to cause chronic sub-lethal infection than high mortality. At least two have been "experimentally" introduced in Maryland and one is probably established. The literature reviewed contained little on non-target impacts, if there would be any. Impacts to native caterpillars have been investigated in the 1990s, and some species can host these pathogens at least in the lab. Results have not been reviewed for this document, and it is unknown if these biocontrols would have significant negative impacts on native species. Nothing on these organisms appears in Appendix G of the 1995 FEIS.
The fungus Entomophaga maimaiga has sometimes been directly introduced into gypsy moth outbreak areas but from at least West Virginia and New Jersey northward it is now well established naturally and it is spreading. At this time its application as an insecticide is not legal under EPA regulations. This fungus was introduced to New England from Japan in 1909 and considered a failure. Whether from this introduction or some other undocumented event, in 1989 Entomophaga collapsed a huge incipient outbreak in Connecticut and most of that state has not had severe defoliation since 1981 or 1982. This event got a lot of press coverage including articles in the New York Times, Science News and Discover magazine. The USFS Gypsy Moth News volume for April 1993 was appropriately titled "Maimaiga Mania" and is still a useful practical reference, for example if one wants to determine if virus or fungus is killing gypsy moth larvae on a site (see also Reardon et al., 1996). Spectacular collapses of gypsy moth populations continued south through New Jersey and into at least West Virginia. This fungus is now clearly established and likely to be a permanent major factor in regional forest ecology, although it is unlikely it will completely prevent all outbreaks. As discussed previously nontarget impacts range trivial to minor (some native Lymantriidae only) and this fungus this is safe for application on virtually any preserve. Caution would be suggested only if a rare native Lymantriid were present--perhaps possible on the outer coastal plain with Orgyia detrita but based on congeners impacts should be minor. Reported impacts to Dasychira species suggest little cause for concern assuming impacts to the rarer species would be comparable to commmon ones.
For biodiversity-oriented managers if Gypchek and pheromone flakes are not available, very often the only options are no action or aerial application of Bacillus thuringiensis variety kurstaki (=BTK) formulations. BTK occurs naturally in some soils, but not on foliage. Like chemical biocides, BTK is used for short-term control, but unlike Diflubenzuron the lethal crystal exotoxins are non-persistent and are nearly specific to some caterpillars. However the spores are much more persistent. Generally lethal effects from BTK persist a week or less after application (e.g. 1985 and 1995 FEIS; Sample et al., 1996; many others), but. Johnson et al. (1995) shows lethal effects persist for at least a month with swallowtails. High susceptibility to spores alone probably explains such lethal residual effects (Richard Reardon, pers. comm., August 2002). Most studies do not find persistent effects implying that most native Lepidoptera are not highly susceptible to the spores alone. Miller (1990a,b) demonstrates that even from two or three applications impacts disappear before August. Wagner et al. (1996) found differences between plots once treated once in May and untreated plots disappeared in June. Wagner's results strongly imply impacts for only about a week or less since the sampled caterpillars required several weeks of growth. The data in Butler et al. (1995) suggest BTK applied in May had little or no impact on a wide array of summer caterpillars. Schweitzer noted large numbers of large nests of Hyphantria cunea and defoliation by the catalpa sphinx (Ceratomia catalpae) less than two months after BTK spraying in Port Norris, New Jersey in the 1990s, implying little impact to hatchlings of either species by late May or early June. Still while very many or most Lepidoptera are not greatly impacted by spores months after applications, the possiblity of persistent lethal effects such as reported for swallowtails exists for other species. More information on impacts to caterpillars such as Saturniidae from the persistent spores is needed.
BTK will generally prevent defoliation or high numbers of gypsy moth caterpillars when properly applied. Like chemical biocides it will not aid in long term control. A single application should result in about 60 to 90% reduction of gypsy moth larvae (R. Reardon, pers. comm., 2003) which is usually sufficient to provide adequate foliage protection. A second application is often used in eradication projects or against unusually dense larval populations. Mortality from two or three applications is generally at least 80 to 90% and this strategy is commonly used in USFS eradication projects, often for two seasons. Biodiversity managers should carefully consider BTK and balance its impacts against those of defoliation. Since gypsy moth outbreaks and BTK have different negative impacts, from a biodiversity perspective a case can often be made for partial treatment of outbreak areas especially in large forested tracts.
BTK has two modes of action. Toxic protein crystals formed by the bacteria damage the gut and enter the body, and/or the caterpillar may succumb to a massive sepsis upon ingesting the spores. For some species both the spores and the crystal must be ingested for high mortality to occur (R. Reardon, pers. com., 2003). Sublethal effects also commonly occur (Peacock et al., 1998 and numerous other studies). All strains of BT must be ingested to kill. In terrestrial systems, non-target impacts from BTK will be limited or very nearly so to Lepidoptera larvae. Appendix G of the 1995 FEIS reviews investigations of BTK impacts to other organisms. It can kill some aquatic insects, primarily midges and blackflies, which is not surprising since another BT strain is often used to kill these. Less expected was mortality to some species of stoneflies (FEIS, Table 5-1). Biodiversity managers should consider buffers to minimize BTK in ponds and streams. Some other strains of BT kill leaf chewing larval or adult beetles, but apparently this impact has not been reported with BTK. It apparently does not affect Orthoptera and definitely does not kill at least most sawfly larvae. Concern for adult butterflies such as monarchs appear groundless. Data (see FEIS) are clearly adequate to conclude that most terrestrial, aquatic, estuarine and marine Arthropoda (broad sense) are unaffected by BTK, still with so many species and so much variation within Lepidoptera (even within families and genera) and variable effects among stoneflies, it is impossible to be sure there are no potential impacts. There is some chance a few additional Arthropods will prove sensitive to BTK. Note though mortality in sensitive species is not always high enough to be of conservation concern, even among Lepidoptera (Peacock et al., 1998; Wagner et al., 1996). For example (FEIS, 1995, Table 5-1) mortality to one of the stoneflies was only 30%. There are (same reference) some other studies showing low field mortality or mortality at high concentrations in artificial diets for other non-Lepidopterous insects. Some field reductions may be from indirect effects. While some of the references cited in the FEIS were not reviewed in preparation of the present document, it appears that significant direct impacts from BTK to terrestrial non-targets other than caterpillars are at worst quite rare.
Within the Lepidoptera impacts of various BT strains including BTK are widespread (Krieg and Langenbruch, 1981; Peacock et al., 1998). Field studies including Miller (1990a, 1990b), Sample et al. (1996), Wagner et al. (1996) and others show reductions (often about 30-70%) in Lepidoptera abundance, and sometimes species richness, following BTK applications. However most field studies do not examine species level effects so findings are invariably cumulative impacts to species that were unaffected, moderately affected, severely affected, or even eradicated. While the study was not optimally designed, the results of the sampling associated with the Highlands North Carolina Eradication Project (Schweitzer, 2000; Adams, 2001) indicate remarkably little impact to a very diverse macromoth fauna containing more than 758 species. See the discussion in Wagner et al. (1996) for a realistic assessment of the technical difficulty of sampling caterpillars, which would be the ideal way to assess BTK impacts. That study contains some useful species level data and Butler et al. (1995) also present useful species level data as do the Schweitzer and Adams Highlands reports. The Peacock et al. (1998) lab assays are the most useful source of species level information for the species included. In contrast to Diflubezuron studies some BTK field studies did include rearing out larvae to determine viability (or verify identity) and the data and discussion in Peacock et al. (1998) and Wagner et al. (1996) address this issue. Many to most, but definitely not all, larvae alive a week after BTK applications are viable (see below).
Within Lepidoptera, species level impacts of BTK vary enormously. Peacock et al. (1998) assayed native non-target species in instars actually present at the time of gypsy moth suppression applications. Effects under ideal laboratory conditions impacts ranged from none through slight developmental delay to 100% mortality in under five days. Mortality was significant for 27 of the 42 species (64%) assayed with Foray 48B, but 14 of these 27 produced viable adults and for some more than a third survived. Thirteen of the 42 species (31%) were considered highly sensitive. Such species would be at some risk of eradication from sprayed areas. Foray 48B and Dipel results were similar. Sensitivity was highly variable within Geometridae, Noctuidae, and Lymantriidae (see also Wagner et al., 1996) and varied greatly within genera Catocala and Lithophane in our data, and also within Orthosia and Dasychira if compared with Wagner et al. (1996). The fact that even for sensitive species a few to many individuals recover from BTK effects is an important difference from Diflubenzuron, and furthermore they will usually have little or no further exposure at least to the crystals. Note however Peacock et al. (1998) report delayed mortality in a few species including some that did not have significant mortality at 5-7 days. One butterfly had significant (100%) mortality only at or after pupation. So reliable claims of insensitivity to BTK really should ideally be based on survival to the adult stage and such claims should be rejected if not based at least on survival to pupation (see Peacock et al. 1998). No Microlepidoptera were included in these lab assays. Most micros gain some protection by feeding within shelters. Wagner et al. (1996) report 16 of the 17 most abundant micros were more frequent in treated plots, but species level results were not significant. Sample et al. (1996) report modest reductions of micros in their adult but not larval samples. However Miller (1990b) reports successful control of spruce budworm (a Tortricid) with BTK. Krieg and Langenbruch (1981) report numerous micros as affected by BT varieties.
BTK sensitivity is a species and/or instar level trait and based on the Peacock et al. (1998) data there appears to be little basis for predicting sensitivity of third and later instars of macromoths even on the basis of data for congenerics. Still some generalizations are justified regarding BTK impacts to caterpillars. In the absence of data to the contrary, first and second instar larvae can be presumed sensitive. All of 18 species assayed by Peacock et al. (1998) as first to third instars had significant mortality, which was 95-100% for 8 (42%) of them. For some species sensitivity was clearly reduced in the third instar. For 25 species evaluated as fourth to last instar, 10 had significant mortality, which exceeded 95% in five species (20%). While some species are highly sensitive in all instars, consecutive instars of other species differed substantially. By the third instar the only generalization apparently justified by the lab assay data is that Xylenini other than some Lithophane species are insensitive or only moderately sensitive.
The claim by Wagner and Miller (1995) that Saturniidae are highly sensitive to BTK is neither justified nor refutable based on the data in Peacock et al. (1998) because of high control mortality for two of the three species, in contrast to other assays in which the median control mortality at day seven was zero. Kaya (1974) found that 13% of fourth and fifth instar but virtually no early instars of another Saturniid, Anisota senatoria, survived BT var. alesti applications. Except for Hemileuca maia (which is highly sensitive to BTK), no other eastern US forest Saturniidae would have high exposure within a week (or often even a month) of typical single gypsy moth suppression applications. Assuming they are not at risk from lingering residual spores a month or more later, most Saturniidae and other summer feeding moths such as most Acronicta, Notodontidae, Limacodidae, actually should benefit from gypsy moth suppression applications of BTK. Treated areas will probably not be defoliated and thus their larvae will have normal food resources and lower generalist parasitoid levels in summer. Furthermore reduction of gypsy moth larval numbers might reduce impacts later in the season to native summer species from Compsilura.
Herminiine Noctuidae whose larvae feed in leaf litter (Hohn and Wagner, 2000) have been suggested as a potentially severely impacted group--especially if BTK spores were to persist and germinate in the litter possibly contaminating their food supply for several years. Unpublished laboratory data (D.Wagner and L. Butler, pers. comm.to Schweitzer in 2002) show several Herminiinae to be BTK sensitive and it is even possible species in some genera are consistently so contrary to the findings of Peacock et al. (1998) for other genera. Field samples of adults in the Highlands North Carolina Eradication project reports (Schweitzer, 2000; Adams 2001; and some additional interpretation herein) are highly consistent with sensitivity for many of the genera and species despite some sampling design issues. Idia species appeared unaffected but Renia species were rather uniformly represented at all sites early in the first season but for adults from larvae present at treatment time, frequencies per sample averaged 5.0 for the untreated samples and 0.5 at both treated sites. Polypogon (aka Zanclognatha) also appeared affected. Numbers of Renia and Polypogon were lowest the year of treatment suggesting there probably is not a large lingering impact from BTK spores in the litter or active bacteria in the soil. Actually the data in the final report (Adams. 2001) seem very definitive on that important issue. Herminiinae numbers clearly did not continue to decline further in 2000 and 2001 from the 1999 spraying, which is inconsistent with a severe lingering impact due to BTK persistence or reproduction seriously affecting the larval food supply of Herminiinae for several years. More data are desirable for Herminiinae but current results suggest variable sensitivity the first season with recovery starting the next season, in other words impacts comparable to other Noctuidae. There does not appear to be a basis for concern over severe long-term impacts to Herminiinae other than rare or highly localized species. However it should be stressed Herminiinae were almost unique among macromoths in showing any apparent impacts in these Highlands samples. Since potentially lethal Dimilinresidues do remain on leaf litter at least into the second summer and possibly longer, Herminiinae are likely to be much less impacted by BTK than by Dimilin.
Many species of butterflies, possibly all species, are highly sensitive to BTK. The published literature documents very high (often 100%) mortality to species in the Pieridae, Papilionidae, Lycaenidae and Nymphalidae (Peacock et al., 1998; Wagner et al., 1996; Krieg and Langenbruch, 1981; Herms et al., 1997; Johnson et al., 1995; Wagner and Miller, 1995). Drift from a single application of BTK apparently eradicated the population of Speyeria idalia at Gettysburg National Battlefield in the early 1980s. According to the USFWS Candidate Evaluation forms several populations of the imperiled Euphydryas editha taylori were eradicated by BTK applications aimed at Asian Gypsy Moth. No treated larvae of Speyeria diana, Papilio glaucus, Asterocampa clyton, Limenitis arthemis astyanax produced adults in the Peacock et al. (1998) assays. Eradication of local populations may have occurred with certain Lycaenidae in a Utah gypsy moth eradication effort (data not reviewed). Unpublished reports of high butterfly larval mortality exist for example to Lycaenidae even for substantial distances downwind in that project. No reports of insensitivity or moderate sensitivity to BTK were found for any butterfly, but Krieg and Langenbruch's tabulation suggests only moderate sensitivity for "Vanessa" io exposed to another BT strain. This could be misleading though since most or all BT mortality is delayed until pupation for two other Nymphalidae: Vanessa cardui (Morris, 1969) and Limenitis arthemis (Peacock et al., 1998). Wagner and Miller (1995) referring to the Peacock et al. data reported that the "spring azure" (actual species not stated because only NABA sanctioned common names were allowed) is highly sensitive. The species was Celastrina lucia. Significant differences in larval survival between treatment replicates precluded formal analysis or inclusion in the published results (Schweitzer, pers. obs.). Also few control pupae and no treateds eclosed probably due to inappropriate pupal storage. The sensitivity of this species is unclear other than that some treated larvae did pupate. Concern is clearly justified concerning eradication of localized populations of butterflies with spring feeding larvae such as Asterocampa, regionally localized Polygonia and relatives, azures, elfins, many hairstreaks, and the spring forest Pieridae.
Taken together field studies show that BTK significantly reduces but does not come close to eliminating native Lepidoptera as a whole (e.g. Wagner et al., 1996; Sample et al., 1996; Butler et al., 1995; Miller, 1990a,b). In general effects do not persist on summer foliage, so summer feeding species undoubtedly often benefit from BTK applications if these prevent severe defoliation or even high numbers of gypsy moth caterpillars. A few widely cited studies document that reductions of caterpillars cause emigration or decreased reproduction by vertebrates, including birds (Rodenhouse and Holmes, 1992) and shrews (Bellocq et al., 1992) in BTK treated areas. Some studies such as Holmes (1998) do not show such effects. See the 1995 FEIS for more studies. Such effects are also well known from chemical biocides applied against gypsy moth. Generally though BTK impacts to vertebrate food supplies are much less than those from Diflubenzuron and it is possible that in some cases they are less than those from gypsy moth defoliation, although data are not adequate regarding the last point.
The field studies cited cannot adequately address the issue of impacts to highly localized species (e.g. almost all species tracked by Natural Heritage Programs or ranked as globally rare by NatureServe), although there is discussion in Peacock et al. (1998) and Wagner et al. (1996). A few of these rare species (with varying sensitivities) are covered by the lab assays. Risks to such rare species must be assessed species by species at least for macro-moths. In the absence of actual data, the most prudent suggestions seems to be that managers should assume BTK is lethal to caterpillars of rare species, other than most Xylenini, feeding within a week of application, highly lethal if they are first or second instar. A reasonable working assumption when faced with lack of data for a given species that would be present as larvae, would be that impact to it will be comparable to those to the targeted gypsy moth. First to third instar gypsy moth larvae do appear to be more sensitive to BTK than most native species assayed by Peacock et al. (1998), and some in various field studies, but some caterpillars are more sensitive than gyspy moth larvae.
The data in Butler et al. (1995) suggest various families of sucking insects (Homoptera) benefit from BTK as compared to moderate defoliation. They probably benefit drastically when defoliation is near 100%. While there are no data it would seem very likely total defoliation obliterates these insects. Schweitzer strongly suspects drastic declines of aphids, Membracyids etc. might explain why sometimes (e.g. very late June and July 1981 around New Haven, Connecticut) sugar baiting for moths can be spectacularly good just after near total defoliation. Honeydew from these sucking insects is an adult noctuid food source, possibly an important one. One would expect some other insect herbivores such as katydids, tree crickets, Cynipidae and sawflies to benefit from BTK applications when the alternative would be severe defoliation.
In eradication projects the US Forest Service generally uses two or three BTK applications to a large portion of the treatment area for one or two seasons. It is well known that later treatments kills most gypsy moth larvae that survive the first one. Impacts to non-targets from these projects undoubtedly vary. Miller (1990) is an often-cited study from Oregon that showed substantial impacts, but without species level data little can really be concluded. Schweitzer (2000) and Adams (2001) concluded that the 1999 Highlands, North Carolina eradication project had very little impact on most native moths even in thrice sprayed areas. Impacts based on adult sampling were clearly minor or less for most species, including some that were sensitive in the lab assays. One has to wonder if larval mortality to BTK might have been offset by relaxation of some other density dependent mortality agent in this instance. Exceptions were major reductions of certain Herminiinae and probably Feralia jocosa. In particular there was no hint of any impacts to summer feeding species which is consistent with published studies. The Highlands studies did not look at butterflies.
Dale Schweitzer and Jane Carter at the USFS facility used by them previously (Peacock et al., 1998) and illustrated by Wagner and Miller (1995) assayed six species using two applications of 36 BIU equivalent ten days apart. These results are unpublished. For the geometrid Prochoerodes transversata the second treatment did kill all survivors of the first. For five noctuids the second treatment produced zero or insignificant additional mortality, including Eupsilia vinulenta, Chaetaglaea sericea, and Sericaglaea signata which are essentially insensitive to BTK, but also Egira alternans and Lithophane grotei which had significant first treatment mortality and were sensitive in the published assays.
Management Programs: The USDA-Forest Service and USDA-APHIS coordinate gypsy moth eradication projects and leadership role is determined largely by the number of acres involed. They are activated for all Asian gypsy moth detections in the USA and for detections of European gypsy moth substantially (usually 100 miles) ahead of the advancing front of general infestation. These projects often involve hundreds of thousands of acres. Gypchek may be used in sensitive areas, and pheromone flakes are sometimes used in very low density areas, but the primary treatment is usually two or three applications of BTK, sometimes for more than one season. Dimilinhas also been used. Most such eradication projects were with intitially low numbers of gypsy moths and have been deemed successful. The USFS also coordinates a "slow the spread" program, mostly near the edges of generally infested areas. Tactics include pheromone flakes where gypsy moth densities are low and BTK where numbers are higher. Its goal is self-explanatory. The USFS coordinates a "suppression" program the goal of which is to reduce potentially defoliating populations of gypsy moth in areas where it is already permanently established. BTK and Dimilinare the usual tools.
Most gypsy moth spraying is part of suppression projects, which usually involve federal, state and county agencies. The goal is generally to reduce defoliation and nuisance. Concern for non-target impacts varies greatly depending on actual local decision makers and local agencies. Usually an integrated approach is taken, and treatment is considered only if monitoring such as egg mass counts shows a high potential for defoliation. Suppression programs are responsible for most use of chemical biocides and probably pose more threat to non-target biota than typical USFS eradication efforts. Increasingly though suppression efforts in some states now rely on BTK in some states but over all there has been some increase in use of Dimilinuse in State Cooperative Suppression Programs, presumably because it is less expense and is more effective at killing gypsy moth larvae. The difference in effectiveness is sometimes overstated by applicators and it should be noted that despite any difference, BTK usually does adequately suppress defoliation. Non-target impact sometimes weighed heavily in decision making at the local level. Dimilincannot be used near Delaware and Chesapeake Bays and near estuaries due to impacts to juvenile crabs and other Crustacea. Some states never, or almost never, use chemical biocides. While state regulations vary, generally landowners are notified of intent to spray and have the option to refuse treatment. Biodiversity managers should virtually always refuse Diflubenzuron (Dimilin or other chemical biocides (for now at least including Mimic and insist on at least the minimum buffer currently recommended for preserves, other sensitive areas, vernal ponds, or streams. An attorney should be consulted immediately if there is resistance to excluding a preserve. As discussed elsewhere, whether or not to use BTK is a much more complicated decision and it might be a mistake to simply refuse to allow it without careful site-specific evaluation of risks. Also consider whether Gypcheck would be an option.
Monitoring Programs: USFS, state and sometimes local agencies generally have monitoring programs in infested areas, along the leading edge of advance and often beyond it at least where Cooperative Suppression or Slow the Spread Programs are in place. Monitoring within the established range is usually by egg mass counts. In extreme low or high densities estimation is easy but often assessing defoliation potential is not as easy as one might expect. Size of current year egg masses and presence of old ones are confounding factors. Defoliation can even occur where there were few egg masses if large numbers of hatchlings are blown in from elsewhere. See Fleischer et al. (1992) for discussion of some specific protocols for estimating egg masses. If monitoring is needed for example on a Nature Conservancy preserve, contact local professionals for advice or direct help. Often a gypsy moth control agency can be found in local phone book government listings. Biodiversity managers should always cooperate with local monitoring agencies except if a high count would lead to pressure to apply Diflubenzuron or other unacceptable biocides in sensitive areas. In most cases no spray requests are honored.
Management Research Programs: Research aimed at developing BT strains more lethal to gypsy moth and less lethal to other Lepidoptera such as butterflies would obviously be useful to biodiversity managers. Such research has been considered and may be on-going, but the topic was not reviewed in preparing this document. Obviously research on new biocide and other control measures will continue at least by chemical companies. USFS is continuing to sponsor Linda Butler's research on the impacts of gypsy moth and efforts to control it on non-target species.
Management Research Needs: More research on the impacts, or lack thereof, of gypsy moth outbreaks on native spring and summer feeding Lepidoptera and comparison to BTK treatment effects seems like a high priority. However field studies present some extreme difficulties (See Wagner et al., 1996; Sample et al., 1996). Observations of sleeved larvae of native spring species to quantify maturation dates relative to defoliation would give a better idea of which species are vulnerable to the most severe impact--outright starvation. Native summer species could be sleeved at low density to assay relative suitability of regrowth foliage after defoliation versus normal foliage. One could sample adult Xylenini and some other moths in defoliated and BTK treated areas and compare fecundity directly or indirectly using weights. However treatment areas would have to be at least thousands of acres since these moths are dispersive.
Research seems warranted to determine if there is any possibility of effective control of the parasitoid Compsilura concinnata. This could potentially benefit dozens of summer species which have been greatly reduced in New England and elsewhere and lead to re-establishment of now absent large native moths. Such research might include investigation of indirect impacts of BTK applications on this fly. Native parasitoids might also benefit.
A more general topic for which data seem inconsistent is persistence of lethal BTK residue on foliage. For most species lethal period is clearly only a few days but for a few such as Papilio glaucus it is reportedly much longer. Species that are unusually sensitive to the spores alone could probably be killed to some extent for the rest of the summer. Direct assays using larvae of known sensitive species sleeved in field treatment areas perhaps at ten-day intervals would answer questions about lethal residual impacts to groups like Saturniidae about which concerns have been raised. The USFS also maintains an excellent facility for bioassays with caterpillars at Ansonia, Connecticut. It is critical that any such experiments be designed such that control mortality approaches zero (the median in Peacock et al., 1998). High control mortality (except if with hatchlings which fail to establish in the rearing sleeve) generally raises the possibility of other major stresses to larvae besides BTK. The best way to achieve such low mortality is to sleeve larvae on growing plants as soon as practical.
Some miscellaneous topics follow:
- Better understanding of mortality as a function of dosage would be useful in protecting very sensitive non-targets such as some butterflies from BTK drift.
- A definitive answer to questions regarding sensitivity of native Lymantriidae (Orgyia, Dasychira) to NPV.
- Research aimed at predicting sensitivity of native Lepidoptera to BTK and a basis for predicting species level effects without actually doing assays.
- Research into the possibility of use of BTK as part of a strategy to reduce Compsilura impacts to larger summer moths.
- Documentation on the current status of native parasitoids, especially Tachinidae in areas heavily infested by Compsilura.
- Some effort to establish baseline data and monitor Saturniidae, Sphingidae, Datana and perhaps others as Compsilura continues to expand and increase, including where this fly was introduced in the western USA.