<%BANNER%>

Defense Mechanisms in Termites Against the Infection of Entomopathogenic Fungi

Permanent Link: http://ufdc.ufl.edu/UFE0024715/00001

Material Information

Title: Defense Mechanisms in Termites Against the Infection of Entomopathogenic Fungi
Physical Description: 1 online resource (187 p.)
Language: english
Creator: Chouvenc, Thomas
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

Subjects / Keywords: antifungal, biological, cellular, control, defense, epizootics, metarhizium, pathogen, reticulitermes, termite
Entomology and Nematology -- Dissertations, Academic -- UF
Genre: Entomology and Nematology thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: The use of fungal pathogens for biological control of subterranean termites has attracted particular attention in the past 40 years and several laboratory studies have shown promising results. This approach was based on the concept of classical biological control with the use of a virulent agent that can self-replicate in a termite nest and be transmitted among individuals due to their high social interaction, resulting in an epizootic to kill the entire colony. However, the absence of positive results in field applications raises questions about the potential of fungal pathogens as a realistic approach for subterranean termite control. The study of defense mechanisms in Isoptera against the infection of pathogens started to receive particular attention in the past decade, but the relationship between fungi and subterranean termites remains poorly understood. This study focuses on the subterranean termite Reticulitermes flavipes (Kollar): Rhinotermitidae and its relationship with the entomopathogenic fungus Metarhizium anisopliae (Metsch.) Sorokin at the colony level, the individual level, and the cellular level. This multimodal approach allowed us to show that subterranean termites have the ability to prevent an epizootic from occurring when kept under controlled soil conditions with normal foraging distances. The defense mechanisms involved following high survivorship were identified and documented. From histological observations, the cellular encapsulation against the infection of M. anisopliae was described and quantified in R. flavipes, and the antifungal activity of the alimentary tract was also demonstrated. Such defense mechanisms were also documented in five other termite species (Hodotermopsis sjoestedti (Holmgren): Termopsidae, Hodotermes mossambicus (Hagen): Hodotermitidae, Kalotermes flavicollis Fabr.: Kalotermitidae, Prorhinotermes canalifrons (Sjo umlautstedt): Rhinotermitidae, and Nasutitermes voeltzkowi (Wasmann): Termitidae), and we confirmed that the cellular encapsulation and the antifungal activity of the gut is conserved in all species studied. Finally, the interactions among defense mechanisms in termites were studied and it is suggested that these mechanisms act synergistically to produce an efficient defense against the infection of the fungus at the individual and group level so as to protect the colony from epizootics.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Thomas Chouvenc.
Thesis: Thesis (Ph.D.)--University of Florida, 2009.
Local: Adviser: Su, Nan-Yao.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2010-02-28

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0024715:00001

Permanent Link: http://ufdc.ufl.edu/UFE0024715/00001

Material Information

Title: Defense Mechanisms in Termites Against the Infection of Entomopathogenic Fungi
Physical Description: 1 online resource (187 p.)
Language: english
Creator: Chouvenc, Thomas
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

Subjects / Keywords: antifungal, biological, cellular, control, defense, epizootics, metarhizium, pathogen, reticulitermes, termite
Entomology and Nematology -- Dissertations, Academic -- UF
Genre: Entomology and Nematology thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: The use of fungal pathogens for biological control of subterranean termites has attracted particular attention in the past 40 years and several laboratory studies have shown promising results. This approach was based on the concept of classical biological control with the use of a virulent agent that can self-replicate in a termite nest and be transmitted among individuals due to their high social interaction, resulting in an epizootic to kill the entire colony. However, the absence of positive results in field applications raises questions about the potential of fungal pathogens as a realistic approach for subterranean termite control. The study of defense mechanisms in Isoptera against the infection of pathogens started to receive particular attention in the past decade, but the relationship between fungi and subterranean termites remains poorly understood. This study focuses on the subterranean termite Reticulitermes flavipes (Kollar): Rhinotermitidae and its relationship with the entomopathogenic fungus Metarhizium anisopliae (Metsch.) Sorokin at the colony level, the individual level, and the cellular level. This multimodal approach allowed us to show that subterranean termites have the ability to prevent an epizootic from occurring when kept under controlled soil conditions with normal foraging distances. The defense mechanisms involved following high survivorship were identified and documented. From histological observations, the cellular encapsulation against the infection of M. anisopliae was described and quantified in R. flavipes, and the antifungal activity of the alimentary tract was also demonstrated. Such defense mechanisms were also documented in five other termite species (Hodotermopsis sjoestedti (Holmgren): Termopsidae, Hodotermes mossambicus (Hagen): Hodotermitidae, Kalotermes flavicollis Fabr.: Kalotermitidae, Prorhinotermes canalifrons (Sjo umlautstedt): Rhinotermitidae, and Nasutitermes voeltzkowi (Wasmann): Termitidae), and we confirmed that the cellular encapsulation and the antifungal activity of the gut is conserved in all species studied. Finally, the interactions among defense mechanisms in termites were studied and it is suggested that these mechanisms act synergistically to produce an efficient defense against the infection of the fungus at the individual and group level so as to protect the colony from epizootics.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Thomas Chouvenc.
Thesis: Thesis (Ph.D.)--University of Florida, 2009.
Local: Adviser: Su, Nan-Yao.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2010-02-28

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0024715:00001


This item has the following downloads:


Full Text

PAGE 1

DEFENSE MECHANISMS IN TERMITES AGAINST THE INFECTION OF ENTOMOPATHOGENIC FUNGI By THOMAS CHOUVENC A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2009 1

PAGE 2

2009 Thomas Chouvenc 2

PAGE 3

Now, I know why we failed Minoru Tamashiro, December 11 th 2007, San Diego, CA. 3

PAGE 4

ACKNOWLEDGMENTS First of all, I want to express all my gratitude to Nan-Yao Su for accepting me in his laboratory and allowing me to do my research with such freedom of intellectual thinking. His support, either educational, motivational, (financial!?), critical, intellectual, and/or philosophical, were a tremendous source of inspiration for me and I am grateful to him for allowing me to conduct my research in such a warm, friendly and motivating environment. Finally, his good taste for Burgundy wine proved to me that he is a good guy. Much of the production of this dissertation would not have been possible without the indispensable scientific contribution from the members of my PhD committee. Monica Elliott was a precious source of advice and made the approach of microbiology possible for an ignorant entomologist like myself. Monica helped me to avoid many mistakes and to work with microbes confidently. Robin Giblin-Davis was another important source of advice and motivation. His cheerfulness and his scientific contribution helped to keep me motivated and to force me to go deeper into my scientific thought processes. Rudolf Scheffrhan contributed a lot to my understanding of the termite world and his critical analysis of my work was a great source of inspiration for improvement. Finally, William Kern was critical, for without him, I would not be able to call myself an entomologist. I would like to thank them all deeply and humbly, as they all fulfilled their role with brilliance. I give my gratitude for Rebecca Rosengaus who has been a critical encounter for my research. Her advice helped me start my entire project and her work was a major source of inspiration for the direction that my research took. Her agreement to provide me with a fungal strain to work with, her kindness in answering all my questions and her continuous support were all a remarkable help to me. 4

PAGE 5

I would like to thank my co-workers and fellow students, who provided me with a friendly and motivating environment to work in. Thanks especially to Hou-Feng Li, who tolerated me as his office mate for the last five years. He was always friendly to me and the long discussions with him were always a matter of interest. His seriousness for research and his dedication to study were a critical factor that motivated me to keep going. How can I sit down and waste my time on irrelevant things when the guy sitting next to me was getting gray hair from his research project? Many thanks for Ronald magic arena Peppin and Paul Ban for all their technical and motivational participation in my project. I also thank Tini Bujang, Teresa Ferreira, Paul Bardunias, Yutaka Kobayashi and Rou-Ling Yang for their support and friendship. Thanks to Mr Brian bugboy Bahder for his non-scientific contribution and for the countless hours of recreational activity outside of what a standard Ph.D. schedule should usually allow. I would also like to thank many other people from the FLREC, who contributed to making my life easier and most enjoyable as an international student, with special attention to Barbara Center, Beth DesJardin, Sarah Kern, Nigel Harrison, Jan Krecek and Bill Latham. This adventure would never have been possible without the help of many people in France. I particularly thank Christian Bordereau for his original support that allowed me meet to Dr. Su, and for the continuous support he has provided during the past five years, even though he had no obligation to do so. I deeply acknowledge Alain Robert, my original mentor, who had a key role in the outcome of my project. He has my gratitude for delaying his summer vacation to take time to help me with my experiments in Dijon in summer 2006, and to teach me histological techniques which ended up being the most important aspect of my research. Many students from Dijons laboratory were an source of support prior to my journey to the West and they were very helpful in my preparation for leaving the old continent. Among them, I especially thank 5

PAGE 6

Micheline Grillet, Ben Huot, Edgar Gervasio, Jeremy Skrzypski, David Sillam-Dusss and Mathieu Bourillot. I deeply thank my mother, father and brothers for their unconditional love, patience and support; as I was so far away from them the past five years. Also, many thanks to the rest of my family who welcomed me back so warmly during the Christmas times in Chagnon. They helped me go through the tough times and helped me keep contact with my roots and with reality. I also would give special attention to David Clarke whose support helped me to prepare my journey. Finally, I want to express all my gratitude to Ericka Helmick, who has been the most supportive, the most patient, the most present, the most caring, and the most willing to help carry the burden of this Ph.D. with me in the past five years. This study was supported in part by a grant from USDA-ARS under grant agreement No. 58-6435-8-276. 6

PAGE 7

TABLE OF CONTENTS page ACKNOWLEDGMENTS ...............................................................................................................4 LIST OF TABLES .........................................................................................................................11 LIST OF FIGURES .......................................................................................................................12 ABSTRACT ...................................................................................................................................15 CHAPTER 1 GENERAL INTRODUCTION..............................................................................................17 Termite Biology ......................................................................................................................17 Biological Control in Termites ...............................................................................................18 Epizootiology in Subterranean Termites ................................................................................20 Pathogen Virulence .........................................................................................................21 Pathogen Density .............................................................................................................21 Pathogen Capacity to Survive outside the Host ..............................................................22 Pathogen Capacity to Replicate .......................................................................................22 Pathogen Capacity to Disperse ........................................................................................22 Host Susceptibility ...........................................................................................................23 Host Density ....................................................................................................................23 Host Behavior ..................................................................................................................23 Favorable Environmental Conditions ..............................................................................24 Defense Mechanisms in Termites ...........................................................................................24 The Entomopathogenic Fungus Metarhizium anisopliae .......................................................25 Objectives ...............................................................................................................................26 2 INTERACTION BETWEEN RETICULITERMES FLAVIPES AND METARHIZIUM ANISOPLIAE IN FORAGING ARENAS.............................................................................39 Introduction .............................................................................................................................39 Material and Methods .............................................................................................................41 Collection of Termite Colonies .......................................................................................41 Metarhizium anisopliae Conidia Preparation ..................................................................41 Metarhizium anisopliae Virulence on Reticulitermes flavipes ........................................42 Foraging Arenas ..............................................................................................................42 Survival of Metarhizium anisopliae in Sand ...................................................................44 Transfer and Dispersal of Metarhizium anisopliae Conidia from Termite Cuticle to Surrounding Sand .........................................................................................................45 Survival of Metarhizium anisopliae in Arenas Containing Termites ..............................45 Results .....................................................................................................................................46 Metarhizium anisopliae Virulence on Reticulitermes flavipes ........................................46 Termite Survival in Foraging Arenas ..............................................................................46 7

PAGE 8

Survival of Metarhizium anisopliae in Sand ...................................................................47 Transfer and Dispersal of Metarhizium anisopliae Conidia from Termite Cuticle to Surrounding Sand .........................................................................................................47 Survival of M. anisopliae in Arenas Containing Termites ..............................................48 Discussion ...............................................................................................................................48 3 SUSCEPTIBILITY OF SEVEN TERMITE SPECIES TO METARHIZIUM ANISOPLIAE.........................................................................................................................59 Introduction .............................................................................................................................59 Material and Methods .............................................................................................................60 Termite Species ...............................................................................................................60 Conidial Suspensions and Susceptibility Test .................................................................61 Statistical Analysis ..........................................................................................................61 Results .....................................................................................................................................62 Naive Termites and Control Termites Survival ...............................................................62 Effect of Termite Weight on Survival .............................................................................62 Median Lethal Dosage for each termite species ..............................................................63 Effect of Conidial Concentrations for Each Species .......................................................63 Mastotermes darwiniensis ........................................................................................63 Hodotermopsis sjoestedti .........................................................................................64 Hodotermes mossambicus ........................................................................................64 Kalotermes flavicollis ...............................................................................................64 Prorhinotermes canalifrons .....................................................................................65 Reticulitermes flavipes .............................................................................................65 Nasutitermes voeltzkowi ...........................................................................................66 Comparison of Survival of among Seven Termite Species Exposed to Metarhizium ....66 Discussion ...............................................................................................................................68 4 INHIBITION OF METARHIZIUM ANISOPLIAE IN THE ALIMENTARY TRACT OF RETICULITERMES FLAVIPES.....................................................................................85 Introduction .............................................................................................................................85 Material and Methods .............................................................................................................86 Termite Preparation and Fungal Inoculation ...................................................................86 Histological Preparation ..................................................................................................87 Data Collection and Analysis ..........................................................................................87 Results .....................................................................................................................................88 Control Termites ..............................................................................................................88 Visibly Healthy Individuals after Exposure to Metarhizium anisopliae .........................88 Moribund and Dead Individuals after Exposure to Metarhizium anisopliae ..................89 Metarhizium anisopliae Exposed Termites; Dead Individuals Fixed after Decomposition .............................................................................................................90 Discussion ...............................................................................................................................91 5 ANTIFUNGAL ACTIVITY OF THE GUT IN SIX TERMITE SPECIES.........................101 8

PAGE 9

Introduction ...........................................................................................................................101 Material and Methods ...........................................................................................................102 Results ...................................................................................................................................102 General Observations ....................................................................................................102 Observation of the digestive tract in Five Termite Species ...........................................103 Hodotermopsis sjoestedti .......................................................................................103 Hodotermes mossambicus ......................................................................................103 Kalotermes flavicollis .............................................................................................104 Prorhinotermes canalifrons ...................................................................................104 Nasutitermes voeltzkowi .........................................................................................105 Discussion .............................................................................................................................105 6 ANTIFUNGAL ACTIVITY OF NORHARMANE.............................................................113 Introduction ...........................................................................................................................113 Material and Methods ...........................................................................................................114 Results and Discussion .........................................................................................................115 7 CELLULAR ENCAPSULATION IN RETICULITERMES FLAVIPES...........................118 Introduction ...........................................................................................................................118 Material and Methods ...........................................................................................................119 Termite Inoculation and Histological Preparation ........................................................119 Free Circulating Hemocytes Count ...............................................................................120 Results ...................................................................................................................................120 FCH Count .....................................................................................................................120 Histopathology of Negative and Positive Controls .......................................................121 Cellular Encapsulation in Reticulitermes flavipes .........................................................121 Discussion .............................................................................................................................123 8 CELLULAR ENCAPSULATION IN SIX TERMITE SPECIES........................................134 Introduction ...........................................................................................................................134 Material and Methods ...........................................................................................................135 Results and Discussion .........................................................................................................136 Nodule Formation in Six Termite Species ....................................................................136 Reticulitermes flavipes ...........................................................................................137 Hodotermopsis sjoestedti .......................................................................................137 Hodotermes mossambicus ......................................................................................138 Kalotermes flavicollis .............................................................................................139 Prorhinotermes canalifrons ...................................................................................139 Nasutitermes voeltzkowi .........................................................................................139 Relationships between the Cellular Encapsulation and Other Known Factors .............140 The relative physiological cost ...............................................................................140 Survival trade-off ...................................................................................................141 Conclusion .....................................................................................................................141 9

PAGE 10

9 SYNERGY AMONG DEFENSE MECHANISMS IN SUBTERRANEAN TERMITES AGAINST FUNGAL PATHOGENS...................................................................................153 Introduction ...........................................................................................................................153 Defense Mechanisms in Termites .........................................................................................154 Behavioral Avoidance ...................................................................................................154 Chemicals in Soil Habitat ..............................................................................................156 Grooming Behavior .......................................................................................................157 Gut chemical activity .....................................................................................................158 Cellular and Humoral immunity ....................................................................................158 Necrophagy and Corpse Avoidance ..............................................................................159 Synergy among Defense Mechanisms ..................................................................................160 Role of the Grooming Behavior ....................................................................................160 Role of the Cellular Encapsulation ................................................................................161 Role of the Antifungal Gut Activity ..............................................................................162 Efficacy of the Synergy .................................................................................................162 Conclusion: The Future of Biological Control in Subterranean Termites ............................163 LIST OF REFERENCES .............................................................................................................166 BIOGRAPHICAL SKETCH .......................................................................................................187 10

PAGE 11

LIST OF TABLES Table page 1-1 List of termite species that were tested against various entomopathogens including fungi, bacteria, nematodes and viruses, but not including parasitic organisms. ................28 2-1 Recovery of Metarhizium anisopliae from control sand, sand in contact with infested termites for 60 min, sand in control arenas, and sand in infested arenas ...........................52 3-1 Ecological characteristics of the termite species used for susceptibity test against Metarhizium anisopliae .....................................................................................................73 3-2 Median lethal dose for seven termite species exposed to Metarhizium anisopliae conidia ..............................................................................................................................73 4-1 Presence of conidia on the surface of the cuticle and in the alimentary tract of Metarhizium anisopliae exposed termites at different times of fixation. ..........................97 6-1 Mycelial diameter growth rate (Mean SE) of Metarhizium anisopliae and Aspergillus nomius at different concentrations of norharmane (g ml) when grown in the dark at 27C.. -1 ..........................................................................................................117 7-1 Average number of free circulating hemocytes (MeanSE) found in the open circulatory system of one median histological section (sagittal, 5m, n=30) in control termites and termites treated with Metarhizium anisopliae. ............................................130 8-1 Size of cellular encapsulation in six termite species. .......................................................143 11

PAGE 12

LIST OF FIGURES Figure page 1-1 Curve showing an epizootic wave .....................................................................................36 1-2 Potential defense mechanisms involved in a termites nest against an epizootic event. ...37 1-3 Life cycle of the entomopathogenic fungus Metarhizium anisopliae in an insect host .....38 2-1 Foraging arena description .................................................................................................53 2-2 Survival distribution of groups of 20 Reticulitermes flavipes individuals after exposure of different treatments of Metarhizium anisopliae .............................................54 2-3 Grooming in Reticulitermes flavipes .................................................................................55 2-4 Canibalism in Reticulitermes flavipes ................................................................................56 2-5 Survival inside the two sets of arenas of the treated termites in contact with 900 naive termites .....................................................................................................................57 2-6 Survival inside the two sets of arenas of the naive termites, in contact with 60 treated termites. ..............................................................................................................................58 3-1 Simplified phylogeny of Isoptera .......................................................................................74 3-2 The seven termite species tested for susceptibility against infection of Metarhizium anisopliae ...........................................................................................................................75 3-3 Relationship between the average termite weight of individual termites and their susceptibility to Metarhizium anisopliae exposure ............................................................76 3-4 Survival distribution of groups of 20 Mastotermes darwiniensis after exposure to different concentration of Metarhizium anisopliae conidia ...............................................77 3-5 Survival distribution of groups of 20 Hodotermopsis sjoestedti after exposure to different concentration of Metarhizium anisopliae conidia ...............................................78 3-6 Survival distribution of groups of 20 Hodotermes mossambicus after exposure to different concentration of Metarhizium anisopliae conidia. ..............................................79 3-7 Survival distribution of groups of 20 Kalotermes flavicollis after exposure to different concentration of Metarhizium anisopliae conidia ...............................................80 3-8. Survival distribution of groups of 20 Prorhinotermes canalifrons after exposure to different concentration of Metarhizium anisopliae conidia ...............................................81 12

PAGE 13

3-9 Survival distribution of groups of 20 Reticulitermes flavipes after exposure to different concentration of Metarhizium anisopliae conidia ...............................................82 3-10 Survival distribution of groups of 20 Nasutitermes voeltzkowi after exposure to different concentration of Metarhizium anisopliae conidia ...............................................83 3-11 Survival distribution of groups of 20 individuals from seven termite species after exposure to Metarhizium anisopliae ..................................................................................84 4-1 Digestive track of Reticulitermes flavipes in situ, dorsal view ..........................................98 4-2 Occurrence of Metarhizium anisopliae conidia in Reticulitermes flavipes, 1 d after inoculation.. ........................................................................................................................99 4-3 Occurrence of Metarhizium anisopliae hyphae in Reticulitermes flavipes .....................100 5-1 Occurrence of Metarhizium anisopliae in Hodotermopsis sjoestedti ..............................108 5-2 Occurrence of Metarhizium anisopliae in Hodotermes mossambicus .............................109 5-3 Occurrence of Metarhizium anisopliae in Kalotermes flavicollis ...................................110 5-4 Occurrence of Metarhizium anisopliae in Prorhinotermes canalifrons ..........................111 5-5 Occurrence of Metarhizium anisopliae in Nasutitermes voeltzkowi ................................112 7-1 Cellular interaction between Metarhizium anisopliae and Reticulitermes flavipes .........131 7-2 Cellular interaction between Metarhizium anisopliae and Reticulitermes flavipes. ........132 7-3 Schematized cellular encapsulation process of Metarhizium anisopliae infection in Reticulitermes flavipes .....................................................................................................133 8-1 Visible nodule formation in two termite species .............................................................144 8-2 Cellular interaction between Metarhizium anisopliae and Reticulitermes flavipes .........145 8-3 Cellular interaction between Metarhizium anisopliae and Hodotermopsis sjoestedi ......146 8-4 Cellular interaction between Metarhizium anisopliae and Hodotermes mossambicus ...147 8-5 Cellular interaction between Metarhizium anisopliae and Kalotermes flavicollis ..........148 8-6 Cellular interaction between Metarhizium anisopliae and Prorhinotermes canalifrons .149 8-7 Cellular interaction between Metarhizium anisopliae and Nasutitermes voeltzkowi ......150 8-8 Relationship between the size of encapsulation of Metarhizium anisopliae infection and the wet weight of termites .........................................................................................151 13

PAGE 14

8-9 Relationship between the relative physiological cost the cellular encapsulation of Metarhizium anisopliae infection and the susceptibility of termites ...............................152 9-1 Areas where cellular encapsulations of Metarhizium anisopliae infection occurred in Reticulitermes flavipes .....................................................................................................165 14

PAGE 15

Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy DEFENSE MECHANISMS IN TERMITES AGAINST THE INFECTION OF ENTOMOPATHOGENIC FUNGI By Thomas Chouvenc August 2009 Chair: Nan-Yao Su Major: Entomology and Nematology The use of fungal pathogens for biological control of subterranean termites has attracted particular attention in the past 40 years and several laboratory studies have shown promising results. This approach was based on the concept of classical biological control with the use of a virulent agent that can self-replicate in a termite nest and be transmitted among individuals due to their high social interaction, resulting in an epizootic to kill the entire colony. However, the absence of positive results in field applications raises questions about the potential of fungal pathogens as a realistic approach for subterranean termite control. The study of defense mechanisms in Isoptera against the infection of pathogens started to receive particular attention in the past decade, but the relationship between fungi and subterranean termites remains poorly understood. This study focuses on the subterranean termite Reticulitermes flavipes (Kollar): Rhinotermitidae and its relationship with the entomopathogenic fungus Metarhizium anisopliae (Metsch.) Sorokin at the colony level, the individual level, and the cellular level. This multimodal approach allowed us to show that subterranean termites have the ability to prevent an epizootic from occurring when kept under controlled soil conditions with normal foraging distances. The defense mechanisms involved following high survivorship were identified and documented. From histological observations, the cellular encapsulation against the infection of 15

PAGE 16

M. anisopliae was described and quantified in R. flavipes, and the antifungal activity of the alimentary tract was also demonstrated. Such defense mechanisms were also documented in five other termite species (Hodotermopsis sjoestedti (Holmgren): Termopsidae, Hodotermes mossambicus (Hagen): Hodotermitidae, Kalotermes flavicollis Fabr.: Kalotermitidae, Prorhinotermes canalifrons (Sjstedt): Rhinotermitidae, and Nasutitermes voeltzkowi (Wasmann): Termitidae), and we confirmed that the cellular encapsulation and the antifungal activity of the gut is conserved in all species studied. Finally, the interactions among defense mechanisms in termites were studied and it is suggested that these mechanisms act synergistically to produce an efficient defense against the infection of the fungus at the individual and group level so as to protect the colony from epizootics. 16

PAGE 17

CHAPTER 1 GENERAL INTRODUCTION Termite Biology Termites (Insecta, Isoptera) are a large and important group of social insects that are widely distributed in tropical, subtropical, and temperate regions of the world (Eggleton, 2000). Termites consist of more than 2,600 described species, from seven families (Kambhampati and Eggleton, 2000; Engel and Krishna, 2004) and are a sister taxa to wood roaches (Inward et al., 2007a; Klass et al., 2008). Although all Isoptera are eusocial and share common traits, termites have spread and adapted to a large range of habitats resulting in high diversity of cast morphology, physiology, behavior, nesting ecology, and are associated with a diverse microbial community (Abe et al., 2000). They all feed on cellulose-based material, (wood, leaves, litter, grass, etc.) although most termites cannot efficiently digest cellulose with their endogenous cellulases (Slaytor, 2000). They are all associated with symbiotic organisms that have the ability to break down cellulose and other organic material into sugars and nutrients that cannot be assimilated by the termites, and may also participate in nitrogen fixation (Ohkuma et al., 1996). Depending on the termite taxa, the cellulose metabolism involving exocellullolitic activity is performed by endosymbionts, such as protozoans or bacteria, or exosymbionts such as wood-decaying fungi (Bignell, 2000). Our study mainly focuses on the biology of Reticulitermes flavipes (Kollar) (Rhinotermitidae), a subterranean termite distributed in eastern North America, but also an invasive species in Chile, Uruguy, and in parts of Europe (Austin et al., 2005). Reticulitermes flavipes is an important economic pest for structural damage as a wood destroying organism, and it has received attention for control solutions (Su and Scheffrahn, 1998). The annual cost of 17

PAGE 18

subterranean termites in the U.S. was estimated at $11 billion in 1999. This estimate included the cost of the repair of damaged structures and of termite control practices (Su, 2002). As a subterranean species, R. flavipes mainly lives in soil which is a favorable habitat for microbial growth and, therefore, has to interact with a complex microbial community, either commensal, opportunistic, parasitic or pathogenic (Rosengaus and Traniello, 1997). For this reason and its availability, R. flavipes was chosen as a model to study the interaction of termites with soilborne entomopathogens. The life cycle and the general biology of Reticulitermes sp. was recently reviewed by Lain and Wright (2003) and by Vargo and Husseneder (2009). In order to compare the biology of subterranean termites in relationships with pathogens, we also looked at termite species with diversity of behavior and habitat, representative of Isopteran phylogeny. However, the diversity was limited by the availability of different species in laboratory cultures. These evaluated species are introduced later in Chapter 3. Biological Control in Termites Of the 2,600 described termite species, only a small minority (70-80) are considered to be of structural pests (Edwards and Mills, 1986), or pest of agriculture and forestry (Logan et al., 1990). Control of these pest species, in particular subterranean termites in urban environments, has received a large amount of attention to reduce the losses due to structural damage (Su and Scheffrahn, 1998; Lax and Osbrink, 2003). The development of baits using chitin synthesis inhibitors have recently transformed the control practices (Su, 1994; Su et al., 1995; Grace et al., 1996; Grace and Su, 2001), although large amounts of liquid termiticides are still used in urban pest management (Su and Scheffrahn, 1998; Su, 2005). Considering the environmental impact of such practices, a need to evaluate non-chemical methods of termite control has been recognized as the alternative approach for structural protection (Milner et al., 1996; Grace, 1997; Culliney and Grace, 2000). 18

PAGE 19

Biological control of termites received interest starting in the 1960s with research on fungal pathogens (Lund, 1966, 1971; Lai et al., 1982; Zoberi and Grace, 1990a; Staples and Milner, 2000). The accumulation of data testing potential biological control agents against various termite pest species (Table 1-1) shows the level interest in this field (Culliney and Grace, 2000), in particular with the use of entomopathogenic fungi, but also with bacteria, nematodes and viruses. Interestingly, more than a hundred laboratory studies suggested a great potential for field application of the tested biological agents, but only four studies reported successful field assays and these were limited to colonies of mound-building or arboreal species (Hnel and Watson, 1983; Milner and Staples, 1996; Staples and Milner, 2000; Lenz, 2005) where large amounts of a pathogen formulation were blown directly into the central part of the nest. Such an approach is problematic with subterranean termite colonies, mainly due to their cryptic life style, long foraging distances, and the complex tunneling patterns of their nest (King and Spink, 1969; Su and Scheffrahn, 1988). The use of fungal pathogens for biological control of subterranean termites was based on the concept of classical biological control (Ferron, 1978; Lacey et al., 2001) with the use of a virulent agent that can self-replicate in a termite nest and be transmitted from individual to individual by high social interaction, so as to cause an epizootic and kill the entire colony. The feasibility of such an approach was based on the following assumptions: Humidity and temperature conditions in a termite nest and social behaviors of termites allow rapid and easy transmission of pathogens among individuals within a colony. The soil environment offers conditions highly favorable for sustaining infection and promoting epizootics. Because of the self-replicating nature of the fungus, it has the potential to spread and produce epizootics in termite populations. 19

PAGE 20

These assumptions were formulated by many authors (Toumanoff and Rombaut, 1965; Kramm et al., 1982; Lai et al., 1982; Hnel and Watson, 1983; McCoy, 1990; Wells et al., 1995; Zoberi, 1995; Delate et al., 1995; Boucias et al., 1996; Jones et al., 1996; Grace, 1997; Rosengaus and Traniello, 1997; Milner et al., 1998b; Culliney and Grace, 2000; Wright et al., 2002; Lax and Osbrink, 2003; Myles, 2002a; Chouvenc, 2003; Sun et al., 2003b), to the point that it reached a status of dogma in the field of termite biological control. However, Lund (1971) stated that In field trials, no organism has demonstrated significant pathogenicity, and no author has ever contradicted this statement for subterranean termites since. It is almost paradoxical to reach such a consensus about the potential of entomopathogens, when no natural or artificial epizootics were ever reported in field colonies. Epizootiology in Subterranean Termites From an epizootiological approach (Tanada and Kaya, 1993), the occurrence of an epizootic wave in a population depends on many parameters concerning the three characteristics of the disease triangle: The susceptible host The pathogen The environmental conditions The relationship between the pathogen and the host can be highly complex and variable depending on environmental conditions, and the possibility for a pathogen to replicate and spread in a given host population in order to complete multiple secondary cycling requires optimal conditions for the pathogen. The secondary cycling of a pathogen is indispensable for transmission of the pathogenic agent to all individuals in a given population, in order to ultimately produce an epizootic. 20

PAGE 21

Inoculating a subterranean termite colony with a given entomopathogen formulation through various protocols (Rath, 2000) is the only manipulative stage of the control process and the epizootic depends on the following conditions. Pathogen Virulence The etiological agent must ultimately cause death to the targeted host. However the degree of virulence can be critical for an epizootic to occur and it depends on two parameters that can be measured in the laboratory, the median lethal dose (LD 50 ) and the median lethal time (LT 50 ). A low LD 50 indicates that a small dose of the etiological agent is needed to produce a diseased host and kill it. A low LT 50 indicates that it takes a short time for the etiological agent to kill the host. However, if the pathogen kills the host too fast, it may not have the chance to spread into the termite population, and if the pathogen kills the host too slowly, diseased individuals may be removed or excluded from the population before the pathogen could replicate and be transmited to other individuals of the population. Among all the pathogens tested against various termites species (Table 1-1), many tested strains were considered virulent enough for field test applications (Milner et al., 1998a). Pathogen Density Although the pathogen virulence is critical for an epizootic to occur, the density of the etiological agent in a host population at a given time determines if the disease remains at an enzootic level. Increasing the density of a pathogen in the host population may allow the pathogen to reach lethal concentrations for a large number of individual termites and a high diseased individuals / healthy individuals ratio may trigger the epizootic wave (Figure 1-1). The density of a pathogen can increase in a population by self-replication. Inundative methods could be used for inoculating a massive concentration of the pathogen into a host population in order to bypass the need of secondary cycling by the pathogen. Inundative methods were used to 21

PAGE 22

inoculate large concentrations of fungal pathogens in mound-building termites as a mycoinsecticide (Hnel and Watson 1983; Milner, 2003), but diffuse subterranean termite nest structure renders inundative approaches impossible. The use of baits were considered (Wang and Powell, 2004), but delivering a required density of the pathogen throughout an entire subterranean nest remains problematic. Pathogen Capacity to Survive outside the Host Once introduced into a host population, the pathogen has to survive long enough to infect a potential host. The host environment has to be conducive enough to allow sustainability of the pathogen population and it was assumed that a subterranean termite nest is favorable for microbial sustainability. This assumption was also based on the hypothesis that termite defense mechanisms or other factors in the nest do not reduce the pathogen survivorship. In subterranean termites, the absence of reports of natural or artificial epizootics may question this assumption. Pathogen Capacity to Replicate Although the goal of the use of an etiological agent for biological control is to kill and reduce a host population, the development of the disease in individuals is the direct consequence of the agent processing through its life cycle. The completion of the pathogens life cycle may result in secondary cycling in the host population, which is critical to creating an epizootic wave. A high replication rate of a given pathogen can significantly increase the sustainability of the pathogen in the host population. Pathogen Capacity to Disperse When an etiological agent density is high in a portion of the host population, due to self replication or artificial inoculation, it is necessary for it to disperse in order to be in contact with the majority of the host population. In subterranean termites, the high social interactions among individuals and termite movements within the colony may greatly enhance the dispersal capacity 22

PAGE 23

of an etiological agent. However, this dispersion may result in pathogen dilution, which may keep the disease at enzootic level. It is therefore important for the dispersion to occur simultaneously with pathogen replication to prevent any dilution effect. The efficient spread of a disease in a termite colony depends, therefore, on the etiological agents capacity to replicate, disperse and survive in the nest environment, which was part of the previously stated assumptions. Host Susceptibility The pathogenicity of an etiological agent depends on the host capacity to prevent the infection and, ultimately, its capacity to prevent the pathogen to complete its life cycle. Although host immunity is mainly responsible for the overall degree of susceptibility, many other defense mechanisms can be involved to reduce the risk of infection in a termite population. Our study focuses on host immunity. Host Density The transmission of an etiological agent within a host population strongly depends on the encounter rate among the individuals of the population. The high density of subterranean termites in their underground nest allows frequent contact and this has been a key argument in favor for successful biological control. Host Behavior The host behavior may influence the transmission rate of an etiological agent. Grooming in termites was considered to be an important mode of pathogen dispersal in a colony (Milner et al., 1998; Myles, 2002b); however, it was also shown to reduce the pathogen load of the nest environment (Shimizu and Yamaji, 2003). Also, different behavioral processes may enhance or reduce the etiological agents chances to infect and disperse within a given population, and these processes are further examined in this study. 23

PAGE 24

Favorable Environmental Conditions Although the many previously described factors demonstrate the complexity of a host-pathogen interaction, the ultimate potential for an etiological control agent to complete its life cycle and produce an epizootic within a host population depends on favorable environmental conditions. Abiotic factors such as temperature and humidity are often considered to be critical for pathogen growth and sustainability (Sun et al., 2003a), but many biotic factors have to be taken into account. The associated microbial community inside a termite nest may interact and compete with the etiological agent (Rosengaus et al., 2003). However, termites have great control of their own nest environment which may limit the chances of a pathogen to successfully spread throughout the colony. Defense Mechanisms in Termites Despite the laboratory successes, the apparent absence of reports of a successful epizootic in the field suggests that the transmission and dispersal of the etiological agent in a termite colony is limited, and that some mechanisms within the colony prevent the pathogen from completing its life cycle. Recent works have emphasized the existence of defense mechanisms in termites that are involved in disease resistance (Figure 1-2), and many of these mechanisms are commonly found in other social insects (Cremer et al., 2007). These mechanisms include grooming behavior (Rosengaus et al., 1998b, Yanagawa and Shimizu, 2007), alarm behavior (Rosengaus et al., 1999a; Myles, 2002a), avoidance of cadavers (Milner et al., 1998b), necrophagy (Myles, 2002b), burial of infected cadavers (Jones et al., 1996), volatile chemicals in the nest (Wright et al., 2000; Rosengaus et al., 2000a, 2004), colony demography (Rosengaus and Traniello, 2001), immune response (Rosengaus et al., 1999b, 2007; Lamberty et al., 2001; Thompson et al., 2003; Bulmer and Crozier, 2004; Xu et al., 2009) antimicrobial compounds produced by termites gut (Rosengaus et al., 1998a; Siderhurst et al., 2005b), interactions with 24

PAGE 25

microorganisms (Rosengaus et al., 2003), social organization (Traniello et al., 2002) and nest architecture (Pie et al., 2004). These studies provide valuable information about individual defense mechanisms, however, it is unknown how all these factors interact with each other in a termite colony, and they may differ depending on the termite species and the pathogen involved. The Entomopathogenic Fungus Metarhizium anisopliae The interaction between termites and their potential pathogens has received particular attention in two distinct fields of research. The first is the use of a pathogenic agent for biological control against various pest species, mainly in the Rhinotermitidae (Grace, 1997; Culliney and Grace, 2000). The second is the use of termites as a model for understanding the evolution of disease resistance of eusocial insects in an environment that can favor microbial growth and increase the risk of epizootics (Rosengaus et al., 1998b). Metarhizium anisopliae (Metsch.) Sorokin is probably the most well-studied microorganism in both fields of research (Rath, 2000). Metarhizium anisopliae is a soil fungus commonly found worldwide in tropical, sub-tropical, temperate, and near-arctic regions of the globe (Bidochka and Small, 2005). Although M. anisopliae has the capability to survive and persist in the soil (Milner et al., 2003; St. Leger, 2008), it is considered to be a generalist pathogen, and has been tested against a wide range of insects such as locusts, moths, beetles (Shah and Pell, 2003), flies (Davidson and Chandler, 2005), cockroaches (Patchamuthu et al., 1999), termites (Milner, 2000), and other arthropods including ticks (Arruda et al., 2005). Some strains were considered more host specific than others (Fargues et al., 1976; Bridge et al., 1997; Milner et al., 1998a). Metarhizium anisopliae is a haploid mitosporic fungus assumed to reproduce clonally (Bidochka and Small, 2005). A conidium usually germinates on the surface of the cuticle of an insect host, produces a germinating tube and an appressorial structure to help binding to the cuticle surface, and penetrates the epicuticle by releasing some cuticle degrading enzymes (St. 25

PAGE 26

Leger et al., 1991). Once through the cuticle, the hyphal body releases toxins such as destruxins into the hosts hemocoel, and ultimately kills the host. Destruxins are a family of cyclic peptides with 23 variants known, and all comprise five amino acids and a hydroxyacid (Kershaw et al., 1999) and their secretion can be variable depending on the strain (Kaijiang and Roberts, 1986). After reaching the hemocoel, the fungus kills the host and invades the cadavers body to produce a new generation of conidia that can eventually disperse in the environment (Figure 1-3). Objectives Our study was based on the following question: Why has biological control in subterranean termites failed so far? We found it is important to provide data that could explain the biological mechanisms involved in resistance of termites against a fungal epizootic. Thus, this study investigates the interactions between termites and the entomopathogenic fungus M. anisopliae, from the cellular to the colony level, in order to comprehend the role of the different defense mechanisms involved in a termites disease resistance and to re-evaluate the potential candidacy of this pathogenic fungus for biological control. In Chapter 2, we first tested the susceptibility of R. flavipes to M. anisopliae in a laboratory protocol commonly used in previous studies, and compared these results with the survival rate of fungal-exposed termites in a soil environment, using large two-dimensional arenas. The capacity of the fungus to survive and replicate in such an environment was also tested. In Chapter 3, we conducted susceptibility tests on a large range of termite species to confirm the variability of disease resistance within the Isoptera. This also provided many specimens for histological analysis later in this dissertation. Chapter 4 describes a histological approach to observing the gut activity of R. flavipes against fungal infections, and Chapter 5 provides additional information about the antifungal gut activity in other termite species. In Chapter 6, we tested the antifungal activity of the chemical norharmane, an alkaloid found in the termite gut that is suspected to play 26

PAGE 27

a key role in a termites defense. Chapter 7 describes a histological approach to observing a cellular encapsulation against fungal infection in R. flavipes and Chapter 8 compares this cellular encapsulation among multiple termite species. Finally, Chapter 9 concludes on the interaction of defense mechanisms in termites against entomopathogens to explain the absence of epizootics in field colonies and discusses the potential of biological control in subterranean termites. 27

PAGE 28

Table 1-1. List of termite species that were tested against various entomopathogens including fungi, bacteria, nematodes and viruses, but not including parasitic organisms. Termite species Pathogen References Rhinotermitidae Reticulitermes flavipes Fungi Antennopsis gallica Gouger and Kimbrough, 1969 Aspergillus flavus Beal and Kais, 1962 Lund, 1966, 1971 Beauveria bassiana Bao and Yendol, 1971 Boucias et al., 1996 Grace and Zoberi, 1992 Toumanoff and Rombaut, 1965 Wang and Powell, 2003 Zoberi and Grace, 1990a Conidiobolus coronatus Yendol and Paschke, 1965 Yendol and Rosario, 1972 Metarhizium anisopliae Engler and Gold, 2004 Myles, 2002a, 2002b Ramakrishnan et al., 1999 Shimizu and Yamaji, 2003 Strack, 2000 Wang and Powell, 2004 Zoberi, 1995 Toumanoff and Rombaut, 1965 Paecilomyces sp. Smythe and Coppel, 1966 Penicillium sp. Smythe and Coppel, 1966 Bacteria Bacillus sp. Toumanoff, 1966 Bacillus thuringiensis Smythe and Coppel, 1965 Serratia marcescens Lund, 1971 Nematodes Heterorhabditis heliothidis Mauldin and Beal, 1989 Heterorhabditis indica Wang et al., 2002 Neosteinernema longicurvicauda Nguyen and Smart, 1994 Steinernema carpocapsae Mauldin and Beal, 1989 Trudeau, 1989 Wang et al., 2002 Reticulitermes hesperus Bacteria Bacillus thuringiensis Smythe and Coppel, 1965 28

PAGE 29

Table 1-1. Continued Termite species Pathogen References Rhinotermitidae Reticulitermes lucifugus Fungi Basidiobolus sp. Krejzov, 1972 Conidiobolus coronatus Krejzov, 1975 Reticulitermes speratus Fungi Aspergillus niger Suzuki, 1995 Beauveria bassiana Suzuki, 1995 Metarhizium anisopliae Suzuki, 1995 Paecilomyces fumosoroseus Suzuki, 1995 Nematodes Steinernema feltiae Wu et al., 1991 Reticulitermes tibialis Nematodes Steinernema feltiae Epsky and Capinera, 1988 Reticulitermes virginicus Fungi Antennopsis gallica Gouger and Kimbrough, 1969 Aspergillus flavus Beal and Kais, 1962 Trichoderma virens Heintschel et al., 2007 Bacteria Bacillus thuringiensis Smythe and Coppel, 1965 Reticulitermes sp. Fungi Gliocladium virens Kramm and West, 1982 Metarhizium anisopliae Kramm and West, 1982 Kramm et al., 1982 Beauveria bassiana Kramm and West, 1982 Nematodes Steinernema carpocapsae Georgis et al., 1982 Heterotermes tenuis Fungi Beauveria bassiana Almeida et al., 1997 Moino and Alves, 1998 Moino et al., 2002 Beauveria brongniartii Almeida et al., 1997 Metarhizium anisopliae Moino and Alves, 1998 Moino et al., 2002 29

PAGE 30

Table 1-1. Continued Termite species Pathogen References Rhinotermitidae Heterotermes indicola Bacteria Bacillus thuringiensis Khan et al., 1977a, 1978, 1985 Pseudomonas aeruginosa Khan et al., 1992 Serratia marcescens Khan et al.,1977b Coptotermes formosanus Fungi Aspergillus flavus Jayasimha and Henderson, 2007b Aspergillus niger Suzuki, 1995 Beauveria bassiana Delate et al., 1995 Grace, 1991, 1993 Jones et al., 1996 Lai et al., 1982 Lai, 1977 Sun et al., 2002, 2003a, 2003b Suzuki, 1995 Wang and Powell, 2003 Wells et al., 1995 Wright et al., 2002 Beauveria brongniartii Yanagawa et al., 2008 Yoshimura and Takahashi, 1998 Conidiobolus coronatus Yoshimura et al., 1992 Ko et al., 1982 Krejzov, 1975, Wells et al., 1995 Metarhizium anisopliae Delate et al., 1995 Dong et al., 2007 Engler and Gold, 2004 Grace, 1993 Jones et al., 1996 Ko et al., 1982 Lai et al., 1982 Lai, 1977 Leong, 1966 Meikle et al., 2005 Sun et al., 2002, 2003a, 2003b Suzuki, 1995 Wang and Powell, 2004 Wells et al., 1995 Wright et al., 2002, 2005 Yanagawa and Shimizu, 2007 Yanagawa et al., 2008 Metarhizium flavoviride Wells et al., 1995 30

PAGE 31

Table 1-1. Continued Termite species Pathogen References Rhinotermitidae Coptotermes formosanus Fungi Paecilomyces fumosoroseus Meikle et al., 2005 Suzuki, 1995 Yanagawa et al., 2008 Trichoderma harzianum Jayasimha and Henderson, 2007b Bacteria Bacillus thuringiensis Grace and Ewart, 1996 Serratia marcescens Connick et al., 2001 Osbrink et al., 2001 Nematodes Heterorhabditis indica Mankowski et al., 2005 Steinernema carpocapsae Fujii, 1975 Mankowski et al., 2005 Steinernema feltiae Wu et al., 1991 Coptotermes acinaciformis Fungi Metarhizium anisopliae Rath and Tidbury, 1996 Milner and Staples, 1996 Milner et al., 1997 Milner, 2000 Coptotermes frenchi Fungi Metarhizium anisopliae Milner et al., 1998a Coptotermes curvignathus Fungi Metarhizium anisopliae Sajap and Kaur, 1990 Sajap and Jan, 1990 Beauveria bassiana Sajap and Jan, 1990 Conidiobolus coronatus Sajap et al., 1997 Coptotermes gestroi Fungi Metarhizium anisopliae Krutmuang and Mekchay, 2005 Maketon et al., 2007 Nematodes Heterorhabditis indica Mankowski et al., 2005 Steinernema carpocapsae Mankowski et al., 2005 Coptotermes heimi Bacteria Pseudomonas aeruginosa Khan et al., 1992 31

PAGE 32

Table 1-1. Continued Termite species Pathogen References Rhinotermitidae Coptotermes lacteus Fungi Metarhizium anisopliae Staples and Milner, 2000 Milner and Staples, 1996 Milner et al., 1998a Milner, 2000 Virus ABPV Gibbs et al., 1970 Hodotermitidae Hodotermes mossambicus Fungi Metarhizium anisopliae Langewald et al., 2003 Beauveria bassiana Langewald et al., 2003 Termopsidae Zootermopsis angusticollis Fungi Metarhizium anisopliae Rosengaus and Traniello, 1997 Rosengaus et al., 1998b Rosengaus and Traniello, 2001 Bacteria Bacillus thuringiensis Smythe and Coppel, 1965 Nematodes Steinernema carpocapsae Wilson-Rich et al., 2007 Zootermopsis sp. Fungi Paecilomyces fumosoroseus Krejzov, 1976 Nematodes Steinernema carpocapsae Georgis et al., 1982 Porotermes adamsoni Virus ABPV Gibbs et al., 1970 Kalotermitidae Bifiditermes beesoni Bacteria Bacillus thuringiensis Khan et al., 1985 Serratia marcescens Khan et al., 1977b Cryptotermes brevis Fungi Metarhizium anisopliae Nasr and Moein, 1997 Leong, 1966 Verticillium indicum Nasr and Moein, 1997 32

PAGE 33

Table 1-1. Continued Termite species Pathogen References Kalotermitidae Glyptotermes dilatatus Nematodes Heterorhabditis sp. Dantharanarayana and Vitarana, 1987 Incistitermes immigrans Fungi Metarhizium anisopliae Leong, 1966 Kalotermes flavicollis Virus NPV Al Fazairy and Hassan, 1988, 1993 Neotermes connexus Fungi Metarhizium anisopliae Leong, 1966 Neotermes rainbowi Fungi Metarhizium anisopliae Lenz, 2005 Termitidae Ancistrotermes guineensis Nematodes Heterorhabditis bacteriophora Rouland et al., 1996 Steinernema carpocapsae Rouland et al., 1996 Benmoussa-Haichour et al., 1998 Steinernema kushidai Rouland et al., 1996 Benmoussa-Haichour et al., 1998 Cornitermes cumulans Fungi Beauveria bassiana Fernandes and Alves, 1991 Neves and Alves, 1999, 2000a, 2000b, 2004 Metarhizium anisopliae Fernandes and Alves, 1991 Neves and Alves, 1999, 2000a, 2000b, 2004 Macrotermes michaelseni Fungi Metarhizium anisopliae Gitonga et al., 1995 Beauveria bassiana Gitonga et al., 1995 Macrotermes Bellicosus Nematodes Steinernema carpocapsae Benmoussa-Haichour et al., 1998 Steinernema kushidai Benmoussa-Haichour et al., 1998 33

PAGE 34

Table 1-1. Continued Termite species Pathogen References Termitidae Macrotermes subhyalinus Fungi Beauveria bassiana Langewald et al., 2003 Cordycepioideus bisporus Ochiel, 1995 Ochiel et al., 1997 Metarhizium anisopliae Langewald et al., 2003 Paecilomyces fumosoroseus Ochiel, 1995 Macrotermes sp. Fungi Metarhizium anisopliae Maniania et al., 2002 Microcerotermes championi Bacteria Bacillus thuringiensis Khan et al.,1977a, 1985 Pseudomonas aeruginosa Khan et al., 1992 Serratia marcescens Khan et al.,1977b Microcerotermes sp. Fungi Metarhizium anisopliae Krutmuang and Mekchay, 2005 Microtermes sp. Fungi Metarhizium anisopliae Maniania et al., 2002 Nasutitermes exitiosus Fungi Absidia coerulea Hnel, 1982a Conidiobolus obscurus Hnel, 1982a Entomophthora virulenta Hnel, 1982a Entomophthora destruens Hnel, 1982a Metarhizium anisopliae Hnel, 1981, 1982a, 1982b Hnel and Watson, 1983 Milner and Staples, 1996 Milner et al., 1997, 1998a Milner, 2000, 2003 Rath and Tidbury, 1996 Paecilomyces fumosoroseus Hnel, 1982a Virus ABPV Gibbs et al., 1970 Nasutitermes corniger Nematodes Steinernema carpocapsae Laumond et al., 1979 34

PAGE 35

Table 1-1. Continued Termite species Pathogen References Termitidae Odontotermes brunneus Fungi Metarhizium anisopliae Khan, 1991 Verticilium lecanii Khan, 1991 Odontotermes formsanus Fungi Metarhizium anisopliae Dong et al., 2009 Odontotermes latericus Fungi Beauveria bassiana Langewald et al., 2003 Metarhizium anisopliae Langewald et al., 2003 Odontotermes obesus Fungi Aspergillus flavus Sannasi, 1968 Beauveria bassiana Khan et al., 1993 Metarhizium anisopliae Khan et al., 1993 Metarhizium flavoridae Khan et al., 1993 Paecilomyces fumosoroseus Khan et al., 1993 Paecilomyces lilacinus Khan et al., 1993 Pseudacanthotermes spiniger Nematodes Heterorhabditis bacteriophora Rouland et al., 1996 Benmoussa-Haichour et al., 1998 Steinernema carpocapsae Rouland et al., 1996 Benmoussa-Haichour et al., 1998 Steinernema kushidai Rouland et al., 1996 Benmoussa-Haichour et al., 1998 35

PAGE 36

PrevalenceTime Epizootic Post-epizooticPre-epizooticEnzooticEnzootic Epizootic wave Figure 1-1. Curve showing an epizootic wave. Prevalence: number of hosts expressing a disease at any given point of time. (Modified after Steinhaus, 1949) 36

PAGE 37

conidiaMetarhizium anisopliae ?Termite nestRepellency / Behavioralavoidance / Alarm responseVolatile Chemicals produced by the termites or associated microbes in the nestGrooming activity Behavioral processInsect Immunity-Hemocytic(Encapsulation)-Humoral (Defensins)Necrophagy/ Burial /Corpse avoidanceMicrobial competition Epizootic Figure 1-2. Potential defense mechanisms involved in a termites nest against an epizootic event. 37

PAGE 38

Host ABCDE Figure 1-3. Life cycle of the entomopathogenic fungus Metarhizium anisopliae in an insect host. A) The conidium binds to the surface of the cuticle of the host. B) The conidium germinates and the germinating tube penetrates through the cuticle into the hemocoel. C) The fungus releases toxins that kill the host. D) The hyphal body invades the cadaver. E) The fungus releases a new generation of asexual spores in the environment (Modified after Tanada and Kaya, 1993). 38

PAGE 39

CHAPTER 2 INTERACTION BETWEEN RETICULITERMES FLAVIPES AND METARHIZIUM ANISOPLIAE IN FORAGING ARENAS Introduction Many studies have investigated the effect of the entomopathogenic fungus M. anisopliae on subterranean termites under laboratory conditions (Chapter 1). Although the virulence of different strains of M. anisopliae was extensively studied in the laboratory (Milner et al., 1998a), very few successful field assays were reported and even these successes were limited to field colonies of mound-building species (Hnel and Watson, 1983; Milner and Staples, 1996). For mound-building species, termite colonies were killed when large quantities of pure, dry conidia were blown directly in the nursery region (Milner and Staples, 1996). Unfortunately, it is extremely difficult to directly introduce a massive amount of conidia in subterranean termite colonies, mainly due to their cryptic life style, long foraging distances, and complex tunneling patterns of their gallery system (King and Spink, 1969; Su and Scheffrahn, 1988). Moreover, behavioral avoidance (repellency) and grooming by termites are two important factors that may inhibit an epizootic in subterranean termite nests (Zoberi, 1995; Staple and Milner, 2000; Myles, 2002a; Shimizu and Yamaji, 2003; Yanagawa and Shimizu, 2007;). Rath (2000) reviewed several potential uses of entomopathogenic fungi: (1) a short-term control measure to eliminate active termites using fungal dust or liquid application directly into the active galleries of the infested structure, (2) soil application to prevent termite infestation in a given area using a repellent formulation of the conidia, and (3) a long-term control measure where the fungus is used to suppress or eliminate the active colony by means of baiting programs or direct dosing of termites in the nest. In the scope of controlling an active termite colony, the first and second uses do not eliminate the termite colony through an epizootic because repellency is mainly used in these strategies to protect the structure (Milner et al., 1996). Therefore, the use 39

PAGE 40

of baits or the direct application on the population appear to be the only strategy that would spread the conidia among the entire population and cause an epizootic (Culliney and Grace, 2000). Wang and Powell (2004) used a cellulose bait to improve the effectiveness of M. anisopliae in laboratory conditions against R. flavipes and Coptotermes formosanus Shiraki, but their results did not confirm the non-repellent nature of the formulation at high conidia concentrations. Thus, biological control of subterranean termites using M. anisopliae in baits remains problematic without the availability of a palatable and truly non-repellent formulation. Milner et al. (1996) suggested the trap and treat system as one of the approaches in bait technology. This approach consists of trapping part of a termite population, treating the individuals with a pathogenic agent and releasing them back to their nest, thus allowing them to spread the agent to the entire colony. In this study, we used a foraging arena to simulate the trap and treat application of M. anisopliae against laboratory groups of R. flavipes. Previous laboratory studies with fungal pathogens typically exposed a small number of termites (<100) in Petri dishes that had little resemblance to their subterranean habitat. Termites in the Petri dishes were forced into fungal exposure with few choices for retreat, avoidance, or to fully perform their hygienic behavioral processes (Myles, 2002a). Termites in our large two-dimensional foraging arena were allowed to establish their galleries before the fungal agents were applied, thus simulating field conditions, and were given the choices to retreat to untreated tunnels. Moreover, the arenas allowed us to continuously monitor the termite populations with minimum disturbance for extended periods of time. In this study, we first determined the virulence (LD 50 and LD 90 ) of the chosen M. anisopliae strain against R. flavipes individuals. The trap and treat approach was tested using 40

PAGE 41

two-dimensional foraging arenas by contaminating a group of individuals and releasing them inside the arenas. Ninety d after M. anisopliae infected termites were released into the arenas, the arenas were dismantled and sand samples were obtained to estimate the survival rate of M. anisopliae in subterranean termite foraging galleries. Material and Methods Collection of Termite Colonies Termites were collected from three field colonies of R. flavipes in Fort Lauderdale, FL, USA, by using underground bucket traps (Su and Scheffrahn, 1986) containing a bundle of spruce (Picea sp.) wood. Before testing, termites were kept at 28C in 1-liter cylindrical plastic containers with pieces of moist wood. Termites were used in experiments 10 to 15 d after collection. Caste proportion was established from the collected samples: 2% soldiers, 4% undifferentiated larvae of second instar or less and 94% workers (undifferentiated larvae of at least the third instar). Metarhizium anisopliae Conidia Preparation The M. anisopliae strain used was ATCC 90448 (Rosengaus et al., 1998b). Conidia were spread for germination on 1/5 strength potato dextrose agar (1/5 PDA) and incubated at 27 C in the dark. After 48 h, single-spore colonies were transferred to 1/5 PDA plates containing one dead termite that had been killed and surface-sterilized for 1 h in a vial saturated with chloroform vapors. Inoculated plates were then incubated at 27 C for 14 d in the dark. Fresh conidia were harvested from these plates with a 0.1% Tween 80 solution and a stock suspension of 10 8 conidia/ml was prepared. Concentrations were determined with a hemocytometer. The conidia stock suspension was stored at 4C and used within 15 d for the experiment described herein. 41

PAGE 42

Metarhizium anisopliae Virulence on Reticulitermes flavipes Eight treatments were tested on R. flavipes: 0.1% Tween 80 solution only, 10 3 conidia/ml, 10 4 conidia/ml, 10 5 conidia/ml, 10 6 conidia/ml, 10 7 conidia/ml, 10 8 conidia/ml, and an untreated control. Termites were chilled to 4 C for 20 min, placed on a sterile glass microscope slide and then individually treated on the dorsal area with a 1l droplet of treatment suspension, which immersed the termite in the droplet, to inoculate 0, 1, 10, 100, 1,000, 10,000, and 100,000 conidia per termite, respectively. Termites were then returned to 4 C for 20 min in order to prevent immediate grooming. Groups of 20 treated termites (19 workers + 1 soldier) that received the treatment were placed in a Petri dish (5-cm diameter and 1.5-cm high) with moistened filter paper on the bottom. Petri dishes were stored in an incubator at 27C for 8 d. Mortality was recorded and dead termites were removed daily. Each of eight treatments consisted of three replicates (one replicate per colony of origin) of 20 termites each. A Cox proportional hazard regression analysis was performed on 480 individuals in order to estimate the effect of each treatment on R. flavipes survivorship. Through the analysis the Wald statistic was generated. The resulting hazard function defines the instantaneous rate of death at a particular time, while controlling for the effects of other variables on survival (Rosengaus et al., 2000b). The LD 50 and LD 90 and their 95% Fiducial limits (FL) were determined by Probit analysis (SAS Institute, 2002). Foraging Arenas The foraging arena was composed of two sheets of transparent Plexiglas (60 by 60 by 0.6 cm in thickness) separated from each other by Plexiglas laminates (5 cm in width and 0.2 cm in thickness) on the four sides, creating a 50 by 50 by 0.2 cm space inside the arena. A piece of wood (Picea sp., 10 by 8 by 0.2 cm) was placed in the arena, 2 cm away from the upper-right corner. An introduction chamber (5-cm diameter and 7-cm tall) was placed in the center of the 42

PAGE 43

top Plexiglas sheet and was connected with the inside of the arena by four 1-cm diameter access holes (Figure 2-1A). The arena was filled with 700 g of sifted sand (150-500 m sieves, rinsed with deionized water, oven dried at 60C for 24 h) and moistened with 175 ml of sterile deionized water. A total of 900 termites was placed into the introduction chamber with the same caste ratio as the field colony, and a 5-cm diameter Plexiglas lid covered this chamber. The arena was stored in a dark room at 26 2C for 10 d, allowing the termites to forage and to tunnel inside the arena, excavate some sand from the arena into the introduction chamber, and start feeding on the piece of wood (Figure 2-1B). These groups of 900 termites were referred to as the naive termites (NT) for the rest of the study. On the same day that the 900 NT were introduced in the arena, 60 nestmates (all workers) were kept in a Petri dish (9-cm diameter and 1.5-cm high) and placed in an incubator at 27C for 10 d. During this period, the 60 termites were fed with filter paper stained with a 0.1% aqueous solution of Nile blue, which resulted in the termites turning blue (Su et al. 1991). On the 10 th day, the 60 blue termites were individually treated with 1 l of a M. anisopliae suspension (10 7 conidia/ml) with the same method described previously. Each of the treated termites was therefore exposed to 10,000 conidia. The 60 blue termites exposed to M. anisopliae (BEMa termites) were placed in the introduction chamber of the arena. Within 5 min, these 60 BEMa termites spread through the arena among the 900 NT. Once the 60 BEMa termites entered the arena, the introduction chamber was removed and the access holes were covered. Therefore, 6.25% of the 960 termites in the arena were inoculated with M. anisopliae. The arenas were replicated three times per colony origin (n = 9). In these nine arenas, NTE referred to the groups of 900 naive termites in contact with the 60 BEMa (exposed termites). 43

PAGE 44

For a control treatment, 60 blue control termites (BC) were treated with 1 l of a conidia-free solution (0.1% Tween 80) each. They were then introduced to 900 naive termites in a foraging arena, and there were three control arenas per colony origin (n = 9). In these nine arenas, there were 60 blue termites treated with a control solution (BC) in contact with 900 naive termites (NTC). After the introduction of the blue termites (BC or BEMa), each of the 18 arenas were stored in a dark room at 26 2C. Digital images of the arenas were taken daily during the first 20 d and every 5 d afterward, for a total period of 90 d. The number of live termites, both treated + naive, was counted using the images projected on a computer monitor. A Cox proportional hazard regression analysis (using the program R-Project for statistical computing, version 2.4; http://cran.r-project.org ) was performed to determine the effect of the 60 termites treated with M. anisopliae on the survivorship of the 900 naive termites. Survival of Metarhizium anisopliae in Sand Three Petri dishes (5-cm diameter and 1.5-cm high) were filled with 10 g of dry sand moistened with 2.5 ml of deionized water (same moisture content as in arenas). A suspension of M. anisopliae conidia was added to each Petri dish (10,000 conidia per Petri dish) and homogenized with the sand. Petri dishes were sealed with Parafilm to prevent the sand from drying and incubated at 27C in the dark. After 30 d, two sand samples (1 g each) were randomly obtained from each Petri dish and serial dilutions of these samples were plated (three replicates per dilution) on a M. anisopliae selective medium (Veen and Ferron, 1966). Plates were incubated at 27C in the dark, and the number of colony forming units (CFU) of M. anisopliae was counted after 10 d. To confirm that the sand used in the experiments had no natural occurrence of M. anisopliae, a negative control was also prepared with the same protocol, but 44

PAGE 45

instead of adding a conidial suspension into the sand, a solution of 0.1% Tween 80 was added and mixed with the sand. Transfer and Dispersal of Metarhizium anisopliae Conidia from Termite Cuticle to Surrounding Sand In order to determine if the conidia could be transferred from the cuticle of termites to the surrounding sand of the tunnel walls and beyond, an additional control was prepared. Ten workers were treated with 1,000 conidia each as described above and transferred into a Petri dish (5-cm diameter and 1.5-cm high) containing 10 g of dry sand that had been moistened with 2.5 ml of deionized water. Artificial cavities were prepared on the surface of the sand to reproduce the tunnel environment of the arena and the multiple contacts between the sand walls and the termite cuticles. Termites were allowed to move on the surface of the sand for 60 min and then were removed from the Petri dish. The Petri dish was sealed with Parafilm, replicated (n = 3) and incubated at 27C in the dark for 30 d before it was opened for sampling. Two types of samples were obtained from each Petri dish: (1) two samples (1 g each) from the surface of the sand (no more than 3 mm away from the surface) that was exposed to infested termites and, (2) two samples (1 g each) from sand that was not in direct contact with the termites (4-6 mm away from the surface of the sand). Serial dilutions of each sample were plated on the M. anisopliae selective medium (three replicates per dilution) and plates were incubated at 27C in the dark. CFU of M. anisopliae were counted after 10 d. Survival of Metarhizium anisopliae in Arenas Containing Termites At the end of 90 d of the arena experiment, two untreated controls (arenas originally with 60 BC) and two infected arenas (arenas originally with 60 BEMa) were randomly chosen for isolation of M. anisopliae. Each arena was opened and six sand samples, each of two sample types per arena, were obtained (1 g per sample), including six samples from the termite tunnel 45

PAGE 46

walls (sand mixed with termite fecal material) and six samples from undisturbed areas (not tunneled by termites) of the arena. Serial dilutions of samples were plated on the M. anisopliae selective medium (three replicates per dilution). Plates were incubated at 27C in the dark, and CFU of M. anisopliae were counted after 10 d. The average recovery of M. anisopliae from the arenas was determined using only the portion of the sand that was in direct contact with the termites (tunnel walls, 123 g of sand per arena) after confirming that M. anisopliae did not disperse in the undisturbed areas of the arena. Results Metarhizium anisopliae Virulence on Reticulitermes flavipes Survivorships of R. flavipes exposed to M. anisopliae conidia are shown in Figure 2-2. No significant survivorship difference was observed between untreated R. flavipes workers and those treated with a control solution (Wald statistic = 0.17, df = 1, P = 0.68), with one conidium (Wald statistic = 0.15, df = 1, P = 0.69), with 10 conidia (Wald statistic = 1.1, df = 1, P = 0.29) or with 100 conidia (Wald statistic = 2.34, df = 1, P = 0.12). However, termites exposed to 1,000, 10,000 and 100,000 conidia had 14.7 (Wald statistic = 19.7, df = 1, P < 0.0001), 44.6 (Wald statistic = 40.1, df = 1, P < 0.0001) and 54 (Wald statistic = 43.8, df = 1, P < 0.0001) times the hazard ratio of death of the controls, respectively. Moreover, termites treated with 10,000 and 100,000 conidia had similar survival (Wald statistic = 1.7, df = 1, P = 0.30). The LD 50 was determined to be 1,030 conidia per termite (n = 420, 95% FL = 843-1,253) and the LD 90 was 9,885 conidia per termite (n = 420, 95% FL = 7,525-13,390). Termite Survival in Foraging Arenas The groups of 900 naive termites (NT) remained for 10 d in the arenas before the introduction of the 60 blue termites. During these 10 d, mortality occurred and at the moment of the introduction of the 60 blue termites, the number of NT in the 18 different arenas was lower 46

PAGE 47

than 900 individuals (n = 14,977 instead of 16,200). Once the blue termites were introduced in the arenas, alert and aggressive behaviors (as described by Myles, 2002a) were observed from the naive termites (NTC and NTE) toward the blue termites (BC and BEMa, respectively) irregardless of the fungal inoculation. These behaviors ceased within 10 min after the introduction of the blue termites, but were followed by extensive grooming behavior of the blue termites in both sets of arenas (Figure 2-3). Cannibalism of weak and dead termites was observed several times and no isolated cadaver was found in the arenas during the entire experiment (Figure 2-4). Within the first 20-d experimental period, blue termites exposed to 10,000 conidia (BEMa) had six times the hazard ratio of death than the blue termites treated with a control solution (BC) (n = 1080, Wald statistic = 454, df = 1, P < 0.0001). Within 5 d, only 10% of BEMa termites survived the arenas, while the survivorship of the BC termites was 70% (Figure 2-5). Within the 90-d experimental period, the survivorship of the 900 naive termites (NTE) that were exposed to 60 termites infested with M. anisopliae was not different from that of the 900 naive termites (NTC) that were exposed to 60 termites treated with a control solution (Figure 2-6), (n = 14,977, Wald statistic = 1.2, df = 1, P = 0.26). Survival of Metarhizium anisopliae in Sand The sand used in this study did not contain naturally occurring M. anisopliae (Table 2-1), while 49,500 2,260 (SE) CFU/g (n = 18) were isolated from the sand mixed with a conidial suspension (1,000 conidia/g of sand) after 30 d. Transfer and Dispersal of Metarhizium anisopliae Conidia from Termite Cuticle to Surrounding Sand For sand that was in contact with 10 termites exposed to M. anisopliae for 60 min, 219 19 (SE) CFU/g (n = 18) were recovered from the sand layer that was in direct contact with the 47

PAGE 48

termites (< 3 mm away from contact surface) after 30 d. However, no M. anisopliae was isolated from the sand layer 4-6 mm away form the contact surface after 30 d. Survival of M. anisopliae in Arenas Containing Termites In the sand samples from the arenas, M. anisopliae was isolated only from samples that were in direct contact with termites infested with M. anisopliae (BEMa) (Table 2-1). In the control arenas containing 900 naive termites (NTC) and 60 blue termites treated with a control solution (BC), no M. anisopliae was isolated from samples of the tunnels or from samples of undisturbed areas of the arenas. In the arenas with 900 naive termites (NTE) and 60 termites treated with M. anisopliae (BEMa), M. anisopliae was not isolated from the undisturbed area of the arenas, but an average of 3.24 0.55 (SE) CFU/g sand were isolated from the tunnel samples (n = 12). A total of 600,000 conidia of M. anisopliae were originally introduced in these arenas via the 60 BEMa termites, and an average of 17.6% (123 g of sand) of the arenas was foraged by the termites within the 90-d period. It was estimated that the average recovery of M. anisopliae from these arenas was 399 67 (SE) CFU per arena, which is only 0.066% of the 600,000 conidia originally introduced in these arenas. Discussion The results of this study confirmed that M. anisopliae ATCC90448 is a pathogenic fungus against subterranean termites in laboratory conditions using a Petri dish bioassay. However, termites treated with up to 100 conidia did not exhibit differential mortality from that of the controls. This result suggests that a relatively high dosage is needed to induce the lethal effect of M. anisopliae on R. flavipes. Once the LD 90 was determined, the trap and treat approach was subsequently tested using large foraging arenas to provide a soil environment for the termites 48

PAGE 49

and also to study the transfer of the fungus in a group of termites that have the ability to forage and tunnel. For groups of 960 R. flavipes living in sand-filled arenas, our results showed that when 6.25% of the termites were infected with M. anisopliae, the fungus killed only termites that were originally inoculated with the conidia. The mortality of the naive termites was not significantly different from those in control arenas after 90 d. We conclude that there was minimal transfer of the fungus and no observable disease development in the naive termites under these conditions. The absence of epizootics in the foraging arena experiment may be due to several factors: (1) dilution of the conidia in the arena environment, (2) reduction of conidial load via grooming, and (3) the inability of M. anisopliae to grow and reproduce as dead termites were cannibalized before M. anisopliae produced new conidia. In the arena experiment, 600,000 conidia were originally introduced via 60 termites treated with 10,000 conidia (BEMa). However, a portion of these conidia were dispersed in the arena by simple contact of the cuticle from innoculated termites with the sand surface, although it was also demonstrated that M. anisopliae neither dispersed nor produced mycelium away from the contact area. Moreover, another portion of these conidia were ingested by grooming. According to Shimizu and Yamaji (2003) with R. speratus (Kolbe), 3 h after grooming a termite infested with M. anisopliae conidia, 90% of the conidia were found in the gut of the groomers. This suggests that shortly after the introduction of the 60 BEMa into the arena, a majority of their conidial load was ingested and went through the gut of the groomers, subsequently reducing the number of available conidia in the environment. Finally, all dead BEMa termites were rapidly cannibalized, suggesting that most of the remaining M. anisopliae units also had to go through the gut of their nestmates. 49

PAGE 50

Several studies have shown that germination rates of M. anisopliae and Beauveria bassiana (Balsamo) Vuill. were reduced in the termites gut and did not invade the general cavity through the termites gut (Bao and Yendol, 1971; Kraam and West, 1982; Boucias et al., 1996; Rosengaus et al., 1998a; Yanagawa and Shimizu, 2007). We demonstrated that in sand samples without termites, 50 times more CFU of M. anisopliae were recovered than were introduced, confirming that conidia are viable in the sand (Milner et al., 2003) in the absence of termites, and had the ability to germinate and produce mycelium. However in our arena study, the minimal (0.066%) recovery of M. anisopliae from the infested arenas showed that the survival and growth of the fungus was extremely reduced in the termite tunnels. This supports the idea that the presence of subterranean termites inhibits M. anisopliae and that the passage of the conidia through their gut helps preventing the occurrence of an epizootic. Therefore, instead of the fungus killing the termite population, the termite population inhibited fungal growth and viability. M. anisopliae infestation of a subterranean termite colony with a trap and treat approach may be more difficult than expected, as suggested by Milner (2000). A survey by Su et al. (1993) suggested that, on average, only 2.75% of a field colony was captured with field traps and this number was reduced to 0.8% for the largest colonies ( > one million termites). The foraging arenas described herein had 6.25% of infested individuals, yet no epizootic was observed in any of the arenas. If 900 termites could not be controlled with 6.25% infested termites in a foraging arena, then the chance of controling a termite population in a large field colony with less than 1% lethally infected termites is very low for this strain of M. anisopliae. The transmission of the fungus and the occurrence of an epizootic depend on three major factors, the host population, the pathogen population, and the environment. Most studies about 50

PAGE 51

the relationship between termites and M. anisopliae have focused on the first two factors by using a Petri dish bioassay. By giving the termites foraging distances in sand, the arena bioassay provided a simplified field-simulated environment that may partially explain the currently limited success of using M. anisopliae as a biological control agent against subterranean termites in the field (Rath, 2000). Disease may occur at an enzootic level in field colonies, but so far no natural epizootic has been observed (Milner et al., 1998b). Some studies (Yaginuma, 1990; Stenzel, 1992; Zimmermann, 1993; Vanninen, 1996; Bidochka et al., 1998; Bidochka and Small, 2005) have shown that M. anisopliae is largely present in the soil environment. Subterranean termites may be regularly exposed to this fungus, and we suggest that the long term interaction between these two groups of organisms have led to the evolution of disease resistance in subterranean termites (as listed in Chapter 1). Our results challenge the potential use of M. anisopliae as a biological control agent against subterranean termites, because the disease remained at an enzootic level in laboratory sand arenas and the fungus showed a reduced survival rate in the presence of termites. 51

PAGE 52

Table 2-1. Recovery of Metarhizium anisopliae from control sand, sand in contact with infested termites for 60 min, sand in control arenas, and sand in infested arenas Origin of sample conidia exposure a # CFU recovered b Control sand (no termites) exposed sand 1,000 49,500.00 2,260 unexposed sand 0 0.00 Contact sand sand in contact with infested termites 1,000 219.00 19 sand >3 mm away from contact to infested termites 0 0.00 Control arena tunnel 0 0.00 undisturbed sand 0 0.00 Infested arena tunnel c 4,878 3.24 0.55 undisturbed sand 0 0.00 a Theoretical number of conidia originally exposed (per g of sand) to the sample. b Number of colony forming units of Metarhizium anisopliae (Mean SE) recovered per g of sand. c The conidia exposure number is based on the estimation that 600,000 conidia were potentially in contact with only 123 g of the sand in the arena (termite tunnels). 52

PAGE 53

ABWood Introductionchamber Moistenedsand ABWood Introductionchamber Moistenedsand Figure 2-1. Foraging arena. A) The foraging arena was composed of two sheets of transparent Plexiglas (60 by 60 by 0.6 cm in thickness) separated from each other by Plexiglas laminates (5 cm in width and 0.2 cm in thickness) on the four sides, creating a 50 by 50 by 0.2 cm space inside the arena. A piece of spruce wood (10 by 8 by 0.2 cm) was placed in the arena 2 cm away from the upper-right corner. An introduction chamber (5-cm diameter by 7-cm tall) was placed in the center of the top Plexiglas sheet and was connected with the inside of the arena by four 1-cm diameter access holes. The arena was filled with 700 g of sifted sand (150-500 m sieves) and moistened with 175 ml of sterile deionized water. B) Close view of one quarter of the arena, 10 d after the release of 900 naive termites, showing galleries and termites feeding on the piece of wood. 53

PAGE 54

050100012345678Days afer inoculation% Surviva l No treatment Control 1 conidium 10 conidia 100 conidia 1,000 conidia 10,000 conidia 100,000 conidia ab b Figure 2-2. Survival distribution of groups of 20 Reticulitermes flavipes individuals after exposure of different treatments of Metarhizium anisopliae. Treatments with the same letter did not have a significant difference in their survival rate (Cox proportional regression analysis, pairwise comparisons, adjusted by Bonferroni method, =0.05). 54

PAGE 55

Figure 2-3. Grooming in Reticulitermes flavipes. Extensive grooming of some naive termites on blue individuals infested with 10,000 conidia of M. anisopliae (BEMa) 10 min after the introduction of the BEMa termites inside the foraging arena. Arrows indicate BEMa termites. 55

PAGE 56

Figure 2-4. Canibalism in Reticulitermes flavipes. Cannibalism of a dead BEMa termite by some naive termites. Arrow indicates a dead BEMa termite. 56

PAGE 57

025507510005101520Days after introduction of blue termites into Arena% Surviva l Control Infected BC BEMa Figure 2-5. Survival inside the two sets of arenas of the treated termites, in contact with 900 naive termites (not shown on this figure), monitored in the first 20 d of the experiment. BC represents the average survivorship of the groups of 60 blue termites treated with a control solution (n = 9). BEMa represents the average survivorship of the groups of 60 blue termites exposed to a suspension of Metarhizium anisopliae conidia (10,000 conidia per termite, n = 9). Cox proportional regression, n = 1080, Wald statistic = 454, df = 1, P < 0.001. 57

PAGE 58

02550751000153045607590Days after introduction of blue termites into Arena% survival NTC NTE NTC NTE Figure 2-6. Survival inside the two sets of arenas of the naive termites, in contact with 60 treated termites (not shown on this figure). NTC represents the average survivorship of the groups of 900 naive termites that were in contact with a group of 60 blue termites treated with a control solution (BC) (n = 9). NTE represents the average survivorship of the groups of 900 naive termites that were in contact with a group of 60 blue termites exposed to a Metarhizium anisopliae conidia suspension (BEMa) (n = 9). The two populations of nave termites did not have a significant difference of survivorship (Cox proportional regression, n = 14977, Wald statistic = 1.2, df = 1, P = 0.26). 58

PAGE 59

CHAPTER 3 SUSCEPTIBILITY OF SEVEN TERMITE SPECIES TO METARHIZIUM ANISOPLIAE Introduction Termites have spread and adapted to a large range of habitats resulting in high diversity of cast morphology, physiology, behavior and nesting ecology in association with a diverse microbial community (Abe et al., 2000). Due to their ecological success and diversity, theoretical questions have been formulated about how termites evolved to resist disease from a solitary life of a wood roach-like ancestor to eusociality in an environment with intense selective pressure by pathogens (Fefferman et al., 2007). The diversity of disease resistance in termites has received particular attention although it is unknown how all these factors interact with each other in natural conditions, and more importantly, how they evolved within the isopteran lineage. A comparative analysis of termite disease susceptibility relative to phylogeny could provide critical information for understanding the evolution of disease resistance in this group. However, as suggested by Fefferman et al. (2007), such an analysis is rendered difficult by the diversity of the nesting and foraging ecology that accompanied social evolution. There would be potential differences in selective pressures during adaptive radiation from a solitary wood roach-like ancestor to eusocial termites. Previously, termite susceptibility against several strains of M. anisopliae was tested for a wide range of species in different studies (Kramm et al., 1982; Lai et al., 1982; Hnel and Watson, 1983; Milner et al., 1998a; Delate et al., 1995; Zoberi, 1995; Rath and Tidbury, 1996; Rosengaus and Traniello, 1997; Myles, 2002b; Milner, 2003; Neves and Alves, 2004; Wright et al., 2005; Dong et al., 2007; Maketon et al., 2007). However, due to the variability of the fungal strain used, experimental conditions, and pathogen inoculation protocols, it is impossible to compare the results of these different studies and interpret the variability of disease susceptibly among termite species. 59

PAGE 60

In this experiment, we tested the susceptibility of seven termite species to a single strain of M. anisopliae under standardized conditions. Based on the results, we discuss the potential role of nesting ecology as a selective agent on the evolution of disease resistance mechanisms against soil fungal pathogens within the Isoptera. Material and Methods Termite Species All termites used in the experiments were collected from the laboratory colonies maintained at the Universit de Bourgogne in Dijon, France, for several years. Termite species were chosen to obtain a representative spectrum of the inferred phylogeny (Figure 3-1), the geographical origin and the nesting ecology (Table 3-1). The seven termite species used in our study (Figure 3-2) were, Mastotermes darwiniensis Froggatt (Mastotermitidae), Hodotermopsis sjoestedti (Holmgren) (Termopsidae), Hodotermes mossambicus (Hagen) (Hodotermitidae), Kalotermes flavicollis Fabr. (Kalotermitidae), R. flavipes (syn. R. santonensis, cf. Brugerolle and Bordereau, 2006) and Prorhinotermes canalifrons (Sjstedt) (Rhinotermitidae), and Nasutitermes voeltzkowi (Wasmann) (Termitidae). Due to the difficulties of obtaining and maintaining termite colonies from such a diverse geographical repartition, we were only able to perform our experiments on a single colony for each species. Termites were collected from their respective colony and kept for 1 h in groups of 20 individuals in a Petri dish with moistened filter paper at 25C with 75% relative humidity (RH) before treatments. For each species, the caste ratio (workers/soldiers) was adjusted among the 20 individuals, according to the colony of origin so as to provide all the castes that could be involved in disease resistance. The cast ratios used were, M. darwiniensis (16 workers + 4 soldiers), Hodotermopsis sjoestedti (18 workers + 2 soldiers), Hodotermes mossambicus (10 large workers + 8 small workers + 2 soldiers), K. 60

PAGE 61

flavicollis (20 workers, no soldier found), R. flavipes (19 workers + 1 soldier), P. canalifrons (18 workers + 2 soldiers) and N. voeltzkowi (16 workers + 4 soldiers). The average wet weight of each species was also measured. Conidial Suspensions and Susceptibility Test The preparation of conidial suspension was identical to that described in Chapter 2. Seven conidia concentrations in 0.1% Tween 80 aqueous solution were tested for each termite species: 0 (control), 10 3 10 4 10 5 ,10 6 10 7 and 10 8 conidia/ml. Termites were chilled to 4C for 20 min and individually treated on the dorsum with a 1 l droplet of solution to inoculate 0, 1, 10, 100, 1,000, 10,000, and 100,000 conidia per termite, respectively. Termites were then returned to 4C for 20 min in order to prevent immediate grooming. Groups of 20 treated termites that received the same treatment were placed in a Petri dish provisioned with moistened filter paper. All Petri dishes containing Hodotermes mossambicus received fragments of oven-dried hay, while other species were provided with filter paper as a food source. In addition to the seven treatments, groups of 20 termites that did not receive any treatment (no Tween 80 solution, no cold exposure = naive termites) were prepared for each species and provisioned as above. Petri dishes were stored at 25C with 75% RH for 8 d. All experiments were performed in these unique temperature and relative humidity conditions to avoid a different germination and growth rate of the fungus. Mortality was recorded and dead termites were removed daily, for a total period of 8 d. Each of eight treatments (seven treatments + naive) consisted of three replicates of 20 termites each. Statistical Analysis For each species, the effective lethal time was achieved for all concentrations on the 8 th d, mortality was corrected using Abbotts formula, and the LD 50 and their 95% fiducial limits (FL) were determined by Probit analysis (SAS Institute, 2002). The LD 50 is the median lethal dose 61

PAGE 62

required to kill 50% of the individuals in the tested population and it was considered that there were no significant differences of mortality when the 95% FL overlapped. To confirm the difference of mortality between species, a Cox proportional-hazard regression analysis (using the program R-Project for statistical computing, version 2.4; http//cran.r-project.org/) was performed to calculate the differential death rate of each species. Pairwise comparison of the death rates were adjusted by the Bonferroni Method ( = 0.05). In addition, for each species, a Cox proportional-hazard regression was performed to estimate the effect of the cold treatment and the Tween 80 solution treatment, and also to determine the effect of each conidial concentration exposure on the termite survival. Through the analysis the Wald statistic was generated and the resulting hazard function defines the instantaneous rate of death at a particular time, while controlling for the effects of other variables on survival. Results Naive Termites and Control Termites Survival Independent of the termite species tested, there were no significant differences in survival rates between the naive termites and the control termites (Wald statistic = 1.12, df = 1, P = 0.29) showing that the treatment with the Tween 80 solution combined with the time spent at 4C had no significant effect on the termite survival for all species tested. Mortality observed in the termites treated with a conidia-free 0.1% Tween 80 solution (control) was used in the Abbotts formula to correct for control mortality in Probit analysis. Effect of Termite Weight on Survival Because all the termite species had a different average biomass, multiple regression models analysis was performed to estimate the effect of average termite weight on survivorship (Figure 3-3). However, no significant correlation was found (R 2 <0.03) suggesting that disease 62

PAGE 63

susceptibility was independent of termite weight. Thus, all the LD 50 values were given on an individual number basis and not biomass. Median Lethal Dosage for each termite species The comparison of LD 50 values (Table 3-2) shows that Hodotermes mossambicus was the most susceptible (LD 50 = 22) to an infection of M. anisopliae, followed by K. flavicollis (LD 50 = 107). Reticulitermes flavipes and P. canalifrons had a moderate susceptibility, (LD 50 = 959 and LD 50 = 616 respectively), N. voeltzkowi and Hodotermopsis sjoestedti had a relatively low susceptibility to the fungal infection (LD 50 = 2,907 and LD 50 =3,307, respectively) and finally, M. darwiniensis was virtually not affected by the exposure of high concentrations of M. anisopliae (LD 50 = 256,032). Effect of Conidial Concentrations for Each Species The effect of M. anisopliae conidial concentration was analyzed separately for each species after confirming that the treatment with Tween 80 solution combined with the time spent at 4C had no significant effect on the termite mortality of all species. Mastotermes darwiniensis After controlling the effect of all other variables, the conidial concentration was a significant predictor of M. darwiniensis survival (Wald statistic = 32.5, df = 7, P < 0.001, Figure 3-4). However, pairwise comparisons among concentrations indicated that the survival of control termites was not significantly different from those of termites treated with 1, 10, 100, 1,000 and 10,000 conidia (multiple comparisons, P > 0.40). Only termites treated with 100,000 conidia were significantly different in survival, with 6.2 times the hazard ratio for death when compared with the controls (Wald statistic = 8.46, df = 1, P = 0.003). 63

PAGE 64

Hodotermopsis sjoestedti The conidial concentration of M. anisopliae was a significant predictor of Hodotermopsis sjoestedti survival (Wald statistic = 174, df = 7, P < 0.001, Figure 3-5). However, pairwise comparisons among concentrations indicated that control termites were not significantly different in survival from those of termites treated with one conidium (Wald statistic = 0.93, df = 1, P = 0.33), 10 conidia (Wald statistic = 0.56 df = 1, P = 0.57) or 100 conidia (Wald statistic = 0.34, df = 1, P = 0.49). Meanwhile, termites exposed to 1,000, 10,000, and 100,000 conidia had 13.2 (Wald statistic = 6.16, df = 1, P = 0.001), 80.7 (Wald statistic = 18.8, df = 1, P < 0.001) and 167 (Wald statistic = 25.4, df = 1, P < 0.001) times the hazard ratio of death of the controls, respectively. Hodotermes mossambicus The conidial concentration was a significant predictor of Hodotermes mossambicus survival (Wald statistic = 257, df = 7, P < 0.001, Figure 3-6). However, pairwise comparisons among concentrations indicated that the survival of control termites was not significantly different from those of termites treated with one conidium (Wald statistic = 1.6, df = 1, P = 0.30) or 10 conidia (Wald statistic = 6.94, df = 1, P = 0.008). Meanwhile, termites exposed to 100, 1,000, 10,000, and 100,000 conidia, had (Wald statistic = 6.94, df = 1, P = 0.008), 8.9 (Wald statistic = 38.4, df = 1, P < 0.001), 17 (Wald statistic = 59.1, df = 1, p < 0.001), 17.2 (Wald statistic = 59.5, df = 1, P < 0.001), and 26.6 (Wald statistic = 58, df = 1, P < 0.001) times the hazard ratio of death of the controls, respectively. Kalotermes flavicollis The conidial concentration was a significant predictor of K. flavicollis survival (Wald statistic = 250, df = 7, P < 0.001, Figure 3-7). However, pairwise comparisons among concentrations indicated that survival of control termites did not differ significantly from those 64

PAGE 65

of termites treated with one conidium (Wald statistic = 0.21, df = 1, P = 0.643), or 10 conidia (Wald statistic = 0.68, df =1, P = 0.41). Meanwhile, termites exposed to 100, 1,000, 10,000, and 100,000 conidia had 17.2 (Wald statistic = 15, df = 1, P < 0.001), 103 (Wald statistic = 39.6, df = 1, P < 0.001), 134 (Wald statistic = 42.5, df = 1, P < 0.001) and 147 (Wald statistic = 43.1, df = 1, P < 0.001) times the hazard ratio of death of the controls, respectively. Prorhinotermes canalifrons The conidial concentration was a significant predictor of P. canalifrons survival (Wald statistic = 132, df =7, P < 0.001, Figure 3-8). However, pairwise comparisons among concentrations indicated that the survival of control termites was not significantly different from those of termites treated with one conidium (Wald statistic = 0, df = 1, P = 0.995), 10 conidia (Wald statistic = 0.15, df =1, P = 0.70) or 100 conidia (Wald statistic = 3.08, df = 1, P = 0.07). Meanwhile, termites exposed to 1,000, 10,000, and 100,000 conidia had 3.8 (Wald statistic = 17.4, df = 1, P < 0.001), 5.7 (Wald statistic = 27.9, df = 1, P < 0.001), and 9.4 (Wald statistic = 47.6, df = 1, P < 0.001) times the hazard ratio of death of the controls, respectively. Reticulitermes flavipes The conidial concentration was a significant predictor of R. flavipes survival (Wald statistic = 219, df = 7, P < 0.001, Figure 3-9). However, pairwise comparisons among concentrations indicated that survival of control termites did not differ significantly from those of termites treated with one conidium (Wald statistic = 0, df = 1, P = 0.983), 10 conidia (Wald statistic = 1.39, df = 1, P = 0.238) or 100 conidia (Wald statistic = 1.38, df =1, P = 0.24). Meanwhile, termites exposed to 1,000, 10,000, and 100,000 conidia had 10.5 (Wald statistic = 19.5, df =1, P < 0.001), 32.7 (Wald statistic = 43.9, df =1, P < 0.001) and 37.4 (Wald statistic = 46.8, df = 1, P < 0.001) times the hazard ratio of death of the controls, respectively. Moreover, 65

PAGE 66

termite treated with 10,000 and 100,000 conidia had similar survival (Wald statistic = 1.43, df = 1, P < 0.23). Nasutitermes voeltzkowi The conidial concentration was a significant predictor of N. voeltzkowi survival (Wald statistic = 179, df = 7, P < 0.001, Figure 3-10). However, pairwise comparisons among concentrations indicated that survival of control termites did not differ significantly from those of termites treated with one conidium (Wald statistic = 0, df = 1, P = 0.99), or 10 conidia (Wald statistic = 0.32, df =1, P = 0.58) Meanwhile, termites exposed to 100, 1,000, 10,000, and 100,000 conidia had 12.1 (Wald statistic = 11.3, df =1, p < 0.001), 13.8 (Wald statistic = 12.6, df =1, P < 0.001), 17.1 (Wald statistic = 14.9, df = 1, P < 0.001) and 146 (Wald statistic = 43.4, df = 1, P < 0.001) times the hazard ratio of death of the controls, respectively. Comparison of Survival among Seven Termite Species Exposed to Metarhizium The following analysis compares mortalities among the seven species of individuals that were exposed to relevant concentrations of M. anisopliae (Figure 3-11). A Cox regression analysis was performed to generate the relative ratio of death among the different species exposed to the fungus (pairwise comparisons, adjusted by Bonferroni method, =0.05, significant difference P = 0.0032). Only the individuals exposed to 100, 1,000, and 10,000 conidia were used to generate the values, due to the low variability of mortality among species at 0, 1, 10, and 100,000 conidia. The analysis was therefore performed independently of the conidial concentrations. When exposed to M. anisopliae, Hodotermes mossambicus had 50.2 times the hazard ratio of death of M. darwiniensis (Wald statistic = 127, df = 1, P < 0.001), 6.6 times the hazard ratio of death of Hodotermopsis sjoestedti (Wald statistic = 142, df = 1, P < 0.001), 1.9 times the hazard ratio of death of P. canalifrons, (Wald statistic = 26.1, df = 1, P < 0.001), 1.95 times the hazard 66

PAGE 67

ratio of death of R. flavipes (Wald statistic = 25.3, df = 1, P < 0.001), and 3.68 times the hazard ratio of death of N. voeltzkowi (Wald statistic = 72.2, df = 1, P < 0.001). Hodotermes mossambicus and K. flavicollis that were exposed to M. anisopliae did no have a significantly different hazard ratio of death (Wald statistic = 2.6, df = 1, P = 0.107). When exposed to M. anisopliae, K. flavicollis had 32 times the hazard ratio of death of M. darwiniensis (Wald statistic = 101, df = 1, P < 0.001), 4.48 times the hazard ratio of death of Hodotermopsis sjoestedti (Wald statistic = 91.6, df = 1, P < 0.001), 1.56 times the hazard ratio of death of P. canalifrons, (Wald statistic = 12.1, df = 1, P < 0.001), 1.53 times the hazard ratio of death of R. flavipes (Wald statistic = 10.3, df = 1, P = 0.001), and 2.75 times the hazard ratio of death of N. voeltzkowi (Wald statistic = 15.46, df = 1, P < 0.001). When exposed to M. anisopliae, P. canalifrons had 18.2 times the hazard ratio of death of M. darwiniensis (Wald statistic = 69.6, df = 1, P < 0.001), 2.62 times the hazard ratio of death of Hodotermopsis sjoestedti (Wald statistic = 35.2, df = 1, P < 0.001), and 1.7 times the hazard ratio of death of N. voeltzkowi (Wald statistic = 11.6, df = 1, P < 0.001). Prorhinotermes canalifrons and R. flavipes that were exposed to M. anisopliae did no have a significantly different hazard ratio of death (Wald statistic = 1, df = 1, P = 0.93). When exposed to M. anisopliae, R. flavipes had 15 times the hazard ratio of death of M. darwiniensis (Wald statistic = 60.2, df = 1, P < 0.001), 2.28 times the hazard ratio of death of Hodotermopsis sjoestedti (Wald statistic = 24.7, df = 1, P < 0.001), and 1.66 times the hazard ratio of death of N. voeltzkowi (Wald statistic = 10.2, df = 1, P < 0.001). Finally, when exposed to M. anisopliae, Hodotermopsis sjoestedti had 7.1 times the hazard ratio of death of M. darwiniensis (Wald statistic = 60.2, df = 1, P < 0.001) and N. voeltzkowi had 9.4 times the hazard ratio of death of M. darwiniensis (Wald statistic = 39.9, df = 1, P < 0.001), 67

PAGE 68

but Hodotermopsis sjoestedti and N. voeltzkowi did not have a significantly different hazard ratio of death (Wald statistic = 3.11, df = 1, P = 0.077). Discussion Our study revealed some of the difficulties of standardizing comparative tests of termite susceptibility to fungal pathogens. A standard test is needed for comparison, but each species has a different habitat and may respond differently to the stress imposed by the experimental conditions. We successfully minimized the control mortality (<18%) in the seven tested species, allowing us to provide unique information about relative disease susceptibility for these species. However, due to the laboratory availability of a single termite colony of each species from diverse geographical origin, this study could not provide data about intra-specific colony variability. Rosengaus and Traniello (2001) showed the existence of variability in disease susceptibility among different colonies of Zootermopsis angusticollis (Hagen), Termopsidae. Although such variability in each of the species tested in the present study cannot be confirmed, the following interpretation was made under the assumption that the intra-species (inter-colonial) variability was lower than the observed interspecies variability. Of the seven species tested, M. darwiniensis was the most tolerant to exposure to M. anisopliae. Mastotermes darwiniensis is often considered the termite species with conserved ancestral traits that are common to wood cockroaches, Cryptocercus sp. (Klass et al., 2008). Although no data are available on Cryptocercus susceptibility to M. anisopliae, some experiments were conducted on Blattodea, and the LD 50 observed for Blattella germanica L. by Patchamuthu et al. (1999) and Quesada-Morega et al. (2004) was in the same order of magnitude as that of M. darwiniensis in this study. The reason for the low susceptibility of M. darwiniensis is unknown, but we hypothesize that it could be a form of immune competence inherited from a cockroach-like ancestor. As a matter of fact, a M. darwiniensis transferrin involved in immune 68

PAGE 69

response to infection closely resembles a cockroach transferrin (Thompson et al., 2003). Another factor that could be an important mechanism in disease resistance of M. darwiniensis involves glandular secretions. Although the chemistry of the salivary glands is poorly known in M. darwiniensis, some preliminary studies (Moore, 1968; Prestwitch, 1979a) showed the presence of quinones in soldier salivary secretions. Further studies are needed to characterize these chemicals and quantify their antifungal property in order to confirm their role in M. darwiniensis resistance to fungal infection. Although the disease resistance observed in Hodotermopsis sjoestedti was not as high as that in M. darwiniensis, our results showed that its resistance to M. anisopliae infection was higher than all other termite species tested (except for N. voeltzkowi). This resistance might be partially attributed to the nesting ecology of this species because Hodotermopsis sjoestedti is a dampwood termite that lives in a moist environment in contact with the soil. Such an environment is favorable for the natural occurrence of M. anisopliae, and the long evolutionary history of exposure of Hodotermopsis sjoestedti to the fungus may have promoted tolerance of this termite species. An adaptive immune reaction may have been conserved from an ancestral trait, or may have evolved against the fungus, i.e, an immune reaction such as the one observed in another termopsid, Z. angusticollis (Rosengaus et al., 1999b, 2007). Moreover, Rosengaus et al. (2004) showed that the sternal gland secretion of Z. angusticollis had fungistatic activity. Such secretions have yet to be reported in Hodotermopsis sjoestedti. Disease susceptibility of N. voeltzkowi was similar to that of Hodotermopsis sjoestedti, and was also highly resistant to fungal infection. However, the life habitat and nesting ecology of N. voeltzkowi are drastically different from those of the Termopsidae. Nasutitermes voeltzkowi builds arboreal nests with above-ground galleries and the potential exposure to M. anisopliae is 69

PAGE 70

probably limited. Contrary to Hodotermopsis sjoestedti, the high survivorship observed in N. voeltzkowi can not be explained by an evolutionary adaptation to a constant exposure to the fungus. However, multiple terpenoid compounds that have been described from the frontal secretion of the soldiers in the genus Nasutitermes (Prestwitch, 1979b) may contribute to its resistance. These terpenoids (originally described as components of the glue secreted by the soldiers for defense against potential enemies) may have attained a secondary antimicrobial function, because some of these chemicals have been shown to have antifungal activity (Rosengaus et al., 2000a). However, it is not known if soldiers of Nasutitermes sp. actually spray these chemicals within the nest structure. Although the frontal gland secretion in Nasutitermes could be a major factor for their high survivorship in our experiment, other factors such as antifungal peptides (Bulmer and Crozier, 2004) may be involved and Nasutitermes could be an interesting model for the evolution of alternative defense mechanisms against pathogens in termites. In the Rhinotermitidae, R. flavipes and P. canalifrons are subterranean termites that are in constant contact with soil and may be regularly exposed to M. anisopliae. As we suggested for Hodotermopsis sjoestedti, Rhinotermitidae may also have evolved adaptive defense mechanisms against soil fungi. For this reason, the susceptibility observed in the Rhinotermitidae was expected to be similar to that observed in Hodotermopsis sjoestedti, yet, R. flavipes and P. canalifrons showed moderate susceptibility to M. anisopliae when compared to the species discussed above. As suggested in Chapter 2, subterranean termites may not be at full capacity to inhibit the pathogen when in small groups and in a Petri dish. The LD 50 observed in our study may be underestimated under these standardized conditions. Although the social interactions and 70

PAGE 71

the nest architecture are important in all termite species, their roles might be particularly critical in the Rhinotermitidae. Kalotermes flavicollis is a drywood termite that infests a single piece of wood. Contact with soil fungi is unlikely and we suggest that the relative Metarhizium-free habitat in which this termite evolved had very little selective pressure to maintain a costly immune mechanism. It is therefore possible that K. flavicollis, and potentially other drywood termites species, have lost most of their disease resistance against soil pathogens during the evolutionary process. (Lenz, 2005) Hodotermes mossambicus is a species with an underground nest and forages on the surface of the ground in the East and South African region. Because it has permanent contact with the soil, we expected a similar result to that obtained with the Rhinotermitidae. Instead, Hodotermes mossambicus showed acute mortality to the fungal exposure. This unexpected result may be explained by the habitat of Hodotermes mossambicus which is a dry grass harvesting species, exceptionally well-adapted to survival under extreme climatic conditions of semi-arid grasslands (Leuthold and Bruinsma, 1977). The low moisture content of the climatic conditions in the native range of Hodotermes mossambicus may not favor the persistence of M. anisopliae in the soil or the germination and infection of insects; however, there is no available data of the presence of M. anisopliae from these regions (Bidochka and Small, 2005) to confirm our supposition. In the hypothesis that Hodotermes mossambicus evolved in a soil environment that has very low content of M. anisopliae, it is possible that this termite lost most of the mechanisms involved in resistance of soil fungal pathogens, using the same reasons given for K. flavicollis. A study about the fungal community of the semi-arid soils in South Africa would improve our understanding of Hodotermes mossambicus pathogen susceptibility. 71

PAGE 72

In conclusion, our results suggest that disease resistance is not phylogenetically consistent, although work with more than one colony for each species would be necessary to confirm our findings. Some relatively closely related species differed considerably in their disease susceptibility (i.e. Hodotermopsis sjoestedti and Hodotermes mossambicus), while distant species were similar in their disease susceptibility. We suggest that the resistance to M. anisopliae infection is an ancestral trait, inherited from a cockroach-like ancestor, which was conserved in some species, or lost in others during evolution due to differential selective pressure. However, it appears that some independent species have evolved unique defense mechanisms in addition to the ancestral mechanisms (i.e. Nasutitermes and Mastotermes, using putative chemical secretions). Contrarily, some species appear to have partially lost some of their disease resistance mechanisms (i.e., Hodotermes and Kalotermes). This could be explained by the minimal selective pressure over evolutionary time due to a habitat that was relatively Metarhizium-free. In addition, individual biomass did not appear to be a factor contributing to termite survival. Thus, body size in termite evolution was probably not pathogen-driven. Finally, because the nesting ecology appears to be important in termite survival against pathogens (Rosengaus et al., 2003; Pie et al., 2004), further studies on the evolution of disease resistance mechanisms in relation to their habitats are warranted. 72

PAGE 73

Table 3-1. Ecological characteristics of the termite species used for susceptibity test against Metarhizium anisopliae Species TC a RH a Life habitat b Nesting type b Origin c Mastotermes darwiniensis 30C 90% Subterranean Intermediate Australia Hodotermopsis sjoestedti 25C 75% Dampwood One piece Vietnam Hodotermes mossambicus 28C 40% Subterranean sub-arid Separate South Africa Kalotermes flavicollis 25C 40% Drywood One piece France Prorhinotermes canalifrons 28C 75% Subterranean Intermediate Reunion Isl. Reticulitermes flavipes 22C 75% Subterranean temperate Intermediate France Nasutitermes voeltzkowi 28C 75% Arboreal Separate Mauritius Isl. a average temperature and relative humidity (RH) used in the laboratory. b according to Abe (1987), modified. c area where the original colony was collected Table 3-2. Median lethal dose for seven termite species exposed to Metarhizium anisopliae conidia Species n a LD 50 b (95% FL) b Slope SE 2 Mastotermes darwiniensis 360 256,032 (126,3531,106,424) 1.87 0.41 20.60 Hodotermopsis sjoestedti 360 3,306 (2,079,351) 1.65 0.17 90.88 Hodotermes mossambicus 360 22 (144) 1.91 0.21 83.50 Kalotermes flavicollis 360 107 (7750) 3.19 0.43 55.07 Prorhinotermes canalifrons 360 616 (3521,090) 2.25 0.24 81.91 Reticulitermes flavipes 360 959 (6451,427) 1.17 0.11 104.57 Nasutitermes voeltzkowi 360 2,907 (1,601,303) 1.18 0.12 94.63 a n= number of termites treated with the six different concentrations of conidia. The mortality was corrected using the termites treated with a control solution of Tween 80. b LD 50 and 95% FL are expressed in number of conidia per termite 73

PAGE 74

Mastotermes darwiniensisHodotermopsis sjoestedtiHodotermes mossambicusKalotermes flavicollisReticulitermes flavipesProrhinotermes canalifronsPseudacanthotermes militarisNasutitermes voeltzkowi MaTpHoKaRhTm Figure 3-1. Simplified phylogeny of Isoptera inferred from simultaneous analysis of molecular and morphological data, modified from Inward et al. (2007b). Ma=Mastotermitidae, Tp=Termopsidae, Ho=Hodotermitidae, Ka=Kalotermitidae, Rh=Rhinotermitidae, Tm=Termitidae. 74

PAGE 75

Figure 3-2. The seven termite species tested for susceptibility against infection of Metarhizium anisopliae. 75

PAGE 76

1.E+001.E+011.E+021.E+031.E+041.E+051.E+06010203040506070 100101102103104105106LD50 (conidia/termite) LogH. mossambicusK. flavicolisP. canalifronsR. flavipesN. voeltzkowiH. sjoestedtiM. darwiniensisAverage termite weight (mg) 010203040506070 Figure 3-3. Relationship between the average termite weight of individual termites and their susceptibility to Metarhizium anisopliae exposure. No significant correlation was found. 76

PAGE 77

050100012345678Days afer inoculation% Surviva l Control 1 conidium 10 conidia 100 conidia 1,000 conidia 10,000 conidia 100,000 conidia ab Figure 3-4. Survival distribution of groups of 20 Mastotermes darwiniensis after exposure to different concentrations of Metarhizium anisopliae conidia. The survival distribution resulted from a Cox proportional regression analysis (pairwise comparisons, adjusted by Bonferroni method, =0.05, significant difference P = 0.0032). Treated groups with the same letter were not significantly different in their survival rate. 77

PAGE 78

050100012345678Days afer inoculation% Surviva l Control 1 conidium 10 conidia 100 conidia 1,000 conidia 10,000 conidia 100,000 conidia abc d Figure 3-5. Survival distribution of groups of 20 Hodotermopsis sjoestedti after exposure to different concentrations of Metarhizium anisopliae conidia. The survival distribution resulted from a Cox proportional regression analysis (pairwise comparisons, adjusted by Bonferroni method, =0.05, significant difference P = 0.0032). Treated groups with the same letter were not significantly different in their survival rate. 78

PAGE 79

050100012345678Days afer inoculation% Surviva l Control 1 conidium 10 conidia 100 conidia 1,000 conidia 10,000 conidia 100,000 conidia abc d Figure 3-6. Survival distribution of groups of 20 Hodotermes mossambicus after exposure to different concentrations of Metarhizium anisopliae conidia. The survival distribution resulted from a Cox proportional regression analysis (pairwise comparisons, adjusted by Bonferroni method, =0.05, significant difference P = 0.0032). Treated groups with the same letter were not significantly different in their survival rate. 79

PAGE 80

050100012345678Days afer inoculation% Surviva l Control 1 conidium 10 conidia 100 conidia 1,000 conidia 10,000 conidia 100,000 conidia abc Figure 3-7. Survival distribution of groups of 20 Kalotermes flavicollis after exposure to different concentrations of Metarhizium anisopliae conidia. The survival distribution resulted from a Cox proportional regression analysis (pairwise comparisons, adjusted by Bonferroni method, =0.05, significant difference P = 0.0032). Treated groups with the same letter were not significantly different in their survival rate. 80

PAGE 81

050100012345678Days afer inoculation% Surviva l Control 1 conidium 10 conidia 100 conidia 1,000 conidia 10,000 conidia 100,000 conidia ab cd Figure 3-8. Survival distribution of groups of 20 Prorhinotermes canalifrons after exposure to different concentrations of Metarhizium anisopliae conidia. The survival distribution resulted from a Cox proportional regression analysis (pairwise comparisons, adjusted by Bonferroni method, =0.05, significant difference P = 0.0032). Treated groups with the same letter were not significantly different in their survival rate. 81

PAGE 82

050100012345678Days afer inoculation% Surviva l Control 1 conidium 10 conidia 100 conidia 1,000 conidia 10,000 conidia 100,000 conidia ab b Figure 3-9. Survival distribution of groups of 20 Reticulitermes flavipes after exposure to different concentrations of Metarhizium anisopliae conidia. The survival distribution resulted from a Cox proportional regression analysis (pairwise comparisons, adjusted by Bonferroni method, =0.05, significant difference P = 0.0032). Treated groups with the same letter were not significantly different in their survival rate. 82

PAGE 83

050100012345678Days afer inoculation% Surviva l Control 1 conidium 10 conidia 100 conidia 1,000 conidia 10,000 conidia 100,000 conidia ab c Figure 3-10. Survival distribution of groups of 20 Nasutitermes voeltzkowi after exposure to different concentrations of Metarhizium anisopliae conidia. The survival distribution resulted from a Cox proportional regression analysis (pairwise comparisons, adjusted by Bonferroni method, =0.05, significant difference P = 0.0032). Treated groups with the same letter were not significantly different in their survival rate. 83

PAGE 84

050100012345678Days afer inoculation% Surviva l Mastotermes Hodotermopsis Nasutitermes Reticulitermes Prorhinotermes Kalotermes Hodotermes abcd Figure 3-11. Survival distribution of groups of 20 individuals from seven termite species after exposure to Metarhizium anisopliae (pooled data from individuals exposed to 100, 1,000 and 10,000 conidia). The survival distribution resulted from a Cox proportional regression analysis (pairwise comparisons, adjusted by Bonferroni method, =0.05, significant difference P = 0.0032). Species with the same letter were not significantly different in their survival rate. 84

PAGE 85

CHAPTER 4 INHIBITION OF METARHIZIUM ANISOPLIAE IN THE ALIMENTARY TRACT OF RETICULITERMES FLAVIPES Introduction We showed that large groups of R. flavipes in forging arenas had the ability to prevent an epizootic and, more importantly, reduce the germination rate of M. anisopliae by more than 99% under conditions resembling their natural habitat (Chapter 2). It was previously shown that termites were highly efficient in removing fungal conidia from the surface of the cuticle of their nestmates by allogrooming (Rosengaus et al., 1998b; Shimizu and Yamaji, 2003). It was also shown that ingested conidia were inhibited once they passed through the termites gut, and that the gut content had fungistatic activity against hyphal growth (Bao and Yendol, 1971; Kramm and West, 1982; Boucias et al., 1996; Rosengaus et al., 1998a; Siderhurst et al., 2005c; Yanagawa and Shimizu, 2007). The gut physiology of insects have received particular attention, especially concerning the interaction between the gut and their microbiota in their pathogenic (Dillon et al., 2005) and nonpathogenic or mutualistic relationships (Cruden and Markovetz, 1987; Nalepa et al., 2001; Dillon and Dillon, 2004). Many studies have also focused on the termites gut ecology and physiology (Breznak, 1982, 2002; Brune, 1998, 2006; Brune and Friedrich, 2000; Watanabe et al., 2003; Yang et al., 2005). Although there are several lines of evidence suggesting that among several other defense mechanisms (Cremer et al., 2007), the gut activity is an important factor in a termites defense against soil pathogens. The function of the alimentary system and its physiological basis for antifungal activity are poorly understood. A peptide secreted by the salivary glands is one potential biochemical involved in fungistatic activity in termites (Lamberty et al., 2001; Bulmer and Crozier, 2004), and fungistatic activity has been suspected to originate from gut symbionts (Siderhurst et al., 2005b). However, besides the brief observations made by 85

PAGE 86

Bao and Yendol (1971), no antifungal activity of different parts of the termites gut is described in vivo and the mode of action of the gut activity remains unclear. In this study, we describe the relationship between M. anisopliae and R. flavipes through a histopathological approach, and characterize the antifungal activity of different parts of the gut at different life stages of the fungus. Material and Methods Termite Preparation and Fungal Inoculation Termites were collected from three field colonies of R. flavipes in Fort Lauderdale, FL, USA, by using underground bucket traps (Su and Scheffrahn, 1986) containing a bundle of spruce (Picea sp.) wood. Before testing, termites were kept at 28C in 1-liter cylindrical plastic containers with pieces of moist wood. Termites were used in experiments 10 to 15 d after collection. The M. anisopliae strain used was ATCC 90448. The conidia solution preparation was made with the protocol described in Chapter 2. Termites were chilled to 4 C for 20 min, placed on a sterile glass microscope slide and then individually treated on the dorsal area with a 1 l droplet of treatment suspension of 10 7 conidia/ml, resulting from capillarity by immersing the termite in the droplet, to inoculate approximately 10,000 conidia per termite, a concentration that produced an LD 90 after 7 d for these termites with this method (see Chapter 2). Termites were then returned to 4 C for 20 min in order to prevent immediate grooming. Groups of 20 treated termites (19 workers + 1 soldier) were placed in a Petri dish (5-cm diameter and 1.5-cm high) with moistened filter paper on the bottom. Petri dishes were stored in an incubator at 27 C for 6 d, with three replicates per termite colony. Intensive allogrooming was observed shortly after termites recovered from the cold. For each colony, a control Petri dish was also prepared, using the same protocol as above, except that the controls were treated with a conidia-free solution of 0.1% Tween 80. 86

PAGE 87

Histological Preparation Termites were sampled from the Petri dishes at 3 h, 24 h, 48 h and 72 h (and daily afterward) after inoculation and prepared for fixation. Three different types of sample were taken: 1) visibly healthy individuals, 2) moribund individuals and 3) dead individuals. The discrimination between visibly healthy and moribund individuals was assessed by the reduced mobility and the characteristic apathy of the moribund termites while the rest of the healthy termites showed energetic alarm behavior when disturbed. Immediately after removing heads and legs, all samples were immersed into Bouins aqueous fixing solution (75% aqueous picric acid, 20% formaldehyde, 5% acetic acid) for at least 24 h at ambient temperature (Martoja and Martoja-Pierson, 1967). Five head capsules of moribund specimens were also fixed. Samples were then stored at 4C in the dark before histological preparation. In addition to these samples, some dead specimens were individually stored in Petri dishes containing wet filter paper at 27 C and 75% RH for a period of 1, 2 and 3 d to allow for decomposition before dissection and fixation. All fixed samples were dehydrated with successive baths of increasing concentrations of ethanol with a final bath of pure butanol, and embedded into paraffin blocks, as described by Gabe (1968). Embedded blocks were sectioned at 5 m intervals and successively stained with Periodic Acid Schiff (PAS), Hemalun and Picro-indigocarmin (modified from Gabe, 1968). Histological preparations were observed and photographs were taken under compound microscope with a digital camera (Model BX51 coupled to a DP70, Olympus Optical Co., Ltd., Tokyo, Japan). Fungal hyphae and conidia were PAS-positive. Data Collection and Analysis For all specimens, occurrence of fungal conidia was examined at the different parts of the gut (Figure 4-1) and the external cuticle, and occurrence of hyphal body was checked in the hemocoel and the internal structures of the thorax and abdomen. In all healthy specimens fixed at 87

PAGE 88

3 h, 24 h, 48 h and 72 h (n = 6 for each time of fixation), the numbers of conidia on the surface of the cuticle and in the different parts of the alimentary tract were carefully counted and were subjected to analysis of variance (ANOVA). A Tukey HSD test (post hoc, =0.05) was performed to compare the average number of conidia among the different times of fixation. Results Control Termites None of the individuals sampled from day 1 to day 6 (total n = 12) after the treatment with a control solution showed any sign of mycosis in the hemocoel or internal structures of the thorax and the abdomen through histological observation. The inspection of the foregut, midgut and hindgut showed no occurrence of hyphal bodies, ungerminated conidia or point of infection. Also, no conidia were found on the surface of the external cuticle. Visibly Healthy Individuals after Exposure to Metarhizium anisopliae Similar to the controls, visibly healthy termites that were individually treated with approximately 10,000 conidia of M. anisopliae did not show any sign of mycosis in the thorax and abdomen from day 1 to day 6 after inoculation. However, starting at day 2, areas of melanization were found just under the cuticle in all inspected individuals, implying occurrence of an immune reaction to the fungal penetration (Vey and Gtz, 1986). In the specimens fixed at 3 h after inoculation (n = 6), large numbers of conidia were found in all parts of the digestive tract. The surface of the cuticle of the thorax and the abdomen also carried a large number of conidia (Table 4-1). At 24 h after inoculation, the number of conidia significantly decreased in the foregut, the midgut and the surface of the cuticle in the inspected termites (n = 6). However the number of conidia in the hindgut was significantly higher than those of termites fixed at 3 h. For the conidia observed on the surface of the cuticle from the sample fixed at 24 h and 48 h, all were located in areas where the cuticle folded (Figure 4-2A), between two segments, or at the 88

PAGE 89

insertion of the legs, which are areas inaccessible for allogrooming by nestmates. At 48 h and 72 h after inoculation, the number of conidia in the foregut and the midgut were not different from that of specimens fixed at 24 h after inoculation. However, the number of conidia in the hindgut and the surface of the cuticle significantly decreased with time. Also, the overall conidial load of individual termites decreased with time. The conidia that were found in the foregut were mainly in contact with the esophagus, the cuticular foldings of the crop (Figure 4-2B) and the gizzard (Figure 4-2C) with no sign of germination. The conidia found in the midgut were not in direct contact with the epithelium but usually mixed with food particles within the peritrophic membrane. The conidia found in the hindgut were mainly in the paunch mixed with protozoan and food particles (Figure 4-2D), and in the colon mixed with near-digested food particles (Figure 4-2E), in specimens fixed at 24 h. However, at 72 h after inoculation, most of the conidia located in the hingut were found in the colon or the rectal valve (Figure 4-2F), apparently on their way to being excreted with fecal material. After 72 h, the number of conidia in the termites gut was marginal, with the exception of a few conidia still bound to the cuticular lining of the foregut or in the rectal valve, with no sign of germination. Moribund and Dead Individuals after Exposure to Metarhizium anisopliae Moribund individuals appeared in the Petri dishes at 48 h (n = 6) and 72 h (n = 6) after inoculation. Examinations of histological preparation of the specimens showed that ungerminated conidia were found in the alimentary tract, similarl to what was previously observed in healthy termites at the same times of fixation. A melanized area at infection points of the external cuticle was also observed. The moribund specimens mainly differed from the healthy specimen in that there was at least one penetration point by the fungus that was unsuccessfully encapsulated by the termites immune system, usually in the area of the insertion 89

PAGE 90

of the legs (arthrodial membrane), which resulted in incipient mycosis of the hemocoel (Figure 4-3A). The moribund stage of these specimens was attributed to the effect of the toxins released by the hyphae once it bypassed encapsulation by the termite and reached the hemocoel (Huxham et al., 1989). In addition, infection through the gut integument was never observed and the head capsule showed no sign of infection inside the mouth or on the maxillary and labial palps. The first dead termites were found in the Petri dishes at 48 h after inoculation (n = 3) and more were found at 72 h (n = 6). Observation of dead termites that were fixed shortly after their death (<8 h) showed little differences from that of the moribund termites through histological observation concerning the presence of ungerminated conidia in the gut, and no infection through the gut integument was observed. Several protozoans in the paunch lost their integrity, but no fungal hypha was ever found in any part of the alimentary tract. However, the hyphae deeply penetrated in the hemocoel and into some of the internal structures such as muscles, ganglia of the central nervous system, and fat body. At this stage of invasion of the fungus in the cadaver, no hypha invaded the gut epithelium or gut lumen. Metarhizium anisopliae Exposed Termites; Dead Individuals Fixed after Decomposition When specimens were left for decomposition for 1 d and 2 d post mortem, hyphae invaded all existing structures of the thorax and abdomen, including the salivary glands, the gut epithelium and the muscular structures around the alimentary tract (Figure 4-3B). However, no fungal hypha ever penetrated inside the gut lumen (Figure 4-3C). Most of the protozoans lost their structural integrity and the content of the hindgut comprised a mixture of particles from the food and the dead protozoans. When specimens were left for decomposition for 3 d post mortem, hyphae finally penetrated the gut lumen and the mycosis of the cadavers was systemic. However, the hyphae in all parts of the gut was not as dense as the hyphae in any other structure of the cadaver and the morphology of these filaments was abnormally weak and presented many 90

PAGE 91

irregularities when compared with fungal filaments in other parts of the cadaver (Figure 4-3D). After 3 d post mortem, M. anisopliae completed its life cycle and started producing a new generation of conidia on the external surface of the cadaver. Discussion Histological observations confirmed that a large quantity of conidia on the surface of the cuticle of inoculated termites were removed by grooming within 24 h and ingested by nestmates, as found in R. speratus by Shimizu and Yamaji (2003) and in C. formosanus by Yanagawa and Shimizu (2007). However, the total conidial load in individual termites 3 h after inoculation was about a third of the original inoculation (10,000 conidia). This relatively low number can be attributed to many factors during the experiment: part of the droplet could have leaked on the glass surface during the inoculation due to capillarity; head and legs were removed before fixation so all the conidia on the surface of the legs, cephalic capsule, mouth part, and inside the oral cavity and parts of the esophagus were not accounted for; and the successive washing with fixing solution and alcohols may have removed all conidia that were not yet bound to the cuticle. Although we confirmed that not all inoculated conidia were actually present on the cuticle surface early after inoculation, an identical protocol produced the LD 90 after 7 d (see Chapter 2), and therefore the overdosed solution used in the current study enabled us to observe a large number of conidia in the gut after grooming, and confirm that none of them germinated. Contrary to the observation made by Dillon and Charnley (1986a) in the desert locust Schistocera gregaria, where the conidia remained in the gut for 2.2 h on average and no more than 24 h before being excreted, conidia ingested by R. flavipes may remain in the alimentary tract for up to 72 h in individuals that were feeding on the cellulose (i.e. not starved individuals where the food passage in the gut is slower). Several thousands conidia were found throughout 91

PAGE 92

the entire alimentary tract 3 h and 24 h after inoculation, but this number substantially decreased after 48 h, and was minimal after 72 h. Our results differed slightly from Yanagawa and Shimizu (2007), who found in C. formosanus that most of the conidia were excreted within 24 h after inoculation. Therefore, in our study, a large number of conidia remained in the gut long enough for possible germination (Dillon and Charnley, 1986a). However, no germination was ever observed in any part of the alimentary tract, even for the conidia that were bound to the cuticular lining of the crop, the gizzard, and the rectal valve. Our results support the hypothesis that the termite gut possesses fungistatic properties. Although similar activity was previously shown in many insects (Dillon and Charnley, 1986a, 1995; Allee et al., 1990), including termites (Bao and Yendoll, 1971; Boucias et al., 1996; Rosengaus et al., 1998a) and the origin of this activity was suspected to be from microbial symbionts from the gut (Dillon and Charnley, 1986b, 2002; Siderhurst et al., 2005b), the observations made in the present study suggest the existence of multiple sources of fungistatic activity in the termites alimentary tract. Metarhizium anisopliae conidia have the ability to bind to the surface of the cuticular lining of the foregut. Once bound to the cuticle, conidia should produce a germinating tube in a short period of time (St. Leger, 1993). In our study, the conidia attached to the cuticular lining of the foregut never exhibited signs of germination which implies that the gut environment is not favorable for conidia germination. this could be attributed to nutrient insufficiency for germination, or to some antifungal activity. Such activity in the foregut cannot be attributed to endosymbionts, as they only occur in the hindgut, so it is likely that this inhibitory activity originates from the glandular secretions upstream. In the higher termite Pseudacanthotermes spiniger, salivary secretions contain the antifungal peptide termicin (Lamberty et al., 2001), and it was shown that related peptides were also found in Nasutitermes sp. (Bulmer and Crozier, 92

PAGE 93

2004) and Macrotermes sp. (Xu et al., 2009). If the occurrence of such peptides is an ancestral trait, then it is possible that Rhinotermitidae possess a similar compound in their salivary secretion. With such a hypothesis, the regular production of saliva and its constant flow into the alimentary tract would continuously coat the cuticular lining of the foregut with freshly synthesized fungistatic compounds. Also, Rosengaus et al. (1998a) suggested that during allogrooming, saliva and/or other glandular exudates could be applied to the cuticle of the groomed individual, which would imply that that the mouthparts of the groomers are also exposed to glandular secretions. This would reduce the ability of conidia to infect termites through the mouthparts. Observation of head capsules of exposed termites showed no sign of infection through the mouth cavity, or maxillary and labial palps which suggests that conidia have a poor chance of penetrating at these sites, although occasional infections cannot be excluded. However, the inspection of the cadavers at 1 d and 2 d post mortem showed that the hyphal bodies of M. anisopliae penetrated the salivary glands (Fig. 4-3B) similarly to all other structures (except the lumen of the alimentary tract), suggesting the absence of fungal inhibition within the salivary glands after the termite dies. This result challenges the origin of a putative antifungal compound in the foregut, or its stability. Indeed, it is unknown how long such a compound can remain active and stable after the death of the termite. It is also unknown if this activity can be disrupted by the presence of toxins resulting from the growth of the invasive hyphae (Huxham et al., 1989). The rapid fungal invasion into the salivary glands and the absence of invasion of the foregut 2 d post mortem therefore contradict the hypothesis that the presence of antifungal compounds in the foregut is only derived from salivary glands. Alternatively, such compounds may originate from different exocrine glands, such as the labral glands or the mandibular glands, 93

PAGE 94

whoses roles are still poorly known (Noirot, 1969; Miura et al., 1999). Also, the possibility of having multiple compounds secreted from different glands and acting synergistically in the foregut once they are mixed, cannot be excluded. This could explain why the salivary glands are readily invaded by the hyphal body while the antifungal activity remains in the foregut for at least 2 d post mortem. The marginal occurrence of conidia in the midgut lumen after 24 h may be explained by the absence of a cuticular lining that does not allow the conidia to attach or remain for an extended time and are flushed into the hindgut with the food flow. This may be enhanced by the presence of the peritrophic membrane. Among the several thousands of conidia observed in the hindgut of healthy termites, very few (<0.1%) were found in direct contact with the gut wall. Instead, most of them were found mixed with protozoans and food particles in the paunch and the colon. It is not known if the absence of bound conidia to the gut integument was due to the presence of chemicals inhibiting the binding process, due to competition for adhesion sites with gut microbiota, or simply due to physical constraints of the gut flow and protozoan movement. Meanwhile, the hindgut lumen was the very last structure of dead termites to be invaded by the hyphal bodies of M. anisopliae at 3 d post mortem, especially the paunch, suggesting that this structure was the site with the highest fungistatic activity. Although this activity may be partially explained by the accumulation of antifungal peptides coming from the exocrine glands around the mouthparts and being flushed into the paunch after passing through the foregut and midgut, we suggest that other factors are involved in the paunch, which result in either stability of antifungal compounds, or the addition of the same or other antifungal biochemicals. The origin of such an activity was suspected to be from the gut symbionts (Siderhurst et al., 2005b), producing norharmane, an 94

PAGE 95

active alkaloid against fungi, but it does not explain the full antifungal activity found in a termites gut (see Chapter 6). Also, the anoxic environment of the paunch (Brune et al., 1995a, 1995b; Breznak, 2000) in living termites may influence the ability of conidia to germinate, but at 2 d post mortem, with the death of the symbiotic microbiota of the gut, all H 2 and O 2 potentials in the paunch (P3) were probably disrupted and therefore, could not explain the absence of fungal invasion of the hindgut. The pH value of the gut can also influence the fungal activity, but in Reticulitermes sp., the paunch pH is 6.0 (Bignell and Eggleton, 1995, Brune et al., 1995a), and such a pH would not have much influence on M. anisopliae germination and growth according to Dillon and Charnley (1986b). We therefore suggest that the antifungal activity of the hindgut is still not fully understood and that unidentified chemicals of unknown origins may be involved in the process. The exact nature of the alimentary tract antifungal activity also remains unclear. Our observations confirmed at least the fungistatic activity of the gut in situ, but did not indicate if this activity actually killed the conidia once excreted in the fecal material or if the conidia still had the ability to germinate later in the termite habitat. Yanagawa and Shimizu (2007) indicated that M. anisopliae conidia had a reduced germination rate after the passage in the gut of C. formosanus. Rosengaus et al. (1998a) also confirmed a fungistatic activity by the feces of Z. angusticollis on M. anisopliae, and we previously showed that a minute amount of conidia (<0.1%) from the fecal material sampled in the R flavipes environment had the ability to germinate (Chapter 2). Although the fungistatic activity of the gut is confirmed, these studies suggest that a proportion of the conidia is killed during the passage through the termites alimentary tract, and therefore is partially fungicidal. 95

PAGE 96

Our findings showed that the fungistatic activity of the termites gut, in tight interaction with grooming, provides effective individual and inter-individual defense against entomopathogenic fungi such as M. anisopliae. The fungistatic activity of the gut appears to be of multiple origins and we suggest that more than one biochemical is involved. The fact that termites use their feces as a building material allows them to disperse this antifungal activity throughout the nest (Rosengaus et al., 1998a) and this activity appears to be stable and efficient in time once it is excreted by termites. Also, as shown in Chapter 2, when tested in large groups, termites that died from the fungal infection were rapidly cannibalized by their nestmates, implying that the growing hyphae of M. anisopliae in the cadavers would also be exposed to the gut activity, preventing it from finishing its life cycle. Therefore, the gut activity has an impact beyond the individual level, and provides a defense mechanism against fungal epizootics at a colony level. Although the importance of the gut physiology is now well defined in a termites defense against fungal pathogens, the chemicals and their sources involved in such defenses are still poorly known. Additional work is necessary to identify these compounds and understand their functions and interactions within the termites gut, and outside the gut once they are spread through the nest environment via fecal material. 96

PAGE 97

Table 4-1. Presence of conidia on the surface of the cuticle and in the alimentary tract of Metarhizium anisopliae exposed termites at different times of fixation. Time Number of conidia Foregut Midgut Hindgut Cuticle surface Total load 3 h 851 279a 459 130a 617 178a 1,117 440a 3,044 685a 24 h 32 21b 16 12b 1,170 713b 98 52b 1,316 734b 48 h 26 32b 8 13b 534 331a 21 14b 589 347bc 72 h 28 19b 4 2b 50 46c 0 0c 82 58c Same letters in a column indicate no significant difference in number of conidia at the different times of fixation (ANOVA, HSD post hoc, =0.05, all samples were visibly healthy termites, n = 6 per time of fixation). 97

PAGE 98

CrGMIPCoR Figure. 4-1. Digestive track of Reticulitermes flavipes in situ, dorsal view.Cr = Crop, G = Gizzard, M = Midgut, I = Ileum, P = Paunch, Co = Colon, R = Rectal valve. Modified from Noirot (1995). 98

PAGE 99

Figure 4-2. Occurrence of Metarhizium anisopliae conidia in Reticulitermes flavipes, 1 d after inoculation. Histological preparation stained with PAS-Hemalun-Picroindigocarmin. A) Conidia between two abdominal sternites. B-F) Presence of conidia in the lumen of the alimentary track: B) In contact with the integument of the foldings of the crop. C) Attached to the cuticle of the gizzard. D) Mixed with protozoans in the paunch. E) Mixed with food particle in the colon. F) In the rectal valve. = gut lumen. Arrows indicate conidia. The scale bars represent 20 m. 99

PAGE 100

Figure 4-3. Occurrence of Metarhizium anisopliae hyphae in Reticulitermes flavipes. Histological preparation stained with PAS-Hemalun-Picroindigocarmin. A) Penetration of the hyphal body in the hemocoel of a moribund termite at 2 d after inoculation. B) Invasion of the fungal filament in the body cavity of a cadaver at 2 d post mortem. C) Absence of fungal invasion in the paunch at 2 d post mortem. D) Penetration of the hindgut lumen by the fungus at 3 d post mortem. hy = hyphae, he = hemocyte, sg = salivary gland, ms = muscle, gl = ganglion of the central nervous system, fb = fat body, hgc = hindgut cuticle, ct = cuticle, = gut lumen. The scale bars represent 50 m. 100

PAGE 101

CHAPTER 5 ANTIFUNGAL ACTIVITY OF THE GUT IN SIX TERMITE SPECIES Introduction We previously demonstrated that the gut antifungal activity in R. flavipes was a major defense mechanism against the propagation of M. anisopliae in their habitat (see Chapter 4). The physiochemical conditions of the digestive tract provided inhibitory activity against the germination of fungal conidia and against the growth of fungal hyphae, and the interaction between the grooming activity and the gut fungistatic activity helped to reduce the conidial load of M. anisopliae in the termite habitat. Such mechanisms appear to be critical for avoiding epizootics in the case of massive exposure to the fungus and we hypothesized that it was selected or conserved during the evolution from a primitive cockroach ancestor to eusocial termites. However, the results observed in Chapter 3 showed that different termite species had variable mortality rates when exposed to concentrations of M. anisopliae conidia. The different susceptibilities observed were attributed to physiological adaptations to the termites habitat, and it is possible that the gut activity for defense against entomopathogens is one of these adaptations and may differ among the studied species. In the hypothesis that the gut antifungal activity was an ancestral trait and was maintained and/or reinforced during the radiation of the isopteran lineage, we should be able to observe similar activity throughout the termite phylogeny. Moreover, if such a conserved trait exists, it would also suggest that the environmental pressure during the radiation was important enough to maintain it in all studied species. Otherwise, if the antifungal gut activity varies among species, it would partially explain the difference of susceptibility previously observed. In this study, we 101

PAGE 102

looked at the antifungal gut activity of five other species through a histological approach, and we compared it to the results previously observed in R. flavipes. Material and Methods Several specimens that were used for the susceptibility test in Chapter 3 were fixed for histological preparation, as described in Chapter 4. Due to the differences of mortality among the termite species observed at 6 d after exposure to M. anisopliae, the availability for specimens was unequal among termite species. As a result, the observation of presence of conidia or fungal hyphae in the digestive tract of healthy, moribund, or dead specimens was limited depending on the species. Among all specimens prepared for histological analysis, there were 25 Hodotermopsis sjoestedti specimens (including 4 soldiers), 8 Hodotermes mossambicus specimens (including 3 soldiers), 13 K. flavicollis specimens, 10 P. canalifrons specimens, and 2 N. voeltzkowi specimens. No specimen of M. darwiniensis was fixed and limited amount of data was collected on the 2 specimens of N. voeltzkowi. The amount of conidia in the gut of the specimens was not quantified due to the absence of standard fixing time after exposure, but the presence or absence of conidia in the gut was recorded in all specimens and the presence of fungal hyphae was also examined in moribund and dead specimens. Results General Observations Histological observation showed that M. anisopliae hyphae were present in the hemocoel of all moribund and dead specimen, while healthy individual showed no sign of mycosis, as observed in R. flavipes. Ungerminated M. anisopliae conidia were found in parts of the alimentary tract of most specimens. In dead specimens, the cause of the death was confirmed to 102

PAGE 103

be from infection with M. anisopliae. The detail of the occurrence of conidia and hyphae in exposed specimens is described separately for each studied species. Observation of the digestive tract in Five Termite Species Hodotermopsis sjoestedti Similar to R. flavipes, ungerminated M. anisopliae conidia were found in the alimentary tract of Hodotermopsis sjoestedti. However, because most of the specimens were fixed after 4 d post exposure, very few conidia were found, probably due to the excretion of them along with the fecal material prior to the fixation, as shown for R. flavipes. Moribund specimens showed presence of hyphae in the hemocoel and in the muscle around the alimentary tract, but not in the gut lumen (Figure 5-1A). Shortly after the death of the termite, hyphae invaded most of the cadaver, except for the salivary glands, some part of the fat body and the gut lumen (Figure 5-1B, 1C, 1D). At 1 d post mortem, the salivary glands and the fat body were invaded (Figure 5-1E), and only after 2 d post mortem the fungal hyphae started to penetrate the gut lumen. However, the penetrating hyphae in the hindgut were weakly chitinized when compared with hyphae from the rest of the cadaver (Figure 5-1F), as previously observed in R. flavipes. Hodotermes mossambicus Ungerminated conidia of M. anisopliae were found in the alimentary tract of Hodotermes mossambicus (Figure 5-2A). Most of our findings in Hodotermes mossambicus were similar to Hodotermopsis sjoestedti, with the spread of the fungal hyphae in most of the hemocoel except for the gut lumen (Figure 5-2B, 2C) shortly after the death, penetration of the salivary glands at 1 d post mortem (Figure 5-2D), and penetration of the gut lumen after 2 d post mortem. However, we observed a unique case of fungal growth in the alimentary tract before death in a Hodotermes mossambicus soldier. This specimen was moribund and presented hyphal penetration through the external cuticle, but the fungus found in the gut was apparently contained in the foregut (Figure 103

PAGE 104

5-2E) and did not penetrate through the crop cuticle because of melanotic reaction was found at a point where the fungus attempted to penetrate the cuticular layer (Figure 5-3F). The soldier was apparently starved, as no food was found in the crop and the midgut, and few food particles were found in the hindgut, suggesting that the soldier did not have any food intake for an extended time before fixation. Kalotermes flavicollis As previously observed in other species, ungerminated conidia were found in the gut of K. flavicollis specimens, most of the muscles were invaded by M. anisopliae hyphae shortly after death (Figure 5-3A), the salivary glands were invaded starting 1 d post mortem (Figure 5-3B), and the fungus penetrated the gut lumen after 2 d post mortem. In addition to these results, we observed a fragment of external cuticle from a nestmate found in the midgut of a specimen. This fragment contained a melanized nodule that had encapsulated a fungal infection and was ingested by cannibalism, probably after the death of the nestmate. No fungal growth of M. anisopliae was observed from this encapsulated infection in the midgut lumen, however, the specimen presented some mycosis in the hemocoel, apparently from an independent infection of the external cuticle. Prorhinotermes canalifrons Ungerminated conidia of M. anisopliae were found in the alimentary tract of P. canalifons and one case of a fragment of a conidium was found to be endocytized by a protozoan present in the paunch (Figure 5-4A). The fungal hyphae invaded the hemocoel, part of the fat body and the hindgut muscles just after the death of the termite, but no penetration into the foregut, the midgut or the hindgut was observed at this point (Figure 5-4B, 4C). At 1 d post mortem, the fungus penetrated through the salivary glands (Figure 5-4D), and hyphal penetration of the hindgut lumen occurred after 2 d post mortem (Figure 5-4E, 4F). 104

PAGE 105

Nasutitermes voeltzkowi The 2 healthy specimens of N. voeltzkowi only allowed us to confirm the presence of ungerminated conidia in the gut at 3 d after exposure to M. anisopliae (Figure 5-5). No additional data concerning the growth of hyphae in any part of the specimens was collected. Discussion Overall observations allowed us to confirm that, after exposure to a solution of M. anisopliae conidia, grooming was performed and conidia were subsequently found in the alimentary tract of the groomers for all six termite species. This trait appears to be consistent in all studied species and we suggest that grooming and ingestion of microorganisms from the cuticle of nestmate is found throughout the Isoptera. Also, as none of the conidia found in the gut had germinated, it confirms that the germination inhibition activity was present in all studied species, as shown in R. flavipes. In all specimens, the fungal hyphae did not readily invade the salivary glands, but penetrated the glands only after 1 d post mortem. This observation was also made in R. flavipes and we confirmed that the salivary glands resisted to fungal infection for at least 1 d after the death of the termites. The fact that M. anisopliae hyphae did not penetrate the gut lumen until 2 d post mortem also confirms that the fungistatic activity of the digestive tract was present in all tested species and that this activity was stronger in the alimentary tract than in the salivary glands. Therefore, the antifungal activity of the gut observed in R. flavipes appears to be similar throughout the Isoptera and we suggest that such an activity was an ancestral trait that was conserved during the radiation of Isoptera. As suggested in R. flavipes, because the gut lumen was invaded at least 1 d after the salivary glands, it is possible that the total antifungal activity observed in the gut of all termite species is from different origins. However, we previously showed that the six studied termite species had a variable mortality rate when exposed to M. 105

PAGE 106

anisopliae (Chapter 3). The current results suggest that this difference of susceptibility among species was not due to variable antifungal activity of the gut of the different species, and therefore depended on other physiological factors. The observation made in K. flavicollis of a fragment of cuticle with a nodule in the alimentary tract demonstrated that the termite performed necrophagy on a nestmate that was infected by M. anisopliae, and that the passage through the gut lumen did not allow this ingested fungal particle to grow while in the alimentary tract. This observation provided direct evidence to confirm the importance of cannibalism in termites as a defense mechanism against the spread of entomopathogens in their nest. Cannibalism of an infected cadaver prevents M. anisopliae from completing its life cycle and producing a new generation of conidia, as the passage through the gut of the cannibals prevents the fungus from growing. Grooming and necrophagy in association with the gut antifungal activity provides an environment where the fungus cannot replicate and, without the possibility of replication, the chances for the fungus to produce an epizootic into the termite nest are low. As previously shown, no germinated conidia or fungal growth was observed in a healthy termite worker. The only case of fungal growth in the gut of a living termite was observed in the foregut of a starved Hodotermes mossambicus soldier. The reason for the starving condition of this soldier is that it was one of the last individuals left alive in the Petri dish at 5 d after fungal exposure, and no healthy worker was available to feed it for at least 1 d. Although this gut infection may be atypical and is an experimental artifact, we suggest that the starvation condition of the soldier may have enabled of fungal growth in the foregut. Observation of two additional non-starved soldiers of Hodotermes mossambicus and four Hodotermopsis sjoestedti soldiers suggested that they had antifungal activity in their gut, similar to what was observed in the 106

PAGE 107

workers of this species. However, the production of compounds by the salivary glands and other glands from the mouthparts of the soldiers may differ from the workers, so it is unknown if the foregut antifungal activity in soldiers is from an endogenous origin, or is exogenous, i.e. from material exchanged during the trophalaxis performed by some workers. In the hypothesis that the soldier did not produce antifungal compound in its foregut while starved, the absence of food flow in the alimentary tract may have allowed conidia of M. anisopliae acquired from a previous trophalaxis to germinate. The invasion of the gut by M. anisopliae in a starved insect was previously observed in S. gregaria (Dillon and Charnley, 1986a), and our observation suggests that food flow may also be an important factor that helped prevent infection through the alimentary tract in some termites. However, the origin of antifungal activity of the foregut in termite soldier remains unclear and needs to be addressed in further studies. To conclude, the antifungal activity in termite guts was present in all tested species from five different families and appeared to be a critical mechanism to reduce the active conidial load of fungi in the termite environment when associated with grooming and necrophagy. From a theoretical approach, the occurrence of disease in a host population depends on the three factors of the disease triangle: the susceptible host population, the pathogen and the environment (Tanada and Kaya, 1993). Because termites can directly control their environment and reduce the chances for the fungus to complete its life cycle, we suggest that termites induce a bias into the disease triangle and that epizootiology in termites cannot be approached with a classical epizootiological method. Again, this raises questions about the potential of biological control in termites. 107

PAGE 108

Figure 5-1. Occurrence of Metarhizium anisopliae in Hodotermopsis sjoestedti. A) Penetration of hyphae around the hingut muscle of a moribund termite. B) Invasion of the fungal filament in muscles just after the death of the termite. C) Invasion of the fungal filament in muscles around the enteric valve just after the death of the termite. D) Absence of penetration of the midgut lumen by the fungus at 1 d post mortem. E) Invasion of the salivary glands at 1 d post mortem. F) Hyphal penetration of the hindgut lumen at 2 d post mortem. ct = cuticle, fb = fat body, hgc = hindgut cuticle, hy = hyphae, mep = midgut epithelium, mep = malpighian tubules insertion, mt = malpighian tubule, ms = muscle, pt = protozoan, sg = salivary gland, = gut lumen. The scale bars represent 50 m. 108

PAGE 109

Figure 5-2. Occurrence of Metarhizium anisopliae in Hodotermes mossambicus. A) Presence of conidia in the paunch of a healthy termite. B) Penetration of hyphae between the midgut and hindgut. C) Invasion of the fungal filament in muscles around the midgut and ileum just after the death of the termite. D) Penetration of hyphae in the salivary glands in a dead termite. E) Growth of hyphae in the foregut of a moribund soldier. F) Melanotic reaction of the foregut cuticle. co = conidia, ct = cuticle, fb = fat body, fgc = foregut cuticle, hgc = hindgut cuticle, hy = hyphae, mel = melanotic reaction, mep = midgut epithelium, mt = malpighian tubule, ms = muscle, pt = protozoan, sd = salivary duct, sg = salivary gland, = gut lumen. The scale bars represent 50 m. 109

PAGE 110

Figure 5-3. Occurrence of Metarhizium anisopliae in Kalotermes flavicollis. A) Accumulation of hyphae around the crop but no penetration into the foregut lumen of a termite at 1 d post mortem. B) Penetration of the fungal filament into the salivary glands but not into the foregut in a termite at 1 d post mortem. C) Fragment of a cellular encapsulation from a cannibalized nestmate in the midgut lumen of a moribund termite. bl = basal lamina, ct = cuticle, fgc = foregut cuticle, hy = hyphae, mep = midgut epithelium, ms = muscle, sg = salivary gland, = gut lumen. The scale bars represent 50 m. 110

PAGE 111

Figure 5-4. Occurrence of Metarhizium anisopliae in Prorhinotermes canalifrons. A) Presence of a fragment of a conidium in the paunch of a healthy termite, apparently phagocytized by a protozoan. B-C) Invasion of hyphae into hindgut muscles just after the death of the termite, but no penetration into the hindgut. D) Invasion of the salivary glands at 1 d post mortem. E-F) Hyphal penetration of the hindgut lumen at 2 d post mortem. co = conidia, ct = cuticle, fb = fat body, hgc = hindgut cuticle, hy = hyphae, mt = malpighian tubule, ms = muscle, pt = protozoan, sg = salivary gland, = gut lumen. The scale bars represent 50 m. 111

PAGE 112

Figure 5-5. Occurrence of Metarhizium anisopliae in Nasutitermes voeltzkowi. Presence of a conidium in the cuticular foldings of the crop (foregut). co = conidia, fgc = foregut cuticle, ms = muscle, = gut lumen. The scale bars represent 50 m. 112

PAGE 113

CHAPTER 6 ANTIFUNGAL ACTIVITY OF NORHARMANE Introduction Reticulitermes sp. (Isoptera: Rhinotermitidae) are subterranean termites living in a soil environment that may be a favorable habitat for microbial growth. Factors such as temperature, humidity, and high density of termites can increase transmission potential of pathogens (Rosengaus and Traniello, 1997). Among these pathogens, fungi such as M. anisopliae and Aspergillus sp. have received particular attention, M. anisopliae is one of the candidates for biological control of subterranean termites (Milner and Staples, 1996; Culliney and Grace, 2000). However, despite multiple laboratory studies, there is limited success in using M. anisopliae as a biological control agent against subterranean termites in the field (Rath, 2000). Aspergillus sp. are generally considered to be more facultative pathogens (Boucias and Pendland, 1998) than true pathogenic fungi because they usually show low virulence against insects (Tanada and Kaya, 1993) and can infect a large variety of organisms as an opportunistic saprophyte (St. Leger et al., 2000). In laboratory ketp termites, the occurrence of Aspergillus was observed in both healthy and moribund termites (Beal and Kais, 1962; Zoberi and Grace, 1990b; Jayasimha and Henderson, 2007a), and it is common to observe sporulation of Aspergillus in different insect cadavers (Peterson et al. 2001) A range of defense mechanisms against microbial epizootics has been reported in termites, including behavioral mechanisms (Rosengaus et al., 1999a), immune mechanisms (Rosengaus et al., 1999b, 2007; Lamberty et al., 2001) and antifungal chemical production in termite nests (Rosengaus et al., 1998a, 2000; Wright et al., 2000). One such chemical is norharmane, a -carboline alkaloid, isolated from Reticulitermes sp. (Siderhurst et al., 2005a). -carboline alkaloids are widely found in plants (Allen and Holmstedt, 1980) and in some animal taxa 113

PAGE 114

(Stachel et al., 1999), and norharmane is known to be toxic against a variety of organisms (Oda et al., 1988; Quetin-Leclercq et al., 1995; Rivas et al., 1999). Siderhurst et al. (2005b) suggested that norharmane is produced by endosymbionts in termites, and that the physiological concentrations of norharmane found in the termite hemolymph (10 g ml -1 ) inhibited M. anisopliae conidial germination. It was considered a critical component in a termites fungal disease resistance (Siderhurst et al., 2005c). However, entomopathogenic fungi typically germinate on the surface of the cuticle and penetrate the cuticle into the haemocoel via a germination tube where they may release toxins to kill the host (Vilcinskas and Gtz, 1999). Siderhurst et al. (2005b) showed that norharmane is not present in the cuticle of Reticulitermes sp., therefore, measuring the effect of norharmane on conidial germination is irrelevant because the fungus encounters the chemical only in its hyphal stage after conidial germination and penetration. In the present study, we quantified the effect of norharmane on the mycelial growth of M. anisopliae and A. nomius, and we examined if norharmane has the potential to function as both a direct and indirect defense factor in a termites disease resistance. Material and Methods The M. anisopliae strain used in this study was ATCC 90448. Aspergillus conidia were collected from dead termites of a moribund laboratory group of R. flavipes. It was identified as A. nomius Kurtzman et al. by amplification of ITS region of the 18S, 5.8S and 28S rRNA genes and comparison with sequences from GenBank. A serial dilution of technical norharmane (purity >99%, Sigma-Aldrich, St. Louis, MO) in 95% ethanol was used for the test. Seven separate flasks of potato dextrose agar (PDA) were prepared, autoclaved, and cooled to 50C. In each PDA flask, 4 ml of the appropriate concentration of norharmane were added to obtain final 114

PAGE 115

concentrations of 1, 10, 100, 200, 400 and 1,000 g ml -1 The control PDA medium was made by adding 4 ml of ethanol only. The different media were poured into 9 cm sterile Petri dishes and left for 24 h in the dark at 21C. Conidia suspensions of both fungal species were spread on 1/5 PDA plates (20 conidia per plate) and incubated at 27C in the dark for 48 h for germination. For both fungal species, single-spore colonies (1 mm diameter) were transferred to the center of norharmane-amended PDA plates, with 14 replicates for each concentration of norharmane and the unamended control medium. Plates were incubated at 27C in the dark, and the diameter of each fungal colony was measured daily for 7 d. For each fungal species and each concentration of norharmane, a linear regression was performed (SAS Institute, 2002) with time (day) as the independent variable and fungal colony diameter (mm) as the dependent variable, to confirm the linear growth of the fungi. The regression slopes were the average growth rates (mm d -1 ), and were compared pairwise among concentrations by a Z-test at =0.05. The significance level was adjusted by the Bonferroni method for multiple comparisons. Results and Discussion When exposed to the physiological concentration of norharmane found in a termites haemolymph (10 g ml -1 ), M. anisopliae mycelial growth rate was significantly lower than the untreated control, but was only reduced by 11.9% (Table 5-1). Therefore, we suggest that norharmane is a minor factor in an individual termites defense against M. anisopliae infection than previously described by Siderhurst et al. (2005c), and that it is probably one of the multiple antifungal chemicals potentially involved in a non-cellular immune response (Rosengaus et al., 1998, 2000; Wright et al., 2000; Lamberty et al., 2001). 115

PAGE 116

Mycelial growth rate of A. nomius was not inhibited by the termites physiological concentration of norharmane. Aspergillus rarely infects insects under natural conditions (Boucias and Pendland, 1998) because the insect immune system is normally able to prevent successful infection (St. Leger et al., 2000). However, it appears that norharmane is not involved in a non-cellular immune response in termites against A. nomius and its growth would not be inhibited by norharmane in a saprophytic or parasitic situation when a hosts immune system was previously suppressed by other factors. In a dual infection by M. anisopliae and A. flavus in an insect, Hughes and Boomsma (2004) demonstrated that A. flavus has the potential to out-compete M. anisopliae once the insect was immuno-suppressed by the original infection of M. anisopliae. The out-competing organism needs a competitive advantage for the intra-host selection dynamics (Nowak and May, 1994). Our results demonstrated that (1) Aspergillus grows twice as fast as M. anisopliae, and (2) the presence of norharmane in the haemolymph of the termites would delay the growth of M. anisopliae in the haemocoel, allowing for Aspergillus to develop and out-compete M. anisopliae in a hypothetical case of a dual infection. As a consequence, a pathogen of low virulence would have the advantage of spreading in the colony over a highly virulent but slow growing pathogen, enhancing the chances of survival of the host colony. In conclusion, even if norharmane has a limited impact on survival of termite individuals penetrated by M. anisopliae, it may indirectly benefit the colony survival by moderating the growth of saprophytic organisms, and reducing the possibility of M. anisopliae to complete its live cycle in the termite nest. However, further studies are needed to confirm the hypothesis that norharmane can play a role in superinfections by reducing the risk of epizootics of virulent fungal pathogens in a termites nest. 116

PAGE 117

Table 6-1. Mycelial diameter growth rate (Mean SE) of Metarhizium anisopliae and Aspergillus nomius at different concentrations of norharmane (g ml -1 ) when grown in the dark at 27C. Rates (in mm d 1 ) were calculated as the slopes of linear regression lines*. Norharmane Concentration (g ml -1 ) Metarhizium anisopliae Aspergillus nomius 0 4.60 0.04 a 9.31 0.16 a 1 4.46 0.06 ab 9.30 0.16 a 10 4.05 0.05 b 9.09 0.13 a 100 3.50 0.05 c 8.45 0.18 b 200 3.26 0.08 d 7.64 0.11 c 400 0.22 0.03 e 7.02 0.06 d 1,000 0.00 0.00 f 0.00 0.00 e *Means followed by the same letter within a column are not significantly different (Z-test, =0.05 adjusted by Bonferroni method for multiple comparison). 117

PAGE 118

CHAPTER 7 CELLULAR ENCAPSULATION IN RETICULITERMES FLAVIPES Introduction A successful epizootic in an insect population depends on the host-pathogen-environment interactions and it is necessary to understand the factors responsible for insect susceptibility or resistance to a pathogen to improve the chances for insect biological control (Hajek and St. Leger, 1994). Termite defense mechanisms against fungal pathogens have attracted a great interest in recent years, especially on behavioral processes and social interactions such as repellency (Rosengaus et al., 1999a; Myles, 2002a; Wang and Powell, 2004) allogrooming (Rosengaus et al., 1998a; Shimizu and Yamaji, 2003; Yanagawa and Shimizu, 2007) and necrophagy (Jones et al., 1996). Several reviews have recently emphasized the complex interaction of defense mechanisms in social insects, including some chemical and immunological factors (Cremer et al., 2007; Feldhaar and Gross, 2008; Wilson-Rich et al., 2009). In the mean time, significant progress has been made in the general knowledge of insect immunity in the past few years (Ferrandon et al., 2007), especially in dipteran and lepidopteran insect models such as Drosophila melanogaster Meigen, Anopheles gambiae Giles, Bombyx mori L., Hyalophora cecropia L. and Manduca sexta L., where various aspects of the innate humoral and cellular immunity mechanisms against various bacterial and fungal pathogens were described at the molecular level (Gillespie et al., 1997; Trenczek, 1998; Vilcinskas and Gtz, 1999; Asano and Ashida, 2001; Bulet et al., 2002; Hoffmann and Reichhart, 2002; Hoffmann, 2003; Cerenius and Sderhll, 2004; Stanley, 2006; Feldhaar and Gross, 2008). Although the mechanisms of innate immunity are well documented in Diptera and Lepidoptera, the understanding of immunity in Isoptera remains fragmentary. Cellular encapsulation and nodule formation processes against fungal entomopathogens have been 118

PAGE 119

described in many other insects (Vey and Gtz, 1986; Gunnarsson, 1988; Lavine and Strand, 2002), however, the cellular encapsulation in Isoptera is still poorly documented. Bao and Yendol (1971) and Kramm and West (1982) studied the histopathology of entomopathogens in Reticulitermes, but did not mention the occurrence of cellular encapsulation. More surprisingly, Sajap and Kaur (1990) simply suggested the absence of cellular immunity in Coptotermes (Isoptera: Rhinotermitidae). Boucias et al. (1996) mentioned the existence of a cellular and humoral response to the invasion of the entomopathogen B. bassiana in Reticulitermes, but did not provide a detailed description. Also, Calleri et al. (2007) used the intensity of cellular encapsulation in the reproductive cast of Zootermopsis angusticollis (Isoptera: Termopsidae) to demonstrate the trade-off between cellular immunity and reproduction, but the encapsulation process was not described at a cellular level. In the present study, we histopathologically describe the relationship between M. anisopliae and R. flavipes and characterize the cellular encapsulation of fungal infections in vivo. Material and Methods Termite Inoculation and Histological Preparation Reticulitermes flavipes individuals were prepared for M. anisopliae inoculation with an identical protocol described as in Chapter 4, with a dosage of 10,000 conidia per termites. Termites were sampled at random daily from the Petri dishes and prepared for fixation. Four different sample types were taken: termites treated with a conidia-free solution of 0.1% Tween 80 (negative controls) termites treated, killed, and invaded by M. anisopliae (positive controls) termites treated with M. anisopliae with no sign of the disease (visibly healthy individuals) termites treated with M. anisopliae showing pathological symptoms (moribund individuals) 119

PAGE 120

The discrimination between visibly healthy and moribund individuals was assessed by the reduced mobility and the characteristic apathy of the moribund termites while the rest of the healthy termites showed energetic alarm behavior when disturbed. Positive controls were fungal-killed specimens that were individually stored in Petri dishes containing wet filter paper at 27 C and 75% RH for a period 3 d to allow for fungal growth, before dissection and fixation. The histological preparation and observation was identical as described in Chapter 4. Hyphae and conidia were PAS-positive (pink coloration) and melanization reactions were colored in orange-brown. For all specimens, occurrence of fungal conidia was examined on the external cuticle, and occurrence of hyphae was checked in the hemocoel and the internal structures of the thorax and abdomen. Free Circulating Hemocytes Count Due to the relatively small size of the termites, the difficulty to obtain sufficient quantity of hemolymph, and the availability of the histological sections, the relative number of free circulating hemocytes (FCH) in the hemolymph of the termites was estimated by counting the cells in ten median sagittal sections of the visibly healthy and negative control specimens. These sections, through the dorsal vessel, are usually the largest of each specimen. FCH were counted through the entire open circulatory system of the thorax and the abdomen of the specimens (n=3 per treatment and day) and hemocytes involved in an aggregation for nodule formation were not counted as FCH. A randomized block design was used, with each termite as a block, and the measured variable was the number of hemocytes per section. Results FCH Count The number of FCH found in the sections of the control termites was constant within 6 d after the treatment with a control solution (ANOVA, df = 6, F = 1.87, p = 0.08) (Table 1). For 120

PAGE 121

the M. anisopliae exposed termites, the number of FCH was significantly different in time (ANOVA, df = 6, F = 24.72, p < 0.001). Termites exposed to M. anisopliae had a significant increase of FCH within 24 h after inoculation and this number remained high for at least 3 d before decreasing to the controls levels (Tukey HSD test). This decrease of FCH corresponded with the large amount of hemocytes involved in nodule formation at 3 and 4 d after inoculation. Histopathology of Negative and Positive Controls In our histological observations, none of the individuals sampled from day 1 to day 6 after the treatment with a control solution showed any sign of mycosis in the hemocoel or internal structures of the thorax and the abdomen. No conidia were found on the surface of the external cuticle and the inspection of the integument revealed the absence of point of fungal infection, cellular encapsulation, or melanization reactions. Fungal-killed termites that were left for decomposition for 3 d had M. anisopliae hyphae invading all existing structures of the thorax and abdomen. Also, after 3 d post mortem, M. anisopliae completed its life cycle and started producing a new generation of conidia on the external surface of the cadaver (Figure 7-1A). Cellular Encapsulation in Reticulitermes flavipes Termites exposed to M. anisopliae showed occurrence of conidia on the surface of their cuticle 1, 2, and 3 d after the original inoculation. The total number of conidia found on the surface of the cuticle of these specimens after 24 h was much lower than that of the original inoculation, and never reached over 100 conidia per termite. Most of these conidia were found in cuticular folds (Figure 7-1B) which usually are areas out of reach for the allogrooming by nest mates. The first sign of melanization due to the fungus infection was found on specimens fixed on day 2, at locations where a conidium germinated. This melanization was exclusively localized in the exocuticle and appeared to originate from the cuticle itself, with apparently no direct participation of cells from the underlying epidermis or hemocytes, because the melanization 121

PAGE 122

occurred only in the exocuticle and no sign of structural or chemical modification was observed in the underlying endocuticle and epidermis. In some cases, the germinated conidia did not remain bound to the cuticle but left subtle melanization marks at the points of unsuccessful infection. In other cases, the germinating tube successfully penetrated the cuticle which resulted in the beginning of accumulation of hemocytes under the point of infection (Figure 7-1C). Usually, the hemocyte aggregation occurred before the fungal filament penetrated into the hemolymph or even reached the epidermal cells and these hemocytes were mainly granular cells. Humoral melanization within the hemocoel was also triggered and surrounded the incipient accumulation of hemocytes (Figure 7-1D). We were not able to confirm if the humoral melanization directly originated from the degranulation of the present hemocytes. At 3 d after inoculation, nodule formations had different outcomes depending on the type of sample; i.e. either visibly healthy or moribund at the moment of the fixation. In moribund specimens, several points of infection were found and at least one of them showed disorganized granular cell aggregation around fungal filaments that reached the hemocoel. In a few cases, we observed that hyphae grew quickly, before any humoral melanization occurred (Figure 7-1E). In most cases the fungal hyphae successfully penetrated the incipient nodule formation, which was poorly sclerotized at the time (Figure 7-1F). In visibly healthy termites, the degree of humoral melanization increased in the hemocoel underneath the point of infection and the original multilayered aggregation of flattened hemocytes were deeply sclerotized, to the point that these granular cells were no longer recognizable (Figure 7-2A). In all observations of healthy specimens at 3 d, several points of infection were also found, but contrary to the moribund specimens, the fungal germinating tube or hyphae were usually restricted to the cuticle, deeply embedded in the sclerotized nodule (Figure 7-2B). 122

PAGE 123

After 4 d post inoculation, most of the moribund specimens showed similar unsuccessful encapsulation when compared with the earlier description, with disorganized aggregation of granular cells and deep penetration of the hyphae into the hemocoel. Dead specimens were also examined, and we confirmed that the hyphae invaded most of the cadavers. In visibly healthy specimens, nodules were highly sclerotized (Figure 7-2C) and the size of the nodule could reach up to 100 m. However, in the majority of our observations, the nodules had a final size of 40-50 m. After 6 d post inoculation, the successful encapsulations did not recruit further hemocytes and the fungus was no longer visible at the point of infection (Figure 7-2D). Two observations at 9 d post inoculation were noted; in most instances, an epidermis-like layer of cells formed under the sclerotized nodule (Figure 7-2E); alternatively the nodule detached from the cuticle into the hemocoel and was covered with a layer of non-granular and poorly melanized hemocytes (Figure 7-2F). The overall nodule formation process was observed more than 150 times in a total of 78 specimens at different stages of the fungal infection and cellular reaction in R. flavipes. Figure 7-3 summarizes the general patterns of the encapsulation process that were described above, at the average times where these processes were observed. Discussion The histological observation in this study confirmed the existence of cellular immunity in R. flavipes against the infection of M. anisopliae. However, among all the specimens that were exposed to 10,000 conidia, 1 to 11 nodule formations were observed in individual termites and most commonly, 2 to 3 encapsulations were found per termite. This suggests that among all the potential infectious conidia present on the surface of the cuticle just after the exposure, very few successfully penetrated the cuticle (mean = 0.0002%) to pose a threat to termite survival. This 123

PAGE 124

result emphasizes the existence of defense mechanisms prior to the fungal penetration that critically reduced the chances of the conidia to successfully infect a termite. One of the major pre-penetration mechanisms is the allogrooming performed by nest mates. Several studies have emphasized the importance of the allogrooming as a disease resistance mechanism in several termite species, which results in the removal of a large quantity of conidia from the surface of the cuticle of exposed termites (Boucias et al., 1996; Rosengaus et al., 1998b; Myles, 2002b; Shimizu and Yamaji, 2003; Yanagawa and Shimizu, 2007; Yanagawa et al., 2008). In our experiment, the termites were kept in groups of 20 individuals, and were allowed to perform their allogrooming activity. In the specimens fixed 24 h after the fungal inoculation, there was a low occurrence of conidia on the surface of the cuticle and the conidia were mainly found in areas where the cuticle folds, out of reach from the groomers. This result confirms the importance of grooming for conidia removal in R. flavipes. Conidia must remain in contact with the cuticle long enough for germination and penetration to occur (Charnley, 1984) and termites such as R. flavipes prevent this mainly by allogrooming. However, it is important to mention that the low occurrence of conidia on the termite cuticle after 24 h may also be partially attributed to other mechanisms. Although our histological study cannot confirm the importance of other pre-penetration factors, we suggest that some of the observations made in other insects (St. Leger, 1991) could be applied to R. flavipes. Low humidity on the cuticle surface can be a limiting factor for conidial germination and infection (Ferron, 1977; Roberts and Humber, 1981), low exogenous nutrients on the cuticle surface can reduce the initial fungal growth (Boucias and Penland, 1984), the presence of some fatty acids on the epicuticle may have some antifungal activity (St. Leger, 1991), and the eventual deposition of saliva on the cuticle surface during the grooming activity may also provide antifungal 124

PAGE 125

properties (Lamberty et al., 2001). Our observation also showed that a melanization reaction occured in the cuticle at the point of infection and appeared to originate from the cuticle itself, with no direct participation of epidermal cells. Butt et al. (1988) reported a similar observation in a leafhopper (Hemiptera) and they attributed the melanization to the presence of prophenoloxidase in the cuticle, as shown by Andersen, 1985. It was shown in B. mori that prophenoloxydase can be synthesized in hemocytes, released into the hemolymph (Iwama and Ashida, 1986), and be actively transported into the cuticular matrix through the epidermal cells by transcytotic transportation (Ashida and Brey, 1995; Asano and Ashida, 2001). Cuticular melanization is known to play a role in the sequestration of fungal pathogens as they attempt to penetrate the cuticle, as it is a physiochemical barrier against fungal infection (Golkar et al., 1993), and it appears adaptively advantageous for an insect to recognize and induce melanotic reaction before the cuticle is damaged (St. Leger, 1991). This adaptation appears to be particularly beneficial for a subterranean termite that has been continuously exposed to soil entomopathogens during their evolutionary history. The existence of such a complex interaction of pre-penetration defense mechanisms against entomopathogens may reduce the chances for an individual conidium to successfully attach and germinate on the surface of the cuticle of the termite and penetrate into the hemocoel. This may explain why the mortality of termites due to conidia exposure is dose-dependent (Rosengaus and Traniello, 1997). One of the consequences of heavy conidia exposure of the termites in this study (10,000 conidia per termite) is that it provided enough chance for some conidia to bypass all the pre-penetration defense mechanisms, germinate, and successfully penetrate into the host hemocoel. Although we documented the relative importance of pre-penetration mechanisms for termite survival, this study mainly focused on the critical importance of cellular encapsulation, 125

PAGE 126

because this immune reaction represents the last line of defense in case all other pre-penetration mechanisms fail to prevent the fungal infection. In our experiment, R. flavipes showed an increase of hemocytes in the hemolymph within 24 h after the exposure to M. anisopliae and decreased after 3 d. Although the reduction of hemocytes after exposure to an entomopathogen is a common observation in insects (Gunnarsson, 1988; Hung and Boucias, 1992) due to their recruitment in nodule formation (Brookman et al., 1989; Gillespie et al., 2000), the original increase of hemocytes 24 h after fungal application was only previously described in Schistocerca gregaria (Forskal) (Gillespie et al., 2000). The augmentation of circulating hemocytes in the hemolymph may prepare the termite to prevent the follow-up fungal penetration by enhancing its immune capability within 3 d post exposure. Interestingly, the expression of an immune-inducible transferrin after the exposure to M. anisopliae was described in the primitive termite M. darwiniensis (Thompson et al., 2003) and such an exposure also induced the production of a protein in the hemolymph of Z. angusticollis that rendered the termites less susceptible to infection (Rosengaus et al., 1999b, 2007). We suggest that the increase of hemocytes and the production of immunocompetent proteins in termites hemolymph (as part of the humoral immunity) may have a common origin in term of cascade of activation. Although the activation of humoral and cellular immunity in an insect is the result of the molecular recognition of the entomopathogen (Gillespie et al., 1997), our observation suggests that the origin for the activation of cellular aggregation and nodule formation occurs directly in the cuticle, early after conidia germination. A similar observation was reported by Gunnarsson (1988) in S. gregaria where hemocyte aggregation was activated before the fungus reached the innermost layers of the cuticle. The microphysiological changes (chemical and physical) at the cuticular level due to fungal penetration may trigger an immune activation pathway in the 126

PAGE 127

underlying epidermal cells (Brey et al., 1993), and most likely a signal is sent to the hemolymph for recruitment of hemocytes at the infection site. If the fungus successfully reached the epidermis 2 d after exposure, the survival of the infected termite was dependent on the ability of the immune system to quickly encapsulate the foreign body. Hajek and St. Leger (1994) suggested that the outcome of an infection depends on the pathogens potential to grow rapidly, to penetrate host-induced barriers and to resist toxic chemicals. Moribund termite specimens were always found with hyphae developed in the hemocoel and with a typical absence of structured hemocyte aggregation or intense humoral melanization. Metarhizium sp. produce destruxins, toxins that inhibit cellular aggregation, alter hemocyte morphology and have immunosuppressive properties (Vey et al., 1982; Samuels et al., 1988; Huxam et al., 1989; Gupta et al., 1993; Vilcinskas et al., 1997; Kershaw et al., 1999). The prophenoloxidase cascade, occurring at the cellular and humoral levels, is responsible for melanization around the point of infection and is known to have a toxic effect against fungal entomopathogens (Gillespie et al., 1997, Wilson et al., 2001; Cerenius and Sderhll, 2004). Therefore, the success of the fungal infection depends on its ability to release the toxins into the hemolymph before encapsulation and melanization occurs, but also depends on the number of fungal penetrations at the same time. The number of encapsulations resulting from fungal infection in visibly healthy specimens was generally lower than four, while the number of encapsulations could reach up to 11 in moribund specimens. This suggests that the recruitment of hemocytes for successful nodule formation is limited and an individual termite can only encapsulate a limited number of infectious bodies in such a short period of time. In visibly healthy termites, the encapsulation process was successful and was usually completed 4-9 d after fungal exposure. The accumulation of hemocytes and the sclerotization of 127

PAGE 128

the nodule were similar to those previously described in other insects (Vey and Gtz, 1986). Hemocytes aggregated and flattened around the point of infection to form a multilayer sheath of sclerotizing cells, with the central layers being more sclerotized than the outer layers (Ratcliffe and Rowley, 1979). The formation of an epidermis-like layer at the end of the process suggests that the nodule is excluded from the integument during the next molt. Also, in a few cases where the sclerotized area was relatively small, the nodule detached from the cuticle and was found drifting in the hemocoel. We suggest that these nodules may be metabolized with time. In addition to our observations, we suggest that other humoral mechanisms can be involved in Reticulitermes immunity. Antimicrobial peptides were described in some termitid termites (Lamberty et al., 2001; Bulmer and Crozier, 2004; Xu et al., 2009) and similar peptides are considered to be a major part of the humoral immunity in Drosophila (Leclerc et al., 2006; Ferrandon et al., 2007). The existence of such peptides in rhinotermitid termites is still unknown but we hypothesize that the permanent exposure of subterranean termite to soil pathogens has probably lead to the evolution of the production of such compounds. In conclusion, our study showed that besides being a social insect with complex behavioral interactions that reduce the risk of infection by entomopathogens (Rosengaus et al., 1998b; Yanagawa and Shimizu, 2007), termites posses fully functional individual cellular immunity. Although it was shown that immunity in termites can also be enhanced by social interactions (Traniello et al., 2002), we suggest that the individual cellular immunity in a subterranean termites such as R. flavipes is still a critical factor in termite survival and one of the multiple components in their defense. Reticulitermes flavipes is consistently exposed to pathogen pressure during their underground foraging and may commonly be infected by such pathogens. The chance of an infectious agent bypassing all pre-penetration defense mechanisms is always a 128

PAGE 129

possibility. Therefore, the activation of the cellular response at an individual level is necessary before the group facilitation of disease resistance may occur (Traniello et al., 2002). Our results confirm that the use of fungal pathogens as biological control agents in subterranean termites is more complex than originally expected (Rath, 2000) and the interaction among the different defense mechanisms (i.e. behavioral, chemical, immunological) suggest that an alternative approach for biological control is necessary. The analysis of such mechanisms and interactions, and the potential ways to bypass them need to be addressed in future studies. 129

PAGE 130

Table 7-1. Average number of free circulating hemocytes (MeanSE) found in the open circulatory system of one median histological section (sagittal, 5m, n=30) in control termites and termites treated with Metarhizium anisopliae. Days after treatment Treated with a control solution Treated with M. anisopliae 0 3.10 0.24a 3.67 0.30a 1 4.10 0.39a 13.83 1.49b 2 4.20 0.32a 14.33 1.47b 3 3.37 0.23a 15.43 1.84b 4 3.93 0.33a 8.23 0.66c 5 4.00 0.29a 4.03 0.36a 6 4.17 0.36a 3.00 0.39a Identical letters in a same column indicate no significant difference (Tukey HSD test, = 0.05) 130

PAGE 131

Figure 7-1. Cellular interaction between Metarhizium anisopliae and Reticulitermes flavipes. A) Conidial germination on the surface of a cadaver. B) Presence of conidia in a fold of the cuticle of the termite. C) Fungal penetration with cuticular melanization and early hemocyte aggregation. D) Beginning of humoral melanization and encapsulation. E-F) Fungal penetration into the hemolymph in a moribund termite with poor humoral melanization. co = conidia, ct = cuticle, hc = hemocyte, hy = hypha, mel = melanization. The scale bars represent 20 m. 131

PAGE 132

Figure 7-2. Cellular interaction between Metarhizium anisopliae and Reticulitermes flavipes. A-C) Intensification of cellular aggregation and sclerotization around the fungal infection point. D-E) Completed encapsulation with formation of an epidermis-like layer. F) Detached nodule in the hemocoel. co = conidia, ct = cuticle, ep = epidermis-like, hc = hemocyte, hy = hypha, mel = melanization, nod = nodule. The scale bars represent 20 m. 132

PAGE 133

Figure 7-3. Schematized cellular encapsulation process of Metarhizium anisopliae infection in Reticulitermes flavipes. A) Conidium attachment to the insect cuticle. B) Conidium germination. C) Melanotic reaction of the cuticle and elimination of the fungal elements. D) Fungal penetration of the cuticle and incipient aggregation of hemocytes. E) Fungal penetration through the epidermis into the hemocoel and recruitment of hemocytes and incipient humoral melanization. F) Failed encapsulation and invasion of the hemocoel by the hyphae. G-H) Successful encapsulation and intensification of the sclerotization. I) Detachment of the sclerotized nodule into the hemocoel. J) Formation of an epidermis-like layer around the sclerotized nodule. 133

PAGE 134

CHAPTER 8 CELLULAR ENCAPSULATION IN SIX TERMITE SPECIES Introduction In Chapter 3, we previously showed that different termite species had a variable mortality rate when exposed to M. anisopliae. We tried to explain the differences in susceptibility by the type of habitat each termite species evolved in, but many other factors for disease resistance may be involved (Cremer et al., 2007) and specific physiological changes may have occurred during the radiation of the isopteran lineage. As previously shown in Chapter 5, the antifungal gut activity appears to be highly conserved in the termite species we tested, which suggests that the gut physiology is a critical factor in termite survival against entomopathogenic fungi. Therefore, the differential mortality among the termite species tested was probably caused by other physiological factors than the gut antifungal activity. The cellular encapsulation in Isoptera against the infection of M. anisopliae was previously described and quantified in R. flavipes (Chapter 7) as a reference model for our study. We demonstrated the importance of cellular encapsulation in this termite species as part of its defense against M. anisopliae at individual level, because it prevents the fungus from penetrating into the host and releasing the toxins into the hemocoel once it successfully bypassed any other pre-penetration mechanisms (St. Leger, 1991). Although the cellular encapsulation appears to be beneficial for the survival of an individual termite, the intensity of the nodule formation may have a physiological cost for the host due to the allocation of limited resources to the immune reaction. In R. flavipes, we showed that a healthy individual exposed to M. anisopliae had few nodules (1 to 4) formed at potential points of penetration of the fungus. In moribund specimens, we observed up to 11 nodules which suggests two possibilities for the proximal cause of the death of the termite: 134

PAGE 135

The multiple point of fungal infection increased the chance for one infective hypha to release the toxin early enough to bypass the cellular encapsulation and kill the host. Formation of multiple nodules consumed all the available resources involved for nodule formation (i.e. hemocytes, phenols, etc.) and the physiological exhaustion of the host allowed additional hyphae to invade the hemocoel. Although the first possibility may be the case for a highly virulent agent infecting the termite, the second possibility may be a better explanation with more moderate strains, where a high dosage is necessary to cause mortality, as is the case for our experiment with M. anisopliae (Chapter 3). Therefore, the chance for a given termite to survive multiple infections of M. anisopliae may partially depend on its ability to successfully encapsulate each infection at a low physiological cost. The cost of a cellular immune reaction can be estimated and quantified by the amount of hemocytes required by the host to encapsulate a penetrating hyphal body. If the formation of a single nodule requires a large quantity of resources, then the termite may only encapsulate a very limited number of penetrating hyphae at the time. The first objective of this study was to describe the cellular encapsulation in five termite species and compare these observations with what was previously observed in R. flavipes. The second objective was to estimate the physiological cost of cellular encapsulation in each termite species. Finally, the third objective was to check if this physiological cost was dependent on the termite weight or on their susceptibility to M. anisopliae. Therefore, in comparison with our observations made in R. flavipes, we estimated the relative cost of nodule formation in five other termite species in the following experiment. Material and Methods Several specimens that were used for the susceptibility test in Chapter 3 were fixed for histological preparation, as described in Chapter 4. Due to the differences of mortality among the termite species observed at 6 d after exposure to M. anisopliae, the availability for specimens 135

PAGE 136

was unequal among termite species. As a result, the observation of cellular encapsulation formation was limited depending on the species. Among all specimens prepared for histological analysis, there were 25 Hodotermopsis sjoestedti specimens (including four soldiers), eight Hodotermes mossambicus specimens (including 3 soldiers), 13 K. flavicollis specimens, ten P. canalifrons specimens, and two N. voeltzkowi specimens. In addition, histological preparations of 20 randomly selected specimens of R. flavipes used in Chapter 7 were re-used in the current experiment. For each specimen, all histological sections were observed for cellular encapsulation and each nodule was measured at its largest diameter. The relative physiological cost of cellular encapsulation in all 6 species was estimated by calculating the ratio of the average nodule size by the average weight of the species (in m / mg). The size of cellular encapsulation and the relative physiological cost of each species were subjected to analysis of variance (ANOVA). A Tukey HSD test (post hoc, =0.05) was performed to compare the averages size of encapsulation and the relative physiological cost among the six tested species. Regression analyses were also performed to look for correlations between the size of the encapsulations and two other factors: the average wet weight for each of the termites and the susceptibility to M. anisopliae. Finally, a survival trade-off value was generated by calculating the ratio of the LD 50 by the relative physiological cost. Survival trade-off estimates how efficient the physiological investment for cellular encapsulation is on the survival in each species, with high values indicating relatively high efficiency, and low values indicating low relative efficiency. Results and Discussion Nodule Formation in Six Termite Species Before fixation, the formation of nodules in the termites differed in size and intensity. The sclerotization was easily visible in Hodotermopsis sjoestedti and Kalotermes flavicollis (Figure 8-1) while a microscope was necessary to observe the nodules in R. flavipes and P. canalifrons. 136

PAGE 137

The observations of the histological preparation of termites inoculated with M. anisopliae revealed differences in the process of cellular encapsulation among the six tested species. The observations were described and compared to what was previously shown in R. flavipes. Differential size of encapsulation, relative physiological cost, and survival trade-off are summarized in Table 8-1. Reticulitermes flavipes The full description of the cellular encapsulation in R. flavipes was provided in Chapter 7. Among the 20 specimens randomly chosen from the previous study, we were able to find 49 different encapsulations of various sizes. Most of them were found in areas where the cuticle folds, or at the legs-thorax junction (Figure 8-2). The average size of the nodules was 78.51 43.51 m and the relative physiological cost was estimated at 28.77 15.94 m / mg. Hodotermopsis sjoestedti Before M. anisopliae could penetrate the procuticle, some observations allowed us to confirm that the cuticle of Hodotermopsis sjoestedti had the ability to trigger the prophenoloxydase cascade within the cuticle itself, without the involvement of hemocytes from the hemocoel. This resulted in a localized sclerotization of the cuticle (Figure 8-3A). Although this mechanism exists, it was only observed a couple of times, just like what was previously observed in R. flavipes, which suggest that this early sclerotization is not a major mechanism in a termites defense against fungal infection. Hodotermopsis sjoestedti was the largest termite tested in our study (Figure 8-3A). The cuticle thickness at some point was the thickest among all the species studied and ranged from 4 m (in the spiracle and the sclerites) to 40 m (at the articulation points or muscle insertions point). This particularity, when compared with R. flavipes, implies that the penetration peg of M. anisopliae had to go through a longer distance before reaching the hemocoel, which had two 137

PAGE 138

direct consequences. First, M. anisopliae had enough space to grow laterally in the cuticle, assimilate the nutrients obtained form the cuticle degradation, and spread between the exocuticle and the endocuticle without being in direct contact with the hemocytic activity of the hemocoel (Figure 8-3B). Second, the thickness of the cuticle provided additional time for the hemocytic reaction to take place and localize of the fungal spread (Figure 8-3C). Therefore, the large cuticle thickness in Hodotermopsis sjoestedti provided a different relationship between the cellular reaction and the infective hypha when compared with R. flavipes and the dynamic in this relationship resulted in massive aggregations of hemocytes and the large formation of large melanized nodules (Figure 8-3D, 3E). The average size of the nodules was 370.54 303.40 m and the relative physiological cost was estimated at 5.40 4.42 m / mg. The survival trade-off values for Hodotermopsis sjoestedti was the highest among all tested species, which suggests that besides having large nodules, the cellular encapsulation was efficient for survivorship at a relatively low physiological cost in this large species. In addition, we observed one case of fungal infection in a spiracle, which is the only case of fungal penetration through the respiratory tract observed through the entire study (Figure 8-3F). Hodotermes mossambicus The cellular encapsulation observed in Hodotermes mossambicus was similar to the one observed in Hodotermopsis sjoestedti in terms of size and intensity (Figure 8-4). The average size of the nodules was 223.43 127.62 m and the relative physiological cost was estimated at 3.23 1.84 m / mg. Although the relative physiological cost was low in comparison with other termite species, the resource investment in Hodotermes mossambicus was one of the least efficient because the survival trade-off value was among the lowest. As a mater of fact, most of the nodules failed to successfully encapsulate the penetrating hypha which confirmed that the large encapsulations in Hodotermes mossambicus were not efficient in preventing the infective 138

PAGE 139

hyphae from reaching the hemocoel. When comparing the survival trade-off values, the cellular encapsulation in Hodotermes mossambicus was 90 times less efficient than the cellular encapsulation observed in Hodotermopsis sjoestedti and 15 times less efficient than the cellular encapsulation observed in R. flavipes. Kalotermes flavicollis Although K. flavicollis is much smaller than Hodotermes mossambicus, their size of cellular encapsulation were similar (221.98 177.04 m for K. flavicollis) which resulted in an elevated relative physiological cost for K. flavicollis (48.73 38.87 m / mg), and was the highest relative physiological cost among all termite species tested. As observed for Hodotermes mossambicus, most of the nodule formations failed to stop the progression of the fungus into the hemocoel (Figure 8-5), so even if the cost to produce such cellular encapsulation was high, the trade-off for survival was minimal. The cellular encapsulation in K. flavicollis was 278 times less efficient than the one of Hodotermopsis sjoestedti, which suggests that the cellular encapsulation of K. flavicollis is poorly adapted to the infection of M. anisopliae. Prorhinotermes canalifrons The cellular encapsulation of Prorhinotermes canalifrons was very similar in shape, size and intensity when compared with R. flavipes (Figure 8-5), which resulted in a similar relative physiological cost and identical survival trade-off values. Nasutitermes voeltzkowi Although the low number of available specimens of N. voeltzkowi for histological preparation limited our observations for cellular encapsulation (Figure 8-7), the available data suggested that the nodule size was similar to that observed in R. flavipes and P. canalifrons, and that the resulting relative physiological costs were also not significantly different. However, due to the relatively low susceptibility of N. voeltzkowi to M. anisopliae, the survival trade-off value 139

PAGE 140

suggested that the cellular encapsulation in N. voeltzkowi was four times more efficient than the cellular encapsulation observed in R. flavipes and P. canalifrons. Relationships between the Cellular Encapsulation and other Known Factors Throughout our study, we accumulated three different types of values in six termite species. The three values obtained from empirical data were: The average wet weight of the species. The median lethal dosage (LD 50 ) of each species when exposed to concentrations of M. anisopliae. The average size of the cellular encapsulation for all species. We were able to look for correlations among these factors and we previously showed in Chapter 3 that the susceptibility was independent of the mass of the termite. With the acquisition of the cellular encapsulation size values, we can now check if this factor was dependent on the weight or the LD 50 of the termite species. The relative physiological cost The relative physiological cost value was generated by dividing the size of cellular encapsulation by the termites weight. This ratio can be interpreted as the amount of resources attributed for immunity relatively to the available resources of the termite. If this value was similar for all termite species, it would suggest that the size of encapsulation was mainly dependent on the weight of the termite. However, as shown previously, this value significantly varied from species to species, and a linear regression analysis found a weak correlation (R 2 < 0.18, F = 50.28, df = 219, P < 0.001) between the termite weight and the size of cellular encapsulation (Figure 8-8). Therefore, the size of cellular encapsulation in termites was not strongly dependent on their weight and the amount of resources allocated for nodule formation mainly depended on other factors. 140

PAGE 141

Survival trade-off The survival trade-off value was generated by dividing the LD 50 value by the relative physiological cost. The resulting ratio is considered as an indicator of relative profitability of the cellular encapsulation. A high ratio suggests that the physiological investment by a termite for cellular encapsulation was highly profitable on survival and that the resource allocation was efficient. If the trade off values were similar among the six species, it would suggest that the survivorship of a termite was dependent on the relative amount of resource allocated for encapsulation. However, as shown previously, this ratio strongly varies from species to species, and a linear regression analysis found a weak correlation (R 2 < 0.14, F = 37.68, df = 219, P < 0.001) between the termite susceptibility and the relative physiological cost for cellular encapsulation (Figure 8-9). Therefore, the susceptibility of a termite was not strongly dependent on the relative physiological cost for cellular encapsulation, and termite investment in cellular immunity had a different profitability in survivorship among all six tested species, with K. flavicollis cellular encapsulation being the least profitable and Hodotermopsis sjoestedti being the most profitable. Conclusion All studied termite species presented a cellular encapsulation reaction to the infection of the entomopathogenic fungus M. anisopliae. Most of the observations about the occurrence of such an immune reaction was similar to that previously observed in R. flavipes (Chapter 7). However, the profitability of the resource investment for nodule formation varied depending on the species. Among those species, Hodotermopsis sjoestedti was the one with the most efficient at cellular encapsulation, followed by N. voeltzkowi, R. flavipes, P. canalifrons, Hodotermes mossambicus, and finally K. flavicollis with the least efficient cellular encapsulation. 141

PAGE 142

This result confirms that the expression of an immune reaction may not reflect the actual resistance to a disease, as stated by Adamo (2004). In a major review of the immunity in bivalve mollusks, Oliver and Fisher (1999) concluded that there is insufficient evidence to support the assumption that the immune measurements such as hemocytic density and the production of cytotoxic molecules correlate with disease resistance. They concluded that characteristics measured at the suborganismal level (e.g. immune assays) may not translate into real effects in the whole organism. One reason lies with the complexity of the immune system that has many interconnected pathways capable of eliminating invaders (Apanius, 1998), and this is particularly true in social insects such as termites with their many behavioral mechanisms involved in resistance to disease. Using a large array of immune measures is necessary because it exposes the complexities of the immune function and prevents overly simplistic interpretations (Adamo, 1999) The understanding of the evolution of disease resistance in termites depends on a multimodal approach, which would include behavioral, chemical and physiological factors. 142

PAGE 143

Table 8-1. Size of cellular encapsulation in six termite species. Average Weight (mg) LD 50 (Conidia / termite) Cellular encapsulation size (m) mean SD Relative physiological cost (m / mg) mean SD Survival trade-off* Number of encapsulationsobserved Hodotermopsis sjoestedti 68.50 3,306 370.54 303.40 a 5.40 4.42 a 612.22 54 Hodotermes mossambicus 69.17 22 223.43 127.62 b 3.23 1.84 a 6.81 23 Kalotermes flavicollis 4.55 107 221.98 177.04 b 48.73 38.87 b 2.19 55 Prorhinotermes canalifrons 4.25 616 77.59 62.59 c 18.26 14,72 c 33.73 34 Reticulitermes flavipes 2.73 962 78.51 43.51c 28.77 15.94 c 33.43 49 Nasutitermes voeltzkowi 5.11 2,907 108.47 108.47 bc 21.20 8.49 c 137.00 6 *Survival trade-off estimates how efficient the physiological investment for cellular encapsulation is on the survival in each species, with high values indicating relatively high efficiency. Identical letters in the same column indicate no significant difference (ANOVA, HSD post hoc, =0.05). Average wet weight and LD 50 values are from Chapter 3. 143

PAGE 144

Figure 8-1. Visible nodule formation in two termite species. A-C) Hodotermopsis sjoestedti. D-F) Kalotermes flavicollis. 144

PAGE 145

Figure 8-2. Cellular interaction between Metarhizium anisopliae and Reticulitermes flavipes. Histological preparation stained with PAS-Hemalun-Picroindigocarmin. A) Complete encapsulation with formation of an epidermis-like layer. B-C) Point of infection between a coxa and trochanter. D) Nodule on an abdominal pleuron. E) Infection on an abdominal sternite. F) Encapsulation on a thoracic pleuron. ct = cuticle, ep = epidermis, he = hemocytes, hy = hyphae, mel = melanization, ms = muscle, Bar = 20m. 145

PAGE 146

Figure 8-3. Cellular interaction between Metarhizium anisopliae and Hodotermopsis sjoestedi. Histological preparation stained with PAS-Hemalun-Picroindigocarmin. A) Melanization induced in the thick cuticle. B-C) Hyphal growth, within the cuticle. D-E) Large melanization with massive hemocyte aggregation. F) Melanized nodule in a tracheole. ct = cuticle, fb = fat body, he = hemocytes, hy = hyphae, mel = melanization, ms = muscle, sp = spiracle, = gut lumen. Bar = 50m. 146

PAGE 147

Figure 8-4. Cellular interaction between Metarhizium anisopliae and Hodotermes mossambicus. Histological preparation stained with PAS-Hemalun-Picroindigocarmin. A) Hyphae imbedded within the mesopleuron cuticle. B) hyphae infection and nodule formation on the edge of the mesothorax. C) Fungal infection and nodule formation between two abdominal segments. D) Intense melanization at the mesopleuron area. E) Melanized nodule at an abdominal sternite folding. F) Large encapsulation with massive hemocyte recruitment. ct = cuticle, ep = epidermis, fb = fat body, he = hemocytes, hy = hyphae, mel = melanization, ms = muscle, sp = spiracle. Bar = 50m. 147

PAGE 148

Figure 8-5. Cellular interaction between Metarhizium anisopliae and Kalotermes flavicollis. Histological preparation stained with PAS-Hemalun-Picroindigocarmin. A-D) Infection at divers point of the exosqueleton. E) Large encapsulation with massive hemocyte recruitment. F) Large encapsulation with massive hemocyte recruitment in a dead indivudal. ct = cuticle, gl = glanglion of the central nervous system, he = hemocytes, hy = hyphae, mel = melanization. Bar = 50m. 148

PAGE 149

Figure 8-6. Cellular interaction between Metarhizium anisopliae and Prorhinotermes canalifrons. Histological preparation stained with PAS-Hemalun-Picroindigocarmin. A) Lateral growth of the hyphae under the epicuticle. B) Hyphal penetration through the encapsulation. C-E) Fungal encapsulation at diverse point of the cuticle. F) Failed encapsulation with proliferation of hyphae. ct = cuticle, he = hemocytes, hy = hyphae, mel = melanization, Bar = 50m. 149

PAGE 150

Figure 8-7. Cellular interaction between Metarhizium anisopliae and Nasutitermes voeltzkowi. Histological preparation stained with PAS-Hemalun-Picroindigocarmin. A-D) Hemocytes aggregation and melanization occurring at a potential fungal infection point. ct = cuticle, he = hemocytes, hy = hyphae, mel = melanization. Bar = 50m. 150

PAGE 151

010020030040050060070080090010000102030405060708 0 Reticulitermes flavipes Prorhinotermes canalifrons Nasutitermes voeltzkowi Kalotermes flavicollis Hodotermopsis sjoestedti Hodotermes mossambicus Size of cellular encapsulation (m)Average weight of the termites (mg)R2= 0.18 Figure 8-8. Relationship between the size of encapsulation of Metarhizium anisopliae infection and the average wet weight of termites. No strong correlation was found (R 2 = 0.18, F = 50.28, df = 219, P < 0.001). 151

PAGE 152

0204060801001201401601802000500100015002000250030003500 Reticulitermes flavipes Prorhinotermes canalifrons Nasutitermes voeltzkowi Kalotermes flavicollis Hodotermopsis sjoestedti Hodotermes mossambicus Relative physiological cost (m / mg)LD50 (Conidia / termite)R2= 0.14 Figure 8-9. Relationship between the relative physiological cost the cellular encapsulation of Metarhizium anisopliae infection and the susceptibility of termites. No strong correlation was found (R 2 = 0.14, F = 37.68, df = 219, P < 0.001). 152

PAGE 153

CHAPTER 9 SYNERGY AMONG DEFENSE MECHANISMS IN SUBTERRANEAN TERMITES AGAINST FUNGAL PATHOGENS Introduction The use of entomopathogens to control termite colonies started more than 40 years ago when Lund (1966) patented fungal strains as biological control agents against subterranean termites. Despite the absence of success in field applications (Culliney and Grace, 2000), the field of biological control still received great interest (Rath, 2000), mainly due to the economical importance of the termites and their status as major structural pests (Su, 2002), and because an ecologically friendly control method was an attractive concept for consumers and the pest control industry. The few cases of successful field application were limited to mound building or arboreal termite species where a large quantity of the pathogenic agent formulation was directly introduced into the central part of the nest (Hnel and Watson, 1983; Milner and Staples, 1996; Milner, 2000; Lenz, 2005). These studies employed fungi as mycoinsecticides and most of the individuals in the nest had to be exposed to a large amount of the pathogenic agent at the same time. However, the mycoinsecticide inundation did not require the secondary cycling of the pathogen (Shah and Pell, 2003), therefore, no replication of the agent was necessary to achieve colony control. Such an approach bypasses the need for an epizootic to occur and can be applied against various pest targets as long as the majority of the insect population is accessible for direct application. In subterranean termites, the inundative application of a pathogen to a nest is technically limited due to its complex tunneling patterns (King and Spink 1969, Su and Scheffrahn 1988) and the dispersion and replication of the biological control agent is necessary to reach the majority of individuals and to produce an epizootic wave. Therefore, an inundative 153

PAGE 154

approach for biological control of subterranean termites is not realistic with the current technology. The occurrence of epizootics to achieve colony mortality is therefore mandatory for a biological control approach for subterranean termites. To produce an epizootic in a subterranean termite colony, the following conditions are required: the virulent etiological agent once introduced into the termite nest has to be able to survive, disperse, replicate, transmit to all individuals of the host population, and maintain a high enough density to produce the desired lethal effect. The occurrence of such epizootics was considered to be possible by many authors (see Chapter 1) as it was assumed that the soil environment of a subterranean termite nest offers conditions highly favorable for sustaining infection and promoting epizootics. However, the absence of epizootics in field studies showed that some mechanisms prevented the etiological agent from spreading and controlling the termite colony. Defense Mechanisms in Termites The relationship between entomopathogenic fungi and termites can be described as a multilevel interaction where each defense mechanism is a barrier that hinders the fungus from advancing forward in the process of completing its life cycle and sustaining an epizootic. Behavioral Avoidance The first step for a fungus such as M. anisopliae to create an epizootic in a subterranean termite colony is to come in contact with the cuticle of the termites. Although this statement appears to be obvious, it is probably one of the major unresolved issues in the methodology of creating epizootics in a subterranean nest. Several studies showed that the presence of the fungus produces avoidance and repellency behavior in termites (Boucias et al., 1996; Rosengaus et al., 1999a; Myles, 2002a; Rath, 2000; Staples and Milner, 2000), and Lai (1977) observed that 154

PAGE 155

foraging sites of C. formosanus were vacated by the termites after the introduction of a formulation of M. anisopliae. The introduction of conidia into the nest remains problematic, because among virulent isolates of M. anisopliae, all those tested were repellent at high dosage (Milner et al. 1998a), and as we have shown in Chapters 2 and 3 of our study, a relatively high dosage is necessary to obtain a lethal effect in most tested species. Few studies tried to reduce the repellency effect by formulating the fungi with various methods, such as the use of a cellulose matrix in baits (Wang and Powell, 2004), or the mix of non-repellent strains to hide the repellent effect of the virulent strain (Milner, 2000), but the results of these approach were inconclusive. Our trap and treat approach described in Chapter 2 bypassed the repellency factor in R. flavipes by directly dosing M. anisopliae conidia on the termites and releasing them into a nave termite group. Although their presence triggered an alarm behavior in nave termites, the infested termites were able to disperse throughout the foraging arena and enter in contact with most of the individuals, which performed grooming behavior on the exposed termites once the alarm behavior faded. The dosage used was high enough to kill the directly exposed termites, but the nave termites did not suffer significant mortality from the presence of the M. anisopliae-exposed termites. Therefore, even though the repellency was avoided with this approach, the required dosage to kill the entire group of termites remains a major problem as the fungus did not replicate, nor survive, in the presence of the termites. The mycoinsectide approach in mound building termite nests (Hnel and Watson, 1983; Milner and Staples, 1996, Milner, 2000) also bypassed the repellency issue by forcing high concentrations of the fungi to contact most of the termite population, but similar applications at foraging sites of subterranean termite colonies would only expose a limited number of termites to 155

PAGE 156

the agent and would result into the avoidance of the site while the colony would dispose of the infected individuals (as demonstrated in Chapter 2). A pathogenic agent that would still be virulent and produce a high mortality at concentrations that are below the threshold level for repellency would greatly enhance the possibility for biological control in termites, but such microorganisms have yet to be identified. Chemicals in Soil Habitat Once introduced into the nest, fungal conidia have to survive and persist in the termite habitat long enough to come in contact with the external cuticle of the termites and germinate. However, several studies showed the existence of various chemicals in the soil that can inhibit the growth of fungal pathogens. These chemicals may be produced by competitive microbes (Osborne and Boucias, 1985) that could be opportunistic, commensualistic, or symbiotic in a termite nest. These chemicals may also originate from the termites themselves. Rosengaus et al., (1998a) showed that the fecal material of Z. angusticollis possessed some fungistatic activity, and we demonstrated that the fecal material of R. flavipes also inhibited M. anisopliae growth once this material was used for nest construction (Chapters 2 and 4). The frontal gland secretion of soldiers of some termite species as well as other glands such as sternal glands may also provide antifungal activity in the termites habitat (Rosengaus et al., 2000a, 2004; Wright et al., 2000; Quintana et al., 2003). The overall antifungal activity in a termite nest is probably a mixture of many different compounds from multiple origins, although their concentrations may be heterogeneous throughout the termite nest, and their potential as fungal inhibitors may vary depending on the nest structure (Pie et al., 2004) and the microniches for microbes within the termites habitat. However, observations of healthy laboratory colonies of many termite species showed no occurrence of pathogenic or opportunistic fungi in the galleries or in the nest structure (Pers. 156

PAGE 157

obs.), and Jouquet et al. (2005) showed that the microbial community inside the nest of some Macrotermitinae termites was significantly different from the soil surrounding the nest. Therefore, it appears the termites have a dynamic relationship with their habitat that may result in a sanitary environment. Grooming Behavior One of the major arguments for the use of pathogens for biological control in termites was that the high level of social interactions among the individuals of a nest would enhance the transmission and dispersion of the pathogenic agent throughout the population (which is one of the requirements for epizootics to occur). Although several studies have shown that the transmission of the etiological agent occurs during grooming activity (Kraam and West. 1982; Grace and Zoberi, 1992; Zoberi, 1995; Wright et al., 2002, Wang and Powell; 2004, ), there is evidence showing that the consequences of the grooming are beneficial to the termite colony as it reduces the pathogenic agent load from the termite habitat and reduces the survivorship of the pathogen (Zoberi and Grace, 1990a; Boucias et al., 1996; Jones et al., 1996; Rosengaus et al., 1997, 1998b, Rosengaus and Traniello, 2001; Shimizu and Yamaji 2003; Yanagawa and Shimizu, 2007). We also showed in Chapter 7 that the grooming activity successfully removed most of the M. anisopliae conidial load from the termite cuticle and prevented them from germinating and penetrating the hemocoel of an exposed individual. Moreover, we showed in Chapter 4 that the ingested conidia were inhibited by the gut activity. Although conidia were dispersed in the termite habitat with the fecal material, a minute amount of conidia retained the potential to germinate (Chapter 2). Therefore, the dispersion of the etiological agent in the termite habitat is irrelevant in such situations because its survivorship may not be good enough to allow an epizootic to occur. 157

PAGE 158

Gut chemical activity As previously shown in Chapter 4, fungal conidia that are ingested during the grooming activity of termites are inhibited by antifungal activity of the alimentary tract and their dispersion through fecal material in the termite nest resulted in a low survivorship (Chapter 2). The gut activity is a critical component of the termites defense against fungal pathogens because it allows the termite to dramatically reduce the amount of the etiological agent that can potentially infect a termite. By efficiently decreasing the overall pathogen density in their habitat, the termites may prevent an epizootic. In addition, the continuous use of fecal material as building material for nest construction may reinforce the fungal inhibition throughout the entire nest and may provide a long term protection against fungal growth. Cellular and Humoral immunity In the case that fungal conidia bypassed early stage defense mechanisms and successfully penetrated the termite cuticle, the last line of defense for an individual termite to survive is to produce an immune reaction that would prevent the growth of the fungal hyphae into the hemocoel. Lamberty et al. (2001) and Bulmer and Crozier (2004) showed the existence of antifungal peptides present into the hemolymph of some termites. Also, the expression of an immune-inducible transferrin after the exposure to M. anisopliae was described in the primitive termite M. darwiniensis (Thompson et al., 2003) and such an exposure also induced the production of a protein in the hemolymph of Z. angusticollis that rendered the termites less susceptible to infection (Rosengaus et al., 1999b, 2007). Presence of such humoral immunity in termites inhibits the fungal growth and enhances the chance of the termite to survive the infection. However, our study demonstrated the importance of the cellular encapsulation in termites as a defense mechanism against the propagation of the fungal hyphae into the host hemocoel 158

PAGE 159

(Chapter 7). The cellular encapsulation was found to be omnipresent in cases of cuticle penetration by the fungus and the sclerotized nodule resulting from the encapsulation was efficient into blocking the fungus from invading the hemocoel. However, in all species tested (Chapter 8), it was shown that the cellular encapsulation has a physiological cost and that an individual termite can only encapsulate a limited quantity of infective agent at the same time (Chapter 7). Therefore, cellular encapsulation is an efficient mechanism that prevents the death of individual termites, only if all previous defense mechanisms reduced the conidial load low enough to avoid overwhelming the termite immunity. Necrophagy and Corpse Avoidance Once termites died from the fungal infection, the surviving nestmates may dispose of the cadaver in various ways. Myles (2002a) suggested the importance of necrophagy in removing fomitic cadavers from the termite population. In our study (Chapter 2), all dead termites were cannibalized by their healthy nestmates, which imply that all the growing fungal masses were ingested and inhibited during the passage through the alimentary tract (Chapters 4 and 5). The behavioral avoidance of areas that contain cadavers is also a way for the healthy members of a colony to be physically separated from a large amount of growing fungi (Jones et al., 1996). Su (1982, 2005) showed that parts of a termite nests in C. formosanus containing a large number of dead individuals were sealed off by living individuals, and Chouvenc (2003) also described cadaver burial behavior in Pseudacanthotermes spiniger Sjestedt (Termitidae, Macrotermitinae). The consequence of avoidance behavior in termitese, burial or necrophagy is that it prevents the fungus from completing its life cycle inside the termite habitat, either by being inhibited through the alimentary tract (social stomach), or by being physically excluded away from the termite population. This prevents the fungus from replicating in the presence of the 159

PAGE 160

termites, and the absence of replication of the pathogen within the termite population is a critical factor that prevents an epizootic from occuring. Synergy among Defense Mechanisms The description of the various defense mechanisms against entomopathogens suggested that the critical requirements for an epizootic are missing in a subterranean termite nest. Throughout our study, we showed that when M. anisopliae was inoculated into a group of R. flavipes, the virulent etiological agent, although it dispersed into the population, was not able to survive, to replicate and maintain a high enough density to trigger an epizootic wave. Our study showed that an independent defense mechanism may inhibit a particular step of the fungal life cycle, but taken individually, a defense mechanism cannot efficiently provide protection for the termites at the colony level. We suggest that various defense mechanisms constantly interact with each other to produce an efficient response against fungal pathogens in a synergistically manner. We identified three major defense mechanisms, grooming behavior, cellular encapsulation, and gut activity, that may be very efficacious individually in protecting the colony from fungal diseases, but the absence of one of these three mechanisms would render the other two virtually useless and a colony could be highly susceptible to fungal infections. Role of the Grooming Behavior The grooming behavior of termites allowed R. flavipes to reduce the cuticular load of fungal conidia in individuals exposed to M. anisopliae. Therefore, the grooming behavior prevented the immune system of individual termites from being overwhelmed by numerous simultaneous infections. It also reduced the overall conidial survivorship and density in the termite nest, because the ingestion of the conidia during the grooming were exposed to the antifungal activity of the alimentary tract. The absence of grooming behavior in termites would render all termites of a colony highly susceptible to M. anisopliae infections, even at low 160

PAGE 161

concentrations because the pathogen survivorship, density, and capacity to replicate could attain levels high enough to produce an epizootic. Role of the Cellular Encapsulation As demonstrated in Chapter 7, the cellular encapsulation in R. flavipes allows individual termites to survive infection of M. anisopliae. This mechanism appears to be the last line of defense for the termite in the case that all other defense mechanisms failed in preventing the fungus from penetrating the host cuticle. We previously observed that nodule formation usually takes place in areas where the cuticle folds which are locations of the cuticle that are inaccessible from the grooming of nestmates. By combining all the data acquired from Chapter 7, we were able to map out all of the occurrences of nodule formation in 25 specimens of R. flavipes that were exposed to M. anisopliae (Figure 9-1). Out of 116 cases of nodule formation observed, 111 (96%) of them were located in areas that were inaccessible from grooming. Although this confirms how important the grooming activity was for removing most of the conidia from the termite cuticle, it demonstrates that the conidia removal was not 100%. Moreover, in the case that a conidium reaches an area where the cuticle folds, it cannot be removed from the cuticle surface by grooming. Although such a situation is unlikely to happen in natural conditions and our observations were due to forced exposure with a liquid solution, there is always a possibility for a single conidium to bind on the surface of the cuticle long enough to penetrate the host. Without the existence of an efficient cellular encapsulation mechanism (in interaction with the humoral immunity), it would only require a single conidium to infect and kill the termite. Because not all conidia can be removed by grooming, cellular encapsulation becomes a critical factor in R. flavipes survivorship when exposed to M. anisopliae. The absence of such hemocytic reaction would make the termite extremely susceptible to exposure because 161

PAGE 162

even with a low fungal dosage, the odds that a single conidium binds to the cuticle are high enough to create an infection. Role of the Antifungal Gut Activity The grooming activity in R. flavipes of termites that were exposed to M. anisopliae conidia resulted in the ingestion of thousands of conidia (Chapter 4). In the absence of antifungal activity of the alimentary tract, such a large quantity of conidia in the termite gut would expose all groomers to an internal infection by the fungus. Therefore, without the antifungal gut activity of termites, M. anisopliae would be able to disperse through the termite colony, survive and replicate by infecting the groomers or the cannibals through the cuticle of the foregut or the hindgut. The antifungal activity of the termite gut appears to be critical in the survivorship of a termite colony, and its absence from the termites defense toolbox would dramatically increase the susceptibility of the termites to infection of entomopathogens such as M. anisopliae. Efficacy of the Synergy The theoretical approach of the role of three major defense mechanisms in termite (grooming behavior, cellular encapsulation, and antifungal gut activity) suggested that, taken individually, the efficacy of each mechanism would be very low in the absence of one of the other two mechanisms. Therefore, the ability of the termites to successfully prevent fungal epizootics in the colony is hypothesized to be the result of the interaction of three mechanisms. However, because the absence of a single of these mechanisms would probably result into the death of the colony, it appears that the final efficacy of the termite in preventing epizootics is not the simple addition of the individual effects of these mechanisms, but results from a synergy among them. This type of interaction efficiently prevents the fungus from completing its life cycle inside the subterranean termite nest, although some of the other defense mechanisms discussed previously may also improve the total disease resistance of the colony. 162

PAGE 163

Conclusion: The Future of Biological Control in Subterranean Termites Defense mechanisms in subterranean termites against entomopathogenic infections protects the colony from the occurrence of disease at an epizootic level. The efficacy of preventing epizootics results from the synergy among these defense mechanisms and the potential of fungal pathogens as biological control agents against subterranean termites is therefore compromised. With the current knowledge of the termite-fungus interaction and with the available technologies, it is not realistic to control subterranean termite colonies with a classical biological control approach because the conditions in the termite habitat does not allow the fungus to survive, replicate and disperse within the population. The conditions to render biological control possible in subterranean termites would depend on the availability of a pathogenic agent that has the ability to bypass some of these defense mechanisms or to break down the synergistic effect among these mechanisms. Such a pathogenic agent would need some of the following characteristics: High virulence with low repellency. Lethal concentrations of the pathogenic agent should be able to contact most of the termite population without triggering behavioral avoidance. Resistance to the gut antifungal activity. The pathogenic agent should be able to infect and produce the disease in the gut of the termites. It should also survive the passage through the gut to facilitate the dispersion of the agent throughout the nest via the fecal material for transmission among nestmates. Escape of the immune reaction. The pathogen should infect and bypass the cellular and humoral immunity with a single infection, so a low pathogen density would still be enough to trigger the epizootic wave. Replicate and disperse before the death of the host. This is a critical factor because dead termites may be physically excluded from the colony which would prevent the secondary cycling of the pathogen within the colony. The effort that has been expended in testing multiple strains from various pathogens during the past four decades suggests that no potential candidates fulfill such a complicated set of requirements. The existence of such a pathogen has yet to be proven, whether it be a fungus, 163

PAGE 164

bacterium, nematode or virus, but such an ideal pathogenic agent may not have coevolved with termites or may simply not exist. The reason for this is that if termites were ever exposed to the pathogen during evolutionary time, it would have triggered a selection of termite colonies that have disease resistance mechanisms that prevent epizootics. Alternatively, if too virulent and successful, such a specific pathogen would have gone extinct along with its host. So far, the intense screening effort to find an efficiacious pathogen for biological control in subterranean termites has yielded poor results, and the continuation of such research may not be fruitful. An alternative to finding an ideal pathogen for subterranean termite control is the manipulation of termite colony parameters that favors the potential of biological control with already available pathogens. If the development of a technology that allows the inhibition of one or more defense mechanisms is possible, it would eventually break the synergistic activity among the different defense mechanisms of subterranean termites. For example, if the immune system of all individuals of a colony could be compromised to prevent the formation of cellular encapsulation, this might allow naturally occurring pathogens in the colony habitat to trigger an epizootic. In conclusion, in spite of all the promising studies that were done in an attempt to develop biological control in subterranean termites during the past four decades, the statement made by Lund (1971) that In field trials, no organism has demonstrated significant pathogenicity, remains unchallenged. Our study confirms that in the current state, a biological control approach is not yet a feasible control method for subterranean termites. However, the development of a technology that inhibits defense mechanisms in subterranean termites would render a biological control approach possible, with a variety of naturally occurring pathogens. 164

PAGE 165

Figure 9-1. Areas where cellular encapsulations of Metarhizium anisopliae infection occurred in Reticulitermes flavipes. Observation of 116 nodules from 25 specimens. 111 nodules occurred in areas where the cuticle folds. 165

PAGE 166

LIST OF REFERENCES Abe, T., 1987. Evolution of life types in termites. In: Evolution and Coadaptation. In: Kawano, S., Connell, J.H., Hidaka, T. (Eds.), Biotic Communities. University of Tokyo Press, Tokyo. pp. 125. Abe, T., Bignell, D.E., Higashi, M., 2000. Termites: Evolution, Sociality, Symbiosis, Ecology. Kluwer Academic Publisher, Netherlands Adamo, S.A., 2004. How should behavioral ecologists interpret measurements of immunity? Anim. Behav. 68, 1443. Al Fazairy, A.A., Hassan, F.A., 1988. Infection of termites by Spodoptera littoralis nuclear polyhedrosis virus. Insect Sci. Appl. 9, 37. Al Fazairy, A.A., Hassan, F.A., 1993. Histopathology of termite Kalotermes avicollis Fabr. infected with a nuclear polyhedrosis virus. Insect Sci. Appl. 14, 127. Allee, L.L., Goettel, M.S., Goldberg, A., Whitney, H.S., Robert, D.W., 1990. Infection by Beauveria bassiana of Leptinotarsa decemlineata larvae as a consequence of fecal contamination of the integument following per os inoculation. Mycopathologia 111, 1724. Allen, J.R.F., Holmstedt, B.R., 1980. The simple beta-carboline alkaloids. Phytochem. 19, 15731582. Almeida, J.E.M., Alves, S.B., Pereira, R.M., 1997. Selection of Beauveria spp. isolates for control of the termite Heterotermes tenuis (Hagen, 1858). J. Appl. Entomol. 121, 539. Andersen, S.O., 1985. Sclerotization and tanning of the cuticle. In: Kerkut, G.A., Gilbert, L.I. (Eds.), Comprehensive Insect Physiology, Biochemistry and Pharmacology. Pergamon Press, Oxford. Vol. 12, pp. 59. Apanius, V., 1998. Stress and immune defense. Adv. Stud. Behav. 27, 133. Arruda, W., Lubeck, I., Schrank, A., Vainstein, M.H., 2005. Morphological alterations of Metarhizium anisopliae during penetration of Boophilus microplus ticks. Exp. Appl. Acarol. 37, 231. Asano, T., Ashida, M., 2001. Cuticular prophenoloxidase of the silkworm, Bombyx mori. Purification and demonstration of its transport from the hemolymph. J. Biol. Chem. 276, 11100. Ashida, M., Brey, P.T., 1995. Role of the integument in insect defense: pro-phenol oxidase cascade in the cuticular matrix. Proc. Natl. Acad. Sci. USA 23, 10698. 166

PAGE 167

Austin, J.W., Szalanski, A.L., Scheffrahn, R.H., Messenger, M.T., Dronnet, S., Bagnres A.-G., 2005. Genetic evidence for the synonymy of two Reticulitermes species Reticulitermes flavipes (Kollar) and Reticulitermes santonensis (Feytaud). Ann. Entomol. Soc. Am. 98, 395. Beal, R.H., Kais, A.G., 1962. Apparent infection of subterranean termites by Aspergillus flavus. Insect Pathol. 4, 488. Bao L.-L., Yendol, W.G., 1971. Infection of the eastern subterranean termite, Reticulitermes flavipes (Kollar) with the fungus Beauveria bassiana (Balsamo) Vuill. Entomophaga 16, 343. Benmoussa-Haichour, D., Reversat, G., Rouland, C., 1998. Sensitivity of the dierent castes of higher termites to an infestation with entomopathogenic nematodesrole of the lipids composition. Actes Coll. Insectes Soc. 11, 45. Bidochka, M.J., Small, C.L., 2005. Phylogeography of Metarhizium, an insect pathogenic fungus. In: Vega, F.E, Blackwell, M. (Eds.), InsectFungal Associations, Ecology and Evolution. Oxford University Press, New York. pp. 28. Bidochka, M.J., Kasperski, J.E., Wild. G.A.M., 1998. Occurrence of the entomopathogenic fungi Metarhizium anisopliae and Beauveria bassiana in soils from temperate and near-northern habitats. Can. J. Bot. 76:1198-1204. Bignell, D.E., Eggleton, P., 1995. On the elevated intestinal pH of higher termites (Isoptera: Termitidae). Insect. Soc. 42, 57. Bignell, D.E., 2000. Introduction to symbiosis. In: Abe, T., Bignell, D.E., Higashi, M., 2000. Termites: Evolution, Sociality, Symbiosis, Ecology. Kluwer Academic Publisher, Netherlands. pp. 189. Boucias, D.G., Pendland, J.C., 1984. Nutritional requirements for conidial germination of several host range pathotypes of the entomopathogenic fungus Nomuraea rileyi. J. Invertebr. Pathol. 43, 288. Boucias, D.G., Pendland, J.C., 1998. Principles of insect pathology. Kluwer Academic Publisher. USA. Boucias, D.G., Stokes, C., Storey, G., Pendland, J.C., 1996. Effect of imidacloprid on the termite, Reticulitermes flavipes and its interaction with insect pathogens. Pfanzenshutz-Natrichten Bayer 49, 103. Brey, P.T., Lee ,W.J., Yamakawa, M., Koizumi, Y., Perrot, S., Franois, M., Ashida, M., 1993. Role of the integument in insect immunity: epicuticular abrasion and induction of cecropin synthesis in cuticular epithelial cells. Proc. Natl. Acad. Sci. USA 90, 6275. Breznak, J.A., 1982. Intestinal microbiota of termites and other xylophagous insects. Ann. Rev. Microbiol. 36, 323. 167

PAGE 168

Breznak, J.A., 2000. Ecology of prokaryotic microbes in the guts of wood and litter feeding termites. In: Abe, T., Bignell, D.E., Higashi, M. (Eds.),Termites: Evolution, Sociality, Symbioses, Ecology. pp. 209. Kluwer Academic Publisher, Netherlands. Breznak, J.A., 2002. Phylogenetic diversity and physiology of termite gut spirochetes. Integ. Comp. Biol. 42, 313. Bridge, P.D., Prior, C., Sagbohan, J., Lomer, C.J., Carey, M., Buddie, A., 1997. Molecular characterization of isolates of Metarhizium from locusts and grasshoppers. Biodivers. Conserv. 6, 177. Brookman, J.L., Rowley, A.F., Ratcliffe, N.A., 1989. Studies on nodule formation in Locusta following injection of microbial products. J. Invertebr. Pathol. 53, 315. Brugerolle, G., Bordereau, C., 2006. Immunological and ultrastructural characterization fo spirotrichonymphidagellates from Reticulitermes grassei and R. avipes (syn. R. santonensis), with special reference to Spirotrichonympha, Spironympha and Microjoenia. Org. Divers. Evol. 6, 109. Brune, A., 1998. Termite guts: the worlds smallest bioreactors. Trends Biotechnol. 16, 16. Brune, A., 2006. Symbiotic associations between termites and prokariotes. Prokariotes 1, 439474. Brune, A., Friedrich, M., 2000. Microecology of the termite gut: structure and function on a microscale. Curr. Opin. Microbiol. 3, 263. Brune, A., Emerson, D., Breznak, A., 1995a. The termite gut microflora as an oxygen sink: Microelectrode determination of oxygen and pH gradients in guts of lower and higher termites. Appl. Environ. Microbiol. 61, 2681. Brune, A., Miambi, E., Breznak, A., 1995b. Roles of oxygen and intestinal microflora in the metabolism of lignin-derived phenylpropanoids and other monoaromatic compounds by termites. Appl. Environ. Microbiol. 61, 2688. Bulet, P., Charlet, M., Hetru, C., 2002. Antimicrobial Peptides in Insect Immunity. In: Ezekowitz, R.A.B., Hoffmann, J.A., (Eds.), Infectious Disease: Innate Immunity. Humana Press Inc., Totowa, NJ, pp. 89. Bulmer, M., Crozier, R.H., 2004. Duplication and diversifying selection among termite antifungal peptides. Mol. Biol. Evol. 21, 2256. Butt, T.M., Wraight, S.P., Galaini-Wraight, S., Humber, R.A., Roberts, D.W., Soper, R.S., 1988. Humoral encapsulation of the fungus Erynia radicans (Entomophthorales) by the potato leaf hopper Empoasca fabae (Homoptera: Cicadellidae). J. Invertebr. Pathol. 52, 49. 168

PAGE 169

Calleri II, D.V., Rosengaus, R.B., Traniello, J.F.A., 2007. Immunity and reproduction during colony foundation in the dampwood termite, Zootermopsis angusticollis. Physiol. Entomol. 32, 136. Cerenius, L., Sderhll, K., 2004. The prophenoloxidase-activating system in invertebrates. Immunol. Rev. 198, 116. Charnley, A.K., 1984. Physiological aspects of destructive pathogenesis in insects by fungi: A speculative review. In: Anderson, J.M., Raynor, A.D.M., Walton, D.W.H. (Eds.), InvertebrateMicrobial Interactions. Cambridge University Press, Cambridge, pp. 229270. Chouvenc, T., 2003. Corpse burial behavior of the growing-fungus termite, Pseudacanthotermes spiniger (Termitidae, Macrotermitinae): Necrophoric behavior and its induction. MS thesis, Universit de Bourgogne, Dijon, France. 25p. Connick Jr. W.J., Osbrink, W.L.A., Wright, M.S., Williams, K.S., Daigle, D.J., Boykin, D.L., Lax, A.R., 2001. Increased mortality of Coptotermes formosanus (Isoptera: Rhinotermitidae) exposed to eicosanoid biosynthesis inhibitors and Serratia marcescens (Eubacteriales: Enterobacteriaceae). Environ. Entomol. 30, 449. Cremer, S., Armitage, S.A.O., Schmid-Hempel, P., 2007. Social immunity. Curr. Biol. 17, 693702. Cruden, D.L., Markovetz, A.J., 1987. Microbial ecology of the cockroach gut. Ann. Rev. Microbiol. 41, 617. Culliney, T.W., Grace, J.K., 2000. Prospects for the biological control of subterranean termites (Isoptera: Rhinotermitidae), with special reference to Coptotermes formosanus. Bull. Entomol. Res. 90, 9. Danthanarayana, W., Vitarana, S.I., 1987. Control of the live-wood tea termite Glyptotermes dilatatus using Heterorhabditis sp. (Nemat.). Agri. Ecosyst. Environ. 19, 333. Davidson, G., Chandler, D., 2005. Laboratory evaluation of entomopathogenic fungi against larvae and adults of onion maggot (Diptera: Anthomyiidae). J. Econ. Entomol. 96, 18481855. Delate, K.M., Grace, J.K., Tome, C.H.M., 1995. Potential use of pathogenic fungi in baits to control the Formosan subterranean termite (Isopt., Rhinotermitidae). J. Appl. Entomol. 119, 429. Dillon, R.J., Dillon, V.M., 2004. The gut bacteria of insects: Nonpathogenic interactions. Ann. Rev. Entomol. 49, 71. Dillon, R.J., Charnley, A.K., 1986a. Invasion of pathogenic fungus Metarhizium anisopliae through the guts of germ-free desert locusts, Schistocerca gregaria. Mycopathologia. 96, 59. 169

PAGE 170

Dillon, R.J., Charnley, A.K., 1986b. Inhibition of Metarhizium anisopliae by the gut bacterial flora of the desert locust: evidence for an antifungal toxin. J. Invertebr. Pathol. 47, 359360. Dillon, R.J., Charnley, A.K., 1995. Chemical barriers to gut infection in the desert locust: In vivo production of antimicrobial phenols associated with the bacterium Pantoea agglomerans. J. Invertebr. Pathol. 66, 72. Dillon, R.J., Charnley, A.K., 2002. Mutualism between the desert locust Schistocerca gregaria and its gut microbiota. Res. Microbiol. 153, 503. Dillon, R.J., Vennard, C.T., Buckling, A., Charnley, A.K., 2005. Diversity of locust gut bacteria protects against pathogen invasion. Ecol. Lett. 8, 1291. Dong, C., Zhang, J., Chen, W., Huang, H., Hu, Y., 2007. Characterization of a newly discovered China variety of Metarhizium anisopliae (M. anisopliae var. dcjhyium) for virulence to termites, isoenzyme, and phylogenic analysis. Microbiol. Res. 162, 53. Dong, C., Zhang, J., Huang, H., Chen, W., Hu, Y., 2009. Pathogenicity of a new China variety of Metarhizium anisopliae (M. Anisopliae var. Dcjhyium) to subterranean termite Odontotermes formosanus. Microbiol. Res. 164, 27. Edwards, R., Mill, A.E., 1986. Termites in buildings: their biology and control. East Grinstead, United Kingdom, Rentokil Ltd. Eggleton, P., 2000. Global patterns of termite diversity. In: Abe, T., Bignell, D.E., Higashi, M. (Eds.),Termites: Evolution, Sociality, Symbioses, Ecology. Kluwer Academic Publisher, Netherlands. pp. 25. Engel, M.S., Krishna, K., 2004. Familly-group names for termites (Isoptera). Am. Mus. Nov. 3432, 1. Engler K.M., Gold, R.E., 2004. Effects of multiple generations of Metarhizium anisopliae on subterranean termite feeding and mortality (Isoptera: Rhinotermitidae). Sociobiology 44, 211 Epsky, N.D., Capinera, J.L., 1988. Efcacy of the entomogenous nematode Steinernema feltiae against a subterranean termite, Reticulitermes tibialis (Isoptera: Rhinotermitidae). J. Econ. Entomol. 81, 1313. Fargues, J., Robert, P.H., Vey, A., 1976. Rle du tgument et de la dfense cellulaire des coloptres htes dans la spcifit des souches entomopathognes de Metarhizium anisopliae (Fungi imperfecti). C. R. Hebd. Seances Acad. Sci. Ser. D. 282, 2223. Fefferman, N.H., Traniello, J.F.A., Rosengaus, R.B., Calleri II, D.V., 2007. Disease prevention and resistance in social insects: modeling the survival consequences of immunity, hygienic behavior, and colony organization. Behav. Ecol. Sociobiol. 61, 565. 170

PAGE 171

Feldhaar, H., Gross, R. 2008. Immune reactions of insects on bacterial pathogens and mutualists. Microbes Infect. 10, 1082. Fernandes, P.M., Alves, S.B., 1991. Controle de Cornitermes cumulans (Kollar, 1832) (Isoptera:Termitidae) com Beauveria bassiana (Bals.) Vuill. e Metarhizium anisopliae (Metsch.) Sorokin Em condies de campo. An. Soc. Entomol. Brasil. 20, 46-49. Ferrandon, D., Imler, J.-L., Hetru, C., Hoffmann, J.A., 2007. The Drosophila systemic immune response: sensing and signaling during bacterial and fungal infections. Nature Rev. 7, 862874. Ferron, P., 1978. Biological control of insect pests by entomogenous fungi. Ann. Rev. Entomol. 23, 409. Ferron, P., 1977. Influence of relative humidity on the development of fungal infection caused by Beauveria bassiana (Fungi Imperfecti, Moniliales) in imagines of Acanthoscelides obtectus (Coleoptera: Bruchidae). Entomophaga 22, 393. Fujii, J.K., 1975. Effects of an entomogenous nematode, Neoaplectana carpocapsae Weiser, on the Formosan subterranean termite, Coptotermes formosanus Shiraki, with ecological and biological studies on C. formosanus. PhD dissertation, University of Hawaii, Honolulu, Hawaii, USA. 163 pp. Gabe, M., 1968. Techniques histologiques, Masson, Paris. Georgis, R., Poinar Jr., G.O., Wilson, A.P., 1982. Susceptibility of dampwood termites and soil and wood-dwelling termites to the entomogenous nematode Neoaplectana carpocapsae. IRCS Med. Sci. 10, 563. Gibbs, A.J., Gay, F.J., Wetherly, A.H., 1970. A possible paralysis virus of termites. Virology 40, 1063. Gillespie, J.P., Kanost, M.R., Trenczek, T., 1997. Biological mediators of insect immunity. Ann. Rev. Entomol. 42, 611. Gillespie, J.P., Bunett, C., Charnley, A.K., 2000. The immune response of the desert locust Schistocerca gregaria during mycosis of the entomopathogenic fungus, Metarhizium anisopliae var. acridum. J. Insect Physiol. 46, 429. Gitonga, W., Maniania, N.K., Eilenberg, J., Ocheil, G.R.S., 1995. Pathogenecity of indigenous enthomopathogenic fungi Metarhizium anispliae and Beauveria bassiania to Macrotermes michaelseni in Kenya. Proceedings of the 2nd Regional Workshop on Termite Research and Control, 7th 9th March, 1995, Nairobi. pp. 36-44. Golkar, L., LeBrun, R.A., Ohayon, H., Gounon, P., Papierok, B., Brey, P.T., 1993. Variation of larval susceptibility to Lagenidium giganteum in three mosquito species. J. Invertebr. Pathol. 62, 1. 171

PAGE 172

Gouger, R.J., Kimbrough, J.W., 1969. Antennopsis gallica Heim and Buchli (Hyphomycetes: Gloeohaustoriales), an entomogenous fungus on subterranean termites in Florida. J. Invertebr. Pathol. 13, 223. Grace, J.K., 1991. Termitefungal associations and manipulations for termite control. 24th annual meeting, Society for Invertebrate Pathology, Northern Arizona University, August 4, 1991. Flagstaff, Arizona, Society for Invertebrate Pathology. p. 29. Grace, J.K., 1993. Microbial termite control: effects of entomogenous fungi on the Formosan subterranean termite (Isoptera: Rhinotermitidae). Proceedings of the 1st International Conference on Insect Pests in the Urban Environment, St Johns College, Cambridge, 30 June July 1993. p. 474. Grace, J.K., 1997. Biological control strategies for suppression of termites. J. Agric. Entomol. 14, 281. Grace, J.K., Ewart, D.M., 1996. Recombinant cells of Pseudomonas uorescens: a highly palatable encapsulation for delivery of genetically engineered toxins to subterranean termites (Isoptera: Rhinotermitidae). Lett. Appl. Microbiol. 23, 183. Grace, J.K., Su, N.-Y., 2001. Evidence supporting the use of termite baiting systems for long-term structural protection (Isoptera). Sociobiology 37, 301. Grace, J.K., Zoberi, M.H., 1992. Experimental evidence for transmission of Beauveria bassiana by Reticulitermes avipes workers (Isoptera: Rhinotermitidae). Sociobiology 20, 23. Grace, J.K., Tome, C.H.M., Shelton, T.G., Oshiro, R.J., Yates III, J.R., 1996. Baiting studies and considerations with Coptotermes formosanus (Isoptera: Rhinotermitidae) in Hawaii. Sociobiology 28, 511. Gunnarsson, S.G.S., 1988. Infection of Schistocerca gregaria by the fungus Metarhizium anisopliae: Cellular reactions in the integument studied by scanning electron and light microscopy. J. Invertebr. Pathol. 52, 9. Gupta, S., Krasnoff, S.B., Renwick, J.A., Roberts, D.W., Steiner, J.R., Clardy, J., 1993. Viridoxin A and Viridoxin B novel toxins from the fungus Metarhizium flavoviride. J. Organic Chem. 58, 1062. Hajek, A.E., St. Leger, R.J., 1994. Interactions between fungal pathogens and insect hosts. Ann. Rev. Entomol. 39, 293. Hnel, H., 1981. A bioassay for measuring the virulence of the insect pathogenic fungus Metarhizium anisopliae (Metsch.) Sorok. (fungi imperfecti) against the termite Nasutitermes exitiosus (Hill) (Isoptera, Termitidae). Zeitschrift fr Angewandte Entomologie 92, 9. Hnel, H., 1982a. The life cycle of the insect pathogenic fungus Metarhizium anisopliae in the termite Nasutitermes exitiosus. Mycopathologia 80, 137. 172

PAGE 173

Hnel, H., 1982b. Selection of a fungus species, suitable for the biological control of the termite Nasutitermes exitiosus (Hill). Zeitschrift fr Angewandte Entomologie 94, 237. Hnel, H., Watson, J.A.L., 1983. Preliminary field tests on the use of Metarhizium anisopliae for the control of Nasutitermes exitiosus (Hill) (Isoptera: Termitidae). Bull. Entomol. Res. 73, 305. Heintschel, B.P., Kenerley, C.M., Gold, E., 2007. Effects of Trichoderma virens fungus on the feeding behaviors of subterranean termites (Reticulitermes virginicus) Isoptera: Rhinotermitidae. Sociobiology 50, 223. Hoffmann, J.A., 2003. The immune response of Drosophila. Nature 426, 33. Hoffmann, J.A., Reichhart, J.M., 2002. Drosophila innate immunity: An evolutionary perspective. Nature Immunol. 3, 121. Hughes, W.O.H., Boomsma, J.J., 2004. Let your enemy do the work: within-host interactions between two fungal parasites of leaf-cutting ants. Proc. R. Soc. Lond. B 271, 104. Hung, C.Y., Boucias, D.G., 1992. Influence of Beauveria bassiana on the cellular defense response of the beet armyworm, Spodoptera exigua. J. Invertebr. Pathol. 60, 152. Huxham, I.M., Lackie, A.M., McCorkindale, N.J., 1989. Inhibitory effects of cyclodepsipeptides, destruxins, from the fungus Metarhizium anisopliae, on cellular immunity in insects. J. Insect Physiol. 35, 97. Inward, D.J.G., Beccaloni, G., Eggleton, P., 2007a. Death of an order: a comprehensive molecular phylogenetic study confirms that termites are eusocial cockroaches. Biol. Let. 3, 331. Inward, D.J.G., Vogler, A.P., Eggleton, P., 2007b. A comprehensive phylogenetic analysis of termites (Isoptera) illuminates key aspects of their evolutionary biology. Mol. Phyl. Evol. 44, 955. Iwama, R., Ashida, M., 1986. Biosynthesis of prophenoloxidase in hemocytes of larval hemolymph of the silkworm, Bombyx mori. Insect Biochem. 16, 547. Jayasimha, P., Henderson, G., 2007a. Fungi isolated from integument and guts of Coptotermes formosanus and their antagonistic effect on Gleophyllum trabeum. Ann. Entomol. Soc. Am. 100, 703. Jayasimha, P., Henderson, G., 2007b. Effect of Aspergillus flavus and Trichoderma harzianum on survival of Coptotermes formosanus (Isoptera: Rhinotermitidae). Sociobiology 49, 135141. Jones, W.E., Grace, J.K., Tamashiro, M., 1996. Virulence of seven isolates of Beauveria bassiana and Metarhizium anisopliae to Coptotermes formosanus (Isoptera: Rhinotermitidae). Environ. Entomol. 25, 481. 173

PAGE 174

Jouquet, P., Ranjard, L., Lepage, M., Lata, J.C., 2005. Incidende of fungus-growing termites (Isoptera, Macrotermitinae) on the structure of the soil microbial communities. Soil Biol. Biochem. 37, 1852. Kaijiang, L., Roberts, D.W., 1986. The production of destruxins by the entomogenous fungus, Metarhizium anisopliae var. major. J. Invertebr. Pathol. 47, 120. Kambhampati, S., Eggleton, P., 2000. Taxonomy and phylogeny of termites. In: Abe, T., Bignell, D.E., Higashi, M. (Eds.),Termites: Evolution, Sociality, Symbioses, Ecology. Kluwer Academic Publisher, Netherlands. pp. 1. Kershaw, M.J., Moorhouse, E.R., Bateman, R., Reynolds, S.E., Charnley, A.K., 1999. The role of destruxins in the pathogenicity of Metarhizium anisopliae for three species of insect. J. Invertebr. Pathol. 74, 213. Khan, K., 1991. Mycopathogens for biological control of Odonotermes brunneus (Hagen). J. Biol. Control 5, 32. Khan, K.I., Fazal, Q., Jafri, R.H., 1977a. Pathogenicity of locally discovered Bacillus thuringiensis strain to the termites: Heterotermes indicola (Wassman) and Microcerotermes championi (Snyder). Pakistan J. Sci. Res. 29, 12. Khan, K.I., Fazal, Q., Jafri, R.H., Ahmad, M., 1977b. Susceptibility of various species of termites to a pathogen, Serratia marcescens. Pakistan J. Sci. Res. 29, 46. Khan, K.I., Fazal, Q., Jafri, R.H., 1978. Development of Bacillus thuringiensis in a termite, Heterotermes indicola (Wassman). Pakistan J. Sci. Res. 30, 117. Khan, K.I., Jafri, R.H., Ahmad, M., 1985. The pathogenicity and development of Bacillus thuringiensis in termites. Pakistan J. Zool. 17, 201. Khan, K.I., Jafri, R.H., Ahmad, M., Khan, K.M.S., 1992. The pathogenicity of Pseudomonas aeruginosa against termites. Pakistan J. Zool. 24, 243. Khan, H.K., Jayaraj, S., Gopalan, M., 1993. Muscardine fungi for the biological control of agroforestry termite Odontotermes obesus (Rambur). Insect Sci. Appl. 14, 529. King, E.G., Spink, W.T., 1969. Foraging galleries of the Formosan subterranean termite, Coptotermes formosanus in Louisiana. Ann. Entomol. Soc. Am. 62, 536. Klass, K.-D., Nalepa, C., Lo, N., 2008. Wood-feeding cockroaches as models for termite evolution (Insecta: Dictyoptera): Cryptocercus vs. Parasphaeria boleiriana. Mol. Biol. Evol. 46, 809. Ko, W.H., Fujii, J.K., Kanegawa, K.M., 1982. The nature of soil pernicious to Coptotermes formosanus. J. Invertebr. Pathol. 39, 38. 174

PAGE 175

Kramm, K.R., West, D.F., 1982. Termite pathogens: effects of ingested Metarhizium, Beauveria, and Gliocladium conidia on worker termites (Reticulitermes sp.). J. Invertebr. Pathol. 40, 7. Kramm, K.R., West, D.F., Rockenbach, P.G., 1982. Termite pathogens: transfer of the entomopathogen Metarhizium anisopliae between Reticulitermes sp. termites. J. Invertebr. Pathol. 40, 1. Krejzov, R., 1972. The pathogenic effect of Basidiobolus sp. on Coptotermes formosanus (Shiraki) and Reticulitermes lucifugus (Rossi) (Isoptera). Vestn. Cesk. Spol. Zool. 36, 253255. Krejzov, R., 1975. Enhancement of pathogenicity of Conidiobolus coronatus for the termites Coptotermes formosanus and Reticulitermes lucifugus by precultivation on an insect host. Vestn. Cesk. Spol. Zool. 39, 13. Krejzov, R., 1976. Pathogenicity of a strain of Paecilomyces fumosoroseus isolated from Zootermopsis sp. (Isoptera). Vestn. Cesk. Spol. Zool. 40, 286. Krutmuang, P., Mekchay, S., 2005. Pathogenicity of entomopathogenic fungi Metarhizium anisopliae against termites. Abstract, Conference on International Agricultural Research for development, Stuttgart-Hohenhein, Germany. Lacey, L.A., Frutos, R., Kaya, H.K., Vail, P., 2001. Insect pathogens as biological control agents do they have a future? Biological Control 21, 230. Lai, P.-Y., Tamashiro, M., Fujii, J.K., 1982. Pathogenicity of six strains of entomogenous fungi to Coptotermes formosanus. J. Invertebr. Pathol. 39, 1. Lai, P.-Y., 1977. Biology and ecology of the Formosan subterranean termite, Coptotermes formosanus, and its susceptibility to the entomogenous fungi, Beauveria bassiana and Metarrhizium anisopliae. PhD dissertation, University of Hawaii, Honolulu, Hawaii, USA. 140p. Lain, L.V., Wright, D.J., 2003. The life cycle of Reticulitermes spp.: what do we know? Bull. Entomol. Res. 93, 267. Lamberty, M., Zachary, D., Lanot, R., Bordereau, C., Robert, A., Hoffmann, J.A., Bulet, P., 2001. Insect immunity constitutive expression of a cysteine-rich antifungal and a linear antibacterial peptide in a termite insect. J. Biol. Chem. 276, 4085. Langewald, J., Mitchell, J.D., Maniania, N.K., Kooyman, C., 2003. Microbial control of termites in Africa. In: Neuenschwander, P., Borgemeister, C., Langewald, J. (Eds.), Biological Control in IPM Systems in Africa. Cromwell Press, Trowbridge, UK. pp. 227. Laumond, C., Maulon, H., Kermarrec, A., 1979. Donnes nouvelles sur le spectre dhtes et le parasitisme du nmatode entomophage Neoaplectana carpocapsae. Entomophaga 24, 1327. 175

PAGE 176

Lavine, M.D., Strand, M.R., 2002. Insect hemocytes and their role in cellular immune responses. Insect Biochem. Mol. Biol. 32, 1237. Lax, A.R., Osbrink, W.L., 2003. United States Department of Agriculture-Agricultural Research Service research on targeted management of the Formosan subterranean termite Coptotermes formosanus Shiraki (Isoptera: Rhinotermitidae). Pest Manag. Sci. 59, 788800. Leclerc, V., Pelte, N., El Chamy, L., Martinelli, C., Ligoxygakis, P., Hoffmann, J.A., Reichhart, J.M., 2006. Prophenoloxydase activation is not required for survival to microbial infections in Drosophila. EMBO reports 7, 231. Lenz, M., 2005. Biological control in termite management: the potential of nematodes and fungal pathogens. Proceedings of the fifth international conference on Urban pests (ICPU). July 10-13 th 2005, Singapore. Leong, K.L.H., 1966. Infection of the Formosan subterranean termite, Coptotermes formosanus Shiraki by the fungus, Metarrhizium anisopliae (Metch.) Sorokin. MS thesis, University of Hawaii, Honolulu, Hawaii, USA. 53 p. Leuthold, R.H., Bruinsma, O., 1977. Pairing behavior in Hodotermes mossambicus (Isoptera). Psyche 84, 109. Logan, J.W.M., Cowie, R.H., Wood, T.G., 1990. Termite (Isoptera) control in agriculture and forestry by nonchemical methods: a review. Bull. Entomol. Res. 80, 309. Lund, A.E., 1966. US Patents 3,249,492; 3, 249,439; 3,249, 494; 3,249,500; 3,249,501. Lund, A.E., 1971. Microbial control of termites. In: Burges, H.D., Hussey, N.W. (Eds.), Microbial Control of Insects and Mites. London, Academic Press. pp. 385 Maketon, M., Sawangwan, P., Sawatwarakul, W., 2007. Laboratory study on the efficacy of Metarhizium anisopliae (Deuteromycota: Hyphomycetes) in controlling Coptotermes gestroi (Isoptera: Rhinotermitidae). Entomol. Gener. 30, 203. Maniania, N.K., Ekesi, S., Songa, J.M., 2002. Managing termites in maize cropping systems with the entomopathogenic fungus, Metarhizium anisopliae. Insect Sci. Appl. 21, 41-46. Mankowski, M., Kaya, H.K., Grace, J.K., Sipes, B.S., 2005. Differential susceptibility of subterranean termite castes to entomopathogenic nematodes. Biocontrol Sci.Technol. 15, 367. Martoja, R., Martoja-Pierson, M., 1967. Initiation aux techniques de lhistologie animale. Masson, Paris. Mauldin, J.K., Beal, R.H., 1989. Entomogenous nematodes for control of subterranean termites, Reticulitermes spp. (Isoptera: Rhinotermitidae). J. Econ. Entomol. 82, 1638. 176

PAGE 177

McCoy, C.W., 1990. Entomogenous fungi as microbial pesticides. In: Baker, R.R., Dunn, P.E. (Eds.), New Directions in Biological ontrol: Alternatives for Suppressing Agricultural Pests and Diseases. Liss, A.R. Inc., N.Y. pp. 139. Meikle, W.G., Mercadier, G., Rosengaus R.B., Kirk, A.A., Derouan, F., Quimby, P.C., 2005. Evaluation of an entomopathogenic fungus, Paecilomyces fumosoroseus (Wize) Brown and Smith (Deuteromycota: Hyphomycetes) obtained from Formosan subterranean termites (Isop., Rhinotermitidae). J. Appl. Entomol. 129, 315. Milner, R.J., 2000. Improved formulation of Metarhizium for biological control of termites. CSIRO Division of Entomology Technical Report no. 86. pp. 37. Milner, R.J., 2003. Application of biological control agents in mound building termites (Isoptera: Termitidae) Experiences with Metarhizium in Australia. Sociobiology 41, 419. Milner, R.J., Staples J.A., 1996. Biological control of termites: results and experiences within a CSIRO project in Australia. Biocontrol Sci. Tech. 6, 3. Milner, R.J., Staples, J.A., Lenz, M., 1996. Options for termite management using the insect pathogenic fungus Metarhizium anisopliae. International Research Group on Wood Preservation IRG/WP 9610142, pp. 1. Milner, R.J., Staples, J.A., Lutton, G.G., 1997. The effect of humidity on germination and infection of termites by the Hyphomycete, Metarhizium anisopliae J. Invertebr. Pathol. 69, 64. Milner, R.J., Staples, J.A., Lutton, G.G., 1998a. The selection of an isolate of the Hyphomycete fungus, Metarhizium anisopliae for control of termites in Australia. Biol. Control. 3, 240247. Milner, R.J., Staples, J.A., Harley, T.R., Lutton, G.G., Driver F., Watson, W.A.L., 1998b. Occurrence of Metarhizium anisopliae in nests and feeding sites of Australian termites. Mycol. Res. 102, 216. Milner, R.J., Samson, P., Morton, R., 2003. Persistence of conidia of Metarhizium anisopliae in sugarcane fields: effect of isolate and formulation on persistence over 3.5 years. Biocontrol. Sci. Technol. 13, 507. Miura, T., Kamikouchi, A., Sawata, M., Takeuchi, H., Natori, S., Kubo, T., Matsumoto, T., 1999. Soldier caste-specific gene expression in the mandibular glands of Hodotermopsis japonica (Isoptera: Termopsidae). Proc. Natl. Acad. Sci. USA 96, 13874. Moino, A., Alves, S.B., 1998. Efeito de imicacloprid e fipronil sobre Beauveria bassiana (Bals.) Vuill. e Metarhizium anisopliae (Metsch.) Sorokin. e no comportamento de limpeza de Heterotermes tenuis (Hagen). An. Soc. Entomol. Brasil. 27, 611. 177

PAGE 178

Moino, A., Alves, S.B., Lopes, R.B., Oliveira, P.M., Neves, J., Pereira, R.M., Vieira, S.A., 2002. External development of the entomopathogenic fungi Beauveria bassiana and Metarhizium anisopliae in the subterranean termite Heterotermes tenuis. Sci. Agric. 59, 267. Moore, B.P., 1968. Studies on the chemical composition and function of the cephalic gland secretion in Australian termites. J. Ins. Physiol. 14, 33. Myles, T.G., 2002a. Alarm, aggregation, and defense by Reticulitermes flavipes in response to a naturally occurring isolate of Metarhizium anisopliae. Sociobiology 40, 243. Myles, T.G., 2002b. Laboratory studies on the transmission of Metarhizium anisopliae in the eastern subterranean termite, Reticulitermes flavipes (Isoptera: Rhinotermitidae), with a method for applying appropriate doses of Conidia to trapped termites for release. Sociobiology 40, 265. Nalepa, C.A., Bignell, D.E., Bandi, C., 2001. Detritivory, coprophagy, and evolution of digestive mutualisms in Dictyoptera. Insect. Soc. 48, 194. Nasr, F.N., Moein, S.I.M., 1997. New trend of the use of Metarhizium anisopliae (Metsch.) Sokorin and Verticillium indicum (Petch) Gams as entomopathogens to the termite Cryptotermes Brevis (Walker) (Isoptera, Kalotermitidae). Anzeiger fur Schadlingskund e Pxanzenschutz Umweltschutz 70, 13. Neves, P.J, Alves, S.B., 1999, Controle associado de Cornitermes cumulans (Kollar, 1832) (Isoptera: termitidae) com Metarhizium anisopliae, Beauveria bassiana e Imidacloprid. Sci. Agric. 56, 313. Neves, P.J., Alves, S.B., 2000a. Grooming capacity inhibition in Cornitermes cumulans (Kollar) (Isoptera: Termitidae) inoculated with entomopathogenic fungi and treated with imidacloprid. An. Soc. Entomol. Brasil. 29, 537. Neves, P.J., Alves, S.B., 2000b. Selection of Beauveria bassiana (Bals.) Vuill. and Metarhizium anisopliae (Metsch) Sorokin strains for control of Cornitermes cumulans (Kollar).Braz. Arch. Biol. Technol. 43, 373. Neves, P.J, Alves, S.B., 2004. External events related to the infection process of Cornitermes cumulans (Kollar) (Isoptera: Termitidae) by the entomopathogenic fungi Beauveria bassiana and Metarhizium anisopliae. Neotrop. Entomol. 33, 51. Nguyen, K.B., Smart Jr, G.C., 1994. Neosteinernema longicurvicauda n. gen., n. sp. (Rhabditida: Steinernematidae), a parasite of the termite Reticulitermes flavipes (Koller). J. Nemotol. 26, 162-174. Noirot, C., 1969. Glands and secretions. In: Krishna K., Weesner, F.M. (Eds.), Biology of Termites, Vol. 1. Academic press, NY. pp 89. Noirot, C., 1995. The gut of termites (Isoptera). Comparative anatomy, systematics, phylogeny. I. Lower termites. Ann. Soc. Entomol. Fr. 31, 197. 178

PAGE 179

Nowak, M.A., May, R.M., 1994. Superinfection and the evolution of parasite virulence. Proc. R. Soc. Lond. B 255, 81. Ochiel, G.R.S., 1995. Biology and biocontrol potential of Cordycepioideus bisporus Stier and Paecilomyces fumosoroseus (Wize) Brown and Smith on the higher termite Macrotermes subhyalinus (Rambur) in Kenya. PhD dissertation, Royal Veterinary and Agricultural University, Copenhagen, Denmark. 210 pp. Ochiel, G.R.S., Evans, H.C., Eilenberg, J., 1997. Cordycepioideus, a pathogen of termites in Kenya. Mycologist 11, 7. Oda, Y., Nakamura, S., Oki, I., 1988. Harman and norharman induce SOS responses and frameshift mutations in bacteria. Mutat. Res. 208, 39. Ohkuma, M., Kudo, T., 1996. Phylogeneic diversity of the intestinal bacterial community in the termite Reticulitermes speratus. App. Env. Microbiol. 62, 461. Oliver, L., Fisher, W., 1999. Appraisal of prospective bivalve immunomarkers. Biomarkers 4, 510. Osborne, L.S., Boucias, D.G., 1985. A review of chemical antagonist to mycopathogens of citrus root weevils. Florida Entomol. 68, 409. Osbrink, W.L.A., Williams, K.S., Connick, W. J., Wright, M.S., Lax, A.R., 2001. Virulence of bacteria associated with the Formosan subterranean termite (Isoptera: Rhinotermitidae) in New Orleans, LA. Environ. Entomol. 30. 443. Pachamuthu, P., Kamble, S.T., Yuen, G.Y., 1999. Virulence of Metarhizium anisopliae (Deuteromycotina: Hyphmycetes) strain ESC-1 to the German cockroach (Dictyoptera: Blatettellidae) and its compatibility with insecticides. J. Econ. Entomol. 92, 340. Peterson, S.W., Ito, Y., Horn, B.W., Goto, T. 2001. Aspergillus bombycis, a new aflatoxigenic species and genetic variation in its sibling species, A. nomius. Mycologia 93, 689. Pie, M.R., Rosengaus, R.B., Traniello, J.F.A., 2004. Nest architecture, activity pattern, worker density and the dynamics of disease transmission in social insects. J. Theor. Biol. 226, 4551. Prestwich, G.D., 1979a. Chemical defense by termite soldiers. J. Chem. Ecol. 5, 459. Prestwich, G.D., 1979b. Interspecific variation in the defence secretions of Nasutitermes soldiers. Biochem. Syst. Ecol. 7, 211. Quesada-Moraga, E., Santos-Quiros, R., Valverde-Garcia, P., Santiago-Alvarez, C., 2004. Virulence, horizontal transmission, and sublethal reproductive effects of Metarhizium anisopliae (Anamorphic fungi) on the German cockroach (Blattodea: Blattellidae). J. Invertebr. Pathol. 87, 51. 179

PAGE 180

Quetin-Leclerc, J., Favel, A., Balansard, G., Regli, P., Angenot, L.,1995. Screening in vitro antifungal activities of some indole alkaloids. Planta Med. 61, 475. Quintana, A., Reinhard, J., Faure, R., Uva, P., Bagneres, A.G., Massiot, G., Clement, J.L., 2003. Interspecific variation in terpenoid composition of defensive secretions of European Reticulitermes termites. J. Chem. Ecol. 29, 639. Ramakrishnan, R., Suiter, D.R., Nakatsu, C.H., Humber R.A., Bennett, G.W., 1999. Imidacloprid-enhanced Reticulitermes flavipes (Isoptera: Rhinotermitidae) susceptibility to the entomopathogen Metarhizium anisopliae. Environ. Entomol. 28, 1125. Ratcliffe, N.A., Rowley, A.F., 1979. Role of hemocytes in defence against biological agents. In: Gupta, A.P., (Eds.) Insect Hemocytes: Development, Forms, Functions and Techniques. Cambridge university press, Cambrige, UK, pp. 331. Rath, A.C., 2000. The use of entomopathogenic fungi for control of termites. Biocontrol Sci. Technol. 10, 563. Rath, A.C., Tidbury, C.A., 1996. Susceptibility of Coptotermes acinaciformis (Isoptera: Rhinotermitidae) and Nasutitermes exitiosus (Isoptera: Termitidae) to two commercial isolates of Metarhizium anisopliae. Sociobiology 28, 67. Rivas, P., Cassels, B.K., Morello, A., Repetto, Y., 1999. Effects of some [beta]-carboline alkaloids on intact Trypanosoma cruzi epimastigotes. Comp. Biochem. Physiol. C 122, 2731. Roberts, D.W., Humber, R.A., 1981. Entomogenous fungi. In: Cole, G.T., Kendrick, B. (Eds.), Biology of conidial fungi. Academic Press, New York, pp. 201. Rosengaus, R.B., Traniello, J.F.A., 1997. Pathobiology and disease transmission in dampwood termites [Zootermopsis angusticollis (Isoptera: Termopsidae)] infected with the fungus Metarhizium anisopliae (Deuteromycotina: Hypomycetes). Sociobiology 30, 185. Rosengaus, R.B., Traniello, J.F.A., 2001. Disease susceptibility and the adaptive nature of colony demography in the dampwood termite Zootermopsis angusticollis. Behav. Ecol. Sociobiol. 50, 546. Rosengaus, R.B., Guldin, M.R., Traniello, J.F.A., 1998a. Inhibitory effect of termite fecal pellets on fungal spore germination. J. Chem. Ecol. 24, 1697. Rosengaus, R.B., Maxmen, A.B., Coates, L.E., Traniello, J.F.A., 1998b. Disease resistance: a benefit of sociality in the dampwood termite Zootermopsis angusticollis (Isoptera: Termopsidae). Behav. Ecol. Sociobiol. 44, 125. Rosengaus, R.B., Lefebvre, M.L., Jordan, C., Traniello, J.F.A., 1999a. Pathogen alarm behavior in a termite: A new form of communication in social insects. Naturwissenschaften 86, 544. 180

PAGE 181

Rosengaus, R.B., Traniello, J.F.A., Chen, T., Brown, J.J., Karp, R.D., 1999b. Immunity in a social insect. Naturwissenschaften 86, 588. Rosengaus, R.B., Lefebvre, M.L., Traniello, J.F.A., 2000a. Inhibition of fungal spore germination by Nasutitermes: Evidence for a possible antiseptic role of soldier defensive secretions J. Chem. Ecol. 26, 21. Rosengaus, R.B., Traniello, J.F.A., Lefebvre, M.L., Carlock, D.M., 2000b. The social transmission of disease between adult male and female reproductives of the dampwood termite Zootermopsis angusticollis. Ethol. Ecol. Evol. 12, 419. Rosengaus, R.B., Moustakas, J.E., Calleri II, D.V., Traniello, J.F.A., 2003. Nesting ecology and cuticular microbial loads in dampwood (Zootermopsis angusticollis) and drywood termites (Incisitermes minor, I. schwarzi, Cryptotermes cavifrons). J. Insect. Sci. 3, 31. Rosengaus, R.B., Traniello, J.F.A., Lefebvre, M.L., Maxmen, A.B., 2004. Fungistatic activity of the sternal gland secretion of the dampwood termite Zootermopsis angusticollis, Insect. Soc. 51, 259. Rosengaus, R.B., Cornelisse, T., Guschanski, K., Traniello, J.F.A., 2007. Inducible immune proteins in the dampwood termite Zootermopsis angusticollis. Naturwissenschaften 94, 25. Rouland, C., Ben Moussa, D, Reversat, G., Laumond, C., 1996. Etude de la sensibilit de sexus de termites Macrotermitinae une infestation par des nematodes entomopathogens de genres Heterorhabditis et Steinernema. C.R. Acad. Sci. Fr. 319, 997. Sajap, A.S., Jan, P.T., 1990. Pathogenicity of two entomogenous fungi, Beauveria bassiana and Metarhizium anisopliae on the termite Coptotermes curvignathus. Proceedings of IUFRO Workshop on Pests and Diseases of Forest Plantations, Bangkok (Thailand), 5-11 Jun 1988. pp. 266-271. Sajap, A.S., Kaur, K., 1990. Histopathology of Metarhizium anisopliae, an entomopathogenic fungus, infection in the termite, Coptotermes curvignathus. Pertanika 13, 331. Sajap, A.S., Atim, A.B., Husin, H., Wahab, Y.A., 1997. Isolation of Conidiobolus coronatus (Zygomycetes: Entomophthorales) from soil and its effect on Coptotermes curvignathus (Isoptera: Rhinotermitidae). Sociobiology 30, 257. Samuels, R.I., Charnley, A.K., Reynolds, S.E., 1988. The role of destruxins in the pathogenicity of 3 strains of Metarhizium anisopliae for the tobacco hornworm Manduca sexta. Mycopathologia 104, 51. Sannasi, A., 1968. Apparent infection of queens and drones of the mound-building termite Odontotermes obesus by Aspergillus avus. J. Invertebr. Pathol. 10, 434. SAS Institute. 2002. SAS/STAT users guide, version 9.1. SAS Institute, Cary, NC. 181

PAGE 182

Schmid-Hempel, P., 2005. Evolutionary Ecology of insect immune defenses. Ann. Rev. Entomol. 50, 529. Shah, P.A., Pell, J.K., 2003. Entomopathogenic fungi as biological control agents. Appl. Microbiol. Biotechnol. 61, 413. Shimizu, S., Yamaji, M., 2003. Effect of density of the termite, Reticulitermes speratus Kolbe (Isoptera: Rhinotermitidae), on the susceptibilities to Metarhizium anisopliae. Appl. Entomol. Zool. 38, 125. Siderhurst, M.S., James, D.M., Rithner, C.D., Dick L.D., Bjostad, L.B., 2005a. Isolation and characterization of norharmane from Reticulitermes termites (Isoptera: Rhinotermitidae). J. Econ. Entomol. 98, 1669. Siderhurst, M.S., James, D.M., Blunt, T.D., Bjostad, L.B., 2005b. Endosymbiont biosynthesis of norharmane in Reticulitermes termites (Isoptera: Rhinotermitidae). Sociobiology 45, 687705. Siderhurst, M. S., James, D.M., Blunt, T.D., Bjostad, L.B., 2005c. Antimicrobial activity of norharmane against the entomopathogenic fungus Metarhizium anisopliae (Metsch.) and the caste and phylogenetic distribution of this defense in termites (Insecta: Isoptera). Sociobiology 46, 563. Slaytor, M., 2000. Energy metabolism in the termite and its gut microbiota. In: Abe, T., Bignell, D.E., Higashi, M., 2000. Termites: Evolution, Sociality, Symbiosis, Ecology. Kluwer Academic, Dordrecht/Norwell. pp. 307. Smythe, R.V., Coppel, H.C., 1965. The susceptibility of Reticulitermes avipes (Kollar) and other termite species to an experimental preparation of Bacillus thuringiensis Berliner. J. Inverteb. Pathol. 7, 423. Smythe, R.V., Coppel, H.C., 1966. Pathogenicity of externally occurring fungi to Reticulitermes avipes. J. Invertebr. Pathol. 8, 266. St. Leger, R.J., 1991. Integument as a barrier to microbial infections. In: Binnington, K., Retnakaran, A. (Eds.), Physiology of the Insect Epidermis. CSIRO, Australia, pp. 284306. St. Leger, R. J., 1993. Biology and mechanisms of invasion of deuteromycete fungal pathogens. In: Beckage, N.C., Thompson, S.N., Federici, B.A. (Eds.), Parasites and Pathogens of Insects, Vol. 2. Academic Press, N.Y. pp. 211. St. Leger, R.J., 2008. Studies on adaptations of Metarhizium anisopliae to file in the soil. J. Invertebr. Pathol. 98, 271. St. Leger, R.J., Screen, S.E., Shams-Pirzadeh, B., 2000. Lack of host specialization in Aspergillus flavus. Appl. Environ. Microbiol. 66, 320. 182

PAGE 183

Stachel, S.J., Stockwell, S.A., Van Vraken, D.L., 1999. The fluorescence of scorpions and catarctogenesis. Chem. Biol. 6, 531. Stanley, D., 2006. Prostaglandins and other eicosanoids in insects: biological significance. Ann. Rev. Entomol. 51, 25. Staples, J.A., Milner, R.J., 2000. A laboratory evaluation of the repellency of Metarhizium anisopliae conidia to Coptotermes lacteus (Isoptera: Rhinotermitidae). Sociobiology 36, 133. Steinhaus, E.A.,1949. Principles of insect pathology. McGrow-Hill, NY. Stenzel, K., 1992. Natural occurrence and population density of Metarhizium anisopliae in cultivated and non-cultivated soil in North-Rhine-Westfalia, FRG. Abstracts of the XXV AnnualMeeting of the Society of Invertebrate Pathology, 16 August 1992, Heidelberg, Germany. p. 251. Strack, B.H., 2000. Biological control of termites by the fungal entomopathogen Metarhizium anisopliae. MS Thesis.University of Toronto, Ontario, Canada. Su, N.-Y., 2002. Novel technologies for subterranean termites control. Sociobiology 40, 95. Su, N.-Y., 2005. Response of the Formosan subterranean termites (Isoptera: Rhinotermitidae) to baits or nonrepellent termiticides in extended foraging arenas. J. Econ. Entomol. 98, 21432152. Su, N.-Y., Scheffrahn, R.H., 1986. A method to access, trap, and monitor field populations of the Formosan termite (Isoptera: Rhinotermitidae) in the urban environment. Sociobiology 12, 299. Su, N.-Y., Scheffrahn, R.H., 1988. Foraging population and territory of the Formosan subterranean termite (Isoptera: Rhinotermitidae) in an urban environment. Sociobiology 14, 353. Su, N.-Y., Scheffrahn, R.H., 1998. A review of subterranean termite control practices and prospects for integrated pest management programmes. Integrated Pest Management Rev. 3, 1. Su, N.-Y., Ban, P.M., Scheffrahn R.H., 1993. Foraging populations and territories of the eastern subterranean termite (Isoptera: Rhinotermitidae) in southeastern Florida. Environ. Entomol. 22, 1113. Su, N.-Y., Thoms, E.M., Ban, P.M., Scheffrahn, R.H., 1995. Monitoring/baiting station to detect and eliminate foraging populations of subterranean termites (Isoptera: Rhinotermitidae) near structures. J. Econ.Entomol. 88, 932. Sun, J., Fuxa, J.R., Henderson G., 2002. Sporulation of Metarhizium anisopliae and Beauveria bassiana on Coptotermes formosanus and in vitro. J. Invertebr. Pathol. 81, 78. 183

PAGE 184

Sun, J., Fuxa, J.R., Henderson G., 2003a. Effects of virulence, sporulation, and temperature on Metarhizium anisopliae and Beauveria bassiana laboratory transmission in Coptotermes formosanus. J. Invertebr. Pathol. 84, 38. Sun, J., Fuxa, J.R., Henderson, G., 2003b. Virulence and in vitro characteristics of pathogenic fungi isolated from soil by baiting with Coptotermes formosanus (Isoptera: Rhinotermitidae). J. Entomol. Sci. 39, 342. Suzuki, K., 1995. Biological control of termites by pathogenic fungi. In: Forestry and Forest Products Research: Proceedings of the Third Conference, Kepong, Malaysia, 3 October 1995, volume 2. Kepong, Forest Research Institute of Malaysia. pp. 146156. Tanada, Y., Kaya, H.K., 1993. Insect pathology. Academic Press, New York. Thompson, G.J., Crozier, Y.C., Crozier, R.H., 2003. Isolation and characterization of a termite transferrin gene up-regulated on infection. Ins. Mol. Biol. 12, 1. Toumanoff, C., 1966. Action de divers champignons entomophages sur Reticulitermes santonensis: apercu gnral. Ann. Parasitol. Hum. Compar. 40, 611. Toumanoff, C., Rombaut, J., 1965. Action de certains champignons entomophages, cultivs sur les milieux appropris attractifs, sur le Termite de Saintonge Reticulitermes santonensis (de Feytaud). Ann. Parasitol. Hum. Compar. 40, 605. Traniello, J.F.A., Rosengaus, R.B., Savoie, K., 2002. The development of immunity in a social insect: evidence for group facilitation of disease resistance. Proc. Natl. Acad. Sci. USA 99, 6838. Trenczek T. 1998. Endogenous defense mechanisms of insects. Zoology 101, 298. Trudeau, D., 1989. Selection of entomophilic nematodes for control of the eastern subterranean termite, Reticulitermes avipes (Kollar) (Isoptera: Rhinotermitidae). MS thesis. University of Toronto, Ontario, Canada. 93 p. Vanninen, I., 1996. Distribution and occurrence of four entomopathogenic fungi in Finland: effect of geographical location, habitat type and soil type. Mycol. Res. 100, 93. Vargo, E.L., Husseneder, C., 2009. Biology of subterranean termites: Insights from molecular studies of Reticulitermes and Coptotermes. Annu. Rev. Entomol. 54, 379. Veen, K.H., Ferron, P,. 1966. A selective medium for the isola-tion of Beauveria tenella and Metarrhizium anisopliae. J.Invertebr. Pathol. 8, 268. Vey, A., Fargues, J., Robert, P., 1982. Histological and ultrastructural studies of factors determining the specificity of pathotypes of the fungus Metarhizium anisopliae for scarabeid larvae. Entomophaga 27, 387. 184

PAGE 185

Vey, A., Gtz, P., 1986. Antifungal cellular defense mechanisms in insects. In: Gupta, A.P. (Ed.), Hemocytic and Humoral Immunity in Arthropods. John Wiley & Sons, Inc. pp. 89115. Vilcinskas, A., Gtz, P., 1999. Parasitic fungi and their interactions with the insect immune system. Adv. Parasitol. 43, 267. Vilcinskas, A., Matha, V., Gtz, P., 1997. Effects of the entomopathogenic fungus Metarhizium anisopliae and its secondary metabolites on morphology and cytoskeleton of plasmatocytes isolated from the greater wax moth, Galleria mellonella. J. Insect Physiol. 43, 1149. Wang, C.L., Powell, J.E., 2003. Isolation and evaluation of Beauveria bassiana for control of Coptotermes formosanus and Reticulitermes flavipes (Isoptera: Rhinotermitidae). Sociobiology 41, 369-381. Wang, C.L., Powell, J.E., 2004. Cellulose bait improves the effectiveness of Metarhizium anisopliae as a microbial control of termites (Isoptera: Rhinotermitidae). Biological Control 30, 523. Wang, C.L., Powell, J.E., OConnor, B.M., 2002. Mites and nematodes associated with three subterranean termite species (Isoptera: Rhinotermitidae). Florida Entomol. 85: 499. Watanabe, Y., Shinzato, N., Fukatsu, T., 2003. Isolation of actinomycetes from termites guts. Biosci. Biotechnol. Biochem. 67, 1797. Wells, J.D., Fuxa, J.R., Henderson, G., 1995. Virulence of four fungal pathogens to Coptotermes formosanus (Isoptera: Rhinotermitidae). J. Entomol. Sci. 30, 208. Wilson, K., Cotter, S.C., Reeson, A.F., Pell, J.K., 2001. Melanism and disease resistance in insects. Ecol. Let. 4, 637. Wilson-Rich, N., Stuart, R.J., Rosengaus, R.B., 2007. Susceptibility and behavioral response of the dampwood termite Zootermopsis angusticollis to the entomopathogenic nematode Steinernema carpocapsae. J. invertebr. Pathol. 95, 17. Wilson-Rich, N., Spivak, M., Fefferman, N.H., Starks, P.T., 2009. Genetic, individual, and group facilitation of disease resistance in insect societies. Ann. Rev. Entomol. 54, 405. Wright, M.S., Lax, A.R., Henderson, G., Chen, J., 2000. Growth response of Metarhizium anisopliae to termite nest volatiles nephthalene and fenchone. Mycologia 92, 42. Wright, M.S., Osbrink, W.L.A., Lax, A.R., 2002. Transfer of entomopathogenic fungi among formosan subterranean termites and subsequent mortality. Appl. Ent. 12, 20. Wright, M.S., Raina, A.K., Lax, A.R., 2005. A strain of the fungus Metarhizium anisopliae for controlling subterranean termites. J. Econ. Entomol. 98, 1451. 185

PAGE 186

Wu, H.J., Wang, Z.N., Ou, C.F., Tsai, R.S., Chow, Y.S., 1991. Susceptibility of two Formosan termites to the entomogenous nematode, Steinernema feltiae Filipjev. Bull. Inst. Zool. Academia Sinica 30, 31. Xu, P., Shi, M., Chen, X.-X., 2009. Positive selection on termicins in one termite species, Macrotermes barneyi (Isoptera: Termitidae). Sociobiology 53, 739. Yaginuma, K., 1990. Detection of fungi pathogenic to peach fruit moth, Carposina niponensis Walsingham from soil. Bull. Fruit Tree Res. Stn. 17, 77. Yanagawa, A., Shimizu, S., 2007. Resistance of the termite, Coptotermes formosanus Shiraki to Metarhizium anisopliae due to grooming. BioControl 52, 75. Yanagawa, A., Yokohari, F., Shimizu, S., 2008. Defense mechanism of the termite, Coptotermes formosanus Shiraki, to entomopathogenic fungi. J. Invertebr. Pathol. 97, 165. Yang, H., Schmitt-Wagner, D., Stingl, U., Brune, A., 2005. Niche heterogeneity determines bacterial community structure in the termite gut (Reticulitermes santonensis). Environm. Microbiol. 7, 916. Yendol, W.G., Paschke, J.D., 1965. Pathology of an Entomophthora infection in the eastern subterranean termite Reticulitermes avipes (Kollar). J. Invertebr. Pathol. 7, 414. Yendol, W.G., Rosario S.B., 1972 Laboratory Evaluation of Methods for Inoculating Termites with Entomophthoraceous Fungi. J. Econ. Entomol. 65, 1027. Yoshimura, T., Takahashi, M., 1998. Termiticidal performance of an entomogenous fungus, Beauveria brongniartii (Saccardo) Petch in laboratory tests. Jap. J. Environ. Entomol. Zool. 9, 16. Yoshimura, T., Tsunoda, K., Takahashi, M., Katsuda, Y., 1992. Pathogenicity of an entomogenous fungus, Conidiobolus coronatus Tyrrell and Macleod, to Coptotermes formosanus Shiraki. Jap. J. Environ. Entomol. Zool. 4, 11. Zimmerman, G., 1993. The entomopathogenic fungus Metarhizium anisopliae and its potential as a biocontrol agent. Pestic. Sci. 37, 375. Zoberi, M.H., 1995. Metarhizium anisopliae, a fungal pathogen of Reticulitermes flavipes (Isoptera: Rhinotermitidae). Mycologia 87, 354359. Zoberi, M.H., Grace, J.K., 1990a. Isolation of the pathogen Beauveria bassiana from Reticulitermes avipes (Isoptera: Rhinotermitidae). Sociobiology 16, 289. Zoberi, M.H., Grace, J.K., 1990b Fungi associated with the subterranean termite Reticulitermes flavipes in Ontario. Mycologia 82, 289. 186

PAGE 187

BIOGRAPHICAL SKETCH Thomas Chouvenc was born in France in 1980 and lived in a countryside village of Burgundy where he grew up with his two brothers. At the age of 12, the reading of a science fiction novel Les fourmis by Bernard Werber, that depicted the world from the subjective point of view of an ant worker, stimulated his interest for social insects and he decided to aim for studies in biology. In 1998, he graduated from high school and moved to Dijon to enter the undergraduate program in biology at the University of Burgundy. In 2000, he started an internship in the laboratory of chemical communication with Christian Bordereau and Alain Robert, where his first contact with termites switched his interests from ants. He stayed in this laboratory until 2004, where he helped in taking care of the termite colonies during the week and weekends and performed experiments with termite attractants. He graduated from a DEUG in general biology in 2000, obtained a licence (B.Sc.) in cell biology and physiology in 2001, and graduated from a maitrise in population ecology and evolution in 2002. In 2002-2003, he moved to the University of Paris XIII to complete his M.Sc. in ethology. For his M.Sc. thesis, Thomas studied the chemical induction of necrophobic behavior in Macrotermitinae termites. Thomas was accepted at the University of Florida in Ft. Lauderdale in Aug. 2004 and started a Ph.D. program with Dr. Nan-Yao Su in urban entomology, and he kept contact with his previous laboratory in Dijon, France where he collaborated throughout the years of his degree for research. During the five years at University of Florida, he was awarded the president prize for oral presentations at the Entomological Society of America annual meetings in 2006 and 2007. He graduated with his Ph.D. from University of Florida in the Aug. 2009. 187