Search

Several researchers have developed a one-generational computer model that simulates infection prevalence of gypsy moth, Lymantria dispar , caterpillars by its fungal pathogen, Entomophaga maimaiga . Inputs required are temperature, humidity, and rainfall records, a measure of fungus resting spore load in the soil, and an estimate of gypsy moth larval density. In a previous study, the model accurately tracked fungal-induced host mortality as long as airborne fungal conidia were allowed to disperse freely over a local area. In 2002, dispersal of conidia and its influence on the impact of the fungus on the gypsy moth was investigated. Gypsy moth densities and fungus resting spore loads were measured in 15 plots within a 3 km area. In 7 of the plots, prevalence of fungal disease was determined weekly by collecting and rearing gypsy moth larvae. Different strategies were used to disperse conidia within the model, and resulting simulated prevalence rates were compared to actual data. Model output was most accurate when airborne conidia were permitted to disperse equally to all plots. Thus, to accurately assess the impact of the fungus in one location, it is necessary to take into account fungal activity throughout the local area.

A computer model developed to simulate the interaction between the pathogenic fungus, Entomophaga maimaiga, and the gypsy moth, Lymantria dispar, was used to investigate airborne dispersal of fungal conidial spores. The model used data on egg mass density gathered from 32 sites in an area of Connecticut and information on fungal prevalence, gypsy moth abundance, fungus resting spore load in the soil, and detailed weather records from six intensively sampled plots in the same area. It calculated seasonal survival rates of gypsy moths at the six plots using a variety of dispersal distributions, or kernels, for the conidia. Distributions ranged from normal densities to the exponential, a Bessel and power distribution, along with a “kinked” linear distribution, which had two parts. All kernels gave good fits to the data so long as the dispersion parameter caused each standardized distribution to have an abscissa value near 0.05 at a distance of 1.25 km from the source. However, significant dispersal ...

A computer model of the interaction between the fungal pathogen Entomophaga maimaiga Humber, Shimazu & Soper and its host, the gypsy moth, Lymantria dispar (L.), had previously been developed to simulate infection by fungus resting spores (azygospores) that are present in soil and germinate in spring if moisture is adequate. This model is now expanded to include infection caused by conidia that are produced on dead, infected caterpillars when relative humidity is near 100%. As inputs, the model uses temperature, relative humidity, rainfall, abundance of resting spores in the soil, and gypsy moth population density. From plots in 1999, 2000, and 2001, data on weather conditions (the first year from daily weather station data and the last two from in-plot data loggers) were obtained, and gypsy moth density and resting spore abundance were determined. Also, collections of gypsy moth larvae were taken weekly to determine proportion of infection by the fungus. These data were used to find best-fitting parameters for the model. Using the same parameter values, model output was close to data for all years. When conidial dispersal was incorporated in the model, output and data were even closer. A validated model should be useful for evaluating and predicting gypsy moth-fungus interactions in forests.

Des tests et des denombrements directs ont permis d'estimer l'abondance des spores de reserve du champignon Entomophaga maimaiga Humber, Shimazu et Soper, un pathogene de la Spongieuse Lymantria dispar (L.), dans des sols forestiers. Les spores de reserve recoltees en octobre, janvier et mars et refrigerees ont germe aussi facilement que les spores recoltees en avril, mais la germination etait plus rapide chez ces derniers. Des tests biologiques sur les spores de reserve de plusieurs sites du Connecticut etaient fortement relies aux resultats de denombrements directs des sols et en correlation faible avec un test anterieur a l'echelle de la foret. Le nombre de spores de reserve a un site etait inversement proportionnel au nombre d'annees ecoule depuis la derniere defoliation par la Spongieuse, ce qui semble devoir etablir implicitement a approximativement 10 ans la viabilite maximale des spores de reserve. Cette estimation de la longevite maximale correspond a la valeur obtenue par mesure directe de la viabilite a long terme des spores de reserve. Les resultats indiquent que la quantite de spores de reserve contenue dans le sol peut etre un facteur determinant important de l'efficacite du pathogene dans la lutte contre la Spongieuse.

Numbers of ants moving on trunks of oak trees and successive visits of workers of the forest ant Formica neogagates Emery to leaves of black birch trees were recorded in the forest. Ants were found to search systematically because they tended to move in the same direction on a twig before and after visiting a leaf. However, many leaves were skipped, and the probability of visiting an encountered leaf was about 0.5. Data were used to develop a computer model of ant foraging on leaves and twigs of trees. It was found that reasonable model outputs of numbers of leaves skipped between visits and the number of new leaves visited, as the total number of visits increased, could be obtained only if ants were assumed to partially avoid leaves that they had previously visited. Model results implied that 100 ants in a tree foraging for about 3 h should be able to visit about one quarter of the leaves in a tree.

Foraging trails of the ant Formica neogagates Emery on leaves of trees and shrubs in the forest or on artificial leaves in a laboratory arena were recorded and digitized. About the same proportion of a leaf was explored by ants whatever the plant species or size of leaf. The portions and numbers of leaves revisited by ants were consistent with the hypothesis that ants do not recognize areas they have recently explored. Trail characteristics of ants foraging on artificial laboratory leaves were similar whether or not leaves had a drop of honey at their base, except that on leaves with honey, more trails occurred and large turning angles were more frequent. By continually watching ants as they climbed up and down tree trunks in the forest, I was able to deduce that up to 150 individuals of F. neogagates may be in a tree at any one time, and that some ants forage in trees for at least 4 h at a time. At this foraging intensity, a significant fraction of the leaves on a tree could be explored by these ants in a day.

Neonate and larger gypsy moth, Lymantria dispar (L.), larvae were collected in a forest and exposed in the leaf litter and trees to determine proportions infected by the fungal pathogen Entomophaga maimaiga Humber, Shimazu & Soper. Results were used to refine a previously developed model of fungus infection of gypsy moths. The revised model only allowed infection of 4th- and larger instar gypsy moths through the germination of resting spores. Infection of neonates via germinating resting spores and of larger gypsy moths through release of conidia from sporulating cadavers was not included because adequate information about transmission rates were not available. This model produced a generally good fit to weekly infection data obtained from collected early to midsized gypsy moth larvae in 1990, 1991, and 1998, so long as a multiplying parameter that did not change the shape of the curve was permitted to vary. This parameter was correlated with a measure of infection caused by resting spores. The intensity of spore germination and resulting infection in the forest is an important predictor of infection by this fungus of early to midsized gypsy moth larvae, and ways to measure the infection potential of such spores before gypsy moth larvae hatch in the spring would greatly aid efforts to predict the impact of this biological control agent.

A model based on weather records and gypsy moth and pathogen developmental times was used to investigate environments favorable for development of the fungus Entonwphaga maimaiga Humber, Shimazu & Soper, a pathogen of the gypsy moth, Lymantria dispar (L.), in the northeastern United States. The fungus was first noticed in 1989, and model results were obtained using data from that year from almost 400 weather stations from Pennsylvania and New Jersey to Maine. The most favorable areas in that year were parts of Connecticut, the northern third of Pennsylvania, and the southern half of New York State. The distribution of the fungus during that year, and the known ability of the fungus to spread, is consistent with an origin for the epizootic in Connecticut. When the model was run using weather records in Connecticut from 1969 to 1995, the most favorable year for development of E. maimaiga was 1989, the year in which the pathogen was first discovered in North America. However, extensive defoliation and at least moderately favorable weather conditions suggested that the fungus would also have been observed in 1971 if it had been present then. Also, model results using weather records in Massachusetts showed that the year 1945 was very favorable for development of E. maimaiga in that state. Because there is no evidence that the fungus was present at these times, it is probable that E. maimaiga became established in North America relatively recently.

A deterministic model was developed that predicts dispersal distance and survival of neonate gypsy moth, Lymantria dispar (L.), larvae as functions of host tree acceptability , tree density, tree canopy size, and intensity of ant predation. This model was compared with output from a more realistic, individual-based computer model. The outputs of both models were similar, implying that the deterministic model assumptions were reasonable. In both models, the number of larvae killed by ants and the distance larvae dispersed increased as tree separation increased. Conversely, predation and dispersal decreased as the size of tree canopies increased. Predation also decreased as trees became more palatable to larvae, and increased as ant numbers and their effectiveness as predators increased. Model outputs were quite sensitive to some ranges of predator effectiveness. Changes in number of dispersal episodes, tree acceptability, ant predation intensity, and unimpeded dispersal distance (i.e., theoretical dispersal distance its the absence of trees) had little effect on actual distances larvae dispersed. Only forest geometry (tree density and canopy size) had substantial effects on dispersal distance. These effects on dispersal were unexpected, boot suggest that information about natural dispersal of neonate gypsy moth larvae can be inferred from simple forest measurements.

From 1985 through 1995, burlap flaps were stapled to 100 trees in a 10 by 10 grid in a 4-ha forest plot. Each year, the number of fourth- to six-instar larvae of gypsy moth, Lymantria dispar L., under each burlap flap was recorded at least weekly in June and into July. From 1987 through 1990,100 000–120 000 hatching gypsy moth larvae were released near the center of this plot in early May at the time of natural hatch. In the 4 years that releases occurred, there was a distinct, single cluster of large larvae under the burlap flaps near the center of the plot. This pattern did not occur in other years. Results are interpreted to mean that total dispersal of larvae from first to ultimate instar is limited, even though first instars are readily dispersed by wind.

A logistic regression model was developed that predicts the probability of forest defoliation by the gypsy moth, Lymantria dispar L., in Connecticut. Data used to develop the model were obtained from sketch maps of defoliation in Connecticut from 1982 to 1994, interpolations (using ordinary point kriging) of fixed area egg mass counts from permanent plots in Connecticut from 1986 to 1994, digital soil maps, and digital elevation models. Subjective comparisons of predicted and actual defoliation maps over years used to fit model parameters were similar. The model was validated with data from years not used in developing the model (1975 to 1985 and 1995). When defoliation was averaged over all locales for validation years, average predicted defoliation tracked actual average defoliation, although the association between them was not significant. Correlations between predicted and actual defoliation, when pairs being compared were at the same locations, were below 0.2 for years in which 2% of the state was defoliated, correlations ranged up to 0.9. The model developed here should be useful in providing probabilities of defoliation that can be used in population suppression programs.

Granulated sucrose, ground-up dry cat food, or sucrose plus dry cat food were distributed weekly in forest plots during the spring and summer of 1995. Counts of ants, particularly Formica subsericea Say and F. neogagates Emery (Hymenoptera: Formicidae), seen in 5-min walks through each plot were similar in the plots receiving supplemental foods to those in control plots. However, numbers of ants coming to canned cat food baits were significantly higher in plots receiving sucrose than in the controls. Ants in sugar plots were probably searching for protein-rich foods to feed larvae. The implications these results have for using forest ants as control agents for the gypsy moth, Lymantria dispar L., and other forest pests are discussed.

Cannibalistic immatures of the gypsy moth (Lymantria dispar L. [Lepidoptera: Lymantriidae]) predator Calosoma sycophanta L. (Coleoptera: Carabidae) must usually be reared individually. This paper describes a method of rearing C. sycophanta larvae in groups by placing them in plastic containers having 3 cm of moist peat moss in the bottom. Larvae are fed gypsy moth pupae placed on top of the peat moss. Under these conditions, larval mortality is about 50%. By varying the number of prey, it was found that larval mortality increased if less than 0.5 pupae/day/larvae were provided. However, when not starved, at least 40 C. sycophanta larvae can be reared per container with no increase in mortality. Larvae were seen to burrow through the peat moss extensively, which may reduce cannibalism. Higher larval mortality occurred when using paper towelling (wet or dry) as the substrate. Group rearing will make it easier to supply beetles for augmentative releases against the gyspy moth.

Thirty forested plots throughout Connecticut were established to determine the associationof forest site conditions with population sizes of common ants (Hymenoptera: Formicidae).The plot characteristics investigated were plot slope, elevation,longitude,latitude, basal area of trees and number of understory stems of species associated with moist conditions, and total length of intersection of rocks along 4 plot transects.Two plot characteristics were consistently related to ant abundance: plot latitude and basal area of trees that prefer mesic or wet sites. Ant numbers were less at high latitudes and in moister sites. Implications of these results for improving biological control by ants are discussed.

Forest ants in Connecticut, especially Formica neogagates Emery and F. subsericea Say (Hymenoptera: Formicidae), feed on first- to third-instar larvae of gypsy moths, Lymantria dispar L. (Lepidoptera: Lymantriidae), but mainly forage on the ground while most young caterpillars are in trees. It would be helpful to obtain information about the individual ants that do climb trees. Thus, numbers of ants climbing up and down different species of forest trees were determined. Numbers varied with both tree species and trunk diameter. Some oaks, plus hickories, tended to have the most ant traffic. These trees were also the most likely to have trunk diameters positively correlated with ant activity. Ants appeared to obtain honeydew or perhaps nectar from the trees, as the crop contents of some ants tested positive for sugar. A possible way to induce more ants to climb trees, and thus to attack caterpillars in trees more frequently, would be to increase numbers of honeydew-producing insects in trees. It should also be possible to encourage ants by spraying trees with sugar solutions or other materials.

The spatial distribution of the eulophid Sympiesis marylandensis Girault was examined in Connecticut in October 1988 and 1989 by vacuum-sampling apple trees in three zones of two apple orchards infested by its gracillariid host, the apple blotch leafminer, Phyllonorycter crataegella (Clemens). The abundance of females of S. marylandensis decreased with increasing distance of the zone or the sampled apple tree from the forest. The abundance of males in different regions of the orchard was similar. In two of three samples, the density of mines of leafminers was significantly higher near the edge than in the center of orchards. Parasitization of leafminers of the second generation and the number of female parasitoids vacuum-sampled during the third generation did not have a similar distributional pattern. We suggest that the amount of parasitism by S. marylandensis would be affected negatively by spraying trees at the border of orchards to kill immigrating pests.

In a Connecticut forest, foraging distances were measured and nests of ants (Hymenoptera: Formicidae) were located by following rewarded individuals back to their nests. Nests were also located by removing all forest litter within an area and counting numbers of ants per nest. Of the three most common species, Formica subsericea Say foraged further from its nest than did either F. neogagates Emery or Aphaenogaster fulva Roger. A computer-intensive nearest-neighbor analysis showed that nests are distributed without relation to other nests of the same or different species. Also, there was no consistent correlation between a nest's nearest neighbor and the combined size of both nests, but some ants appeared to avoid foraging in the same areas. At baits, F. subsericea was much more common than would be expected from its population density. The implications these results have for ant effectiveness as biological control agents are discussed.

In a previous study, ants had a significant effect on gypsy moth, Lymantria dispar (L.), numbers in a forest where the trees were only about 3 m high. The present study was carried out similarly, but trees were about 20 m high. In some plots, ant numbers were suppressed by a combination of barriers and poison baits. In other plots, 24% sucrose sprays were applied with a back-pack mist blower in an attempt to increase ant abundance, but no increases occurred. There was a nonsignificant trend for caterpillar numbers to be higher in the plots where ants were suppressed, and gypsy moth larval counts on leaves of trees were negatively correlated with ant counts on leaves. Thus, there is evidence that ants have an effect on gypsy moth caterpillars in mature forests as well as in young ones.

Adult Calosoma sycophanta L. (Carabidae: Coleoptera) male and female pairs in a laboratory arena were observed for up to 30 days with a video recorder. Beetles were visible (i.e. not hiding) or in soil (i.e. female ovipositing) especially during night hours. Feeding on gypsy moth larvae occurred any time of the day or night. Male and female beetles were most active during the time when female beetles were ovipositing. Females spent more time eating and staying in soil than did males, but males tended to move around the arena more than did females. Based on results from this study, it is expected that, in the forest, C. sycophanta females will spend most of their time hidden in leaf litter or ovipositing. Males are more likely to be seen because they spend more time moving about, probably seeking mates.

Ants in 12 by 12 m plots in a young, regrown mixed oak-maple forest were attracted by sugar sprays or suppressed using barrier bands and poison baits. Sampling for ants using cat food baits and direct observations confirmed that ant abundance was changed by these treatments. Predation of tethered gypsy moth, Lymantria dispar (L.), larvae followed a similar pattern. Direct counts of young gypsy moth larvae were lower in plots where ants were attracted with sprays and higher in plots where ants were suppressed with baits. The implications these results have for control of gypsy moths by ants are discussed.

Sampling for the gypsy moth fungus, Entomophaga maimaiga , was carried out in 10 plots in Connecticut to determine the importance of overwintering fungal load, rainfall, gypsy moth abundance, and physiographic site conditions to fungal prevalence. At one of the plots, gypsy moths were also exposed daily in cages on the soil and at 2 m in shrubs to examine the pathogen transmission potential of germinating resting spores and of conidia produced by diseased larvae. Gypsy moth population levels and an index of site moisture (as determined by tree composition) were marginally related to fungus prevalence. However, the relationship between fungal prevalence and population was an inverse one, opposite that found in a previous study. Resting spore load did not appear to affect prevalence; it is possible that a threshold effect occurs. Daily sampling showed that germination of resting Spores is necessary for infection of early-stage larvae, but that conidia produced from diseased larvae are mainly responsible for infection of late-stage caterpillars. Maximum germination of resting spores occurred 1 or 2 d after significant precipitation, whereas the greatest production of conidia from diseased caterpillars took place on the day of rainfall. We hypothesize that in areas where overwintering resting spores are abundant, the most important determinant of fungal effectiveness will be rainfall amount and pattern.

The effect of the fungus, Entomophaga maimaiga, on gypsy moth populations was investigated by assessing the prevalence of disease in caterpillars collected from forest plots and in laboratory-reared, sentinel caterpillars exposed in wire cages in the same plots. The importance of moisture was evaluated by spraying water on portions of two sites four times a day. Large amounts of rainfall in May appeared to be critical for development of the fungus. Under favorable conditions, infection rates averaged 5% or less in May, but reached 40% or more in June. Infective conidia were produced in May, while resting spores became increasingly common within dead cadavers in June. The prevalence of E. maimaiga was slightly higher in forest-collected larvae from sprinkled portions of the sites, and was significantly higher in caged larvae in the irrigated areas. Infection rates also varied between plots, and were correlated with gypsy moth density and tree composition. Using a degree-day model, estimates of daily recruitment and death of fungus-infected caterpillars were obtained. Recruitment in May was not related to precipitation, possibly because of abundant rainfall, but in June recruitment was correlated with the sporadic rainfall that occurred then. Data on recruitment and death were used to construct survival curves for larvae. Only a few to over half the larvae survived the fungus epizootic in different plots.

Publisher Summary This article emphasizes the diverse ways in which predaceous insects live. Many insect predators carry out active hunting. The adults of tiger beetles (Cicindellidae), ground beetles (Carabidae), and many ants (Formicidae) actively run over the ground to capture prey. Many walking or crawling predators, including ladybird beetles (Coccinellidae), lacewing larvae (Chrysopidae), syrphid fly larvae (Syrphidae), and predatory bugs (Hemiptera) feed on sedentary insects such as aphids and scale insects or on eggs and young larvae of more mobile species. Other insect predators are agile flyers and actively snatch insects out of the air. Dragonflies and damselflies (Odonata) have large eyes and strong wings and feed mainly on mosquitoes and other small flying insects. Robber flies (Asilidae) also catch their prey on the wing, but will often tackle and subdue insects at least their own size, even bees and wasps. Their legs are very strong and they have piercing mouthparts to suck up prey juices. Other insects, such as many social wasps (Vespidae), pluck insects from the ground or vegetation while flying. Many predaceous insects live in fresh water and pursue prey by swimming. The giant water bugs (Belostomatidae) are excellent swimmers and often capture and subdue small fish and tadpoles. Further, insect predators are more important in the regulation of their prey species than is commonly thought. Because they often leave no identifiable remains after devouring prey, the extent of predation is hard to quantify.

Publisher Summary This chapter discusses how parasitoids seek their hosts. The chemicals produced by insect hosts that serve as cues to parasitoids are kairomones because the parasitoid exploits them to the host's detriment. Usage and context are important here. A sex pheromone may be beneficial in the context of reproduction to attract a male moth to a female, but if a parasitoid cues on this chemical, the substance also functions in the context of interactions between hosts and parasitoids as a kairomone. A female parasitoid may find herself far from potential hosts. This could occur if the host stage she emerges from is different from the one attacked. Also, many parasitoid females have a preoviposition period before eggs are ready to be laid. During this interval of a few days to several weeks, the parasitoid may leave the vicinity of the host to mate and obtain nourishment. For example, the ichneumonid wasp, Pimpla ruficollis , is a parasitoid of the European pine shoot moth, Rhyacionia buoliana . Yet for the first few weeks of her adult life, she is repelled by the odor of pine, and thus avoids the forest where the host is located. As a consequence, the initial stage in host seeking in many parasitoids is to search for locations where the host is likely to occur. Parasitoids often respond to general stimuli such as light, humidity, or vegetation form, leading them to meadows, forests, swamps, ponds, soil, or different vegetation strata. These behaviors considerably narrow the areas that must be actually searched for hosts.

The length of the time young gypsy moth, Lymantria dispar L. (Lepidoptera: Lymantriidae), larvae remain in litter after failing from trees would help in understanding the impact of predation by ants, as ants appear to be important predators of gypsy moths only in litter. To obtain information on caterpillar litter residence times, visual activity of larvae was investigated in the laboratory by presenting them with vertical dark silhouettes of various widths in a 1 m wide cylindrical arena. Their limit of angular resolution was about 1°. This value helped to determine distances to trees caterpillars could be expected to orient toward from randomly-located points in a forest. These data were used with results on litter crawling speed of instars 1, 2, and 3 in a simulation model that generated litter residence times averaging several hours. By combining results with a previously-developed model of ant predation in forest litter, it was estimated that about half of unfed first instars falling into the litter would survive, whereas only about 10% of 2nd and 3rd instars would. The implications these results have for control of the gypsy moth are discussed. Resume Simulation du temps de presence des jeunes chenilles de Lymantria dispar dans la litiere et consequences sur leur predation par les fourmis La predation par les fourmis du genre Formica spp. a une influence importante sur les premiers stades de L. dispar (Lepido., Lymantriidae) en Amerique du Nord. La duree du temps passe dans la litiere par les jeunes chenilles apres etre tombees des arbres doit permettre de comprendre l'impact de la predation par les fourmis. Pour connaitre le temps de sejour des chenilles dans la litiere, l'acuite visuelle des chenilles a ete examinee au laboratoire en leur presentant des silhouettes noires verticales de differentes largeurs dans une arene circulaire de 1 m de diametre. La limite de leur angle de resolution etait d'environ 1 °. Cette valeur a ete utilisee pour determiner les distances (jusqu'a 5 m) quelles que soient les orientations dans la foret, d'arbres supposes susceptibles d'orienter les chenilles. Les arbres aux silhouettes acceptables etaient distribues uniformement entre 1 et 5 m. Les vitesses de deplacement dans la litiere des stades 1, 2 et 3 ont ete mesurees par observation directe dans la foret (elles variaient de 0,9 a 1,54 cm/min). La distance des arbres et la vitesse de deplacement ont servi a construire un modele de simulation qui a donne un temps moyen de sejour de plusieurs heures. En combinant ces resultats avec un modele precedemment publie de la predation par les fourmis dans la litiere de la foret, qui utilisait le temps de contact des chenilles avec les fourmis et les taux de captures victorieuses pour estimer la mortalite, nous avons determine que la moitie environ des chenilles non alimentees de premier stade tombees dans la litiere pourraient survivre, tandis que la survie pour le second et le troisieme stade ne serait que de 10%. La discussion porte sur l'importance de ces resultats dans la lutte contre L. dispar.

Releases of limited numbers of Ca/osoma sycophanta L. adults into high populations of gypsy moths, Lymantria dispar L., were made in 1988 and 1989. In 1988, 0-40 wing-clipped beetle adults were released into four plots. Larval survival of gypsy moths was least in plots receiving the most beetles, but based on gypsy moth pupal mortality and numbers of immature beetles seen, adults did not reproduce. In 1989, a release with four replicates was conducted in which 0 beetles, 20 intact beetles, or 20 beetles with clipped wings were released into separate plots. Larval survival rates for gypsy moths could not be adequately assessed, because of overwhelming mortality due to an Entomophaga fungus, but there were significantly more beetle larvae in plots receiving intact beetles than in the other treatments, and pupal mortality occurred earlier and was greater in release plots than in check plots. The releases show that a small number of adult beetles, through progeny production, can have an effect on gypsy moth populations.

Data were gathered during two years on the death and removal of gypsy moth. Lymantria dispar L.. larvae tethered by threads in various forest microhabitats where gypsy moth densities were low. These data were used to estimate larval mortality rates due to predation, after first correcting for several sources of bias. Values were similar for young and old larvae. but were higher and more variable in the second year. The estimates were used to run Monte Carlo simulations to determine variations in daily predation rates. The average daily predation rate for one of the years, 0.121, was very close to the daily mortality rate found from declines in population estimates between early and late-larval instars obtained from mark-recapture studies. In this low-density gypsy moth population, predation was by far the most important mortality factor.

An entomopathogenic fungus, Entomophaga maimaiga, was found causing an extensive epizootic in outbreak populations of the gypsy moth, Lymantria dispar, throughout many forested and residential areas of the northeastern United States. This is the first recognized occurrence of this or any entomophthoralean fungus in North American gypsy moths, and its appearance was coincident with an abnormally wet spring. Most fungal-infected gypsy moth larvae were killed in mass during the fourth and fifth stadium and were characteristically found clinging to the trunks of trees with their heads pointed downward. The fungus produces thick-walled resistant resting spores within dried gypsy moth cadavers and infectious conidia when freshly killed larvae are held in a wet environment. The morphology and development of the fungus are described. The fungus appears to have had its origin in Japan, and the current epizootic may have resulted from the survival and inapparent spread of an early introduction in 1910-1911.

Anastatus kashmirensis Mathur (Hymenoptera: Eupelmidae), an egg parasite imported from India for release against the gypsy moth in North America, attacks other gypsy moth parasites. It develops on one of them ( Apanteles melanoscelus (Ratzeburg) (Hymenoptera: Braconidae) more readily than does the related, established species A. disparis Ruschka. Based on this fact, and because A. kashmirensis appears to be no more able to parasitize deeper eggs in gypsy moth egg masses than the established egg parasites, we suggest that it should not be released in North America. This report emphasizes the importance of prerelease assessment of exotic parasites and cautions against hasty evaluations based on comparisons with related species.

Observations of premating behavior of Apanteles melanoscelus (Ratzeburg) (Hymenoptera: Braconidae) males exposed to parts of female abdomens suggested that the sex pheromone gland of females was not Dufour's gland as previously reported but paired epidermal glands on the 8th abdominal tergum. The columnar cells of these glands contained lipid deposits, microvilli, and numerous mitochondria, which are typical of pheromone gland cells in other insects.

I studied dispersal of gypsy moths, Lymantria dispar (L.), released as eggs into a naturally low gypsy moth population, and retention of small larvae on isolated oak terminals in the same area. Most larvae recovered after eggs hatched were close to the release point and dispersed randomly shortly thereafter. Larvae were lost from branch terminals, but usually at least 80% were retained from one day to the next. I also carried out a mark-recapture study of large larvae in this same area, as well as in another area with a moderately high infestation. Caterpillars were marked by selectively removing prolegs. Dispersal patterns of larvae at both sites were similar and population densities were about 10,000 and 400,000 larvae per ha at the low- and high-population sites, respectively. Survival rates were highest at the high-population site. Estimated population values were closely correlated with the number of large caterpillars found under burlap bands, suggesting that the latter may usefully indicate absolute population levels

In a 2-month expedition to Beijing and northeast People's Republic of China (PRC) to investigate natural enemies of the gypsy moth, Lymantria dispar (L.), much was learned about the pest itself. Its distribution is wide in the PRC but it is more abundant in the northeast where local outbreaks occurred in 1980 and/or 1981. Known outbreak history is tabulated. Gravid adult females are capable of sustained flight and often lay eggs on structures near outdoor lights, to which they are attracted. Larval behavior is similar to that of larvae in North America. At low population levels much feeding is done at night and large larvae rest under bark flaps, stones, or leaf litter during the day. Larval food plants are listed and are similar to that of New World larvae except that persimmon, Diospyros sp., in the Beijing area, and larch, Larix spp., in the northeast, are highly preferred. Larval developmental phenology based on our collected larvae is presented. Larval color variation is pronounced (greater quantities of yellow) but this varies within sites and geographically. About 0.3% of larvae collected were black-backed mutants, an example of which is illustrated.

Observations of behavior and dissections of hosts were carried out to investigate the ability ofApanteles melanoscelusRatzeburg (Hymenoptera: Braconidae) to distinguish between parasitized and nonparasitized gypsy moth larvae,Porthetria dispar(L.) (Lepidoptera: Lymantriidae). Fewer host examinations occurred and fewer eggs were laid in hosts which had been parasitized 0–9 days previously than in those which had never been parasitized. However, superparasitism occurred readily especially when few unparasitized hosts were available. By dissecting hosts which were parasitized more than once it was possible to investigate competition between supernumerary parasitoids. First instars are eliminated by some mechanism (possibly active combat), and older parasitoids eliminate younger ones.

Young adults of Ooencyrtus kuvanae (Howard) were held at different photoperiods and periodically supplied with gypsy moth, Lymantria dispar (L.), eggs while their progeny production and longevity were noted. Females reared at 10:14 (L:D) and 12:12 photoperiods produced fewer progeny but lived longer than those held at 14:10 or 16:8 photoperiods. Switching females from a 10:14 to a 16:8 photoperiod after 12 days (or vice versa) did not influence progeny production; females continued to reproduce as they would have if photoperiods had not been changed. This suggests that the initial photoperiod environment of the adult is important. Dissections showed that number of mature eggs in ovaries of females held at a 16:8 photoperiod increased more than that in ovaries of females held at a 10:14 photoperiod, corroborating progeny production results. Females preconditioned to long or short photoperiods for 2 weeks did not differ in longevity when held subsequently at low temperatures.

Over a 3-yr period, diapausing larvae within cocoons of Apanteles melanoscelus (Ratzeburg), a gypsy moth ( Lymantria dispar (L.)) larval parasite, were placed in various microhabitats in trees, left for one wk, retrieved, and mortality assessed by examining cocoons for evidence of predation and holding intact cocoons for emergence of hyperparasites. At least 13 species of hyperparasites were recovered in the study, and cocoon predators were prevalent in 2 of the years. Each species of hyperparasite had a particular seasonal trend in abundance. Each also had a particular distribution with respect to location where cocoons were attacked within a tree. Data suggest that visible cocoon predation, which was substantial, was probably caused by vertebrates—possibly birds or small mammals.

Various bioassays were used to determine which forest microhabitats contained the most nuclear polyhedrosis virus (NPV) of the gypsy moth ( Lymantria dispar ) and which of these microhabitats most readily led to infection of newly hatched gypsy moth larvae. Soil just under the leaf litter layer and gypsy moth pupal exuviae on tree trunks had significantly more virus than tree bark. In the laboratory, neonates became infected by walking over field-collected pupal exuviae, but contaminated larvae did not transmit the virus to other larvae when confined for a day with them in the same container. Washing larvae with disinfectant immediately after they walked over pupal exuviae reduced the rate of infection, suggesting that polyhedra from exuviae adhere to caterpillars' integuments and are ingested later. However, less than 7% of larvae that walked over contaminated soil became infected, even though 60% of caterpillars died from NPV in bioassays of water extracts of soil, demonstrating the abundant presence of viable virus in soil. Thus, of the environmental surfaces tested, pupal exuviae were the most important source for horizontal transmission of virus to newly hatched caterpillars.

Gypsy moth larvae, Lymantria dispar L., were fed various doses of Bacillus thuringiensis Berliner and parasitized by Apanteles melanoscelus (Ratzeburg) before or after infection. The pathogen did not directly influence, nor was it influenced by, the parasitoid. Development of those host larvae which survived doses of B. thuringiensis was delayed compared to that of uninfected larvae. Hosts treated in the 2nd instar and held 10 days were attacked to a greater extent than similarly held, nontreated hosts of the same age, probably because more of the former had not yet molted to the 4th instar. The pathogen-induced developmental lag may explain the relatively high amount of parasitism by A. melanoscelus in B. thuringiensis field plots.

Courtship movements of Apanteles melanoscelus (Ratzeburg) (Braconidae) males include wing elevation and fluttering, mounting, and copulation. Copulation lengths of 20 sec and more are necessary for sperm transfer. Males older than one wk did not mate readily, but old females were as receptive as young ones, as long as they had not previously mated. Conditions where females tend to be quiescent increase mating success. The female sex pheromone is relatively non-volatile and does not diminish in activity after mating or as females age. Non-pheromonal stimuli from females are necessary as well as pheromone for the complete expression of mating behavior in males.

Apameles melanoscelus (Ratzeburg) females examined gypsy moth ( Lymantria dispar L.) silk more intensely in the afternoon than they did in the morning, but they examined it for decreasing lengths of time the longer they were exposed to it. This decrease was not due to a buildup of deterrent chemicals deposited by the female. Parasites were reactivated by holding them without silk for 60 min, indicating that habituation was involved. Parasites also remained active longer when increasing amounts of silk or hosts in conjunction with silk were presented to them.

Gypsy moth instars four through six move down trees at dawn and climb back up at dusk. This daily migration appears to be influenced by environmental conditions, including abiotic factors. In this study, gypsy moth caterpillars were given choices of abiotic conditions to see what their preferences are. Different gypsy moth instars responded similarly in a temperature gradient, were indifferent to direct infrared radiation, and moved upward in a uniformly illuminated tube. All instars tended to accumulate at the lower of two relative humidities when given a choice. Small larvae, but not large ones, were strongly attracted to light. Under a laboratory photoperiod of 16 h light per day, fifth instars migrated away from a food source during the lighted period and toward the food when it was dark. I conclude that sunlight is probably responsible for stimulating large caterpillars to migrate up and down trees.

Laboratory tests with the acaricide Plictran (tricyclohexylhydroxystannane), suggest that this gypsy moth feeding deterrent, like Bacillus thuringiensis, has minimal direct effects on larval survival, adult longevity, and progeny production of the parasite, Cotesia melanoscelus. Plictran and very low doses of B. thuringiensis (down to 1/25 the recommended dose) sprayed on apple trees did not significantly reduce gypsy moth numbers but did retard caterpillar development. Thus these materials should increase parasitization by C. melanoscelus in the field because they increase the length of time the parasite, which attacks only small hosts successfully, can parasitize caterpillars.

Artificial and modified natural hosts were exposed to females of the gypsy moth [Lymantria dispar (L.)] hyperparasite,Eurytoma appendigaster (Swederus), to investigate its host recognition behavior on the primary host, which are cocooned larvae of the gypsy moth parasite,Cotesia melanoscela (Ratzeburg). Material(s) which caused drilling behavior by the hyperparasite on host cocoons were extracted with both polar and non-polar solvents. However, cocoons washed with large volumes of solvent still caused substantial drilling activities by females, suggesting that additional cues may be important. Results suggest that host recognition in this hyperparasite involves a variety of host characteristics.

Mark/recapture surveys were carried out to evaluate the relationship between counts of caterpillars under burlap-band resting sites and gypsy moth, Lymantria dispar (L.), populations. Principal component analysis showed that rain on the day of sampling was related to decreases in numbers counted under burlap. Caterpillars did not readily move between burlap resting sites placed high in trees and those lower down, and larvae found under high burlap developed slower and had greater mortality than those under low burlap. Survival rates calculated from the mark/recapture data were higher for larvae under low burlap. Calculations of population levels from the mark/recapture data showed that the part of the population sampled by low burlap was 10-fold as large as that sampled by high burlap. Estimates of daily survival rates of the first four instars were close to the corresponding rates for large larvae. There was a correlation of 0.95 between mark/recapture log population estimates for large caterpillars and log counts over four orders of magnitude, but comparisons with independent data suggest that the relationship between log counts and log populations may be nonlinear.

In the field, ichneumonid hyperparasites were active around Apanteles melanoscelus (Ratzeburg) (Hymenoptera: Braconidae) cocoons over a larger temperature range than were chalcidoids. Observations of cocoons on leaves and trunks of trees showed that larval (intrinsic) competition by the hyperparasite, Eurytoma verticillata (F.) (Hymenoptera: Eurytomidae), partially excluded other hyperparasites from the trunk microhabitat. In laboratory studies E. verticillata larvae consistently destroyed larvae of another hyperparasite, Gelis tenellis (Say) (Hymenoptera: Ichneumonidae), when both were associated with the same host. The implications that the results have concerning ecology and control of these hyperparasites are discussed.

Parasitic Hymenoptera caught on colored sticky panels placed in various forest microhabitats showed patterns of trap catch that could often be related to the hosts they attack. Parasites caught predominately on yellow panels, which mimic foliage, generally attack foliage-frequenting insects such as leaf-chewers or leaf-miners. The few parasites that did not respond predominately to yellow panels attack such hosts as psocids or eggs of stem- or pod-boring Lepidoptera. The observed correlations between patterns of trap catch and host preferences suggest that sticky panels may be useful in revealing host and microhabitat preferences of little-known parasites.

Pupae and adults of Apanteles melanoscelus (Ratzeburg) (Hymenoptera: Braconidae) were released in 5 plots in 3 sites in Connecticut during the summers of 1973 and 1974. Gypsy moth numbers varied from low to high in the different sites. Weekly collections of caterpillars showed consistently greater percent parasitism by A. melanoscelus in the release plots than in the check plots at all 3 sites for at least the first 2–3 wk after release. In later collections no differences occurred in the 2 sites where gypsy moth numbers were low. In the site with high host numbers, percentage parasitism in the release plots were higher than in the checks even at the 7th weekly collection after release.