The d e of social behaviour of Ret i cu l i tnmes~v ipes (Kollu) (Isoptera: Rhinotermitidae) in defence against the fungal pathogen Metarhizium

ansiopliae (Metschnikoff) Sorokin (Deuteromycotina: Hyphomycetes)

A thesis submitted in conformity with the requirements for the degree of Master of Science in Forestry

Graduate Department of Forestry University of Toronto

O Copyright by Barry H. Strack 1998

National Library 191 ma* Bibliothèque nationale du Canada

Acquisitions and Acquisitions et Bibliographie Services services bibliographiques

395 Wellington Street 395. rue WeiIington OttawaON K 1 A O M Ottawa ON K1A O N 4 Canada Canada

The author has granted a non- exclusive licence allowing the National Library of Canada to reproduce, loan, distribute or sell copies of th is thesis in microform, paper or electronic formats.

The author retains ownership of the copyright in this thesis. Neither the thesis nor substantial extracts fiom it may be printed or otherwise reproduced without the author's permission.

L'auteur a accordé une licence non exclusive permettant à la ~ibliotheque nationale du Canada de reproduire, prêter, distribuer ou vendre des copies de cette thèse sous la forme de microfiche/nlm, de reproduction sur papier ou sur format électronique.

L'auteur conserve la proprieté du droit d'auteur qui protège cette thèse. Ni la thèse ni des extraits substantiels de celle-ci ne doivent être imprimes ou autrement reproduits sans son autorisation.

ABSTRACT

The role of social behaviour of Reticul i termesflavipes (Kollar) &optera: Rhinotennitidae) in defence against the fun@ pathogen Metarhiziurn anisopl iae

(Me tsdinikoff) Sorokin (DeuteromycotM: Hyp homyce tes)

Master of Science in Forestry 1998

Barry H. Stradc

Graduate Department of Forestry

University of Toronto

The feasibility of ushg the entomopathogenic fungus, Me ta rhiz ium

a nisopl iae (Metçchnikoff), for the biological control of Re ticuli termesflavipes

(Kollar) was investigated. Individual termites were removed h m their colonies

and dusted with the conidia of M. anisopliae, then immediately rehimed among

their nestmates. Termite interactions resulted in the dissemination of conidia

through populations maintained in petri dishes, resulting in termite mortality.

These interactions, however, were mostly antagonistic. It was observed that

individu& carrying the conidia elicited aggressive behaviour in their nestmates.

Additional studies reveaied that termites are able to recognize individu& dusted

with the conidia of M. anisopliae and prevent them from integrating into the

colony. In a soi1 environment, dusted termites were frequently buried alive by

their nestmates or groomed untii no conidia remained on the5 cutide. Further

investigations showed that termite agonism is elicited by the presence of M.

anisopliae on termite cutide, but not by the symptoms associateci with fungal

infection.

I would first like to thank my supervisor Dr. Timothy G. Myles. Without his

guidance, encouragement and encyclopedic knowledge of termites, this research

project would not have been possible. Collectively, the members of my cornmittee,

Dr. Dave Malloch, Dr. h d y Kenney, and Dr. Çandy Smith, were a weU-spring of

knowledge and experience, thereby ensuring that my researdi remained focused and

of the hghest caliire. Individually, each was a source of inspiration and insight and

1 extend to them my çincere gratitude. I would also like to thank the biinistry of the

Environment for Ontario and the Rosamind Gilles Fellowship from the University

of Toronto for providing financial support during the initial stages of this project.

Desenring of special adaiowledgement and thanks are Ralph Toninger,

Dr. Devaiah Munivanda, and John Sisson, three friends at the University of

Toronto who frequently provided invaluable assistance during innumerable

moments of crisis. 1 must also thank John IMcCarron and Ian Kennedy for

extending to me their expertise, with very litüe grumbling, whenever needed.

My wife, Susan Strack, is to be thanked most of dl. Her love, support and

kindness is magical and it is entirely because of her that 1 was able to undertake and

complete this most interesting investigation.

iii

TABLE OF CONTENTS

TITLE PAGE ABSTRA- ACKNOWLEDGMENTS LIST OF TABLES

CHAPTER ONE Introduction to the biologicd control of the eastem subterranean termite with the h g a l pathogen

Metarhiz ium anisopliae (Metschnikoff) Sorokin

Termite biology and distribution Conventional termite control Biological control Mection of insect hosts by entomopathogenic fun@ h e c t responseç to huigal infection Subterranean termite problem in Ontario The research project

CHAPTER TWO The transmission of the fungal pathogen Meta rh iz ium a n isopl iae (Deuteromycontina: Hyphomycetes) Sorokin within laboratory colonies of the eastem subterranean termite (Isoptera: Rhinotermitidae)

ABSTRACT

INTRODUCTION

Sub terranean termite colonies Termite control nie huigal biocontrol agent Metarhiz iu m anisopliae (Metschniko fO Termite social behaviour

I . . Il

iii vii

Termite collection

Termite populations in bare pehi dishes Termite populations in petri dishes with agar Termite populations in soil cups Behâviourd defences in the soi1 environment Data analysis

RESULTS

Termite populations in bare petri dishes Termite populations in petri dishes with agar Termite populations in soil cups Behavioural defences in the soil environment

DISCUSSION

CONCLUSION

CHAPTER THREE Termite recognition of nestmates infected with the fungai pathogen, M e tarhizium anisopl iae (Metschnikoff) Sorokin

ABSTRACT

INTRODUCTION Termite agonism Interspeafic agonsim Agonistic behaviour in termites of the same colony Recognition factors Research objectives

Termite collection Coating of termites in conidia pathogenic and non-pathogenic conidia Infection of termites by M. anisopliae

The effects of conidia on nestmate recognition The effects of in€edion on nestmate recognition

DISCUSSION

CHAPTER FOUR The influence of density and group behaviour in the dissemination of pathogenic conidia in laboratory colonies of the eastern subterranean termite (Içoptera: Rhino termi tidae)

ABSTRACT

INTRODUCTION

Disease in termite populations Research objective

Density experiment Group effect on fungal growth

RESULTS Density experiment Group effed on fungal growth

DISCUSSION

CONCLUSION

CHAPTER FIVE Thesis stllzunary and conclusions

LIST OF TABLES

TABLE 21 Types of behavioural responses displayed by populations of ReficulitermesfIaoipes following the introduction of a nestmate dusted with the conidia of Meta rhiziurn a nisopl iae.

TABLE 2.2 Frequency of behaviour in petri dishes (n=5 dishes) containing ReticulitermesJlaoipes following the introduction of two termites dusted with increasing quantities of Metarhiz ium anisopliae conidia.

TABLE 2.3 Mean mortality among groups of Re ticrilitermesflav ipes after 14 days in bare petri dishes following the introduction of two termites dusted with increasing quantities of Mefarhiz ium misopl iae conidia.

TABLE 2.4 Frequency of behaviour observed in petri diçhes of sterile agar (n=5 dishes) containing Reticuliterrnesflavipes following the introduction of two termites dusted with increasing quantities ofMetarhizium anisopliae conidia.

TABLE 2.5 ;Mean rnortality among groups of Re t icu litermes flaa ipes after 14 days in petri dishes of sterile agar following the introduction of two termites dusted with increasing quantities of

Me ta rhiz i u m anisopliae conidia.

TABLE 2.6 Mean mortality of groups of Ret iculitermesflaoipes after 14 days in soi1 cups foliowirtg the introduction of 12 termites dusted with different doses of Meta rhizium anisopl iae conidia.

TABLE 27 Mean number of different behavioural responses displayed by populations of Reticulitermesflnaipes following the introduction of termites dusted with either the conidia of iMetnrhizium anisopliae orPenicillium breoicompactum, or with talcum powder.

vii

TABLE 2 8 Frequenqr of Metent types of behaviour disphyed in observational chambers (n=5 Chambers) containing ReticulitermesfIauipes following the introduction of two termites dusted with eitherhfetn rhiziurn anisopliae or Penicilliurn breoicompactum, or with talcum powder (conttol).

CHAPTER THREE

TABLE 3.1

TABLE 3.2

TABLE 3.3

TABLE 3.4

Types of behavioural responses displayed by populations ofRet icu litermesflaaipes following the introduction of a nestmate either coated with the conidia of Meta rhiz iu m an isop line, suffering from mycosis, or dead h.om fun@ infection. 58

Frequency of d m behaviour observed in petri dishes (nt5 dishes) of Ret iculi t e r m e s f l a ~ i p e s foIlot\.ing the introduction of a single termite dusted with either the conidia of Metarhiz ium anisopliae, Penicillium brevicompactum, or with talcum powder.

The mean number of Merent behavioural responses displayed by populations of Ret iculitermesfIavipes in petri dishes (n=5 dishes) following the introduction of a single termite either coated in the conidia of Metarhiziurn an isop 1 iae, moribund from fungal infection, or dead.

Frequency of behaviour displayed by Reticuli terrnesJlav ipes in petri dishes (n=5) foUowing the introduction of a single treated nestmate.

CHAPTER FOUR

TABLE 4.1 Mean mortality among groups of Reticul i termesflaaipes under different density regimes for 14 days following the introduction of a single termite dusted with Metarhiz ium anisopliae.

viii

TABLE 4.2 The percentage of termite cadavers in pehi dishes (n=10 dishes) that were either buried under termite feces, supporting abundant fungal growth or showing signs of deliquescence after king iniected with the h g a l pathogen Meta rhiziu m anisopliae. 76

Introduction to the biologicd control of the eastern subtemnean termite with the

fun& pathogen Metarhirium ansiopliae (Metsdurikom Sorokin

Termites play an important role in the decomposition of wood and the

cyding of nutrients through their consumption of lignoce11ulosic materiai (Martius

1994). While they are an integral component of terrestriai ecosystems, their dietary

habits frequently place them in connict with human interesb. Approximately 300

speaes of termites are considered pests of economic importance, due to the damage

they inflict upon tree crops, lumber, utility p les and residential properties (Gay

1969; Logan et al. 1990).

Termite biology and distribution

There are over 2300 species of termites, in seven fiunilies, comprising the

order Isoptera (Shellman-Reeve 1997). They are approximately distributeci between

4S0N and 4 5 5 and below elevations of 3000 m (Wood 1988). Largely considered

tropical insects (Krishna 1970), termites are most diverse in the Ethiopian (Bouillon

1970), Oriental (Roonwal1970) and Australian regions (Gay and Calaby 1970).

Relatively few speaes have become established in the cooler Nearctic and Palearctic

regions (Weesner 1970; Harris 1970). Only two genera, Reticuliterrnesand

Zoo ter mopsis, have naturd distributions which extend into Canada, and both are

reshided to the pa&c coast in southem British Columbia (Weesner 1970; Myles

1995). Accidental introduction of termites occur on a large scaie through shipments

of wood products, and it is likely in this way that the eastern subterranean termite,

Reticuliterrnesflaoipes (Kollar), was introduced to Ontario during the 193û's (Gay

1

1969).

Termite colonies are often established through date pairs. Mate termites

disperse in large tlights fmm mature colonies at certain times of the year and

following a brief flight, select a mate and enter into wood or soil to copulate, lay eggs

and rear brood (Nutting 1969). Alternatively, some subterranean termites are able to

establish colonies through a process referred to as budding whereby a cohort of

neotenic and worker termites separate from the primary colony and create their

own nest (Myles 1988). This type of colony formation cm result in a number of

intercomected nests, or caü, (Noirot 1970) extending over a large area and

supporting a population that can exceed a million individuals ( F o d e r and

Townsend 1996). Controlling infestations of subterranean termites is notoriously

difficult due to the possibility that if disturbed, a single large colony might divide

into severai smder colonies, each capable of sustaining itself and requiring separate

treatment (Pawson and Gold 1996).

Conventional termite control

The method used in controllhg termite infestations depends on the moishue

requirements of the pest species (Spear 1970). Dry-wood termites are able to

establish colonies within pieces of wood that have very little mois&, such as

packkg mates, furnihue and the foundations of homes. The entire colony is self-

contained, without requiring outside sources of moisture. These infestations are

generally treated with chernical insediades that can either be pumped into the

galleries created by termites within infesteci wood or deployed in vapour form as

funiigants that penetrate the entire home (Spear 1970; Scheffrahn et al. 1997).

Conversely, subterranean termites have high moisture requirements and

must dways remain in contact with the soi1 so that they have access to water. As

they begin excavating their galleries within the foundation of a home, they line the

galleries with soil, which is cemented together with saliva and feces. There iç a

constant movement of termites between the surrounding soii and the infested

structure as they replenish the moisture within their galleries. This system allows

them to maintain high humidity while within a dry environment. Consequently,

only a portion of the adud colony may be within the structure, the remainder may

be in the surrounding soil and in adjacent buildings. Because their colonies

encompass large areas, treatrnent for subterranean termites is largely preventative

in nature. Chernical barriers are usually establiçhed in the soil around vulnerable

structures, which prevents the immigration of termites from nearby sites of

infestation (French 1991; Su et al. 1997).

Research has also been conducted on the feasibility of non-chernical barriers,

using materials which subterraneat termites cannot burrow through. Finely

crushed cord, of a diameter ranging between 1 .l8 mm to 2.8 mm, has been

succesçful in laboratory trials at exduding Coptotermes formosanus Shiraki, (Su et

ai. 1991). Stainless steel alloy meshes are available in Australia and are reportedly

very effective (Lenz and Runko 1994), but their cost can be prohibitive.

Uiatomaceous earth, which is us& as an abrasive desiccant in pest control, has also

been investigated, but it does not appear to prevent penetration by termites (Grace

and Yamamoto 1993).

Once subterranean termites become established in a home, it rnay become

necessary to replace all parts of the structure that are infested or have sustained

major damage. The sumundmg soi1 is then treated with chernical insecticides to

block the entry of termites (Spear 1970). However, public resistance to the use of

toxic pestiades in the urban environment has resulted in research into controhg

termite infestations with naturally-occurring pathogens.

Biological control

Numerous organisms have been investigated for controlling populations of

insect pests. Unlike chemical insectiades, pathogenic organisms have the ability to

be self-replicating and can be moved throughout the environment by the behaviour

of their host (Villani and Wright 1990). Entomopathogenic nematodes (Berry et al.

1997), insect parasitoids (Sweezey and Vasquez 1991) and fun@ (Gottwald and

Tedders 19û4) have been used with varying degrees of success to control pests in

agricultural systems. Interest in the biological control of subterranean termites haç

led researchers to assess the kasibility of deploying pathogenic nematodes (Trudeau

1989; Logan et al. 1990), ants and ant-derived compounds (Comafius and Grace 1995;

1996; 1997) to suppress termite populations. Entomopathogenic fun@ have been

widely investigated for controlling subterranean termites (Kramm et al. 1982; Hanel

and Watson 1983; Delate et al. 1995; WeUs et al. 1995; Zoberi 1995; Jones et al. 1996)

and show considerable promise because the social behaviour of termites may

fadtate h g a l dissemination through termite colonies.

The social nahm of termites demands that colony members participate in

activities that require frequent body-to-body contact. For instance, the coordination

of foraging, bail-following and colony defense relies upon the exchange of

semiochemicals between nestmates (Grace 1991) and the transfer of these chernicals

is likely accomplished through mutual grooming and trophollaxis (M&Iahan 1969;

Wilson 1971). Such interactions have been effective at transferring fungd

pathogens between nestmates in laboratory experiments (Myles 1992; Rosengaus and

Traniello 1997). Fungi are also excellent candidates for termite control because the

soi1 environment inhabited by subterranean termites is ideal for the propagation of

certain pathogenic fungi (Zimmerman 1981; Inglis et al. 1997).

Infection of insect hosts by entomopathogenic fungi

Fungal infections, or mycoses, often follow a typical pattern of development.

The h g a l spore, or conidium, is the infective agent and adheres to the insecYs

cutide through a combination of electrostatic forces and mucous production

(Bidodika and Khachatourians 1994). Before conidial germination can occur, the

fungus must recognize the insect as an appropriate host. This requiTes that the host

cuticie meet the physiological (Tanada and Kaya 1994; Hajek and S t teger 1994) and

nutritional requirements (St. Leger et al. 1994; hiilner et al. 1996) of the conidia. If

the h g u s were to corne into contact with an incompahile host species, the fungus

may fail to develop (St. Leger 1993).

Penetration of the host's cutide is accomplished through a combination of

mechanical pressure and cuticle-degrading proteases secreted by the fungus (Goettel

et al. 1989). Once the body cavity, or hemocoel, of the insect has been readied, the

fungus invades the intemal organs and the host evenhially dies, usually from

damage to vital organs or by the various toxins that the fungus produces (St. Leger

1993). The Me cycle of the h g u s is complete when it emerges through the host's

cutide and b e g h sponilating (Hanel 1982). At this stage, the insect cadaver is a

source of infective propagules and is capable of spreading the disease. Lacking a

suitable host, the conidia of some entomopathogenic fungi are able to persiçt in the

soil for up to six years, all the while retaining their infective capabilities (Keller and

Zimmerrnan 1989; Weseloh and Andreadis 1997).

Insect responses to fungal infection

The insect immune system consists of several layers of mechanical and

cellular defenses designed to prevent infection by pathogenic fungi. The first

b d e r encountered by a huigus is the array of numerous hairs, or setae, whidi

cover the inçects exoskeleton. These structures can suspend an infectious conidium

above the cuticle and its supply of nuhients, where it will eventually die (St. Leger

1991). If the surface of the host is reached, germination and hyphal growth c m be

inhibited by the presence of certain lipids found within the the layer of wax on the

cutide (Lecuona et al. 1997). In addition, hyphal invasion of the cutide can initiate

melanization directly below the developing fungus, preventing penetration (St.

Leger 1991; Tanada and Kaya 1993). Furthemore, the extrusion of hemolymph into

the wound created by the hyphae can seal the route of penetration and initiate

cellular and humoral immunoresponses (Ratdiffe 1993). If the hingus succeeds in

breaching the integument and reacheç the hemocoel, it is likely to be recognized by

free-floating hemocytes, which rnay destroy it through phagocytosis or

encapsdation (Gillespie et al. 1997). The recent discovery of antifmgal proteins in

insect hemolymph rnay provide additional defence against invasion (Gillespie et al.

1997).

The behaviour of insects rnay change dramatically when they become infected

by parasites or pathogens, possibly affecting the dessemination of disease (Horton

and Moore 1993). Inçects that are n o m d y nochunal rnay become active during the

day, or insects that are predominantly subterranean rnay move to the soil surface.

These authors are of the opinion that such behaviour rnay benefit the overd fitness

of the host, as it results in the removal of the infected individual from the viQnity

of its kui, reducing the lïkelihood k t they will become infected.

Social insects rnay take an active role in maintainhg the health of their

colonies by removing nestmates that are near death. Wilson (1971) and Holdobbler

and \ ' i l i n (1990) des& how ants are able to recognize dead or dying nestmates

and wiu remove them from the colony. Similar observations of necrophoric

behaviour in dry-wood termites by Pearce (1984) reveal that healthy termites bury

si& and dead nestmates within chambers in their nests, usually after covering

them in feces or resinow secretions. The apparent ability of termites to recognize

and isolate morbid and dead nestmates may h t r a t e attempts at controlling

termites with infectious disease as it may M t the exposure that healthy termites

have to diseased members of the colony.

Subterruiean termite problern in Ontario

The eastem subtemean termite, K. j la v ipes (Isoptera: Rhuiotermitidae), is

found in 32 muniQpalities in southern Ontario (Myles and Grace 1991). By 1988,

almost 5000 homes in Toronto had been treated for termite infestations. It has been

estimated that termite colonies in Toronto encompass entire city blocks and may

contain over a million individu& (Grace 1990; Myles 1996). Traditional treatments

in Toronto entail the use of soi1 termiticides, which a d as diemical barriers to

foraging termites. However, these barriers do not suppress termite populations and

are susceptible to termite penetration as the chernicals leach into the soi1 or degrade

over time. The use of entomopathogenic fun@ for suppressing termite populations

may be an alternative to cherrtid treatments, though it is necessary to determine

whether termites carrying the conidia of pathoge~c fimgi are able to mix freely with

the colony.

The research projed

This thesis assesses the feasibilïty of using the h g a l entomopathogen

Me fa r h i z i u m a n isop l iae (Metsdmikoff) as a biological control agent against the

eastern subterranean termite, Reficulitermesflaoipes (KoUar). The research is

8

divideci into three areas. First, investigations were carrieci out to determine if

termite social behaviour is effective ai tansferring conidia between nestmates. Also

investigated was the question of whether termites are able to recognize individuals

that are infected by pathogenic h g i and prevent their mixing with the rest of the

colony. The second area of researdi determined how termites are able to recognize

nestmates that are carrying the conidia of M. a nisopliae. The final topic investigated

the influence of termite density and group behaviour on the epizootiology of

h g a l diçeases in termite coIonies.

Berry, RE, Liu, J. and E. Groth 1997. Efficacy and persistance of Heterorhabditis mareln t us (Rhabditida: Heterorhabditidae) against root weeviIs (Coleoptera: Curculionidae) in strawberry. Environ. Entomol. 26: 465-470-

Bidodùca, M.J. and G.G. Khachatourians. 1994. Basic proteases of entomopathogenic fun@ ciiffer in theu absorption properties to insed cutide. J. Invertebr. Paîhol. 64: 26-32.

Bouillon, A. 1970. Termites of the Ethiopian region. pp. 154-273. I n K. Krishna and F.M. Weesner (eds) Biology of termites, vol- 2. Academic Press, New York.

Cornelius, M.L. and JX. Grace. 1995. Laboratory evaluations of three ant species with the formosan subterranean termite (Isoptera: Rhinotermitidae). Sociobiology 26: 291-298.

Cornelius, M.L. and J.K. Grace. 1996. Effect of two ant speties (Hymenoptera: Fromicidae) on the foraging and survival of the Formosan subtemanean termi te (Isoptera: Rhuiotermitidae). Environ. Entomol. 25: 85-89.

Cornelius and Grace. 1997. Effect of termite soldiers on the foraging behavior of Copto te rmes formosanus (Isoptera: Rhinotermitidae) in the presence of predatory ants. Çociobiology 29: 247-253.

Delate, KM., Grace, J.K., and C.H.M. Tome. 1995. Potential use of pathogenic fun@ in baits to control the Formosan subterranean termite (Isopt., Rhinotermi tidae) J. Appl. Ent. 119: 429-433.

Forschler, B.T. and M.L. Townsend. 1996. Mark-release-recapture estimates of Re t ic u 1 i ter m es spp. (Isoptera: Rhinotermitidae) colony foraging populations from Georgia, USA. Environ. Entomol. 25: 952-962.

French, R.J.R. 1991. Physical barriers and bait toxicants: the Romeo and Juliet of future termite control. Proceedings of the international research group on wood preservation, Kyoto, Japan. IRG/ WP no. 1503.

Gay, F.J. 1969. Species inhPduced by man. pp. 459-491. In K. Krishna and F.M. Weesner (eds) Biology of termites, vol. 1. Academic Press, New York

Gay, FJ. and J.H. Calaby. 1970, Termites of the Australian region. pp. 393-447. In K. Krishna and F.M. Weesner (4s) Biology of termites, vol. 2 Academic Press, New York.

Gillespie, JP., Kanost, MX., and T. Trenaek 1997. Biological mediators of insea immunity. h. Rev. Entornol. 42611-643.

Gottwald, T.R and W.L. Tedders. 19W Colonization, transmission and longevity of Beauverin bassiann and Meta rhizium anisopliae (Deuteromycotina: Hyphomycetes) on pecan weevil larvae (Coleoptera: Curdonidae) in the soil. Environ. Entomol. 13: 557-560.

Goettel, US., St Leger, RJv Rizzo, N.W., Staples, RC. and D.W. Roberts. 1989. Ultrastructural localization of a cutide-degrading protease produced by the entomopathogenic h g u s Metarhizium anisopliae duMg penetration of host (Mnnd uca sexta) cutide. J. Gen Microbio. 135: 2233- 2239.

Grace, J.K 1990. Mark-recapture studies with Re t ic ul i ter mesj7ao ipes (Isoptera: Rhinotermitidae). Sotiobiology 16: 297-303.

Grace, J.K. 1991. Semiochemicd meàiation and manipulation of Re t ic u li ter mes behaviour (isoptera: Rhinotermitidae). Sociobiology 1 9 147-162.

Grace, *.K. and R.T. Yamamoto. 1993. Diatomaceous earth is not a barrier to forrnoçan subterranean termites (Isoptera: Rhinotermitidae). Sociobiology 23: 25-29.

Hajek, A.E. and RJ. St Leger. 1994. Interactions between h g a l pathogens and insect hosts. h u . Rev, Entomol. 39: 293-322.

Hanel, H. 1982. The Me cycle of the insed pathogenic h g u s Metarhizium anisopliae in the termite Nasutitermes exitiosus. Mycopathologia 80: 137-145.

Hael , H. and A.L. Watson. 1983. Prelimhary field tests on the use of Metnrhiziurn anisopliae for the control of Nasufitermes exitiosus (Hill) (Isoptera: Termitidae). Bull. ent. Res. 73: 305-313.

Harris, W.V. 1970. Termites of the Palearctic region pp. 295312 1 n K. Krishna and F.M. Weesner (eds) Biology of termites, VOL 2. Academic Press, New York.

Holdobbler, B. and E.O. Wilson. 1990. The an&. Belknap Press, Cambridge. PP- 644-

Horton, D.R. and J. Moore. 1993. Behavioural effects of parasites and pathogens in insect hosts. pp. 107424. N E Bedcage, SN. Thompson and B.A. Federia (eds) Parasites and pathogenç of insects, vol. 1. Academic Press. New York

hglis, GD., Johnson, D.L. and M.S. Gwttel. 1997. Effects of temperature and d g h t on mycosis (Bea U V e r ia bassiana) (Hyphomycetes: Sympodulosporae) of grashoppers under field conditions. Environ. Entomol. 26: 400-409.

Jones, W. E., Grace, J.K., and M. Tamashiro. 1996. Virulence of seven isolates of Beauoeria bassiana and Metarhizium anisopliae to Coptotermes fo rmo sa n u s (Isoptera: Rhinotermitidae). Environ. Entomol. 25: 481-487.

Keller, S. and G. Zimmermm. 1989. Mycopathogens of soil inçects. In N. Wilding, N.M. Collins, P.M. Hammond, and JE. Webber (eds) Insed-fungus interactions. Academic Press, London. pp 240-265.

Kranun, KR., West, D.F., and P.G. Rockenbach. 1982. Termite pathogens: transfer of the entomopathogen Metarhizium anisopliae between Reticulitermes spp. termites. J. Invert. Pathol. 40: 1-6.

Wb, K 1970. Taxonomy, phylogeny, and distribution of termites. pp. 127-150. l n K. Krishna and F.M. Weesner (eds) Biology of termites, vol. 2. Academic Press, New York.

Lecuona, R, Qement, J., Riba, G., Joulie, D. and P. Juarez. 1997. Spore germination and hyphal growth of B ea uve ria sp. on insect üpids. J. EC& ~ncomol. 90: 119-123.

Lenz, M. and S. Runko. 1994. Protection of buildings, other structures and materials in ground contact frorn attadc by subterranean termites (Isoptera) with a physical barrier-a fine mesh of high grade staUlless steel. Sociobiology 24: 1-15.

Logan, J.W.M., Cowie, R H . and T.G. Wood. 1990. Termite (Isoptera) control in agriculture and forestry by non-chernical methods: a review. Bull. Entomol. Res. 80: 309-330.

Martius, C. 1994. DiversiSr and ecology of termites in Amazonian forests. Pedobiologia 38: 407-428.

McMahan, E.A. 1969. Feeding relationships and radioisotope techniques. pp. 387- 402. In K. Krishna and F.M. Weesner (eds) Biology of termites, vol. 1. Academic Press, New York.

Mher, RJ., Staples, J.A. and G.G. Lutton. 19%. The effect of hurnidity on germination and infection of termites by the Hyphomyce te, Me ta rh iz i u m an isopl iue. J. hvertebr. Pathol. 69: 64-69.

Myles, T.G. 1988. Social evolution front termites to man pp 379-413. I n C.N. Slobodchikoff (ed) The ecology of social behaviour. Academic Press, Boston.

Myles, T.G. 1992. Laboratory studies on the transmission of green muscardine disease in the eastern subtemanean termite with a method for applying appropriate doses of conidia to trappeci termites for release. RAC project No. 5536. Ontario Ministry of the Environment.

Myles, T.G. 1995. New records of drywood termite introduction, interception and extirpation in Ontario. Proc ent. soc. Ont. 126: 77-83.

Myles, T.G. 1996. Development and evduation of a transmissable coating for control of subterranean termites. Sociobiology 28: 373-257.

Myles, T.G. and J.K. Grace. 1991. Behavioural ecology of the eastem subterranean termite in Ontario as a baçis for control. Roceedings of the Ontario Ministry of Environment technology transfer conference, Vol. 2: 547-554.

Noirot, C.H. 1970. The nests of termites. pp. 73-120. I n K. Krishna and F.M. Weesner (eds) Biology of termites, vol. 2. Academic Press, New York.

Nutting, W.L. 1969. Flight and colony foundation. pp. 233 276. In K Krishna and F.M. Weesner (eds) Biology of termites, vol. 1. Academic Press, New York.

Pawson, B.M. and RE. Gold. 19%. Caste differentiaticm and reproductive dynamics of three subterranean termites in the gents Reticulitermes (Isoptera: Rhino termitidae). Çoaobiology 28: 241-251.

Pearce, M.J. 1987. Seals, tombs, mummies and tunnelhg in the drywood termite C ry p tu f er m es (Içoptera: Kalotermitidae). Sociobiology 13: 217-226.

Ratcliffe, N.A. 1993. Cellular defense responses of insects. pp. 267-297. In NE. Beckage, S.N. Thompson and B.A. Federici (eds) Parasites and pathogens of insects, vol. 1. Academic Press, New York.

Roonwal, M.L. 1970. Termites of the Oriental region. pp. 315-W. In K. Krishna and F.M. Weesner (eds) Biology of termites, vol. 2. Academic Press, New York.

Rosengaus, RB. and J.F.A. Traniello. 1997. Pathobioiogy and disease transmission in darnpwood termites, Zooterrnopsis angusficollis (Isoptera: Rhino termitidae), infected with the fungus Me ta r h i z i u m a n isopl iae (Deuteromycotina: Hyphomycetes). Sociobiology 30: 185195.

Scheffrahn, RH., Su, N. and P. Busey. 1997. Laboratory and field evaluations of selected chernid treatments for control of drywood termites (Isoptera: Kalotedtidae). J. Econ. Entomol. 90: 492-502.

Shehan-Reeve, J.S. 1997. The spectnun of eusociality in tennites. pp. 52-93. I n J-C. Choe and B.J. Gespi ( 4 s ) The evolution of social behaviour in insects and arachnidS. Cambridge University Press, Cambridge.

Spear, P.J. 1970. Prinaples of termite control. pp. 577-602. I n K. Krishna and F.M. Weesner (eds) Biology of termites, vol. 2. Academic Press, New York.

St. Leger, RI. 1991. Integument as a b h e r to microbial infections. pp. 284-301. I n K. Biruiington and A. Retnakaran (eds). Physiology of the insect epidermis. CSIRO, Australia.

St. Leger, RJ. 1993. Biology and mechanisms of insect cuticle invasion by Deuteromycete fungal pathogens. pp. 211-225 I n N.E. Beckage, S.N. Thompson and B.A. Federici (eds). Parasites and pathogens of insects, vol. 2. Academic press, New York.

St. Leger, R.J., Bidochka, M.J., and D.W. Roberts. 1994. Germination triggers of Metarhiziurn anisopliae conidia are related to host species. Miaobiology 140: 1651-1660.

Su, N.-Y., Scheffrahn, R.H., and P.M. Ban. 1991. Unifom size particle banie~: a physical exdusion device against subterranean tennites (lsoptera: Rhinotermitidae). J. Econ Entomol. 84: 912-916.

Su, N.-Y., Ban, PM. and R.H. Scheffrahn. 1993. Foraging populations and temtories of the eastern subterranean termite (Isoptera: Rhinotermitidae) in southeastern Florida. Environ. Entomol. 22: 1113-1117.

Su, N.-Y., Chew, V., Wheeler, GS. and R.H. Scheffrahri. 1997. Cornparison of tunnelling responses into insecticide treated soi1 by field populations and labotatory groups of the subterranean tennites (Isoptera: Rhinotermitidae). J. Econ. Entomol. 90: 505-509.

Sweezey, S.L. and E.C. Vaquez. 1991. Biological control of citrus blackay (Homoptera: Aleyrodidae) in Nicaragua. Environ. Entomol. 20: 1691-1698.

Tanada, Y. and H.K. Kaya. 1993. Insect pathology. Academic press, San Diego.

Tmdeau, D. 1989. Selection of entomophilic nematodes for contrd of the eastem subterranean termite, Ret icu 1 i termes Pau ipes (KoUar) (Isoptera: Rhuiotermitidae). Mas ter's thesis, University of Toronto, Toronto, Onbrio. pp 93.

Villani, MM. and R.J. Wright. 1990. Environmental influences on soil maao- arthropod behaviour in agricultural systems. Ann Rev. Entomol. 35: 249-269.

Weeçner, F.M. 1970. Termites of the Nearctic region pp. 477-522. In K Krishna and F.M. Weesner (eds) Biology of termites, vol. 2. Academic Press, New York-

Wells, J.D., Fuxa, J.R., and G. Henderson 1995. Vinilence of four fungal pathogens to Coptotermes formosanus (Isoptera: Rhinotermitidae). J. Entomol. Sci. 30: 208.215.

Weseloh, R.M. and T.G. Andreadis. 1997. Persistance of resting spores of Entomophaga mairnaiga, a fun@ pathogen of the gypçy moth, Lymantria dispar. J. Invertebr. Pathol. 69: 1951%.

W k n , E.O. 1971. The insed societies. Belknap Press, Cambridge. pp 548.

Wood, T.G. 1988. Termites and the soi1 environment. Biol. Fertil. %ils 6: 228-236.

Zimmerman, G. 1981. Effect of high temperatures and artifiaal sunlight on the viability of conidia of Metarhizium anisopliae. J. Invertebr. Path. 40: 3640.

Zoberi, M. 1995. Meta rhiziurn anisopliae, a fungal pathogen of Reticulitermes fIa a ipes (Isoptera: Rhinotermitidae). Mycologia 87: 354-359.

The transmission of the fungal pathogen Metarhizium anisupliae (Metschnikoff) Sorokin (Deuteromycotina: Hyphomycetes) within laboratory colonies of the eastem

subterranean termite asoptera: Rhinotermitidae)

ABSTRACT

The transmission of the h g a l pathogen Me ta rhizium a nisopl iae

(Metschnikoff) through laboratory colonies of the eastem subtmanean termite,

Ref ictîlit ermesflaoipes (KoLiar) was hvestigated. Termites dusted with different

quantities of pathogenic conidia were introduced into groups of termites

maintained in bare petn dishes, petri dishes filled with sterile agar, and cups of soil.

The behaviour of the resident termites towards those that were dusted with conidia

was observed and the mortality recorded. In the bare and agar-filled petri dishes,

resident termites displayed darm behaviour and either groomed or attad<ed the

dusted individu&. In the petri dishes with high doses of conidia, termite

mortality exceeded 95%. In the soil cup tests, mortality was l e s than 30%.

Extensive observations of the interactions between dusted termites and undusted

termites were made in two-dimensional, sand-fUed glas observation chambers.

Termites carrying pathogenic conidia triggered alarm behaviour and the

recruitment of nestmates in all resident populations. Each of the dusted termites

was groomed by their nestmates and two of the five dusted termites were buried

alive in the sand. One dusted termite was kiUed by it's nestmates and then

immediately cannabalized. Termites dusted with the conidia of Penicil liu m

breo icom pactum (Dierckx) did not eliat alarm behaviour or 0th- agonisitic

responçes in the resident population.

Subterranean termites are serious structural pests in the wban environment

and are responsible for hundreds of millions of dollars in control efforts every year

(Su and Scheffrahn 1990). The eastern subterranean termite, Ret iculi termes

flavipes (Kollar), is conçidered the most etonomically important termite in North

America. In 1983, over $470 d o n was spent in nine southeastem states in the

U.S. to control infestations of this termite (Su and Scheffrahn 1990). It was first

introduced into Toronto in the 193û's (Gay 1969) and is now found in 32

muniap&ties throughout çouthem Ontario. By 1988, over 5 000 homes in

Toronto, Ontario, had been treated for termite infestations (Myles and Grace 1991).

Subterranean termite colonies

Subterranean termites forage for food through a diffuse network of galleries

excavateci in the upper 50 cm of the soil column (Delaphane 1991). A typical

subterrmean colony may consist of several interconnected nests, or di, (Noirot

1970), extenduig over an area cxceeding 2000 square metres and supporthg a

population of 1 to 5 million individu& (Grace 1990; Su et al. 1993; Forschler and

Townsend 1996; Myles 1996). Estimates of colony size, however, are frustrateci by

- the ayptic nature of these insects and our lack of information concerning many

aspects of their foraging behaviour (Thorne et al. 1996).

Termite activity in temperate regions is seasonal, presumably due to

temperature and moisture conçtraints (Esenther 1969; Husby 1980). In Toronto,

termites become adive during the early spring, when workers emerge from their

overwintering sites and begin foraging (Grace 1996). Activity demeases in the fall

and the termites begin moving down into the soi1 (MyIes and Grace 1991) or into

17

large pieces of surface wood to oveminter (Grace 1996).

Termite control

The contrd of subterranean termites relies heavily upon the use of

insecticides (Su and Scheffrahn 1990; Scheffrahn et al. 1997; Su et al. 1997). Most

often chemical barriers are established around vulnerable stnictures to prevent the

emigration of termites from nearby infestations (Su et al. 1997). However, this can

result in a substantial infl.ux of insecticide into the environment. Other techniques

uüiize chemical baits which termites feed upon (French 1991) or transmissible

coatings applied directly to the termite cutide (Myles 1996).

The use of nahirally o c m g organisrns and compounds for controbg

subterranean termites has attracted substantial interest in recent yearç.

Entomopathogenic nematodes (Trudeau 1989; Logan et al. 1990) and ants (Cornelius

and Grace 1995,19%, 1997) are king explored as options for controhg termite

populations. Chernical compounds extracted from trees and plants are known to

repel foraging termites and rnay prove to be effective at deterring termites when

applied to wood products (Kard and Mailette 1997) or when used in the soi1 as a

barrier to foraging populations (Cornelius et al. 1997). Entomopathogenic fungi are

potential candidates for controhg subterranean termites. Over 40 species of fun@

have been recovered from colonies of the eastem subterranean termite, some of

which have demonstrateci high Ievels of pathogenitity againçt termite hosts (Zoberi

and Grace 1990a; 1990b).

The h@ biocontrol agent, Metarhizium anisopliae (Metsdurkoff)

The entomopathogenic fungus, Metarhizium nnisopl iae (Metschnikoff), is

known to infect over 200 species of insects (Roberts and Yendol1971) and has been

investigated for controlhg insect pests of sugar cane (Çamuels et al. 1990; Samson

et al- 1994), pecan weevils (Gottwald and Tedders 1984), coduoaches (Kaakeh et al.

1997) and ants (Kelley-Tunis et al. 1995). It has also been tested for use against

termites (Kramm et al. 1982; Hanel and Watson 1983; Delate et al. 1995; Wells et al.

1995; Zoberi 1995; Jones et al 1996; Milner et al. 1997), though mostly under

laboatory conditions.

The infection process begins when the h g a l conidium adheres to the

cutide of the termite. Infection occurs when the conidium germinates and

penetrates the exoskeleton, readiing the host's body cavity, or hemocwl, 18-66

hours later (Hanel 1982). Death of the host is a result of the toxinç released by the

fungus (Roberts and Yendol1971) and from the damage associated with the fungal

invasion of vital organs (Hajedc and St. Leger 1994).

The fungus can be uitroduced into termite colonies by removing a number

of termites from the colony and applying conidia directly to their cutide. The dosed

termites are then immediately retunied to the colony so that they can interact with

their nestmates. The transmission of the pathogen throughout the colony occurs

through normal sociai interactions (Kramm et al. 1982; Zoberi 1995).

Termite social behaviour

Trophallaxis, mutual groorning and brood care are routine interactions

between termites within colonies. However, the intimate nature of these activities

may facilitate the transmission of pathogenic organisms between individu&.

Under controlled conditions, researchers have been able to intiate fun@ epizootics

in laboratory colonies by infecting a srnail portion of the population; these

individuais then spread the pathogen to their nestmates (Kramm et ai. 1982;

Rosengaus and Tranieilo 1997). However, it has been obsemed that termites have

the apparent ability to recognize and avoid sick or deceased nestmates (Su and

Scheffrahn 1990; Zoberi 1995; Jones et al. 1996). Pearce (1987) observed that dead

termites are Irequently covered with fecal materiai or cannabalized, which may

prevent the spread of disease within the colony. Given the prevelance of infedious

organisms within the soil environment and the social nature of termites, it is

highly likely that they have evolved a means of protecting their nests against

outbreaks of h a s e .

Colony defense is adiieved through a complex orchestration of termite

interactions. Alarm behaviour is one such interaction exhibited by termites when

they are exposed to potentially dangerous stimuli, çuch as vibrations or bright lights

(Leis et al. 1994). It can be charaderized as a rapid, vibratory motion of the termite

or by the termite striking irs head capsule against the sides of the nest or gaE!ery

(Sbrenna et al. 1992). This behaviour is often followed by the recnlltment of

nestmates and displays of aggression towards the stimulus (Stuart 1969). Alarm

behaviour and aggression have also been obsemed when termites dusted with the

- c o d a of M. anisopliae are introduced into laboratory populations of R.flavipes

(personal observation). This may have an effect on attempts at the biologieal

control of termites, as the elicitation of such behaviour by termites carrying conidia

may adversely affect the dissemination of the pathogen within the colony.

Thus, the purpose of this set of experiments was to detennine whether the

social behaviour of termites was effective at tramferring pathogenic conidia

between nestmates. It also determmeà whether the expression of agonistic

behaviour towards conidia-carrying termites affecteci their ability to integrate into

the colony and partiapate in nomial social activities.

Termite collection

The eastem subterranean termite, R. fIaoipes, was collecteci during the

sufnmer of 1995 from the Unwin Street composthg faality, a muniapal

composthg site operated by the City of Toronto in Ontario, Canada The site is used

primarily for composting wood waçte and is known to be infested with termites.

Holes were dug into the soil (approximately 1 m deep) and filled with rok of

comigated cardbwd. These were mvered with soil and pieces of wood and left

undistiwbed for two weeks. The site was visited every two weeks and the termites

that had begun feeciing on the cardboard rolls were removed and transported to the

laboratory. They were maintained in the laboratory in transluœnt plastic boxes,

40~28x15 an, (Tamor Rastics Corp. MA, USA) at room temperature and with sheets

of moisteneci comgated cardboard for food.

Fungus culture and harvest

The entomopathogenic fungus, M. an isop 1 iae, used in this study was initially

isolated from an infeded termite cadaver in Toronto, ON. The fungus was

cultured on potato dextrose agar and maintained in a growth chamber

(86%RH and 26OC, Forma Scienafic, OH) for 17 days. The conidia were diçlodged

h m the cultures by inverting the petri dish and gently tapping it on the bottom.

The conidia were then transferred to a glas via1 and stored at 4OC until needed.

Application of conidia to termites

To apply an approximate dose of conidia to the termites, a mass of conidia

was weighed and placed in a 30 ml glas petri diçh. A number of equai-sized

termites (worker caste) were placed in the dish with the conidia and gently swirkd

for two min. This action distrîbuted the conidia evenly over the cutide of the

termites, as verified by observation with a dissecting microscope. To achieve a dose

of 0.1 mg per termite, 1 mg of conidia was placed in the petri dish and 10 termites

were introduced and gentiy swirled. Higher and lower doses were obtained by

adjusting the amount of conidia in the pehi dish. Virtudy no conidia remained

in the petri dish d e r dos- indicating it had adhered to the termites.

Termite populations in bare petri dishes

Populations of 50 termites (worker caste) were placed in petri dishes

containing a moistened (Whatman's #1, 7 cm) fdter paper for food and a saturated

piece of cotton (2x2 cm) to maintain humidity. Two termites canying the conidia

of M. anisopliae were introduced into each petri dish, establishing a ratio of one

dusted termite for every 25 termites in the petri diçh. Six treatments were

performed with termites carrying approximate doses of 0.1, 0.05, 0.025, 0.013, or 0.006

mg of conidia, with O mg as a control. The 0.1 mg treatment was considered a

heavy dose, the 0.05 mg treatment a medium dose, and any quantity l e s then 0.05

mg, a light dose. Each treatment consisted of five petri dishes. The reaction of the

termites towards the individual carrying the conidia was monitored for one hour

and the frequency of alarm behaviour (displayed for at least three min), grooming

(dusted termite groomed continuously for at least three min), and attack behaviou

was recorded. The petri dishes were then transferred to an environmental

chamber (Forma Scientific, OH) and maintaineci (86%RH, 26OC) for 14 days, after

which the average mortality for each of the treatments was determined.

Termite populations in petri dishes with a g u

Populations of 50 termites (worker caste) were introduced into petri dishes

that had been filleci with sterile water agar. A (Whatman's #1,7an) filter paper for

food was placed in each pehi dish before the agar hardened. The petri dishes were

then transferred to an environmental chamber (Forma Scientinc, OH) and

maintaineci (86%RH, 26°C) for seven days to ailow the termites to excavate galleries

within the hardened aga. Following this pend, two dusted termites were

introàuced into each of the petri dishes, establisiung a ratio of one dusted termite

for every 25 termites in the petri dish population Five treatments with termites

carrying approximate doses of either 0.2, 0.1 mg (both considered a heavy dose), 0.05

m g (medium dose), 0.025 fight dose) and O mg of conidia were conduded. Each

treatment consisteci of five petri dishes. The readion of the termites towards the

individuals canying the conidia was obsemed for one hour and the frequency of

d m behaviour (displayed continuously for at least three min), grooming (dusted

termite gmomed continuously for at least 3 min) and attack behaviour was

recorded. The petri dishes were then retumed to the environmental chambers for

14 days, after which the average mortaiity for each treatment was caldated.

Termite populations in soil cups

Populations of 300 termites were estabLished in plastic cups (500 ml, 10

replicates) c o n m g 300 g of saturated, sterile brick sand. Three small watering

holes were created at the bottom of each cup by heaüng a probe and poking it

through the side of the cup. The cups rested in dishes of distilled water, which

was able to enter the cups through the w a t e ~ g holes. The bottom of each cup

24

contained a layer of b h e r Stones (1.4-2.0 mm in diameter) to prevent the

termites h m escaping through the holes. As a source of food, a strip of cardboard

(5x20 cm) was rolled up, secured with a mbber band and buried in each cup with

approximately 30% of the cardboard visi-b1e above the sand. A petri dish was placed

loosely over the top of each cup to maintain humidity. The termites were allowed

to excavate their galleries over a period of 11 days. Following this period, 12

termites dusted with conidia were introduced into each of the cups, establishing a

ratio of 1 dusted termite for every 25 termites in the cup population. The doses of

conidia were 0.1 mg (heavy dose) and 0.05mg (medium dose), and each treahnent

was replicated ten times. The cups were maintained in the laboratory at room

temperature (-Z°C) for 14 days, after which they were disassembled and the

mortality for each treatment was deteRnined.

Behavioural defences in the soi1 environment

Populations of R. flan ipes were established in glas observational chambers

(n=5 chambers, 250 termites per chamber), sirnilar to those desCnbed by Becker

(1%9). The chambers (235x240~7 mm) were £illed with steerile bridc sand, overlaying

a 40 mm high row of barrier sand (2-4 mm in diameter). The barrier sand

prevented the termites h m escaping through four drainage holes situated at the

bottom of the chamber and each drainage hole was plugged with a small quantity of

steel wool. Distilled water was poured into each chamber until the sand was

sahuated and water began to seep through the drainage holes. A strip of cormgated

cardboard (100x20 mm) was bUTied vertidy in the sand, dong one edge of the

chamber, as a source of food. A s m d pit (-30 mm deep) was dug on the side

opposite that of the food. An additional strip of cardboard (2ûxîO mm) was placed

verticaliy in the pit and the termites were gently poured on top of the cardboard

quare. O f the 250 termites per chamber, five were of the soldier caste and the

remahder were workers. The top of each chamber was then fitted with a secure

plastic fid. The obsemation Chambers were maintained at 86%M and 26°C for

seven days to d o w the termites suffiCient time to excavate a network of galleries.

Following the perïod of gallery excavation, three treatments were initiated.

In the firçt treatment, two termites, each duçted with 0.1 mg of M . an isop l iae were

introduced into the observation chambers. In the second treatment, two termites

were dusted with the conidia of the common, non-pathogenic soi1 fungus,

Penicillium breoicompactum (Dierckx). This fungus was used to d e t e d e

whether the presence of non-pathogenic conidia on termite cuticle affected the

behaviour of !5e nestmates. Because the conidia of P. breo icom pact um do not

easily separate £rom the fungal hyphe, termites were introduced into sporulating

cultues and gently rolled until they were evenly coated. For the third treatment,

which was the control, two termites were dusted with talcum powder (Life Brand

O), then introduced into the chambers. Preliminary investigations revealed that

talcum powder did not cause mortalïty in termites, nor did it appear to have an

effed on termite behaviour. Ifs use in this experiment was to ensure that ail

treated termites camed a material on their cuticle that was easily removed through

mutual grooming behaviour (allogrooming). Each of the treatments was observed

for a three hour period and the number of different behavioural interactions

between the introduced termites and the resident population was recorded. The

categories of behavioural interactions (Table 2.1), were sindar to those described by

Stuart (1969) and Adams (1991). Both researchers found that the number of

behavioural responses displayed by termites increased as the intensity of aIarm

stimulus they were exposed to rose. Consequently, by counting the number of

different types of behaviour displayed in the observational Chambers, the relative

26

degree of alarm stimulus couid be detemiined for each of the treatments.

Following the observation period, the chambers were placed in an environmental

chamber (86%RH, 26OC) for 14 days, to mure that there was sufficient time for dl

aspects of termite behaviour to be expressed and to determine the health of the

colony through direct observation.

Data analysis

Termite mortality in the petri dishes and soil cups was calculateci after 14

days. Mortality data was transformed by an arcsine transformation. Both mortaliv

data and the behavioural data collecteci from the obse~ational chambers were

subjected to an analysis of vaniance (ANOVA, SPSS 1995). A Stüdent-Newman-

Keuls test at a level of a=0.05 was performed for each of the experiments to

determine significant ciifferences between treatments.

Termite populations in bare petri dishes

At the highest dose (0.1 mg), alam behaviour was displayed in each of the

petn dishes following the introduction of the termites carrying the conidia (Table

22). This behaviour continued for approximately 30 minutes. During the

obse~ation pend, one dusted termite was attacked and killed. Termite

aggregations formed around both dusted individuals in each dish, which were then

groomed for the entire hour. After being groomed the dusted termites carrieci few

observable conidia.

Identical results of alarm diçplay and groomhg behaviour were obtained in

27

the medium dose (0.05 mg) treatment, though there were two incidents in which

the dusted termites were attacked by the petri population (Table 22). With the

lower doses, (0.025,0.013 and 0.006 mg), the display of d m , grooming or attack

behaviour was not observeci in any of the diçhes (Table 2.2). In a l l instances, the

introduced termites rnixed freely with the petri dish populations without eliciting

agonis tic behaviour.

Mortality was highest in the 0.1 mg and 0.05 mg treatments (TabIe 23), which

were not significantly different h m each other. Levels of mortality decreased in

the remaining treatments, with the lowest mortality (30% + 18 SD) o c d g in the

treatment that used the least amount of conidia (0.006 mg). The mortality levels

between the higher and lower doses were sigmficantly different (P<0.0005; F=59;

df=5) as shown in Table 2.3. and all treatments had a significantly higher mortality

than the control.

Termite populations in petri dishes with agar

Alatm behaviour was displayed in both of the high dose treatments (0.2 and

0.1 mg) following the introduction of the individu& dusted with conidia (Table

2.4). GrooIIYng, however, was not obsemed. Dusteci termites were killed by their

nestmates in both the 0.2 and 0.1 mg treatments.

There was no display of alarm, grooming or attack behaviour in the 0.05 or

0.025 mg treatment. In each of the dishes, the dusted termites were able to mix

freely with the populations . There was no signifiant difference between the

mortality recorded for the 0.2,O.l and 0.05 mg treatments (Table 2.5) and the low

dose (0.025 mg) had a level of mortality that was si@cantfy lower Uian that of the

higher doses (Pc0.0005; F-59.08; df4 ) .

Termite populations in soi1 cups

Behaviour could not be observeci in the soil cup tests because many of the

galleries constructeci by the termites were hidden from view withui the soil. There

was no significant clifference between the rnortality of the 0.1 and 0.05 mg

treatments (Table 2.6), but both treatments had a level of mortality that was

significantly higher that of the control (Pc0.0005; F=19.17; d62).

Behavioural defences in the soii environment

The termites dusted with M. anisopl iae eliated a response that was greater

then that eliated by the termites dusted with P. breoicompactum (Table 2.7;

F=17.0526; ~0.0a05; dk2). Initial encouniers behveen the termites dusted with M.

anisopliae and a member of the colony consisted of a brief period of contact, after

which the resident termite recoiled, displayed alarm behaviour and immediately

left to r e d t nestmates (Table 2.8). While unattended, the dusted termites

remained relatively motionless in all five replicates, not moving more than two

centimeires into the colony. An aggregation of no more than 10 termites formed

around the dusted individuals in a l l five replicates. Grooming of the dusted

termites was observed in al l replicates and was performed by the termites

comprising each aggregation. During the observation period alann behaviour was

king displayed by the termites that were participating in the grooming and by

termites that had no contact with the dusted individuals. In four of the five

replicates members of the soldier caste were among the first to form the

aggregations, though they did not appear to have contact with the dusted termites.

They were a h the k t to leave the aggregations, generally within 30 minutes of

arriving. AlI dusted termites remaineci motionless during the groorning

procedure. In two of the five replicates, grooming continued for approximately one

29

heur, after which the dusted termites were largely free of conidia. The aggregatiom

dissolved at this point and the dean tennites entered the colony unimpeded. In

two of the replicates, grooming behaviouc subsided and the termites in the

aggregation began burying the dusted termites under grains of sand. Within 30

nin, the dusted termites in both replicates were completely buried within the

substrate. Throughout and immediately following the burial procedure, it was

possible to observe movements of the legs and and antennae of the dusted termites,

indicating they were still alive. However, there were no attempts made by the

dusted termites to resist being buricd or to extricate themselves. In a single

replicate, a dusted termite was k i k i and irmnediately cannabalized by ifs

nestmates. After the 14 day incubation period, al1 five diambers appeared to be

healthy, with no indication that a fungal epizootic had occurred.

Termites dusted with the conidia of P. brevicompactum eliated a

behavioural response that was not signiscantly different from the response

observed in the control (Table 2.7). Encounters between the tennites dusted with P.

brevicompactum and the resident termites did not result in alarm behaviour in

any of the replicates (Table 2.8). Workers were recruited in two of the five

replicates, which then proceeded to groom the dusted termites. No soldiers were

reavited in any of the replicates, nor were there any incidents of attack behaviour.

The period of grooming was of approximately 30 minutes duration, after which the

deaned termites entered the network of galleries. The behaviour displayed in the

controls was simiiar, in that only workers were recniited and the dusted termites

groomed (Table 2.8).

Disasion

The results of the trials conducted in the bare petri dishes suggest that high

levels of mortality are a result of large quantities of conidia being introduced into

the termite populations. As fewer conidia were used, there was a corresponding

deaease in mortality. There was no significant difference between the mortality

observed in the 0.1 and 0.05 mg treatments (97% and 88%, respectively), suggesting

that both doses are suffiâent to transfer conidia to most of the termites in the petri

dishes. It would appear that agonistic behaviour, involving aggregations and the

intense grooming of the dusted termites, was a key factor in the dispersal of conidia

in the petri dish populations and the resulting high levels of mortaüty. This,

however, can be considered artificial transmission, as it depends upon the dusted

termite king recognized as a source of alarm stimulus. Ideaily, the dusted termite

should be able to mix freely with the rest of the colony, as this would maximize the

number of termites that would be exposed to the conidia. That high levels of

mortality occurred within bare pehi dishes ïs not surprising, considering they are a

highly artificial environment. For instance, there was little opporhinity for the

termites in the dishes to escape from the individu& dusted with conidia. Without

king able to escape, confrontations between the dusted termites and the petri dish

populations may have been unavoidable.

The lower doses (0.013 and 0.006 mg) did not cause similarily high levels of

mortality (only 40% and 30%, respectively). However, this may not be a result of

insufficient conidia to cause disease, but rather an insufficient quantity to initiate

the agonistic behaviour needed to hansfer the fungus to a larger number of

termites. Agonistic behaviour was infrequent or non-existent at doses lower than

0.025 mg, implying that agonism is dose dependent. The grooming that occurred

when smaller quantities of conidia were used did result in significant mortality

relative to the controls, but significantly tess then that at higher doses.

The narrow diameter of the termite galleries in the aga. dishes may have

prevented aggregations from forming mund the dusted termites. Intense

groomhg (exceeding 3 min) by nestmates was not observed, though occasional

grooming was. Considering that alarm behaviour did occur and that the conidia-

carrying termites were attacked and killed in three of the five aga petri dishes, the

dusted individuals were obviously a source of alarm stimulus. The high levels of

mortaiity recorded for the 0.2,0.1, and 0.05 mg doses (100%, 98% and 94%.

respectively) may refiect the incidental transfer of conidia between termites as they

passed one another in the restricted network of gaileries and during the numerous

periods of allogrooming.

The results of the soil cup tests were surprising because the levels of

mortality recorded for the 0.1 and 0.05 mg dose treatments (23% and 26%,

respectively) were considerably lower than the levels recorded when similar doses

were used in the aga- petri dish tests (98% and 94%, respectively). As well, in both

experiments the ratio of dusted termites to undusted termites remained the same (1

dusted termite to every 25 undusted termites), so that there was a comparable

amount of c o ~ d i a available to infect the healthy termites. The major difference

- between the soi1 cup and petri dish triais was that the cups provided a more

natural? three dimensional substrate for the termites to burrow in. The cups also

contained a greater termite population.

These observations reveal that termite behaviour can r e d t in the

dissemination of pathogenic conidia among populations of termites, but the

behaviour is largely agonistic and effective at tramferring conidia only when

termites are maintained in petri dishes. In a more naturai soil environment,

agonistic behaviour can prevent nestmates dusted with pathogenic conidia h m

penetrating termite colonies. In ail of the observational diambers, individu&

carrying the conidia of M. anisopl iae were recognked by the resident termite

population, as indicated by the rapid and wide-spread display of d a m behaviour,

and prevented from moving freely throughout the network of galleries. The ody

termites exposed to conidia were those that participated in the groomhg or bund

of the dusted termite. Smce aggregations arodd these individuals consiçted of

fewer than ten texmites, less than 10% of the colony within each chamber was

exposed to the conidia. While some of the dusted termites in the chambers did

eventually penetrate the galleries, this occurred only after they were groomed and

mostly kee of conidia.

It remainç unknown as to why the dusted termites in the observational

chambers rarely moved after encomtering a member of the resident population,

even while being buried alive. One might speculate that the resident termites

exerted control over the dusted individual, possibly through pheromones, or that

the dusted termites were behaving altniistically and allowed the colony to bbury

them. The burial of the live, dusted termites suggests that either the colony

perceived the termites as dead or dying and therefore, in need of being buried or

that they were a threat of such magnitude that they had to be inunediately isolateci

from the rest to the colony. The fact that members of the soldier caste were part of

the aggregation further indicates that the dusted individuals were a source of alarm

stimuls, as soldiers are normally recniited when the colony needs defence

(Prestwich 1984).

The la& of termite agonism in the P. breoicompactum treatment suggests

that the termite agonism was speciflcally in response to M. nnisopliae, and not to

c o d a in general. While there were incidents of grooming, it was Srpical of the

normal cleaning that occurs between termites. This raises the possibility that

termites might have evolved the ability to detect h g a l pathogens in their

environment. This hypothesis couid be tested by exposing termite populations to

other types of pathogenic fun@ and comparing the behavioural response to that of

M. unisopliae. The results of such an investigation may have important

implications for the biological control of subterranean termites with

entomopathogm.

Conciusion

Termite social behaviour appears to facilitate the dissemination of h g a l

conidia in populations of termites maintained in petri dishes. This behaviour,

however, is largely antagonistic and evidently the result of direct contact with the

pathogenic conidia of M. unisopliae. In a more natud, soi1 environment, and

with larger populations, termite agonism prevents the dissemination of pathogenic

conidia. The recognition, arrest, grooming and burial of termites dusted with

conidia effectively suppresses h g a l epizootics by preventing the infected

individuals from travelling very far into the colony and mXnimlzhg theit contact

with colony members.

Table 2.1 Types of behavioural responses displayed by populations of Re t iculitermes

Paoipes following the introduction of a nestmate dusted with the conidia of

Metarhiz ium anisopliae.

Termite behaviour

i) Marm behaviour is displayed by resident population

ii) Worker termites are recruited to the dusted termite

iii) Dusted termite is groomed continuously for more than 3 min.

iv) Soldier termites are reariited to the dusted termite

V) Dusted termite is burieci aiive in the soil by nestmates

vi) Dusted termite is attadced, dismembered or killed by nestmates

Table 2.2 Frequency of behaviour in petri dishes (n=5 dishes) containing

Reticulitermes/laoipes following the introduction of two termites dusted with

increasing quantities of Me tarhizzu m n nisopliae conidia.

Dose (mg) Marm Grooming Attack

Table23 Mean mortality among groups of Reticulitermesflaoipes after 14daysin

bare petri dishes following the introduction of hYo termites dusted with inceasing

quantities of Me tarhizium anisopliae conidia.

0.1

0.05

0.025

0,013

0.006

O .O (con trol)

'Means followed by the same letter are not signihcantly different based on Student-

Newman-Keuls cornparisons at an a value of 0.05 (MOVA of mortality data; n=5

dishes; 50 termites per diçh; ÇPSS 1995).

Table24 Frequency of behaviour observed in petri dishes of sterile aga. (n-S dishes)

contaùiing ReticuliterrnesfIavipes EoUowing the introduction of two termites

dusted with inceasing quantities of Meta rhizium anisopliae conidia.

Dose (mg) Alarm Grmming Attack

0-2

0.1

0.05

0.025

0.0 (control)

Table25 hfean mortality among groups of Reticulifermesflnoipes after 14 days in

petri dishes of stede agar following the introduction of two termites dusted with

increasing quantities of Metarhizium anisopliae conidia.

Dose (mg) Mean mortality (% + m)

0.2 100 a*

0-1 98 2 02.8 a

0.05 94 t 05.8 a

0,025 68 + 20.6 b 0.0 (control) 08 + 06.6 c

'Means foilowed by the same letter are not significantly different based on Student-

Newman-Keuls cornparisons at an a value of 0.05 (ANOVA of mortality data; n=5

dishes; 50 termites per dish; SES 1995)

Table 2.6 Mean mortality of groups of Ref iculi termesflav ipes after 14 days in soil

cups following the introduction of 12 termites dusted with different doses of

Metarhizium anisopliae conidia.

- -- - -

Dose (mg) Mean mortality (% + SD) - -- .- -

0.1

0.05

0.0 (control)

* Means followed by the same letter are not signhcantly different based on Student-

Newman-Keuls cornparisons at an a value of 0.05 (ANOVA of mortality data; n=10

cups; 300 termites per cup; S E S 1995)

Table 27 Mean nurnber of different behavioural responses displayed by populations

of Ret i cu l i t e rmes~ao ipes Eollowing the introduction of termites dusted with either

the conidia of Metarhiz ium anisopliae or Penicil l ium brevicompacfum, or with

tdcum powder.

Treatment Behavioural response (mean + SE)

M . anisopliae 4.6 + 0.5 a*

P. breoicornpactum 1.0 + O 3 b

Tdcum powder (control) 1.0 2 0.6 b

'Means followed by the same letter are not significantly different based on Student-

Newman-Keuls cornparisons at an a value of 0.05 (iWOVA of mortaliy data; n=5

dishes; 50 termites per dish; SPÇS 1995).

Table 2.8 Frequency of different types of behaviour displayed in observational

chambers (n=5 chambers) containhg Reticuli termesflaoipes following the

introduction of two termites dusted with the conidia of either Me ta rhit iu m

anisopliae or Penicillium brenicompactum, or with t a l m powder (control).

Behavioural response M. anisopliae P. breaicompactum Control

Display of alann 100

Recrvitment of workers 2 0 0

Grooming of dusted termite 100

Recruitment of soldiers 80

Burial of dusted termite 40

Dusted termite attacked 20

Adams, ES. 1991. Nest-mate recognition based on hentable odours in the termite Microcerotermes arboreus. Roc. Natl. Acad. Çci. 88: 2031-2034.

Becker, G. 1969. Rearing of termites and testing methods used in the laboratory, pp. 351-385. In K. Krishna and F.M. Weesner, (eds) Biology of termites, vol 1. Academic Press, New York.

Cornelius, M.L. and J.K. Grace. 1995. Laboratory evaluatiom of three ant species with the formosan subterranean termite (Isoptera: Rhinotermitidae) . Soàobiology 26: 291-298.

Cornelius, M L . and J.K. Grace. 1996. Effect of two ant speBes (Hymenoptera: Formicidae) on the foraging and survival of the formosan subterranean termite (Isoptera: Rhinotermitidae). Environ. Entomol. 25: 8589.

Cornelius and Grace. 1997. Effect of termite soldiers on the foraging behavior of Cop totermes formosa n u s (Isoptera: Rhinotermitidae) in the presence of predatory anS. Soaobiology 29: 247-253.

Cornelius, ML., Grace, J.K. and J.R. Yaks. 1997. Toxiaty of monoterpenoids and other naturd products to the formosan subterranean termite (Isoptera: Rhinotermitidae). J. Environ. Entomol. 90: 320-325.

Delaplane, KS. 1991. Foraging and feeding behaviour of the formosan subterranean termite (Isoptera: Rhinotermitidae). Soaobiology 19: 101-114.

Delate, KM, Grace, JX., and C.H.M. Tome. 1995. Potential use of pathogenic fungi in baits to control the Formosan subterranean termite (Isopt, Rhinotermitidae) J. Appl. Ent. 119: 429-433.

Esenther, G R . 1969. Termites in Wisconsin. Ann. Entomol. Soc. Amer. 62: 1274-1283.

Forschler, B.T. and M.L. Townsend. 1996. Mark-release-recapture estimates of Re t ic u lite rm es spp. (Isop tera: Rhinotermitidae) colony foraging populations from Georgia, USA. Environ Entomol. 25: 952-962.

French, R. J.R 1991. Physical barriers and bait toxicants: the Romeo and Juliet of future termite control Proceedings of the international research group on wood preserva tion, Kyoto, Japan.

Gay, F.J. 1969. Species introduced by man. pp. 439-491. In K. Krishna and F.M. Weesner, (eds) Biology of termites, vol 1. Academic Press, New York.

Gottwald, T.R. and W.L. Tedders. 1984. Colonization, transmission and longevity of Bea uoerin bassiana and Metarhiz ium anisop liae (Deuteromycotina: Hyphomycetes) on pecan weevii larvae (Coleoptera: Curdonidae) in the soil. Environ. Entornol. 13: 557-560.

Grace, J.K. 1990. Mark-recapture studies with Reticulite rmes flavipes (Isoptera: Rhino termitidae). Sociobiology 16: 297-303.

Grace, J.K. 1994. Protocol for teshg effeds of microbial pest control agents on non-target sub terranean termites (hop tera: Rhino termitidae). J. Econ. Entomol. 87: 269-274.

Grace, J.K. 1996. Absence of overt agonistic behaviour in a northem population of Re ticu l i ter mes Jan ipes (Isoptera: Rhinoterrnitidae). Soaobio!ogy 28: 103-110.

Hanel, H. 1982. The He cyde of the insed pathogenic fungus Metarhiziurn anisopliae in the termite Nasutitermes exifiosus. Mycopathologia 80: 137-145.

Hanel, H. and A.L. Watson. 1983. Preliminary field tests on the use of Metarhizium anisopliae for the control of Naçu titermes ex i tiosus (Hill) (Isoptera: Termitidae). Bull. Ent. Res. 73: 305-313.

Hajek, A.E. and R.J. St. Leger. 1994. Interactions between fungal pathogens and insect hosts. Ann. Rev. Entomol. 39: 293-322.

Husby, D.W. 1980. BioIogical studies on Reticuli termes flaaipes (Kollar) (Dictyop tera, Rhinotermitidae) in southern Ontario. Master's thesis, University of Guelph, ON. 154 pp.

Jones, W. E, Grace, J.K., and M. Tamashiro. 1996. Viniience of seven isolates of Beauoeria bassiana and Metarhizium nnisopliae to Coptotermes formosanus (Isoptera: Rhinotennitidae). Environ. Entomol. 25: 481-487.

Kaakeh, W., Reid, B.L., Bohnert, T.J. and G.W. Bennett. 1997. Toxiaty of unidadoprid in the German cockroach (Didyoptera: Blattellidae), and the synergism between Imidadoprid and Meta rhiziu m anisopl iae (hperfed Fungi: Hyphomycetes). J. Econom. Entomol. 90: 47482

Kard, B.M. and E.J. Mallette. 1997. Resistance of six wood products used in paneling to Reticulitermesflaoipes (Isoptera: Rhinotermitidae). J. Environ. Entomol. 90: 178-182.

Kelley-Tunis, K.K., Reid, B.L. and M. Andis. 1995. Activity of entomopathogenic fungi in free-foraging workers of Camponof u s pennsyloanicus (Hymenoptera: Formiadae). J . Econ. Entomol. 88: 927-943.

Kramm, KR., West, D.F., and P.G. Rockenbah 1982 Termite pathogens: Transfer of the entomopathogen Metarhiziurn anisopl iae between Reticuii termes spp. termites. J . Invert Pathol. 40: 14.

Leis, M., Angelini, I., Sbrenna-M1cciarellî, A., and G.S. Sbrema 1994. Further observations on intercaste communication in Knlo ter mesfla~icoll is: frequency of vibratory movements in different experimental conditions. Et.01. Eco~. Evol. 3: 1-7.

Logan, J.W.M., Cowie, R.H. and TG. Wood. 1990. Termite (Isoptera) control in agriculture and forestry by non-chemical methods: a review. Bull. Entomol. Res. 80: 309-330.

m e r , R.J., Staples, J.A. and G.G. Lutton. 1997. The effect of humidity on germination and infection of termites by the hyphomycete, Me ta rh izi u m an isop 1 ine. J . hvertebr. Pathol. 69: 6469-

bfyles, T.G. 1996. Development and evaluation of a transmissable coating for control of subterranean termites. ÇoQobiology 28: 373-257.

Myles, T.G. and J.K. Grace. 1991. Behavioural ecology of the eastem subterranean termite in Ontario as a basis for control. Proceedingç of the Ontano Ministry of Environment technology transfer conference, Vol. 2: 547-554.

Noirot, C.H. 1970. The nets of termites. pp. 73-120. I n K. Krishna and F.M. Weesner (eds) Biology of termites, vol. 2. Academic Press, New York

Pearce, M.J. 1987. Seals, tombs, mummies and tunnelhg in the drywood termite C ry p t o ter mes (Isoptera: Kalotermitidae). Sociobiology 13: 217-226.

Prestwich, G.D. 19û4. Defense mechanisms of termites. Ann. Rev. Entomol. 29: 201-232.

Roberts, D.W. and W.G. Yendol. 1971. Use of fungi for microbial control of insects. pp. 125-149. I n H.D. Burgess and N. W. Hussey (eds) Microbial control of insects and mites. Academic Press, London.

Rosengaus, R.B. and J.F.A. Tranielio. 1997. Pathobiology and disease transmission in dampwood termites, Z o o termopsis angusticollis (Isoptera: Rhinotermitidae), infected with the fungus Metarhizium anisopliae (Deu teromycotina: Hyphomycetes). Sociobiology 30: 185-195.

SPSS Inc. 1995. SPSS User's Guide. ÇPSS Inc. Chicago, Ill.

Samson, P-R, m e r , RJ., and P.D. McLennan. 1994. Field trials of Metarhizium anisopl iae (Deuteromycoîïna: Hyphomycetes) against ln o p us rubriceps (Diptera: Stratiomyidae) in sugarcane. Environ. Entomol. 23: 749-754.

Samuels, K.D.Z., Pinnock, D.E., and R.M. Bull. 1990. Çcarabeid larvae control in sugarcane using Meta rhiziu rn an isopl ine. J. Invert. Path. 55: 135-137.

Sbrenna, G., S b r e ~ a MicBareIli, A., Leis, M. and G. Pavan. 1992. Vibratory movements and sound production in Ka10 termesjZaoico1 lis (Isoptera: Kalotermitidae). pp. 233-238. I n J. Billen (ed) Biology and evolution of social insects. Leuven University Press, Leuven

Scheffrahn, R.H., Su, N. and P. Busey. 1997. Laboratory and field evaluations of selected chefnical treatments for contml of drywood termites (Isoptera: Kalotermitidae). J. Econ. EntomoL 90: 492-502.

Shiart, A.M. 1969. Social behavior and communication pp 193-229. In K. Krishna F. M. Weesner (eds) Biology of termites, vol. 2. Academic Press, New York.

Su, N.-Y. and RH. Scheffrahn 1990. Economically important termites in the United States and their control. Çoaobiology 17: 77-94.

Su, NA., Ban, P.M. and R.H. Scheffrahn. 1993. Foraging populations and temtorïes of the eastern subterranean termite (Isoptera: Rhinotermitidae) in southeastern Horida. Environ. Entomol. 22: 1113-1 117.

Su, N.-Y ., Chew, V., Wheeler, G.S. and R.H. Scheffrahn. 1997. Cornparison of tunnelhg responses into insecticide treated mil by field populations and laboratory groups of the subterranean termites (Isoptera: Rhinotermitidae). J. Econ. Entomol. 90: 503-509.

Thorne, B.L., Russek-Cohen, E., Forschler, B.T., Breisch, N.L. and J.F.A. Traniello. 1996. Evaluation of mark-release-recaphrre methods for estimating forager population size of subterranean termite (Isoptera: Rhinotermitidae) colonies. Environ. Entomol. 25: 93&.9Sl.

Trudeau, D. 1989. Selection of entomophilic nematodes for control of the eastern subterranean termite, Re ticuli f ermes flan ipes (Kollat) (Isoptera: Rhinotermitidae). Master's thesis, University of Toronto, Toronto, Ontario. pp 93.

Wells, J.D., Fuxa, J.R, and G. Henderson 1995. V ï m c e of four fungal pathogens to Cop to ter mes fo rm osa n us (Isoptera: Rhinotermitidae). J. Entomol. Sci. 30: 208-215.

Zoberi, M. 1995. Me ta rhizium anisopl iae, a fungal pathogen of Reticulitermes flan ipes (Isoptera: Rhuiotermitidae). Mycologia 87: 354-359.

Zoberi, M. and J.K. Grace. 1990a. Fungi associated with the subterranean termite Reticuliterrnesflaoipes inchtario. Mycologia 82: 289-294.

Zoberi, M. and JX. Grace. 1990b. Isolation of the pathogen Beauveria bassiann from Reticulitermes/lavipes (Isoptera: Rhinotermitidae). Soaobiology 16:289-296.

Termite recognition of nestmates infected with the fungai pathogen, Metarhiziurn anisopliae (MetschnikofO Somkin

Populations of the eastern subterranean termite, ReticulitermesfIaoipes

(Kollar) were exposed to termites dusted with the conidia of Me ta rhizi u rn

anisopfiae (Metschnikoff), Penicilliurn breoicompact um @ierdo<) and talcum

powder. Termites dusted with M. a n isopl iae elicited a greater behavioural response

from there nestmates than did termites dusted with P. bre~ icompac turn .

Populations of termites were also exposed to individuals that were coated in viable

or dead pathogenic conidia, and to tennites that were either rnoribund or dead h m

h g a l infection. Agonistic behaviour \vas observed in treatments where termites

were dusted in viable and dead conidia. No agonistic behaviour was observed in

the treatments where termites were rnoribund or dead £rom fungal infection,

showing that the agonistic response is associated with conidia and not hyphd

invasion.

Subterranean termite colonies are vulnerable to predation by insects (Wilson

1971; Cornelius and Grace 1995,1996) and to disease from micro-organisms within

the soil environment (Keller and Zimmerman 1989). The termite soldiers, with

their large mandib1es and toxic secretiom, are well-equipped for defense (Prestwich

1984; Bordereau et al. 1997), but the response of the colony to danger involves the

participation of both soldiers and workers. Any termite, upon detecting an alarm

stimulus, can initiate a cascade of complex behavioural interactions that prepares all

members of the colony for defence (Stuart 1969).

Termite agonism

Agonistic behaviour is a broad category describirtg the condud obsenred when

t e d e s are responding to danger (Thome and Haverty 1991; Shelton and Grace

1996). It indudes all aspects of aggression, uicluduig lunging, biting and the

recruitment of nestmates. Mann behaviour, which rnay initiate or precede

agonism, c m be caused by numerous stimuli, induding bright lights, vibrations or

the presence of a predatory insect (personal observation). This behaviour is

characterized by a set of rapid, viiratory motions (Leis et al. 1994; Stuart 1988) whkh

cm be transmitted to other members of the colony through pheromones, vibrations

(Sbrema et al. 1992), or t a d e stimulation (Stuart 1969). Nestmates c m a h be

reauited in this marner for colony defence. Agonistic behaviour with

Reticuliterrnesflaoipes (KoIlar) has been observecl by Myles (1992) in laboratory

colonies where individual termites were coated with pathogenic hrngi. Under such

circumstances, intracolony agonism might play an important role in prevmting

outbreaks of hfectious disease within termite colonies and may occur frequently

49

when members of a colony become infected by pathogenic org&ms.

Interspecific agonism

Agonistic behaviour is important in the maintenance of termite foraging

territones, particularly when there is cornpetition for resources from other termite

species. Levings and Adams (19&4) observed that the foraging territories established

by two species of Nasu t i te rmes completely occupied an area within a s m d tract of

mangrove forest, yet eadi species maintaineci a discrete temtory and in no place did

they overlap. The authors found that when members of one colony were forced

into a confrontation with members of the other, fighting between the species

ensued. High levels of mortality resulting h m such interspeafic termite

encounters is not unusual (Kettler and Lenthold 1995).

Cornpetitive exdusion may be the outcome when an existing foraging

temtory is encroached upon by a more aggressive termite species. Su and

Scheffrahn (1988) observed that Coptotermes formosanus Shiraki, recently

uitroduced from Hawaii to Florida, was able to successhtlly supplant established

colonies of Reticulitermesflaoipes (Kollar).

Inter-cofony agonism

The expression of hostility towards members of the same species may occur

when individuals from different colonies meet. Howidc and Creffield, (1980),

colieded Cop to ter mes acinaciJorm is (Froggatt) from several different colonies

throughout Austraiia and brought them together under labora tory conditions. They

found Lhat populations from different colonies would attack each other so

ferociously that some of the groups in their çtudy were completely eliminated.

Su and Scheffrahn (1988), however, obse~ed that separate colonies of

51

C. formosanus were able to fuse into a single colony without apparent animosity.

Sinularly, laboratory experïments by Grace (1996) in Toronto fded to find evidence

of intercolony agonism between several colonies of the eastern subterranean

termite, R. paoipes.

A study of agonism in subterranean termites by Clement, (1986), revealed that

levels of termite agoniçm in R. lucifugus (Rossi) v d e d according to season. Duruig

the summer months, when food resources were abundant and the clirnate amiable,

agonism between termite colonies was low and there were signs of inter-colony

fusion. As the winter approached, however, and the environment cooled and

resources became scarce, the colonies began to aggressively defend their respective

foraging territories

Agonistic behaviour in termites of the same colony

A review of termite agonism by Thome and Have* (1991) suggests that

intracolony agonism is quite rare and may be restricted towards termites that are

injureci or ill. However, the observations by Myles (1992) suggest that under certain

conditions, agonistic behaviour between members of the same colony may be

corrunon. In his studies, termites were removed from their laboratory colonies and

coated with the conidia of M. anisopliae. They were then retumed among their

nestmates under the premise that tennite social behaviour (e.g. trophdaxis and

grooming) would spread the fungus throughout the population and initiate an

epizootic. It was observed that termites carrying the conidia lriggered wide-spread

alarm behaviour and agonism in their nestmates, with the termites coated with

conidia frequently king at-tacked and dismembered.

The fungus itself does not appear to have an effect on the behaviour of

subterranean termites. Delate et al. (1995) obsewed that the Formosan subterrartean

52

termite did not appear to avoid contact with the conidia of M. anisopliae when it

was growing on cardboard rab used in termite bait experimenh. Yet, termites

apparently recognize nestmates that carry conidia on their cuticle.

Recognition factors

The recognition of nestmates within a termite colony depends on the abilïty

of termites to perceive and distuiguish a speciûc set of signals. Recognition is likely

a combination of several factors (Thorne and Haverty 1991), one of the rnost

important being termite odour (Stuart 1969). Shelton and Grace (1996) suggest that

termite odour may vary according to the environmental characteristics of the

foraging temtory and the type of food resources located within i t As a result, odour

rnay Vary sigdicantly between termite colonies, allowing termites to make

intercolony distinctions.

The chemistry of the termite cuticle rnay be another recognition factor.

Howard et al. (1982) and Bagneres et al. (1991) found that the cutidar hydrocarbons

of termites vary between species and between castes. The authors were of the

opinion that termites are able to deted these differences in cuticular chemistry and

in doing so, are able to distinguish one termite from another. Adams (1991), found

that when termites encountered conspecifics, they were less aggressive to unfamiliar

relatives than to non-relatives, suggestirtg that some recognition factors may have a

heritable component that is maintained in termite colonies.

Alarm behaviour and agonism may arise when termites fail to deted the

appropriate recognition factors during termite-tetermite encouniers. It is possible

that the presence of h g a l conidia on the cuticle of termites may interfere with

recognition factors, resulting in the elicitation of agonistic behaviour in nestmates.

Research objectives

This set of scperiments explores the means by whidi termites are able to

idenhfy nestmates carrying the conidia of M. annisopliae. The k t experiment

determines if conidia on termite cuticle interferes with the ability of nestmates to

recognize members of their colony.

The second experiment investigates whether termite agonism is eliated by

the pathogenic effects that M. anisopliae has on ifs host. Termites were exposed to

individuals that were: 1) carrying pathogenic conidia, 2) individuals that were

suffering from mycoses and 3) termites that were dead from fungal infection.

Methods

Termite collection

The eastem sub terranean termite, Re t ic u 1 i ter mes flaoipes (KoUar), was

collected from the Unwin Street composüng faQlity, a munigpal composting site

operated by the City of Toronto, during the summer months of 1996.

Termites were trapped in rolls of comgated cardboard that had been placed in

several pits (-1 m deep) dug on site. The traps were visited at roughly two week

intervals and the termites were taken to the laboratory, where they were maintained

in plastic tubs with sheeh of comigated cardboard until needed.

Coating of termites in pathogenic m d non-pathogznic conidia

Termites were dusted with the conidia of Penicil lium breoicornpactum, a

common soil fungus that preliminary studies indicated was not pathogenic to

termites. Because the conidia of P. breoicompactum do not easily separate from the

fungal hyphae, termites were intrduced into a spodating culture and gently rolled

54

until they were covered in conidia.

Termites were also coated in talcum powder (Life Brand O), by plaüng 2 mg of

powder in a clean glas petri dish, to whidi 10 termites (worker caste) were added.

The dish was gently swirled for two min to evenly coat the termites in powder.

Termites were coated with the conidia of M. anisopliae (0.1 mg) by placing 1

mg of conidia in a glas pehi dish and introducing 10 workers. These were gently

swirled for two min to evenly coat the termites in conidia. A control treatment was

also conducted in whkh the termites were swirled gently for two min in a bare petri

dis h.

Eadi treatment was represented by five dishes of 10 termites and each dish

contauied a moistened (Whatman's no. 1) filter paper. Into each dish a single

dusted termite was introduced into each dish and the kequency of alarm behaviour

and grooming was recorded over a three hotu period.

Infection of tennites by M. amsopliae

To determine if termite agonism is elicited by the presence of pathogenic

conidia on termite cutide, termites were coated with the active conidia of M.

a n isopliae and conidia that had been bidogically deactived. To deactivate conidia,

they were soaked in ethyl alcohol(75%) for two ho=, after which the alcohol had

evaporated. An earlier trial hdicated that conidia were effectively killed in this

marner. Two treatments were performed, with termites being coated with dead

conidia in one treatment and viable conidia in the other. Dosing was done by

plaàng 10 termites in a glass petri dish that contained either 1 mg of the dead or 1

mg of viable conidia. The dishes were gently swirled for 2 min to evenly coat the

termites in conidia.

To determine if agonistic behaviour was eiicited by symptoms associated with

infection, termite populations were exposed to uidividuals that were dead or

dying h m h g a l infection. Termites were infected with M. a n isopl iae 24 h prior

to introducing them into the petri dish populations by swirling them in a glas petri

dish with a small quantity of conidia. Termites were also infeded two weeks earLier

in a sirnilar manner and kept in the laboratory until they had died and sponilating

fun@ had emerged h m the cadavers. For a control, termites were exposed to

individuals that had died from exposure to freezing temperatures (-12OC).

For each treatment a single termite was introduced into a population of 10

termites (five dishes per treatment) that were maintained in plastic petri dishes

containing a moistened (Whatman's no. 1) Eiter p a p e The petri dishes were

observed for a period of three hous and the presence of alarm, grooming and attack

behaviour was recorded (Table 3.1). Data was analyzed by perfomiing an analysis of

variance ( S B 1995) on the number of different types of behaviour observed in each

treatmen t.

Effects of conidia on nestmate recognition

N o alarm behaviour was displayed in the treatment where termites were

coated with talcum powder (Table 3.2). However, the dusted termites were groomed

in three of the five replicates. Alarm behaviour was observed in the one of the

replicates where the introduced termites were coated with the conidia of P.

breoicompactum. Grooming of the coated termite occurred in three replicates.

Alarm behaviour and grooming was displayed in all replicates where the termites

were coated with M. a nisop liae. There was no response in the control, except for the

display of alarm behaviour in a single replicate.

Effects of infection by M. nnisopliae on nestmate recognition

There was a signiscant difference in the behavioural response observed

between the five treatments (F-12.187; df--4; pcO.ûûûl), however there was no

merence between the response recorded for the live and dead conidia treatments

(Table 3.3). Similarly, there was no signiscant difference between the behavioural

scores recorded for the treatments involving moniund termites, termites that were

dead h m mycosis, and termites that had died from freezing (Table 3.3). Groomhg,

alarm and attadc behaviour was eliated in both of the conidia treatrnents following

the introduction of the dusted termites, but attack behaviour was absent in the

remaining trea tments (Table 3.4).

Discussion

It would appear that the conidia of P. breuicompactum do not interfere

with termite recognition factors, as their presence on termite cuticle did not once

eliat alarm behaviour in nestmates. This implies that termites are able to

distinguish between M. anisopliae and P. breoicompactum, raising the possibility

that agonism is elicited only by pathogenic conidia. The dusting of termites with

talcum powder also failed to have an affect on termite behaviour, suggesting that

termite odours are not easily masked by absorptive powders, or that recognition

factors involve more than olfactory mes (Thorne and Haverty 1991). It is not

surprising that grooming behaviour was observed in all treatrnents, as social insects

are fastidious in keeping their colony and nestmates dean (Wilson 1971). However,

it is intereshg to note that it was only in the treahnent with M. anisopl iae that

alarm behaviour and grooming occurred in aI3 replicates. This suggests that M.

anisopliae is a source of considerable alarm stimulus.

There was no significant difference in the levels of agoniçm exhibited by the

petri dish populations in which the termites were dusted with either viable or

dead conidia. This suggests that termites do not deted factors associateci with

conidia germination or the initial stages of host penetration. Furthermore, the low

agonism levels in the treatment where the introduced termites were moribund

from mycosis indicates that termites do not appear to respond as strongly to

symptoms associated with h g a l invasion as they do to the presence of conidia on

termite cuticle.

The lack of agonism in the treatment in which the cadaver had sporulating

fun@ emerging frorn it may be a function of its mobility. Stuart (1969) maintains

that a non-moving alarm stimulus is of a lower intensity than that of a stimulus in

motion. Consequently, termites that were dusted with pathoge~c conidia, but

immobile, would not elicit as strong as a reçponse as would dusted termites that

were freely moving throughout the petri dish.

Conclusion

Alarm behaviour and termite agonism appear to be elicited specifically by the

conidia of M. a nisopliae on a termite host. The la& of agonism eliated in the

treatment where the individu& were coated in P. b revicom pa c t u m suggests that

the mere presence of c o d a do not interfere with termite recognition factors or that

conidia, per se, do not eiicit termite agonism. Furthermore, the lack of alarm

behaviour in the treatment in which the termites were moribund from fun@

infection implies that agonism is not elicited by the pathology associated with

mycosis.

Table 3.1 Types of behavioural responses displayed by populations of Ret ic u l i termes

flavipes following the introduction of a nestmate either coated with the conidia of

Me ta rh i z iu m a nisop liae, suffering from mycoses, or dead from h g a l infection.

Termite behaviour

i) Introduced termite is groorned by nestmates for at least 3 min

ii) Narm behaviour is displayed by nestmates for at least 3 min

iü) Introduced termite is attacked or killed by nestmates

Table 32 Frequency of damt behaviour observeci in petri dishes (n=5 dishes) of

Re ticul itermesflao ipes followhg the introduction of a single termite dusted with

either the conidia of Metarhizium anisopliae, Penicillium breoicompactum, or with

t a l m powder.

Alarm behaviour (%)

M . anisopliae

P. breoicompactum

Talcum powder

Control

Table 3.3 The mean number of different behavioural responses displayed by

populations of Reticulitermesflavipes in petri dishes (n=5 dishes) following the

introduction of a single termite either coated in the conidia of Me tarh izi u m

an isop liae, moribund from fun@ infection, or dead.

Treatment Response (mean + SE)

Termite coated with live conidia

Termite coated with dead conidia

Termite morbid from infection

Termite corpse with sponilating fungi

Tennite corpse from freezing

'Means followed by the same letter are not significantly different based on Student-

Newman-Keuls cornparisons at an a value of 0.05 (ANOVA on the number of

different behaviours observed in petri dishes; n=5 diçhes; 10 termites per dish; SES

1995)

Table 3.4 Frecpency of behaviour displayed by Re ticuli termesflao ipes in petri

diçhes (nd) following the introduction of a single treated nestmate.

Trea tment Alarm Grooming Attack

termite coated with live conidia

termite coated with dead conidia

termite infected with h g i

termite corpse covered with conidia

frozen corpse (control)

Adams, ES. 1991. Nest-mate recognition based on heritable odours in the termite Microceroter mes a rboreus. Proc. Nati. Acad. Scï. 88: 2031-2034.

Bagneres, A., KîiIian, A., Clement, J., and C. Lange. 1991. Interspecinc recognition among termites of the genus Reticuli termes: evidence for a role for the cutidar hydrocarbons. J. Chem. Ecol. 17: 2397-2420.

Bordereau, C , Robert A., Van Tuyen, V. and A. Peppuy. 1997. Suitidal defensive behaviour by frontal gland dehiscence in Globitermes sulphureus Haviland soldiers (Isoptera). Insectes. Soc 44: 289-296.

Cornelius, M.L. and J.K. Grace. 1995. Laboratory evaluatiom of three ant species wi th the formosan sub terranean termite (Isoptera: Rhinotermitidae). Sociobiology 26: 291-298.

Cornefius, M L . and JK. Grace. 1996. Effect of two ant species (Hymenoptera: Formiadae) on the foraging and survivd of the formosan subterranean termi te (Isop tera: Rhinotermitidae). Environ. Entomol. 25: 85-89.

Clement, J. 1986. Open and closed societies in Re ticu 1 iter mes termites (Isoptera: - &

Rhin0 termi tidae): geographic and seasonal variations. Sociobiology 11: 311-

Delate, K M , Grace, JK, and C.H.M. Tome. 1995. Potential use of pathogenic fungi in baits to control the Formosan subterranem termite (Isopt., Rhinotermitidae) J. Appl. Ent. 119: 429-433.

Grace, J.K. 1996. Absence of overt agonistic behaviour in a northem population of Ref icul itermesflaoipes (Isoptera: Rhinotermitidae). Çociobiology 28: 103110.

Howard, R.W., McDaniel, C.A., Nelson, D.K., Blomquist, G.J., Gelbaum, L.T. and L.H. Zalkow. 1982. Cutidar hydr~carbons of Reticulitermesoirginicus (E3anks) and their rok as potential species-and-caste recognition cues. J. Chem. Ecol. 8: 1227-1239.

Howick, C.D. and J.W. Creffield. 1980. IntraspeQfic antagonism in Cop to ter mes acinaci form is (Froggatt) ( h p tera: Rhino termitidae). Bull. Ent. Res. 70: 17-23.

Keller, S. and G. Zimmerman. 1989. Mycopathogens of soi1 insects. 1 n N. Wilding, N.M. Collins, P.M. Hammond, and J.F. Webber (eds) Insect- fungus interactions. Academic Press, London. pp 240-265.

Kettler, R. and R.H. Leuthold. 2995. Inter-and intraspeci€ic alarm response in the termite Macrofermes subhyalinus (Rambu). Insectes. Soc. 42: 145-156.

Leis, M., Angelhi, 1, Sbrenna-MicciareUi, A., and G.S. Sbrema. 1994. Further observations on intercaste communication in K n l o t ermesf7a~ico i 1 is: frequency of vibra tory movements in different experimental conditions. Ethol. Ecol. EvoL 3: 1-7.

Levings, S.C. and ES. Adams. 1984. Intra-and interspeQfic temtoriality in Nas u t i ter m es (Isoptera: Termitidae) in a Panamanian mangrove forest. J. Anim. Ecol. 53: 7û5-714.

hdyles, T.G. 1992. Laboratory studies on the transmission of green muscardine diçease in the eastem subterranean termite with a method for applying appropriate doses of conidia to trapped termites for release. RAC Project No. 5536. Ontano Ministry of the Environment.

Prestrvich, G.D. 1984. Defense rnedianisms of termites. Ann. Rev. Entomol. 29: 201-232.

Sbrenna, G , Sbrenna Micciareili, A., Leis, M. and G. Pavan. 1992. Vibratory movements and sound production in Ka 1 o t e r m es P a vicoll is (Isoptera: Kalotermitidae). pp. 23S238. In J. Billen (ed) Biology and evolution of social insects. Leuven Universiiy Press, Leuven.

Shelton, T.G. and J.K. Grace. 1996. Review of agonistic behaviours in the Isoptera. Sotiobiology 2 8 155-176.

Stuart, A.M. 1969. Social behaviour and communication. pp 193-229. In K. Krishna F. M. Weesner (eds) Biology of termites, vol. 2 Academic Press, New York.

Stuart, A.M. 1988. Preliminary studies on the sigru6cance of head-banging movements in termites with special reference to Zoo ter m ops is angus tic01 lis (Hagen) (Isop tera: Hodo termitidae). SoQobi010gy 14: 49-60.

Su, N.-Y. and R.H. Scheffrahn. 1988. Intra-and interspecific cornpetition of the formosan and the eastem subtemanean termite: evidence £rom field observations (Isoptera: Rhinotennitidae) . Sotiobiology 14: 157-164.

Thome, B.L. and M.I. Haverty. 1991. A review of intacolony, intraspecific and interspeaEic agonism in termites. Soaobiology 19: 115145.

Wilson, E.O. 1971. The insect societies. Behap Press, Cambridge. pp 548

The influence of density and group behavioar in the dissemination of pathogenic conidia in laboratory colonies of the eastem subterranean termite

Usoptera: Rhinotermitidae)

Populations of the eastern subterranean termite, Re t ic u li ter mes /lao ipes

(Kollar) were maintained at different densities and exposed to termites dusted with

the pathogenic conidia of M e ta h iziu m an isopiiae (Metsduiikoff) to determine if

mortality is affeded by host density. There was no significant difference between

the mortality recorded from the high, medium and low density regimes. The

number of soldier termites that survived varied according to density. Under the

high density condition, soldier survival was 10%. under medium 47.5% and under

low density 57.5%. Termites were also dusted with conidia and introduced into

populations of termites to determine whether the presence of nestmates affected the

colonization of the cadaver by fungi. Fewer termite cadavers were covered with

sporulating h g i under the group condition. This was likely because termite

cadavers were covered in feces by their nestmates, which appeared to inhibit hgal

growth.

INTRODUCTION

There are over 2200 species of termites in seven families comprishg the

order Isoptera (Shellman-Reeve 1997). Ml species are considered eusocial, with each

colony exhibithg cooperative brood care, a division of labour, and overlapping

generations (Wilson 1971). This social behaviour requires frequent body- to-body

contact, an activity which rnay faditate the transmission of infectious disease in

termite colonies (Kramm et al. 1982; Rosengaus and Traniello 1997) and place

colonies at considerabIe nsk.

Colonies are often founded by a single pair of alate termites that disperse from

nearby nests ai certain h e s of the year (Nutting 1969; Noirot 1970). h i e termitesf

however, experimce very high rates of mortaiity and in some instances fewer than

1% may ever produce a viable colony (Myles 1988). Rosengaus and Tranieilo (1993)

found that over 40% of the founding termites in incipient colonies of Zoo ferrnopsis

angusticollis died hom infection by fun@ and bactena. According to Myles (1988),

even when alate pairs do succeed in mating and manage to rear brood, a very high

percentage of newly establiçhed colonies fail. S m d populations may be partidarly

susceptible to failure because they la& the resources to endure even brief periods of

hardship. Furthemore, s m d colonies may not have the defensive capabilities of

larger termite groups (Shellman-Reeve 1997), making them especially vulnerable to

infectious disease or predation. In a study of social facilitation, Mirmontes

and DeSouza (1996), found that s m d groups of termites (fewer than five

individu&), deprived of water and food, surviveci for only a few days. However, as

the researchers increased the group size, the termites lived longer.

They concluded that soaal interactions within groups may have an affect on

termite physiology, in this case reducing their rate of metaboikm during a period of

67

68

resource deprivation According to Becker (1969) the minimum population

needed to &tain a stable colony in a laboratory environment is approximately 50

individuals, though larger populations are more stable.

Subterranean colonies established through colony fission, or budding, may

have an iniierently more stable population than their terrestrial counterparts.

According to Myles (1988), colonies established through budding enjoy several

advantages. Notably, they start out with large populations of siblings and would be

l a s vulnerable to the hazards encountered by very srnall inapient colonies.

Furthemore, budded colonies inherit valuable ressources from the parent colony,

such as a preexistïng network of galleries, identifid sources of food, and a

famüiarity with the local environment.

In addition to group sue, density has been shown to have an affect on the

performance of termite colonies. Lenz et al. (1984) and Lenz (1985) found that

Cop t o termes lacteus (Froggatt) maintained stable populations and consued more

food when they were at a density of 0.011 grams of tennite per millilitre of substrate.

Colony performance decreased when density fell below or exceeded this value. It is

unlikely that the optimal density of C. lacteus can be directïy extrapolated to other

species, as the spatial needs of termites are bound to vary according to numerous

environmental and ecological conditions. However, it is reasonable to assume that

termite colonies must meet certain density criteria in order to achieve a stable

population.

Disease in termite populations

Wi th respect to infectious disease, recently budded sub terranean colonies

may have the capacity to recover from occasional pathogen outbreaks because of

their high populations. These colonies may have an additional advantage in that

they experience frequent, cydic inbreeding (Shehan-Reeve 1997), which may

affect the immunoresponse of termites. Accordhg to Rosengaus and TrarUeIIo

(1993) incipient colonies founded by inbred primary reproductives are less EeIy to

be affected by infectious disease than are colonies where the primary pair are

outbred. The researchers suggest that mating pairs which originate from within the

same colony share an acquired immunity to the pathogens in their immediate

environment, making it less likely that one would infect the other with a pathogen

to which both are not immune.

The development of epizootics in insect populations depends upon several

key factors. The most important are the presence of a sufficiently Wulent pathogen

and a high host population and density (Keller and Zimmerman 1989). Mature

colonies of subterranean termites may contain over a million ùidividuals (Grace

1990; Su et ai. 1993; Forschler and Townsend 1996) and they inhabit a pathogm-rich

environment. It would be reasonable to assume that they meet the aiteria needed

for the establishment of an epizootic. Yet, there are no reports of large outbreaks of

diçease o c c u r ~ g in subterranean termite colonies. Observations have been made,

however, that suggest termites prevent disease in their colonies by isolathg dead or

infected nestmates (Zoberi 1995). Pearce (1987) observed in dry-wood termites that

dead or dying termites were kequently buried under piles of feces or placed within

galleries that were sealeci by nestmates.

The removai of dead or dying individuals from the nest is common in social

insect societies (Wilson 1971; Holldobler and Wilson 1990). Termites caryhg

pathogenic conidia on their cuticle have been obsewed to elicit a series

of agonistic responses from their nestmates. The ability of the colony to quiddy

recognize and respond to a termite c+g a pathogen would be aiticai in

preventing outbreaks of infectious disease. The reauitment of nestmates and the

transmission of d m behaviour is communicated through a combination of

pheromones and tactile stimulation (Stuart 1%9), making it possible that the

proximity of termites to one another (their density) may affect their ability to

reqmnd to nestmates infected with pathogens. Consequently, termite populations

at relatively high density may be able to respond more rapidly to a termite carrying

pathogenic conidia than would termites at lower density, and thus be better able to

recognize infeded nestmates and suppress outbreaks of infectious disease.

Furthermore, termites rnay bury or cannibalize infected or dead nestrrtates so that

the cadavers do not become a source of disease in the colony, as pathogenic fungi

can emerge from their hosts.

Research ob jertive

This set of experiments was designed to determine whether the abiüty of

termites to protect their colonies from epizootics was affected by the density of the

individu& wi thin the colony . Furthemore, i t inves tigated whether nes tma tes

were able to prevent the secondary growth of pathogenic fun@ from termite

cadavers.

Matends and Methods

Density experiment

Laboratory populations of R./laoipes were introduced into petri dishes of

three different sizes: 3470, and 190 ml, representing different volumes for

termite foraging. Each population consisted of 98 workers and two soldiers.

Termites in the 30 ml dish were considered to be at high density (0.0087g/ml), those

in the 70 ml dish at medium density (0.0041g/ml), and the termites in the 190 ml

dish were considered to be at low density (0.0015g/ml). Each dish contained

0.1 g of aspen wood and a 4.25 an filter paper as €4. Prior to introduchg the

termites, the dishes were fiUed with sterilized brick sand that had been moistened to

80% saturation. After the termites had been introduced, the petri dishes were placed

in an environmental chaniber (86%RH, 26OC) for a period of seven days to allow the

termites to excavate a network of galleries. FolIowing thk pend, a single termite

coated in the conidia of M. anisopliae was added to each of the replicates

(20 replicates for each treabent). The petri dishes were retumed to the chamber for

an additional 14 days, after which they were dismantled and the number of

surviving termites recorded.

Group effect on fungil growth

Five worker temites were placed in a 30 ml glas petri dish (ten repliates)

containing a moistened 4.25 cm filter paper. To each population a single live

termite dusted with the conidia of M. anisopliae was added to determine the effects

the group would have on the resulting termite cadaver. For cornparison, a single

dusted termite was placed alone in a petri dish (ten replicates) containùis a

moistened 4.25 cm filter paper. The petri dishes were rnaintained in an

environmental chamber (%%RH; 26°C) for a period of 10 days.

Results

Density experiment

There was no significant difference between the levels of mortality recorded

in the hi& medium and low density treatments (F=1.88; &=2; p=0.162). Mortality

ranged between 30 and 40% in each of the treaiments following the introduction of a

single termite carrying M. anisopliae (Table 4.1). There was a siuficant

dioerence between the total number of soldier termites that survived in each of the

treatments (F-9.96; df=2; p=û.0002; Table 4.1). In the high density treatment, only

four soldiers out of a total of 40 survived, representing a soldier mortality level of

90%. At medium density, 19 soldiers çurvived (47.5% mortality) and at low density,

23 soldiers surviveci (57.5% mortality). Dunng periodic inspections of the petri

dishes, it was obsemed that large numbers of termites were found aggregating

around the food sources in each of the treatments.

Group effect on fun@ growth

At the end of the experiment (day 10) all termites in the control replicates

were dead and fun@ growth was abundant on six of the ten cadavers. The

remaining four cadavers were deliquescent (Table 4.2). In the treatment group, of

the termites had died and of these, ten had abundant fungal growth, eight were

covered with fecal material with no evidence of h g i , and 15 were deliquescent

(Table 4.2).

Discussion

Density is considered an important factor in the development of epizootics in

insect populations (Keller and Zimmerman 1989). Thmefore, it was surprising that

there was no significant difference in mortality between the three density regimes.

The largest petri dish had a capaaty that was 6x greater than that of the srnailest

dish, so the relative volume available to each termite varied substantiaUy between

treatments. However, a seemingly large nurnber of termites were regularly

observeci aggregathg around the food sources in al of the treatments, implying that

relatively few termites were actually moving throughout the galleries at any given

73

tirne. The dustering of workers at the food sources may reflect a tendency for

termites to niaintain an optimal density in their colonies. Consequently, the density

of termites between each of the treatments rnay not have varied. The comparable

levels of rnortaüty between the density regimes may reflect the number of termites

in the galleries, which rnay have been the same.

However, it is still difficult to account for the higher Ievel of soldier mortality

under the high density regime. The extent of the gdery network excavated in the

soil of the dishes may have had an effed on the speed at which termites were able to

commUNcate alam signals and reauit nestmates. The termites in the smaiiest

dishes (highest density) rnay have had the least extensive gallery network and might

have been able to recnrit soldiers more rapidy than the termites in the larger dishes.

The rapid reauitment of soldiers rnay have increased the IikeLthood that they

would have confronted the termites carrying the conidia and become infected.

The completion of the fungus life cyde within infecteci termites appears to be

affected by the presence of nestmates. Several of the termites that were dead from

mycosis and maintained in isolation supported abundant fungal growth or were

deliquescent, whereas this was not the case in the group condition. It is interesthg

to note that of the dead termites in the treatment group, 25% of them were covered

with fecal material and showed no signs of fungal growth or deliquescence. This

suggests that termites feces rnay have anti-microbial properties.

Conclusion

The tendency for termites to aggregrate around sources of food rnay have

resulted in simüar tennite densities within a l l treatments, and hence, termite

mortality did not Vary between the high, medium and low density regimes. This

finding raises the possibility that termites maintain an optimal density within their

74

colonies even when they have the opportunity to forage in a larger area.

Optimal densities may allow termites to collunmicate rapidly and respond quickly

to alam stimulus, such as nestmates that are infeded with pathogenic organisms or

the presence of predatory insects. When nestmates do become uifected and die, the

burial of the cadaver under fecal material rnay intempt the development of the

pathogen within the host and prevent it hom becoming a site of secondary

infection.

Table 4.1 Mean mortality among groups of Re t iculitermesfZaoipes under different

density regimes for 14 days following the introduction of a single termites dusted

with Metarhizium anisopliae.

Mean mortality (% + SD)

Density regime Workers Soldiers

Medium

Low

* Meam foliowed by the same letter are not significantiy different based on Student-

Newman-Keuls cornparisons at an a value of 0.05 (ANOVA of mortality data; n=20

dishes; 98 workers and two soldiers per dish; SPSS 1995)

Table42 The percentage of termite cadavers in petri dishes (n=10 dishes) that are

either 'inuied mder termite feces, supporthg abundant fungal growth or showing

signs of deliquescence after being infecteci with the fungal pathogen M e f n r h iz i u m

an isopl iae.

Single termite Cadavers within group cadavers of termites

Condition of cadavers (%) (%)

Covered with fungi 60

Covered with feces O

Deliquescen t 40

Becker, G. 1969. Rearing of termites and testing rnethods used in the laboratory. pp. 351-385. In K. Krishna and F.M. Weesner (eds) Biology of termites, vol. 1. Academic Press, New York.

Forschler, B.T. and M.L. Townsend. 1996. Mark-release-recapture estimates of Re t ic u lit er mes spp. (Isoptera: Rhinotermitidae) colony foraging populations fmm Georgia, USA. Environ. Entomol. 25: 952-962

Grace, J.K. 1990. bfark-recaphire çtudies with Reticuli ter mesflaoipes (Isoptera: Rhinotermitidae). Çociobiology 16: 297-303.

Holdobbler, B. and E. O. Wilson. 1990. The ants. Belknap Press, Cambridge. PP- 6Ci4-

Keller, S. and G. Zimmerman. 1989. Mycopathogens of soi1 insectç. I n N. Wilding, N.M. Collins, P.M. Hammond, and J.F. Webber (eds) Insect-fungus interactions. Academic Press, London. pp 240-265.

Kramm, K.R., West, D.F., and P.G. Rockenbach. 1982. Termite pathogens: Transfer of the entomopathogen Metarhiz ium anisopl iae between Reticulitermes sp. termites. J . Invert. Pathol. 40: 1-6.

Lenz, M. 1985. Variability of vigour between colonies of Copto termes acinaciformis (Froggatt) (Isoptera: Rhinotennitidae) and its implications for laboratory experimentation. Bull. Entomol. Res. 75: 13-21.

Lenz, M., Barrett, RA. and ER. Williams. 19W Implications for cornparability of laboratory experiments revealed in shidies on the effects of population dençity on vigour in Coptotermes lacteus (Froggatt) and Nasutitermes exit ios us (HiU) (Isoptera: Rhinotermitidae and Termitidae) . Bull. Entomol. Res. 74: 477-485.

Miramontes, 0. 0. DeSo~za. 1996. The non-linear dynamics of sumival and social facilitation in termites. J. Theor. Biol. 181: 373-380.

Myles, T.G. 1988. Sockd evolution from termites to man. pp 379-413. I n C.N. Slobodchikoff (4). The ecology of social behaviour. Academic Press, Boston.

NoUot, C.H. 1970. The nests of termites. pp. 73-120. In K. Krishna and F.M. Weesner (eds) Biology of termites, vol. 2. Academic Press, New York.

Nutting, W.L. 1969. nght and colony foundation. pp. 233-276. In K. Krishna a . F M . Weesner (eds) Biology of termites, vol. 1. Academic Press, New York.

Pearce, M.J. 1987. Se&, tombs, mummies and tunnebg in the drywood termite Cry p to ter rn es (Isoptera: Kalotermitidae). Sociobiology 13: 217-226.

Rosengaus, R.B. and J.F.A. TranieUo. 1993. Disease risk as a cost of outbreeding in the termite Zoo ter mopsis angusticollis. Proc. Natl. Acad. Sa 90: 6641-6645

Rosengaus, R.B. and J.F.A. TranieIlo. 1997. Pathobiology and disease transmission in dampwood termites, Zoo terrnopsis angus ticollis (Isoptera: Rhinotermitidae), infected with the h g u s hdetarhizium anisopliae (Deuteromycoüna: Hyphomycetes). Sociobiology 30: 185-195.

SPSS Inc. 1995. SPS User's guide. SPÇS Inc. Chicago, m.

Shellman-Reeve, J.S. 1997. The spectrum of eUSOaality in termites. pp. 52-93. In J.C. Choe and B.J. Crespi (eds). The evolution of social behaviour in insects and arachnids. Cambridge University Press, Cambridge.

Stuart, A.M. 1969. Social behaviour and communication. pp 193-229. In K Krishna F, M Weesner (eds) Biology of termites, vol. 2. Academic Press, New York

Su, N., Ban, PM. and R.H. Sdieffrahn. 1993. Foraging populations and temtories of the eastem subterranean termite (Isoptera: Rhino tenni tidae) in southeastem Florida. Environ. Entomol. 22: 1213-1117.

Wilson, E.O. 1971. The insect societies. B e h a p Press, Cambridge. pp 548

Zoberi, M. 1995. Meta rhiz ium anisopl iae, a fungal pathogen of Reticulitermes jiav ipes (Isoptera: Rhinotermitidae). Mycologia 87: 354-359.

Thesis summary and conclusions

The h g a l pathogen, Meta rh i z ium a nisopl iae (Metschnikoff) is considered a

candidate for the biocontrol of the eastem subterranean termite, Re t i c u li termes

flaoipes (Kollar). It is conceivable that termite social behaviour may be effective aC

tramferring pathogenic conidia between nestmates, resulting in high levels of

disease. However, it is &O known that termites dusted with the conidia of M.

a n iso p l iae elict a l m behaviour in their nestmates, implying that the fungus rnay

be a source of alarm stimulus. The focus of this thesis was to determine whether

termite soaal behaviour may prevent the dissemination of pathogenic conidia in

sub terranean termite colonies,

ï h e results reported in chapter two dearly indicate that, in petri dishes, M.

anisopliae is capable of causing high levels of mortality in populations of

R.flaaipes. The high levelç of mort&& however, were Likely the result of

agonistic interactions which brought unexposed termites into contact with those

that were dusted with conidia. The termites duçted with M. anisopliae were

frequently attacked, killed or burieci by their nestmates.

That the results of the petri dish experiments could not be replicated in cups

fïlled with soil suggested that termite behaviour, in a more natural soil

environment, may limit the dissemination of conidia. This hypothesis was

confinneci by the outcome of the experiment conducted within the observational

chambers. It was found that termites dusted with the c o d a of M. a n isopl iae

triggered a cascade of agonistic behaviour in their nestmates. This behavioural

responçe, which induded the display of alarm behavim, the retniihnent of

workers and soldiers, and the grooming or burial of the dusted termite, prevented

the termites rvith conidia from entering the network of galleries. This response was

not elicited by termites thaï were dusted with non-pathogenic conidia, suggesting

that the response was s p d ï c to M. anisopl iae.

80

81

The results of chapter three confirmed that agonistic behaviour in termites

is in response to the conidia of M. a nisopliae. Tefanites dusted with either viable or

dead conidia eliQted agonistic behaviour in their nestmates. Agonistic behaviour

was not elicited by termites that were moribund or dead from mycosis, indicating

that it was the conidia that eliated agonism, and not the symptoms assoaated with

h g a l infection. The possibility that the presence of conidia on termite cutide may

interfere with termite recognition factors, thus triggering the agonistic response,

appearç unlikely, since the response to termites dusted with non-pathogenic conidia

and talcum powder was minimal.

Termite density does not seem to have an affect on the epizootiology of

fungai diseases in termite colonies. The experiments in chapter four found no

difference in the levels of mortaüty recorded from termites populations maintained

at three dif ferent density regimes, following the introduction of termites dusted

with the conidia of M. an isopl iae. However, the results of the experiment may

have been affected by the tendency termites have to duster. Though petri dishes of

Merent volumes were provided, the termites within them appeared to maintain

similar densities for each of the density regimes. Termite soldier mortality was

affecteci by density, or at least by the different volumes of the petri dishes. The

number of soldiers that died under the high density regime (the smallest dish) was

significantly higher then the number that died under the medium and low density

conditions. It is possible that following the introduction of the dusted termites, the

colonies in the smdest dishes were able to reaiiit soldiers more quiddy. This may

have incseased the chances of the soldiers coming into contact with the dusted

termites, which would have resulted in their becoming infected and dying.

Chapter four also revealed that termites defecate on nestmates that have

died from fungal infection. Furthermore, termite cadavers covered with feces did

not support the growth of fungi, nor did they become deliquescent (from

bacterial or yeast invasion). These results imply that termite feces may have anti-

miaobial properties.

Effective termite biocontrol with M. a nisopl iae requires that termites dusted

with conidia be able to mix freely with the colony in order to distribute conidia

among nestmates. The research conducted for this thesis found that in a soil

environment, termites were able to recognize nesbates that were carryhg the

conidia of M. anisopliae and prevent them from entering into the colony. It can be

conduded that the social behaviour of termites, with respect to fungal disease, may

be a formidable obstacle to termite biocontrol with M. anisopliae. Before the

remedial use of fun@ for termite control c m be undertaken, additional research

will need to be conducted to h d other strains of 113. a nisopliae, or other h g a l

pathogens, that do not elicit similar agonistic behaviour in termites.

IMAGE EVALUATION TEST TARGET (QA-3)

APPLIED IMAGE, lnc 1653 East Main Street - -. - Rochester. NY 14609 USA -- --= Phone: 71 6M82-0300 -- -- - - Féuc 7161288-5989