Non-Target Effect of Imidacloprid on the Predatory Arthropod Guild on Eastern Hemlock, Tsuga Canadensis (L.) Carriere, in the Southern Appalachians

Home , Anthicidae, Lambdina, Lambdina fiscellaria, Mantispidae

University of Tennessee, Knoxville TRACE: Tennessee Research and Creative Exchange

Masters Theses Graduate School

5-2008

Non-Target Effect of Imidacloprid on the Predatory Arthropod Guild on Eastern Hemlock, Tsuga canadensis (L.) Carriere, in the Southern Appalachians

Abdul Hakeem University of Tennessee - Knoxville

Follow this and additional works at: https://trace.tennessee.edu/utk_gradthes

Part of the Plant Pathology Commons

Recommended Citation Hakeem, Abdul, "Non-Target Effect of Imidacloprid on the Predatory Arthropod Guild on Eastern Hemlock, Tsuga canadensis (L.) Carriere, in the Southern Appalachians. " Master's Thesis, University of Tennessee, 2008. https://trace.tennessee.edu/utk_gradthes/360

This Thesis is brought to you for free and open access by the Graduate School at TRACE: Tennessee Research and Creative Exchange. It has been accepted for inclusion in Masters Theses by an authorized administrator of TRACE: Tennessee Research and Creative Exchange. For more information, please contact [email protected]. To the Graduate Council:

I am submitting herewith a thesis written by Abdul Hakeem entitled "Non-Target Effect of Imidacloprid on the Predatory Arthropod Guild on Eastern Hemlock, Tsuga canadensis (L.) Carriere, in the Southern Appalachians." I have examined the final electronic copy of this thesis for form and content and recommend that it be accepted in partial fulfillment of the requirements for the degree of Master of Science, with a major in Entomology and Plant Pathology.

Jerome F. Grant, Major Professor

We have read this thesis and recommend its acceptance:

Paris L. Lambdin, Frank A. Hale, James R. Rhea

Accepted for the Council: Carolyn R. Hodges

Vice Provost and Dean of the Graduate School

(Original signatures are on file with official studentecor r ds.) To the Graduate Council:

I am submitting herewith a thesis written by Abdul Hakeem entitled “Non-Target Effect of Imidacloprid on the Predatory Arthropod Guild on Eastern Hemlock, Tsuga canadensis (L.) Carriere, in the Southern Appalachians.” I have examined the final electronic copy of this thesis for form and content and recommend that it be accepted in partial fulfillment of the requirements for the degree of Master of Science, with a major in Entomology and Plant Pathology.

______Jerome F. Grant, Major Professor

We have read this thesis and recommend its acceptance:

______Paris L. Lambdin

______Frank A. Hale

______James R. Rhea

Accepted for the Council:

______Carolyn R. Hodges, Vice Provost and Dean of the Graduate School

(Original signatures are on file with official student records) “Non-Target Effect of Imidacloprid on the Predatory Arthropod Guild

on Eastern Hemlock, Tsuga canadensis (L.) Carriere,

in the Southern Appalachians”

A Thesis Presented for the

Master of Science

Degree

The University of Tennessee, Knoxville

Abdul Hakeem

May 2008

DEDICATION

the Prince Karim Aga Khan, 49th Hereditary Imam of Shia Imami Ismaili Muslims for His Untiring Efforts to Improve Socio-Economic Conditions of the People of Developing Countries Without Discriminating Race, Color, Sex or Religion

iii

ACKNOWLEDGEMENTS

I express sincere appreciation to all of those who assisted me in this research project. I would like to thank my major advisor, Dr. Jerome Grant, for his guidance, invaluable support, patience, statistical support and for his time and knowledge. His open door policy helped me to meet my deadlines. I extend my heartiest thanks to my committee members, Dr. Paris Lambdin, Dr. Frank Hale and Rusty Rhea, for their input, valuable suggestions and guidance.

I am thankful to Greg Wiggins, for his guidance from my first day at the

University of Tennessee and also for his help during identification of predators. I am thankful to Renee Follum for also helping in identification of predators, David

Paulsen and Toni Conatser for field assistance, Carla Dilling for her patience during field visits, Nick Reynolds for his company, and Coleman Timberlake for sorting specimens.

I would also like to thank the management personnel at the Tellico Plains

Ranger Station for their assistance. I am thankful to Dr. Ann Reed for statistical analysis.

For me, this research project would have been a nightmare without funding from Fulbright and U.S.D.A. Forest Service. I am thankful to Department of Agriculture, Gilgit, and Government of Pakistan for allowing me to study in the

United States.

At last, but not least, I am also thankful to my parents and family members who supported me throughout life.

ABSTRACT

Imidacloprid, a neonicotinoid insecticide, is commonly applied on eastern hemlock to reduce populations of Hemlock woolly adelgid (HWA). A large number of other herbivorous and transient insects also are associated with eastern hemlock. These herbivorous insects may acquire imidacloprid through feeding on treated plants. Predatory insects may acquire imidacloprid when they feed on insecticide-contaminated prey. To investigate this phenomenon, a study was conducted at Indian Boundary Campground, Cherokee National Forest,

2005-2007. This study was conducted to: 1) ascertain the effect of imidacloprid used against HWA on the predatory guild associated with eastern hemlock, 2) determine seasonal abundance of the predatory guild on eastern hemlock, and

3) assess influence of vertical stratification on spiders and other predators.

During this study, 4,917 predators representing 75 families and 10 orders were collected. Spiders were the most dominant predator group, and the most abundant spider families were Mimetidae (1,038), Salticidae (736), Araneidae

(733), Gnaphosidae (517), Philodromidae (330), Theridiidae (168),

Tetragnathidae (161) and Thomisidae (142). The most abundant insect predator families were Vespidae (132), Ichneumonidae (50), Braconidae (31),

Pentatomidae (25), Reduviidae (24), Coccinellidae (15), and Elateridae (15).

Predator densities were not significantly different between pesticide application times (Fall and Spring). In both years, predator densities in control treatments and horticultural oil treatments were significantly (p<0.05) greater than those in imidacloprid treatments. However, predator densities were not significantly v

(p<0.05) different among soil drench, soil injection, and tree injection treatments or between control and horticultural oil treatments. Predator densities were at least 1.5-3X greater in the imidacloprid-treated plots in 2007 than in 2006, possibly suggesting a rebound in predator densities 1-1½ years after treatment.

Predator densities were significantly (p<0.05) greater in the top and middle canopy than in the lower canopy. Imidacloprid concentration level declined progressively from the bottom stratum to the top stratum of the tree canopy.

Highest levels were observed in the bottom stratum which shows that higher concentrations of imidacloprid lead to lower numbers of predators and lower concentrations of imidacloprid lead to higher numbers of predators.

TABLE OF CONTENTS

Chapter Page

I. Literature Review ...... 1 Eastern Hemlock ...... 1 Hemlock Woolly Adelgid ...... 7 Importance of Predators ...... 18 Non-target Impact of Imidacloprid on Predators ...... 20 Research Objectives ...... 23 II. Effect of Imidacloprid and Horticultural Oil on Abundance and Diversity of Non- Target Predators Associated with Eastern Hemlock, Tsuga canadensis (L.) Carriere ...... 24 Introduction ...... 24 Materials and Methods ...... 26 Results and Discussions ...... 35 Summary...... 49 III. Effect of Vertical Stratification on Predators Associated with Eastern Hemlock, Tsuga canadensis (L.) Carriere ...... 51 Introduction ...... 51 Materials and Methods ...... 52 Results and Discussions ...... 54 Summary...... 61 IV. Conclusions ...... 63 Literature Cited ...... 65 Appendix ...... 74 Vita ...... 78

vii

LIST OF TABLES

Table Page

1. Eastern hemlock as a minor component of 18 forest types (Godman and Lancaster 2003)………………………………………4

2. Pre-treatment and post-treatment infestation of hemlock woolly adelgid in Cherokee National Forest, Tennessee…………………………………………………………….28

3. Abundance of spiders and opilionids collected from eastern hemlock, Indian Boundary Campground, Cherokee National Forest, Tennessee, 2006-2007 ………………………………………….39

4. Abundance of insect predators collected from eastern hemlock, Indian Boundary Campground, Cherokee National Forest, Tennessee, 2006-2007……………………………………………………………..41

viii

LIST OF FIGURES

Figure Page

1. Native range of eastern hemlock, (T. canadensis) and Carolina hemlock (T. caroliniana) in North America (Thompson et al. 2006)……..2

2. Distribution of the hemlock woolly adelgid, Adelges tsugae Annand in the eastern United States (USDA 2006)………………………………….8

3. HWA life stages: A) crawler (1st instar nymph), and B) winged adult…………………………………………………………………………….10

4. Study site: A). Indian Boundary/Recreational Area, B) Cherokee National Forest, Tennessee……………………………………………………………………...26 5. Imidacloprid applied as a tree injection……………………………………..29 6. Imidacloprid applied as a soil injection………………………………………29 7. Imidacloprid applied as a soil drench………………………………………..30

8. Horticultural oil applied as a foliar spray. …………………………………...30

9. High pressure hydraulic sprayer used for foliar spray and drench …….32 10. Beat-sheet sampling from hemlock trees …………………………………..33 11. Modified malaise traps on hemlock trees……………………………..…….34 12. Abundance of predators on eastern hemlock at Indian Boundary Campground, Cherokee National Forest, Tennessee, 2006…………...…37

13. Abundance of predators on eastern hemlock at Indian Boundary Campground, Cherokee National Forest, Tennessee, 2007……………...37

14. Abundance of predators on eastern hemlock at Indian Boundary Campground, Cherokee National Forest, Tennessee, 2006…………...…43

15. Abundance of predators on eastern hemlock at Indian Boundary Campground, Cherokee National Forest, Tennessee, 2007…………...…43

16. Influence of pesticide application time (Fall or Spring) on predator abundance Indian Boundary Campground, Cherokee National Forest, Tennessee, 2006……………………………………………………………....44

17. Influence of pesticide application time (Fall or spring) on predator abundance at Indian Boundary Campground, Cherokee National Forest, Tennessee, 2007……………………………………………………………....45

18. Influence of pesticide treatments on monthly abundance of predators on eastern hemlock, Indian Boundary, Cherokee National Forest, Tennessee, 2006, Fall treatments, B) Spring Treatments….…………….47

19. Influence of pesticide treatments on monthly abundance of predators on eastern hemlock, Indian Boundary Campground, Cherokee National Forest, TN, 2007, A) Fall treatments, B) Spring treatments………………………………………………………………………48

20. Vertical sampling of the canopy of hemlock tree using a bucket lift truck………………………………………………………………………….….53

21. Influence of strata on predator densities on eastern hemlock, Indian Boundary Campground, Cherokee National Forest, Tennessee, 16-19 August 2006…………………………………………………………….55

22. Influence of strata on predator densities on eastern hemlock, Indian Boundary Campground, Cherokee National Forest, Tennessee, 5-11 September 2007…………………………………….…….55

23. Influence of time of pesticide application, averaged across all three strata and pesticide treatments, on predator densities on eastern hemlock, Indian Boundary Campground, Cherokee National Forest, Tennessee, A) 16-19 August 2006, B) 5-11 September 2007……….…..57

24. Influence of pesticide treatments, averaged across all strata and pesticide application times, on predator densities on eastern hemlock, Indian Boundary, Cherokee National Forest, Tennessee, 16-19 August 2006………………………………………………………….…58

25. Influence of pesticide treatments, averaged across all strata and pesticide application times, on predator densities on eastern hemlock, Indian Boundary, Cherokee National Forest, Tennessee, 5-11 September 2007………………………………………………………...58

26. Influence of strata on predator densities among pesticide treatments on Fall-treated eastern hemlock, Indian Boundary Cherokee National Forest, Tennessee, 16-19 August 2006………………………………….…59

27. Influence of strata on predator densities among pesticide treatments on spring-treated eastern hemlock, Indian Boundary Campground, Cherokee National Forest, Tennessee, 16-19 August 2006……………...60

28. Influence of strata on predator densities among pesticide treatments on Fall-treated eastern hemlock, Indian Boundary Cherokee National Forest, Tennessee, 5-11 September 2007…………………...... 60

29. Influence of strata on predator densities among pesticide treatments on Spring-treated eastern hemlock, Indian Boundary Cherokee National Forest, Tennessee, 5-11 September 2007…………………………………61

I. Literature Review

Eastern Hemlock

Distribution and Range of Hemlock (Tsuga spp.)

At least 9-13 Tsuga species are known worldwide, and four of these hemlock species are native to the United States. The mountain hemlock, Tsuga mertensiana Carriere, and the western hemlock, T. heterophylla Sargent, occur in the western United States. The Carolina hemlock, T. caroliniana Engelmann, which is limited to small areas in North Carolina and South Carolina, and the eastern hemlock, T. canadensis (L.), are found in the eastern United States and adjacent Canadian provinces (Fig. 1) (Welch and Haddow 1993, Robert 2001).

Eastern hemlock, which ranges from Nova Scotia to Georgia and westward to Minnesota (Fig. 1) (Ward et al. 2004), is found from sea level to

1,520 m. In the southern Appalachians, eastern hemlock is mostly found from

610 to 1,520 m (Godman and Lancaster 2003) and grows on north and east facing slopes (Ward et al. 2004). Eastern hemlock is generally confined to areas with a cool climate (-12.2 Cº in winter and 16.0 Cº in summer). Precipitation in these areas ranges from 740 mm to 1,270 mm per year (Godman and Lancaster

2003, Ward et al. 2004). Eastern hemlock grows well in moist and well drained soils but also grows in shallow to bedrock drained soils (Yamasaki et al. 1999).

Sandy loam, loamy sands and silt loams with coarse rocky material also are suitable for hemlock development.

In the United States, eastern hemlock accounts for 7% of timberland area

Fig. 1. Native range of eastern hemlock (T. canadensis) and Carolina hemlock (T. caroliniana) in North America (Thompson et al. 2006).

in northeastern states and 57% in northcentral states (Robert 2001). Hemlock is an important component of the eastern forest ecosystem due to its economic, ecological, and aesthetic value (Table 1) (Wallace and Hain 2000). Eastern hemlock, also referred to as Canada hemlock (Ward et al. 2004), is long lived

(988 years or more) and may take 250 to 300 years to reach maturity. The largest eastern hemlock tree recorded has a diameter of 213.0 cm and is 49.0 m tall (Godman and Lancaster 2003).

In North America, eastern hemlock is associated with 29 forest cover types (Eyre 1980, Godman and Lancaster 2003). It is the dominant tree species in four forest cover types [white pine-hemlock (Type 22), eastern hemlock (Type

23), hemlock-yellow birch (Type 24) and yellow-poplar-eastern hemlock (Type

58)]. It is a common component in seven other forest cover types [white pine- northern red oak-red maple (Type 20), eastern white pine (Type 21), red spruce- 2

yellow birch (Type 30), red spruce-sugar maple-beech (Type 31), red spruce

(Type 32), red spruce-balsam fir (Type 33) and red spruce-fraser fir (Type 34)]. In addition, eastern hemlock occurs as a minor component in 18 forest cover types

(Table 1) (Godman and Lancaster 2003).

Biology of Eastern Hemlock

Eastern hemlock grows on ca. 7.6 million hectares of forests in the United

States. Maximum photosynthetic rate occurs during the winter which helps hemlock to develop under the canopies of deciduous hardwoods (Ward et al.

2004). Flowers of eastern hemlock are found in separate clusters, and flowering time lasts from late April to early June. Pollen, which is usually dispersed by the wind, is sensitive to weather conditions and may dry quickly. Fertilization is complete in about 6 weeks, cones open fully in mid-October, and seed dispersal extends into the winter. Eastern hemlock is one of the most frequent cone producers among the eastern conifers. The seed is about 1.6 mm long with a slightly longer terminal (Godman and Lancaster 2003).

Stress Absorption

Eastern hemlock is the most shade tolerant species in eastern forests and can survive with as little as 5% full sunlight. Hemlock does not require as much space as hardwoods. While thinning, no more than one-third of the total basal area of hemlock should be removed at one time. Excessive cutting may contribute to reduced growth and increased mortality (Godman and Lancaster 2003).

Table 1. Eastern hemlock as a minor component of 18 forest types (Godman and Lancaster 2003)*

Type Number Forest Type

5 Balsam Fir

17 Pin Cherry

18 Paper Birch

25 Sugar Maple-Beech-Yellow Birch

26 Sugar Maple-Basswood

27 Sugar Maple

28 Black Cherry-Maple

35 Paper Birch-Red Spruce-Balsam Fir

37 Northern White-Cedar

39 Black Ash-American Elm-Red Maple

44 Chestnut Oak

52 White Oak-Black Oak-Northern Red Oak

53 White Oak

57 Yellow-Poplar

59 Yellow-Poplar-White Oak-Northern Red Oak

60 Beech-Sugar Maple

97 Atlantic White-Cedar

108 Red Maple

* Forest types recognized by the Society of American Foresters.

Importance of Hemlock

The dense, evergreen canopy of hemlock forests provides a unique habitat for many animal and plant species and is important in augmenting terrestrial habitat diversity in forest ecosystems (Godman and Lancaster 2003). It supports more than 120 vertebrate species and also provides thermal cover and forage for a variety of mammals and birds (Ward et al. 2004). Besides these, many aquatic species, such as trout, are found in streams sheltered by eastern hemlock trees. Eastern hemlock shaded streams have lower summer temperatures and are less likely to dry (Ward et al. 2004, Johnson et al. 2005).

Eastern hemlock, also grown as an ornamental plant (Godman and Lancaster

2003), is valued for its aesthetic appeal. Its presence in forest ecosystems supports outdoor recreation and provides habitat for wildlife and fish.

Primary uses of eastern hemlocks between 1890 and 1910 in the United

States were light framing, sheathing, roofing, subflooring, boxes, crates and general millwork. It was also used as a source of tannin in the leather industry.

Currently, eastern hemlock contributes to ca. 22% of the total volume of softwood growing stocks in the northeastern United States. Eastern hemlock is used in the pulp and paper industry, and as lumber and mulch. It is also used in structures/frames of shelters and mostly used as raw wood (Ward et al. 2004).

Damaging Agents of Hemlock

A total of 293 insect species representing 226 genera, 101 families and 10 orders has been collected from eastern hemlock. Among these, 92% of the insects found were canopy-dwelling species comprising eight guilds which

include transient, scavenger, predator, detritivore, phytophagous, parasitoid, haematophagous, and fungivore (Buck et al. 2005, Dilling et al. 2007). Insect orders collected include Coleoptera, Diptera, Lepidoptera, Hymenoptera and

Hemiptera. Buck et al. (2005) also documented the incidence of elongate hemlock scale, Fiorinia externa Ferris (Hemiptera: Diaspididae), on eastern hemlock.

Eastern hemlock is fed upon by at least 24 insect species. Among these, the hemlock borer, Melanophila fulvoguttata (Harris) (Coleoptera: Buprestidae) bores into the trunk producing holes in the bark and weakening trees. Galleries are filled with dark excrement and shoot tips become yellow (Godman and

Lancaster 2003). The hemlock looper, Lambdina fiscellaria fiscellaria (Guenee)

(Lepidoptera: Geometridae), feeds on needles which later turn brown. The hemlock scale, Abgrallaspis ithacae (Ferris) (Hemiptera: Diaspididae), damages young trees. Besides insect pests, several diseases affect eastern hemlock. The rust caused by Melampsora farlowii (Arthur) is among the most damaging diseases (Godman and Lancaster 2003). These pests and diseases contribute to weakening eastern hemlock trees and possibly make them more susceptible to feeding by other pests, such as hemlock woolly adelgid (HWA), Adelges tsugae

Annand (Hemiptera: Adelgidae). By excessive feeding, HWA indirectly effects structure, composition, and ecosystem function in hemlock-dominated forests

(Kizlinski et al. 2002).

Hemlock Woolly Adelgid

Origin and Distribution

Eastern hemlock populations have dramatically declined in the eastern

United States because of defoliation caused by HWA (Keller 2004, Wise 2006).

This exotic hemipteran was accidentally introduced into western North America from Asia in 1920. Two native species of hemlock, T. heterophylla and T. mertensiana, are believed to be resistant to HWA because populations do not reach damaging levels on these two tree species. Because of this resistance and possible other factors, HWA did not become a problem on hemlocks in this area.

HWA was then discovered in the eastern United States near Richmond, Virginia in 1951 (Orwig et al. 2002, Ward et al. 2004). A. tsugae can feed on all species of hemlock, but feeding causes the highest mortality to T. canadensis and T. caroliniana.

Eastern hemlock and Carolina hemlock of all sizes and ages are susceptible to

HWA infestation at all growth stages (Wise 2006, Kohler 2007). Since its introduction into the eastern United States, HWA has spread to most of the hemlock growing areas of the eastern United States and Canada (Fig. 2). Heavy infestations of the adelgid on eastern and Carolina hemlocks can result in premature drop of foliage, bud abortion, and death within 2 to 12 years after infestation (Ward et al. 2004). Due to decreased populations of living eastern hemlock trees in highly infested areas, the aesthetic value of forest landscapes is diminishing. HWA infestations on eastern hemlock in Virginia, New Jersey and

Fig. 2. Distribution of the hemlock woolly adelgid, Adelges tsugae Annand, in the eastern United States (USDA 2006).

Connecticut have been severe with mortality ranging from 42-90% among stands

(SAMAB 2005).

Biology and Lifecycle of Hemlock Woolly Adelgid

The HWA is a small (ca 1.5 mm long), dark reddish-brown to purplish- black insect with piercing-sucking mouthparts. Nymphs and adults produce a woolly secretion or mass which protects eggs and aldegids from desiccation and from natural enemies and also protects from pesticide contact. Except for active crawlers, all subsequent stages including adults are covered with the wool-like material. The eggs, contained in an “ovisac”, hatch into first instars (crawlers) which insert their stylets into the plant at the base of needles and feed on sap

(Ward et al. 2004, Deal 2007).

In Tennessee, the HWA has a bivoltine lifecycle and has two peaks of abundance in the parthenogenic generations on eastern hemlock (sistens, progrediens and sexaparae). The lifecycle of HWA is slightly different in

Tennessee (Deal 2007) than in Connecticut (McClure 1989). In Tennessee, sistens are present throughout the year except May, mature into adults (Fig. 3B) in January and begin oviposition in February. After hatching in March or April, reddish-brown first-instar nymphs or crawlers (Fig. 3A) seek feeding sites and most often settle on the new growth. HWA become adults in late May to early

June. In Connecticut, sistens are present throughout the year except May and

June. In March, sistens begin to oviposit while progrediens start to oviposit in

June (Ward et al. 2004). Sistens begin aestivation in July and end in October

A B

Fig. 3. HWA life stages: A) crawler (1st instar nymph), and B) winged adult.

which is the same as documented in Tennessee (Charles 2002, Ward et al.

2004, Deal 2007, PDCNR 2007) while progrediens begin to lay eggs in April and the peak emergence of sexaparae occurs in May (Ward et al. 2004, Deal 2007).

In North America, due to unavailability of favorable hosts especially red spruce,

Picea rubens (Sarg), winged generations (sexaparae) die. For more detailed information on the lifecycle of HWA in Tennessee or Connecticut, see Deal

(2007) or McClure (1989), respectively.

Hemlock Woolly Adelgid Dispersal

Since 1951, HWA has gradually spread into 16 states in the eastern

United States (Fig. 2). Adelgid eggs and crawlers are spread by wind, wildlife, humans and through infested nursery stock. Along the northeastern United

States, HWA populations have rapidly increased along watercourses and coastal

areas, probably due to migratory birds (Ward et al. 2004). Initially, HWA spread slowly in the eastern United States but in the last decade, its populations have been expanding rapidly in the southern, western and northern ranges of its establishment. The dispersal rate of HWA is 20 to 30 km per year (Ward et al.

2004).

Damage to Hemlock by Hemlock Woolly Adelgid

Adelges tsugae usually attacks the youngest twigs and needles, resulting in desiccation, discoloration, and eventually needle drop. HWA populations increase rapidly, feed together in large numbers and deplete starch reserves which cause the hemlock to lose its ability to grow and produce new growth. Eastern hemlock decline symptoms include reduction of new shoots, followed by needle drop, branch tip dieback, thinning of foliage and finally death of the tree (Ward et al.

2004). There are two possible explanations of mortality in T. canadensis. First, following each molt, salivary toxins may increase during stylet bundle reinsertion because successive saliva sheaths remain in the intercellular space. Second, poor growing conditions lead to less storage of nutrients which may have a negative affect on hemlocks (Kohler 2007).

Control of Hemlock Woolly Adelgid

Control of HWA is a challenging task due to its unusual lifecycle, presence of its susceptible host, and lack of natural enemies (Hain 2005). These factors contribute to a rapid increase in its population and spread into an area. Due to decreased populations of living hemlock trees in highly infested areas, the aesthetic value of forest landscapes is diminishing. HWA control is possible if

control practices are initiated before HWA populations increase to a high level. In landscapes, control measures are also more economical and effective at initial stages of infestation (Ward et al. 2004). After HWA populations have increased to high levels, it is difficult to control. Generally, the following measures are implemented to reduce infestations of HWA.

Cultural Control

Eastern hemlock trees can tolerate higher infestations of HWA populations if trees have enough vigor. Weak trees are more at risk of infestation than healthy trees; therefore, emphasis should be taken to supply enough nutrients to the trees. Mulching may be helpful to preserve moisture in extended drought conditions which may also provide a source of organic matter to hemlock (Ward et al. 2004). Isolated infested trees, as well as infested branches, can be cut, removed and disposed to reduce further infestation and spread of HWA from infested to uninfested trees and/or areas (Dilling 2007). Quarantine measures should be implemented to prevent HWA from spreading. Firewood and logs which are moved from infested areas should go through a quarantine facility to ensure that HWA is not present. Tourists may also be a source of HWA dispersal from infested to uninfested areas and should be educated on ways to reduce

HWA spread. Care should be taken while moving agricultural commodities and machinery from infested to uninfested areas to avoid further infestation.

Biological Control

Few established natural enemies, which may be helpful to reduce populations of HWA in North America, have been identified (Wallace and Hain

2000). Field surveys in Hanging Rock State Park, Stokes County, NC., North

Creek, Jefferson National Forest, Botetourt County, VA and Cave Mountain

Lake, Jefferson National Forest, Rockbridge County, VA on eastern hemlock documented 22 predators representing three species, three genera, three families and two orders collected from beat samples from three sites over four sampling dates. Predators on eastern hemlock were Coccinellidae (Coleoptera),

Derodontidae (Coleoptera), Cecidomyiidae (Diptera), Chamaemyiidae (Diptera),

Syrphidae (Diptera), Chrysopidae (Neuroptera), and Hemerobiidae (Neuroptera).

Overall predator numbers were low during 1997 and 1998 (Wallace and Hain

2000). Predator density on eastern hemlock was low as compared to Chinese hemlock or western hemlock. On western hemlock, 55 species in 14 families were found representing a diverse complex of predators potentially attacking A. tsugae. Predator families on western hemlock were Derodontidae,

Chamaemyiidae, Hemerobiidae, Coccinellidae, Cantharidae, Reduviidae,

Miridae, Syrphidae, Chrysopidae, Coniopterygidae, Staphylinidae, Anthocoridae,

Nabidae, and Raphidiidae. Most of these predators on western hemlock were abundant at the time when two generations of A. tsugae eggs were present

(Kohler 2007).

Several predators, however, feed on HWA in its native habitat in China and Japan. More then 60 species of predaceous insects are associated with

Chinese hemlock. At least 54 species of coccinellids were found on hemlocks in

China (Yu et al. 2000). This large number of predators may be a key factor to explain why HWA is not a problem either on Chinese hemlock or western

hemlock. Among these, three predatory beetles recorded on Chinese hemlock were sevenspotted lady beetle, Coccinella septempunctata L., globose dune beetle, C. trifasciata L., and multicolored Asian lady beelte, Harmonia axyridis

(Pallas) (Yu et al. 2000). Although these species are not implicated in reducing populations of HWA, several other predatory species have been investigated as biological control agents of HWA.

Because HWA is not a problem on hemlock in its native range, studies have been conducted to investigate the possible use of these predatory beetles on eastern hemlock. Several coccinellid beetles were imported into the United

States from the native ranges of HWA. Among these, the first predator introduced, reared, and released in the United States was a small beetle

[Sasajiscymnus tsugae (Sasaji and McClure) (St), Coleoptera: Coccinellidae] introduced from Japan. More than two million St beetles have been released in the eastern United States. Another predator, Laricobius nigrinus Fender

(Coleoptera: Derodontidae), native to western North America where it feeds on introduced HWA on western hemlock, also has been released (PDCNR 2007).

So far, tens of thousands of L. nigrinus either reared in the laboratory or field collected from British Columbia and the western United States have been released in the eastern United States against HWA. Three other species,

Scymnus sinuanodulus, S. camptodromus S. ningshanensis (Coleoptera:

Coccinellidae), were imported into the eastern United States and are under investigation (Zilahi-Balogh et al. 2002). The effectiveness of these introduced natural enemies has not yet been confirmed. It may take three to six years for

these introduced biological control agents to become established after release in their new habitat and begin to reduce populations of HWA.

Insect-killing Fungi

Insect pathogens (entomopathogens) can dramatically reduce insect populations. Most of these pathogens are fungi, bacteria, viruses, protozoa, rickettsia and microsporidia. A laboratory study was conducted to verify insect- killing fungi of HWA (Cheah et al. 2004). Isolates were identified, and these isolates readily germinated and infected HWA at normal temperature ranges where hemlock commonly grows. The effect of fungi on non-target predators and mass development technology of the fungi are currently being studied (Cheah et al. 2004). HWA populations were significantly reduced by a single application of

Verticillium lecanii Zimmermann, when field trials were conducted in

Massachusetts to select the most virulent fungal isolates and determine the best time and concentration for fungal applications. Application of fungal pathogens on HWA in summer could provide the best results because wool is absent during late summer and HWA are aestivating as first instar sistens (Cheah et al. 2004).

Horticultural Oils/Insecticidal Soaps

Horticultural oils and insecticidal soaps have been found to be effective in controlling HWA (Hale 2004). When horticultural oils are applied, they coat the bodies of HWA and disrupt cell membranes or suffocate them, eventually leading to death.

Horticultural oils and insecticidal soaps are relatively safe and are not highly toxic to predators, parasitoids and the environment so their use may help

to control HWA without greatly impacting the environment. However, their application on tall trees is difficult and they may have to be applied several times during a season. Horticultural oils and insecticidal soaps can provide effective control of HWA even when the waxy covering is present.

To get maximum control, a high pressure spray is necessary to make an even cover on HWA. These oils and soaps have to be applied frequently as compared to traditional insecticides (Hale 2004). Horticultural oils may cause phytotoxicity; therefore, a 1% solution should be applied from May through

September, and a 2% solution should be applied from October to April (Hale

2004).

Chemical Insecticides

Several chemical insecticides are used to reduce populations of HWA on eastern hemlock. Chemical control, however, is not a permanent solution for

HWA and its wide-scale application in forests is almost impractical and economically expensive (Grant et al. 2005). The high cost of pesticides and their potential impact on non-target organisms, as well as secondary pest emergence after pesticide application, make this method controversial. Insecticide selection is important in a pest management strategy directed against HWA. Pyrethroids are effective against HWA but are toxic to predatory mites and other beneficial organisms (Ward et al. 2004). These beneficial organisms include spiders, lady beetles, preying mantids, lacewings, parasitic wasps and other predatory beetles.

Although pesticides are important to control HWA, they may have detrimental impacts on predators associated with eastern hemlock.

Imidacloprid

Imidacloprid is the primary chemical insecticide used to control HWA.

Imidacloprid, a chloronicotinyl insecticide, evolved from the naturally occurring insecticide nicotine which interferes with normal nerve impulse transmission by binding to the post-synaptic nicotinergic acetyleholine receptor (Charles 2002).

This binding to the receptor causes death of the insect. Formulations of imidacloprid, such as Merit® 75 WP, may be used as a soil drench, or a foliar spray. Other formulations, such as Imicide®, are available to use as a trunk injection (Charles 2002). Imidacloprid is effective against foliage-feeding insects for several reasons. First, it is a systemic which can be injected in the soil or in plant tissue. Once injected, it flows within the tree sap. Its application via an injection method is comparatively less expensive and effective as compared to other application methods. With this method, a small amount of insecticide is needed to apply. Soil injection provides longer protection (Charles 2002), allows uptake of the insecticide by the roots, and reduces the chances of runoff of the insecticide from the application site (Townsend and Rieske-Kinney 2007).

Second, imidacloprid possesses excellent contact activity, can be applied by several methods, has low application rates and provides long residual control.

Imidacloprid applied as a soil injection provides the longest duration of HWA control, but also the slowest acting and may need at least two months to achieve results (Charles 2002).

Drawback of Pesticides

Although pesticides play an important role in suppressing pests, they may have negative impacts on non-pests and also induce secondary pests. For example, a high abundance of mites was observed on imidacloprid-treated hemlock and it may be due to lethal impact on natural enemies associated with eastern hemlock (Raupp et al. 2004).

Importance of Predators

Predators are important stabilizing elements of various agricultural ecosystems, and their influence on pest species often reaches the extent at which they can be considered to be significant biological control agents (Samu et al. 2003). Several studies have been conducted to document insect density in tree canopies. The total arthropod density was 34 individuals/m2 of leaf area while spiders were abundant within crowns of Argyrodendron aetinophyllum

(Sterculiaceae) in a subtropical rainforest in Australia. The most diverse predator families included Staphylinldae and Theridiidae (Basset and Arthington 1992).

Arthropod fauna of Bornean lowland rainforest trees comprises Coleoptera,

Diptera, Formicidae and Hymenoptera consistently across the tree canopy (Stork

1991). A total of 5,233 spiders comprising 149 species was recorded in an

African montane forest in the Uzungwa Mountains of Tanzania. Three spider families (Linyphiidae, Oonopidae and Pholicidae) contributed to 45% of all samples collected. Theridiidae, Salticidae and Araneidae were rich in species

(Sorensen 2004). In another study, at least 89 morph species and 19 orders of

arthropods were collected from a lowland forest of the Barro Colorado National

Monument in Panama (Stuntz et al. 2002) while in a tropical rainforest,

Coleoptera, particularly Staphylinidae, Hymenoptera and Araneae appear to be most abundant in the tree canopy (Basset 2001). A total of 2,270 coleopteran specimens was associated with northern red oak, Quercus rubra L., in the Great

Smoky Mountains National Park, Tennessee. Of these, five predator families

(Anthicidae, Cantharidae, Coccinellidae, Derodontidae, and Staphylinidae) were recorded from northern red oak during 1992-95 (Trieff 1992). In a similar study on southern magnolia Mangolia grandiflora L., five predator families (Anthicidae,

Carabidae, Coccinellidae, Lampyridae and Staphylinidae) were recorded (Werle

2002).

Generalist predators play an important role in biological pest control in many crops including maize (Meissle and Lang 2005). On maize, the most abundant generalist predators captured by sticky traps were lady beetles, green lacewings and damsel bugs (Bhatti et al. 2005). Natural enemies play an important role in agricultural ecosystems where they influence the yield of many crops. On average, ground-living natural enemies of aphids in barley helped to increase yield by 23% (Ostman et al. 2003). The solitary parasitoid, Urophora quadrifasciata, was almost 19.4% of all insects collected from spotted knapweed,

Centaurea stoebe L., in eastern Tennessee (Kovach 2004).

On eastern hemlock, the most abundant predators in beat-sheet and twig samples were brown and green lacewings followed by flower or hover fly larvae and tooth-necked fungus beetles, Laricobius rubidus LeConte (Wallace and Hain

2000). Spiders are amongst the most dominant generalist predators in agricultural fields, and a number of studies indicated their effectiveness in biological pest control. For example, in orchards, spiders were the most abundant and diversified natural enemy groups and their impact on pests of forests is not as well documented as for agricultural crops (Stano and Haddad

2005).

Classical biological control attempts against balsam woolly adelgid (BWA),

Adelges piceae (Ratzeburg), are well documented. About 25 species of predators were released, which resulted in the establishment of eight predator species, including Laricobius erichsonii Rosenhauer. Despite establishment of these predators, BWA could not be controlled successfully. Reasons may be poor synchronization between predators and BWA, limited effectiveness of predators, inability of predators to adapt to the harsh climate, inability of host trees to withstand even low populations of BWA, and lack of an appropriate natural enemy complex (Zilahi-Balogh et al. 2002).

Non-target Impact of Imidacloprid on Predators

Control of A. tsugae on ornamental hemlocks, as well as those in forest settings, requires treatment with horticultural oils, insecticidal soaps, and more conventional chemical insecticides (such as malathion, diazinon and imidacloprid) (McClure 1992). Pesticides can modify invertebrate movement and feeding behavior which has been shown to reduce predation in agroecosystems.

Non-target arthropods of agroecosystems can be adversely affected by pesticide

applications as part of crop management or indirectly in the form of spray drift.

Species of spiders that hunt prey or capture prey in webs are susceptible to many pesticide formulations within agricultural crops. Similar findings may be possible in forest ecosystems.

Several potential pathways may enable agrochemicals to influence non- target fauna; for example, through direct contact with pesticide droplets or consumption of contaminated prey. Paralysis of treated individuals were indicative of sub-lethal chemical effects (Shaw et al. 2006). Highly mobile invertebrates were strongly affected by neuroactive insecticides in an apple orchard. In an Oregon apple orchard, high mobile invertebrates were strongly affected by broad-spectrum neuroactive insecticides, and significantly lower numbers of predators (ground beetles, centipedes, earwigs, harvestmen and spiders) were recorded (Epstein et al. 2000). Foliar application of insecticides in maize significantly and consistently decreased the abundance of lady beetles, green lacewings and damsel bugs compared with no treatment (Bhatti et al.

2005). Total number of predators was higher in untreated plots independently of the year (Albajes et al. 2003). Similar types of results may be seen with some beneficial organisms and predators in forest systems.

Imidacloprid may affect facultative predators through direct ingestion of insecticide (Albajes et al. 2003). Several pyrethroid or chlorinated insecticide families have negative effects on spider populations through direct mortality or reduced mobility. When winter wheat was sprayed with dimethoate, non-target arthropods were reduced by 85% (Vickerman and Sunderland 1977).

Broad-spectrum insecticides have severe effects on many groups of non- target arthropods, including spiders (Meissle and Lang 2005). Spiders are exposed to chemicals in the field as their herbivorous prey is likely to ingest and pass the toxin to predators (Dutton et al. 2002, Godman and Lancaster 2003).

Single applications of pyrethroids against European corn borer, Ostrinia nubilalis

(Hubner), in corn fields in spring had a significant negative impact on the spider community that lasted over the whole growing season (Meissle and Lang 2005).

Pesticide residues cause considerable mortality and affect predator behavior. In laboratory experiments, spiders could recognize fresh residues of pesticides but could not detect one-day-old pesticide residue (Stano and Haddad

2005). Pesticides may repel predators despite their low toxicity, thus causing them to either leave the treated area or starve to death (Stano and Haddad

2005). Predators are susceptible to secondary poisoning when they eat contaminated prey (Epstein et al. 2000).

Imidacloprid and chlorpyrifos were found to be toxic to coccinellid beetles in an apple orchard (Filho et al. 2002). Seven other insecticides, such as bifenthrin, diazinon, dimethoate, methomyl, carbaryl, malathion and phosmet, caused 100% mortality of H. axyridis at 0.01, 0.125, 0.06, 0.11, 0.24, 0.075 and

0.06% ai, respectively. Imidacloprid (0.013% ai) and chlorpyrifos (0.125% ai) were highly toxic and caused 80% mortality (James 2003). Such pesticides constitute a potential threat to natural enemies (Filho et al. 2002). Ideally insecticides should have a component that is repellent to natural enemies so that contaminated surfaces are avoided by predators. In an apple orchard in Czech

Republic, spider abundance was low after insecticide application in late spring

(Pekar 1999). Ambush spiders (Philodromidae and Thomisidae) were directly exposed to chemicals as they were active during the day and had no protection against pesticide droplets (Pekar 1999). Azadirachtin (an extract from neem trees) when applied in cotton also caused temporary repellency to some spiders

(Salem and Matter 1991).

Imidacloprid applied against HWA in the eastern United States may have a negative impact on predatory guilds associated with eastern hemlock.

Pesticides applied on eastern hemlock trees may have a long residual effect which can be harmful to predators, such as spiders. A large number of spiders are associated with eastern hemlock and these predators feed on herbivorous insects which may have fed on insecticide-treated hemlock. Thus, these spiders may become contaminated with insecticides indirectly by feeding on the prey.

Research Objectives

Herbivorous insects, especially those with piercing-sucking mouthparts, acquire imidacloprid through feeding on plants. When predators feed on insecticide-contaminated prey, they also could be affected. However, little information is known about the impact of imidacloprid on non-target foliage- dwelling predatory arthropods on eastern hemlock. Thus, this study was conducted to: 1) ascertain the effect of imidacloprid used against HWA on the predatory guild associated with eastern hemlock, 2) determine seasonal abundance of the predatory guild on eastern hemlock, and 3) assess influence of vertical stratification on spiders and other predators. 23

II. Effect of Imidacloprid and Horticultural Oil on Abundance and

Diversity of Non-Target Predators Associated with Eastern

Hemlock, Tsuga canadensis (L.) Carriere

Introduction

Broad-spectrum insecticides, horticultural oils and insecticidal soaps are extensively used against agricultural pests (McClure 1992) and are important in controlling hemlock woolly adelgid (HWA), Adelges tsugae Annand (Cowles et al.

2006) in landscape and forest settings. Some pesticides (i.e., malathion, diazinon, imidacloprid, etc.), however, also have high toxicity to non-target arthropods (Vickerman and Sunderland 1977, Pekar 1999, Albajes et al. 2003,

Meissle and Lang 2005). In an apple orchard in Czech Republic, imidacloprid and chlorpyrifos were found to be toxic to coccinellid beetles in apple. Ambush spiders, which were active during day time (i.e., Philodromidae and Thomisidae), were affected adversely, and their density was higher in untreated orchards as compared with orchards which were sprayed with broad-spectrum insecticides

(Pekar 1999, Miliczky et al. 2000). Azadirachtin, a botanical insecticide, caused temporary repellency to some spiders (Salem and Matter 1991). Foliar application of certain insecticides significantly reduced abundance of lady beetles, green lacewings and damsel bugs as compared with no treatment in corn (Bhatti et al. 2005), while higher predator density was recorded in untreated plots in field corn as compared with imidacloprid seed treatment.

Predators are important elements of agricultural ecosystems because they feed mainly on other insects. However, they are not as well documented in forests, where they also play an important role as biological control agents to keep pest populations low to avoid losses (Nyffeler et al. 1994, Samu 2003,

Meissle and Lang 2005, Stano and Haddad 2005, Torres and Ruberson 2005).

Spider guild structure in the United States is complex and includes hunters like Oxyopidae, Salticidae, Clubionidae, Thomisidae, and Lycosidae which have extended diet requirements. Spider density depends upon crop structure, cultural practices and suitable ground cover (Nyffeler and Sunderland 2003). Harvestmen

(Opiliones) are also important predators and widely distributed in almost all terrestrial environments, living on rocks, litter, and on vegetation (Bragagnolo et al. 2007).

Imidacloprid, a neonicotinoid insecticide, is one of the most commonly used insecticides against HWA. Imidacloprid interferes with normal nerve impulse transmissions in insects and is effective at reducing populations of HWA

(Charles 2002). Herbivorous insects, especially those with piercing-sucking mouthparts, acquire imidacloprid through feeding on plants (Dutton et al. 2002).

When predators feed on these insecticide-contaminated prey, they may lose their ability to hunt, or may die due to secondary poisoning. However, little information is known about the impact of imidacloprid and horticultural oils on non-target foliage-dwelling arthropods on eastern hemlock. Thus, this study was initiated to determine the effect of imidacloprid and horticultural oils used against HWA on the predatory guild associated with eastern hemlock.

Materials and Methods

Study Site

The study was conducted at Indian Boundary Campground (Fig. 4A) in the southern part of Cherokee National Forest (Fig. 4B) (35° 23‟N, 84° 06‟W; elevation ca. 543 m) in Monroe Co. Tennessee. This part of Cherokee National

Forest is ca. 140 km south of Knoxville, Tennessee. The forest consists of a substantial number of eastern hemlocks of various sizes and ages. Cherokee

National Forest is spread over an area of 8,000 hectares, the largest tract of public land in Tennessee. It lies in the southern Appalachian mountain range, one of the world's most diverse areas which serves as home to more than 20,000 species of plants and animals, including ca. 1,500 black bears.

Cherokee National Forest provides an outdoor recreational area to the public, as well as a habitat for wildlife. It has majestic mountains, tumbling

A B

Fig. 4. Study site: A) Indian Boundary Recreation Area, B) Cherokee National Forest, Tennessee. 26

streams, and diverse vegetation with well developed campgrounds, picnic areas, and several hundred kilometers of trail and cold water streams which provide habitat for fish.

Eastern hemlocks (n=30) were selected for use in this research on 5

November 2005. Infestation levels of HWA (Appendix A and B) and several characteristics (height, girth, crown, etc.) (Appendix C) describing the condition of selected trees were recorded before application of treatments (Table 2). The incidence of HWA was generally low, but the intent of this study was to measure impact of insecticides on predators populations, not on HWA. The study was arranged in a split plot 2 X 5 factorial design (2 treatment times and 5 treatments) with three blocks. This study was established at 35° 23” N, 84° 06” W with an elevation of 543-565 m. Each block contained 10 trees in five pairs. One tree in each pair was treated on 29-30 November 2005 (Fall treatments) and the remaining tree was treated on 16 April 2006 (Spring treatments).

Application of Insecticides and Horticultural Oil

Five treatments were evaluated in this study. These included a control (no treatment), horticultural oil, and imidacloprid which was applied as a tree injection

(Fig. 5), soil injection (Fig. 6), and soil drench (Fig. 7). Horticultural oil was applied as a foliar spray (Fig. 8).

Tree injection was accomplished using a Mauget® injector and 3 ml 10%

Imicide capsules with a feeder tube (J. J. Mauget Co. Arcadia, CA) and applied at one capsule per 15 cm (0.15 ml ai/2.5 cm DBH). A 1.2 cm hole was made ca.

Table 2. Pre-treatment and post-treatment infestation of hemlock woolly adelgid in Cherokee National Forest, Tennessee.

Treatments Pre-treatment Post-treatment percent infestation* percent infestation* (November 2005) (January 2007)

Soil Drench Absent Absent Soil Drench Absent Absent Soil Drench 25-50% <25% Soil Drench <25% <25% Soil Drench Absent Absent Soil Drench <25% Absent Horticultural Oil Spray <25% Absent Horticultural Oil Spray Absent Absent Horticultural Oil Spray Absent Absent Horticultural Oil Spray <25% Absent Horticultural Oil Spray <25% 50-75% Horticultural Oil Spray 25-50% Absent No Treatment Absent <25% No Treatment 25-50% 25-50% No Treatment <25% >75% No Treatment <25% >75% No Treatment Absent <25% No Treatment <25% 50-75% Soil Injection Absent Absent Soil Injection Absent Absent Soil Injection Absent Absent Soil Injection Absent <25% Soil Injection <25% <25% Soil Injection <25% <25% Tree Injection <25% 50-75% Tree Injection Absent 50-75% Tree Injection <25% <25% Tree Injection <25% >75% Tree Injection <25% 50-75% *Percent infestation rating (three branches of 30.48 cm branch per tree)

Fig. 5. Imidacloprid applied as a tree injection.

Fig. 6. Imidacloprid applied as a soil injection.

Fig. 7. Imidacloprid applied as a soil drench.

Fig. 8. Horticultural oil applied as a foliar spray.

20.5 cm above ground level using a 0.4 cm drill bit with a slanting angle. The feeder tube was inserted into the hole and a capsule was attached to the tube

(Fig. 5). Capsules were equally distributed on the tree trunk as much as possible for even distribution of insecticide. The capsules were left for 1-5 hours and then collected when all insecticide was taken up by the hemlock.

Imidacloprid was applied via soil injection using a Kioritz® soil injector

(Kioritz Corp. Tokyo, Japan) (Fig. 6). Imidacloprid (Merit® 75 WP) was applied at

1 g ai/2.5 cm dbh in 60 ml of water. Soil injection was applied within 45 cm from the base of the trunk and spaced evenly as much as possible. The insecticide was injected at a depth of 7 cm below soil surface at 30 ml per injection.

Imidacloprid (Merit 75 WP) was applied as a soil drench at 1.5 g ai/2.5 cm dbh using a high pressure hydraulic sprayer (Figs. 7 and 9) (FMC Corporation,

Jonesboro, AR). The standard recommended dose of 50 g imidacloprid was mixed with 379 liters (100 gall) of water. The soil beneath the drip line of the canopy was sprayed with at least 124.9 liters (33 gall) of mixture. The same procedure was followed for Fall (November 2005) and Spring (April 2006) treatments.

Sunspray® horticultural oil was applied using a FMC high pressure hydraulic sprayer (Figs. 8 and 9). The tree was sprayed at the rate of 7.6 liters ai/379 liters of water and sprayed to runoff to ensure that the tree was fully covered.

Fig. 9. High pressure hydraulic sprayer used for foliar spray and drench.

Sampling Methods

Several sampling methods were used to assess the impact of treatments on predator populations on eastern hemlock. These included beat sheet, direct observation/vacuuming of main trunk and malaise traps.

Beat-Sheet Sampling

A cloth beat sheet (1 x 1 m) with white background was used to collect predators (Fig. 10). Sampling was done monthly. Four branches from each cardinal direction on each tree were randomly selected. A beat sheet was placed underneath the branch and struck five times using a wooden stick (1 m).

Fig. 10. Beat-sheet sampling from hemlock trees.

Specimens that fell onto the beat sheet were collected using tweezers. Soft bodied predators were collected by dipping a finger into alcohol and touching it to the specimen. All specimens collected were placed into 6 dram vials which contained 75% alcohol (ca. ½ full). Each vial was properly labeled (date collected, tree number and collecting method) and stored until further processing.

Direct Observation/Vacuuming

A battery operated vacuum was used to collect predators from the trunk.

Direct observation/vacuuming of trunk was done for at least 15 minutes up to distance at breast height from each tree. Samples were then placed in vials which contained 75% alcohol.

Malaise Traps

Malaise traps were installed on each tree and monitored monthly. A modified malaise trap with a bottom pan was used (Fig. 11). To install the trap on the tree, a rope was used to hoist and hang the malaise trap on the hemlock.

Modified malaise traps (60 cm x 60 cm x 60 cm) were constructed of PVC pipe and covered with No-Thrips® insect screen. The pan (15 cm wide x 65 cm long x 12 cm deep) was filled with ca. 1,000 ml of 50% propylene glycol and water. A collection cup was used on the top of the trap (6 cm diameter and 6.5 cm deep). The cup was filled with ca. 30-60 ml of 50% propylene glycol and water. Specimens collected from malaise traps were placed in a labeled cup (ca.

60 mm x 65 mm deep), taken to the laboratory, processed and identified.

Fig. 11. Modified malaise traps on hemlock trees.

Processing and Identification of Specimens

Specimens collected from direct observation of the main trunk were placed in vials. These specimens were then taken to the laboratory. Specimens collected from malaise traps were drained and placed in 75% alcohol.

Spiders and other predators collected from all sampling methods were separated from other non-predators and identified using existing taxonomic keys

(Howel and Jenkins 2004, Ubick et al. 2005). Voucher specimens (spiders and other arthropod predators) were organized into Cornell drawers and will be deposited in the Insect Museum, University of Tennessee.

Data Analysis

Treatments were analyzed using mixed model analysis of variance

(ANOVA) in SPSS 15.0. Data entered into Excel® consisted of order, family, collection method and number of specimens collected. Significant differences were determined by using LSD, and output values were represented as significant at P=0.05.

Results and Discussion

Abundance of Predators

During this two-year study, 4,917 predators were collected, representing

75 families and 10 orders. Spiders were dominant in all treatments and seasons as compared to other predators. Forty two spider families and 33 other predatory families were collected. The most abundant spider families were Mimetidae

(1,038), Salticidae (736), Araneidae (733), Gnaphosidae (517), Philodromidae

(330), Theridiidae (168), Tetragnathidae (161) and Thomisidae (142) (Table 3).

The most abundant insect predator families were Vespidae (132),

Ichneumonidae (50), Braconidae (31), Pentatomidae (25), Reduviidae (24),

Coccinellidae (15), and Elateridae (15) (Table 4). In a two-year study to document insect communities on eastern hemlock, 671 insect predators, representing 32 species, were collected (Buck 2004, Buck et al. 2005, and Dilling et al. 2007) which is similar to the total number collected in this study. The previous study also included pitfall sampling, which was not included in this study, and did not include an analysis of spider populations. Dilling et al. (2007) reported that the predator guild on eastern hemlock was higher than that reported on western hemlock (Schowalter and Ganio 1998). However, in a more comprehensive study, Kohler (2007) reported 55 species of predators representing 14 families on western hemlock.

Seasonal Abundance of Predators

The time of year significantly impacted predator abundance with greater numbers of predators found from April to November in 2006. In 2006, predator densities were not significantly different among January, February, March, April,

May, June, October and December samples. However, those densities were significantly less than those found in July, August, September and November

(Fig. 12). The highest densities were recorded in August and September, while the fewest predators were collected in February and March. The higher densities in summer months may be due to general climatic conditions while there were few predators on the trees during winter.

0 Mean Mean number of predators per tree Jan Apr Oct Feb May June July Aug Sep Nov Dec March Fig. 114.2. Abundance of predators on eastern hemlock at Indian Boundary Campground, Cherokee National Forest, Tennessee, 2006.

4 Mean number of predators per tree per predators of number Mean 2

0 Feb March Apr May June July

Fig. 13. Abundance of predators on eastern hemlock at Indian Boundary Campground, Cherokee National Forest, Tennessee, 2007.

In 2007, predator densities were recorded only for six months. However, predators during this time were much higher in 2007 than they were in 2006.

During 2007, predator densities in February, April and June were significantly lower than those collected in March, May and July. No significant differences in predator densities were observed between March, May and July samples (Fig.

13).

Influence of Treatment on Abundance of Predators

In 2006, predator densities in control treatments and horticultural oil treatments were significantly greater than those in imidacloprid treatments (soil drench, soil injection and tree injection). However, predator densities were not significantly different among soil drench, soil injection and tree injection treatments or between control and horticultural oil treatments (Fig. 14). Densities of predators were ca. 2-3 times greater in control and horticultural oil treatments than in imidacloprid treated plots.

In 2007, predator densities in the control treatments, horticultural oil treated treatments, and soil drench treatments were significantly greater than those in soil injection and tree injection treatments. However, no significant differences were found between soil drench, horticultural oil and control treatments or between soil injection or tree injection treatments (Fig. 15).

Predator densities in the imidacloprid treatment plots in 2007 were 1.5-3 X higher than in 2006, possibly suggesting a rebound in predator densities 1 to 1 ½ year after treatment.

Table 3. Abundance of spiders and opilionids collected from eastern hemlock, Indian Boundary Campground, Cherokee National Forest, Tennessee, 2006- 2007.

Number of Order Family specimens collected Collection method

Araneae Agelenidae 7 bs, do, m Amaurobiidae 6 bs, do Anapidae 1 bs Antrodiaetidae 2 bs, m Anyphaenidae 51 bs, do,m Araneidae 733 bs, do, m Caponiidae 4 bs, do, m Clubionidae 92 bs, do, m Corinnidae 41 bs, do, m Ctenidae 47 bs, do Cyrtaucheniidae 15 bs, do Desidae 9 bs, do Dictynidae 61 bs, do Dipluridae 3 bs, do Dysderidae 1 do Filistatidae 5 bs, Gnaphosidae 517 bs, do, m Homalonychidae 4 bs, do Hypochilidae 7 bs Linyphiidae 27 bs, do, m Lycosidae 4 bs Mimetidae 1038 bs, do , m Mysmenidae 3 bs Nesticidae 6 bs Oecobiidae 24 bs Oxyopidae 6 bs, do Philodromidae 330 bs, do, m Pholcidae 2 bs, do Pisauridae 5 bs, m Plectreuridae 28 bs, do Prodidomidae 6 bs, m Salticidae 736 bs, do, m Segestriidae 2 bs, m Sparassidae 12 bs, do Tetragnathidae 161 bs, do, m Theraphosidae 7 bs, m, Theridiidae 168 bs, do, m

Table 3. Continued.

Number of Order Family specimens collected Collection method Thomisidae 142 bs, do, m Uloboridae 3 bs, do, m Zoridae 9 bs Zorocratidae 13 bs, do Zoropsidae 3 bs, do Opiliones 203 bs, do, m bs=Beat-sheet, do=Direct observation, m=Malaise trap

Table 4. Abundance of insect predators collected from eastern hemlock, Indian Boundary Campground, Cherokee National Forest, Tennessee, 2006-2007.

Number of Collection Order Family specimens collected method

Coleoptera Anobiidae 1 m Coccinellidae 15 bs, do, m Elateridae 15 m Histeridae 1 m Staphylinidae 13 m

Dermaptera Anisolabididae 1 bs

Diptera Asilidae 1 m Empididae 1 bs, m Mycetophilidae 5 bs Phoridae 3 m Syrphidae 2 m Tachinidae 1 m

Hemiptera Nabidae 2 bs, do Pentatomidae 25 bs, do Reduviidae 24 bs, do, m

Hymenoptera Bethylidae 1 m Braconidae 31 bs Chalcididae 8 bs, m Chrysididae 1 do Encyrtidae 2 do, m Eupelmidae 1 do Figitidae 4 m Formicidae 6 bs, do Ichneumonidae 50 bs, m Mymaridae 4 bs, m Perilampidae 2 bs Pteromalidae 12 m Tiphiidae 1 bs Vespidae 132 do, m

Table 4. Continued.

Number of collection Order Family specimens collected method

Neuroptera Chrysopidae 3 bs, do Mantispidae 1 m

Odonata Libellulidae 2 m

Orthoptera Gryllidae 2 m bs=Beat-sheet, do=Direct observation, m=Malaise trap

10 9 Control Horticultural Oil 8 a Tree Injection 7 a Soil injection 6 Drench 5

4 tree b b 3 b 2 1

0 Fig. 14. Abundance of predators on eastern hemlock at Indian Boundary

Campground, Cherokee National Forest, Tennessee, 2006. Mean number of predators per treatment per per treatment per of predators number Mean

10 a Control 9 a Horticultural Oil Tree Injection 8 Soil injection 7 Drench b b 6 5 b

4 treatment per tree per treatment

3 Mean number of predators per per predators of number Mean 2

Fig. 15. Abundance of predators on eastern hemlock at Indian Boundary Campground, Cherokee National Forest, Tennessee, 2007.

Effect of Pesticide Treatments on Abundance of Predators

Effect of Time

Predator densities were not significantly different between pesticide application times (Fall 2005 and Spring 2006) in 2006 (p>0.05, df=1, 250, F=

4.768) (Fig. 16) or 2007 (p>0.05, df=1, 167, F=0.105) (Fig. 17). In other words, treatment with imidacloprid or horticultural oil in the Fall or Spring had similar effects on predator densities.

Effect of Pesticide Treatment on Seasonal Abundance of Predators

During most months in 2006, predator densities on the Fall treated trees were much higher on control and horticultural oil treated trees than on any of the

7 a 6 a 5

Mean numbers of predators per tree ofpredators numbers Mean 2

1 Fall Spring

Fig. 16. Influence of pesticide application time (Fall or Spring) on predator abundance Indian Boundary Campground, Cherokee National Forest, Tennessee, 2006.

6 a a 5

Mean numbers of predators per tree per predators of numbers Mean 2

Fall Spring 0

Fig. 17. Influence of pesticide application time (Fall or Spring) on predator abundance at Indian Boundary Campground, Cherokee National Forest, Tennessee, 2007.

imidacloprid-treated trees (Fig. 18A). Predator densities on the control and horticultural oil treated trees followed a typical seasonality curve, with populations gradually increasing early season, peaking, and then declining in the cooler months. Predator densities on the imidacloprid-treated trees, however, were more erratic, with populations varying throughout the season. The highest predator density in any month (ca. 42) was documented in August on the horticultural oil treated trees. In general, predator densities in most months were highest on the horticultural oil treated trees.

In 2006, predator densities on the Spring treated trees were much more erratic than those observed on the Fall treated trees (Fig 18B). It should be noted

that the Spring treatments were applied in April 2006, which explains the erratic densities observed in predator populations following the pesticide applications.

As observed with the Fall treated trees, predator densities in most months were generally much higher on control and horticultural oil treated trees than on any of the imidacloprid-treated trees (Fig. 18B). Even though horticultural oil had been applied in April 2006, it did not have a negative impact on predator populations during the remainder of the year. Predator populations rebounded quickly and, in

August and November, predator densities on horticultural oil treated trees had exceeded predator populations on the control trees (Fig.18B).

In 2007, the highest monthly predator densities were documented on control treatments followed by horticultural oil treated trees for Fall treatments (Fig. 19A) while higher monthly densities were observed on horticultural oil treated trees followed by soil drench treatments for Spring treatments (Fig. 19B). Predator densities from February to July 2007 on imidacloprid-treated trees varied greatly.

The highest cumulative predator densities for Fall and Spring treatment application times were found on control trees (57.0) and horticultural oil treated trees (48.33). Based on monthly and cumulative predator densities observed in

2007, it does not appear that the reductions in predator densities in 2006 led to a continued reduction in 2007. In other words, imidacloprid treatments may cause a short-term reduction of predator populations, but these reductions may not continue long-term.

Control A Horticultural Oil Tree Injection Soil Injection Soil Drench

Mean numberMeanpredatorsof per tree

50 45 Control B 40 Horticultural oil 35 Tree injection

30 Soil Injection 25 Soil Drench 20 15

10 Mean number of predators number of per tree predators Mean 5

Fig. 18. Influence of pesticide treatments on monthly abundance of predators on eastern hemlock, Indian Boundary Campground, Cherokee National Forest, Tennessee, 2006, A) Fall treatments, B) Spring treatments.

50 Control

45 Hort. Oil A

Tree Injection 40 Soil Injection 35 Soil Drench 30

25 20

Mean number of predators per tree tree per predators of number Mean 10

Feb March April May June July

50 Control B 45 Hort. Oil 40 Tree Injection Soil Injection 35 Soil Drench

25 A

20 15

tree per of predators number Mean 10 5

0 Feb March April May June July

Fig. 19. Influence of pesticide treatments on monthly abundance of predators on eastern hemlock, Indian Boundary Campground, Cherokee National Forest, TN, 2007, A) Fall treatments, B) Spring treatments.

Impact of Treatment on Hemlock Woolly Adelgid

Although the intent of this research was not to assess the impact of imidacloprid on HWA, it is important to note the response of HWA to imidacloprid in the study. HWA infestation on control trees and imidacloprid-injected trees increased on five trees and stayed the same on one tree. On the soil-injected trees, HWA stayed the same on the five trees and was reduced on one tree. On the soil-drenched trees, HWA infestations stayed the same on four trees and were reduced on two trees. The horticultural oil treatments provided the best control as five trees were infested with HWA; however, infestations had increase on one tree. From these results, HWA infestations were reduced or slowed by horticultural oil treatments or by imidacloprid applied as a soil drench or soil injection. Imidacloprid applied via trunk injection did not reduce or retard HWA populations.

Summary

During this two-year study, 4,917 predators were collected, representing

75 families and 10 orders. Spiders were the most dominant predator group in all treatments and seasons. The five most abundant spider families were Mimetidae

(1,038), Salticidae (736), Araenidae (733), Gnaphosidae (517), and

Philodromidae (330). Spiders were present on eastern hemlock throughout the year, with the highest densities generally occurring during the warmer months.

Predator densities were not significantly different between pesticide application times (Fall 2005 and Spring 2006) in 2006 (p>0.05, F=4.768) or 2007 (p>0.05,

F=0.105). In other words, treatment with imidacloprid or horticultural oil in the

Fall or Spring had similar effects on predator densities.

In 2006, predator densities in control treatments and horticultural oil treatments were significantly (p<0.05) greater than those in imidacloprid treatments (soil drench, soil injection, and tree injection). However, predator densities were not significantly (p<0.05) different among soil drench, soil injection, and tree injection treatments or between control and horticultural oil treatments. Predator densities in the imidacloprid-treated plots in 2007 were 1.5-

3X higher than in 2006, possibly suggesting a rebound in predator densities 1-1

½ years after treatment.

Predator densities were not affected by application time, but they were affected by imidacloprid, regardless of application method, for as long as 12-18 months after application. Predator densities were low following imidacloprid application but gradually increased to levels similar to those on control trees by

12-18 months after treatments. Long term affects of imidacloprid on diversity and densities of predators associated with eastern hemlock should be investigated to more fully understand its role as a management tool against HWA.

III. Effect of Vertical Stratification on Predators Associated with

Eastern Hemlock, Tsuga canadensis (L.) Carriere

Introduction

Hemlock woolly adelgid (HWA), Adelges tsugae Annand (Hemiptera:

Adelgidae), severely impacts eastern hemlock, Tsuga canadensis (L.) Carriere, and Carolina hemlock, Tsuga caroliniana Engelmann, throughout eastern North

America (McClure 1990). Imidacloprid and horticultural oil are used to control

HWA. Imidacloprid is used as a soil injection, tree injection or soil drench, while horticultural oil is used as a foliar spray. Effectiveness of imidacloprid may depend upon application method. Both soil injection and soil drench methods of applications have been proven to be successful in HWA control (Rhea 1995,

Cowles et al. 2006). Imidacloprid application as soil injection, soil drench and trunk injection made in fall and the spring had no significant effect on HWA control but all imidacloprid applications reduced HWA population between 50 and

100% (avg. 80%). Trunk injection treatments did not result in reduction of HWA

(Cowles et al. 2006).

Although imidacloprid is effective against HWA, little is known about its impact on predatory arthropods, especially throughout the tree canopy. Thus, no relevant data regarding the impact of pesticides on vertical stratification of the predatory guild on eastern hemlock in a forest setting are available. However, in a similar study, peach canopy height had a significant impact on the egg parasitism rate of oriental fruit moth, Grapholita molesta (Busck) (Lepidoptera:

Tortricidae) (Atanassov et al. 2003). The objective of this study was to determine the effect of vertical stratification on the diversity and abundance of predators associated with eastern hemlock.

Materials and Methods

To evaluate the effect of vertical stratification on predators associated with eastern hemlock, the same procedures described in Chapter II, in addition to those outlined in this chapter, were used. This vertical stratification study was conducted using the same two treatment times (Fall and Spring), five treatments

(imidacloprid as tree injection, soil injection, soil drench and horticultural oil as foliar spray and control trees), and eastern hemlock trees (n=30) as described in

Chapter II.

Vertical Stratification

For vertical stratification, data were collected using a bucket lift truck two times. The first samples were collected 16-19 August 2006 and the second samples were collected 5-11 September 2007. A bucket truck was used to allow samples to be collected from three strata in each tree (Fig. 20). Each tree was divided into three strata: top (upper 1/3 of canopy), middle (middle 1/3 of canopy), and bottom (lower 1/3 of canopy) (Fig. 20). From each stratum, three beat-sheet samples were collected and direct observations were made for at least 15 minutes. The branch was struck five times using a wooden stick (1 m) as described in Chapter II. Specimens collected were placed in vials containing 75%

Fig. 20. Vertical sampling of the canopy of eastern hemlock tree using a bucket lift truck.

alcohol, collected, and taken to the laboratory for processing and identification as described in Chapter II.

Data Analysis

Treatments were analyzed using mixed model analysis of variance

(ANOVA) in SPSS 15.0. Means were separated using least significant differences (LSD) procedure (p=0.05).

Results and Discussion

In 2006, predator populations, averaged across all application times and treatments, were equally distributed throughout the tree canopy. Although predator densities were slightly higher in the top canopy than in the bottom canopy, there were no significant (P>0.05) differences among the three strata

(Fig. 21).

In 2007, however, predator densities were significantly (p<0.05) different among strata. Predator densities in the middle and top strata were almost 2x greater than those in the bottom strata. Carla (2007) found lower levels of imidacloprid concentrations on top stratum as compared to middle and bottom stratum and it may be a factor of lower densities of predators on bottom as compared to top stratum and middle stratum. Densities were not significantly different between the middle and top stratum (Fig. 22).

When averaged across all strata and pesticide treatments, no significant differences (p>0.05) in predator densities between Fall and Spring application

16 a 14 Bottom a Middle a 12 Top 10

8 per tree 6 4

Mean Mean number of predators 2

0 Fig. 21. Influence of strata on predator densities on eastern hemlock, Indian Boundary Campground, Cherokee National Forest, Tennessee, 16-19 August 2006

Bottom b 14 Middle b 12 Top 10

8 a

per tree 6

4 Mean Mean number of predators 2

0 Fig. 22. Influence of strata on predator densities on eastern

hemlock, Indian Boundary Campground, Cherokee National Forest, Tennessee, 5-11 September 2007

times were documented in 2006 or 2007 (Fig.23 A and B). Predator densities on

Fall-treated trees were slightly greater than those on Spring-treated trees in both years (Fig. 23 A and B). These results were similar to those documented on eastern hemlock throughout the year (see Chapter II).

The densities of predators collected in all three strata were influenced by pesticide treatment, in a similar manner as discussed in Chapter II. In 2006 (Fig.

24) and 2007 (Fig. 25), significantly (p<0.05) greater numbers of predators were collected from control and horticultural oil treatments than from any of the three imidacloprid treatments. No significant differences in predator densities were observed between control and horticultural oil treatments or among tree injection, soil injection, or drench treatments in either year (Figs. 24 and 25).

Predator densities collected in each strata were compared among pesticide treatments within Fall and Spring treatment application times (Fig. 26 -

Fall treatments, 2006, Fig. 27 - Spring treatments, 2006, Fig. 28 - Fall treatments,

2007, Fig. 29 - Spring treatments, 2007). Regardless of the strata, predator densities were almost always higher in control and in the horticultural oil treatments than in any of the imidacloprid treatments.

The highest number of predators on Fall-treated eastern hemlock in 2007 was observed in the middle strata on horticultural oil-treated trees and in the middle and top strata on control trees. The lowest number of predators was recorded from the bottom strata in soil drench treatments. Relatively higher numbers of predators were observed in the top and middle strata regardless of treatments, as compared to the bottom strata (Fig. 28).

16 a Fall 14 A Spring

12 a

8 per tree per 6

4 Mean number of predatorsofnumber Mean 2

14 B a Fall 12 a Spring

10 8

per tree 6

Mean number of predators 2

Fig. 23. Influence of time of pesticide application, averaged across all three strata and pesticide treatments, on predator densities on eastern hemlock, Indian Boundary Campground, Cherokee National Forest, Tennessee, A) 16-19 August 2006, B) 5-11 September 2007.

25 a Control Horticultural Oil 20 a Tree Injection Soil injection

15 Drench

per tree b 10 b

ean numberean of predators 5 M

Fig. 24. Influence of pesticide treatments, averaged across all strata and pesticide application times, on predator densities on eastern hemlock, Indian Boundary Campground, Cherokee

National Forest, Tennessee, 16-19 August 2006.

Control Horticultural Oil 20 a a Tree Injection Soil injection 15 Drench

per tree 10 b b b

5 Mean Mean number of predators

Fig. 25. Influence of pesticide treatments, averaged across all strata and pesticide application times, on predator densities on eastern hemlock, Indian Boundary, Cherokee National Forest,

Tennessee, 5-11 September 2007.

30 Control Horticultural Oil 25 Tree Injection Soil injection Drench 20

strata 10

Bottom Middle Top Fig. 26. Influence of strata on predator densities among pesticide treatments on Fall-treated eastern hemlock, Indian Mean number of predators per treatment per tree pertree per treatmentper predatorsof number Mean Boundary Campground, Cherokee National Forest, Tennessee, 16-19 August 2006.

Control 30 Horticultural Oil Tree Injection Soil injection 25 Drench

10 strata per

0 Bottom Middle Top Fig. 27. Influence of strata on predator densities among pesticide treatments on Spring-

Mean number of predators per treatment per treeper per treatment predatorsof number Mean treated eastern hemlock, Indian Boundary Campground, Cherokee National Forest, Tennessee, 16-19 August 2006.

35 Control 30 Horticultural Oil Tree Injection 25 Soil Injection Drench 20

strata pertree per

0 Bottom Middle Top

Fig. 28. Influence of strata on predator densities among Mean number of predators pertreatment predatorsof number Mean pesticide treatments on Fall-treated eastern hemlock, Indian Boundary Campground, Cherokee National …

The highest number of predators was recorded on the Spring-treated eastern hemlock in 2007 in the top strata on control trees and in the middle strata on control and horticultural oil treated trees. The lowest number of predators was observed in the bottom strata of soil injected trees (Fig 29).

Summary

During this study, a total of 2,155 predators representing 7 orders and 33 families were collected from the canopies of eastern hemlock. Predator densities were significantly (p<0.05) greater in the top and middle canopy than in the lower canopy. In 2007, predator densities in the middle and top strata were almost 2X greater than those in the bottom stratum. These differences may be partially attributed to the impact of imidacloprid on herbivorous prey at different strata.

35 Control 30 Horticultural Oil Tree Injection 25 Soil injection Drench 20

tree per strata per tree 5

0 Bottom Middle Top Fig. 29 .nfluence of strata on predator densities among

Mean number of predators per treatment perper treatment predatorsof number Mean pesticide treatments on Spring-treated eastern hemlock, Indian Boundary Cherokee National Forest, Tennessee,…

35 Control 30 Horticultural Oil Tree Injection 25 Soil Injection Drench 20

5 per tree per strata per tree per

0 Bottom Middle Top

Fig. 28. Influence of strata on predator densities among Mean number of predators per treatmentper predatorsof number Mean pesticide treatments on Fall-treated eastern hemlock, Indian Boundary Cherokee National Forest, Tennessee, 5-11…

Control 30 Horticultural Oil Tree Injection 25 Soil injection Drench 20

tree per strata pertree 5

0 Bottom Middle Top

Fig. 29. Influence of strata on predator densities among

ofnumberMeanpredatorsper treatment per pesticide treatments on Spring-treated eastern hemlock, Indian Boundary Campground, Cherokee National Forest, Tennessee, 5-11 September 2007.

Because concentration levels of imidacloprid declined progressively from the bottom stratum to the top stratum of the tree canopy, the herbivorous also may have been impacted in a similar manner (Dilling 2007).

This research demonstrates that predator populations are distributed throughout the tree canopy. Imidacloprid treatments, regardless of the application method, decreased predator densities in each strata following imidacloprid treatment. These reductions may be due to a decline in prey populations or due to consumption of contaminated prey. Predator densities rebounded 1-1½ years after imidacloprid application. Although there is a short-term impact of imidacloprid on predators throughout the canopy, the long-term effect of

imidacloprid on predator densities should be investigated to better understand the use of imidacloprid as a management tool against HWA.

IV. Conclusions

During this two-year study, 4,917 predators were collected from eastern hemlock, representing 75 families and 10 orders. Spiders were the most dominant predator group in all treatments and seasons. The five most abundant spider families were Mimetidae (1,038), Salticidae (736), Araenidae (733),

Gnaphosidae (517), and Philodromidae (330). Spiders were present on eastern hemlock throughout the year, with the highest densities generally occurring during the warmer months.

Predator densities were not significantly different between pesticide application times (Fall 2005 and Spring 2006) in 2006 (p>0.05, F=4.768) or 2007

(p>0.05, F=0.105). In other words, treatment with imidacloprid or horticultural oil in the Fall or Spring had similar effects on predator densities. In 2006, predator densities in control treatments and horticultural oil treatments were significantly

(p>0.05) greater than those in imidacloprid treatments (soil drench, soil injection, and tree injection). However, predator densities were not significantly (p>0.05) different among soil drench, soil injection, and tree injection treatments or between control and horticultural oil treatments. Predator densities in the imidacloprid-treated plots in 2007 were 1½ -3X higher than in 2006, possibly suggesting a rebound in predator densities 1-1½ years after treatment.

Predators were found throughout the tree canopy. For the two-year period, predator densities were significantly (p<0.05) greater in the top and middle canopy than in the lower canopy. In 2007, predator densities in the middle and top strata were almost 2X greater than those in the bottom stratum. 63

This research demonstrates that imidacloprid treatments, regardless of the application method, decrease predator densities throughout the canopy following treatment. These reductions may be due to a decline in prey populations or due to consumption of contaminated prey. Predator densities rebounded 1-1 ½ years after application. The long-term effect of imidacloprid on predator densities should be investigated.

Literature Cited

Albajes, R., C. Lopez, and X. Pons. 2003. Predatory fauna in cornfields and response to imidacloprid seed treatment. J. Econ. Entomol. 96: 1805- 1813. Atanassov, A., P. W. Shearer, and G. C. Hamilton. 2003. Peach pest management programs impact beneficial fauna abundance and Grapholita molesta (Lepidoptera: Tortricidae) egg parasitism and predation. Environ. Entomol. 32: 780-788. Basset, Y. 2001. Invertebrates in the canopy of tropical rainforests. How much do we really know? Plant Ecol. 153: 87-107. Basset, Y., and A. H. Arthington. 1992. The arthropod community of an Australian rainforest tree: Abundance of component taxa, species richness and guild structure. Australian J. Ecol. 17: 89-98. Bhatti, M. A., J. Duan, G. P. Head, C. Jiang, M. J. Mckee, T. E. Nickson, C. L. Pilcher, and C. D. Pilcher. 2005. Field evaluation of the impact of corn rootworm (Coleoptera: Chrysomelidae) protected Bt corn on foliage- dwelling arthropods. Environ. Entomol. 34: 1336-1345. Bragagnolo, C., A. A. Nogueira, R. Pinto-da-Rocha, and R. Pardini. 2007. Harvestmen in an Atlantic forest fragmented landscape: Evaluating assemblage response to habitat quality and quantity. Bio. Cons. 139: 389- 400. Buck, L., P. Lambdin, D. Paulsen, J. Grant, and A. Saxton. 2005. Insect species associated with eastern hemlock in the Great Smoky Mountains National Park and environs. J. Tennessee Acad. Sci. 80: 60-69. Charles, A. S. 2002. Using imidacloprid to control hemlock woolly adelgid, pp. 280-287. In B. Onken, R. Reardon and J. Lashomb [eds.], Proceedings; Hemlock Woolly Adelgid in the Eastern United States Symposium, February 5-7, 2002. Rutgers University, East Brunswick, New Jersey. Cheah, C., M. E. Montgomery, S. Salom, B. L. Parker, S. Costa, and M. Skinner. 2004. Biological control of hemlock woolly adelgid.

http://www.invasive.org/hwa/pathogens.cfm. In R. Reardon and B. Onken [eds.]. USDA Forest Service, Morgantown, WV, FHTET-2004-04. Cowles, R. S., M. E. Montgomery, and C. A. S. J. Cheah. 2006. Activity and residues of imidacloprid applied to soil and tree trunks to control hemlock woolly adelgid (Hemiptera: Adelgidae) in forests. J. Econ. Entomol. 99: 1258-1267. Deal, I. K. 2007. Life history of hemlock woolly adelgid, Adelges tsugae Annand, on eastern hemlock, Tsuga canadensis (L.) Carriere, in the southern Appalachians and assessment of egg releases of Sasajiscymnus tsugae (Sasaji and McClure) for its management. M.S. thesis, University of Tennessee, Knoxville, 67 pp. Dilling, C., P. Lambdin, J. Grant, and L. Buck. 2007. Insect guild structure associated with eastern hemlock in the southern Appalachians. Environ. Entomol. 36: 1408-1414. Dilling, C. I. 2007. Impact of imidacloprid and horticultural oil on non–target phytophagous and transient canopy insects associated with eastern hemlock, Tsuga canadensis (L.) Carrieré, in the southern Appalachians. M.S. thesis, University of Tennessee, Knoxville, 108 pp. Dutton, A., H. Klein, J. Romeis, and F. Bigler. 2002. Uptake of Bt-toxin by herbivores feeding on transgenic maize and consequences for the predator Chrysoperla carnea. Ecol. Entomol. 27: 441-447. Epstein, D. L., R. S. Zack, J. F. Brunner, L. Gut, and J. J. Brown. 2000. Effects of broad-spectrum insecticides on epigeal arthropod biodiversity in Pacific Northwest apple orchards. Environ. Entomol. 29: 340-348. Eyre, F. 1980. Forest cover types of the United States and Canada. Soc. Amer. Foresters. Washington, DC, 148 pp. Filho, M. M., T. M. C. D. Lucia, I. Cruz, and R. N. C. Guedes. 2002. Response to the insecticide chlorpyrifos by arthropods on maize canopy. Int. J. Pest Mgt. 48: 203-210.

Godman, R. M., and K. Lancaster. 2003. Eastern Hemlock-Tsuga canadensis Pinaceae-Pine family. United States Department of Agriculture, Forest Service. http://wildwnc.org/trees/Tsuga_canadensis.html. Grant, J. F., D. Palmer, R. Rhea, G. Taylor, P. L. Lambdin, and I. Deal. 2005. Assessment of egg releases for establishment of Sasajiscymnus tsugae on eastern hemlock, pp. 297-303. In B. Onken and R. Reardon [eds.], Third Symposium on Hemlock Woolly Adelgid in Eastern United States, February 1-3, 2005, USDA Forest Service, Asheville, North Carolina. Hain, F. P. 2005. Overview of the third hemlock woolly adelgid symposium (including balsam woolly adelgid and elongate hemlock scale), pp. 3-5. In B. Onken and R. Reardon [eds.], Third Symposium on Hemlock Woolly Adelgid in Eastern United States, February 1-3, 2005, USDA Forest Service, Asheville, North Carolina. Hale, F. A. 2004. The Hemlock Woolly Adelgid: A Threat to Hemlock in Tennessee, The University of Tennessee, Agricultural Extension Service, 4 pp. SP503-G. Howel, W. M., and R. L. Jenkins. 2004. Spiders of the eastern United States: A photographic guide. Pearson Education, 263 pp. James, D. G. 2003. Pesticide susceptibility of two Coccinellids (Stethorus punctum picipes and Harmonia axyridis) important in biological control of mites and aphids in Washington hops. Bio. Sci. Tech. 13: 253-259. Johnson, K., G. Taylor, and T. Remaley. 2005. Managing hemlock woolly adelgid and balsam woolly adelgid at Great Smoky Mountains National Park, pp. 232-233. In B. Onken and R. Reardon [eds.], Third Symposium on Hemlock Woolly Adelgid in Eastern United States, February 1-3, 2005, USDA Forest Service, Asheville, North Carolina. Keller, D. A. 2004. Associations between eastern hemlock (Tsuga canadensis) and avian occurrence and nest success in the southern Appalachians. M.S. thesis, University of Tennessee, Knoxville, 98 pp.

Kizlinski, M. L., D. A. Orwig, R. C. Cobb, and D. R. Foster. 2002. Direct and indirect ecosystem consequences of an invasive pest on forests dominated by eastern hemlock. J. Biogeography 29: 1489-1503. Kohler, G. R. 2007. Predators associated with hemlock woolly adelgid (Hemiptera: Adelgidae) infested western hemlock in the Pacific Northwest. M.S. thesis, Oregon State University, 116 pp. Kovach, A. L. 2004. Assessment of insects, primarily impacts of biological control organisms and their parasitoids, associated with spotted knapweed (Centaurea stoebe L. s. l.) in eastern Tennessee. M.S. thesis, University of Tennessee, Knoxville, 108 pp. McClure, M. S. 1989. Evidence of a polymorphic life cycle in the hemlock woolly adelgid, Adelges tsugae (Homoptera: Adelgidae). Ann. Entomol. Soc. Amer. 82: 50-54. McClure, M. S. 1990. Role of wind, birds, deer, and humans in the dispersal of hemlock woolly adelgid, Adelges tsugae (Homoptera: Adelgidae). Environ. Entomol. 19: 36-43. McClure, M. S. 1992. Effects of implanted and injected pesticides and fertilizers on the survival of Adelges tsugae (Homoptera: Adelgidae) on the growth of Tsuga canadensis. J. Econ. Entomol. 85: 468-472. Meissle, M., and A. Lang. 2005. Comparing methods to evaluate the effects of Bt maize and insecticide on spider assemblages. Agric. Ecosystems and Environ. 107: 359-370. Miliczky, E. R., C. O. Calkins, and D. R. Horton. 2000. Spider abundance and diversity in apple orchards under three insect pest management programmes in Washington State, USA. Agric. and Forest Entomol. 2: 203-215. Nyffeler, M., and D. K. Sunderland. 2003. Composition, abundance and pest control potential of spider communities in agroecosystems: a comparison of European and U.S. studies. Agric. Ecosystems and Environ. 95: 579- 612.

Nyffeler, M., W. L. Sterling, and D. A. Dean. 1994. Spiders make a living. Environ. Entomol. 23: 1357-1367. Orwig, D. A., D. R. Foster, and D. L. Mausel. 2002. Landscape patterns of hemlock decline in New England due to the introduced hemlock woolly adelgid. J. Biogeography 29: 475–1487. Ostman, O., B. Ekbom, and J. Bengtsson. 2003. Yield increase attributable to aphid predation by ground-living polyphagous natural enemies in spring barley in Sweden. Ecol. Econ. 45: 149-158. PDCNR. 2007. DCNR management of hemlock woolly adelgid. Department of Conservation and Natural Resources. Forest Health Fact Sheets. http://www.dcnr.state.pa.us/forestry/woollyadelgid/pestmanagement.aspx. Pekar, S. 1999. Effect of IPM practices and conventional spraying on spider population dynamics in an apple orchard. Agric. Ecosystems and Environ. 73: 155-166. Raupp, M. J., R. E. Webb, A. Szczepaniec, D. Booth, and R. Ahern. 2004. Incidence, abundance, and severity of mites on hemlocks following applications of imidacloprid. J. Arboriculture. 30: 108-113. Rhea, S. R. 1995. Preliminary results for the chemical control of hemlock woolly adelgid in ornamental and natural settings, pp 113-125. In S. M. Salom, T. C. Tigner and R. C. Reardon [eds.], Proceedings of the First Hemlock Woolly Adelgid Review, 12 October 1995, Charlottesville, VA. United States Forest Service. Robert, T. B. 2001. Effects of the removal of overstory hemlock from hemlock- dominated forests on eastern redback salamanders. Forest Ecol. and Mgt. 149: 197-204. Salem, S. A., and M. M. Matter. 1991. Relative effect of neem seed oil and deenate on the cotton leaf worm, Spodoptera littoralis, and the most prevalent predators in cotton fields at Menoufyia governorate. Bull. Fact Sci. Cairo Univ. 42: 941-952.

SAMAB. 2005. Environmental assessment of hemlock woolly adelgid (HWA) control strategies. http://www.saveourhemlocks.org/pdf_docs/HWA_EA1.pdf. Samu, F. 2003. Can field-scale habitat diversification enhance the biocontrol potential of spiders? Pest Mgt. Sci. 59: 437-442. Samu, F., A. Sziraunyi, and B. Kiss. 2003. Foraging in agricultural fields: local „sit-and-move‟ strategy scales up to risk-averse habitat use in a wolf spider. Animal Behav. 66: 939-947. Shaw, E. M., M. Waddicor, and A. M. Langan. 2006. Impact of cypermethrin on feeding behaviour and mortality of the spider Pardosa amentata in arenas with artificial „vegetation‟. Pest Mgt. Sci. 62: 62-64. Sorensen, L. L. 2004. Composition and diversity of the spider fauna in the canopy of a montane forest in Tanzania. Biodiversity and Conserv. 13: 437-452. Stano, P., and C. R. Haddad. 2005. Can agrobiont spiders (Araneae) avoid a surface with pesticide residues? Pest Mgt. Sci. 61: 1179-1185. Stork, N. E. 1991. The composition of the arthropod fauna of Bornean lowland rainforest trees. J. Tropical Ecol. 7: 161-180. Stuntz, S., C. Ziegler, U. Simon, and G. Zotz. 2002. Diversity and structure of the arthropod fauna within three canopy epiphyte species in central Panama. J. Tropical Ecol. 18: 161-176. Thompson, R. S., K. H. Anderson, and P. J. Bartlein. 2006. Atlas of relations between climatic parameters and distributions of important trees and shrubs in North America. U.S. Department of Interior, United States Geological Survey. http://pubs.usgs.gov/pp/p1650-a/. Torres, J. B., and J. R. Ruberson. 2005. Canopy and ground-dwelling predatory arthropods in commercial Bt and non-Bt cotton fields: Patterns and mechanisms. Environ. Entomol. 34: 1242-1256. Townsend, L., and L. Rieske-Kinney. 2007. Meeting the threat of the hemlock woolly adelgid. http://www.ca.uky.edu/entomology/entfacts/ef452.asp.

Trieff, D. D. 1992. Composition of the Coleoptera and associated insects collected by canopy fogging of Northern red oak (Quercus rubra L.) trees in the Great Smoky Mountains National Park and the University of Tennessee Arboretum. M.S. thesis, University of Tennessee, Knoxville, 87 pp. Ubick, D., P. Paquin, P. E. Cushing, and V. Roth. 2005. Spiders of North America: An identification manual. Amer. Arachnological Soc. 377 pp. USDA. 2006. Hemlock woolly adelgid infestation. http://na.fs.fed.us/fhp/hwa/maps/hwa_2006.pdf. Vickerman, G. P., and K. D. Sunderland. 1977. Some effects of dimethoate on arthropods in winter wheat. J. Appl. Ecol. 14: 767-777. Wallace, M. S., and F. P. Hain. 2000. Field surveys and evaluation of native and established predators of the hemlock woolly adelgid (Homoptera: Adelgidae) in the southeastern United States. Environ. Entomol. 29: 638- 644. Ward, J. S., M. E. Montgomery, C. A. S.-J. Cheah, B. P. Onken, and R. S. Cowles. 2004. Eastern hemlock forests: Guidelines to minimize the impacts of hemlock woolly adelgid. USDA Forest Service Northeastern Area State and Private Forestry, NA-TP-03-04, 27 pp. Welch, H., and G. Haddow. 1993. The world checklist of conifers, Landsman‟s Bookshop Ltd. Herefordshire, England. pp. 405-426. Werle, C. T. 2002. Insects associated with southern magnolia (Magnolia grandiflora L.) in east Tennessee. M.S. thesis, University of Tennessee, Knoxville, 78 pp. Wise, D. H. 2006. Cannibalism, food limitation, interspecific competition, and the regulation of spider populations. Annu. Rev. Entomol. 51: 441-465. Yamasaki, M., R. M. DeGraaf, and J. W. Lanier. 1999. Wildlife habitat associations in eastern hemlock-birds, smaller mammals, and forest carnivores, pp. 135-143. In K. A. McManus, K. S. Shields and D. R. Souto [eds.], Proceedings, Symposium on Sustainable Management of Hemlock

Ecosystems in Eastern North America, June 22-24, 1999. USDA Forest Service, Durham, NH. Yu, G., M. E. Montgomery, and D. Yao. 2000. Lady beetles (Coleoptera: Coccinellidae) from Chinese hemlocks infested with the hemlock woolly adelgid, Adelges tsugae Annand (Homoptera: Adelgidae). The Coleopterists Bull. 54: 154-199. Zilahi-Balogh, G., L. T. Kok, and S. M. Salom. 2002. A review of world-wide biological control efforts for the family Adelgidae, pp. 129-140. In B. Onken, R. Reardon and J. Lashomb [eds.], Proceedings; Hemlock Woolly Adelgid in the Eastern United States Symposium, February 5-7, 2002, Rutgers University, East Brunswick, New Jersey.

Appendix

Appendix A. Woolly mass counts (0.3 m branch) from eastern hemlock trees (Fall treatments)

Treatments Total % No of woolly number of Infested Woolly masses per terminals Infested terminals terminals masses terminal No treatment 1542 46 3.0 105 0.07 Horticultural oil 1601 21 1.3 48 0.03 Soil drench 1699 59 3.5 68 0.04 Tree injection 1407 11 0.8 12 0.009 Soil Injection 1485 37 2.5 84 0.06

Appendix B. Woolly mass counts (0.3 m branch) from eastern hemlock trees (Spring treatments)

Treatments Total % No of woolly number of Infested Woolly masses per terminals Infested terminals terminals masses terminal No treatment 1584 47 2.9 101 0.03 Horticultural oil 1690 32 1.8 70 0.04 Soil drench 1640 45 2.7 120 0.07 Tree injection 1673 33 1.9 59 0.04 Soil Injection 1537 3 0.19 70 0.04

Appendix C. General description of eastern hemlock used to assess effect of imidacloprid and horticultural oil on predators, Indian Boundary Campground, Cherokee National Forest, Tennessee, 2005-2007.

Tree Time Height Crown Ratio Foliage Crown No. Treatment (F/S)* (meters) (%) Color Appearance Condition

93 No Treatment F 17.0 95 DG* healthy Full 5 No Treatment S 17.9 100 DG fair Full 78 Tree Injection F 10.3 80 DG healthy Full 73 Tree Injection S 10.6 80 DG healthy Full 97 Horticultural Oil Spray S 18.5 95 DG healthy Full 77 Horticultural Oil Spray F 19.6 95 DG healthy Full 82 Soil Drench F 18.2 95 DG healthy Full 83 Soil Drench S 18.7 95 LG healthy Full 7 Soil Injection S 17.0 95 DG healthy Full 9 Soil Injection F 17.3 95 DG healthy Full 94 No Treatment S 12.1 90 DG healthy Full 98 No Treatment F 12.8 90 DG healthy Full 74 Horticultural Oil Spray F 20.7 95 DG healthy Full 6 Horticultural Oil Spray S 22.2 95 DG healthy Full 84 Tree Injection S 6.4 95 DG healthy Full 80 Tree Injection F 6.5 95 DG healthy Full 99 Soil Injection F 12.4 95 DG healthy Full 11 Soil Injection S 18.7 95 DG healthy Full 96 Soil Drench F 13.8 90 DG healthy Full 75 Soil Drench S 9.9 90 DG healthy Full 70 Horticultural Oil Spray F 10.2 100 DG healthy Full 86 Horticultural Oil Spray S 7.6 100 DG healthy Full 92 No Treatment F 7.4 80 DG poor Full 10 No Treatment S 8.8 80 DG poor Full 76

Tree Time Height Crown Ratio Foliage Crown No. Treatment (F/S)* (meters) (%) Color Appearance Condition 81 Soil Injection S 14.6 80 DG healthy Full 76 Soil Injection F 16.1 80 DG healthy Full 88 Soil Drench F 13.7 80 DG healthy Full 95 Soil Drench S 14.3 90 DG healthy Full 87 Tree Injection S 10.0 90 DG healthy Full 85 Tree Injection F 9.4 90 DG healthy Full *F=Fall, S=Spring **DG=Dark green, LG=Light green

Vita

Abdul Hakeem was born in 1972 in Gilgit, Pakistan and raised in Danyore,

District Gilgit where he received his primary and secondary education. He received a Master of Science Degree in Agriculture from North West Frontier

Province (NWFP) Agricultural University Peshawar. Abdul Hakeem worked with

North South Seeds, Aga Khan Rural Support Program Gilgit for three years. He has worked as an Agriculture Officer in Department of Agriculture, Gilgit since

2001 and currently is on study leave. He received a Fulbright scholarship from

Department of States to pursue Master of Science Degree in Entomology and

Plant Pathology. In January 2006, he started his Master of Science Degree at the

University of Tennessee in Department of Entomology and Plant Pathology under supervision of Dr. Jerome Grant. During his M.S. program at the University of Tennessee, he presented several oral and poster presentations. Abdul

Hakeem is a member of the Entomological Society of America, Entomological

Society of America, Southeastern Branch, Tennessee Entomological Society, and Gamma Sigma Delta Agricultural Honor Society.