Jí^íS. United States fil X m Department of m Agriculture Budworms Forest Service Handbook Cooperative State Research Service Agriculture Handbook No. 644 Predators of the^ Spruce Budworm

i 864305

In 1977, the United States Depart- strategies for controlling the spruce ment of Agriculture and the Canadian budworms and managing budworm- Department of the Environment susceptible forests to help forest agreed to cooperate in an expanded managers attain their objectives in an and accelerated research and develop- economically and environmentally ment effort, the /United States acceptable manner. The work Spruce Budworms Program (C ANUS A), represented in this pubUcation was aimed at the spruce budworm in the wholly or partially funded by the East and the western spruce budworm Program. This manual is one in a in the West. The objective of series on the spruce budworm. CANUSA was to design and evaluate

Canada United States Spruce Budworms Program

November 1985 Contents

Introduction 1 Life History of the Spruce Budworm 2 Host Trees 2 Susceptibility of Life Stages to Prédation 3 Predator Groups 5 Overview of Predators 5 Predators of Eggs 6 Phalangids 6 Mites 6 8 Plant Bugs 10 Lacewings 10 10 11 Birds 12 Predators of Small Larvae 13 Phalangids 13 Spiders 14 Beetles 17 Ants 17 Miscellaneous 18 Birds 18 Predators of Large Larvae 19 Spiders 19 20 Beetles 21 Spruce Coneworms 24 Ants 25 Wasps 27 Fish 30 Mammals 30 Birds 31 Predators of Pupae 34 Spiders 34 Beetles 34 Spruce Coneworms 35 Budworm Larvae 37 Syrphids 37 Ants 37 Mammals 38 Birds 38 Predators of Adults 41 Spiders 41 Dragonflies 42 Beetles 45 Robber 45 Ants 46 Birds 47 Predators as Natural Regulators 48 Endemic Spruce Budworm Populations 49 Epidemic Spruce Budworm Populations 52 Methods for Studying Prédation 56 Exclusion Techniques 56 Direct Assessments 57 Serological Tests 57 Radioisotope Tagging 57 Direct Observations 58 Conservation and Enhancement of Predators 59 Habitat Requirements 59 Food Requirements 59 Nesting Requirements 60 Summary 61 Acknowledgments 62 Selected References and Guides for Identifying Predators 62 Phalangids 62 Spiders 62 Mites 63 Dragonflies 63 Carabid Beetles 63 Coneworms 63 Robber Flies 64 Ants 64 Wasps 64 Fish 64 Birds 64 Manmials 64 Literature Cited 65 Predators of the Spruce Budworm by Daniel T. Jennings and Hewlette S. Crawford, Jr.^

Introduction The spruce budworm, outbreaks. These natural agents of fumiferana (Clemens) (: control, also called natural enemies, ), is the most destructive include various diseases, parasites, forest pest in eastern North and predators. America. Millions of acres of spruce- forests have been damaged Enhancement of predators of the by the spruce budworm in eastern spruce budworm is a desirable feature Canada and in the North-central and of integrated pest management; Northeastern United States from Min- however, before we can incorporate nesota to Maine. Recorded outbreaks predator enhancement we need to of the budworm date back to the know which predators are important, early 1700's; but in the 20th century, how effective they are in regulating outbreaks are increasing in frequency, budworm populations, and what their extent, and severity (Biais 1983). habitat requirements are. Ultimately, such information and understanding The spruce budworm is a native, will lead to the identification and natural component of the spruce-fir development of forest-management forest. Managers need to know how strategies that enhance predators of to manage the insect to prevent or the spruce budworm. minimize damage to the forest. Applied control of the spruce bud- The goals of this handbook are to worm has been largely chemical. A (1) review available information on more integrated approach is needed predators of the spruce budworm; that includes safer, longer lasting, (2) provide information on predator less costly methods of pest manage- biologies, including life stages, modes ment. Such methods should be com- of attack, and general importance; patible with multiple uses of all forest (3) summarize the importance of resources, including fiber, wildlife predators and prédation in population and fish production, soil and water dynamics of the spruce budworm; conservation, and recreation. (4) review forest-management prac- tices that enhance predator popula- The spruce-fir forest has various tions; and (5) list basic literature natural agents that help keep bud- sources for identifying predators of worm populations in check between the spruce budworm.

'Daniel Jennings is a research forest ento- mologist with the USD A Forest Service, North- eastern Forest Experiment Station, Orono, ME. Hewlette Crawford, Jr., is a research wildlife biologist with the Station at Amherst, MA. Life History of the Spruce Budworm

The spruce budworm has four life two or more shoots together to form stages: egg, , , and adult. a feeding tunnel or shelter. During The adult stage is a with a this active feeding period, the larvae wingspan of about three-quarters of molt three more times to reach the an inch (2 cm). The female lay sixth by late June or early July. their eggs in masses on needles of The fifth and sixth cause most host trees in mid-July to early of the feeding damage to host trees. August. The number of eggs per Destruction of current shoots causes mass varies considerably but averages considerable larval movement, and about 20. The eggs hatch in about 8 many larvae drop from their host to 12 days, and the newly emerged trees and are exposed to ground- first-instar larvae do not feed but inhabiting predators. disperse and seek overwintering sites under bark scales and in old After development, the larva stops staminate flower bracts. After spin- feeding and transforms into a pupa. ning a silken hibernaculum, the larva The pupal stage lasts about 10 days, molts to the second instar, enters after which the moth emerges. The , and overwinters. female moth emits a that attracts the male for mating. In the spring of the following year, Moths live about 2 weeks and are the second-instar larvae emerge from found from early July to August. their hibernacula, disperse, and seek Shortly after mating, the female moth feeding sites. Newly opened staminate begins to lay eggs, completing the flowers are the preferred food on life cycle. There is only one genera- balsam fir, (L.) tion of the spruce budworm per year, Miller, during the initial feeding but the life cycle spans 2 calendar stage. When staminate flowers are years. not available and vegetative buds are "tight" and unexpanded, larvae mine Host Trees the previous year's needles. Usually, only one needle is mined and the Balsam fir is the principal host of the larva molts to the third instar within spruce budworm in eastern North or soon after leaving the needle. By America; it is the host that suffers the mid- to late May or early June, greatest damage and tree mortality depending on temperatures, most of (Miller 1963a). Outbreaks are most the larvae leave mined needles and likely when mature stands of balsam begin feeding on newly opened fir cover extensive areas (Biais 1983). vegetative buds (Miller 1963a). The budworm also attacks and feeds on white spruce, Feeding continues on the new foliage (Moench) Voss; red spruce, of developing shoots and on old P. rubens Sarg.; black spruce, P. foliage once the current foliage is mariana (Mill.) B.S.P.; and blue consumed. The larvae typically web spruce, P. pungens Engelm. During outbreaks there may be feeding on to the forest floor, where ants, tamarack, Larix lancina (Du Roi) K. carabid beetles, spiders, and small Koch; eastern white pine, Pinus mammals often are numerous. strobus L.; and eastern hemlock, Tsuga canadensis (L.) Carr. Mature Conversely, the inability to move and and overmature trees are most escape predators increases the suscep- susceptible to attack, though trees of tibility of pupae to prédation. Pupae all sizes can be damaged or killed generally are found near the apexes during an epidemic. of branches in silken shelters con- structed by the larvae before pupa- Susceptibility of Life Stages tion. Although capable of wiggling to Prédation and squirming, pupae usually remain attached to these silken threads by All life stages of the spruce budworm means of cremaster hooks. are susceptible to attack by predators. However, some stages are more sus- Spruce budworm moths may escape ceptible than others. For example, some predators by taking flight; during both larval dispersal periods, however, flying moths are exposed to the small larvae are exposed and sub- aerial-searching predators such as ject to attack by arboreal and epigeal dragonflies, robber flies, and birds. predators. Once needle or bud mining There are temporal and sexual dif- begins, the larvae are protected and ferences in susceptibilities of moths to are less likely to be eaten by pred- prédation. Male spruce budworm ators. Overwintering larvae in hiber- moths are more active, rapid fliers nacula are protected but not immune than female moths. Female moths to prédation. must lay some of their eggs before engaging in long-distance flights. Many predators rely on movements to Moths are subject to prédation by detect potential prey. Because egg foliage-searching predators such as masses are deposited on needles and birds and spiders. During coitus, both are relatively stable, the egg masses sexes are susceptible and vulnerable are not likely to be detected except to attack because they represent a by foliage-searching predators such as relatively large target and their coccinellid beetles and birds. movements are somewhat impaired.

Large larvae of the spruce budworm Habitats occupied by the different life may escape foliage-searching stages of the spruce budworm and predators by dropping from their host searching behaviors of potential trees. This dropping behavior may predators also influence prey suscep- have survival value; but dropping tibility. Most spruce budworm eggs also exposes the larvae to other are deposited on peripheral shoots in predators (fig. 1), especially when the the upper crowns of host trees (Miller larvae drop to nonhost vegetation or 1958, 1963a). Such sites are exposed Figure 1—Cape May warbler and dropping spruce budworm larva. to predators that concentrate their Vertebrate predators of the spruce searches near branch apexes. After budworm include birds, mammals, egg hatch, the young larvae disperse, and fishes. Birds are the best-known some falling to lower crown levels, and probably most important where other predators are found. The predators of the spruce budworm. small larvae also disperse to nearby Predaceous mammals include the trees, nonhost vegetation, and the rodents (Rodentia) such as voles and forest floor. During dispersal the lar- the (Insectívora) such as vae are exposed to numerous arboreal shrews. Vertebrates have a high and epigeal predators. attack potential because they are warm blooded and have relatively Similarly, large larvae dropping from high metabolic rates that require tree crowns are susceptible to preda- ample food supplies. tors. Large larvae falling to interven- ing vegetation or the forest floor are Overview of Predators especially vulnerable to attack by ants and spiders. In this section we review what is generally known about predators of Starvation also contributes to move- the spruce budworm. This review is ment of large larvae, and movement restricted largely to predators of both within and between tree crowns Choristoneura fumiferana, though exposes the larvae to a variety of reference is made to the jack pine predators. budworm (C pinus Freeman) and the western spruce budworm (C occiden- Although all life stages of the spruce talis Freeman). bud worm are susceptible to prédation, susceptibilities vary according to life We have taken the life-table stage, location, and behavior of both approach, presenting what is known predator and prey. about predators of each prey life stage, beginning with the egg stage Predator Groups and progressing to the adult or moth stage. This approach introduces some Predators of the spruce budworm overlap among the predators because include both and some predator groups (birds, spiders) vertebrates. The largest guilds of prey on all life stages of the spruce predators are found budworm. And within each predator among the insects and spiders. With group, some species probably are few exceptions, most of the insect more important than others as preda- orders (Coleóptera, , tors. Predators of large larvae and Diptera, etc.) contain predaceous pupae of the spruce budworm are species; all spiders are predaceous. most important in terms of generation survival (Morris 1963). Predators of Eggs

Despite an extensive review of the Predators of spruce budworm eggs spruce budworm literature and efforts include phalangids, mites, spiders, to collate unpublished information plant bugs, lacewings, beetles, ants, from colleagues, our knowledge and and birds. understanding of this complex preda- tor-prey system remains incomplete. Phalangids Certainly, not all of the predators and potential predators of the spruce bud- Little is known about these predatory worm have been identified or studied. and their food habits. Some For example, virtually nothing is species are scavengers and feed on known about the reptiles and amphi- dead insects (Todd 1950); others are bians that possibly attack and feed on predaceous and feed on lepidopterous spruce budworms. And nocturnal- larvae, pupae, and adults (Bishop flying bats are potential predators of 1949, Edgar 1971). Presumably, they spruce budworm moths, but they also attack and feed on lepidopterous remain unstudied in northeastern eggs, including eggs of the spruce spruce-fir forests. Perhaps the scar- budworm. Neilson (1963) recognized city of information about some only two groups (mites and insects) predator groups will help stimulate as predators of budworm eggs; Varty future investigations. and Titus (1974) included phalangids among the that influence budworm abundance in the egg stage. However, virtually nothing is known about their specific egg-feeding habits.

In Maine, significantly more indi- viduals and species of phalangids were trapped in uncut residual strips and in dense spruce-fir stands than in clearcut strips (Jennings et al. 1984). Peaks in seasonal activities of phalan- gids coincided with the egg and early-larval stages of the spruce bud- worm. Because budworm eggs are relatively small and immobile, we suspect they are susceptible to préda- tion by phalangids.

Mites

Mites (Arachnida: Acari) are among the most abundant arthropods in ter- restriai ecosystems. They rival insects average interval between feedings of in numbers of species and habitats 2.3 days. Individual mites consumed occupied (Borror et al. 1976). Many as many as five eggs in 1 day, and are parasitic; others are predaceous, feeding generally was irregular. phytophagous, or scavengers. They are of considerable biological interest Serological tests for detecting préda- and importance (Treat 1975). tion in natural populations showed that 22 to 24 percent of the mite Both parasitic and predaceous mites population sampled in the Green attack the spruce bud worm. Parasitic River area had fed on spruce bud- mites infesting spruce budworm worm eggs. Loughton et al. (1963) moths may indirectly affect budworm concluded that this level of mite egg production by reducing fecundity prédation was of considerable sig- (Houseweart et al. 1980). nificance because mites commonly fed on two or more eggs. The sero- Most of our knowledge about logical test detected egg proteins in predaceous mites concerns mites mites for only 24 hours after feeding; feeding on budworm eggs. Bennett thus, total prédation by mites may (1952a) reported that Anystis agilis have been underestimated. was observed preying on budworm eggs in . Neilson In New Brunswick, Varty (1977) (1963) noted that mite populations observed predaceous mites feeding on were estimated during foliage exami- budworm eggs and on aphids, scales, nations for budworm eggs; the most collembola, and other mites. Popula- common species was A. agilis Banks. tions of mites were very high in August and were estimated at Morris (1963) reported that egg 404,695 per acre (1 million/ha). He prédation of the spruce budworm was noted that at least eight species of caused largely by mites (species predaceous mites (mostly trom- undetermined). Red mites frequently bidiform taxa) are found on balsam were observed feeding on eggs in the fir in New Brunswick. Green River area of New Brunswick, and mites were extremely abundant In July 1979, David M. Kendall of during the budworm's egg-laying our staff observed red mites feeding period. on green, unhatched eggs of the spruce budworm in Washington These observations of mite prédation County, ME. The eggs were laid on on budworm eggs prompted further small understory balsam fir. The investigations by Loughton et al. mites were later determined to be an (1963). "Known-feeding" tests with undescribed species of Balaustium red mites in the laboratory showed (Family Erythraeidae). We have that mites consumed an average of subsequently made additional collec- 0.68 budworm egg per day, with an tions of Balaustium mites from balsam fir foliage examined for Spiders generally feed on mobile spruce budworm Qgg masses. prey, chiefly insects. They use two principal methods of prey capture: Mites (fig. 2) show both numerical web building and hunting. Both and functional responses to increases methods involve visual and tactical in budworm prey density. On the cues triggered by mobile, flying, or Green River Watershed, Neilson walking insects. Few spiders have (1963) noted that mite populations been observed feeding on immobile showed numerical responses to egg prey and even fewer on insect eggs. densities over the entire range of den- sities studied. He concluded that these Morris (1948) reported that several numerical responses may partially ex- species of spiders found on mature plain the increase in percent prédation balsam fir trees in New Brunswick observed with increasing egg density prey on both larvae and eggs of the up to 9.3 eggs per ft^ (100 eggs/m^) spruce budworm. The spiders were of foliage. Mites also showed func- not idenfified. Both Neilson (1963) tional responses, i.e., numbers of and Loughton et al. (1963) concluded eggs consumed per mite increased that spiders usually do not attack with egg density, especially at prey immobile prey because in cage densities below 9.3 eggs per ft^ of experiments spiders could not be foliage. induced by starvation to prey on bud- worm eggs. Hence, spiders were not Spiders tested with antiegg sera to determine predators of budworm eggs. Spiders (Arachnida: Araneae) are among the most abundant predaceous In 1977, a jumping , Metaphi- arthropods in northeastern spruce-fir dippus flavipedes (G. & E. forests. Morris (1963) estimated Peckham), was observed feeding on a populations of 75,000 spiders per green, uneclosed egg mass of the acre (185,325/ha), not including spruce budworm (Jennings and species that restrict their hunting to Houseweart 1978). The egg mass was the ground or to low vegetation. This deposited on foliage of a young estimate probably is conservative under story balsam fir. The spider was because some of the more active disturbed and the feeding was inter- forms escape during collection of rupted; however, once the eggs foliage samples. Varty (1980) hatched, the spider readily captured reported average spider densities of and fed on the first instars. 0.4 to 0.8 individual per ft^ (5 to 10 individuals/m^) of branch surface; The extent of spider prédation on occasionally, there were as many as budworm eggs is not known; we 242,817 per acre (600,000/ha) of suspect that it is minor. balsam fir. Figure 2—Red mite, Balaustium sp., preying on spruce budworm eggs. Plant Bugs spruce budworm. Most likely these were brown lacewings (Family Plant or leaf bugs (Insecta: Hemerobiidae) or green lacewings Hemiptera) belong to the family (Family Chrysopidae); the larvae of Miridae, the largest group of true both families are predaceous. Brown bugs. Most mirids feed on the juices lacewings generally are found in of plants, but a few prey on other wooded areas; green lacewings are insects. common in grass and weeds and on the foliage of trees and shrubs During investigations of budworm (Borror et al. 1976). predators on the Green River Water- shed, New Brunswick, mirids were Virtually nothing is known about the recognized as possible predators of importance of neuropteran prédation budworm eggs (Dominion Department on spruce budworm eggs. The of Agriculture 1950). However, little Neuroptera associated with spruce-fir is known about the species associated trees have scarcely been studied. with northeastern spruce-fir forests and their importance as predators of Beetles spruce budworm eggs. Presumably, the prey is sucked dry, leaving little Beetles (Insecta: Coleóptera) are the evidence of predator activity. largest order of insects; over a quarter million species have been Lacewings described (Borror et al. 1976). These insects are found in almost every Lace wings (Insecta: Neuroptera) are kind of available habitat. Their small, soft-bodied insects with four feeding habits are varied; many are membranous wings. The wings phytophagous, many are predaceous, usually have a great number of cross some are scavengers, and others feed veins and extra branches of the on mold and fungi. The longitudinal veins; hence, the order families Carabidae, Staphylinidae, name (Borror et al. 1976). The front and contain members and hind wings are held rooflike over that are predaceous, both as larvae the body. and adults.

Larval and adult neuropterans live in At least two species of ladybird a variety of habitats, and many are beetles (Family Coccinellidae) are predaceous. The adults are weak known to feed on eggs of the spruce fliers, but most are predaceous and budworm. In cage experiments, adults feed on relatively weak prey. of the ladybeetle Mulsantina hudsonica (Casey) readily accepted Neilson (1963) reported that larvae of budworm eggs as prey (Varty 1969). Neuroptera (species undetermined) This small brown coccinellid is the had been observed eating eggs of the most abundant ladybeetle on balsam

10 fir in New Brunswick. It feeds main- species. In addition to M hudsonica, ly on the balsam twig aphid, Varty (1969) listed the following six Mindarus ahietinus Koch, but eggs species of coccinellids found on and early-instar larvae of the spruce balsam fir in New Brunswick: budworm are subject to prédation. mali (Say), Adalia bipunctata (L.), A. frigida (Sehn.), Coccinella montícola Adults of Anatis mali (Say) have been Mulsant., Chilocorus stígma (Say), observed in Maine feeding singly and and Coccinella transversoguttata in groups of three to four on newly Falderman. Little is known about the deposited egg masses of the spruce food habits of most of these budworm. Field and laboratory coccinellids. studies indicate that the life history of A. mali is well synchronized with the Ants ovipositional period of the spruce budworm. 2 Egg masses of the spruce Ants (Hymenoptera: Formicidae) are budworm are deposited over a period the most widely distributed of the of about 27 days beginning in late social insects (Francoeur 1979); often June or early July (Houseweart et al. they are locally abundant in forest 1982). The last 2 weeks of the coc- habitats, including northeastern cinellid's larval period coincides with spruce-fir forests that are infested the beginning of the budworm's tgg with the spruce budworm.^ ac- period. Newly emerged A. mali tivity begins early in the spring and adults are present for about 3 weeks continues into midfall. Their feeding of the spruce budworm's egg-mass habits are diverse: some species are period. carnivorous; others are herbivorous. Many feed on sap, nectar, and During investigations of the spruce honey dew secretions; and many are budworm on the Green River Water- able to switch their diets to take ad- shed in New Brunswick, Neilson vantage of abundant foods, such as (1963) noted that larvae of Coc- lepidopterous defoliators. Because cinellidae (species undetermined) ants prey on numerous species of were observed feeding on budworm forest pests, they are considered eggs (fig. 3). potentially important as agents of population mortality. Little is known about the predatory beetle fauna associated with north- Although we have no direct observa- eastern spruce-fir forests. Few tions of ants preying on spruce bud- studies have dealt with the beetles worm eggs, we strongly suspect that found on various budworm host-tree ^Jennings, Daniel T.; Houseweart, Mark W.; ^Lawrence, Robert K.; Houseweart, Mark W.; Francoeur, Andre. Ants (Hymenoptera: For- Jennings, Daniel T. Anatis mali (Say), a coc- micidae) associated with strip clearcut and dense cinellid predator of spruce budworm egg spruce-fir forests of Maine. [Unpublished masses. [Unpublished manuscript.] manuscript.]

11 Figure 3—Coccinellid beetle, Anatis mali, prey- ing on spruce budworm eggs.

budworm eggs are subject to ant and species that forage in nearby tree prédation. Finnegan (1974) noted that crowns. prédation by ants was not limited to a particular prey life stage; ants prey Birds on eggs, larvae, pupae, cocoons, and adults. Varty and Titus (1974, p. 17) All life stages of the spruce budworm included ants among the arthropods are subject to prédation by birds; that "exercise a light restraint on however, little is known about birds budworm abundance in the egg and preying on budworm eggs. George small-larval instars. ..." However, and Mitchell (1948) reported that the they gave no quantitative data. larvae, pupae, adults, and eggs of the spruce budworm provide excellent In northern Maine, peaks in pitfall food for insectivorous birds but gave catches of ants coincided with the egg no specific information about species stage of the spruce budworm, par- preying on eggs. Neilson (1963) con- ticularly in dense spruce-fir stands.^ cluded that budworm egg masses are Captured ants included individuals too small to constitute worthwhile

12 Predators of Small Larvae

prey for birds, and Morris (1963) Predators of small larvae (L1-L2) of surmised that wood warblers probably the spruce budworm include do not prey on budworm eggs since phalangids, spiders, beetles, ants, the warblers did not accept small lar- miscellaneous insects, and birds. vae as prey. Phalangids During our bird studies in Maine, we collected an adult male pine siskin, Varty and Titus (1974) included Carduelis pinus (Wilson), that had phalangids among the arthropods that fed on egg masses of the spruce bud- exercised a light restraint on bud- worm (Jennings and Crawford 1983). worm abundance in the egg and Only new, unhatched egg masses small-larval stages. However, they were eaten by the pine siskin, and in- concluded that phalangids have vir- dividual eggs destroyed by one bird tually no importance in the survival totaled 2,162. Apparently the bird of large larvae, pupae, or adult bud- stripped the egg masses from host- worms. No doubt, this conclusion is tree needles because there were no based on the general inability of needles or fragments of needles in the phalangids to subdue large, active crop or gizzard. prey.

Egg prédation by birds is one source A determination of the phalangid of mortality not recognized or ac- fauna associated with northeastern counted for in conventional life tables spruce-fir forests has received more (Morris 1963). Such mortality is dif- attention than their predatory roles. ficult to detect and evaluate because Carter and Brown (1973) reported six the egg masses and eggs are removed species—CööWö agilis Banks, and consequently are "missing" from Sabacon crassipalpe (L. Koch), the population. Odiellus pictus (Wood), Leiobunum calcar (Wood), L. bicolor (Wood), and L. ventricosum (Wood)—from pitfall traps in a mature red spruce stand in New Brunswick. Five of these species, all except L. bicolor, were collected in budworm-infested spruce-fir stands of New Brunswick by Varty and Carter (1974). Because of misidentification and nomenclatural change, S. crassipalpe is no doubt S. cavicolens (Packard) and L. bicolor is L. elegans (Weed).

Five genera and at least seven species of phalangids were collected by pitfall

13 traps in strip-clearcut and dense spruce budworm in Maine. He spruce-fir forests of northern Maine observed spiders preying on budworm (Jennings et al. 1984). More than 90 larvae, both in the laboratory and in percent of the specimens were the field, shortly after the larvae Leiobunum calcar (Wood). The emerged from egg masses. Five phalangids generally preferred the species of spiders, all web spinners, more closed, shaded habitats of dense were collected from spruce foliage on spruce-fir stands and of uncut which egg masses of the spruce bud- residual strips to the more open, worm were abundant. He concluded cleared habitats of cut strips. that **two spiders . . . were quite Significantly more individuals and capable of exterminating the several species were trapped in uncut residual hundred newly hatched little larvae strips and in dense spruce-fir stands which emerged from the dozen or than in clearcut strips. For both study more egg masses," placed on a caged years, phalangids were most abundant balsam fir in the laboratory and mean catches per pitfall trap (Johannsen 1913, p. 24). were greatest during the egg and first-instar periods of the spruce bud- Tothill (1923) included spiders among worm. The egg and early-larval in- the "checks" causing mortality to stars probably are the stages most budworm progeny. In New susceptible to attack and prédation by Brunswick in 1918, he estimated that phalangids. Strip clearcutting con- of the 150 eggs laid by the typical tributes to dispersal losses of early- budworm, 8 of the resulting larvae instar larvae (Jennings et al. 1983), (5.3 percent) would be eaten by exposing the larvae to numerous spiders, but he gave no indication of predators including phalangids the species or prey larval size. (fig. 4). During investigations on the Green Special techniques are needed to River Watershed, New Brunswick, determine the predator-prey relation- F. C. Hirtle observed several species ships involving phalangids. Loughton of spiders preying on both budworm et al. (1963) included phalangids larvae and eggs (Morris 1948). The among the predators that could be species were not identified, but assessed serologically for their préda- populations averaged 20 per mature tion on spruce bud worm. balsam fir tree.

Spiders In New York, Jaynes and Speers (1949) placed 1,100 first-instar larvae Spiders are opportunistic predators of the spruce budworm on a 6-ft that feed on a variety of prey, (1.8-m) balsam fir. They commonly including lepidopterous larvae. observed spiders seizing larvae as the Johannsen (1913) first noted that larvae spun down from one branch to spiders prey on first instars of the another; however, no exact count of

14 Figure 4—Phalangid feeding on first-instar spruce budworm.

mortality was made. The spiders Hirtle (1951) concluded that spiders were not captured and identified. probably are important predators of the budworm during the early-larval Conversely, Miller (1958) concluded instars. Mott (1963) noted that préda- that very few early instars fall prey tion by spiders takes place during to spiders or to other predators. dispersal of first and second instars Apparently, this conclusion was based but that little is known about its on earlier observations (Bennett importance. In New Brunswick, 1952b) where spiders were confined Morris (1963) noted that spiders in jars or cages and potential prey generally were active from early larvae were introduced. Only May, before larval emergence from sp. was observed prey- hibernacula, until early November. ing on a larva. In some instances, the Spiders were abundant and active budworm larvae had spun webbing during both egg hatch and dispersal that the spiders did not disturb. of first instars (Morris) 1963.

15 Loughton et al. (1963) gives some second-instar larvae. They noted that indication of the magnitude of spider G. angusta is well adapted to take prédation on small bud worm larvae. advantage of prey that are extremely Using serological techniques and abundant for only a short time, e.g., antisera prepared against first and the situation during larval dispersal. second instars, they found that 13 and The spider is very resistant to 7 percent of the spiders collected dur- starvation. ing the springs of 1959 and 1960 had fed on spruce budworm larvae. These Renault and Miller (1972) designed percentages were based on field col- and conducted field experiments to lections made between larval assess the predatory behavior of emergence and needle mining; préda- Dictyna phylax Gertsch and Ivie, a tion dropped as expected during small web-building spider, on emerg- needle mining. With the appearance ing spruce budworm larvae in the of first-instar larvae in July, spider spring. In these experiments, second- prédation was estimated to be at instar larvae were "planted" on about the same level as before pupa- branches of balsam fir trees, some tion, i.e., in 1959, 26 percent of the with D. phylax juveniles and adults spiders had fed on early-instar larvae; and some spider-free. Results over a similarly, 25 percent had fed on early 3-year period showed that 60 percent instars in 1960. of the larvae survived on the control foliage, whereas only 3 percent These investigators noted that pro- survived on foliage with a spider teins (antigens) of first- and second- predator. The authors concluded that instar larvae were detectable for only D. phylax is extremely efficient in 1 day after feeding by spiders; hence, capturing small spruce budworm estimates of field prédation are con- larvae that are attempting to establish servative. Small larvae of the spruce feeding sites at the tips of branches. budworm were consumed at more frequent intervals than large larvae, Laboratory feeding tests showed that and more small larvae were eaten. partially starved D. phylax females Confined species offered 10 first- consumed an average of 15 second- instar larvae averaged 7.5 to 9 larvae instar larvae in a 6-hour period consumed per day over a 4-day before changes were noted in attack period (Loughton et al. 1963). response, handling time, and utiliza- tion of prey (Renault and Miller In laboratory studies on the predatory 1972). The web of this dictynid behavior of Grammonota angusta spider is spun near the periphery of Dondale, a species frequently found spruce and fir branches, ideal sites on foliage in New Brunswick, for capturing migrating first- and Haynes and Sisojevic (1966) found second-instar spruce budworm larvae. that the spider attack rate was propor- However, the authors concluded that tional to prey density up to eight the probability of a predator-prey en-

16 counter was extremely low at the first instars in July. Strip clear- endemic bud worm densities. cutting contributes to dispersal losses of these early-stage larvae (Jennings In northern Maine, the jumping et al. 1983) by exposing the larvae to spider M. flavipedes was observed numerous predators, including carabid capturing and feeding on first-instar beedes. Significantly more carabid larvae of the budworm after the lar- beetles are found in uncut residual vae emerged from eggs (Jennings and strips than in clearcut strips or dense Houseweart 1978). stands."^

Beetles Ants

Our knowledge of beetles preying on Our knowledge of ants preying on early instars of the spruce budworm small larvae of the spruce budworm is very limited. Because the larvae is limited. Most observations concern are small, mobile, and somewhat prédation on large larvae, pupae, and secretive, observations of predators adults of the budworm. However, feeding on them are rare. During Varty and Titus (1974) included ants dispersal, which occurs in both first among the arthropods that prey on and second instars, larvae are subject the small-larval instars. The authors to prédation by arboreal and epigeal concluded that predatory arthropods predators, including predaceous exercised a light restraint on bud- beetles. worm abundance in the egg and small-larval stages. Varty (1969) noted that hungry adults of the coccinellid beetle M hudsonica Although not fully studied and quan- Casey may prey on second-instar tified, young spruce budworm larvae budworms when the budworms leave are susceptible to ant prédation dur- their hibernacula in early May. He ing the first- and second-instar disper- concluded that if a large population sal periods. Numerous larvae are lost of hungry ladybeetles occupies the during both dispersals (Morris and same habitat as the migrating bud- Mott 1963). They spin down from worm larvae, there is a prospect for host-tree crowns and alight on significant pest mortality. intervening surfaces, including nonhost vegetation and the forest In northern Maine, seasonal activity floor. During these active, mobile of carabid beetles (including periods, the larvae are exposed to predaceous species) was greatest dur- numerous predators, including ants. ing the early- and late-larval stages of the spruce budworm."^ Activity generally declined as the summer '•Jennings, Daniel T.; Houseweart, Mark W.; Dunn, Gary A. Carabid beeties (Coleóptera: Carabidae) progressed, but carabid beetles were associated with strip clearcut and dense spruce-fir abundant during spring dispersal of forests of Maine. [Unpublished manuscript]

17 Finnegan (1978) noted that because predaceous insects that feed externally young larvae are concealed in foliage, on budworm larvae in their hiber- prédation by the introduced red wood nacula. They refer to a few ant, Formica lugubris Zett., had been Tetrastichus sp. (Hymenoptera: light on the first three instars of the Eupelmidae) and several cecidomyid spruce bud worm. Apparently these individuals (species undetermined) observations were made after larval feeding on larvae placed on small dispersal, when the young larvae had balsam fir trees. Cecidomyids are gall established feeding sites in old midges or gall gnats (Diptera: needles and new, expanding buds. Cecidomyiidae). Larvae of about two- thirds of the more than 1,200 North In northern Maine, ants generally American species cause galls on were active during most of the bud- plants (Borror et al. 1976); and a few worm's developmental stages.^ For species are predaceous on aphids, both study years, ants were especially scale insects, and other small insects. active during dispersal of the first in- Varty (1977) reported that the stars in July. Strip clearcutting in- cecidomyiid larva Lestodiplosis sp. is creases nonhost vegetation and con- a predator of small insects, including tributes to dispersal losses of these small spruce budworms and aphids. early-stage larvae (Jennings et al. 1983) by exposing the larvae to ants. Undoubtedly there are many other predaceous insects that feed on small Additional studies are needed to larvae of the spruce budworm. evaluate the importance of ants as Predators of overwintering larvae predators of small larvae of the (Miller 1958) and of dispersing larvae spruce bud worm. Finnegan (1974) have received little attention (Mott noted that the period of ant activity 1963). was long in , extending from mid-April to mid-October. This Birds period spans both dispersal periods of the early instars. No doubt, ants Small larvae of the spruce budworm prey on first instars before the larvae generally are not considered impor- spin hibernacula for overwintering, tant food for birds. Cheshire (1959) and again the following spring before reported that birds do not feed on the second instars establish feeding small larvae. However, Miller (1958) sites. Because both predator and prey included prédation by chickadees and are small, special techniques are nuthatches among the possible factors needed for study and evaluation. responsible for losses of over- wintering larvae in hibernacula. Such Miscellaneous Insects prédation destroys all trace of the hibernaculum. Dowden et al. (1950) noted that C. F. Speer discovered two

18 Predators of Large Larvae

Early-instar larvae are susceptible to Predators of large larvae (L4-L6) of prédation during both larval dispersal the spruce budworm include spiders, periods, i.e., in the summer after egg dragonflies, beetles, spruce cone- hatch (when the first instars seek worms, ants, wasps, fish, mammals, overwintering sites) and again in the and birds. following spring (when second instars seek feeding sites). Mott (1963) Spiders indicated that prédation during these dispersals probably was due to Large larvae of the spruce budworm spiders and predaceous insects, not are susceptible to prédation by birds. Kendeigh (1947) considered spiders (Arachnida: Araneae). Larvae budworm larvae as important food fall prey to both foliage-searching and items for birds only after the larvae web-spinning spiders. Most prédation reached an appreciable size (about on large larvae probably occurs when one-quarter to one-half inch [0.6 to the larvae leave their feeding shelters 1.3 cm]). Morris et al. (1958) also in search of food. In reference to indicated that the budworm does not dropping larvae of the western spruce become attractive to most species of budworm, TumbuU (1956) noted that birds until the fourth instar is larvae sometimes were intercepted by reached. spider webbing. An ensnared larva usually is subdued quickly by the host During population dynamics studies spider. on the Green River Watershed, New Brunswick, data were collected on Watt (1963) estimated about a bird consumption of spruce bud- threefold increase in spider density as worms (Mook 1963). Of the total the spruce budworm increased from 2 bud worms eaten, fewer than 1 per- to 180 larvae per 10 ft^ (1.9 to cent were in the fourth instar or 168/m2) of foliage during the 1949 to smaller. 1959 budworm outbreak on the Green River Watershed in New Brunswick. In Maine, we collected two female Despite these increases, he concluded downy woodpeckers, Picoides that spiders and other insects show pubescens (L.), and one black-capped essentially no numerical response to chickadee. Parus atricapillus L., (sex budworm numbers. He estimated that undetermined) in mid-March before a thirtyfold increase in attack rate of second instars emerged from hiber- individual predators would be nacula. However, none of these birds required to suppress population had fed on spruce bud worms. growth of the spruce budworm. However, at low larval densities, only 0.46 larva per 10 ft^ (0.42/m2) of foliage would have to be eaten by predators to account for a decrease in budworm survival rates.

19 Loughton et al. (1963) noted a func- Additional studies are needed to tional response of spiders to fluctua- determine the important species of tions in budworm populations on the spiders preying on large larvae of the Green River Watershed in New spruce budworm in spruce-fir forests. Brunswick. Using serological tech- Because of their abundance, diversity, niques, they estimated that 20 percent and predatory capabilities, they are of the foliage-collected spiders gave undoubtedly important agents of bud- positive tests when the budworm worm mortality. Renault and Miller population was high; only 8 percent (1972) concluded that spiders might gave positive tests when the popula- play a significant role in determining tion was signiñcantly lower. During endemic densities of budworm late June to mid-July, when sixth populations between outbreaks but instars were present, 21 and 26 per- have little influence in regulating ex- cent of the spiders tested gave posi- plosive outbreaks. tive results. Antigens of large larvae were detectable for longer periods Dragonflies than antigens of small larvae, i.e., regularly for 4 or 5 days and com- Dragonflies (Insecta: Odonata) are monly for 6 days. Laboratory feeding voracious predators. Both the aquatic rates indicated that on average most nymphs and the terrestrial adults feed spiders will feed on a fourth-instar on a variety of prey, chiefly insects. budworm every third day. Most species are commonly associated with aquatic habitats; On the basis of percentages of ñeld- however, many are found in ter- coUected spiders giving positive restrial habitats, including spruce-fir serological tests, Loughton et al. forests. Adult dragonflies often patrol (1963) concluded that species of territories along forest roads and Theridiidae were the most effective trails, where they "hawk" flying predators. Both adult and immature insects. theridiids can attack and successfully subdue large larvae of the spruce The prey of adult Odonata include budworm. In Maine, we have several different orders of insects. observed Theridion pictum Clausen (1940) found that Diptera, (Walckenaer) preying on late-instar Lepidoptera, and Hymenoptera con- spruce budworms. Jumping spiders stituted the bulk of the prey. Bell and (Family Salticidae) also can attack Whitcomb (1961) reviewed the large larvae. Loughton et al. (1963) literature for dragonflies preying on concluded that the Salticidae should Lepidoptera. be considered important predators of the budworm at all stages of larval Apparently only one observation has development. been made of dragonflies feeding on larvae of the spruce budworm. Liscombe and Lejeune (1949)

20 reported that dragonflies (species terrestrial communities. Their abun- undetermined) were predators of dance in northeastern spruce-fir spruce budworm larvae in the Spruce forests infested with the spruce bud- Woods Forest Reserve of Manitoba. worm has been documented by Varty In 1948 they observed "hordes" of and Carter (1974) in New Brunswick, dragonflies preying on spruce bud- by Freitag et al. (1969) and Freitag worm larvae and concluded that and Poulter (1970) in , by dragonflies were responsible for Krall (1977) and Jennings and others'* much of the unknown mortality to in Maine, and by Reeves et al. spruce budworm. Unfortunately, the (1983) in New Hampshire. Although manner of prédation was not many species are arboreal, most described—we do not know if the studies concern the ground-inhabiting dragonflies captured dropping bud- fauna. worm larvae or actively picked the larvae from foliage. Carabid beetles are chiefly oppor- tunistic predators. Some species are In Maine we have observed both predatory and phytophagous; dragonflies patrolling and hawking others are strictly phytophagous spruce budworm moths near tree (Lindroth 1969, Johnson and crowns, but we have not observed Cameron 1969, Kulman 1974). them capturing budworm larvae. Predaceous species feed chiefly on Because mortality factors operating other insects, including large larvae during the late-larval stage often of the spruce budworm. influence generation survival (Watt 1963), additional observations of Using radioactive tagging techniques. these predators of large larvae are Krall (1977) identified the following needed. nine species that had fed directly on spruce budworm larvae or secondarily Beetles on other predators of spruce bud- worm larvae: adstrictus Most of our knowledge about beetle Eschz., P. coracinus (Newm.), P. predators of large budworm larvae adoxus (Say), P. rostratus (Newm.), concerns the carabids or ground P. scrutator Lee, P. pensylvanicus beetles (Coleóptera: Carabidae) and Lee, Synuchus impunctatus (Say), the coccinellids or ladybeetles (Col- Calathus ingratus Dej., and eóptera: Coccinellidae). Other beetle Sphaeroderus canadensis Chd. In all, families also contain predaceous 133 beetles were radioactive, or 16 species, but little is known about percent of the 824 carabid beeties their feeding habits in northeastern collected in pitfall traps. The authors spruce-fir forests. concluded that vulnerability of spruce budworm larvae to prédation by Carabid beetles are among the domi- carabid beetles is extremely great nant predatory arthropods in many once the larvae are on the ground. Of

21 the nine radioactive species, P. Calosoma frigidum Kirby, S. adstrictus apparently was the most canadensis Chd., S. lecontei Dej., important predator of spruce bud- Harpalus herbivagus Say, and H. worm larvae, with 24 and 23 percent pleuriticus Kirby. Both P. pen- of the total beetles of that species sylvanicus and P. decentis were abun- radioactive in -treated and dant in mid- and late June, when control plots, respectively (Krall large larvae of the budworm were 1977). present. Species with population peaks in mid-June were S. lecontei, Sanders and van Frankenhuyzen S. canadensis, and C frigidum. (1979) observed Calosoma frigidum Kirby eating late instars of the spruce Although synchrony of predator-prey budworm in two white spruce planta- activities is important, food habits tions near Sault Ste. Marie, ON. and feeding preferences also must be These beetles are also predaceous on considered. For example, both H. the forest tent caterpillar, herbivagus and H. pleuriticus are her- Malacosoma dis stria Hübner, and the bivorous; S. canadensis and S, saddled prominent, Heterocampa gut- lecontei are snail feeders (Reeves et tivitta (Walker). More beetles were al. 1983). Thus, the list of potentially seen on fully foliated trees than on important predators of late-instar bud- trees without foliage or with only worms narrows to three species: P. peripheral foliage. The observers pensylvanicus, P. decentis, and C estimated that as many as 40 beetles frigidum. Future studies should con- may inhabit a 49-ft (15-m) spruce. centrate on these species, particularly The authors concluded that these C. frigidum, which is a well-known beetles may have played an important predator of lepidopterous larvae role in reducing budworm populations (Gidaspow 1959). To our knowledge, in spruce plantations because of their C. sycophanta L., an exotic species in- size, numbers, and manner of search- troduced to combat the gypsy moth in ing current foliage (fig. 5). New England, has not been observed feeding on the spruce budworm. On the basis of five criteria—number of individuals, habitat preference, In northern Maine, 13 genera and 23 seasonal activity, size, and food species of carabid beetles were col- habit—Reeves et al. (1983) identified lected by pitfall trapping in a eight species of carabid beetles that bud worm-infested forest.'* Of the are potentially important predators of species caught, P. adstrictus and P. the spruce budworm in northern New decentis, both potential predators of Hampshire. However, based solely on spruce budworm, were the most seasonal activity, species that are abundant. For both study years, potential late-larval feeders on spruce seasonal activity of carabid beetles budworm were P. pensylvanicus was greatest during the early- and Lee, Platynus decentis Say, late-larval stages of the spruce bud-

22 /

Figure 5—, Calosoma fiigidum, with spruce budworm larva.

23 worm. Activity generally declined (MacDonald and Webb 1963). Popu- after budworm pupation and moth lations of other predaceous insects flight. C. frigidum, a known predator declined immediately following insec- of budworm larvae, was collected in ticide treatment. The reasons for residual stands of strip clearcuts and these apparent differential effects are in dense stands but not in clearcut not known; presumably, some natural strips. Apparently this carabid beetle enemies of coccinellid beetles were prefers habitats with little ground affected by insecticidal spraying. cover or with abundant humidity (Kulman 1974). Prédation by staphylinid beetles (Coleóptera: Staphylinidae) on spruce Coccinellids or ladybeetles are budworm larvae apparently has not predaceous on large larvae of the been observed; however, these beetles spruce budworm. Coccinellid larvae are abundant in northeastern spruce- (species undetermined) were included fir forests. Most species of among the known predators of spruce staphylinid or rove beetles are budworm larvae during investigations predaceous, and the larvae usually of the budworm's population are found in the same habitats as the dynamics on the Green River Water- adults (Borror et al. 1976). Un- shed, New Brunswick (Dominion doubtedly larvae of the spruce bud- Department of Agriculture 1950). worm are susceptible to prédation by Smith (1966) collected individuals of staphylinid beetles, but the bud- Anatis mali (Say) that had fed on lar- worm's active, often secretive habits vae of the spruce budworm. Collec- make direct observations of prédation tions of this species were made in difficult. June from fir and spruce in Ontario, presumably during the late-larval Spruce Coneworms stage of the budworm. Larvae of the spruce cone worm, In Maine, larvae of A. mali were Dioryctria reniculelloides Mutuura observed from early June to mid-July and Munroe (Lepidoptera: ), in synchrony with the late-larval are predaceous on large larvae of the stages of the spruce budworm. ^ spruce budworm. The spruce con- Although late instars of the predator eworm, also called "spruce foliage are more voracious than earlier worm" and the "spruce needle- instars, the extent of prédation on worm," previously was known as D. budworm larvae is unknown. reniculella Grote. Cone worm larvae feed on both foliage and cones of Interestingly, populations of coc- . Their habits were described cinellid beetles were invariably found by McKay (1943) and more recently at higher densities in areas sprayed by McLeod and Daviault (1963). for spruce budworm suppression than in unsprayed areas of New Brunswick

24 In a mixed infestation of both spruce pressure, i.e., when populations of budworm and "spruce foliage worm" spruce coneworms increase, popula- in the Spruce Woods Forest Reserve, tions of spruce budworm decrease Manitoba, Barker and Fyfe (1947) correspondingly (Liscombe and observed that considerable mortality Lejeune 1949, Warren 1954). Warren was inflicted on budworm larvae by (1954) also showed that foliage age larvae of the foliage worm. However, influences prédation. In a replicated prédation took place only when an laboratory experiment, many more insufñcient supply of spruce foliage larvae of the spruce budworm were was available to larvae of the foliage destroyed when only old foliage was worm. In a campsite experiment, provided than in the presence of new equal numbers of budworm and foliage. foliage worm larvae were placed in separate containers with varying MacKay (1943) noted that the spruce amounts of spruce foliage. Prédation foliage worm was predaceous on jack by the foliage worm was observed pine budworm in northeastern On- only in jars without foUage or with a tario. McLeod and Daviault (1963) scant supply, but not in jars contain- summarized records of D. reniculella ing an adequate supply of foliage. feeding on the spruce budworm and After 9 days of extreme food short- also noted that the young larvae are age and starvation, the only surviving occasional predators of a spruce nee- larvae were foliage worms. dle miner, Eucordylea piceaella (Kerfott) (now known as From 1946 to 1949, the annual drop piceaella [Kerfott]). in spruce budworm populations in the Spruce Woods Forest Reserve ranged Ants from 86 to 97 percent (Liscombe and Lejeune 1949). Most of this mortality Ants have long been recognized as occurred between the time of early- potential biological control agents of larval emergence and the pupal forest pests. They have been studied period. The authors concluded that for more than 60 years in Europe, prédation probably accounted for where elaborate techniques have been much of the mortality. They reported developed for collecting, rearing, and that the '* spruce foliage worm" and propagating ants used in forest-pest dragonflies were the most important control (Finnegan 1971). Only within predators of the spruce budworm. the past 20 to 30 years has much attention been devoted to the pros- Spruce coneworms not only prey on pects of using ants to control forest larvae of the spruce budworm but pests in . also compete for food and shelter (Liscombe and Lejeune 1949). Finnegan (1974) listed several Relative abundances of both species qualities possessed by predaceous red are possible indicators of predator wood ants, which are not commonly

25 found among other predators of forest Ants are among the most important pests: predators of large larvae of the 1. They can attain very high spruce budworm. Records of ants population densities. Red wood ants preying on various Choristoneura are not host dependent but change species, including C. fumiferana, their diet according to available prey. were summarized earlier (Jennings 2. Their foraging area or hunting 1971). One early record (Dominion ground covers all levels of the forest Department of Agriculture 1950) from the forest floor to the uppermost should be added; ants (species branches of tree crowns. undetermined) were observed carry- 3. Their period of activity is ing larvae of the spruce budworm very long, about 180 days in Quebec. from a canvas mat in the Green River Activity begins near the nest before Watershed, New Brunswick. More the last snow melts in the spring and recently, Finnegan (1978) and continues until the ground starts to McNeil et al. (1978) have shown that freeze in the fall. Activity generally the introduced red wood ant is an is continuous—24 hours a day— effective predator of the spruce bud- though reduced at night. worm in Quebec. Lab tests showed 4. The more desirable species that this species was highly aggres- are polygynous, i.e., there are many sive in searching for and attacking queens per nest. This feature assures fourth, fifth, and sixth instars of the a long life to the nest because old spruce budworm. After importation, queens are replaced continuously. release, and establishment of col- 5. Desirable species form col- onies, the seasonal predatory activity onial nests, i.e., the nests are not was observed; more than 95 percent isolated socially from neighboring of the prey were insects. Lepidoptera nests. This creates stability and per- were the major prey during two manence in the ant population over peaks of prédation, and ftilly 80 per- large areas. cent of the tortricid prey were spruce 6. Red wood ants do not limit bud worms. At peak prédation, an their prédation to particular prey life estimated 5,298 larvae were brought stage: they may attack eggs, larvae, to the nest per day. During the pupae, cocoons, or adults with equal 20-day period that large spruce bud- vigor. worm larvae were available, an 7. Hunting activity is regulated estimated 43,500 larvae were con- by the nest: individual ants hunt for sumed per nest (McNeil et al. 1978). the queens, their brood, and other workers, not strictly for themselves. The impact of this prédation by red 8. As prey populations increase, wood ants also was measured in ants specialize in hunting the most terms of defoliation. Finnegan (1977) abundant prey, i.e., functional estimated defoliation in 1974 and responses to increasing prey densities. 1975 at 30.0 and 42.8 percent where ants were present; this compares with

26 42.5 and 63.1 percent in neighboring Superfamily Vespoidea, which in- areas without red wood ants. cludes the familiar yellowjackets, hornets, paper wasps, and potter Finnegan (1978) concluded that red wasps. Some vespoids are social and wood ants can be an important con- build large papery nests, where the trol factor at endemic population queen wasp and workers rear levels or during the initial phase of a numerous young. Other vespoids are developing outbreak. McNeil et al. solitary: after mating, the adult (1978) believed that the species could female wasp constructs a nest in the play a role in an integrated control ground or in some natural cavity and program against the spruce bud worm. then provisions cells of the nest with Similarly, Campbell and Torgersen food for her offspring. Adult vespoid (1982) concluded that native wasps generally feed on sap or nec- predaceous ants may play an impor- tar, but their larvae are fed insect tant role in the population dynamics prey. of the western spruce budworm in Washington. Yellowjackets and hornets (Hymenoptera: Vespidae: Vespinae) Ants certainly deserve more attention along with paper wasps (Hymen- and study, particularly in the north- optera: Vespidae: Polistinae) often eastern United States and Canada. are considered pests because they Although Finnegan (1971) concluded sometimes interfere with man's activ- that none of the native species ities, and they possess a sting that showed promise as limiting agents of may cause a serious allergic reaction. forest pests in Quebec, he did in- However, these wasps are beneficial dicate that some species of insects that prey on numerous insect Camponotus and Formica showed pests. The beneficial aspects of these several desirable qualities. A species wasps remain largely unreported. of Camponotus, C. herculeanus (L.), Akre et al. (1980) noted that resear- is one of the most abundant ants in chers frequently have observed strip clearcut and dense spruce-fir yellowjackets preying on defoliators stands of northern Maine. ^ Further in forests, but little has been pub- studies are needed to determine its lished about this prédation and its predatory impact on large larvae of possible values. the spruce budworm. We found no published information Wasps on vespoid wasps preying on large larvae of the spruce budworm; Wasps (Insecta: Hymenoptera) are however, we suspect that such préda- beneficial insects. Many are parasites, tion may be common. During popula- others are predators of various insect tion dynamics studies on the Green pests, and some are pollinators. Most River Watershed, New Brunswick, predaceous wasps belong to the efforts were made to collect all

27 winged insects that visited four tigris (Saussure) {= A. adiahatus balsam fir trees infested with spruce adiahatus [Saussure]), and Rygchium budworai (Morris 1963). Only two leucomelas (Saussure) (= Euodynerus predatory wasps were collected dur- leucomelas [Saussure])—whose ing the 2-hour examination periods, nesting activities coincided with late and the investigators tentatively con- instars of the spruce budworm and cluded that winged insect predators associated spruce-fir defoliator com- were of minor significance compared plex in the Black Sturgeon Lake with spiders. But other observers (D. region, Ontario. Provisions of the Mullen, personal communication) first generation of these wasps, par- have witnessed "swarms" of hornets ticularly E. leucomelas, included the on infested fir and spruce trees, spruce budworm (fig. 6) and the jack apparently searching for and feeding pine budworm. He concluded that on late instars of the spruce solitary wasps may satisfactorily bud worm. sample endemic numbers of important prey species, particularly the spruce Most of our knowledge about budworm. Thus, the trap-nesting predaceous wasps preying on large technique may be a useful tool for larvae of the spruce budworm con- early detection of endemic budworm cerns the potter or eumenid wasps populations. (Hymenoptera: Eumenidae). These solitary wasps also are known as Larvae of the spruce budworm also trap-nesting wasps because they will were included in the diverse prey of accept and provision artificial nests R. leucomelas (= E. leucomelas constructed from blocks of wood with [Saussure]) and A. catskill predrilled borings. Eumenid wasps albophaleratus (Saussure) found in nest in natural cavities of stems, provisioned nests placed in the Black branches, and stumps (Krombein Sturgeon Lake region, Ontario (Fye 1967), or in small holes bored in 1965a). Individual cell data indicated blocks of wood. The foundress that hunting female wasps tend to female wasps construct mud- prey on a given species of tree or partitioned cells in these nests, and plant at one time. For example, not the cells are provisioned with only were larvae of the spruce bud- paralyzed lepidopterous larvae. The worm taken but also larvae of other paralyzed prey larvae serve as food defoliator species of white spruce. for the developing wasp larvae. Because of this consistency in sear- ching habit, Fye (1965) concluded Apparently Fye (1962) was the first that it may be possible to use the to observe and report on eumenid wasp's superior searching ability as a wasps preying on late instars of the technique for sampling populations of spruce budworm. He identified three particular prey species. species—Ancistrocerus catskill albophaleratus (Saussure), A. tigris

28 Figure 6—Trap-nesting wasp witii spruce bud- worm larval prey provisioned in nest.

Krombein et al. (1979) also listed C. leucomelas (Saussure)—that accepted fumiferana as prey of E. leucomelas and provisioned trap-nesting blocks leucomelas (Saussure) and A. catskill placed in a spruce-fir forest. The albophaleratus. wasps clearly preferred the more open habitats of strip clearcuts, which In northern Maine, Jennings and had abundant floral forage, to dense Houseweart (1984) found four species spruce-fir stands. Two species, A. of eumenids—Ancistrocerus adiabatus catskill and E. leucomelas, preyed on (Saussure), A. antilope (Panzer), A. late instars of the spruce budworm catskill (Saussure), and Euodynerus and on other lepidopterous defoliators

29 of northeastern hardwoods and soft- Mammals woods. Spruce budworm larvae accounted for 38 percent of the total Although mammals may reach greater prey observed in 1977 but only 3 population densities than birds, little percent of the total prey observed in attention has been devoted to deter- 1978. Apparently the wasps switched mining the species of mammals prey- to a more preferred or locally abun- ing on spruce bud worms. Morris dant prey the second year. (1963) indicated that the budworm was available to purely terrestrial Fish mammals only when populations were high, resulting in foliage depletion Because of habitat differences, fish and larvae dropping from host trees. are not strictly predators of spruce Some manmials, however, are ar- budworm. However, larvae dropping boreal. C. H. Buckner trapped into forest streams are susceptible to specimens of the deer mouse prédation or scavenging by fish. Lar- Peromyscus maniculatus abietorum val droppage may be natural or Bangs without difficulty in the induced by spraying of chemical crowns of mature balsam fir trees on (Hydorn et al. 1979), and the Green River Watershed, New feeding on larvae spinning out of Brunswick (Morris 1963). It was not streamside trees may be intense. known whether Peromyscus had Kingsbury and Kreutzweiser (1980) discovered an abundant supply of found an average of 26.8 spruce bud- budworm larvae in the crowns or worm larvae in a sample of 10 brook their presence was due to normal trout, Salvelinus fontinalis (Mitchell), foraging activity. collected from a stream in Temiscouata County, Quebec, on Earlier Morris et al. (1958) examined June 14, 1978. Budworm larvae con- possible numerical responses of small tributed over 60 percent of the mammals to changing populations of volume of food items in the trout spruce budworm on the Green River stomachs. Once the larvae fall into Watershed. They found only two the water, they are essentially lost to species that showed direct but weak the population, regardless of predator responses: the short-tailed shrew, activity. Blarina brevicauda (Say); and the rock vole, Microtis chrotorrhinus As a source of predator-induced mor- (Miller). Interestingly, both the deer tality, prédation by fish is indirect mouse P. maniculatus (Wagner) and and secondary to other more impor- the red-backed vole, Clethrionomys tant sources of budworm mortality. gapperi (Vigors), showed possible in- However, the predatory activities of verse responses to increasing bud- fish should be considered in evalu- worm populations. The authors at- ating the environmental impacts of in- tributed these declines in mammal secticides on nontarget organisms. populations to indirect causes, i.e.,

30 severe defoliation and tree mortality dicated a population density of about reduced the supply of balsam fir 0.4 squirrel per acre (1/ha). seed, which is the rodent's main source of winter food. In western Maine and northern New Hampshire, we found evidence that Otvos (1981) reported that three red squirrels prey on endemic popula- small mammals may feed on the tions of the spruce budworm. spruce budworm in Newfoundland: Stomach-content analyses of 31 the meadow vole, M. pennsylvanicus specimens showed that only 2 had (Ord); the masked shrew, Sorex eaten spruce budworms. einer eus Kerr; and the red squirrel, Tamiasciurus hudsonicus (Erxleben). Because of their attack potential, He noted that the first two species selected rodents, particularly the red probably feed only on larvae that squirrel, and insectivores merit fur- have dropped from severely defoli- ther investigation as predators of ated trees, whereas the squirrel also spruce budworm. may feed on budworms in tree crowns. Birds

Red squirrels were implicated as Birds are the best known and prob- possibly causing substantial reductions ably the most important predators of in budworm populations in northern large larvae of the spruce budworm. Maine (Dowden et al. 1953). They rival ants, spiders, carabid Stomach-content analyses of 24 red beetles, and predaceous wasps as squirrels collected in a budworm- budworm predators. More is known infested forest showed that spruce about birds preying on spruce bud- budworms accounted for 51 percent worms than any other predatory of their total food. The remainder group. consisted mainly of spruce cone- worms, another defoliator of spruces. During spruce budworm outbreaks, The investigators estimated that a many species of birds prey on the single red squirrel (fig. 7) could eat abundant larvae (Mitchell 1952, 400 to 500 larvae per day. Dowden et al. 1953). However, birds can consume only about 2 percent of In New Brunswick, W. F. Cheshire an epidemic population (Crawford et observed red squirrels {T. hudsonicus al. 1983). Fewer species of birds gymnicus [Bangs]) in captivity and prey on endemic populations, but estimated a mean food capacity of their influence in limiting the number 600 to 700 mature budworm larvae of larvae can be significant. Crawford or pupae per day (Morris 1963). et al. (1983) presented information on Counts of red squirrels on some plots birds preying on spruce budworms in in the Green River Watershed in- forest stands supporting endemic, transitional, and epidemic populations

31 Figure 7—Red squirrel, Tamiasciurus hud- sonicus, preying on spruce budworm larva. of the spruce budworm. The most nian warblers, Dendroica fusca important bird predators were those (Müller); Nashville warblers, Ver- that maintained high population den- mivora ruficapilla (Wilson); and sities and high feeding rates over the golden-crowned kinglets, Regulus lower ranges of the insect's density, sátrapa Lichtenstein, were among and those that responded to initial those considered the most important rises in endemic populations. Black- predators of large larvae in north- capped chickadees; red-breasted eastern spruce-fir forests. nuthatches, Sitta canadensis L.; white-throated sparrows, Zonotrichia Other important predators of large albicollis (Gmelin) (fig. 8); blackbur- budworm larvae (fifth and sixth

32 í A

Figure 8—White-throated sparrow with spruce bud worm larva.

33 Predators of Pupae

instars) are the solitary vireo, Vireo Known predators of spruce budworm solitarius (Wilson); Swainson's pupae include spiders, beetles, spruce , Catharus ustulatus (Nuttall); cone worms, budworm larvae, syr- black-throated green warbler, D. phids, mammals, and birds. virens (Gmelin); yellow-rumped warbler, D. coronata (L.); Cape May Spiders warbler, D. tigrina (Gmelin); bay- breasted warbler, D. castanea Tothill (1923) estimated that of the (Wilson); magnolia warbler, D. progeny developing from each pair of magnolia (Wilson); and Tennessee spruce budworm moths laying 150 warbler, Vermivora peregrina eggs, one pupa (1 percent) would be (Wilson). The Tennessee, blackbur- eaten by spiders. The spiders were nian, and bay-breasted warblers have not identified, but they probably shown direct numerical responses to represent hunting spiders rather than increasing budworm densities (Morris web spinners. Because pupae of the et al. 1958, Mook 1963). Inverse spruce budworm are relatively responses have been noted for the immobile, they are susceptible mainly yellow-rumped and black-throated to foliage-searching predators. green warblers (Morris et al. 1958), However, spider prédation as a but Gage and Miller (1978) found source of pupal mortality probably is both of these warblers to be more insignificant. abundant under outbreak than under postoutbreak conditions. Beetles

Earlier we reviewed the relationships Predaceous beetles attack and feed on between birds and the spruce bud- pupae of the spruce budworm. In the worm (Crawford and Jennings 1982), Uxbridge Forest, Ontario, Thomson including prédation and bird popula- (1957) found adult Elateridae and tions, methods of determining con- Coccinellidae (species undetermined) sumption of budworms by birds, life feeding on budworm pupae. Of the histories of important predaceous 400 pupal sites examined before moth birds, and predator-prey models. emergence, 73 were attacked by Although considerable information is predators; however, beetles accounted available on birds and spruce bud- for only 2.2 percent of this prédation. worms, additional studies are needed, particularly at the endemic and transi- In spruce-fir stands of northern New tional population levels. Such studies Hampshire, Reeves et al. (1983) will help determine which species of identified several species of carabid birds are most beneficial in keeping beetles that are potential predators of budworm populations in check. The budworm pupae. Five criteria were forest can then be managed to con- used to denote potential predators, serve and enhance populations of im- including synchrony of the beetle's portant insectivorous birds. seasonal activity with the budworm's

34 pupal stage. Five species were indi- During prédation, the coneworm cated as possible predators of bud- spins a loose, silken cocoon that worm pupae based on seasonal partly encloses the budworm pupa activity: Notiophilus aeneus Hbst., (fig. 9). Barker and Fyfe (1947) Pterostichus adoxus (Say), P. noted that as many as four pupae coracinus (Newm.), Calathus in- were eaten by the same coneworm on gratus Dej., and Cymindis cribricollis spruce terminals in the Spruce Woods Dej. Individuals of Agonum retractum Forest Reserve, Manitoba. Liscombe (Lee), P. melanarius 111., and and Lejeune (1949) attributed much Synuchus impunctatus (Say) also were of the budworm's mortality in the abundant during the budworm's pupal Spruce Woods Forest Reserve to period. prédation by coneworms; they noted that coneworms destroy both larvae In spruce-fir stands of northern and pupae of the budworm, but Maine, carabid beetle activity was mainly pupae. greatest during the early and late lar- val stages of the spruce bud worm; In a laboratory experiment. Warren but many species were also active (1954) found that prédation of bud- during late June and early July, when worm pupae was much higher than budworm pupation occurs."^ Many of that of larvae, probably because of the same species, identified as poten- the pupa's inability to escape the tial budworm predators in New predaceous coneworm larva. In the Hampshire (Reeves et al. 1983), also field, as many as five budworm were collected in northern Maine. pupae were found in a single cone- worm's larval web (Warren 1954). Spruce Coneworms In 1949, the budworm population Pupae of the spruce budworm are dropped to low levels on the Spruce subject to prédation by larvae of the Woods Forest Reserve, Manitoba. spruce cone worm. The spruce cone- Warren (1954) indicated that this worm pupates a few days later than drop may have been due to heavy the budworm (McLeod and Daviault prédation by coneworms on the bud- 1963), and this asynchrony in life worm's pupal population. Coneworm cycles makes the budworm suscepti- populations reached a peak in 1948. ble to prédation by coneworms. The cone worm occasionally is abundant, Thomson (1977) examined 400 pupal often in conjunction with epidemics sites of the spruce budworm in the of the spruce budworm (Rose and Uxbridge Forest, Ontario. The ex- Lindquist 1977). These two rival amination was done 3 days before species not only compete for food adult flight began. The spruce cone- and shelter, but the cone worm preys worm was the most abundant pupal on the spruce budworm when new predator and accounted for 5.8 per- foliage becomes scarce. cent of the 73 pupae attacked by

35 Figure 9—Spruce coneworm, Dioryctria reniculelloides, preying on pupa of the spruce budworm.

36 predators. He concluded that late in- percent) were attributed to budworm stars of the coneworm are able to prédation. prey on the relatively defenseless spruce bud worm pupae. This source of pupal mortality is pro- bably insignificant except perhaps In Newfoundland, Otvos (1981) under special conditions of high reported that in spruce stands where disease incidence. the ratio of coneworm to budworm is about 50:50, prédation on budworm Syrphids pupae is likely to be high. Syrphid flies (Diptera: Syrphidae) In laboratory tests, Doganlar and resemble bees or wasps, but they lack Beirne (1978) observed larvae of D. stingers and have only one pair of pseudotsugella Munroe feeding on wings. The larvae of syrphid flies prepupae and pupae of the western vary considerably in habits and ap- spruce budworm. Consumption rates pearance (Borror et al. 1976). Many were one or two prey per larva per are predaceous. day, but when given a choice, this western coneworm preferred to feed Thomson (1957) found dipterous lar- on fresh foliage of Douglas-fir, vae, believed to be larvae of Pseudotsuga menziesii (Mirb.) Syrphidae, feeding on pupae of the Franco, and attacked bud worms only spruce budworm in the Uxbridge if such foliage was not available. Forest, Ontario. These predaceous larvae accounted for only 0.5 percent Budworm Larvae of the total pupae attacked by predators. An interesting case of cannibalism was observed in the Uxbridge Forest, Ants Ontario, and reported by Thomson (1957). Sixth instars of the spruce As omnivorous predators, ants attack budworm fed on new pupae of their and feed on pupae of the spruce bud- own species. The pupae were less worm. Thomson (1957) observed ants than 24 hours old and probably were (species undetermined) feeding on susceptible because of their soft pupal pupae of the spruce budworm in cases. A microsporidian disease that Ontario. During preliminary labor- retards larval development may have atory tests with the red wood ant, contributed to this cannibalistic Finnegan (1978) found that this behavior. More larvae were present species was highly aggressive in sear- than usual after the bulk of the popu- ching for and attacking late-instar lar- lation had pupated. Of the 400 pupal vae, pupae, and adults of the spruce sites examined, 73 were attacked by budworm. The red wood ant was in- predators; and attacks at 10 sites (2.5 troduced into Quebec in 1971, and nests are now well established at Lac

37 Normand and Valcartier (Finnegan Brunswick (Morris 1963). There was 1975, 1977). an abundant supply of spruce bud- worm larvae in the tree crowns. The The seasonal predatory activity of diet of this rodent probably includes these introduced red wood ants was insect pupae as well as larvae. studied in 1976. Two peaks in predatory activity were noted During the Green River studies in (McNeil et al. 1978). At first the New Brunswick, W. F. Cheshire authors believed that the temporary estimated that red squirrels had a drop between the peaks was attri- mean food capacity of 600 to 700 butable to pupation of tortricid lar- mature larvae or pupae of the spruce vae, including the spruce bud worm. budworm per day (Morris 1963). However, both larvae and pupae brought to the nest showed temporary In northern New Hampshire we col- reductions. The authors concluded lected red squirrels that had fed on that the reductions were due to a pupae of the spruce budworm. These drop in mean daily temperature. collections came from areas where spruce budworm populations were In northern Maine, we collected four low. The recovery of larval-pupal genera and nine species of ants by remains in squirrel stomachs indicates pitfall trapping in a budworm-infested that like birds, squirrels can search forest.^ Greater numbers oí Myrmica and find scarce prey. detritinodis Emery and Camponotus herculeanus (L.) were caught than Because budworm pupae are rela- other species in both strip clearcuts tively immobile and few drop from and in dense spruce-fir stands. Ants host trees, they are not susceptible to were active during most of the spruce strictly terrestrial predators, such as budworm's developmental stages, in- insectivorous shrews. Unless dis- cluding the pupal stage (fig. 10). lodged, most pupae remain attached by their cremaster hooks to silk spun Mammals by the larvae. Thus, mammalian prédation on spruce budworm pupae Our knowledge of mammals preying is restricted largely to arboreal on pupae of the spruce budworm is mammals. very limited. Potential mammalian predators of pupae include the deer Birds mouse P. maniculatus (Wagner) and the red squirrel. Numerous birds feed on pupae of the spruce budworm; at least 49 species The deer mouse P. maniculatus have been observed and recorded abietorum Bangs was observed in with pupal remains in their stomachs. crowns of mature fir trees on the Many of the same species that prey Green River Watershed, New on large larvae also prey on bud-

38 Figure 10—Carpenter ant, Camponotus sp., and spruce budworm pupa. worm pupae. Most are species of 11), Cape May, magnolia, yellow- warblers, vireos, kinglets, sparrows, rumped, black-throated green, and grosbeaks. There are some in- blackburnian, and bay-breasted dications that when pupae become warblers. available, fewer late-instar larvae are taken by birds (Mook 1963). During budworm epidemics, many transient birds flock into spruce-fir The species of birds we consider im- stands and consume large numbers of portant predators of endemic-level budworm pupae. These include red- budworm pupae in northeastern winged blackbird, Agelaius spruce-fir forests are the black- phoeniceus (L.); common grackle, capped chickadee, red-breasted Quiscalus quiscula (L.); evening nuthatch, golden-crowned kinglet, grosbeak, Coccothraustes vespertinus solitary vireo, and the Nashville (fig. (Cooper); and pine grosbeak.

39 Figure 11—Nashville warbler searching foliage for spruce budworm pupa.

40 Predators of Adults

Pinícola enucleator (L.). Their Predators of spruce budworm moths feeding often is instantaneous and include spiders, dragonflies, beetles, spectacular. robber flies, ants, and birds.

Interestingly, Johannsen (1913) Spiders reported that the purple martin, Progne subis (L.), fed on pupae of Spruce budworm moths are suscept- the spruce budworm during the ible to prédation by both web- 1911-12 outbreak in Maine; but the spinning and hunting spiders. Web- birds were only locally abundant. spinning species that are abundant in northeastern spruce-fir forests belong In Maine and New Hampshire, we to the families Araneidae, Dictynidae, estimated that birds consumed 941 Erigonidae, , and pupae per acre (2,325/ha) in stands Theridiidae. supporting endemic populations of spruce budworm, 9,016 pupae per Vagrant hunting spiders of the acre (22,279/ha) in stands supporting families Clubionidae, Salticidae, transitional populations, and 14,791 Thomisidae, and Philodromidae also pupae per acre (36,551/ha) in stands capture and feed on spruce budworm supporting epidemic populations moths. Most of these hunters actively (Crawford et al. 1983). These are search conifer foliage for prey; conservative estimates based on a however, the sedentary crab spiders pupal availability period of only 14 (Family Thomisidae) wait in ambush days. for passing prey.

Additional studies are needed to In , Turnbull (1956) determine the effects of pupal préda- observed both web-spinning and hunt- tion by birds, especially at low bud- ing spiders feeding on moths of the worm densities. Gage and Miller western spruce budworm. During (1978) concluded that birds can exert moth flight, webs of 39 argiopid a strong prédation pressure on preout- (Family Araneidae) and 34 theridiid break populations of the spruce bud- (Family Theridiidae) spiders were worm. To fully understand these observed at night. Almost every web regulatory processes, researchers need had captured a budworm moth; 85 to determine the consequences of moths were found in orb webs and 46 pupal prédation on generation in theridiid webs. Turnbull also survival. observed hunting salticid spiders cap- turing and feeding on gravid female moths. Sluggish, fully gravid females were especially vulnerable to attack by salticid spiders.

41 Our knowledge of spiders feeding on Family Dictynidae moths of spruce budworm in the Dictyna foliaceae (Hentz) northeastern United States and Dictyna phylax Gertsch and Ivie Canada is limited. Using serological techniques to detect prédation, Species in Maine most commonly Loughton et al. (1963) estimated that observed with budworm moths were in 1959 about 30 percent and in 1960 Theridion pictum and Frontinella about 25 percent of the spiders col- communis. Both species build their lected and tested had fed on spruce webs in young, understory spruce-fir budworm during the moth stage. The trees, and each species builds a estimates were made when preceding characteristic web. T. pictum con- budworm populations were high (433 structs a tangle web of many ensnar- egg masses per 100 ft^ [46.6/m2] of ing viscid threads (fig. 12). This foliage) and lower (156 egg masses theridiid spider also ties several per 100 ft2 [16.8/m2]), respectively. spruce or fir needles together, form- However, the percentages do not ing a "turret" in which to hide. necessarily refer only to moth préda- Discarded cadavers of budworm tion because developmental stages moths often are incorporated into the overlapped (pupae, moths, and eggs). walls of the turret.

In Maine we have collected the The linyphiid F. communis spins a following species of web-spinning "bowl and doily" web consisting of spiders with spruce budworm moths a cuplike bowl and a horizontal sheet in their webs: spun beneath the bowl. Above the bowl the spider spins several ir- Family Theridiidae (Comb-footed regular strands of silk of varying spiders) length to impede flying insects, many Theridion frondeum Hentz of which fall into the bowl. We com- Theridion murarium Emerton monly observed three or four bud- Theridion pictum (Walckenaer) worm moths in webs of both species (fig 13). Family Linyphiidae (Sheet-web weavers) Dragonflies Frontinella communis (Hentz) In Maine we frequently observed Family Araneidae (Orb weavers) dragonflies hawking and capturing Araniella displicata (Hentz) spruce budworm moths on the wing. Araneus marmoreus Clerck These observations led to a study to Cyclosa cónica (Pallas) determine the species of Odonata associated with spruce-fir forests of Family Agelenidae (Funnel-web Maine and to determine by gut- weavers) content analysis the prey of adult Agelenopsis utahana (Chamberlin Odonata, with special emphasis on and Ivie) 42 i^

■íS

Figure 12—Theridion piclum in web with spruce budworm pupae and moth prey.

43 Figure 13—Bowl and doily spider, Frontinella communis, witii spruce budworm motli prey. identifying remains of the spruce bud- examined, 163 (47 percent) had fed worm (Tsomides et al. 1982). on lepidopterans. Fifty percent of the Odonata representing nine families, Anisoptera had lepidopterous scales in 22 genera, and 39 species were col- their guts, whereas only 19 percent of lected in spruce-fir forests of Maine. the Zygoptera had eaten lepidop- Collecting localities (14) spanned the terans. The libellulids (Family central part of the State from the Libellulidae) were the most numerous northwest to the southeast. All sites Odonata collected, representing 424 had budworm-infested trees. individuals. Fully 58 percent of the libellulids had fed on lepidopterans. Both dragonflies (suborder Anisoptera) and damselflies (suborder Although we were unable to specifi- Zygoptera) were collected in this cally identify spruce budworm study (Tsomides et al. 1982). Of 350 remains in the odonate gut contents, specimens dissected and gut contents we were able to recognize lepidop-

44 teran scales. Comparisons of these potential predators of spruce bud- scales with scales from laboratory- worm adults. The species were reared spruce budworms proved Pterostichus melanarius 111., P. inconclusive. However, many of the coracinus (Newm.), S. impunctatus lepidopterous scales undoubtedly were (Say), Cymindis crihricollis Dej., and those of the spruce budworm because P. decentis Say. Seasonal activities of odonates often were collected while these beetles coincided, with budworm they were hawking budworm moths moth activity in northern New (fig. 14). One was captured Hampshire. while feeding on a spruce budworm moth. Comparison of budworm moth In northern Maine, carabid beetles flight information with gut-content also were active during the bud- analysis indicated that odonates were worm's flight period; but greatest consistently feeding on moths during activity, as evidenced by pitfall June and July, when spruce budworm catches, occurred during the early- flight occurred. and late-larval stages of the budworm. "^ Although the results of our gut- content study are encouraging, other Robber Flies methods are needed to confirm this apparent widespread odonate feeding This group of dipterans (Diptera: on the spruce budworm. Serological ) contains about 850 species techniques offer one possible ap- in North America (Borror et al. proach (Mclver 1981) but may be 1976), but the species found in north- difficult to quantify (Boreham 1979). eastern spruce-fir forests are poorly known. Adult robber flies are found Beetles in a variety of habitats; each species usually is found in a characteristic Our knowledge of beetles preying on habitat (Borror et al. 1976). The spruce budworm adults is limited and adults are predaceous and attack a intuitive at best. Because budworm variety of prey, including Lepidop- moths are highly mobile and can tera. Prey usually is captured on the escape predators by flight, their wing, and robber flies will attack in- susceptibility to foliage-searching sects larger than themselves. beetles is limited. However, freshly emerged moths, moths in coitus, and In New Brunswick, I. W. Varty ovipositing females may be subject to observed and collected an adult rob- attack by beetles. Beetle prédation on ber , Asilus sp., feeding on a budworm moths is probably spruce budworm moth. Apparenfly, inconsequential. this is the only observation of préda- tion on spruce budworm. Reeves et al. (1983) identified at least five species of carabid beetles as

45 «'^äK*'

Figure 14- Dragonfly hawking spruce budworm moth.

Because prey usually are captured on Ants the wing, prédation by robber flies on spruce budworm may be limited to Few observations have been made of moths. However, large larvae drop- ants (Hymenoptera: Formicidae) prey- ping from host trees may be eaten. ing on spruce budworm moths. The Asilidae associated with north- Finnegan (1978) noted that F. eastern spruce-fir forests, their lugubris attacked adults of the spruce predatory habits, and their potential budworm in laboratory feeding tests. for preying on spruce bud worms need Field observations of prey brought to be investigated. back to nests by this species included

46 wing fragments and abdomens, moths in northeastern spruce-fir thought to be those of C fumiferana forests: black-capped chickadee, aduhs (McNeil et al. 1978). In north- golden-crowned kinglet, and Cape ern Maine, ants were very active dur- May, yellow-rumped, and bay- ing the adult stage of the spruce bud- breasted warblers. Most of these worm, but none were observed species have been observed hawking feeding on budworm moths.^ and capturing moths in flight.

We have observed ants removing and Greenbank (1963) noted that bud- scavenging male spruce budworm worm moths represented a small pro- moths caught in pheromone-baited portion of the gizzard content of traps coated with sticky materials. wood warblers and flycatchers. After capturing several moths, the Despite numerous hours of observa- trap's sticky surfaces become coated tion from tree platforms, he never with moth scales and other debris, saw birds capturing moths on the allowing the ants to traverse the wing. Because moth populations were sticky surfaces and remove the moth high in relation to bird populations, bodies, leaving behind only wing Greenbank concluded that bird préda- fragments. tion on the Green River Watershed was probably unimportant. Because spruce budworm moths generally are mobile and capable of The effects of moth prédation by flight, they are less susceptible to ant birds on population dynamics of the prédation than earlier life stages. spruce budworm are unknown. Cer- However, newly emerged moths, and tainly, female moths are susceptible especially egg-laden females, may be to prédation before and during attacked by ants and other foliage- oviposition. Prédation on egg-laden searching predators. If moths are cap- moths may adversely affect the suc- tured before egg laying, prédation on ceeding generation, especially at low adults could be an important source population densities. More observa- of mortality. tions and collections are needed dur- ing the moth stage, particularly over Birds a range of budworm densities.

At least 25 species of birds are known to capture and eat spruce bud- worm moths. Chickadees, thrushes, kinglets, vireos, warblers, grosbeaks, and sparrows are among the predators of spruce budworm moths.

We consider the following species as important predators of budworm

47 Predators As Natural Regulators

Predators are natural regulators of crease in fecundity or searching populations upon which they capacity where food habits are feed (Price 1975). Numerous specific, mathematical models have been • Normally contact, kill, and con- developed to explain the regulatory sume large numbers of prey processes of prédation and predator- organisms; by contrast, parasites prey interactions (Rolling 1959a, spend most of their life on or in a 1959b, 1966; Rolling and single host. Buckingham 1976; Nicholson 1933, 1954; Royama 1971, 1977; Solomon Buckner (1971) reviewed the roles of 1949; Tinbergen 1960; Watt 1959). vertebrate predators in the forest The spruce budworm serves as an ex- ecosystem. Re advocated that forest ample in some models. Prédation acts should take an integrated as one of the most important factors approach that includes the use of in maintaining stability within ecosys- small vertebrates. Re defined the tems (Sailer 1971). ideal predator for encouragement as one that (1) has a high feeding A predator is an organism that kills capacity, (2) prefers the target insect and consumes many animal-food pest, (3) selectively feeds on insects items in its lifespan, whereas a not attacked by other natural control parasite () requires and eats agents (e.g., parasites), (4) maintains only one animal in its life (Price high population densities, (5) re- 1975). Rowever, because the female sponds quickly to other control parasite oviposits in numerous hosts, measures applied simultaneously, and parasites ultimately may be responsi- (6) is unaffected by subsidiary ble for killing many hosts. Sailer treatments such as pesticides. (1971) compared the attributes of in- vertebrate parasites and predators and Predators may respond numerically concluded that in general, predators and functionally (or both) to increases • Tend to be less specific in their in prey populations. Both invertebrate food habits, and vertebrate predators can increase • Are less dependent on a single their reproductive potential (numerical food resource, response) in the presence of abundant • Are better able to maintain stable food and concentrate predatory activ- populations that tend to exploit prey ities (functional response) on selected species in proportion to their relative prey. Because many invertebrates (in- abundance, sects) are multivoltine (i.e., having • Tend to be longer lived and less many generations per year), they may likely to be adversely affected by respond sooner to prey abundances physical factors of the environment, than some vertebrates. Many in- • Rave searching capabilities that vertebrates and vertebrates have an generally are greater, acute searching capability, allowing • Likely have a compensatory in- them to find prey even at low den-

48 sities. Vertebrates have the added ad- iferous foliage and feed on larvae, vantage of forming "search images." pupae, and adult spruce budworms.

Buckner (1967) stressed the impor- Populations of predators and their tance of both numerical and func- feeding behaviors may vary depen- tional responses of avian and mam- ding on densities of potential prey. malian predators on forest insect Populations of the spruce budworm populations. He divided numerical generally are classified as low-level responses into breeding and or endemic populations and high-level behavioral responses. Examples of or epidemic populations. Of course, breeding numerical responses to in- there may be numerous gradations creasing budworm densities have been between these two extremes. "Out- demonstrated for the Tennessee, breaks" generally refer to epidemic blackburnian, and bay-breasted populations. warblers (fig. 15) (Morris et al. 1958, Mook 1963). Behavioral Endemic Spruce Budworm numerical responses often are instan- Populations taneous and spectacular. Roving flocks of colonial nesting birds such It is generally recognized that as grackles and blackbirds that enter predators help maintain insect popula- the forest and feed on budworms tions at low, endemic levels. exemplify a behavioral numerical Predators and other natural enemies response. Such responses may be such as parasites help keep potential significant at high prey densities pest populations in check until they when budworms make up a large are released, usually by some climatic percentage of the diet (Dowden et al. factor or combination of abiotic and 1953). biotic factors. However, the impor- tance of predators and their potential Buckner (1967) divided functional for regulating and maintaining responses into basic components endemic populations of the spruce (reaction to increasing prey density) budworm have not been studied in and subsidiary components (food detail. preferences and feeding behaviors such as hoarding and "sport" kill- Morris (1963, p. 244) concluded that, ing). Ovenbirds, Seiurus aurocapillus ' 'If prédation has any important in- (L.), exhibit a ftinctional response to fluence on the dynamics of budworm outbreaks of the spruce budworm by populations it must, therefore, be ex- changing their feeding behavior (Zach erted during the endemic period, or and Falls 1975). That is, they nor- during the early years of population mally restrict their feeding to or near release . . . ." Other investigators the forest floor; but when budworms have reached similar conclusions reach high densities, they search con- regarding the importance of natural control mechanisms operating against

49 Figure 15—Bay-breasted warbler searching foliage for budworm larval prey.

50 low-level populations of the spruce (1948) noted that another defoliating budworm (Miller and Varty 1975). insect of balsam fir, the eastern However, our knowledge and blackheaded budworm (Acleris understanding of the specific variana [Fernald]), reached a high predators and their regulatory func- population level in the late 1940's. tions during endemic populations of He suggested that these high popula- the spruce budworm is limited. Vir- tions of an alternate food source may tually nothing is known about in- have relieved prédation pressure on vertebrate predators and endemic the spruce budworm, thus aiding in budworm populations. And little in- population release. formation is available on vertebrates and their importance in maintaining In northern New Hampshire, the esti- low-level populations. mated seasonal consumption by birds exceeded 1,336 larvae and 931 pupae Dowden et al. (1953) concluded that per acre (3,300 larvae and 2,300 in light infestations of the spruce bud- pupae/ha) when budworm populations worm, prédation by birds should be were endemic (Crawford et al. 1983). of great economic importance. Gage These estimates were derived by and Miller (1978) concluded that determining daily consumption rates birds can exert a strong prédation (Gage et al. 1970) and multiplying by pressure on preoutbreak populations 20 and 14, the estimated number of of the spruce budworm. Morris et al. days large larvae and pupae were (1958) indicated that birds can be ex- present, respectively. These estimates tremely important when budworm represent prédation by species that populations are at endemic levels of are adapted for and capable of fin- 1,000 or fewer per acre (< 2,471/ha). ding sparsely distributed budworms. Overall, prédation by birds amounted At least two studies point to the to 87.2 percent in spruce-fir stands possible inñuence of prédation on supporting endemic populations. This release of spruce budworm popula- is a conservative estimate because not tions. Graham and Orr (1940) sug- all species of birds were sampled, gested that the 1912 outbreak of the some sample sizes were small, and spruce budworm in Minnesota might estimates of seasonal availability were have been precipitated or caused by moderate, e.g., Mook (1963) the scarcity of small insectivorous estimated that sixth-instar larvae are birds. Thousands of migrating available for 30 days. warblers were killed by late snow and ice storms during the late springs of At low larval densities of the spruce 1907 and 1910. Warbler populations budworm. Watt (1963) estimated that had recovered to only about 10 per- only 0.46 larva per 10 ft^ (0.42/m2) cent of their former numbers when of foliage would have to be eaten by the budworm outbreak started in predators to account for a decrease in 1912. In New Brunswick, Morris survival rate. Predator populations as

51 low as one pair of breeding birds per that when the number of budworms is acre along with one or two spiders reduced by starvation, the percentage per 10 ft2 (0.9 to l.Q/m^) of foliage of budworms destroyed by parasites (300,000 ftVacre = 30,000 to 60,000 and predators reaches 85 to 95 spiders/acre; 68,869 mVha = 74,130 percent. to 143,260 spiders/ha) have con- siderably more feeding potential than In New Brunswick, population Watt's estimated value. Additionally, dynamics studies showed that the we can increase bird populations with numerical responses of all predators appropriate silvicultural modifications to increasing budworm density was in spruce-fir stands (Crawford and limited (Watt 1963, Morris 1963). Titterington 1979, Titterington et al. Spiders and predaceous insects 1979), thus increasing predator apparently showed limited numerical potential. responses to increases in budworm density (Watt 1963), whereas birds In summary, we have very little in- showed marked increases. Buckner formation about predators, prédation, (1971) concluded that small mammals and endemic populations of the play an insignificant role in the spruce bud worm. Much more work dynamics of budworm populations. needs to be done to fully understand Earlier, Morris et al. (1958) indicated the regulatory roles predators play in that two mammals, the short-tailed maintaining budworm populations at shrew, B. brevicauda (Say), and the low levels. rock vole, M. chrotorrhinus (Miller), showed possible direct numerical Epidemic Spruce Budworm responses to increasing budworm Populations populations; however, they discounted these responses as being directly Predators may affect budworm related to the spruce budworm. populations before, during, and after outbreaks. The absence or lack of Morris (1963) noted that most préda- prédation may be instrumental in tion on spruce budworm occurs dur- releasing budworm populations from ing the large-larval period. Prédation endemic to epidemic levels (Graham during this period is especially impor- and Orr 1940, Morris 1948 and tant because it is the period that 1963, Morris et al. 1958). Predators determines generation survival for the and prédation have received much spruce budworm. Both spiders and more attention during an epidemic birds kill appreciable numbers of lar- than before or afterward. Graham and vae during the late-larval period of Orr (1940) concluded that parasites the spruce budworm. and predators play a minor role dur- ing an outbreak, but that at the end Birds exhibit both numerical and of an outbreak they effectively functional responses to increasing destroy most stragglers. They noted populations of the spruce budworm.

52 We have reviewed the literature 2,471,000 budworms/ha). Their describing these responses (Crawford estimates were more than double that and Jennings 1982). Direct and strong of Kendeigh (1947) and represent numerical responses have been 17,000 to 70,000 budworms de- observed for the Tennessee, blackbur- stroyed per acre (42,007 to nian (fig. 16), and bay-breasted 172,970/ha). During an outbreak on warblers (Morris et al. 1958, Mook the Green River Watershed in New 1963). Later, Gage and Miller (1978) Brunswick, populations reached 8 showed more yellow-rumped and million larvae per acre (19,768,000 black-throated green warblers under larvae/ha); however, bird prédation outbreak than postoutbreak condi- was estimated at less than 1 percent tions. The ovenbird exhibits a func- (Morris et al. 1958). tional response to outbreaks of the spruce budworm (Zach and Falls For epidemic populations in Maine, 1975). Ovenbirds generally restrict we estimated that prédation by birds their feeding behavior to and near the amounted to 2.4 percent (Crawford et forest floor. However, once spruce al. 1983), with total seasonal con- bud worms reach high densities, these sumption of more than 21,044 larvae birds frequently were observed and 14,569 pupae per acre (52,000 searching branches of infested con- larvae and 36,000 pupae/ha). The ifers and feeding on larvae, pupae, percentage prédation by birds was and adults. much less in stands supporting epidemic populations than in stands The percentage of budworms de- supporting transitional populations stroyed by birds during epidemics is (23.0 percent) or in stands supporting usually less than 10 percent; how- endemic populations (87.2 percent). ever, up to 40 percent or more of Daily consumption rates for in- their total diet may be spruce bud- dividual species of birds are sum- worms (Mitchell 1952, Dowden et al. marized in Crawford and Jennings 1953). Kendeigh (1947) estimated (1982). that birds destroyed 4.3 percent of a heavy infestation of spruce budworms Several investigators have indicated in Ontario in 1945. He estimated that that birds contribute to the collapse of one breeding pair plus their nestlings spruce budworm outbreaks. During a could consume 16,000 larvae and "waning" outbreak in New pupae per acre (39,536 larvae and Brunswick in 1918, Tothill (1923) pupae/ha) during the period of estimated that birds consumed about availability. In New York, George 13 percent of the larvae. This percen- and Mitchell (1948) calculated that tage is slightly higher than that the degree of control by birds was observed during epidemics. Dowden from 3.5 to 7.0 percent when infesta- et al. (1950) noted that insectivorous tions were 500,000 to 1,000,000 bud- birds, largely warblers, accounted for worms per acre (1,235,500 to a considerable portion of the popula-

53 Figure 16—Blackbumian warbler searching foliage for spruce budworm larva.

54 tion reduction to a collapsing out- In summary, most of our knowledge break in the Adirondacks in 1947 and about predators, prédation, and 1948. Watt (1963) indicated that birds epidemic spruce budworm populations affected larval survival at low bud- concerns only two predator groups, worm densities; Buckner (1971) inter- birds and spiders. The impacts caused preted this to mean that birds may be by other predator groups, both in- of considerable importance in clean- vertebrate and vertebrate, need inves- ing up residual populations of the tigation over a range of predator-prey pest after the major outbreak has run densities. its course. Biais and Parks (1964) in- dicated that residual populations of the spruce budworm were controlled through the predatory actions of grosbeaks that invaded remaining pockets of infestation after insecticide treatment.

Spiders are also abundant during epidemics of the spruce budworm. Estimates of population densities in- dicate that spiders far outnumber all groups of predaceous insects on con- iferous foliage (Morris 1963); popula- tions of spiders in forest stands of medium density were estimated at 75,000 per acre (185,323/ha), not in- cluding species inhabiting the ground or lower vegetation. With a popula- tion density of two or more per 10 ft^ (1.8/m2) of foliage and a feeding rate of one spruce budworm larva every 3 days, Morris (1963) concluded that spiders would have a very high feeding potential compared with birds. However, their capacity to regulate epidemic populations has not been demonstrated and may be hampered by limited numerical responses (Morris 1963) compared with exploding prey populations.

55 Methods for Studying Prédation

Measurement of prédation in field of predator feedings, and identifica- populations is notoriously difficult tion of prey remains in predator (Morris 1963). Unlike parasites and stomachs or in feces. Numerous tech- diseases, predators often leave little niques used include exclusion techni- or no evidence of their feeding activi- ques, direct assessments, serological ty. Many predators are cryptic and tests, tagging with radioisotopes, secretive; others are nocturnal, mak- direct observations, .and photographic ing observations difficult except with recordings. Buckner (1966), Kiritani specially designed equipment. and Dempster (1973), DeBach et al. (1976), and Southwood (1978) re- Buckner (1966) indicated that three viewed pertinent literature available basic measurements must be taken to for assessing prédation. We discuss obtain a true understanding of some of the techniques applicable to predator-prey systems: density of spruce budworm. prey, density of predators, and extent of destruction of prey by predators. Exclusion Techniques Sampling methods generally have been worked out for each life stage These include (1) mechanical exclu- of the spruce budworm (Morris 1955, sion, (2) insecdcidal check method, Miller 1958, Sanders 1980), but (3) biological check method, and (4) methods for determining predator hand-removal techniques. Mechanical densities are often lacking. Estimates exclusion is achieved by constructing of prey consumption require deter- sleeve or larger cages to exclude minations of seasonal availability, predators (Campbell et al. 1981). predator feeding behaviors, and the Disadvantages of this technique are proportion of prey in the predator's that dispersal of prey is prevented, total diet. Buckner (1966) discussed the physical environment is modified five factors to consider when by the cage (DeBach and Bartlett evaluating individual predator species: 1964), and predator identities usually (1) food capacity of the predator, are unknown, though some indication (2) effects of alternate foods, can be gained by varying mesh sizes, (3) prey defense mechanisms, e.g., birds v^. ants. Both insecticidal (4) numerical responses of predators and biological check methods employ to increasing prey populations, and "before and after" observations or (5) functional responses of predators "check-treated" observations in con- to prey populations. junction with an insecticide or biological used selectively to kill Despite numerous limitations, both predators (Kiritani and Dempster direct and indirect methods have 1973, DeBach et al. 1976). Hand evolved for studying the effects of removal of predators probably is predators on prey populations. These the most reliable exclusion tehnique, methods include techniques for obser- but it is extremely time consuming ving predators in action, identification and costly.

56 Direct Assessments precipitate. There have been numer- ous refinements to the basic ''preci- Direct assessments of prédation in- pitin test," and new, more sensitive clude stomach-content analyses, coun- tests have been developed (see Miller ting corpses killed by a predator 1979 for a review). (e.g., insects in spider webs [Turnbull I960]), examining feces or Serological techniques have been used fecal pellets for prey remains, and qualitatively to identify the predators placing known numbers of prey in of particular prey species, though the field. Stomach-content analyses quantitative estimates of prédation require a thorough knowledge of the are possible under certain conditions ñora and fauna where samples are (Kiritani and Dempster 1973, collected (Korschgen 1980), and par- Dempster 1960). The minimum ticularly familiarization with bud- number of prey eaten can be worm morphology of all life stages. estimated by this equation: Some predators thoroughly masticate their prey, making postmortem iden- Prey eaten = (PmT)/t tifications difficult. Counting corpses also has drawbacks because some where P is the number of predators predators bury their prey. Placing present, m is the proportion of those prey in the field has a disadvantage tested that gave positive serological in that the rate of attack may be af- reactions, T is the total time that prey fected by density, position, or ex- are available for prédation, and t is posure of prey (Kiritani and the length of time that a meal re- Dempster 1973). mains detectable by serological methods (Kiritani and Dempster Serological Tests 1973). The latter usually is deter- mined under controlled conditions in Serological tests offer one of the most the laboratory with known predator- convenient, reliable methods for prey feedings. assessing prédation, especially by in- vertebrate predators. This technique Radioisotope Tagging has been used to determine spider and mite prédation on the spruce bud- Radioisotopes can be used to "tag" worm (Loughton et al. 1963). or mark prey and identify predators Basically, prey material is identified feeding on the labeled organisms. in the gut of a predator by its reac- Labeling of prey usually is done by tion with the blood serum from a immersion, by use of sprays or gases, mammal (usually a rabbit) that has by introduction through food or been sensitized against the prey. The water, or by injection (Odom and blood serum contains antibodies Golley 1963). Krall and Simmons (antisera) that react with proteins (1977) used phosphorus-32 injected (antigens) of the target prey to form a into tree roots of balsam fir to label

57 the spruce budworm and identify records can be made of birds, wasps, carabid beetles preying on budworm or ants returning to their nests with larvae. prey.

Kiritani and Dempster (1973) con- Another direct observational tech- cluded that in practice, tagging with nique entails placing predrilled blocks radioisotopes was extremely difficult of wood in the forest for trap-nesting to quantify and use because there is wasps (Krombein 1967). Foundress considerable variation in radioactivity female wasps use the blocks to make of individual prey; many predators their nests, which are provisioned consume only part of their prey; with paralyzed lepidopterous prey. excretion of the radioisotope by a Observations can be made of the predator seems to depend on the wasps bringing prey to their nests, or amount of food it subsequently eats; the blocks can be split open, the nests and it is difficult to ensure that the and cells examined, and the prey predator obtained all of its radioac- counted directly. tivity solely from the tagged prey, i.e., the environment may become In summary, each technique or contaminated or the isotope may be method for studying prédation has passed up the food chain to limitations. When possible, it is best scavengers and secondary predators. to employ more than one technique, e.g., direct field observations of Direct Observations prédation coupled with examination of predator-gut contents. Direct observations of prédation pro- bably are the most reliable method; but they are costly, time consuming, and often difficult or impossible to make. Special techniques and equip- ment such as blinds, telescopes, and night-viewing scopes may be re- quired. If prédation is common, the frequency of observing the predator consuming the prey may be useful (Morimoto 1960); however, prédation usually is difficult to observe in ar- boreal habitats such as tree crowns of budworm-infested trees. Photographic recordings of predators returning to nests with food may be helpful so long as the food can be recognized and identified on film. With proper electronic equipment, permanent

58 Conservation and Enhancement of Predators

Predators and other natural enemies eumenid wasps clearly selected the of the spruce budworm can be pro- more open habitats of strip clearcuts tected, maintained, and enhanced by over dense spruce-fir stands environmental manipulation and (Jennings and Houseweart 1984). cultural practices. The spruce-fir Strip harvesting and other measures forest can be managed to reduce that produce openings in the forest susceptibility to budworm damage favor these predators of late-instar and to provide suitable habitats for spruce budworm larvae. natural enemies of the spruce bud- worm. Effective management requires In northern New Hampshire, stands a thorough knowledge and under- of red spruce had more carabid standing of predator biologies and beetles (species and individuals) than their habitat needs, including food, mixed stands of spruce-fir, fir- cover, and nesting sites. Examples of spruce, or fir (Reeves et al. 1983). these requirements and provisions Many of these carabid beetles occu- follow. pying spruce stands were determined to be potentially important predators Habitat Requirements of the spruce budworm. Thus, silvi- cultural techniques aimed at reducing Bird populations respond to land- the balsam fir component would not management measures (Crawford and only decrease susceptibility of the Titterington 1979, Crawford et al. forest to budworm damage but also 1981, Titterington et al. 1979) and promote greater densities of can be increased by directed forest predaceous carabid beetles. practices. A mature managed forest containing a mix of tree species and Similarly, strip harvesting contributes size classes, and with scattered open- to dispersal losses of early instars of ings and patches of regeneration, sup- the spruce budworm (Jennings et al. ports populations of birds that prey 1983) and provides new habitats for effectively on spruce budworm many predaceous arthropods, in- (Crawford et al. 1983). These find- cluding phalangids, ants, spiders, and ings support the thesis that environ- carabid beetles. mental diversity favors the conserva- tion of and increase in numbers of Food Requirements beneficial organisms (Glen 1954). The availability of food, including Predaceous eumenid wasps prefer alternate food sources, is essential for open areas with abundant light encouragement of beneficial penetration (Fye 1972). Wasp popula- organisms (Coppel and Mertins tions were larger and more diverse in 1977). Most predators feed on a recently disturbed than in nondis- variety of prey organisms; few have turbed forests of northwestern specific, restricted diets. The avail- Ontario. In northern Maine, the ability and abundance of alternate

59 foods can favor and maintain popula- entomophagous insects (Stern et al. tions of predators during periods 1976). when target prey are scarce. Nesting Requirements Forests can be managed to provide a mix of species that, in turn, provide European foresters routinely provide food sources for numerous her- nest boxes to encourage birds (Bruns bivores. The herbivores, including 1959, Franz 1961), but this technique defoliators of spruce-fir, provide has received minimal attention food sources (prey) for polyphagous elsewhere (Goppel and Sloan 1971). predators. For example, eumenid Artificial structures have been used to wasps prey on late instars of the encourage predaceous Polistes wasps spruce budworm and on a variety of in agroecosystems (Lawson et al. other lepidopterous larvae (Fye 1962, 1961, Kirkton 1971). Apparently, this Jennings and Houseweart 1984). Most technique has not been used in north- of the prey are defoliators of north- eastern spruce-fir forests, where eastern hardwoods and softwoods. species of Vespula are common. Managing the forest to promote species diversity can provide alternate Nesting materials can be provided to foods for predators. attract and encourage nesting by eumenid wasps (Collins and Jennings The adults of numerous parasitic in- 1984). Danks (1971) noted that the sects require alternate food sources scarcity of natural nesting sites was a such as pollen and nectar (Syme limiting factor to population buildup 1966, 1975; Leius 1960). Similarly, of these wasps in England. Although predaceous eumenid wasps are at- strip and clearcut harvesting usually tracted to flowering plants (Fye produces abundant logging slash and 1972). In northern Maine, eumenid debris, the number of suitable nesting wasps were most common in open sites (with holes) can be increased habitats with abundant floral forage substantially by placing predrilled (Jennings and Houseweart 1984). blocks of wood in these open Managing the forest to encourage habitats. flowering shrubs and forbs will pro- vide alternate food sources for these Eumenid wasps can be trapped in one predators of budworm larvae. Ar- locality, transported with minimal tificial plantings and sowings of disturbance, and released in a new wildflower seeds also are possibilities locality. Such manipulations are for increasing nectar sources of possible because the wasps seal their preferred species. In the Soviet nest entrances with mud. The nest re- Union, considerable work has been mains closed until the new generation devoted to the use of nectar-bearing emerges. By collecting provisioned plants to increase the effectivness of blocks before adult wasp emergence.

60 Summary

field crews can centrally locate wasp We have reviewed available informa- populations in areas of abundant prey tion on predators of the spruce bud- species. worm, their general importance in the population dynamics of the pest, Other measures that encourage methods of studying prédation, and predators include lopping and scatter- conservation and enhancement of im- ing brush for insectivorous shrews portant predators. Obviously, there and other small mammals (Hamilton are many gaps and voids in our and Cook 1940), leaving dead snags knowledge and understanding of these for cavity-nesting birds (Hardin and agents of pest mortality. Much work Evans 1977), and erecting shelters for remains to be done, particularly in bats (Buckner 1966). identifying important predaceous species, determining their habitat re- In addition to environmental quirements, and assessing their manipulations and cultural practices regulatory roles. Such information is to encourage predators, curtailing the especially needed at low population use of pesticides would help conserve levels of the pest so that we can learn natural enemies of the spruce bud- how to manage the forest to prevent worm. Many parasitic and predaceous future outbreaks. Predators definitely species are highly susceptible to cer- warrant inclusion in future pest- tain pesticides; widespread use has management systems for the spruce resulted in the destruction of natural- budworm. enemy complexes (DeBach 1974, Goppel and Mertins 1977). Curtailing pesticide spraying and developing safer materials, such as insect viruses and Bacillus thuringiensis Berliner, will help conserve natural-enemy complexes of the spruce budworm.

61 Acknowledgments Selected References and Guides for Identifying Predators

We are grateful to Alison Here are general references and LaRochelle, Enfield, CT, for the identification guides for various watercolor paintings. predator groups. The list is by no means exhaustive but includes books We thank the following for supplying and publications that are helpful. Par- information on predators: John B. ticularly helpful are those pertaining Dimond, Mark W. Houseweart, Glen to introductory biologies. Most mam- W. Koehler, Robert K. Lawrence, mals and birds can be identified with K. P. Lim, and L W. Varty. We field guides; identification of in- also thank David M. Kendall, Janet vertebrate predators often requires J. Mel vin, Marcia A. Skratt, and specialized keys found in taxonomic Bonnie B. Grant. We are indebted to monographs. Consultation with a tax- those who reviewed the manuscript: onomist or identification specialist P. D. Kingsbury, Canadian may be advisable. Service, Forest Pest Management In- stitute, Sauh Ste. Marie, ON; R. M. Phalangids Reeves, Entomology Department, University of New Hampshire, Bishop, S. C. The Phalangida Durham; and H. R. Smith, North- (Opiliones) of New York with special eastern Forest Experiment Station, reference to the species in the Forest Insect and Disease Laboratory, Edmund Niles Huyck Preserve, Hamden, CT. Rensselaerville, N.Y. Proceedings of the Rochester Academy of Science. Verification of insect names was 9(3): 159-235; 1949. handled by Terry L. Erwin and Jonathan Coddington of the Depart- Edgar, A. L. Studies on the biology ment of Entomology, Smithsonian In- and ecology of Michigan Phalangida stitution; and Edward W. Baker, (Opiliones). Misc. Publ. 144. Ann Robert D. Gordon, and David R. Arbor, MI: University of Michigan, Smith of the Systematic Entomology Museum of Zoology; 1971. 64 p. Laboratory at the Insect Identification and Beneficial Insect Introduction In- Spiders stitute, USD A Agricultural Research Service. Fred Stormer, USD A Forest Gertsch, Willis J. American spiders. Service, Forest Environment Re- 2d ed. New York: Van Nostrand search, verified the names of birds Reinhold; 1979. 274 p. and mammals. Kaston, B. J. How to know the spiders. 3d ed. The pictured key nature series. Dubuque lA: Wm. C. Brown; 1978. 272 p.

62 Mites Dillon, E. S.; Dillon, L. S. A manual of common beetles of eastern Treat, A. E. Mites of moths and but- North America. Evanston, IL: Row, terflies. Ithaca and London: Cornell Peterson; 1961. 884 p. University Press; 1975. 362 p. Lindroth, C. H. The ground beetles Baker, E. W.; Wharton, G. W. An of Canada and Alaska. Part 2. introduction to Acarology. New Opuscula Entomológica Supplemen- York: Macmülan; 1952. 465 p. tum. 20: 1-200; 1961.

Krantz, G. W. A manual of Lindroth, C. H. The ground beetles acarology. Corvallis, OR: Oregon of Canada and Alaska. Part 3. State University; 1978. 509 p. Opuscula Entomológica Supplemen- tum. 24: 201-408; 1963. Dragonflies Lindroth, C. H. The ground beetles Needham, J. G.; Westfall, M. J., of Canada and Alaska. Part 4. Jr. A manual of the dragonflies of Opuscula Entomológica Supplemen- North America (Anisoptera). tum. 29: 409-648; 1966. Berkeley and Los Angeles: University of California Press; 1955. 615 p. Lindroth, C. H. The ground beetles of Canada and Alaska. Part 5. Walker, E. M. The Odonata of Opuscula Entomológica Supplemen- Canada and Alaska. Vol. 2, part IIL tum 33: 649-944; 1968. The Anisoptera—four families. Toronto, ON: University of Toronto Lindroth, C. H. The ground beetles Press; 1958. 318. p. of Canada and Alaska. Part 6. Opuscula Entomológica Supplemen- Walker, E. M.; Corbet, P. S. The tum34: 945-1192; 1969. Odonata of Canada and Alaska. Vol. 3, part IIL The Anisoptera—three Lindroth, C. H. The ground beetles families. Toronto, ON: University of of Canada and Alaska. Part 1. Toronto Press; 1975. 307 p. Opuscula Entomológica Supplemen- tum. 35: 1-48; 1969. Carabid Beetles Coneworms Arnett, R. H., Jr. The beetles of the United States: a manual for identifica- Rose, A. H.; Lindquist, O. H. In- tion. Washington, DC: Catholic sects of eastern spruces, fir and University of America Press; 1963. hemlock. For. Tech. Rep. 2. Ottawa, 1112 p. ON: Canadian Forestry Service; 1977. 155 p.

63 Lindquist, O. H. Keys to lepidop- Spradbery, J. P. Wasps. An account terous larvae associated with spruce of the biology and natural history of budworm in northeastern North social and solitary wasps. Seattle, America. Sault Ste. Marie, ON: WA: University of Washington Press; Canadian Forestry Service, Great 1973. 416 p. Lakes Forest Research Centre; 1982. 18 p. Fish

Robber Flies Eddy, S. The freshwater fishes. Dubuque, lA: Wm. C. Brown Co.; Curran, C. H. The families and 1969. 286 p. genera of North American Diptera. Mount Vernon, NY: H. Tripp; 1965. Lee, D. S.; Gilbert, C. R.; Hocutt, 515 p. C. H.: Jenkins, R. E.; McAllister, D. E.; Stauffer, J. R. Atlas of Hull, F. M. Robberflies of the North American freshwater fishes. world. The genera of the family Raleigh, NC: North Carolina Asilidae. U.S. National Museum Museum of Natural History; 1980. Bulletin. 224; 1962. 907 p. 854 p.

Ants Birds Creighton, W. S. The ants of North America. Harvard University Bulletin Peterson, R. T. A field guide to the of the Museum of Comparative birds. Boston, MA: Houghton Mif- Zoology. 104; 1950. 585 p. flin; 1980. 384 p.

Wheeler, W. M. Ants, their struc- Scott, S. L., ed. Field guide to the ture, development and behavior. New birds of North America. Washington, York: Columbia University Press; DC: National Geographic Society; 1960. 663 p. 1983. 464 p.

Wasps Mammals Evans, H. E.; Eberhard, M. J. W. Burt, W. H.; Grossenheider, R. P. The wasps. Ann Arbor, MI: Univer- A field guide to the mammals. sity of Michigan Press; 1970. 265 p. Boston, MA: Houghton Mifflin; 1978. 289 p. Krombein, K. V. Trap-nesting wasps and bees. Washington, DC: Smithso- Godin, A. J. Wild mammals of New nian Press; 1967. 576 p. England. Baltimore, MD: Johns Hop- kins University Press; 1977. 304 p.

64 Literature Cited

Akre, Roger D.; Greene, Albert; Biais, J. R. Trends in the frequency, MacDonald, John F.; Landolt, extent, and severity of spruce bud- Peter J.; Davis, Harry G. The worm outbreaks in eastern Canada. yellow jackets of America north of Canadian Journal of Forest Research. Mexico. Agrie. Handb. 552. 13: 539-547; 1983. Washington, DC: U.S. Department of Agriculture; 1980. 102 p. Biais, J. R.; Parks, G. H. Interac- tion of evening grosbeak Barker, R. B.; Fyfe, H. A. A mixed (Hesperiphona vespertina) and spruce infestation in the Spruce Woods budworm {Choristoneura fumiferana Forest Reserve, Manitoba. Bi- (Clem.)) in a localized budworm out- monthly Progress Report 3(4): 3; break treated with DDT in Quebec. 1947. [Ottawa, ON: Dominion De- Canadian Journal of Zoology. 42: partment of Agriculture, Forest Insect 1017-1024; 1964. Investigations.] Boreham, P. F. L. Recent develop- Bell, R.; Whitcomb, W. H. ments in serological methods for Erythemis simplicollis (Say) a dragon- predator-prey studies. Miscellaneous fly predator of the boll worm moth. Publications of the Entomological Florida Entomologist. 44: 95-97; Society of America. 11(4): 17-23; 1961. 1979.

Bennett, C. W. Spiders. In: Annual Borror, Donald J.; DeLong, Dwight technical report., vol. 2, sec. 7.2. M.; Triplehorn, Charles A. An Fredericton, NB: [Canadian Depart- introduction to the study of insects. ment of Resources and Development], 4th ed. New York: Holt, Rinehart Forest Biology Laboratory; 1952a: and Winston; 1976. 852 p. 305. Bruns, H. The economic importance Bennett, C. W. Mites. In: Annual of birds in the forest. Bird Study. 7: technical report., vol. 2, sec. 7.5. 193-208; 1959. Fredericton, NB: [Canadian Depart- ment of Resources and Development], Buckner, Charles H. The role of Forest Biology Laboratory; 1952b: vertebrate predators in the biological 307. control of forest insects. Annual Review of Entomology. 11: 449-470; Bishop, S. C. The Phalangida 1966. (Opiliones) of New York. Pro- ceedings of the Rochester Academy Buckner, Charles H. Avian and of Science. 9: 159-235; 1949. mammalian predators of forest in- sects. Emtomophaga. 12: 491-501; 1967.

65 Buckner, C. H. Vertebrate Clausen, C. P. Entomophagous predators. In: Toward integrated con- insects. New York: McGraw-Hill; trol. Proceedings of the third annual 1940. 688 p. northeastern forest insect work con- ference. Res. Pap. NE-192. Upper Collins, Judith A.; Jennings, Daniel Darby, PA: U.S. Department of T. A simplified holder for eumenid Agriculture, Forest Service, North- nesting blocks (Hymenoptera: eastern Forest Experiment Station; Eumenidae). Entomological News. 1971: 21-31. 95: 58-62; 1984.

Campbell, Robert W.; Torgersen, Coppel, Harry C; Mertins, James Torolf R. Some effects of predaceous W. Biological insect pest suppression. ants on western spruce budworm Advanced series in agricultural pupae in north central Washington. science 4. New York:Springer- Environmental Entomology. 11: Verlag; 1977. 314 p. 111-114; 1982. Coppel, Harry C; Sloan, Norman Campbell, Robert W.; Torgersen, F. Avian prédation, an important Torolf R.; Forrest, Steven C; adjunct in the suppression of larch Youngs, Loma C. Bird exclosures casebearer and introduced pine sawfly for branches and whole trees. Gen. populations in Wisconsin forests. In: Tech. Rep. PNW-125. Portland, OR: Proceedings, Tall Timbers conference U.S. Department of Agriculture, on ecological animal control by Forest Service, Pacific Northwest habitat management. Vol. 2; 1970 Forest and Range Experiment Station, February 26-28; Tallahassee, PL. 1981. 10 p. Tallahassee, PL: Tall Timbers Research Station; 1971: 259-272. Carter, N. E.: Brown, N. R. Seasonal abundance of certain soil Crawford, Hewlette S.; Hooper, arthropods in a fenitrothion-treated R. G.; Titterington, R. W. Song- red spruce stand. Canadian bird population response to silvi- Entomologist. 105: 1065-1073; 1973. cultural practices in central Ap- palachian hardwoods. Journal of Cheshire, W. F. A preliminary Wildlife Management. 45: 680-692; report on the responses of birds to 1981. the Green River spruce budworm infestations, 1949 to 1958 inclusive. Green River annual report, 1958. Fredericton, NB: Canada Department of Agriculture, Forest Biology Division; 1959.

66 Crawford, Hewlette S., Jr.; DeBach, P.; Bartlett, B. R. Jennings, Daniel T. Relationships of Methods of colonization, recovery birds and spruce budworms— and evaluation. In: DeBach, P., ed. literature review and annotated Biological control of insect pests and bibliography. Bibliogr. Lit. Agrie. weeds. New York: Reinhold; 1964: 23. Washington, DC: U.S. Depart- 402-426. ment of Agriculture, Forest Service, Canada/United States Spruce DeBach, Paul; Huffaker, C. B.; Budworms Program; 1982. 40 p. MacPhee, A. W. Evaluation of the impact of natural enemies. In: Crawford, Hewlette S.; Tittering- Huffaker, C. B.; Messenger, P. S., ton, Richard W. Effects of silvi- eds. Theory and practice of biological cultural practices on bird communities control. New York: Academic Press; in upland spruce-fir stands. In: 1976: 255-285. Management of northcentral and northeastern forests for nongame Dempster, J. P. A quantitative study birds: proceedings of the workshop; of the predators on the eggs and 1979 January 23-25; Minneapolis, larvae of the broom beetle, MN. Gen. Tech. Rep. NC-51. St. Phytodecta olivácea Forester, using Paul, MN: U.S. Department of Agri- the precipitin test. Journal of Animal culture, Forest Service, North Central Ecology. 29: 149-167; 1960. Forest Experiment Station; 1979: 110-119. Doganlar, Miktat; Beirne, Bryan P. Natural enemies of budworms, Crawford, H. S.; Titterington, Choristoneura spp. (Lepidoptera: R. W.; Jennings, D. T. Bird préda- Tortricidae), on Douglas-fir near tion and spruce budworm populations. Yale, British Columbia, in 1977. Journal of Forestry. 81: 433-435; Journal of the Entomological Society 1983. of British Columbia. 75: 25-26; 1978. Danks, H. V. Biology of some stem- nesting aculeate Hymenoptera. Tran- Dominion Department of Agricul- sactions of the Royal Entomological ture. Annual technical report. Vol. 2. Society of London. 122: 323-395; Fredericton, NB: Dominion Depart- 1971. ment of Agriculture, Forest Insect Investigations, Dominion Entomo- DeBach, P. Biological control by logical Laboratory; 1950. 556 p. natural enemies. London-New York: Cambridge University Press; 1974. Dowden, Philip B.; Carolin, V. M.; 323 p. Dirks, CO. Natural control factors affecting the spruce budworm in the Adirondacks during 1946-1948. Jour- nal of Economic Entomology. 43: 774-783; 1950. 67 Dowden, Philip B.; Jaynes, H. A.; Francoeur, A. Formicoidea. In: Carolin, V. M. The role of birds Danks, H. V., ed. Canada and its in- in a spruce budworm outbreak in sect fauna. Memoirs of the Entomo- Maine. Journal of Economic logical Society of Canada. 108: Entomology. 46: 307-312; 1953. 502-503; 1979.

Edgar, A. L. Studies on the biology Franz, J. M. Biological control of and ecology of Michigan phalangida pest insects in Europe. Annual (Opiliones). Misc. Publ. 144. Ann Review of Entomology. 6: 183-200; Arbor, MI: University of Michigan, 1961. Museum of Zoology; 1971. 64 p. Freitag, R.; Ozburn, G. W.; Finnegan, R. J. An appraisal of Leech, R. E. The effects of indigenous ants as limiting agents of sumithion and phosphamidon on forest pests in Quebec. Canadian populations of five carabid beetles Entomologist. 103: 1489-1493; 1971. and the spider Trochosa terrícola in northwestern Ontario, including a list Finnegan, R. J. Ants as predators of of collected species of carabid beetles forest pests. Entomophaga. 7: 53-59; and spiders. Canadian Entomologist. 1974. 101: 1328-1333; 1969.

Finnegan, R. J. Introduction of a Freitag, R.; Poulter, F. The effects predaceous red wood ant, Formica of the insecticides sumithion and lugubris (Hymenoptera: Formicidae), phosphamidon on populations of five from Italy to eastern Canada. species of carabid beetles and two Canadian Entomologist. 107: species of lycosid spiders in north- 1271-1274; 1975. western Ontario. Canadian Entomo- logist. 102: 1307-1311; 1970. Finnegan, R. J. Establishment of a predaceous red wood ant, Formica Fye, R. E. Prédation of lepidopterous obscuripes (Hymenoptera: larvae by solitary wasps. Bi-monthly Formicidae), from Manitoba to Progress Reports. 18(2): 2-3; 1962. eastern Canada. Canadian En- [Ottawa, ON: Canadian Department tomologist. 109: 1145-1148; 1977. of Forestry, Forest Entomology and Pathology Branch.] Finnegan, R. J. Prédation by Formica lugubris (Hymenoptera: For- Fye, R. E. The biology of the micidae) on Choristoneura fumiferana Vespidae, Pompilidae, and Sphecidae (Lepidoptera: Tortricidae). Bi- (Hymenoptera) from trap nests in monthly Research Notes. 34(1): 3-4; northwestern Ontario. Canadian 1978. [Ottawa, ON: Fisheries and Entomologist. 97(7): 716-744; 1965. Environment Canada, Canadian Forestry Service.]

68 Fye, R. E. The effect of forest Greenbank, D. O. The analysis of disturbances on populations of wasps moth survival and dispersal in the and bees in northwestern Ontario unsprayed area. In: Morris, R. F., (Hymenoptera: Aculeata). Canadian ed. The dynamics of epidemic spruce Entomologist. 104: 1623-1633; 1972. budworm populations. Memoirs of the Entomological Society of Canada. Gage, S. H.; Miller, C. A. A 31: 87-99; 1963. long-term bird census in spruce budworm-prone balsam fir habitats in Hamilton, W. J., Jr.; Cook, David northwestern New Brunswick. Inf. B. Small mammals and the forest. Rep. M-X-84. Fredericton, NB: Journal of Forestry. 38: 468-477; Department of Fisheries and the 1940. Environment, Canadian Forestry Service; 1978. 6 p. Hardin, Kimberly I.; Evans, Keith E. Cavity nesting bird habitat in the Gage, S. H.; Miller, C. A.; Mook, oak-hickory forest—a review. Gen. L. J. The feeding response of some Tech. Rep. NC-30. St. Paul, MN: forest birds to the black-headed bud- U.S. Department of Agriculture, worm. Canadian Journal of Zoology. Forest Service, North Central Forest 48: 359-366; 1970. Experiment Station; 1977. 23 p.

George, John L.; Mitchell, Robert Haynes, Dean L.; Sisojevic, P. The T. Calculations on the extent of predatory behavior of Grammonota spruce budworm control by insec- angusta Dondale (Araneae: tivorous birds. Journal of Forestry. Erigonidae). Intern. Rep. M-15. 46: 454-455; 1948. Fredericton, NB: Canadian Depart- ment of Forestry, Forest Research Gidaspow, T. North American cater- Laboratory; 1966. 42 p. pillar hunters of the genera Calosoma and Callisthenes (Coleóptera: Hirtle, F. C. Studies on the Carabidae). Bulletin of the American bionomics and natural control of the Museum of Natural History. 116: spruce budworm, (Choristoneura 225-344; 1959. fiimiferana (Clem.)), in the Green River watershed, N.B. In: Annual Glen, R. Factors that affect insect technical report, vol. 3, sec. 4. abundance. Journal of Economic Fredericton, NB: [Canadian Forestry Entomology. 47: 398-405; 1954. Service], Forest Insect Investigations; 1951: 142-161. Graham, S. A.; Orr, L. W. The spruce budworm in Minnesota. Tech. Holling, C. S. The components of Bull. 142. St. Paul, MN: University prédation as revealed by a study of of Minnesota Agriculmral Experiment small-mammal prédation of the Euro- Station; 1940. 27 p. pean pine sawfly. Canadian Entomologist. 91: 293-320; 1959a. 69 HoUíng, es. Some characteristics Jaynes, H. A.; Speers, C. F. of simple types of prédation and Biological and ecological studies of parasitism. Canadian Entomologist. the spruce budworm. Journal of 91: 385-398; 1959b. Economic Entomology. 42: 221-225; 1949. Rolling, C. S. The functional response of invertebrate predators to Jennings, Daniel T. Ants preying on prey density. Memoirs of the En- dislodged jack-pine budworm larvae. tomological Society of Canada. 48: Annals of the Entomological Society 1-86; 1966. of America. 64: 384-385; 1971.

HoUing, C. S.; Buckingham, S. A Jennings, Daniel T.; Crawford, behavioral model of predator-prey Hewlette S. Pine siskin preys on egg functional responses. Behavioral masses of the spruce budworm, Science. 21: 183-195; 1976. Choristoneura fumiferana (Lepidop- tera: Tortricidae). Canadian En- Houseweart, Mark W.; Jennings, tomologist. 115: 439-440; 1983. Daniel T.; Berkett, Lorraine P.; Brann, Thomas B. Parasitic mites Jennings, Daniel T.; Houseweart, (Acari: Erythraeidae) on spruce bud- Mark W. Spider preys on spruce worm moths (Lepidoptera: Tor- budworm egg mass. Entomological tricidae). Canadian Entomologist. News. 89: 183-186; 1978. 112: 193-197; 1980. Jennings, Daniel T.; Houseweart, Houseweart, Mark W.; Southard, Mark W. Prédation by eumenid Susan G.; Jennings, Daniel T. wasps (Hymenoptera: Eumenidae) on Availability and acceptability of spruce budworm (Lepidoptera: spruce budworm eggs to parasitism Tortricidae) and other lepidopterous by the egg parasitoid larvae in spruce-fir forests of Maine. minutum (Hymenoptera: Trichogram- Annals of the Entomological Society matidae). Canadian Entomologist. of America. 77: 39-45; 1984. 114: 657-666; 1982. Jennings, Daniel T.; Houseweart, Hydorn, Susan B.; Rabeni, Charles Mark W.; Cokendolpher, James C. F.; Jennings, Daniel T. Effects of Phalangids (Arachnida: Opiliones) forest spraying with acephate insec- associated with strip clearcut and ticide on consumption of spiders by dense spruce-fir forests of Maine. brook trout (Salvelinus fontinalis). Environmental Entomology. 13: Canadian Entomologist. Ill: 1306-1311; 1984. 1185-1192; 1979.

70 Jennings, Daniel T.; Houseweart, Kirkton, Rodney M. Habitat Mark W.; Dimond, John B. Disper- management and its effects on sal losses of early-instar spruce populations of Polistes and budworm (Lepidoptera: Tortricidae) Iridomyrmex. In: Proceedings, Tall larvae in strip clearcut and dense Timbers conference on ecological spruce-fir forests of Maine. animal control by habitat manage- Environmental Entomology. 12: ment. Vol. 2; 1970 February 26-28; 1787-1792; 1983. Tallahassee, FL. Tallahassee, FL: Tall Timbers Research Station; 1971: Johannsen, O. A. Spruce bud-worm 243-246. (Tortrix fumiferana, Clemens). Bull. 210. Orono, ME: Maine Life Korschgen, Leroy J. Proceedings for Sciences Agricultural Experiment Sta- food-habits analyses. In: Schemnitz, tion; 1913. 31 p. Sanford D., ed. Wildlife management techniques manual. 4th ed., rev. Johnson, Norman E.; Cameron, R. Washington, DC: Wildlife Society; Scott. Phytophagous ground beetles. 1980: 113-127. Annals of the Entomological Society of America. 62: 909-914; 1969. Krall, Jay H. The prédation of litter- dwelling carabidae on larvae of the Kendeigh, S. Charles. Bird popula- spruce budworm, Choristoneura tion studies in the coniferous forest fumiferana, in northern Maine. biome during a spruce budworm Orono, ME: University of Maine; outbreak. Biol. Bull. 1. Ontario, 1977. 92 p. M.S. thesis. Canada: Department of Lands and Forests, Division of Research; 1947. Krall, J. H.; Simmons, Gary A. 100 p. Tree root injection of phosphorus-32 for labeling defoliating insects. Kingsbury, P. D.; Kreutzweiser, Environmental Entomology. 6: D. P. Dosage-effect studies on the 159-160; 1977. impact of permethrin on trout streams. Rep. FPM-X-31. Sault Ste. Krombein, Karl V. Trap-nesting Marie, ON: Department of the wasps and bees: life histories, nests, Environment, Canadian Forestry and associates. Washington, DC: Service, Forest Pest Management Smithso nian Press; 1967. 570 p. Institute; 1980. 51 p. Krombein, Karl V.; Hurd, Paul D., Kiritani, K.; Dempster, J. P. Dif- Jr.; Smith, David R.; Burks, B. D. ferent approaches to the quantitative Catalog of Hymenoptera in America evaluation of natural enemies. Journal north of Mexico. Vol. 2: Apocrita of Applied Ecology. 10: 323-330; (Aculeata). Washington, DC: 1973. Smithsonian Institution Press; 1979: 1472-1510.

71 Kulman, H. M. Comparative Macdonald, D. R.; Webb, F. E. ecology of North American Carabidae Insecticides and the spruce bud worm. with special reference to biological In: Morris, R. P., ed. The dynamics control (1). Entomophaga. 7: 61-70; of epidemic spruce budworm popula- 1974. tions. Memoirs of the Entomological Society of Canada. 31: 228-310; Lawson, F. R.; Rabb, R. L.; 1963. Guthrie, F. E.; Bowery, T. G. Studies of an integrated control MacKay, Margaret R. The spruce system for horn worms on tobacco. foliage worm and the spruce cone Journal of Economic Entomology. 54: worm (Dioryctria spp., Lepidoptera, 93-97; 1961. Pyralidae). Canadian Entomologist. 75: 91-98; 1943. Leius, K. Attractiveness of different foods and flowers to the adults of Mclver, James D. An examination some hymenopterous parasites. of the utility of the precipitin test for Canadian Entomologist. 92: 369-376; evaluation of predator-prey 1960. relationships. Canadian Entomologist: 113: 213-222; 1981. Lindroth, C. H. The ground beeties of Canada and Alaska. Part I. McLeod, J. M.; Daviault, L. Notes Opuscula Entomológica Supplemen- on the life history and habits of the tum. 35: 1-48; 1969. spruce cone worm, Dioryctria reniculella (Grt.) (Lepidoptera: Liscombe, E. A. R.; Lejeune, R. R. Pyralidae). Canadian Entomologist. Natural control of the spruce bud- 95: 309-316; 1963. worm. Bimonthly Progress Reports 5(6): 2; 1949. [Ottawa, ON: Domin- McNeil, Jeremy N.; Delisle, ion Department of Agriculture, Johanne; Finnegan, R. J. Seasonal Science Service, Division of Ento- predatory activity of the introduced mology, Forest Insect Investigations.] red wood ant, Formica lugubris (Hymenoptera: Formicidae) at Valcar- Loughton, B. G.; Derry, C; West, tier, Quebec, in 1976. Canadian En- A. S. Spiders and the spruce bud- tomologist. 110: 85-90; 1978. worm. In: Morris, R. P., ed. The dynamics of epidemic spruce bud- Medler, J. T.; Fye, R. E. Biology worm populations. Memoirs of the of Ancistrocerus antilope (Panzer) Entomological Society of Canada. 31: (Hymenoptera, Vespidae) in trap- 249-268; 1963. nests in Wisconsin. Annals of the Entomological Society of America. 49(1): 97-102; 1956.

72 Medier, J. T.; Koerber, T. Biology Mock, L. J. Birds and the spruce of Dipogon sayi sayi Banks in trap budworm. In: Morris, R. F., ed. The nests in Wisconsin. Annals of the dynamics of epidemic spruce bud- Entomological Society of America. worm populations. Memoirs of the 50: 621-625; 1957. Entomological Society of Canada. 31: 268-271; 1963. Miller, C. A. The measurement of spruce budworm populations and Morimoto, R. Polistes wasps as mortality during the first and second natural enemies of agricultural and larval instars. Canadian Journal of forest pests. 1. Studies on the social Zoology. 36: 409-422; 1958. Hymenoptera of Japan 10. Science Bulletin of the Faculty of Agriculture, Miller, C. A. The spruce budworm. Kyushu University. 18: 110-116; In: Morris, R. F., ed. The dynamics 1960. of epidemic spruce budworm popula- tions. Memoirs of the Entomological Morris, R. F. Green River Project- Society of Canada. 31: 12-19; summary of entomological investiga- 1963a. tions in 1947. Bimonthly Progress Reports. 4(1): 1-2; 1948. [Ottawa, Miller, C. A. The analysis of pupal ON: Dominion Department of Agri- survival in the unsprayed area. In: culture, Science Service—Division of Morris, R. P., ed. The dynamics of Entomology, Forest Insect epidemic spruce budworm popula- Investigations.] tions. Memoirs of the Entomological Society of Canada. 31: 63-70; Morris, R. F. The development of 1963b. sampling techniques for forest insect defoliators, with particular reference Miller, C. A.; Varty, I. W. Biolog- to the spruce budworm. Canadian ical methods of spruce budworm con- Journal of Zoology. 33: 107-223; trol. In: The spruce budworm. 1955. Forestry Chronicle. 51: 150-153; 1975. Morris, R. F. Prédation and the spruce budworm. In: Morris, R. F., Miller, M. C. Preparatory im ed. The dynamics of epidemic spruce munodiffusion for production of budworm populations. Memoirs of specific anti-adult southern pine the Entomological Society of Canada. beetle serum. Annals of the 31: 244-248; 1963. Entomological Society of America. 72: 820-825; 1979. Morris, R. F.; Cheshire, W. F.; Miller, C. A.; Mott, D. G. The Mitchell, Robert T. Consumption of numerical response of avian and spruce budworms by birds in a Maine mammalian predators during a spruce-fir forest. Journal of Forestry. graduation of the spruce budworm. 50: 387-389; 1952. Ecology. 39: 487-494; 1958. 73 Morris, R. F.; Mott, D. G. Disper- Otvos, I. S. Predators. In: Hudak, sal and the spruce bud worm. In: J.; Raske, A. G., eds. Review of the Morris, R. F., ed. The dynamics of spruce budworm outbreak in epidemic spruce budworm popula- Newfoundland—its control and forest tions. Memoirs of the Entomological management implications. Inf. Rep. Society of Canada. 31: 180-189; N-X-205. St. John's, NF: Environ- 1963. ment Canada, Canadian Forestry Service; 1981: 103-104. Mott, D. G. The analysis of the survival of small larvae in the Price, Peter W. Insect ecology. New unsprayed area. In: Morris, R. F., York: John Wiley & Sons; 1975. ed. The dynamics of epidemic spruce 514 p. budworm populations. Memoirs of the Entomological Society of Canada. Reeves, R. Marcel; Dunn, Gary A.; 31: 42-52; 1963. Jennings, Daniel T. Carabid beetles (Coleóptera: Carabidae) associated Neilson, M. M. The analysis of egg with the spruce budworm, Choristo- survival in the unsprayed area. In: neura fumiferana (Lepidoptera: Tor- Morris, R. F., ed. The dynamics of tricidae). Canadian Entomologist. epidemic spruce budworm popula- 115: 453-472; 1983. tions. Memoirs of the Entomological Society of Canada. 31: 37-41; 1963. Renault, T. R.; Miller, C. A. Spiders in a fir-spruce biotype: abun- Nicholson, A. J. The balance of dance, diversity, and influence on animal populations. Journal of spruce budworm densities. Canadian Animal Ecology. 2 (Supplement 1): Journal of Zoology. 50: 1039-1046; 132-178; 1933. 1972.

Nicholson, A. J. An outline of the Rose, A. H.; Lindquist, O. H. dynamics of animal populations. Insects of eastern spruces, fir and Australian Journal of Zoology. 2: hemlock. For. Tech. Rep. 23. 9-65; 1954. Ottawa, ON: Department of the Environment, Canadian Forestry Odum, E. P.; GoUey, F. B. Service; 1977. 159 p. Radioactive tracers as an aid to the measurement of energy flow at the Royama, T. A comparative study of population level in nature. In: models for prédation and parasitism. Schultz, v.; Klement, A. W., eds. Researches on Population Ecology. Radioecology. New York: Rheinhold 13 (Supplement 1): 1-91; 1971 Publ. Corp.; 1963: 403-410. Royama, T. Population persistence and density dependence. Ecological Monographs. 47(1): [reprint p. 1-35]; 1977. 74 Sailer, R. I. Invertebrate predators. Stern, Vernon M.; Adkisson, Perry In: Toward integrated control. Pro- L.; Beingolea, Oscar G.; Viktorov, ceedings of the third annual north- G. A. Cultural controls. In: Huf- eastern forest insect work conference. faker, C. B.; Messenger, P. S., eds. Res. Pap. NE-94. Upper Darby, PA: Theory and practice of biological U.S. Department of Agriculture, control. New York: Academic Press; Forest Service, Northeastern Forest 1976: 593-613. Experiment Station; 1971: 32-44. Syme, P. D. The effect of wild Sanders, C. J., comp. A summary carrot on a common parasite of the of current techniques used for sampl- European pine shoot moth. Bimonthly ing spruce budworm populations and Research Notes. 22(4): 3; 1966. estimating defoliation in eastern [Ottawa, ON: Canadian Department Canada. Rep. O-X-306. Sault Ste. of Forestry.] Marie, ON: Department of the Environment, Canadian Forestry Syme, P. D. The effects of flowers Service, Forest Research on the longevity and fecundity of two Centre; 1980. 33 p. native parasites of the European pine shoot moth in Ontario. Environmental Sanders, C. J.; van Franken- Entomology. 4: 337-346; 1975. huyzen, K. High populations of a carabid beetle associated with spruce Thomson, H. M. A note on the budworm. Bi-monthly Research prédation of spruce budworm pupae. Notes. 35(4): 21-22; 1979. [Ottawa, Bi-monthly Progress Reports. 13(4): ON: Environment Canada, Canadian 2; 1957. [Ottawa, ON: Canadian Forestry Service.] Department of Agriculture, Science Service, Forest Biology Division.] Smith, B. C. Variation in weight, size, and sex ratio of coccinellid Tinbergen, L. The natural control of adults (Coleóptera: Coccinellidae). insects in pine woods. 1. Factors Canadian Entomologist. 98: 639-644; influencing the intensity of prédation 1966. by songbirds. Archives Néerlandaises de Zoologie. 13: 265-336; 1960. Solomon, M. E. The natural control of animal populations. Journal of Titterington, R. W.; Crawford, Animal Ecology. 18: 1-35; 1949. H. S.; Burgason, B. N. Songbird responses to commercial clear-cutting Southwood, T. R. E. Ecological in Maine spruce-fir forests. Journal methods, with particular reference to of Wildlife Management. 43: the study of insect populations. 2d 602-609; 1979. ed., rev. London: Chapman and Hall; 1978. 524 p.

75 Todd, Valerie. Prey of harvestmen Varty, I. W. Long-term effects of (Arachnida, Opiliones). The fenitrothion spray programs on non- Entomologist's Monthly Magazine. target terrestrial arthropods. 86: 252-254; 1950. Proceedings of the symposium on fenitrothion; 1977 April 20-22. Tothill, J. D. Notes on the outbreaks Ottawa, ON: National Research of spruce budworm, forest tent cater- Council Canada; 1977: 343-375. pillar, and larch sawfly in New Brunswick. Proceedings of the Varty, I. W. Forest tree insects and Acadian Entomological Society. 8: spiders. In: Environmental surveil- 172-182; 1923. lance in New Brunswick, 1978-1979, effects of spray operations for forest Treat, Asher E. Mites of moths and protection against spruce budworm. butterflies. Ithaca and London: Cor- Fredericton, NB: Department of nell University Press; 1975. 362 p. Forest Resources, University of New Brunswick; 1980: 44-46. Tsomides, Leónidas; Gibbs, K. Elizabeth; Jennings, Daniel T. Varty, I. W.; Carter, N. E. A Species of Odonata feeding on baseline inventory of litter-dwelling Lepidoptera in spruce-fir forests of arthropods and airborne insects Maine. Research in the Life Sciences. including pollinators in two fir-spruce 39(1): 1-12; 1982. Orono, ME: stands with dissimilar histories of University of Maine, Life Sciences insecticide treatment. Inf. Rep. M- and Agriculture Experiment Station. X-48. Fredericton, NB: Department of the Environment, Canadian TurnbuU, A. L. Spider predators of Forestry Service, Maritimes Forest the spruce budworm, Choristoneura Research Centre; 1974. 32 p. fumiferana (Clem.), at Lillooet, B.C., Canada. Vol. 3A. Proceedings of the Varty, I. W.; Titus, F. A. Effects Eighth Pacific Science Congress, of phosphamidon sprays on non-target 1956: 1579-1593. insects in fir-spruce forest, spruce budworm aduldcide trials 1973. Inf. TurnbuU, A. L. The prey of the Rep. M-X-47. Fredericton, NB: spider Linyphia triangularis (Clerck), Department of the Environment, (Araneae, Linyphiidae). Canadian Canadian Forestry Service, Maritimes Journal of Zoology. 38: 859-873; Forest Research Centre; 1974. 24 p. 1960.

Varty, I. W. Ecology of Mulsantina hudsonica Casey, a ladybeetle in fir- spruce forest. Intern. Rep. M-42. Fredericton, NB: Canadian Depart- ment of Fisheries and Forestry, Forestry Branch; 1969. 28 p. 76 Warren, G. L. The spruce needle worm, Dioryctria reniculella Grt., as a predator of spruce bud- worm. Bi-monthly Progress Reports. 10(3): 2-3; 1954. [Ottawa, ON: Canadian Department of Agriculture, Science Service, Forest Biology Division.]

Watt, K. E. F. A mathematical model for the effect of densities of attacked and attacking species on the number attacked. Canadian Entomologist. 91: 129-144; 1959.

Watt, K. E. F. The analysis of the survival of large larvae in the unsprayed area. In: Morris, R. F., ed. The dynamics of epidemic spruce budworm populations. Memoirs of the Entomological Society of Canada. 31: 52-63; 1963.

Zach, Reto; Falls, J. Bruce. Response of the ovenbird (Aves: Parulidae) to an outbreak of the spruce budworm. Canadian Journal of Zoology. 53: 1669-1672; 1975.

77