Proceedings of the 62^ndAnnual Meeting of the
Acadian Entomological Society
in conjunction with the
Maine Entomological Society
July 21-23, 2002 at
The University of Maine, Machias, Maine

Carole Cheah and Charlene Donahue. Connecticut Agricultural Experiment Station, Hamden, CT and Maine Forest Service Augusta, ME.

Balsam woolly adelgid, Adelges piceae Ratzeburg, is a serious introduced pest of true firs in North America. With recent warming climatic trends, there has been a resurgence of coastal infestations in Maine where mortality and decline of balsam fir, Abies balsamea L. has been steadily increasing. Many attempts at biological control of A. piceae have been unsuccessful in the past. A closely related species, the hemlock woolly adelgid, Adelges tsugae Annand, is also currently attacking and killing native hemlocks in the eastern USA. Almost half the range of eastern hemlock, Tsuga canadensis Carriere is infested and the range is expanding at an alarming rate, threatening northern and southern hemlock forests. Pseudoscymnus tsugae is a tiny coccinellid which has been released in 15 eastern states for biological control of A. tsugae since 1995. This joint study in 2001-2002 between Maine and Connecticut investigated the potential of P. tsugae for biological control of A. piceae and for advance implementation of this predator for A. tsugae, which has not yet established in Maine.

P. tsugae adults were released into bole cages in mid to late July 2001 at two sites: on infested Fraser firs (Great Mountain Forest, Norfolk, Connecticut) and balsam fir (Sebec, Maine). Weekly observations first detected mature larvae in late August in Connecticut and in early September 2001 in Maine and again in June and July 2002 at both sites, confirming reproduction and development on A. piceae. Adult P. tsugae were readily observed throughout the summer into the fall and survived the 2001-2002 winter on the bole of infested trees in both sites, when minimum February temperatures in the bole cages reached -4.4 to -5.8°F (-20.2 to -21 °C) in Maine and 7.7 °F (-13.5 °C) in Connecticut. General trends in caged A. piceae population levels indicated a reduction by October 2001 in live trees caged with P. tsugae in Maine while control populations remained unchanged. In Connecticut, A. piceae populations increased in both types of bole cages in October 2001 but were significantly lower (2.5x) in cages with P. tsugae. These preliminary trends on P. tsugae impact on A. piceae will require further studies but do indicate the potential of this coccinellid for biological control of balsam woolly adelgid. Data from this study supported the first "pre-inoculation" releases of P. tsugae in Maine on balsam woolly adelgid infestations in 2002.

(Gn.)): Assessment and Refinement of Forecasting Methods for Predicting Populations and Defoliation.

Lester Hartling, Janet Proude, Doug Winter, Danny O'Shea, Dan Lavigne and Nelson Carter. New Brunswick Natural Resources and Energy Forest Pest Management Section, N.B.

The eastern hemlock looper (Lambdina fiscellaria fiscellaria (Gn.)) has been historically a significant forest pest in Newfoundland and Québec, with more recent outbreaks in New Brunswick, Nova Scotia and Maine. Although it feeds primarily on hemlock and balsam fir, the latter species is of primary concern in New Brunswick.

Data from New Brunswick's pheromone trapping network and extensive egg sampling in 2000 suggested hemlock looper populations were increasing. Also significant were increases in moth activity noted in other eastern Provinces and Maine. At the time this project was initiated (2001) it was unclear whether the Province was facing the start of a new looper outbreak or simply a temporary spike in populations. If the former, there would be need to consider future controls. Since proper survey forecasting methods are an essential cornerstone of good integrated pest management it was essential to fill knowledge gaps. Current circumstances provided the opportunity to investigate and improve upon existing biological relationships and survey methods to better monitor and forecast hemlock looper outbreaks, and to aid in the decision-making process when contemplating controls. As it happened, hemlock looper populations in New Brunswick did not reach widespread outbreak levels in 1991, with only 760 ha of detectable defoliation mapped during aerial surveys in 2001. (However, in neighboring Maine approximately 80 000 ha of defoliation were mapped). Hence conclusions from this study pertain to the conditions encountered. However, the study did allow the opportunity to help refine existing methods for forecasting and monitoring. Furthermore, the final report integrated data from this study and previous DNRE studies, and compared these findings with that of other jurisdictions. Such information will be invaluable if future foliage protection programs become necessary.

The study was able to build on relationships between pheromone trap catch and egg density, egg and early instar larval counts, egg counts and defoliation, and larval numbers and defoliation. Evaluation of pheromone trap survey data showed that pheromone (10-mg strength lure) trap catches of <1200 moths/trap are unlikely to represent threatening populations because associated egg densities never exceeded 15 eggs/100-cm lower-crown branch. Furthermore, data integrated from this study and previous DNRE studies suggest that counts up to ~25 eggs/100-cm lower-crown branch seldom cause >10% loss of current-year foliage and seldom exceed 25% loss of the 3 older age classes of needles.

This study and previous ones indicate most of the egg parasitism was not accounted for when sampling in the fall and winter months. Significant increases in egg parasitism were caused by spring-attacking parasitoids (i.e. Telenomus nr. alsophilae Viereck). Any population forecast derived solely from sampling in the fall and winter months could underestimate egg parasitism and thus have the potential to overestimate the forecast. When planning control plans it is probably prudent to include spring egg surveys to see if treatment areas can be excluded.

For New Brunswick's operational pheromone trap network, traps are hung in the field with just enough of a window for appropriate lure aging, slightly ahead of predicted moth flight. Results from this study suggested pheromone traps may be able to be placed in the field 1 - 2 weeks earlier than 'normal', thus giving some logistic flexibility to surveys at that time. Furthermore, it may be possible to retrieve traps one or two weeks earlier, a significant advantage in areas of higher elevation where snowfall comes early in the season. Data will be collected over several years to strengthen these observations. Degree-day relationships were further developed to aid in forecasting egg hatch and larval development, both essential to 'timing' and assessing control operations. While not intended as a major component of this project, opportunities arose to evaluate methodologies for sampling eggs and larvae, and assessing feeding damage of the current and older age classes of foliage. Likewise, techniques for measuring larval populations (e.g. branch sizes and crown levels within trees), and assessing defoliation of current-year and previous-years' foliage on balsam fir were examined.

Results from this project are expected to be pertinent to various agencies, landowners and other forest management groups dealing with hemlock looper outbreaks in such jurisdictions as Nova Scotia, Québec, Ontario, Newfoundland and the State of Maine. Direct funding for this project was provided by (1) members of the Spray Efficacy Research Group (SERG project 2001- 25F), specifically Forest Protection Limited, SOPFIM (la Société de protection des fôrets contre les insectes et les maladies) and Newfoundland and Labrador Forest Resources and Agrifoods; and (2) Fundy Model Forest. In-kind support was provided by New Brunswick Department of Natural Resources and Energy (Forest Pest Management Section). A comprehensive report on this study is available upon request.

E. Michael Blumenthal. Pennsylvania DCNR-Forestry, Forest Pest Management, Middletown, PA.

The hemlock woolly adelgid (HWA), Adelges tsugae (Annand), is the most serious pest of Eastern hemlock, Tsuga Canadensis, in Pennsylvania, and about half of the Commonwealth is infested. Releases of the lady beetle predator Pseudoscymnus tsugae (Coleoptera: Coccinellidae) (Pt) have been made in eastern and central PA since 1999, and to date about 118,000 adult beetles have been released at 31 sites in 20 counties. Releases at individual sites have ranged from about 800 to over 10,000 Pt on one tree or distributed on up to a 12 trees. The New Jersey Department of Agriculture's Phillip Alampi Biological Control Laboratory, West Trenton, NJ (NJ- Pt) and Ecoscientific Solutions, Inc, Scranton, PA (ESI-Pt) have supplied beetles reared, stored, and shipped in different ways.

Release results have been highly variable, although the beetle is presumed to be established at all release sites based on recoveries the year after release. Initial releases, made on settled HWA nymphs in high density pest infestations where release trees failed to produce new shoot growth the year following release, have appeared least successful. It is likely, however, and peripheral surveys have confirmed, that the released Pt simply dispersed from these release sites in search of more favorable conditions.

Subsequent to these observations, releases have been made in low-to-moderate infestations where tree growth effects were less severe. At several of these sites, Pt adults and larvae have been found two years post-release, are at least locally dispersed, and are producing three generations per year.

Some of the factors which most appear to influence successful establishment (or determination thereof) are as follows:

· Early releases seem to favor Pt survival and reproduction.

· Pt stored in chill for long periods appear to reproduce more slowly in the field.

· Pt shipped with and/or released on preferred prey stages seem to reproduce more quickly and disperse less.

· Pt reproduction in the second and third generations may be related to availability of long-lived progrediens adults and persistence of non-viable eggs.

· Pt seem to prefer brightly illuminated feeding sites.

· Pt appears to prefer healthy prey on healthy trees.

· Pt reproduction may be independent of prey density.

· Predation on Pt appears to be very high, especially by spiders.

· Intuitive sampling seems to recover more Pt than systematic sampling.

· Pt larvae appear not to be randomly distributed among similar prey densities within lower tree strata.

Perhaps the most remarkable finding this season has been that releases made in PA in late March 2002 (prior to several sub-freezing nights) seem to exhibit the best survival (least dispersal?), reproduction, and synchronization with HWA.

Abell, Kristopher. Department of Biological Sciences, University of Maine, Orono, ME.

Short and long-term effects of herbicide application and clearcut timber harvesting on parasitoid wasp communities were studied in Bingham, Maine. Samples were collected from six different treatment groups. Treatments were as follows: 1) 3-5 year old clearcuts treated with herbicide, 2) 3-5 year old untreated clearcuts, 3)15-16 year old clearcuts treated with herbicide, 4) 15-16 year old untreated clearcuts, 5) 15-16 year old plantations (herbicide clearcut planted with black spruce), and 6) untreated mature forest. There were four replicates of treatments 3 and 4, and three replicates of treatments 1,2,5, and 6 for a total of 20 sites. Samples were collected every two weeks from June 14 to August 30, 2000 and from June 5 to August 28, 2001. Samples were taken using malaise traps, pitfall traps, and flight intercept traps. All wasps were identified to family with wingless wasps, Ichneumonidae, and Brachonidae separated into morphospecies. To investigate a possible correlation between floral resources and wasp abundance, floral density was estimated at each site by four 50 meter transects. Preliminary results indicate that there were harvesting and herbicide effects on both wingless and winged wasp abundance. There was a strong correlation between wingless wasp abundance and floral density. Winged wasp abundance weakly correlated with floral density indicating a complex response to disturbance.

Francis A. Drummond and Charlene Donahue. Department of Biological Sciences, University of Maine, Orono, ME and Maine Forest Service Augusta, ME.

The arborvitae (cedar) leafminer complex in Maine is comprised of four species of micro-lepidoptera; three species of Argyresthia (thuiella, freyella, and aureoargentella) in the family Argyresthiidae and Coleotechnites thujaella in the family Gelechiidae. Based upon a 32 year survey of cedar for leafminer infestation in eight locations in Maine, it was found that of these four species, A. thuiella predominates (ca. 70% of the abundance). The abundances of the four species were highly positively correlated with one another at most of the sites during the survey period, suggesting that interspecific competition is not a significant factor in their population dynamics. Analysis of the complex of the four species suggested that a high degree of spatial correlation exists among their population dynamics, meaning that population synchronies over time were more similar between neighboring sites than sites separated by considerable distance. Time series modeling and key factor analysis suggests that density dependent factors such as parasitoids are significant in regulating populations than are density independent factors such as weather, although trends with weather did exist. Radial tree growth was found to be negatively correlated with leafminer complex fluctuations in abundance for time lags of 0, 1, 2, and 3 years. At a lag of 0 years, a density of 4.0 leafminers / twig resulted in a 50% reduction in radial growth.

Bonnie MacCulloch¹ and Douglas C. Allen. Environmental and Forest Biology, State University of New York-College of Environmental Science and Forestry, Syracuse, NY.

Foresters and land managers interested in developing integrated pest management programs should pay particular attention to native predators capable of regulating insect populations. Understanding the impact and response of natural enemies during an outbreak of the introduced pine false webworm (Acantholyda erythrocephala) in northern New York is essential to developing a successful pest management program. In this study, avian predators were examined to determine their responses across an increasing gradient of pine false webworm densities. Using whole-tree exclosures we manipulated avian predators and demonstrated their ability to reduce pine false webworm densities on individual trees by 53%. Avian species richness and diversity were found to be positively associated with increasing pine false webworm densities. Reproductive success was 43 % greater in stands with higher densities of pine false webworm. Multiple flocks of greater than 500 individuals were observed foraging on pine false webworm during early and mid-summer corresponding with the breeding season. This research demonstrates that avian predators can inflict heavy mortality on populations of pine false webworm and may influence the dynamics of these outbreaks. We suggest that avian predators can effectively limit or reduce populations of tree-damaging insects and are a valuable resource for land managers to consider when developing an appropriate integrated pest management program.

¹ Currently at: USDA, Animal Plant Health Inspection Service, Plant Protection and Quarantine, BIA, 267B Godfrey Blvd., Bangor, ME 04444.

Barry Hicks. College of the North Atlantic, Newfoundland.

A strain of Beauveria bassiana isolated from the pine beauty moth, Panolis flammea (Lepidoptera: Noctuidae) has shown good potential to be used as a biological control agent of this forestry pest in Britain. Preliminary research and the potential of employing B. bassiana against forest defoliators of North American forests will be discussed.

Sweeney, J.¹, Klimaszewski, J.², Gesner, G.¹, Price, J.¹, and Bousquet, Y. ^{3 1} Natural Resources Canada, Canadian Forest Service, Fredericton, NB, ² Natural Resources Canada, Canadian Forest Service, Sainte Foy, QC and ³ Agriculture & Agri-Food Canada, Ottawa, ON.

The effect of different silvicultural treatments on the relative diversity and abundance of ground beetles in red spruce stands was investigated at two sites in the Acadia Experimental Forest near Noonan, NB. Stands were treated in the winter of 1998-99: 1) Uncut, 2) Selective cut (30% trees removed); 3) Strip cut (50% trees removed); and 4) Clear cut (100% trees removed). A 10-year old black spruce plantation was also sampled at one site. Beetles were sampled with open pitfall traps from May to September 1999 and 2000. Data for carabids sampled in 1999 are presented here. Diversity and similarity statistics were estimated using EstimateS freeware (http://viceroy.eeb.uconn.edu/estimates).

A total of 39 species and 1921 specimens were trapped in 1999. Carabid species diversity was greatest in clear cuts and lowest in uncut plots, with select and strip cuts falling in the middle. Carabid abundance in pitfall traps was greatest in the strip cuts and lowest in the 10-year old plantation. Although clear cuts had greater diversity of species, species composition was very different from that in uncut plots. Uncut plots shared many species with select- and strip cuts but shared few species with clear cuts and the 10-year old plantation. Certain species common in the uncut plots (e.g.,Pterostichus coracinus, Sphaeroderus lecontei) were reduced or absent in clear cuts and the plantation, wheras other species, such as Pterostichus adstrictus and Bembidion versicolor were trapped only in clear cuts and the cut portions of strip cuts. Traps in clear cuts captured several species known to prefer open areas dominated by grasses or weeds, e.g., Harpalus, Amara spp., and most of these were single individuals. Selective cutting and strip cutting appeared to have less effect on carabid species composition than did clear cutting.

Richard A. MacKenzie, Wells National Estuarine Research Reserve, Wells, ME. rmac@wellsnerrcec.lib.me.us

Insects emerging from salt marsh habitat were quantitatively sampled during the spring, summer, and fall of 2001 to document: 1) community structure, 2) the spatial and temporal distribution of these insects, and 3) how various human activities might influence these patterns (i.e, development, marsh management techniques). Samples were collected from the marsh vegetated surface and from standing pools of salt and brackish water on the marsh surface known as marsh pannes. Results revealed that more insects emerged from the vegetated marsh surface (1765 ± 784) than from brackish water pannes (1451 ± 178) or salt marsh pannes (277 ± 109) during the sampling period (end of May through October). Distinctive insect communities were observed in all three habitat types. Insects characteristic of marine environments were more common from the salt marsh pannes (Anthomyiidae, Scathophagidae, and Limonia libertia tipulid flies) while chironomids were the most abundant insects emerging from the vegetated marsh (largely Tanytarsus mendax group) and brackish marsh pannes (largely Chironomus sp.). The vegetated marsh supported a greater number of species (31 species) compared to brackish (18 species) and salt water pannes (20 species), including the seaside dragonlet, Erythrodiplax berenice, the salt marsh caddisfly, Limnephilus ademus, and an unidentified coenagrionid damselfly. The high number of chironomids observed emerging from the salt marsh in the springtime coupled with their well documented high production and turnover rates, suggests their importance as a commonly overlooked food source for fish, birds, and other invertebrates that utilize salt marsh systems.

Phillip deMaynadier and Paul M. Brunelle. Maine Dept. of Inland Fisheries and Wildlife, Bangor, ME and Halifax, NS.

Insects in the order Odonata, damselflies and dragonflies, are a significant and conspicuous component of Maine's wildlife diversity. At this time 162 species have been documented in the state, comprising fully 37% of the total North American fauna (435 species). Considering the distribution ofodonates in neighboring states and provinces, five to ten additional species could occur in Maine. The Maine Damselfly and Dragonfly Survey (MDDS) is a multi-year, volunteer-based atlasing initiative designed to improve our knowledge of the distribution, abundance and status of odonates in Maine.

Currently, there are two state-listed dragonflies in Maine, the ringed boghaunter (Williamsonia lintneri; Endangered) and the pygmy snaketail(Ophiogomphus howei; Threatened). Additionally, there are 30 odonate species listed as Special Concern in Maine (collectively ~20% of the state'sfauna), a non-legal status assigned to species that are believed rare, but that require additional survey before status can be confidently assessed.

In addition to raising statewide awareness of damselflies and dragonflies, and invertebrates generally, the MDDS is designed to help our Department more accurately assess status of the state's rare, threatened and endangered odonates. To our knowledge, the MDDS is among the first completely government-sponsored dragonfly atlasing projects of its kind in North

America. In the midst of its fourth field season, progress to date has exceeded expectations and is summarized for the 1999-2001 volunteer field seasons. We conclude that state-initiated wildlife atlases are an effectivetool for both rapidly improving our state of knowledge and raising public awareness of specific non-game taxa.

Ball, George E. Department of Biological Sciences, University of Alberta, Edmonton, AL.

Ellie Groden, Frank Drummond, Jeff Garnas, and Shicai Yan. Dept. of Biological Sciences, University of Maine, Orono, ME.

We have initiated studies on an introduced ant, Myrmica rubra, which because of its aggressive, stinging behavior, has been causing problems in some coastal communities in Maine. M. rubra

nests and forages in people's yards and around buildings, where nest density can average more than one nest per m² in heavily infested areas. M. rubra is native to Europe where it is widely distributed from the British Isles to Scandinavia and Siberia and south to the Medittarean. It was first reported in the United States in a 1908 (Wheeler, J. Econ. Entomol. Vol. 1) at the Arnold Arboretum in Forest Hills, Massachusetts. Samples collected at the Arboretum in 2001 revealed that the ant is still present, though it has not been successful in spreading and achieving the densities observed in coastal areas of Maine. The first catalogued collection of the ant in Maine was in 1952 from Eastport (specimen housed at the Harvard Museum of Comparative Zoology). Surveys conducted in 2001 revealed isolated M. rubra infestations from the southern most coastal community of Kittery to the northeastern most community of Eastport. One non-coastal infestation in the state has been located in the town of Wayne, ME. Research questions being addressed include: 1) what is this ant's potential for further spread; 2) what impact is M. rubra having on native ants, other arthropods, and vertebrates; and 3) how can infestations be mangaged? Preliminary results of foraging studies indicate that M. rubra foragers are active throughout the day and night over a range of seasonal temperatures. Activity appears to decline during heavy rains, and around mid-day in open areas exposed to full sun.

Anthony O'Neal¹, Constance Stubbs¹, Francis Drummond¹, and Howard Ginsberg^2,3. ¹ University of Maine, Orono , ²Patuxent Wildlife Research Center and ³ University of Rhode Island.

Invasive plants are widely regarded as having a negative impact on native flora. This may occur for a variety of reasons. The invasive plants may shade out potential competitors, have a longer photosynthetic period, or excel at drawing trace nutrients from soils. One resource that has not been widely examined as a potential source of invasive plant success is the attraction of insect pollinators, especially the bees (Apoidea), to the plant. High bee visitation rates to these invasive plants may lead these important pollen transportation vectors away from native plants, thus affecting native plant reproductive success. Reduced native plant reproductive success may then give the invasive plant an even larger competitive advantage. The purpose of this two-year investigation is to examine the abundance and diversity of insect pollinators on co-flowering native and invasive plant pairs. It also includes an examination of floral rewards on the pollinator foraging behavior. These studies may provide further insight into why invasive plants could be more successful. This research is being conducted in Acadia National Park, Mount Desert Island, Maine, which has over 30 species of invasive plants that are of concern to the Park Resource Managers. We are examining three invasives that require insect pollinators. Because there are still many areas where these plants have not spread, we can compare bee behavior on native plants in the presence and absence of invasive plants. Results from the 2002 field season are presented with a focus on the early flowering invasive, Berberis thunbergii (Japanese barberry) and native Vaccinium angustifolium (lowbush blueberry). Visitation rates for both 1 min and 10 min observation periods on 10 experimental (invasive present) and 7 control (invasive absent) 2 m x 2 m plots indicate a significant reduction of bee visits to V. angustifolium when B. thunbergii is present, as opposed to when it is absent. Our preliminary investigation found that the proportions of bee visitors representing several genera differed between the plants. The proportion of bees in the genus Andrena was found to be higher in both the control and experimental plots of V. angustifolium. In contrast, visitors belonging to the genus Bombus (Apidae) were found in larger proportion on B. thunbergii. Other infrequent visitors to both V. angustifolium and B. thunbergii were Nomada, Osmia, and members of the family Halictidae. A second field season (2003) should help clarify whether competition for pollinators exists between B. thunbergii and V. angustifolium.

Karen Coluzzi. University of Maine, Orono, ME.

Ground-dwelling arthropods were sampled over one summer (2001) from three potato fields in Maine to determine if scavenging insects have a negative impact on the transmission of the fungal disease caused by Beauveria bassiana in Colorado Potato Beetle (Leptinotarsa decemlineata (Say)) populations. Past studies have indicated an increase in the potential for disease persistence through contact with infected individuals. However, the disappearance of infected sentinel cadavers was observed in the field and the presence of ground beetles (Coleoptera: Carabidae) had been noted in the vicinity. The three sites were chosen for their varying degrees of insect diversity, as well as their differences in abundance of the dominant ground beetle, Harpalus rufipes. Pitfall traps were used to assess the relative scavenger abundances. CPB cadavers were generated in the laboratory by topically spraying a conidial formulation of Beauveria bassiana onto 2^nd instar larvae. The cadavers were placed on plates that were randomly distributed in the three potato fields and monitored for disappearance on an hourly or daily basis. The proportion of cadaver decline adjusted to a 24-hour period was significantly higher in the potato field that contained the highest number of H. rufipes caught per trap per day (P < 0.03). The mean abundance per trap per day of H. rufipes was correlated with the proportion of cadaver decline, but was not found to be significant (P = 0.058).

Ellen Klinger. University of Maine; Orono, ME.

Behavior of newly emerged summer adult Colorado potato beetles (Leptinotarsa decemlineata) in relation to the presence of Beauveria bassiana was studied. Wooden boxes were constructed as to accommodate potato plants and to simulate field conditions. Laboratory-raised newly emerged adult beetles were placed in soil at the center of this arena and were allowed 30 minutes to colonize one of the four plants that had been placed in the box. Beetle movements were recorded using a grid consisting of 5x5cm squares. B. bassiana killed and sporulated adult beetles were then placed in varying patterns surrounding the release point so that one B. bassiana-free grid square remained. Similar measurements were taken for beetles in the presence of the fungus. Analysis of plant colonized showed no preference in control or B. bassiana treated groups for one cardinal point. There was no difference between B. bassiana treated and untreated beetles in pattern of plant colonization. There was no deviation from the expected frequency of interaction with B. bassiana squares. Time for plant colonization did not differ between control and B. bassiana beetles, nor did distance traveled. Relative humidity was a significant factor for both time to colonize and distance traveled to plant, with longer and lengthier travel times as the RH fell. The plant colonization behavior of newly emerged Colorado potato beetles does not seem altered by the presence of B. bassiana in the immediate environment.

Kathleen D. Murray. Maine Department of Agriculture, Food, and Rural Resources, Augusta, ME.

Farmers often cite the lack of demonstration projects as a barrier to adoption of integrated pest management (IPM) methods. This project was conducted to demonstrate the efficacy of augmentative releases of pteromalid parasitoids (Hymenoptera: Pteromalidae) for management of the house fly (Musca domestica) to Maine dairy producers. Using methods based on research from Cornell University, a mixture of two parasitoid species, Muscidifurax raptor and M. raptorellis were released weekly at rates of 1,000 wasps per calf and 200 wasps per cow for 12 weeks on each of four organic farms. Resulting fly parasitism rates were estimated by placing mesh bags of 'sentinel' house fly pupae in each barn. These sentinels were retrieved after one week and were then reared out in the laboratory to determine percentage parasitism. House fly activity was monitored weekly in each barn by counting the number of fly specks deposited each week on strips of masking tape placed on stalls, feeders, and barn walls.

We found parasitism rates were significantly increased in three out of four barns compared with that in the control barn in which no wasps were released. The greatest parasitism was achieved in individual calf hutches where average parasitism was 51%, reaching greater than 70% in late summer. Average parasitism in the control barn was 1%. Fly activity was very low in all but one barn.

Interviews with cooperating farmers indicated that all were pleased with the level of fly control achieved. Three of the four producers felt that fly control was better with the parasitoid releases than without. Project results were shared with dairy farmers in the region via presentations at dairy producer meetings, a Cooperative Extension Bulletin, and an article published in a regional farming magazine

Stephen Woods and Shelly Thomas. Department of Biological Sciences, University of Maine, Orono, ME.

A study was conducted to evaluate the role of downed woody material as habitat for insects. Thirty 1 m² enclosure cages were deployed at 3 recent clearcut and 3 selection harvest stands in the Penobscot Experimental Forest in Bradley Maine. With each stand, enclosures were placed around 1 hardwood, decay class 2 log, 1 hardwood, decay class 4 log, 1 softwood, decay class 2 log, 1 softwood, decay class 4 log, and 1 over bare ground. Two pitfall traps were placed in each and ground beetles collected every two weeks between May and November, 2000. Nine species with 907 individuals were recovered. Across both forest types, decay class 2 softwood logs yielded significantly and substantially fewer ground beetles than the other treatments which were not significantly different from each other. However all log means were lower than the bare ground suggesting that logs do not enhance habitat and may merely take up space or be detrimental. The ranking of the means were consistent with the hypothesis that the logs harbor compounds which are more common in softwoods and leach out or become volatilized over time and that these may serve as repellents to ground beetles.

1. ¹ "Nearly 200,000 acres of defoliation by this pest was mapped in 2001 in Hancock, Washington and York counties" (from newsletter Forest and Shade Tree - Insect and Disease Conditions for Maine, May 13, 2002 Issue of the Insect and Disease Laboratory, Maine Forest Service, Department of Conservation, State of Maine).