1 NEW CONCEPTS IN MANAGEMENT OF DRYWOOD (BLATTODEA: KALOTERMITI DAE) AND SUBTERRANEAN TERMITES (BLATTODEA: RHINOTERMITIDAE) By BENNETT WILLIAM JORDAN 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 2013
2 2013 Bennett William Jordan
3 To Tully, the Jordans and the Clarks
4 ACKNOWLEDGMENTS I woul d like to express sincere gratitude and appreciation to my academic and research adviser, Dr. Phil Koehler, and committee member Dr. Roberto Pereira, for their expertise and willingness to help at all times. Their curious and inquisitive nature allowed me to see things in new ways and ultimately become a better scientist. I would also like to thank my committee members Dr. Rebecca Baldwin and Dr. Nicole LaMee Perez Stedman for their dedication and encouragement throughout the process. Liz Pereira and Tin y Willis deserve special notice for making the urban lab a well stocked, efficient and friendly place to work. I must thank the Departm ent of Entomolog y and Nematology for awarding me a doctoral fellowship which allowed me to work on research that was interesting to me and also afforded me the valuable experience of being a teaching assistant for Principles of Entomology. I would also li ke to thank D r. Clay Scherer ( Syngenta ) Bob Hickman ( BASF ) Jennifer Leggett ( Lindsey Pest Services ) and Termatrac for providing me with guidance, products and resea rch sites which were essential for this research The termites that lost their lives for a good cause a lso deserve special reco gnition and will not soon be forgotten. Finally, I want to thank my family and my wife Tully for being endlessly supp ort ive, helpful and understanding. My parents encouraged the exploration of what ever interested me and my brother, Evan, helped spur my interest in biology. Tully moved across the country to help me pursue my burgeoning interest in entomology and helped me t hrough the doubts and concerns one is bound to feel while pursuing this degree. Also, Eric Garfield Nied of Duluth, MN, is a friend of mine.
5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 7 LIST OF FIGURES ................................ ................................ ................................ .......... 8 ABSTRACT ................................ ................................ ................................ ..................... 9 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 1 1 2 LITERATURE REVIEW ................................ ................................ .......................... 15 Systematics ................................ ................................ ................................ ............ 15 Evolution ................................ ................................ ................................ ................. 16 Caste System and Colony Foundation ................................ ................................ .... 18 Subterranean Termites ................................ ................................ ..................... 21 Drywood Termites ................................ ................................ ............................ 22 Tunneling and Trailing Behavior ................................ ................................ ............. 23 Feeding and Grooming Habits ................................ ................................ ................ 25 Colony Defense ................................ ................................ ................................ ...... 28 Inspection and Treatment Prac tices ................................ ................................ ....... 32 Subterranean Termites ................................ ................................ ..................... 32 Drywood Termites ................................ ................................ ............................ 35 Behavioral Ef fects of Termiticide Exposure ................................ ............................ 40 3 PRE LETHAL AND LETHAL EFFECTS FROM THE TRANSFER OF FOUR NON REPELLENT TERMITICIDES IN THE EASTERN SUBTERRANEAN TERMITE RETICULITERMES FLAVIPES (KOLLAR) (BLATT ODEA: RHINOTERMITIDAE) ................................ ................................ ............................. 47 Methods and Materials ................................ ................................ ............................ 49 Foundation and Maintenance of Laboratory Colonies ................................ ...... 49 Preparation of Termites for Use in Experiment ................................ ................. 50 Soil ................................ ................................ ................................ ................... 50 Termiticides Used ................................ ................................ ............................. 51 Exposure to Termiticide Treated Soil ................................ ............................... 51 Experimental Units ................................ ................................ ........................... 52 Synthetic Pheromone ................................ ................................ ....................... 52 Evaluation Arena ................................ ................................ .............................. 52 Categorization of Termite Health and Performance ................................ ......... 53 Statistical Analyses ................................ ................................ .......................... 54 Results ................................ ................................ ................................ .................... 54
6 Donor Versus Recipient Mortality and Performance ................................ ......... 54 Synthetic Pheromone Trail Completion ................................ ............................ 55 Horizontal Transfer ................................ ................................ ........................... 55 Donors ................................ ................................ ................................ .............. 56 Recipients ................................ ................................ ................................ ......... 57 Discussion ................................ ................................ ................................ .............. 57 4 RAPID CHANGES IN TUNNELING BEHAVIOR IN THE EASTERN SUBTERRANEAN TE RMITE (RETICULITERMES FLAVIPES (KOLLAR)) (BLATTODEA: RHINOTERMITIDAE) AFTER EXPOSURE TO CHLORANTRANILIPROLE ................................ ................................ .................... 70 Methods and Materials ................................ ................................ ............................ 71 Foundation and Maintenance of Laboratory Colonies ................................ ...... 71 Experimental Units ................................ ................................ ........................... 72 Soil ................................ ................................ ................................ ................... 72 Termiticide ................................ ................................ ................................ ........ 73 Experimental Setup ................................ ................................ .......................... 73 Exposure and Experimental Procedure ................................ ............................ 73 Evaluation of Termite Health and Tunneling Behavior ................................ ..... 74 Statistical Analyses ................................ ................................ .......................... 75 Results ................................ ................................ ................................ .................... 75 Health of Termites ................................ ................................ ............................ 75 Measures of Tunneling Behavior ................................ ................................ ...... 76 Discussion ................................ ................................ ................................ .............. 78 5 QUANTIFYING DRYWOOD TERMITE ACTIVITY BEFORE AND AFTER LOCALIZED TREATMENTS WITH TERMIDOR DRY ................................ ............ 90 Methods and Materials ................................ ................................ ............................ 92 Reading the Termatrac ................................ ................................ ..................... 92 Calibration of Activity Readings ................................ ................................ ........ 93 Location and Descri ption of Infestations ................................ ........................... 94 Inspection Protocol ................................ ................................ ........................... 95 Chemical Treatment ................................ ................................ ......................... 95 Post Treatment Inspections ................................ ................................ .............. 96 Statistical Analyses ................................ ................................ .......................... 97 Results ................................ ................................ ................................ .................... 97 Calibration ................................ ................................ ................................ ........ 97 Field Sites ................................ ................................ ................................ ......... 98 Discussion ................................ ................................ ................................ .............. 99 6 CONCLUSIONS ................................ ................................ ................................ ... 110 LIST OF REFERENCES ................................ ................................ ............................. 112 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 128
7 LIST OF TABLES Table page 5 1 Characteristics of field sites, Termidor Dry application, and summary of termite activity ................................ ................................ ................................ ... 106
8 LIST OF FIGURES Figure page 3 1 The experimental arena used to evaluate termite trail following ability. .............. 67 3 2 The mean length of time to complete the synthetic pheromone trail by donor (blue) and recipient (red) termites ................................ ................................ ....... 68 3 3 The percentage of healthy (completed trail), pre lethal and dead termites among donor termites ................................ ................................ ......................... 69 4 1 Aerial view (left) and side view (right) of the tunneling apparatus used to evaluate tunneling behavior. ................................ ................................ ............... 83 4 2 The percentage of dead (A), moribund (B), and healthy termites (C) separated by treatment and length of exposure ................................ ................. 84 4 3 Percentage of replicates in which tunneling occurred for chlorantraniliprole exposed (blue bars) and control (orange bars) R. flavipes after 10, 20, 30 and 40 min exposure times ................................ ................................ ................ 85 4 4 The proportion of replicates in which tunnels reached the bottom of the tube for chlorantraniliprole exposed (blue bars) and control (orange bars) R. flavipe s after 10, 20, 30 and 40 min exposure times. ................................ ......... 86 4 5 Depth of soil reached by tunneling for chlorantraniliprole exposed (blue bars) and control (orange bars) R. flavipes after 10, 20, 30 and 40 min exposure times ................................ ................................ ................................ ................... 87 4 6 Weight of soil (g) excavated by tunneling for chlorantraniliprole exposed (blue bars) and control (orange bars) R. flavipes after 10, 20, 30 and 40 min exposure times ................................ ................................ ................................ ... 88 4 7 Mean number of tunnels constructed per replicate for chlorantraniliprole exposed (blue bars) and control (orange bars) R. flavipes after 10, 20, 30 and 40 min exposure times ................................ ................................ ................ 89 5 1 Sample display from Termatrac T3i. ................................ ................................ 107 5 2 Termatrac activity readings ................................ ................................ ............... 108 5 3 Initial, 6 month and 12 month termite activity readings ................................ .... 109
9 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 NEW CONCEPTS IN MANAGEMENT OF DRYWOOD (BLATTODEA: KALOTERMITI DAE) AND SUBTERRANEAN TERMITES (BLATTODEA: RHINOTERMITIDAE) By Bennett William Jordan December 2013 Chair: Philip G. Koehler Major: Entomology and Nematology Te rmites cause millions of dollars in damage throughout the United States, and millions more are spent on preventative control measures. In this dissertation, newly available control strategies were tested agai nst locally found subterranean, Reticulitermes flavipes and drywood (Blattodea: Kalotermitidae) termite species. R. flavipes workers were exposed to a non repellent soil termiticide, chlorantraniliprole, to determine how limited exposure affected tunneling behavior. Chlorantraniliprole was found t o be a fast acting compound that affected termite behavior with just a 10 min exposure. Compared with control groups, termites exposed to chlorantraniliprole for 10 min tunneled in one third fewer replicates constructed two fifths as many tunnels, and exc avated one fifth the weight of soil. In a second set of experiments, R. flavipes donor termites were exposed to chlorantraniliprole, chlorfenapyr, fipronil and imidacloprid and allowed 1 d, 3 d, 7 d, and 14 d to horizontally transfer termiticides to unex posed recipient termites. Termites were evaluated via a synthetic trail following assay and termite s were categorized as either healthy, suffering from pre lethal effects, or dead. Between 64% and 90% of donor
10 termites were healthy or exhibiting pre leth al behavioral effects 24 h after termiticide exposure. The relatively high percentage of initially pre lethal and healthy donor termites, followed by high donor and recipient mortality, suggested that termites exposed to slow acting termiticides transferr ed termiticides from donor to recipient termites. Localized drywood termite infestations in northern Florida were inspected for activity using the Termatrac T3i, and treated with an injecta ble dust, Termidor Dry. At six and twelve months post treatment, i nitial spots of termite activity were re measured for the amount (number of activity spikes in Termatrac reading ) and intensity (maximum spike of activity ) A significant decline in termite activity was detected one year after treatment, evidenced by a de cline in the mean number of spikes and the maximum spike detected. Protection of structures from both subterranean termites and drywood termites could be achieved using newly available technologies. Chlorantraniliprole was a fast acting, non repellent so il termiticide that killed slowly, allowing for transfer among subterranean termites in a colony. The Terma trac 3i located drywood termite activity in infested wood so localized Termidor Dry treatments could be effectively applied.
11 CHAPTER 1 INTRODUCT ION Termites are among the most successful groups in the animal kingdom. Termites are no longer considered as a separate order, however, their unique ability to break down cellulose and live eusocially has allowed them to dominate environments in which t hey are found, at least from a biomass stand point (Potter 2011). To date, termites have a worldwide distribution and seven families of termites have been described with over 280 genera and 2,600 species between them (Kambhampti and Eggleton 2000). The g reatest diversity of termites is found in tropical and sub tropical regions (Eggleton 2000). Termite colonies have a caste system, or division of labor, which includes workers, soldiers and reproductives. Termites are major decomposer of wood and help cy cle nutrients in the field. However, this wood destroying behavior makes termites a major economic pest throughout much of the world. In the United States, subterranean (Blattodea: Rhinotermitidae) and drywood termites (Blattodea: Kalotermitidae) cause b illions of dollars annually in property damage (Su 2002), and s ubterranea n termites cause more damage annually in the United States than any other structural pest (Derrick et al. 1990, Scheffrahn and Su 1999, Potter 2011). Subterranean termites enjo y a global distribution and live in colonies containing thousands to hundreds of millions of individuals living in nesting and foraging sites connected underground by tunnels. Subterranean termites enter structures while foraging for food and moisture, us ually by exploiting cracks and crevices in the foundation and through plumbing penetrations. The eastern subterranean termite, Recitulitermes flavipes (Kollar), is one of the most important pest termite species in eastern and central North America with $1 billion in economic value annually (Oi et al.
12 2003). In areas where eastern subterranean termites are found, preventative measures must be taken against them to prevent or minimize risks of a destructive infestation. Homeowners may buy pre treate d wood, use a baiting system or have termiticides applied to the soil beneath their foundation and around the perimeter of the house (Su et al. 1993, Lewis 1997, Ripa et al. 2007, Peterson 2010, Koehler et al. 2011 ) Soil termiticides have been the primary method of termite exclusion for decades because they have proven to be effective and long lasting ( Su 2002, Mulrooney et al. 2007 Potter 2011). Unlike subterranean termites, drywood term ites live completely within dry, sound wood, which functions as a colony site and food source (Snyder 1948, Krishna and Weesn er 1969). Several species of d rywood termites are established in a fairly narrow band in the continental United States, spanning from Virginia south through Florida, extending along the Gulf coast, through the southwest and up the Pacif ic coast to Norther n California and the islands of Hawaii (Scheffrahn et al. 1988, Su and Scheffrahn 1990 a Myles 1995, Scheffrahn and Su 1999, Potter 2011). Globally, drywood termites are found in many sub tropical, tropical and coastal locations (Rust et al. 1988, Scheffr ahn et al. 1988, Scheffrahn and Su 1999, Pott er 2011) with several species of drywood termites found in the United States (Scheffrahn et al. 1988, Su and Scheffra hn 1990 a Scheffrahn and Su 1999 ). Drywood termites are often found in structural lumber, fu rniture, utility poles, boats and dead tree limbs (Bess 1971) These termites do not need to maintain a soil contact as they are able to extract all the moisture they need from the wood they co nsume and metabolic processes. Drywood termite colonies are sm aller than native
13 subterranean termite colonies and consume much less wood per colony (Potter 2011). Drywood termite infestations can be treated with fumigation or localized spot treatments with injectable termiticides, extreme heat or cold, and electrocu tion. The first objective in this dissertation is to determine the pre lethal effects associated with limited exposure of R. flavipes to soil treated with four non repellent soil termiticides: chlorantraniliprole, chlorfenapyr, imidacloprid and fipronil. A new method for evaluating termite function was used to measure the effect of termit icide transfer. Most termiticide evaluations categorize exposed term ites as live, moribund or dead however, this research exploit s the trail following behavior of the eastern subterranean termite to better categorize termite health and to detect pre lethal behavioral changes. In addition to determining the effects on termites directly exposed to termiticides, these termites were allowed to horizontally transfer termiticides to chemically naive recipient termites. These experiments investigate the physical functional of termites after direct and indirect exposure to four commonly used termiticides, and provide information on the possibility of horizontal transfer in the field. The second objective is to determine the speed of action of the termiticide ch lorantraniliprole on the tunneling behavior in the eastern subterranean termite. Tunneling through soil is a necessary behavior for subterranean termites, which
14 require s worker termites to physically construct tunnels through soil by moving soil particles with their head and mandibles. Chlorantraniliprole is known to cause delayed mortality but cause behavioral changes shortly after exposure as it affect s th e muscle function of termites. Experiments associated to this objective investigate the effect of brief exposure to soil treated with chlorantraniliprole on quick reduction of the intensity function normally or consume wood, despite a healthy appearance. The fina l objective of this dissertation is to investigate the use of a termite detection device, the Termatrac T3i, to locate drywood termite infestations and quantify the activity of infestations. Spots with detected termite activity were treated with an inject able termiticide, Termidor Dry, and revisited six and twelve months post treatment, when the Termatrac T3i was used to measure for termite activity, which was compared against initial, pre treatment readings. These experiments explores the possibility o r eliably measuring cryptic termite activity using a Termatrac T3i, in conjunction with an injectable termiticide to provide for control of drywood termites.
15 CHAPTER 2 LITERATURE REVIEW Systematics Dictyoptera is a monophyletic group within the class Ins ecta, containing termites, cockroaches and mantids. The classification of these insects has been moved, changed and debated for decades. All members of Dictyoptera release eggs in an ootheca (though this has been lost in termites) and have perforated ten toria (Krishna and Weesner 1969, Bell et al. 2007, Svenson and Whiting 2009). There were several problems with Dictyoptera classification in general. The first is that the evolutionary evidence available (fossil record, molecular data) suggest Dictyopter a radiated quickly (Nalepa and Bandi 2000). When one considers that taxon sampling has been completed with a limited view because of a paucity of specimens and the fossil record, inconsistencies and questionable results are no t surprising (Inward et al. 2 007 ). Phylogenetic studies have also been derailed by questionable character polarity and improper inferences (Dietz et al. 2003, Klass and Meier 2006). Many phylogenetic studies of Dictyoptera were computed using termites as an 'out group' and excluding them from consideration (Inward et al. 2007). Another pertinent problem is that the cockroach genus at the center of it all ( Cryptocercus) is without a fossil record because members of Cryptocercus remain under logs for their entire lives (Bell et al. 20 07). The extant members of Dictyoptera family tree comprise the terminal branches with many unknown or poorly understood extinct groups (Deitz et al. 2003, Klass and Meiers 2006, Klass et al. 2008). There are several recent classification options feat uring a monophyletic framework with termites nested within Blattodea. Inward et al. (2007) produced a model
16 using 5 gene loci and 2500 sampled tr ees where all the termites are in a single family called Termitidae (Inward et al. 2007). Termitidae and Cryp tocercidae are sister families within the order Blattodea and the superorder Dictyoptera. In this classification, all termite groups would need to be lowered one level (family to sub family for example) but the cockroaches would remain the same. Termiti dae and Cryptocercidae are neither basal nor derived (Inward et al. 2007). Roth et al. (2009) produced a model based only on phylogenetic relationships (by using CAPA peptide comparisons, 18000 sampled trees) which only seeks to group insects by how close ly related they are now, without suggesting any changes to the classification scheme. The gene loci model places termites and Cryptocercidae as families within the suborder Blattoidea and as sister superfamilies to Blattidae. The peptide model places ter mites and Cryptocercidae as being more basal than Blattidae (Inward et al. 2007, Roth et al. 2007 Todaka et al. 2010, Cameron et al. 2012 ). Evolution Evidence suggests that termites evolved from omnivorous cockroaches and closely share a common ancestor with wood eating cockroaches in the family Cryptocercidae (Cleveland et al. 1934, Dietz et al. 2003, Bell et al. 2007, Klass et al. 2008, Ware et al. 2008, Davis et al. 2009, Roth et al. 2009). Members of the cockroach genus Cryptocercus contain some of the same gut flagellates with basal termites which led to this classification (Cleveland et al. 1934). Others have challenged this concept even within the past year for one main reason: the passing of endosymbiont protozoans could have occurred very early between these two (or more) groups (Thorne 1990, Triplehorn 2009). McKitterick suggested that similarities in nymph morphology between some termites and Cryptocercus strengthened this grouping but detractors say that he
17 incorrectly believed Cryptocercus to be basal cockroaches when in fact they are derived (McKitterick 1964, Bell et al. 2007). C ockroach ancestors to termites likely had a diploid reproductive system and released ootheca instead of live birth or individual eggs. Extant termites do not re lease eggs inside an ootheca; although Mastotermes darwiniensis (Froggatt) (Blattodea: Mastotermitidae) oviposits eggs in masses that somewhat oothecal in nature (Hill 1942, Emerson 1965). The loss of ootheca can likely be attributed to living within a cl imate controlled nest; since eggs are not at risk to desiccate, an ootheca would be an unnecessary energy expenditure ( Krishna and Weesner 1969). Termites and cockroaches engaged in both coprophagy and gregarious behavior which would have allowed co evolv ed gut symbionts to evolve (Emerson 1965). Ingestion of feces allows for the transmission of a microbial "package" from adult to nymph and subsequently; from one generation to the next. Things changed rapidly once mutualistic flagellates became part of t he gut and allowed wood feeding behavior (Bell et al. 2007, Inward et al. 2007, Pellens et al. 2007). Wood feeding cockroach nymphs would have needed substantial parental contact to facilitate the transfer of flagellates and other beneficial symbionts (Lo et al. 2006). Without some element of parental presence (or at minimum, aggregation amongst con and intraspecifics), it is unlikely that wood feeding would have evolved in cockroaches (Bell et al. 2007). There are currently seven termite families. Two of them, Termitidae and Serritermitidae, are considered higher, or more evolutionarily derived. The five lower termite families include, from most basal to most derived, Mastotermitidae, Termopsidae (dampwood termites), Hodotermitidae (harvester termi tes), Kalotermitidae
18 (drywood and powderpost termites) and Rhinotermitidae (subterranean termites) (Noirot and Pa steels 1998, Inward et al. 2007 Engel 2011 ). Rhinotermitidae is believed to be the link between the lower and higher termites and two genera within Rhinotermitidae, Reticulitermes and Coptotermes, share feeding and nesting strategies with both lower and higher termites (Shellman Reeve 1997). Non termite eusocial insects are found in the order Hymenoptera and include ants and certain bees and wasps. Eusociality in ants (Hymenoptera: Formicidae) evolved from predatory wasps and their caste system is different from that in termites (Thorne 1997) Workers of Hymenoptera are entirely female; whereas termite workers and soldiers are male and fema le (Snyder 1948, Wilson 1971 Thorne et al. 2003) The specific mechanisms of eusociali ty in termites are not well understood but the evidence from Cryptocercus and phylogenetic data make termite evolution appear to have taken a very different course than Formicidae or any other eusocial Hymenoptera ( Shellman Reeve 1997, Thorne 1997, Thorne et al. 2003). Caste System and Colony Foundation Termites are social and colonial organisms. There is a division of labor within a termite colony and individuals are broadly classified as workers, soldiers and reproductives. Castes are separated by morpho logy and function within the colony (Snyder 1948). Eggs and immature termites do not fit into the caste system until maturation into adulthood. Immatures differentiate into one or more of those roles, however, their role is not pre determined. As immatu res molt into adulthood, they will differentiate into a worker, soldier or nymph (pre reproductive) depending on the needs of the colony (Noirot and Pasteels 1987). The caste dynamics within a colony are not
19 entirely understood but it is believed that che mical signaling (Thorne 1997 Matsuura et al. 2010) environmental factors, and genetic inheritance ( Deheer and Vargo 2004, Hayashi et al. 2007) shape the caste structure and ratio. There has been much debate over the issue of colony size and what delineates one colony from another (Bulmer and Traniello 2002) Termite colonies are founded by primary reproductives known as alates which are produced by mature colonies (Hartke and Baer 2011) Alates are more heavily scleroti zed than their nestmates and are the only termites with compound eyes and fully developed and functional wings (Snyder 1948, Noirot and Pasteels 1987). Alates produced within a colony are approximately half males and half females (Hu and Forschler 2012) Termite colonies produce and release alates at different times (seasonally and time of day), depending on the species of termite and environmental factors, though most termite swarms are associated with rain and high humidity. Alates fly away from the colony en masse from specially constructed exit structures. Alates are poor flyers and do not typically fly far or long. In another departure from typical t ermite behavior, alates are attracted to light (Snyder 1948, Ferreira and Scheffrahn 2011, Potter 2011). Shortly after flying from their nest, alates shed their wings. There is a suture point at the base of the wings that allows for dealation to occur (K rishna and Weesner 1969). After wings are shed, females begin to search for a suitable spot to exploit (soft patch of soil, crack, crevice, knothole) to found a new colony (Sun et al. 2007, Ferreira and Scheffrahn 2011) One or more males will follow closely behind the female for up to several minutes until the female has selected a promising site to initiate a new colony (Snyder 1948, Krishna and Weesner 1969, Potter 2011, Hu and Forschler
20 2012). A pair of alates that successfully found a new colony is referred to as the royal pair and they will produce the majority of eggs for t he colony as long as they remain healthy (Noirot and Pasteels 1987) The initiation of a colony involves the excavation of a brood chamber in which the royal pair is sealed from the external environment and then mates. The site selected by the royal pair is usually in association with a food source (Cro sland et al. 1997, Korb et al. 2012) Eggs, which are produced by the queen, are small and often grouped together. The king and queen tend to the eggs and feed immatures until the initial offspring develop and can assume these responsibilities. The ra te at which a colony grows depends on many factors including which family they belong to and environmental conditions present. The range of colony size varies greatly depending on family and even species. Among subterranean termites, the concept of a sin gle centralized nest has fallen out of favor and replaced with a decentralized concept in which there are multiple interconnected nesting and feeding sites in the soil (Su et al. 1993, Potter 2011). Drywood colonies contain with dozens to a few thousand i ndividuals while some subterranean species can have several million individuals (Potter 2011). Colony dynamics are heavily skewed and the majority of individuals are workers ( Snyder 1948, Noirot and Pasteels 1987) Workers, which can be male or female, are lightly sclerotized (e xcept for their mandibles), prone to desiccation, blind and light averse. The production of workers is essential to the expansion of a newly founded colony ( Potter 2011, Scheffrahn 2011) Workers are responsible for feeding the rest of their nestmates, regardless of caste (Huang et al. 2008, Korb et al. 2012) and construct
21 and maintain colony features that are essential for the function of the colony. Subterranean termite workers construct nest features, tunnels below ground and mud shelter tubes above ground. Below ground tunnels are exploratory with termites seeking food, moisture and perhaps potential additional nesting sites. Above ground workers make exploratory shelter tubes, drop tubes and swarm tubes. There are differences in the manifestation of the caste system between families of termites. Roisin and Korb (2010) characterized the developmental pathways of the lower termites as a linear model and the higher termites as a bifurcated model. This dissertation deals with two families of lower termites, drywood termites (Blattodea: Kalotermitidae) and subterraneant termites (Blattodea: Rhinotermitidae). Subterranean Termites Subterra nean termites have a complex developmental pathway (Noirot and Pasteels 1987, Hu and Forschler 2012) Eggs are laid by reproductives (primary, secondary or tertiary) and hatch into larvae, also called white immatures Larvae molt several times (up to 9 molts in Reticulitermes ) and follow an apterous (wingless) or nymphal pathway (Krishna and Weesner 1969). Most termites that have differentiated into the apterous line function will function as true workers for the du ration of their life (Noirot and Pasteels 1987) Workers are easier defined by the morph ological features they lack (wing buds, sclerotized head capsule with large mandibles, compound eyes) (Krishna and Weesner 1969, Potter 2011), than what they have. A small percentage of workers will undergo further molts and become a soldier or an apterou s neotenic reproductive (Thorne et al. 2003) Soldiers and apterous neotenic reproductives are
22 terminal life stages; these termit es are incapable of progressive or regressive molts (Noirot and Pasteels 1987). Termites that have developed along the nymphal line are referred to as nymphs until further differentiation. Nymphs can be distinguished from workers by the presence of wing buds on the thorax. Adults derived from nymphs make up the majority of the reproductive (imaginal) capability within the colony ( Snyder 1948, Matsuura et al. 2010) Most nymphs will undergo additional molts an d become alates (primary reproductives) while a few will develop into brachypterous neotenics (secondary reproductives) (Noirot and Pasteels 1987 Thorne 1998 ) Alates are highly sclerotized and have two equal pairs of long, membranous wings. Brachypterous neotenics are lightly s clerotized, lack wings and have a distended abdomen. Nymphs are capable, though it rarely occurs, of regressively molting. The wing buds are lost and they assume the appearance and function of workers (Noirot and Pas teels 1987, Hu and Forschler 2012) Termites that have reversed their development from nymphs into workers are known as pseudergates. Drywood Termites Drywood termites lack true workers. Larvae molt and mature into pseudergates, also known as false wo rkers, or large immatures (Krishna and Weesner 1969, Korb 2007, Potter 2011, Korb and Buschman 2012, Evans et al. 2013). Pseudergates perform the typical work within the colony, but this is not a terminal life stage within Kalotermitidae. Pseudergates wi ll eventually molt into a soldier, a neotenic reproductive or into a nymph (Snyder 1948, Thorne 1997). Nymphs molt into alates, the winged and sclerotized primary reproductives of the colony. Drywood termite colonies are much
23 smaller than native subterran ean colonies, ranging from several dozen to a few thousand individuals (Snyder 1948, Krishna and Weesner 1969, Potter 2011). There are two major forms of reproductives, primary (winged alates) and replacement (wingless, not heavily sclerotized). Alates l eave the colony to found new colonies, while replacement reproductives remain in the colony in which they developed (Krishna and Weesner 1969, Thorne 1997) The development of neotenic reproductives allows drywood termite colonies to rebound when the roya l pair falters. Neoteny in drywood termite colonies allows for greater rebounding after the population of the colony has been reduced ( Thorne 1997, Ferster et al. 2001, Korb and Lenz 2004). Tunneling and Trailing Behavior Tunneling is an essential behavi or for subterranean termites. All below ground foraging behavior requires termites to physically move soil particles. The construction of the tunnel network is accomplished in two ways: worker termites either push loose soil forward then press the soil p articles from side to side with their body and head or they pick up compacted soil particles with their mandibles and deposit them outside of the tunnel (Ebeling and Pence 1957, Su and Scheffrahn 1992, Remmen and Su 2005 b ). Small soil particles are combin ed with feces and saliva and used to fortify and smooth tunnel walls (Ebeling and Pence 1957). The construction of tunnels is an intensive and elaborate process. Termites construct tunnels in a radiating, three dimensional branching pattern to decrease r epetitive searching. Chemical (conspecific pherom ones, wood and fungal odors), biotic (termite density), and environmental factors (soil particle size, moisture and thermal gradients) influence the direction and intensity o f foraging (Smythe et al. 1967 Jones
24 1990, Su and Scheffrahn 1992, Cornelius et al. 2002, Tucker et al. 2004, Green et al. 2005, Rust and Saran 2006 Gautam and Henderson 2012 Lima and Costa Leonardo 2012 ). Shelter tubes are constructed during the search for cellulose above ground; su bterranean termites do not travel above the surface without building shelter tubes. Mud tubes are made of compacted cellulose containing feces, saliva and soil particles. As this mixture dries, it hardens and provides termites protection from predation a nd desiccation while traveling above ground. As the tubes are being built, soldiers guard the opening to allow the workers to continue construction. The penetration of structures by subterranean termites typically involves the construction of mud tubes i nside or on the structure and is a diagnostic feature of an infestation. Worker termites are responsible for the construction and maintenance of foraging tunnels and shelter tubes (Gautam and Henderson 2012) There has been a documented skewed division of tunneling labor amongst worker termites present during tunnel construction. Older worker termites have been shown to engage in tunnel construction to a greater degree than younger workers (Forschler 1994, Su and Puche 2003). A small percentage of workers of the Formosan termite, Coptotermes formosanus (Shiraki), spent the majority of time tunneling (Cornelius 2012). Tunnels or tubes leading to viable food sources can quickly become termite e 2003, Swoboda and Miller 2004, Su 2005 b Potter 2011). Non alate castes are blind and depend heavily on chemical signaling to accomplish individual and colony level tasks (Krishna and Weesner 1969, Fei and Henderson 2004). Worker termites secrete a tra il
25 pheromone (3Z,6Z,8E) Dodecatrien 1 ol) from their 5 th abdominal sclerite as they forage so other termites pick up on it and are recruited to follow (Matsumura et al. 1968, Krishna and Weesner 1969, Matsumura et al. 1968, Matsumura et al. 1969, Saran et al. 2007 Lacey et al. 2011 Schwinghammer et al. 2011 ). Trails leading to food sources and the colony will be heavy with pheromone and encourage recruitment (Fei and Henderson 2005). Termites that lack the ability to follow a trail pheromone are highly u nlikely to be productive members of their colony (Arab et al. 2004, Saran et al. 2007, Potter 2011). 2 Phenoxyethanol, also known as e thylene glycol monophenyl ether, was discovered to be a synthetic analog to the trail pheromone excreted by Formosan subt erranean termites, Coptotermes formosanus (Shiraki), a close relative to R. flavipes (Fei and Henderson 2004 Fei et al. 2005, Ibrahim et al. 2005). Termites readily follow 2 Phenoxyethanol and the use of it allows for consistent trails to be made and qua ntified in a laboratory setting (Saran at al. 2011). Feeding and G rooming H abits Termites consume live or dead cellulose containing material (Snyder 1948, Krishna and Weesner 1969 Potter 2011). Termites are capable of chewing through non nutritive materi als in their search for food or moisture and subterranean termites have been observed chewing though plastic, foam (including insulation), rubber, electrical cables, drywall, stucc o and plaster ( Smith 1995, Potter 2011). There are differences in the type s of food sources exploited depending on the group of termites. Subterranean termites may feed on decaying wood, sound wood, living trees and plants, paper, sponges and cardboard (Krishna and Weesner 1969,
26 Noirot 1969 Ramakrishnan et al. 2000, Whitman an d Forschler 2007). Drywood termites nest and live within their food source which is often sound wood of furniture or structural timbers; however they have also been observed consuming books (Reierson 1950, Minnick 1970, Smith 1995, Gulmahjamad 1997, Korb 2007, Potter 2011). Cellulose, a continuously produced plant carbohydrate, is a complex organic compound that is very difficult to digest. That termites are able to do so is certainly one of the greatest reasons for their success worldwide. Termite gut s are lined with symbiotic protozoans and bacteria that aid in cellulose digestion (Cleveland 1923, Krishna and Weesner 1969, Suarez and Thorne 2000, Potter 2011). Even with a specialized gut fauna, termites must digest food twice before they are able to access nutrients. Worker termites shear small wood fragments with their mandibles, digest the contents and then defecate (Krishna and Weesner 1969, Potter 2011). Then they must consume their feces in order to re process the cellulose and retain valuable gut symbionts that otherwise may be lost in the process. Food sharing is an essential behavior that occurs between nestmates within a termite colony (Krishna and Weesner 1969, Ibrahim et al. 2003, Shelton and Grace 2003, Rust and Saran 2006, Shelton et a l. 2006, Huang et al. 2008, Spomer et al. 2008). Wood is consumed at the source by workers, who are accompanied by soldiers. Soldiers are unable to feed due to specialized mandible morphology (Prestwich 1984). When workers return to the colony, they begi n to feed and groom nestmates (Krishna and Weesner 1969). Food and fluid transfer, or trophallaxis, between worker and nestmate can be stomodeal (mouth to mouth) or proctodeal (anus of donor to mouth of recipient) (Krishna and Weesner 1969, Suarez and Tho rne 2000, Huang et al. 2008,
27 Potter 2011). Trophallaxis itself is intertwined with grooming and symbiont transfer along with the passing of nutrients. Larvae must acquire symbionts from nestmates upon emergence and symbionts must be replenished after eac h molt (Krishna and Weesner 1969, Suarez and Thorne 2000). Mouthparts are used to groom within and between castes; an important behavior which termites must do often for two reasons. Living in the soil leaves termites at risk of encountering and harbori ng dangerous parasites, fungi, bacteria which must be removed by nestmates if they are unable to do it themselves. Termites also exchange important behavior regulating chemicals during the grooming process (Krishna and Weesner 1969, Suarez and Thorne 2000 Saran et al. 2007). Drywood termites excavate galleries by eating along and across wood grain; spring and summer wood are both consumed. Drywood termites infest structural beams, window frames, wood floors, furniture and any other sound, dry wood they encounter (Snyder 1948, Krishna and Weesner 1 969, Rust et al. 1988, Scheffrahn et al. 1988 Potter 2011). The size and shape of the galleries vary but are smooth and often very close to the surface of the wood, leaving only a thin and fragile layer betwe en their gallery and the outside world (Smith 1995). Galleries can be as narrow as to only accommodate one termite at a time or wide enough to allow several to pass at once and the size of a gallery system is closely related to the size of the colony. Th e gallery system is unpredictable with some large, open cavities suddenly giving way to a narrow tunnel that may reach another cavity or not (Hickman and Forschler 2012).
28 Colony D efense The success of eusocial insects can largely be attributed to the dev elopment of complex mechanisms of colony defense (Prestwich 1984). Termites incorporate defensive strategies from across the animal kingdom including mechanical, chemical, behavioral and architectural defense. The use of these strategies in concert has a llowed termites to flourish. Among all social insects, termite soldier castes are likely more specialized than any other (Deligne et al. 1981). Mechanical defense is best broken down into mandibular modifications and cephalic modifications. Often, such modifications are associated with foraging but this is not true with soldier termites. Each type of morphological adaptations can be found in termite soldiers but there is much greater variety in modifications to the mandibles. Each modification has cos ts and benefits associated with it (Prestwich 1984). Jaws of termite soldiers fit into three categories based on their use: slashing, biting/crushing and snapping. Crescent shaped mandibles for slashing are the primary modifications seen in members of Rh inotermitidae (Prestwich 1984). They are also smooth, which is a departure from the serrated jaws often seen in soldier termites. Slashing mandibles tend to have a wide breadth as well which allows for greater amplitude when subduing predators. This mo dification does not require the musculature necessary to power other jaw shapes. In some species the jaws actually cross (Scheffrahn et al. 1998). These mandibles are suited to deal with predators physically but they also often associated with chemical s ecretions. A soldier will open a wound with their jaws and inject it with painful chemical compounds. The combination of physical and chemical defense is powerful and highly effective.
29 Biting and crushing jaw modifications are common amongst the worker c aste in Rhinotermitidae. The same jaw modifications can be seen in soldiers of primitive families, though they are not nearly as effective as the modifications exhibited by more derived families (Inward et al. 2007). The mandibles are thicker and serrate d, yet lack the strength and range of motion to be very effective as defense mechanisms. The success of termite colonies with low proportions of soldiers (like R. flavipes and R. fukiensis), depends on the ability of workers to provide tangible defense ev en though their jaws are not ideally suited for it ( Minnick 1970 ). Cephalic modifications also are important factors in colony defense amongst some members of Rhinotermitidae (Hermann 1984). The size and shape of termite soldier head s Modifications of head shape and size offer the colony a form of physical (phragmotic) defense (Krishna and Weesner 1969). This type of defense is relatively common in lower termites, like the West Indian drywood termite, Cryptotermes brevis (Wal ker) (Minnick 1970 ). Sold iers of some species use their heads to plug foraging holes order to temporarily block predators from entering the colony (Hermann 1984). This is an effective strategy and when coupled with the jaw modifications it can decrease the risk of injury to t he s oldier (Prestwich 1984). Behavioral defense strategies have been observed in workers and soldiers, with aggression as the most common trait (Krishna and Weesner 1969, Deligne et al. 1981). Behavioral defensive mechanisms in soldiers are manifested throu gh mechanical and chemical adaptations (Stuart 1972). Soldiers with phragmotic head modifications plug holes providing a physical barrier into the colony ( Minnick 1970 Scheffrahn and Su 1994 ). When a soldier puts its head out to plug a nest opening it w ill often bite or gnash
30 its mandibles at the same time (Prestwich 1984). Biting or biting coupled with chemical secretions is the standard mode for behavioral defense in soldiers. Subterranean termites produce nests that are closed off to the environment and largely resistant to fungal and bacterial growth (Krishna and Weesner 1969, Thorne et al. 1997). The most important defense is not against a predator, but rather the environment. Termites are soft bodied and desiccate quickly when they are removed fr om their colony (Hu et al. 2012) Without a good nest construction, the colony will likely fail. The colony is not very susceptible to disease if the desired microclimate is achieved within the nest (Krishna and Weesner 1969). Kalotermitidae and Rhinote rmitidae live in a protected environment, liming the potential predator pool and access to the nest. Defensive behaviors are coordinated and communicated throughout the colony through sound, contact and pheromones ( Prestwich 1984, Chouvenc and Su 2010) With eusocial insects, most defensive communication occur above the individual l evel and below the population level (Hermann 1984). Soldiers are acutely aware of distress signals from workers and other soldiers and act accordingly The detection of chemical odors, usually pheromones, is another important defensive strategy. Alarm ph eromones are the primary defensive pheromones used and termites use two types: general alarms and specific alarms (Deligne et al. 1981). During a general alarm normal activity will cease and workers seek advantageous harborage and often clump together, wh ich is called thigmotaxis (Messenger et al. 2005). Once the threat has passed, the colony will return to normal (Stuart 1972). Specific alarms are often initiated by head banging which causes mass excitation and
31 greater movement amongst the termites. So ldier castes of Rhinotermitidae, including those in Reticulitermes flavipes will bang their heads audibly as an alarm signal to the rest of the colony (Prestwich 1984). Head banging, a poorly understood behavior, is used as a defensive signal, and it ope rates like a positive feedback mechanism (head banging begets more head banging) during threats (Whitman 2006) Once the threat has diminished, the colony will return to its homeostasis (Hermann 1984). Termites have a subgenual organ that is sensitive to substrate vibrations which compensates for poor traditional hearing as termites cannot hear airborne sounds (Snyder 1948, Krishna and Weesner 1969). The detection of vibrations can induce a defensive response depending on the specifics of the vibrations (Cornelius and Grace 1997). This is one way that termites can be prepared for an upcoming attack; it is a chance to ready the defenses of the colony before the threat has materialized. That greater movement indicates the major difference between general and specific alarm because it stimulates workers to build defensive structures more rapidly (Reinhard and Clement 2002 Hertel et al. 2011 ). Workers of Rhinotermitidae are especially aggressive (Krishna and Weesner 1969). Some species of this family hav e as low as a one percent soldier population so workers need to be able to perform defensive duties to allow for colony function and growth (Cornelius and Grace 1997). Workers will bite intruders with their mandibles and do not retreat from larger predat ors. Workers engage in chemical defense by regurgitating on an intruder (Wie et al. 2007). Regurgitated matter acts as an irritant and discourages further predation attempts (Hermann 1984). Workers also defecate on intruders, reducing the severity and d uration of the attack (Deligne et al. 1981). There
32 are times when workers become too vigorous during regurgitation and defecation which can result in fatal abdominal dehisance or autothysis. Abdominal dehisance occurs when the abdominal wall bursts due t o vigorous defecation during defensive bouts (Prestwich 1984). This has been documented in both soldier and workers but it more common in workers. Abdominal dehisance is a deterrent to predators but it is also deadly to the termite (Krishna and Weesner 1 969, Hermann 1984, Prestwich 1984). Inspection and Treatment Practices Subterranean Termites Inspection for evidence o f subterranean termite presence is an important step in termite control. Moisture and cellulose are the most important factors for subte rranean termite infestations, and are often found together in infested structures. Once termites have penetrated a structure and have found wood, they can retrieve the water they need by maintaining contact with the soil (Krishna and Weesner 1969, Potter 2011). As long as that soil connection remains, termites will be able to get the water they require and consume the structure until stopped (Snyder 1948, Krishna and Weesner 1969). Inspections for subterranean termite infestations are intended to identif y conditions conducive to termite presence, including areas with high moisture and poor drainage, dirt filled porches, cracks and crevices in foundations, expansion joints, mulch and vegetation abutting a structure and facade construction extending below g rade, as well as physical indications of an existing infestation, including shelter tubes, damaged wood and alate swarms (Snyder 1948, Smith 1982, Tucker 2008, Potter 2011). In addition to visual inspections, there are tools available to aid during insp ection, including moisture meters, which are perhaps the most beneficial tool. Moisture meters can locate moisture issues in inaccessible locations, an indication of an area
33 susceptible or already exploited by subterranean termites (Thorne 1993 Potter 2011 ) Other available tools include acoustic emissions detectors, microwave radiation detectors, scoping devices, and canines trained to detect the odors associated with termites (Lewis 1993 Potter 201 1, Hickman and Forschler 2012). Even thorough inspections may not determine the presence of an existing infestation, due to the nature in which termites exploit structures. Many of the common structural entry points for termites, including cracks in conc rete slabs, plumbing fixtures, wooden stakes left in concrete during construction, and behind brick and stucco veneer are difficult to inspect. In areas with high termite pressure, preventative treatments are recommended, and in some states, legally requi red with the construction of new structures (Koehler et al. 2011). There are several methods of termite exclusion currently used, including physical barriers to entry, pressure treated wood, treatment of structural voids and timbers, baiting systems and s oil termiticides (Lewis 1997 Hu 2005, Huang et al. 2006, Mulrooney et al. 2006, Mulrooney et al. 2007, Horwood et al. 2012 Koehler et al. 2011, Eger et al. 2012 Jordan et al. 2013 ). Biological control of termites has long been explored, but has been ei ther unsuccessful or impractical ( Wright et al. 2002, Chouvenc et al. 2011). Soil termiticides are the dominant method of termite prevention used in North America, and are primarily formulated as liquids or soluble concentrates. Soil termiticides are app lied pre or post construction to provide a chemical barrier against subterranean termites (Su and Scheffrahn 1998). Soil termiticides have been in use for several decades and the way in which they are applied has changed little. Termiticides are mixed w
34 applications are used underneath foundational slabs, garages, patios and other expansive areas covered by concrete or asphalt and termiticides are applied at 3.79 liters per 3.05 square meters (Potter 2011 ). Vertical applications are made to penetrate deeper into the soil in areas with a narrower band of treatment, like around the perimeter of the foundation, and are applied at 15.14 liters of mixed termiticide per 3.05 linear met ers Ideally, soil termiticides are applied before the construction of a structure, as it is easier to achieve uniform distribution of the active ingredient (Su and Scheffrahn 1998). Termiticide is applied to all soil in the area where the slab will be p oured, around the foundation, and around plumbing and utility conduits. Once the slab has been poured, and the structure is in place, a vertical treatment is made around the sides of the foundation (Potter 2011). Horizontal barriers underneath the slab a re difficult to achieve post construction, so termiticide can either be injected beneath the slab through holes drilled into the foundation or by applying termiticide through rods inserted into the soil from the perimeter of the structure and pointed unde r neath the foundation Soil termiticides are categorized as repellent or non repellent, based on their detectability by foraging termites. Pyrethroid based termiticides are highly repellent to termites; termites are not killed but rather refuse to tunnel into soil treated with a pyrethroid termiticide. Repellent termiticides do not cause mortality of foraging termites, but can still be effective at preventing invasion and consumption of the structure; even at low concentr ations of active ingredient (Bl s ke and Hertel 2001, Gahlhoff and Koehler 2001, Smith et al. 2008, Potter 2011). Repellent termiticides have fallen out of favor in recent years. This is partially due to the availability of more effective non repellent termiticides and also because with a pplication error, repellent termiticides
35 leave gaps that termites exploit (Forschler 1994, Bl ske et al. 2003, Saran 2006, Potter 2011). Even a small (> 2.5 cm ) disruption in a band of repellent termiticide may be wide enough to allow for exploitation (Fo rschler 1994, Kuriachan and Gold 1998). Non repellent termiticides are applied as preventative barrier treatments around the perimeter of structure in the same manner as repellent termiticides (Su and Scheffrahn 1998, Potter 2011). However, they do not at tract nor discourage foraging termites from tunneling into treated soil. The method in which the termiticide kills a termite varies by active ingredient as there are termiticides available with several modes of action. Older termiticides relied on cyclodie nes, like aldrin, heptachlor and chlordane, and organophosphates, like chlorpyrifos, which were highly effective but also presented environmental and health hazards. Beginning in 1988, their use as soil termiticides was phased out and soon were no longer a llowed to be used at all (Potter 2011). Drywood Termites In North America, economic losses associated with drywood termites have been estimated at $300 million annually, with most activity occurring in California, Florida and Hawaii (Scheffrahn et al. 19 88, Scheffrahn and Su 1999). The cryptic nature of drywood termites allows for the use of fewer control strategies than are available against subterranean termites. Preventative control measures used against drywood termites are limited to physically exc luding alates from entering structures, by reducing entry points, and preventing alates that have entered a structure from penetrating wood by using pressure treated wood and treating structural voids and the exterior surfaces of wood with insecticides lik e silica aerogel and borates (Ebeling and Wagner 1959, Potter 2011). The use of these preventative control measures is not nearly as common
36 against drywood termites as subterranean termites (Potter 2011), so drywood termite control usually occurs after an infestation has been detected. Drywood termite infestations seldom jeopardize the structural integrity of a structure but can cause significant and visible damage. In the years before a colony can produce alates, detection is difficult and infestations often persist for many years unnoticed. The presence of a drywood termite infestation is typified by, in order of which is most likely to occur first, fecal pellets, visible damage on exterior surfaces of wood, and a swarm of alates inside the structure. Before fecal excretion, termites use highly specialized rectal pads to squeeze as much water out of their feces as possible and retain it (Krishna and Weesner 1969). What remains are tiny, hard, oval shaped pellets with 6 indentations on the sides, whi ch are a diagnostic feature of a drywood termite infestation (Smith 1995, Potter 2011). Pellets range in color from cream to red to black, and their color is not necessarily representative of the food they have been eating (Krishna and Weesner 1969, Smith 1995, Potter 2011). The lack of water in the pellets helps preserve them for a long time, making it diff icult to determine how long it has been since they were kicked out. Pellets from today are indistinguishable from ones that are years old (Smith 1995 ). Drywood are round and tiny, less than 1mm in diameter (Smith 1995, Potter 2011). Pellets found on the ground can be traced to the kick out hole from which they ca me. The higher the kick out hole is from the floor, the more difficult it usually is to find. Pellets falling from several feet hit the ground with greater force and as a result the pellets are scattered and are not likely to be concentrated in a single pile. Pellets may also be pushed
37 upward from below (Smith 1995). However, fecal pellets also may not be kickout out of the colony or excluded into a place in which they are undetected. The way in which drywood termites excavate galleries often leaves o nly a thin and fragile layer of wood between the external environment and the colony, typified by a bubbled or translucent appearance (Smith 1995, Hickman and Forschler 2012). Tapping on infested wood will often produce a hollow sound, compared to sound w ood, or break through the surface of the wood (Potter 2011). Alate swarms are the greatest indication of an infestation but due to the slow rate of colony expansion, it typically takes 5+ years from penetration of wood (through cracks, joints or other irr egularities in the wood) by the royal pair until they are able to develop and mature the colony enough to produce alates (Krishna and Weesner 1969, Smith 1995, Potter 2011). Specially configured termite detection devices include devices designed to detec t sound, acoustic emissions detectors, motion, microwave radiation detectors and gallery configuration, resistograph and X ray (Rust et al. 1988, Lewis 1997, Lewis 2003, Hickman and Forschler 2012). X ray devices are expensive and have only been used by r esearchers in drywood termite infestations (Lewis 2003) Resistographs Moisture meters and thermal imaging devices are not especially helpful as drywood termites maintain similar levels as their surroundings (Potter 2011). Acoust ic emissions detectors have been shown to reliably detect sound produced by termites shearing and consuming wood (Scheffra hn et al. 1993, Thoms 2000, Mankin et al. 2002) although they are not reliable when used through drywall (Potter 2011). Thoms (200 0) reported a decline in drywood termite activity by using an acoustic emissions (AE) detector after structural treatments with the insecticidal active ingredient spinosad.
38 Lewis (2003) reported that AE detectors were 80 94% successful compared to 86% fo r microwave detection devices (Peters and Creffield 2002). The microwave detection device used in this study, Termatrac T3i, was a newly released device that performs additi on, the ease of use of the Termatrac device (lower sensitivity, accelerometer to distinguish between device and termite movement, high portability, and readability) along with the ability for it to detect activity through materials other than wood make it a more attractive option than an AED for use in the field. Microwave radiation devices, like the Termatrac (TermaTrac, Queensland, Australia), detect movement of termites by the reflection of microwaves off of moving termites. Microwave radiation detec tion devices have been available for several years and are best used to determine the presence or absence of drywood termite infestations (Lewis 2003, Hickman and Forschler 2012) Hickman and Forschler (2012) were reliably able to detect termite activity and confirm elimination of termite colonies in wooden boards after treatment with a termiticide. There are two primary options for the treatment of drywood termites: fumigation and spot treatment. The decision about which method to use is typically determined by the scope of the infestation. Multiple or wide spread infestations within a structure are typically treated b y fumigation, while an infestation believed to be isolated or localized, a spot treatment may be administered (Lewis 2003, Potter 2011). When executed properly, all life stages of drywood termites in a structure (or piece of furniture) will be killed duri ng fumigation (Gray 1960, Su and Scheffrahn 1986 Lewis 1997). S ulfuryl fluoride is the primary fumigant used in structural pest control in the Unit ed States
39 ( Stewart 1957, Osbrink et al. 1987, Scheffrahn et al. 1997, Potter 2011). Localized drywood termite treatments include application of extreme heat, extreme cold, electrocution, and most commonly, introduction of termiticide into the gallery syst em ( Pence 1966, Ebeling 1983, Forbes and Ebeling 1986, Quarles 1993, Lewis and Haverty 1996, Su 1996, Ebeling 1997, Lewis 1997 Quarles 1997, Quarles 1999, Lewis 2003 ) which can be ve ry complicated and rely on accurate inspections and targeted application of termiticide. Holes are drilled into infested wood, with the selection of drilling sites guided by termite detection devices or physical evidence of an infestation and a termiticid e is introduced into the gallery system. Termiticides used in localized drywood termite treatments include non repellent foams and dusts formulated with a variety of organic and inorganic active ingredients (Lewis and Haverty 1996, Lewis 1997, Scheffrahn et al. 1997a, Scheffrahn et al. 1997b, Potter 2011). D rill and treat products are designed to fill or coat galleries, not necessarily contact termites directly. The termiticides used in drill and treat methods have changed, but this methodology has been used for nearly 100 years (Snyder 1948, Rust et al. 1988, Smith 1995, Lewis 2003, Potter 2011). Drywood termites are confined to a network of connected galleries, so a termiticide that hits part of the gallery may eventually wipe out the colony. However the variability and unpredictable way in which drywood termites excavate galleries make treatment difficult (Rust et al. 1988, Smit h 1995, Scheffrahn et al. 1997 Hickman and Forschler 2012). Drilling right where activity was found does not necessarily guarantee that a gallery was contacted and application of termiticide into wood that does not
40 intersect a gallery results in a failed treatment (Potter 2011, Hickman and Forschler 2012). Lewis (2003) reported an efficacy range from 0 100% control for all localized drywood termite treatments. Behavioral Effects of Termiticide Exposure The development of highly varied non repellent termiticides in the past twenty years has elucidated many elements of termite behavior that were previously unknown. Currentl y available products affect the nervous system, energy production, growth and metamorphosis, muscular function, cellular function and the insect cuticle. Soil termiticides and baits are introduced into the buccal cavity during tunneling and feeding, respe ctively (Osbrink and Lax 2002, Mao et al. 2011, Potter 2011 Eger et al. 2012 ). Soil termiticides have been shown to have different binding affinities depending on the type of soil to which they are applied (Blaeske et al. 2003, Spomer et al. 2009, Potter 2011). Soil with high organic matter content tend to bind more tightly to termiticides than do soils with low organic matter content (Saran and Kamble 2008, Bagneres et al. 2009, Spomer et al. 2009). Sandy soil has been shown to allow for easy uptake by termites which results in a more toxic and thus better performing treatment (Gold et al. 1996, Ramakrishnan et al. 2000, Cornelius and Osbrink 2010, Cornelius and Osbrink 2011, Gautam and Henderson 2011). Tunneling worker termites pick up treated soil par ticles with their mandibles and physically move them, introducing the toxin into their buccal cavity (Krishna and Weesner 1969, Whitman and Forschler 2007, Brown et al. 2009, Potter 2011). Eventually, the termiticide moves to the rest of the body, ultimat ely resulting in death.
41 There is not significant contact exposure through the cuticle (Osbrink and Lax 2002, Ibrahim et al. 2003, Mao et al. 2011). Since non repellent soil termiticides have become the standard preventative termite treatment, there has been a lot of research conducted on the ways in which these active ingredients affect individuals and groups of termites (Shelton and Grace 2003, Shelton et al. 2006, Rust and Saran 2006, Gautam and Henderson 2011, Neoh et al. 2012). Novel modes of action have driven the success of chlorfenapyr (trade name Phantom; BASF Corp., Ludwigshafen, Germany), fipronil (trade name Termidor; BASF Corp., Ludwigshafen, Germany) and imidacloprid (trade name Premise; Bayer AG, Leverkusan, Germany) and chlorantraniliprole (trade name Altriset; Syngenta AG, Basel Switzerland). Fipronil is a relatively fast acting toxin that is labeled to be used against termites, cockroaches, ants, wasps, and field and turf pests in the class phenylpyrazole (Ibrahim et al. 2003, Potter 20 11). Fipronil blocks the passage of chloride ions through the GABA receptor and glutamate gated chloride channels (GluCl) in the central nervous system in insects (Ibrahim et al. 2003, Raymond Delpech et al. 2005, Huang et al. 2006, Tomalski et al. 2010). Eventually, the nervous system is shut down due to hyper excitation of nervous and muscle tissue and feeding, tunneling and other behavior cease (Remmen and Su 2005 a Saran and Rust 2007). The intoxicated effects after fipronil exposure have been report ed as uncontrollable twitching of legs and antennae (Saran and Rust 2007) The speed of action for fipronil on subterranean termites has been reported as 24 h for Formosan termites (Henderson 2003) and 1 h for the western subterranean termite, Reticuliter mes Hesperus ( Banks ) (Saran and Rust 2007).
42 Several authors have used simulated field studies to demonstrate that termites encountering fipronil treated were killed quickly and the accumulation of dead termites around the treatment prevented conspecifics from subsequent exposure (Su et al. 1982 Su 2006 Sa ran and Rust 2007 ). Saran and Rust (2007) reported that most mortality occurred within 1.5 m from a fipronil treated zone t ermites that tunnel ed into a treated zone of fipronil die d quickly. Imidacl oprid was approved by the EPA in 1994 and has a wide range of uses as an insecticide (Ramakrishnan et al. 2000, Osbrick et al. 2005, Potter 2011) Imidacloprid is a systemic chloronicotinyl in the neonicotinoid class; by interfering with the transmission o f nervous impulses it functions as a neurotoxin ( Raymond Delpech et al. 2005, Potter 2011). Eastern subterranean termites exposed to soil treated with imidacloprid may cease feeding and tunneling (Ramakrishnan et al. 2000, Parman and Vargo 2010, Tomalski e t al. 2010). The neurological effects of imidacloprid exposure were described by Schroeder and F reported to occur within one day of exposure (Osbrink et al. 2005, Mao et al. 2011) th ough Thorne and Breisch (2001) reported that termites exposed to imidacloprid have study is best classified as sub lethal rather than pre lethal. T ermites that are exposed to imidacloprid by tunneling through it would likely be rendered motionless while still in contact with the treated soil and be unable to recover enough to return to a colony site (Mulrooney and Gerard 2009). Chlorfenapyr is a pro insecticide classified as a halogenated pyrrole (Shelton et al. 2006). It is considered a pro insecticide because it is 'inactive' until it has been taken
43 up by the termite. Once consumed, chlorfenapyr is metabolized and 'activated' which results in shutting down the production of adenosine triphosphate (ATP), the primary source of cellular energy ( Bignell et al. 2004, Rust and Saran 2006, Potter 2011). When ATP production decreases or ceases, cellular and then organismal death follow. Mortality occurs within one to fourteen d ays after exposure (Burkart et al. 2002, Shelton et al. 2006, Mao et al. 2011), however, loss of locomotion can be affected within hours (Rust and Saran 2006). If colony elimination were to occur due to chlorfenapyr exposure, it would be because workers a re unable to feed nymphs (Parman and Vargo, 2010). Chlorantraniliprole is a recently registered active ingredient in the chemical class anthranilic diamide that binds tightly to the ryanodine receptors in invertebrate muscle tissue (Spomer et al. 2009, Gau tum and Henderson 2011, Potter 2011). Chlorantraniliprole prevents the muscle tissue from functioning and paralysis occurs before death (Spomer et al. 2009, Gautum and Henderson 2011, Saran et al. 2011). The mandibles of both workers and soldiers are powe red by muscles so exposure to chlorantraniliprole disables them from eating, tunneling and protecting the colony (Prestwich 1984, Hannig et al. 2009, Spomer et al. 2009, Gautum and Henderson 2011, Neoh et al. 2012). Buczkowski et al. (2012) described pre lethal effects for termites exposed to chlorantraniliprole as an aggregation of motionless termites that can no longer feed nor eat. Gautam and Henderson (2011) found 0% mortality fo r donor termites, those termites directly exposed to termiticide, and rec ipient termites termites indirectly exposed through contact with donor termites, 24 h after 4 h exposure to soil treated with 50 ppm of chlorantraniliprole, and approximately 50% donor and recipient
44 mortality 21 d after exposure. Buczkowski et al. (2012) found 100% mortality 7 d after 4 h exposure t o 50 ppm of chlorantraniliprole. Compounds like chlorantraniliprole, chlorfenapyr and imidacloprid have shown delayed toxicity to termites, especially at low concentrations (Hu 2005, Rust and Saran 2006, Sa ran and Rust 2007, Gautam and Henderson 2011, Mao et al. 2011). Slow acting insecticides were defined by Su et al. (1987) as causing 90% mortality to exposed insects within 14 days of treatment. Pre lethal effects, that may harm the colony, are among the benefits associated with slow acting termiticides. Pre lethal effects are physiological and behavioral changes occurring after termiticide exposure, but before death, and can include reduced or inhibited tunneling, grooming, and feeding behavior, as well as abnormal aggregation behavior, all of which may be caused by partial or full paralysis (Thorne 2001, Henderson 2003, Saran and Rust 2007, Saran et al. 2011, Buczkowski et al. 2012). An insect does not need to be directly exposed to an insecticide to pick up a sub lethal or lethal dose ; a principle demonstrated amongst sub social and eusocial insects ( Hu et al. 2005, Shelton et al. 2006 Buczkowski et al. 2008, Choe and Rust 2008 Neoh et al. 2012 ). This finding has led researchers to investigate the ability of these active ingredients to be passed from termite to termite through social behaviors like grooming, trophallaxis and corpse relations. These behaviors can enhance the transfer of termiticides between nestmates (Su et al. 1982, Surez and Thor ne 2000, Soeprono and Rust 2004, Huang et al. 2008, Bagnres et al. 2009). In order to maximize transfer the active ingredient should be easily picked up from soil particles and relatively slow to kill an individual. A fast acting termiticide may kill a worker before they are able to
45 return to a nesting site T unnels leading to or through treated soil may become fill ed with dead workers which provide a physical barrier to termite pass age and that tunnel may be abandoned, preventing transfer from occurri ng (Su et al. 1982, Su 2006 Sa ran and Rust 2007, Forschler 2012). The principle of horizontal transfer of non repellent soil termiticides is a well accepted idea that has been reliably demonstrated in laboratory studies ( Suarez and Thorne 2000, Ibrahim et al. 2003, Shelton and Grace 2003, Rust and Saran 2006, Saran and Rust 2008, Spomer et al. 2008, Bagneres et al. 2009, Gautum et al. 2012 ). Radioactive fipronil has been shown to be transferred to recipient termites via trophallaxis (Suarez and Thorne 2000) and body contact and grooming (Ru st and Saran 2006), after donor termites were expos ed to fipronil. Termites that die quickly cannot engage in trophallaxis and any termiticide transfer would have to occur from body contact or cannibalism. In the pa st few years, there has been disagreement as to what degree does transfer of non repellent soil termiticides occur in the field (Vargo and Parman 2012) Soil termiticides, unlike baiting systems, were not designed to eliminate colonies, but, depending on the characteristics of each compound, transfer may provide secondary termite mortality beyond direct mortality due to exposure in the field. In order to maximize transfer among termites, the termiticide active ingredient should be easily acquired and rela tively slow to kill an individual termite, so the donor will live long enough to transfer it. The termiticide should also be relatively easily transferred from donor termites to recipient termites though social interactions and physical contact. The abil ity of donor termites exposed to a termiticide to follow a trail and subsequently transfer sufficient active ingredient to cause mortality in recipient termites one to several
46 days after exposure suggests an increased likelihood that foraging termites can pick up a lethal dose and yet remain locomotive long enough to return to a nesting site and transfer termiticide to nestmates. The response of nestmates to moribund termites can determine the level of termiticide transfer that occurs. Termites have bee n shown to isolate nestmates infected with an entomopathogen (Kramm et al. 1982, Wright et al. 2002 ). However, once a termite has died, it is recognized as dead within minutes (Sun et al. 2013). Burial behavior and cannibalism have both been documented i n R. flavipes Burial is a costly behavior and cannibalizing dead termites is a means of preserving and cycling nutrients within the colony (Moore 1969, Kramm et al. 1982 Jouquet et al. 2011 Neoh et al. 2012b ). If cannibalism occurs, consuming the corp ses of termites previously exposed to termiticides can lead to horizontal transfer.
47 CHAPTER 3 PRE LETHAL AND LETHAL EFFECTS FROM THE TRANSFER OF FOUR NON REPELLENT TERMITICIDES IN THE EASTERN SUBTERRANEAN TERMITE RETICULITERMES FLAVIPES ( KOLLAR ) (BLATT ODEA: RHINOTERMITIDAE) The eastern subterranean termite, Reticulitermes flavipes ( Kollar ) is the most widely distributed subterranean termite in North America and the most important economical pest species in the United States with economic loss estimates near one billion dollars annually (Scheffrahn et al. 1988, Su 2002, Potter 2011). Control of subterranean termites is a massive industry with several methods of termite exclusion currently available including physical barriers, borate and pressure treate d wood, baiting systems and most commonly, soil termiticides ( Su et al. 1991, Lewis 1997, Hu 2005, Huang et al. 2006, Mulrooney et al. 2007, Horwood et al. 2010 Rust and Su 2012 ). Subterranean termites live below ground, and any foraging activity requir es termites to tunnel by excavating soil (Snyder 1948, Ebeling and Pence 1957, Su and Scheffrahn 1992). Non repellent soil termiticides, which are applied as preventative chemical barrier treatments around the perimeter of a structure and injected below the foundation, do not attract nor discourage termites from tunneling into treated soil (Potter 2011, Rust and Su 2012). Foraging termites contact termiticide treated soil while tunneling which introduces the toxins into their buccal cavity. There are several active ingredien ts and chemical classes found in non repellent termiticides including chlorantraniliprole (oxadiazine), chlorfenapyr (prryole), fipronil (phenylpyrazol), and imidacloprid (chloronicotinyl). Slow acting insecticides were defined by Su et al. (1987) as cau sing 90% mortality to exposed insects within 1 4 days of treatment. The array of available non repellent soil termiticides in use today have been categorized as being slow acting
48 (Mulrooney and Gerard 2009, Saran et al. 2011); especially at low concentrati ons. Pre lethal effects that may harm the colony are among the benefits associated with slow acting termiticides. Pre lethal effects are physiological and behavioral changes occurring after termiticide exposure, but before death and can include reduce d or inhibited tunneling, grooming and feeding behavior as well as abnormal aggregation behavior, all of which may be caused by partial or full paralysis (Thorne 2001, Henderson 2003, Saran and Rust 2007, Buczkowski et al. 2012) A termite does not nee d to be directly exposed to treated soil in order to pick up a lethal dose of termiticide (Ibrahim et al. 2003, Shelton and Grace 2003, Rust and Saran 2006, Bagneres et al. 2009). Horizontal termiticide transfer can occur when termites exposed to soil ter miticides engage in social behaviors necessary for colony function including grooming, trophallaxis, nest construction and corpse removal. All of these behaviors could enhance the degree to which the transfer of termiticides between nestmates occurs (Surez and Thorne 2000, Soeprono and Rust 2004, Huang et al. 2008, Bagnres et al. 2009) Worker and soldier termite are blind and depen d heavily on chemical signaling to accomplish individual and colony level tasks (Krishna and Weesner 1969, Fei and Henderson 2004). Worker termites secrete a trail pheromone (3Z,6Z,8E) Dodecatrien 1 ol) from their 5 th abdominal sclerite as they forage whi ch are detected and followed by conspecifics (Krishna and Wees ner 1969, Matsumura et al. 1968 Tai et al. 1969, Tai et al. 1971, Saran et al. 2007). Trails leading to food sources and the colony will be heavy with pheromone and encourage recruitment (Fei and Henderson 2005 Jaffe et al.
49 2012 ). Termites that lack the ability to follow a trail pheromone are highly unlikely to be productive members of their colony (Arab et al. 2004, Saran et al. 2007). This study investigated the changes in termite behavior that occurred after exposure to non repellent termiticides that are known to cause delayed mortality in subterranean termites ( Thorne 2001, Hu 2005, Rust and Saran 2006, Saran et al. 2011, Gautam and Henderson 2011) Several authors have studied the transfer effects of chlorantraniliprole, chlorfenapyr, fipronil, and imidacloprid, but these active ingredients have not been compared against each othe r under the same conditions (Ibrahim et al. 2003, Rust and Saran 2006, Shelton et al. 2006, Spomer et al. 2008, Bagneres et al. 2009 Buczkowski et al 2012). The purpose of this study was to directly compare the horizontal transfer of four commonly used, non repellent termiticides among the worker caste of the eastern subterranean termite and to evaluate changes in behavior due to pre lethal termiticide exposure. Methods and Materials Foundation and M aintenance o f Laboratory C olonies Reticulitermes flavip es were field collected from University of Florida campus in Gainesville, FL. Wood stakes were placed in ground throughout campus and sites with termite activity were selected for trap placement. Termite traps consisted of a polyvinyl chloride bucket (20 cm in height by 20 cm diameter; 811192 4, Ventura Packaging Inc., Monroeville, OH), with 11 6 cm diameter holes drilled in the sides and the base, allowing termite access beneath the surface. Buckets were placed vertically in the ground to a depth of rou ghly 19 cm, and covered with a polyvinyl chloride lid. Two rolls of single faced corrugated cardboard (20 cm in length by 10 cm in diameter) were dipped in water and placed into the bucket side b y side as a food source. From spring
50 to fall termites were collected weekly from traps by removal and replacement of cardboard rolls. Termites were removed from the cardboard rolls and maintained at room temperature (23 C) 80% RH and a photoperiod of 0:24 (L: D) hours. Colonies were maintained in plastic boxes (27 by 19 by 9.5 cm; Sterilite Co., Townsend, MA) provisioned with soil, wooden blocks (12.7 cm wide by 12.7 cm tall by 2.5 cm thick) and moistened sponges (O Cel O; 3M Co., Maplewood, MN) to provide termites with food and moisture. Traps greater than 100 m from each other are assumed to represent different colonies and were maintained in different containers in the lab. Termites were kept in the laboratory < 45 days before use in experiments. Preparation of T er mites for Use in E xperiment Seven hundred eastern subterranean termites were removed from each of the selected colonies and then split into two Petri dishes per colony, which determine if they were to be donors or recipients. Both dishes contained 50 g of Alachua County fine sand mixed with water to achieve 10% moisture (wt:wt). Dishes for donors were provisioned with moistened filter paper dyed with Nile Blue A (80% dye content, Sigma Aldrich, St Louis, MO) while recipients received plain white filter paper. All dishes were sealed with Parafilm (Pechiney Plastic Packaging Co., Chicago, IL) to retain moisture. Consumption of dyed filter paper colored termites blue in approximately 1 week, at which point the transfer experiment began. Nile Blue A is a fat soluble dye which leaves termites visibly dyed for the duration of the experiment (Surez and Thorne 2000, Evans 2006, Huang et al. 2006). Soil Bonneau fine soil used in this experiment was collected near the Entomology and Nematology Department on the campus of the University of Florida (Gainesv ille,
51 FL). Soil was sieved through a #35 (0.5 mm) stainless steel test sieve (Fisher Scientific Inc, Hampton, NH) and baked at 150C for 24 hours for sterilization. Soil was analyzed by Institute of Food and Agricultural Sciences Analytical Services Labo ratories Extension Soil Testing Laboratory at the University of Florida and found to contain 1.48% organic matter, 6.8 pH and high levels of phosphorus (53 ppm) and magnesium (51 ppm). Termiticides Used Four non repellent soil termiticides were used in th is study. Commercially formulated chlorantraniliprole (50 ppm) (Tradename Altriset; Syngenta AG, Basel, Switzerland), chlorfenapyr (125 ppm) (Tradename: Phantom; BASF Corp., Ludwigshafen, Germany), fipronil (60 ppm) (Tradename Termidor; BASF Corp., Ludwig shafen, Germany) and imidacloprid (50 ppm) (Tradename Premise 2; Bayer AG, Leverkusen, Germany) E xposure to Termiticide Treated S oil Groups of 25 dyed worker termites (donors) were placed in 60 ml souffl cups containing 5 g of compacted, untreated soil (control) or soil treated with one of four non repellent termiticides. Stock solutions of each product were prepared just prior to experimentation and the appropriate amount of each solution was mixed with water and combined with 90 g of dried and sieved s and to achieve the desired concentration in soil with 10% moisture. Soil was treated with either the label rate, or the label rate, of commercially formulated chlorantraniliprole (50 ppm), chlorfenapyr (125 ppm), fipronil (60 ppm), and imidacloprid (50 ppm). Concurrently, groups of 25 non dyed worker termites (recipients) were placed in 60 ml souffl cups containing 5 g of compacted, untreated soil. All soil used contained 10% moisture (wt:wt). After an exposure period
52 of 1 or 4 hours, termites were g ently aspirated from cups containing soil and placed together (25 donors and 25 recipients) into 60 ml souffl cups. Any soil particles the termites were aspirated with were removed from transfer cups and discarded. Each cup contained 50 termites and one circle of moistened filter paper (# 2 qualitative circles, Whatman PLC, Kent, UK) and was sealed with a tightly fitting lid (# 200 PCL, Dart Container Corp., Mason, MI). Filter paper was re moistened with water every 3 days. Experimental Units Colonies were used as replicates and twelve replicates were tested for each termiticide, concentration, exposure length and time to evaluation for a total of 960 souffl cups and 48,000 termites. Depending on availability of termites, replicates were from differen t field colonies, with a total of 10 colonies used for experiments and a minimum of 6 distinct colonies used per treatment. Synthetic P heromone 2 Phenoxyethanol (2PE), also known as e thylene glycol monophenyl ether, was discovered to be a synthetic analog to the trail pheromone excreted by Formosan subterranean termites, Coptotermes formosanus Shiraki, a close relative to R. flavipes (Fei and Henderson 2005, Fei et al. 2005, Ibrahim et al. 2005). A mixture of 10% 2 phenyoxyethanol (>90%, Aldrich Chemical Co., Milwaukee, WI) and 90% ethyl alcohol was used as the synthetic pheromone. Evaluation A rena The effects of termiticide transfer were evaluated after 24 h, 72 h, 7 d and 14 d. Termite health was evaluated by a trail following assay. Termites were remo ved from souffl cups via camel hair brush and placed onto the trial arena which consisted of two inverted plastic cups (Dart Container Co, Mason, MI) on a sheet of new, white copy
53 paper (Office Depot, Boca Raton FL). Just prior to use, a 20 cm line comp osed of synthetic trail pheromone was applied to the paper with a 0.18 mm Rapidograph pen (Faber Castell, Nuremberg, Germany). Each trail line contained an average of 3.29 mg (0.16 mg/cm) of synthetic trail pheromone mixture (N=45, std =0.0011). A new she et of paper with a new synthetic pheromone trail was used for each replicate. A 1 cm x 1 cm x 1 cm triangle was cut out of the lip of the larger plastic cup (59 ml, Dart Container Co., Mason, MI) so when the cup was inverted, it allowed the termites a narr ow and directed exit point from the introduction site. A smaller plastic cup (30 ml, Dart Container Co., Mason, MI) was inverted and placed insi de the larger cup for the 30 s in which the termites were allowed to acclimate to the trial arena and locate th e synthetic pheromone trail, which extended 2 cm into the introduction site (Figure 3 1). After 30 s the larger cup was removed, which allowed the termites to follow the synthetic pheromone trail down the paper. Categorization of Termite Health and Perfo rmance Each replicate, containing 50 termites (25 donors, 25 recipients) was allowed three minutes to traverse the 18 cm synthetic pheromone trail. For termites that completed the trail, completion times were recorded for each termite in 1 min blocks of time (completed in 0 60 s, 61 120 s, and 121 180 s). These termites were put in the category of 'completed trail'. Termites that left the introduction site but did not complete the trail will be classified as 'left introduction site'. Individuals that d id not leave the introduction site were subjected to further analysis. These termites were gently flipped using featherweight forceps and allowed 10 seconds to right themselves. Those able to do so were put in the category 'able to right' and those incap able into the category 'moribund'. Slight movement or twitching was deemed sufficient in determ ining an
54 both groups. If the termite did not exhibit any movemen t then it was considered dead. Three categories were considered pre lethal effects in this study, from least affected to most affected: left introduction site, able to right, moribund. Statistical Analyses All data analyses were performed using JMP Pro 10 statistical software (SAS Institute 2012). Comparisons among treatments (chlorantraniliprole, chlorfenapyr, fipronil, and imidacloprid), concentration (ppm), exposure times and post evaluation ti mes were evaluated using analysis of variance (ANOVA) on arcsine square root transformed data. The level of significance was set at 0.05 for all statistical tests. A test was used for separation of means for donors v er s us recipients; ANOVA an icant difference (HSD) test was used for comparisons of insecticide treatments. Results Donor V ersus R eci pient Mortality and Performance Manipulation of termites during transfer experiments did not affect the health of the termites Among termites exposed to untreated soil, there were no significant differences in the number of dead plus moribund termites between the donor and the recipient termites ( F = 2.0; df = 1; P = 0.1587). The number of termites completing the synthetic pher omone trail (d.f. = 1, F = 2.65, P = 0.1042) and the average time to complete the trail (d.f. = 1, F = 1.82, P = 0.1870) was not significantly different between donors and recipients in the control treatment.
55 Synthe tic Pheromone Trail Completion Termites i n the control group followed the synthetic pheromone trail significantly faster (donors: d.f. = 4, F = 8.59, P < 0.001; recipients: d.f. = 4, F = 9.07, P < 0.001) and more often (donors: d.f. = 4, F = 333.72, P < 0.001; recipients: d.f. = 4, F = 207.64, P < 0.001) than termites exposed to chlorantraniliprole, chlorfenapyr, fipronil or imidacloprid. Donor and recipient termites exposed to fipronil were the slowest and least likely to finish the pheromone trail. Over 37x as many donors in the control group were able to complete the trail compared to fipronil exposed termites, and those that were able to complete the trail did so much faster in the control group (1.40 .01 min) as opposed to fipronil (1.81 .08 min) (Figure 3 2) For all compounds with su fficient data (not enough donor termites exposed to fipronil completed the trail), more recipients completed the pheromone trail than donors (chlorantraniliprole: d.f. = 1, F = 20.86, P < 0.0001; chlorfenapyr: d.f. = 1, F = 25.56, P < 0.0001; imidacloprid: d.f. = 1, F = 9.34, P = 0.0024). Donor termites were slightly slower than recipient termites at completing the pheromone trail, even in the control groups (control: 0.03 min; chlorantraniliprole: 0.01 min; chlorfenapyr: 0.03 min; fipronil: 0.08 min; imid acloprid: 0.04 min). Horizon tal Transfer Direct exposure (donor termites) to chlorantraniliprole (d.f. = 1, F = 46.71, P < 0.0001), fipronil (d.f. = 1, F = 28.62, P < 0.0001), chlorfenapyr (d.f. = 1, F = 32.91, P < 0.0001) and imidacloprid (d.f. = 1, F = 26.05, P = 0.0087) caused significantly higher levels of dead plus moribund termites than indirect termiticide exposure to all four termiticides (recipient termites)
56 The number of dead and moribund termites increased with the length of post exposure ti me independent of treatment (Figure 3 3) however, the rate at which termites were negatively affected varied by product. Donor and recipient termites exposed to fipronil were killed most effectively, regardless of concentration (50% ppm, 100% ppm) or exp osure time (1 or 4 h). Fipronil resulted in the highest percentage of dead and moribund at each of the four evaluation periods (one, three, seven or fourteen days post exposure period) compared to the other termiticides. One day after a 4 h exposure to the label rate of fipronil, > 97% of donor termites were dead or moribund, which was > 1.3 times as many dead and moribund termites exposed to chlorantraniliprole, nearly twice as many as chlorfenapyr and three times the percentage of dead and moribund for imidacloprid exposed termites. Despite causing rapid mortality in donors, fipronil was transferred very effectively to recipients. For termites with three days of post exposure time, >80 of recipients were dead or moribund no matter the exposure length or concentration of fipronil. Fipronil was the only compound to cause 100% mortality in donors and recipients at the fourteen day evaluations. Donors M ortality and morbidity significantly increased for donor termites with an increase in post exposure time (chlorantraniliprole: d.f. = 3, F = 35.76, P < 0.0001; chlorfenapyr: d.f. = 3, F = 20.80, P < 0.0001; fipronil: d.f. = 3, F = 36.47, P < 0.0001; imidacloprid: d.f. = 3, F = 4.86, P = 0.0028). Higher termiticide concentrations caused significantly more mo rtality and morbidity in donor termites exposed to chlorantraniliprole (d.f. = 1, F = 17.58, P < 0.0001) and chlorfenapyr (d.f. = 1, F = 8.23, P = 0.0046). For fipronil exposed termites, concentration was not significant (d.f. = 1, F = 1.78, P = 0.2792).
57 Donor termites expose d to treated soil for 4 h had higher rates of mortality and morbidity than termites expos ed to chlorantraniliprole for 1 h (d.f. = 1, F = 19.62, P < 0.0001), chlorfenapyr (d.f. = 1, F = 26.17, P < 0.0001) and fipronil (d.f. = 1, F = 3 4.50, P < 0.0001), but not to imidacloprid (d.f. = 1, F = 1.87, P = 0.1805). Recipients M ortality and morbidity significantly increased for recipient termites as post exposure time increased (chlorantraniliprole: d.f. = 3, F = 58.90, P < 0.0001; chlorfen apyr: d.f. = 3, F = 39.05, P < 0.0001; fipronil: d.f. = 1, F = 34.50, P < 0.0001; imidacloprid: d.f. = 3, F = 23.49, P < 0.0001). The higher concentration levels significantly increased recipient termite mortality and morbidity in chlorantraniliprole (d.f = 1, F = 16.14, P < 0.0001), chlorfenapyr (d.f. = 1, F = 5.09, P = 0.0250) and imidacloprid (d.f. = 1, F = 5.09, P = 0.0273) but was not significa nt for fipronil, as observed for donor termites. Discussion Trail following is a normal and essential sub terranean termite behavior that can be affected by termiticide exposure. Termites that contact termiticide treated soil have been shown to possess near normal trail following ability immediately after exposure (Saran et al. 2011) but as post exposure time increases, the speed traveled decreases (Rust and Saran 2006, Saran and Rust 2007). The average trail completion times in this study did not follow a clear pattern. The number of termites completing the pheromone trail in this study decreased w hen termi tes were exposed for 4 h, at the label rate and as time increased post exposure, while completion times were not longer with increased time post exposure, nor were they linked to expos ure length. As post exposure time increased those termites that were h ealthy enough to follow the
58 pheromone trail became more affected and by the next evaluation period, termites that had previously shown pre lethal effects or had died. Completion times we re recorded in 1 m in blocks, so potential differences between treatm ents may hav e been lost in such large time intervals By pro rating trail completion times reported by Saran et al. (2011), it would have taken an individual termite from the control group less tha n 11 s to complete the 18 cm trail from this study. In sofar as horizontal termiticide transfer is concerned, following a pheromone trail is ultimately more important than how long it takes a termite do it. A termite exposed to termiticide treated soil that is able to return to a nesting site may be capable o f transferring termiticide to a chemically naive nestmate, regardless of how long it takes for the affected termite to return to the nest Quarcoo (2009) and Quarcoo et al. ( 2010) reported that termites exposed to 50 ppm of fipronil ceased all trailing ac tivity 6 h after treatment but in this study, 16% of don or termites (exposed to 50X the concentration of h) wer e able to complete the trail 24 h after treatment. Possible explanations include higher moistures conten t, nearly double the moisture levels were used by Quarcoo (2009) as in the current study, or a slight recovery of physical fun ctions that occurred between 14 h and 24 h after exposure. However, by 72 h after treatment, < 1% of donor termites and 10% of re cipient termites completed the pheromone trail. Over 40% of donor termites exposed to chlo rfenapyr and imidacloprid for 1 h wer e able to complete the trail 72 h after exposure; a strong indication that termiticidal effects were either sub lethal or very s low acting, pre lethal effects ; the distinction between sub lethal and pre lethal effects can only be made once a termite has died.
59 One difference in the current study is that the length of time post exposure, and pre evaluation period were significantly longer than in other trailing studies, that either used less than 8 h (Rust and Saran 2006, Saran and Rust 2007, Saran e t al. 2011) or 16 h (Quarcoo 2010) as post exposure, pre evaluation periods The percentage of termites successfully completing a pher omone trail 24 h, 72 h and even 7 d effectively show the slow acting and delayed mortality associated with chlorantraniliprole, chlorfenapyr and imidacloprid. This suggests that pre lethal effects persisted for longer periods of time than reported by othe r researchers (Mulrooney and Gerard 2009, Quarcoo 2009, Quarcoo 2010) Termites that left the introduction site in the trail following assay were often visibly shaking and slow moving, that is, exhibiting pre lethal symptoms due to termiticide exposure. However, a termite incapable of completing the pheromone trail in this study may still be able to trail back to a nesting site and contact and groom nestmates, if given enough time. If not, the termite may still move far enough from the treated zone to p revent blocking the tunnel before dying (Remmen and Su 2005 a Su 2006). Affected termites able to leave the introduction site, but not follow the pheromone trail, might suggest that a more random distribution of these pre lethal termites throughout the tu nnel and nest network may occur in field situations If intoxicated termites do not block tunnels, this allows conspecifics to reach the treated area; even if no transfer of termiticides occur between insects. This allows increased termiticide exposure to the colony. Healthy termites have been observed walling off and avoiding tunnels that are filled with dead and dying termites (Su et al. 1982). According to Henderson (2003), C. formosanus exposed to imidacloprid did not
60 beha ve normally and acted sick ly 24 h after exposure and were unable to travel very far. Moderately affected termites in this study were physically capable of righting themselves but did not leave the introduction site during the 3 min allotted for evaluation perhaps because they w ere not physically able to follow a trail. The able to right category decreased dramatically between the 1 d and 3 d evaluation periods, and presumably, many of those that had been able to right themselves at day one were unable to do so at day three, e ither because they were moribund or dead. There was a lot of variability in the behavior of termites in the able to right category; some immediately righted themselves while others struggled before weakly righting themselves. Simple categorization of t Moribund termites were no longer physically functional enough to right themselves, but some termites vigorously attempt ed to do so for several seconds. T his was especially prevalent in termites exposed to chlorantraniliprole. In a colony, these termites might be jostled into righting themselves and be able to walk or groom to some extent, increasing the likelihood of transferring termiticide to nestmates. The response of nestmates to moribund termites can determine the level of termiticide transfer that occurs. Termites have been shown to iso late nestmates infected with an entomopathoge n (Kramm et al. 1982, Wright et al. 2002 ). However, once a termite has died, it is recognized as dead within minutes (Sun et al. 2013). Burial behavior and cannibalism have both been documented in R. flavipes Burial is a costly behavior and cannibalizing dead termites is a means of preserving a nd cycling nutrients within the
61 colony (Moore 1969, Kramm et al. 1982 Neoh et al. 2012b ). If cannibalism occurs, consuming the corpses of termites previously exposed to termiticides can lead to horizontal transfer. Termites were confined on termiticid e treated soil so the route of exposure for donors was likely from mandibular contact (either directly from tunneling in exposure cups or grooming soil particles from the tarsi). Soil termiticides have been shown to have different soil binding affinities depending on the type of soil. Soil with high organic matter content tend to bind more tightly to termiticides than do soils with high sand content (Blaeske et al. 2003, Saran and Kamble 2008, Bagneres et al. 2009, Spomer et al. 2009). Based on the high levels of mortality in donor termites, it does not seem that soil termiticides were too tightly bound to the substrate, which was high in sand and low in organic matter. Individual insecticidal dose and the mechanism in which transfer occurred were not d etermined in this study. R adioactive fipronil has been shown to be transferred to recipients via trophallaxis (Suarez and Thorne 2000) and body contact and grooming ( Saran and Rust 2007 ) after donors were exposed to fipronil. Termites that die quickly c annot engage in trophallaxis and any termiticide transfer would have to occur from body contact or cannibalism. Termites were live counted in this study so cannibalism was not noted. The high rate of mortality amongst recipients secondarily exposed to fi pronil suggests that body contact is likely the primary transfer mechanism. The mode of action for each active ingredient determines how termites are affected and can help explain why groups of termites behaved as they did. Chlorantraniliprole prevents the muscle tissue from functioning and paralysis occurs
62 long before death (Spomer et al. 2009, Gautum and Henderson 2011, Mao et al. 2011). Buczkowski et al. (2012) described pre lethal effects for termites expo sed to chlorantraniliprole as manifested in aggregation s of motionless termites that can no longer feed nor eat. In this study, over 2/3 of termites directly exposed to chlorantraniliprole were physically functional after 24 h compared to over 1/3 after 72 h. Differing rates of transfer have been reported for chlorantraniliprole. Gautam and Henderson (2011) found 0% mortality for d onors and recipient termites 24 h after a 4 h exposure to soil treated with 50 ppm of chlorantraniliprole, and approximately 50% d onor and recipient mortality 21 d after exposure. The low levels of mortality report ed by Gautam and Henderson (2011 ) was attributed to the high organic content in the soil used (18.6%) as exposure to sand treated with 50 ppm of chlorantraniliprole on sand caused 90% morta lity in donors and re cipients 5 d after exposure. The organic content in the soil used in this study (1.5%) was much closer to the sand used by Gautam and Henderson (2011), as were the mortality rates. Buczkowski et al. (2012) found 100% donor and recipient mortality 7 d aft er a 4 h exposure to 50 ppm of chlorantraniliprole while the mortality numbers from this study were slightly lower. Horizontal transmission of soil termiticide should occur to a greater extent in regions with sandier soils, like the southern United States Mortality after chlorantraniliprole exposure is delayed with mandibular function as the first area to be affected (Hirooka et al. 2007, Lahm et al. 2009). Muscle failure limits a termite's ability to tunnel and follow a chemical trail (Pitts Singer a nd Forschler 2000, Spomer et al. 2009); among pre lethal symptomatic termites, those exposed to chlorantraniliprole were the most visibly affected as their trailing behavior w as
63 uncoordinated and uneven. Despite suffering from the initial stages of muscul ar paralysis, chlorantraniliprole exposed termites followed the synthetic pheromone trail closely, when capable. The tunneling behavior of R. flavipes has been shown to be inhibited after only 10 min of exposure to chlorantr aniliprole treated soil (Chapter 4 ). However, a termite that is unable to tunnel after termiticide exposure may still be able to follow trails and return to a nesting site to interact with conspecifics, therefore transferring termiticide active ingredient to other colony members. The time to mortality for fipronil on subterranean t ermites has been reported as 24 h for Formosan termites (Henderson 2003) and 1 h for the western subterranean termite, Reticulitermes Hesperus Banks (Saran and Rust 2007). The intoxicated effects, includ ing uncontrollable twitching of legs and antennae, reported by Saran and Rust (2007) were not observed in this study. Fipronil exposed donors experienced pre lethal and lethal eff ects of fipronil well before 24 h, as evidenced by the high number of mor ibu nd and dead recipients at 24 h after exposure. The fast action of fipronil led to an incredibly efficient rate of transfer in this study. Over 56% of recipients exposed to donors that had been confined on 50 ppm of fipronil treated soil for 1 h were dead or moribund 24 h later; much higher than other termiticides tested but well below the 100% r ecipient mortality after 24 h reported by Saran and Rust (2007). The high levels of pre lethal and dead recipients within a short time after exposure indicate s th at most termitic ide transfer occurred within 24 h of exposure to donor termites, consistent with the findings of Saran and Rust (2007). As expected the concentration of fipronil was insignificant in this study as both rates caused mortality quickly.
64 Alth ough horizontal transfer was high amongst termites exposed to fipronil, the inability of donors to complete the pheromone trail or leave the introduction site suggests they were immobilized quickly, and transfer in the field may be limited to a small area near the treatment. Potential donor termites that contacted fipronil treated soil in previous studies (Su et al. 1982, Saran and Rust 2007, Su 2006) were killed quickly and the accumulation of dead donor termites around the treatment prevented conspecific s from being exposed to fipronil Saran and Rust (2007 ) reported that most mortality occurred within 1.5 m from a fipronil treated zone and that horizontal transfer was not a significant factor in the field. Chlorfenapyr was the only pro insecticide us ed in this study. It is activated by biological processes once it has entered a termite and prevents the production of energy throughout the body (Black et al. 1994). Compared with the other termiticides tested, termites exposed to chlorfenapyr were most likely to be healthy (completing the pheromone trail) or dead; pre lethal effects were less pronounced, likely due to the mode of action. In line with Shelton et al. (2006), higher concentration of chlorfenapyr caused higher mortality and recipient morta lity never reached 100%, even 14 d after exposure to donors. Chlorfenapyr produced slow acting mortality in this study suggesting that it may be effectively transferred in a field colony. I midacloprid performed poorly in this study with only 60% of d onors and recipients categorized as moribund or dead after 14 days perhaps due to the short exposure times and method of exposure Donor termites confined on imidacloprid treated soil appeared to be quickly affected and were largely motionless when time came for donors to be mixed with recipients. While donors exposed to
65 chlorantraniliprole, chlorfenapyr and fipronil readily tunneled into treated soil, those confined on imidacloprid treated soil did not engage in tunneling behavior which may have limited the dosage to a sub lethal level. The neurological effects of imidacloprid Breisch (2001) reported that termites exposed to imidacloprid have been shown to m exposure. The response to imidacloprid in the current study is best classified as sub lethal rather than pre lethal. This appeared to be happening, as stationary and near motionless termites during imidacloprid exposure were quite active when evaluatio ns were conducted and readily completed the synthetic pheromone trail. However, termites that are exposed to imid acloprid by tunneling through treated soil would likely be rendered motionless while still in contact with the treated soil and be unable to r ecov er enough to return to a colony site (Mulrooney and Gerard 2009). The principle of horizontal transfer of non repellent soil termiticides is a well accepted idea that has been reliably demonstrated in laboratory studies (Ibrahim et al. 2003, Shelton and Grace 2003, Rust and Saran 2006, Saran and Rust 2008, Spomer et al. 2008, Bagneres et al. 2009, Gautum and Henderson 2011). In the past few years, there has been disagreement as to what degree does transfer of non repellent soil termiticides occur in the field. Soil termiticides, unlike baiting systems, were not designed to eliminate colonies, but, depending on the characteristics of each compound, transfer may provide secondary termite mortality beyond direct mortality due to exposure in the field. In order to maximize transfer among termites, the termiticidal active ingredient should be easily acquired and relatively slow to kill an individual termite so the donor will live long enough to transfer it. The ability of donor termites
66 exposed to a ter miticide to follow a trail and subsequently transfer sufficient active ingredient to cause mortality in recipient termites one to several days after exposure suggests a likelihood that foraging termites can pick up a lethal dose and yet remain mobile long enough to return to a nesting site and transfer termiticide to nestmates. This study set up a scale of termite health that follows a logical progression from healthy to pre lethal to dead. Healthy termites follow pheromone trails, pre lethal termites ret ain limited function that degrades as they approach death. Understanding how non repellent termiticides affect the healthy and pre lethal behavior can help determine the likelihood for horizontal transfer to occur and harm the colony.
67 Figure 3 1. The experimental arena used to evaluate termite trail following ability. A. Small, inverted, 3 cm diameter souffl cup which held termites for the 30 sec orientation period. B. Introduction site : 4 cm diameter inverted sou ffl cup with a 1 cm x 1 cm x 1 cm triangle (1 cm base, 0.86 cm height) cut out of the lip to allow directed exit of termites down the synthetic pheromone trail. C. Synthetic pheromone trail, 2 cm of it extending into the introduction site.
68 Figure 3 2 The mean length of ti me to complete the synthetic pheromone trail by donor (blue) and recipient (red) termites in treatments with donor termites directly exposed to either chlorantraniliprole, chlorfenapyr, fipronil or imidacloprid at 50% of t he label rate (A, B) and 100% of t he label rate (C, D) for 1 h (A, C) and 4 h (B, D) over the duration of the experiment Replicates in which fewer than 10 termites completed the trail were omitted from this figure
69 A. B. Figure 3 3 The percentage of healthy (completed trail), pr e lethal and dead termites among donor termites directly exposed to either chlorantraniliprole, chlorfenapyr, fipronil or imidacloprid at 50% of the label rate (A) and 100% of the label rate (B) and recipient termites exposed to donor termites.
70 CHAPTER 4 RAPID CHANGES IN TUNNELING BEHAVIOR IN THE EASTERN SUBTERRANEAN TERMITE (RETICULITERMES FLAVIPES (KOLLAR)) (BLATTODEA: RHINOTERMITIDAE) AFTER EXPOSURE TO CHLORANTRANILIPROLE Tunneling is an essential behavior for subterranean termites that requires term ites to physically move soil particles. The construction of the tunnel network is accomplished in two ways: in loose soil, worker termites either push soil forward then press the soil particles from side to side with their body and head or, in compacted s oil, they pick up soil particles with their mandibles and deposit them outside of the tunnel (Ebeling and Pence 1957, Su and Scheffrahn 1992, Yang et al. 2009) Small soil particles are combined with feces an d saliva and used to fortify and smooth tunnel walls (Ebeling and Pence 1957, Krishna and Weesner 1969). The construction of tunnels is an intensive and elaborate process. Termites construct tunnels in a radiating, three dimensional branching pattern to decrease repetitive searching (Puche and Su 2001) The direction and intensity of foraging behavior is influenced by chemical (conspecific pheromones, wood and fungal odors) and environmental factors (soil compaction, soil particle size, moisture and thermal gradients (Smythe et al. 196 7, Grace and Wilcox 1988, Su and Scheffrahn 1992, Puche and Su 2001, Tucker et al. 2004, Green et al. 2005, Su 2005b, Rust and Saran 2006) Any disturbance or interruption in tunneling activity can severely affect the termite colony as tunnels lead to nesting and foraging sites (Krishna and Weesner 1969, Gautam and Henderson 2011) A significant disturbance in normal tunneling behavior occurs when termites contact soil treated with a non repellent soil termiticide s the predominant method of termite control used around structures in the United States ( Su et al. 1982, Su and
71 Scheffrahn 1990b, Potter 2011). Termites are unable to detect non repellent termiticide in soil and readily tunnel into it using their mouthparts as they are constructing explorat ory tunnels (Shelton and Grace 2003, Spomer and Kamble 2011). Tunneling into treated soil can lead to a lethal dose of a termiticide and the accumulation of dead termites in a tunnel can cause nestmates to abandon or seal the tunnel against further entry (Fei and Hender s on 2005, Su 2005 a ). A recently registered non repellent termiticide, containing the anthranilic diamide chlorantraniliprole (trade name Altriset; Syngenta AG, Basel Switzerland), is known to affect the muscle function of insects, including termites (Hirooka et al. 2007, Lahm et al. 2009, Saran et al. 2011), which may affect tunneling ability. Chlorantraniliprole is toxic through oral and dermal exposure even at very low doses (Gautam and Henderson 2011, Mao et al. 2011, Saran et al. 2011, Buczkowski et al. 2012, Neoh et al. 2012 a ) The primary objective of this experiment was to de termine the speed of action for chlorantraniliprole and how different lengths of exposure to treated soil affect the occurrence and quality of tunnel building behavior in the eastern subterranean termite, Reticulitermes flavipes We hypothesized there wou ld be a reduction in soil excavation and tunneling behavior with increased length of chlorantraniliprole exposure. Secondarily, we examined effects of chlorantraniliprole exposure on termite morbidity and mortality after limited confinement on treated soi l and expected there to be low mortality yet high morbidity among termites. Methods and Materials Foundation and M aintenance of L aboratory C olonies Reticulitermes flavipes were field collected from University of Florida campus in Gainesville, FL. Wood sta kes were placed in ground throughout campus and sites with
72 termite activity were selected for trap placement. Termite traps consisted of a polyvinyl chloride bucket (20 cm in height by 20 cm diameter; 811192 4, Ventura Packaging Inc., Monroeville, OH), wi th 11 6 cm diameter holes drilled in the sides and the base, allowing termite access beneath the surface. Buckets were placed vertically in the ground to a depth of roughly 19 cm, and covered with a polyvinyl chloride lid. Two rolls of single faced corr ugated cardboard (20 cm in length by 10 cm in diameter) were dipped in water and placed into the bucket side by side as a food source. From spring to fall, cardboard rolls containing termites were collected weekly from traps. Termites were removed from t he cardboard rolls and maintained at room temperature (23 C), 80% RH, and a photoperiod of 0:24 (L: D) hours. Colonies were maintained in plastic boxes (27 by 19 by 9.5 cm; Sterilite Co., Townsend, MA) provisioned with soil, wooden blocks (12.7 cm wide by 12.7 cm tall by 2.5 cm thick) and moistened sponges (O Cel O; 3M Co., Maplewood, MN) to provide termites with food and moisture. Traps greater than 100 m from each other are assumed to represent different colonies and were maintained in different contain ers in the lab. Termites were kept in the laboratory < 45 d before use in experiments. Experimental Units Three colonies were sub sampled three times in this experiment for a total of 180 replicates per treatment. 51 termites were used per replicate (50 workers, 1 soldier) for a total of 18,360 termites. Soil Bonneau fine soil used in this experiment was collected near the Entomology and Nematology Department on the campus of the University of Florida (Gainesville, FL). Soil was sieved through a #35 ( 0.5 mm) stainless steel test sieve (Fisher Scientific
73 Inc., Hampton, NH) and baked at 150 C for 24 h for sterilization. Soil was analyzed by Institute of Food and Agricultural Sciences Analytical Services Laboratories Extension Soil Testing Laboratory at the University of Florida and found to contain 1.48% organic matter, 6.8 pH and high levels of phosphorus (53 ppm) and magnesium (51 ppm). Term iticide Commercially formulated chlorantraniliprole (20% AI) was provided by Syngenta AG (Basel Switzerland) a nd 1,000 ppm stock solutions of chlorantraniliprole were prepared by adding 0.25 ml of concentrate to 49.75 ml water. Stock solutions were prepared just prior to use in experimentation. Stock solution (15 ml) and water (15 ml) were combined with 270 g of dried and sieved soil to achieve the desired concentration (50 ppm) in soil with 10% moisture. Dried and sieved soil (270 g) was mixed with 30 ml of water to make untreated control soil. Experimental Setup The tunneling apparatus consisted of a 100 mm x 15 mm (diameter by height) Petri dish (Fisher Scientific Inc., Hampton, NH) with a 32 mm circular hole in the center. A plastic cylinder (85 mm tall) with a 32 mm diameter opening was placed flush into the hole in the Petri dish and the outside juncture of the Petri dish and the tube were sealed with hot glue (Fig. 4 was filled to the top with 100 g of 10% soil moisture. Exposure and E xperimental P rocedure Groups of 51 eastern subterranea n termites (50 workers and 1 soldier) were gently aspirated from colony boxes prior to use and exposed to either untreated soil (0 ppm) or soil containing 50 ppm of chlorantraniliprole (label rate), in 120 ml souffl cups (Dart Container Co., Mason, MI) co ntaining 25 g of compacted soil with 10% moisture
74 (wt:wt). Termites were expo sed for 10, 20, 30 or 40 min and then gently aspirated from the cups. Any soil particles that were aspirated with termites were removed from transfer cups and discarded. After t ermites were removed from exposure cups containing control or treated soil, the insects were subject to a variety of post exposure times (30 min, 60 min, 120 min and 180 min) before being placed onto the tunneling apparatus. Termites were transferred from exposure cups and placed into different 60 ml souffl cups containing plain, moistened filter paper (#2 qualitative circles, Whatman PLC, Kent, UK) and were sealed with a tightly fitting lid (#200 PCL, Dart Container Corp., Mason, MI). After the desired p ost exposure time had elapsed, termites were moved to the tunneling apparatus by camelhair brush. Termites were introduced into the Petri dish that was then closed with its cover. Tunneling apparatuses were kept at constant conditions (23 C, 80% RH, photo period of 0:24 (L: D) hours). Termites were undisturbed for 24 hours prior to evaluation. Evaluation of T ermite H ealth and T unneling Behavior Evaluations were conducted by determining the health of these termites found above the surface of the soil. These termites were gently flipped using featherweigh t forceps and allowed 10 sec to right themselves. Those able to do so were categorized movement, then it was c onsidered 'dead'. Slight movement or twitching was deemed All soil excavated above the surface of the dish was scraped and weighed as a measure of tunneling volume. Other data re corded included the number of tunnels that were constructed from the surface of the tube, whether termites were able to reach the
75 bottom of the tube, the maximum depth reached, and the number of termites found below the surface of the dish. S tatistical A nalyses All data analyses were performed using JMP Pro 10 statistical software (SAS Institute 2012). Comparisons were made between treatments (chlorantraniliprole and control) and factors (exposure times, post exposure times) for all variables (morbidity, mortality, weight of soil excavated, number of tunnels excavated and depth of tunnels) using analysis of variance (ANOVA) of means. Percent morbidity and mortali ty were subjected to an arcsine square root transformation while number of tunnels constructe d and weight of soil excavated were normalized with a square root transformation. The level of significance was set at 0.05 for all statistical tests. Significant means were test. For overall measure of health, termites cate gorized Results Health of T ermites Neither exposure ti me (d.f. = 3, F = 1.16, P = 0.3251) nor post exposure time (d.f. = 4, F = 0.87, P = 0.4843) significantly affected the number of dead and moribund termites across the control replicates. The majority of control termites (84.2%) were found below the surfac e of the soil in the tunneling apparatus at the time of evaluation compared to 18.3% of termites exposed to chlorantraniliprole The effect of treatment was strong as exposure to chlorantraniliprole treated soil significantly increased termite morbidity (d .f. = 1, F = 1126.66, P < 0.0001) and mortality
76 (d.f. = 1, F = 184.96, P < 0.0001) compared to control groups (Fig. 4 2). There were over 23 times as many dead and moribund termites in the chlorantraniliprole group. Even the shortest exposure to chlorantr aniliprole (10 min) resulted in over 20 times the number of dead and moribund termites compared to control. The number of dead and moribund termites increased with longer exposure to chlorantraniliprole, with 30 and 40 minute exposure having significantly more affected termites than 10 and 20 minute exposures (d.f. = 3, F = 17.66, P < 0.0001). The majority of affected termites were categorized as moribund (89.0%) as opposed to dead (11.0%) and no exposure period resulted in more than 6.7% mortality. Post exposure time did not significantly affect termite health amongst termites exposed to chlorantraniliprole (d.f. = 4, F = 2.18, P = 0.0730). Measures of Tunneling Behavior Tunneling occurred in 100% of control replicates (n=180). Neither exposure time nor post exposure time were significant in any measure of tunneling behavior for untreated termites. Exposure to chlorantraniliprole treated soil significantly decreased all measures of tunneling behavior compared with the control group. Across the experim ent, termites exposed to chlorantraniliprole tunneled less than half as often as control termites (d.f. = 1, F = 191.34, P < 0.0001), constructed fewer tunnels per replicate (d.f. = 1, F = 377.74, P < 0.0001), tunneled to 1/3 the depth (d.f. = 1, F = 736.0 4, P < 0.0001) and reached the bottom of the test tube arena at 1/8 the rate of control termites (d.f. = 1, F = 1125.62, P < 0.0001). On average, the control termites excavated over ten times the weight of soil as the chlorantraniliprole exposed group (d. f. = 1, F = 1515.28, P < 0.0001).
77 As observed with termite health, even brief exposure (10 min ) to chlorantraniliprole significantly reduced all tunneling behavior compared with the control group (Fig. 4 3 ) Chlorantraniliprole exposed termites tunneled in nearly 1/3 fewer reps (d.f. = 1, F = 19.9, P < 0.0001) ( (Fig. 4 3 ) reached the bottom of the tube 1/3 as often (d.f. = 1, F = 108.31, P < 0.0001) (Fig. 4 4), and averaged one half the tunneling depth of the control group (d.f. = 1, F = 66.97, P < 0.00 01) (Fig. 4 5) The control group was able to excavate over 5 times the weight of soil (d.f. = 1, F = 223.53, P < 0.0001) (Fig. 4 6) and construct over 2.5 times the number of tunnels (d.f. = 1, F = 55.20, P < 0.0001) (Fig. 4 7) observed in the chlorantra niliprole treatments. The significant reduction in tunneling behavior between chlorantraniliprole exposed termites and control groups cont inued at 20 30 and 40 min exposure times. Groups of exposed termites tunneled in fewer reps (d.f. = 1, F = 169.4 0, P < 0.0001), reached the bottom of the tube significantly less often than non exposed termites (d.f. = 1, F = 1197.15, P < 0.0001), averaged a significantly lower maximum tunneling depth (d.f. = 1, F = 921.84, P < 0.0001), excavated less soil (d.f. = 1, F = 1487.00, P < 0.0001) and constructed fewer tunnels (d.f. = 1, F = 339.80, P < 0.0001). Within chlorantraniliprole exposed termites, exposure time was a significant factor as increased exposure time reduced tunneling behavior (reps in which tunneling o ccurred: d.f. = 3, F = 6.41, P = 0.0004; percentage depth reached: d.f. = 3, F = 13.90, P < 0.0001; percentage of reps in which tunnels reached the bottom of the tube: d.f. = 3, F = 7.78, P < 0.0001; number of tunnels constructed d.f. = 3, F = 11.48, P = 0 .0025; weight of soil excavated: d.f. = 3, F = 11.48, P < 0.0001). Termites exposed to
78 ch lorantraniliprole for 30 min tunneled less often and less effectively than termites in the 10 and 20 min exposure groups. The percentage of dead termites was very simi lar between 10, 20 and 40 min exposure time groups ; the maximum number of dead from one replicate was 26%. Termites exposed for 30 min had significantly higher percentage of dead termites than the other exposure time groups. In 18.3% of chlorantranil iprole exposed replicates, 60% of termites were categorized as healthy, yet no tunneling occurred. In 3.8% of replicates, tunneling occurred while less than 40% of the termites were deemed healthy. Discussion The natural behavior of R. flavipes as ev idenced by the control group, was to excavate soil and continue tunneling beneath the surface until they reached the bottom of the tunnel ing apparatus. The termiticide chlorantraniliprole limited the ability of termites to tunnel normally, even with just brief 10 min exposure, the effects of chlorantraniliprole persisted after contact with treated soil, and there was no observed recovery among termites after exposure. Mortality of termites exposed to chlorantraniliprole treated soil was 7% at each exposure length. Other authors have also repor ted low mortality rates 24 h after limited exposure, including < 10% after 5 min exposure on treated sand (Saran et al. 2011) and < 1% (Gautam and Henderson 2011) and < 10% (Buczkowski et al. 2012 ) for 1 h exposure on chlorantraniliprole treated sand. The persistence of chlorantraniliprole within termites combined with low mortality may allow for an increased amount of termiticidal transfer between healthy and exposed termites (Gautam and Henderso n 2011, Buczkowski et al. 2012, Neoh et al. 2012 a ).
79 Despite being known as a slow acting insecticide in terms of the termite mortality (Gautam and Henderson 2011, Mao et al. 2011) chlorantraniliprole significantly reduced all termite tunneling activity with very short exposure of the insects to the termiticide. The speed of action, represented by the length of time required to chlorantraniliprole to elicit persistent, irrev ersible changes to termite behavior, was high. This was perhaps best illustrated by the weight of soil excavated after 10 min of chlorantraniliprole exposure. Despite 74% of termites categorized as healthy, termites were able to excavate only 1/5 the wei ght of soil as the control group. These results clearly indicate that termite tunneling behavior was rapidly affected by chlorantraniliprole, even before clear signs of intoxication were observed. These results support the classification of chlorantranil iprole as a compound with some fast acting properties ( Buczkowski et al. 2012, Neoh et al. 2012 a ). However, these authors did not explicitly define speed of action. Based on the tunneling depth and amount of soil excavated in this experiment, a much shor ter exposure time than those used by Buczkowski et al. (2012) and Neoh et al. (2012 a ) can affect termites and prevent normal feeding behavior. Previous studies regarding chloran traniliprole focused on the termite mortality and noted the delayed mortality after chlorantraniliprole exposure (Yeoh and Lee 2007, Gautam and Henderson 2011, Mao et al. 2011, Saran et al. 2011) Based on the limited tunneling ability of termites reported i n this study, despite > 47% of termites able to right themselves at all exposures, pre lethal behavioral effects rather than termite death may have prevented soil penetration.
80 S tudies by Yeoh and Lee ( 2007 ) Gautam and Henderson ( 2011 ) Mao et al. ( 2011 ) a nd Saran et al. ( 2011) did not acc ount for, or did not address the immediate effect of chlorantraniliprole exposure on termite behavior reported here. Mortality is delayed because chlorantraniliprole inhibits muscle function, with mandibular function as the first area to be affected, and ultimately paralysis ensues (Hirooka et al. 2007, Lahm et al. 2009). While delayed mortality is a defining characteristic of this compound, the fast acting behavioral effects may be more important in how chlorantranilipr ole functions in protecting a structure. The inability of termites to tunnel towards a structure and consume wood has the same effect as rapid mortality in terms of structural protection from termites. Termites in these experiments were exposed to chloran traniliprole in a less direct way than expected for field colonies. The route of exposure to chlorantraniliprole in this experiment was likely from grooming treated soil particles from the tarsi into the buccal cavity because termites were confined on top of treated soil. Preliminary tunneling activity may have occurred in the exposure cups, but there were not any tunnels constructed nor any soil visibly excavated; likely due to the short length of time in the exposure cups. In the field, soil termiticid es are applied as a band around a structure and termites tunnel directly into the soil, as opposed to being confined on top of treated soil (Miller 2001, Shelton and Grace 2003). During tunneling, termites pick up soil particles containing the insecticide Minimal tu nneling occurred in some chlorantraniliprole replicates Although it is not clear why this occurred, it may be that delayed effects started somewh ere between 180 min and 24 h after exposure. This would have allowed tunneling behavior to beg in
81 at a normal, or near normal, rate which then slowed until it stopped. Also, because t ermites exposed to non repellent termiticides do not necessarily pick up equal doses (Shelton e t al. 2006, Forschler 2009), t here may have been a few individuals who d id not pick up a sufficient dose of chlorantraniliprole to have reduce d tunneling ability, while other termites did and were unable to assist in soil excavation. Given the narrow circumference of constructed tunnels, termites could not be extracted from tu nnels in the tunneling bioassay areas without compromising their health. Termites below the surface of the assay at the time of evaluatio n were assumed to be healthy, because they were, or at l east had been behaving apparently normally. The method for d etermining health of termites in these experiments provided the most conservative number of dead and moribund termites because any cadavers that might have existed below the soil surface were not counted. If there was any delayed onset effect of chlorantr aniliprole exposure, there may have been substantial numbers of dead and moribund termites that had tunneled, or followed tunneling conspecifics below the surface before being rendered ineffective. Post exposure time was not a significant factor in termi te health or tunneling ability in this study, suggesting that the effects of chlorantraniliprole arise quickly in termites. Some delayed effects may take longer than the maximum post exposure time (180 min) and, therefore, were undetectable in this study. The purpose of soil termiticides is structural protection. Repelling or killing termites quickly are not the only means to achieve successful structural protection. The ability of termites to tunnel and consume wood is what determines the threat a colon y poses t o a structure. Pre lethal changes in behavior caused by chlorantraniliprole are
82 likely to prevent damage to structures before the termites are actually killed by the insecticide. The substantial reduction in tunneling ability observed in this st udy combined with previous studies demonstrating reduced food consumption (Buczkowski et al. 2012, Neoh et al. 2012 a ) suggests that chlorantraniliprole has the potential to protect a structure within a short time after a relatively reduced exposure of t he termites to the termiticidal active ingredient.
83 Figure 4 1. Aerial view (left) and side vi ew (right) of the tunneling apparatus used to evaluate tunneling behavior.
84 Figure 4 2 The percentage of de ad (A), moribund (B ), and healthy term ites (C ) separated by trea tment and length of exposure for t ermites exposed to chlorantraniliprole (blue bars) and untreated control soil ( orange bars )
85 Figure 4 3 P ercentage of replicates in which tunneling occurred for chlorantraniliprole expos ed (blue bars) and control (orange bars) R. flavipes after 10, 20, 30 and 40 min exposure times
86 Figure 4 4. T he proportion of replicates in which tunnels reached the bottom of the tube for chlorantraniliprole exposed (blue bars) and control (orange b ars) R. flavipes after 10, 20, 30 and 40 min exposure times.
87 Figure 4 5. D epth o f soil reached by tunneling for chlorantraniliprole exposed (blue bars) and control (orange bars) R. flavipes after 10, 20, 30 and 40 min exposure times
88 Figure 4 6. W eight of soil (g) excavated by tunneling for chlorantraniliprole exposed (blue bars) and control (orange bars) R. flavipes after 10, 20, 30 and 40 min exposure times
89 Figure 4 7. M ean number of tunnel s constructed per replicate for chlorantra niliprole exposed (blue bars) and control (orange bars) R. flavipes after 10, 20, 30 and 40 min exposure times
90 CHAPTER 5 QUANTIFYING DRYWOOD TERMITE ACTIVITY BEFORE AND AFTER LOCALIZED TREATMENTS WITH TERMIDOR DRY Drywood termites cause millions of dollars of damage annually across the United States with heavy distribution in southern, western and coastal areas (Rust et al. 1988, Scheffrahn et al. 1988, Derrick et al. 1990, Scheffrahn and Su 1999, Potter 2011). Drywood termites nest within wood whic h makes them difficult to detect and control (Lewis 2003, Potter 2011, Hickman and Forschler 2012). Drywood termites infest structural beams, window frames, wood floors, furniture and any other sound, dry wood they encounter and excavate galleries by eati ng spring and summer wood along and across grain (Smith 1930, Krishna and Weesner 1969, Rust et al. 1988, Scheffrahn et al. 1988, Rust and Su 2012). The size and shape of drywood termite galleries vary, but they are often smooth and close to the surface of the wood. The gallery system is unpredictable with some parts narrow to only accommodate one termite at a time, while other parts may be open cavities which suddenly give way to a narrow tunnel that may reach another cavity or not (Hickman and Forschler 2012). There are multiple species of drywood termites, including Incisitermes snyderi I. minor I. schwarzi Kalotermes approximates and Cryptotermes brevis but determining which species is present is usually unimportant as treatment protocols are the s ame for all species and it may not be possible to make a species determination prior to treatment (Scheffrahn 1988, Scheffrahn and Su 1992, Scheffrahn et al. 1997, Potter 2011). When an infestation is believed to be isolated or localized, a spot treatmen t may be administered. Termiticides used in spot treatments include non repellent foams and and
91 ingredient. Although the products have changed, this method has been use d for nearly 100 years (Smith 1930, Rust et al. 1988, Potter 2011). Drywood termites are confined to a network of connected galleries, so a product that hits part of the gallery may eventually eliminate the colony. The unpredictable way in which drywood termites excavate galleries makes treatment difficult, because drilling where activity is detected may not intersect a gallery (Rust et a l. 1988, Scheffrahn et al. 1997 ). Application of product to wood without reaching a gallery results in a failed treatm ent and wasted termiticide (Hickman and Forschler 2012). Lewis (2003) reported an efficacy range from 0 100% control for all localiz ed drywood termite treatments. Insecticides used against drywood termites are intended to coat the gallery, not necessarily contact termites directly. Fipronil, a heavily used active ingredient for subterranean termite control, has recently been reformulated and impregnated in a micro cellulose product (Termidor Dry; BASF Corp., Ludwigshafen, Germany) intended to treat drywo od and subterranean temites. The dry formulation of this product is intended to coat the gallery system upon introduction, and to last over time without degrading or spoiling, providing residual protection. A thorough inspection is essential to diagnosin g and treating any drywood termite infestation. Many tools for detection are commercially available and function in a variety of ways (e.g. microwave radiation, acoustic emission, drilling resistance and X ray technology) but are not often used by pest co ntrol operators (Rust et al. 1988, Thoms 2000, Lewis 2003, Woodrow et al. 2006, Potter 2011). The commercially available termite detector TermaTrac T3i (TermaTrac, Queensland, Australia) has a radar function which allows gathering of real time evidence of locations and relative termite
92 activity (movement) levels, and has been shown to reliably detect termites and other insects in wood by using microwave radiation (Peters and Creffield 2002, Mankins 2004, Hickman and Forschler 2012). The objective of this study was to quantify drywood termite activity in existing localized drywood termite infestations using the Termatrac. Galleries were treated with a micro cellulose product formulated with fipronil (Termidor Dry; BASF Corp., Ludwigshafen, Germany) and act ive spots were reevaluated for activity six and twelve months after treatment. I hypothesized that using the TermaTrac would allow for effective diagnosis of the scope of localized infestations and allow for successful drill and treat procedures without d estructive sampling. Methods and Materials Reading the Termatrac The PDA display featured three components (Figure 5 1): 1) A n adjustable gain setting that amplified the activity reading as the gain was increased. In order to prevent activity readings fro m exceeding the maximum value (10 notches) the gain was set at large horizontal graph that displayed the activity readings. As activity was detected, a blue bar extended from left to right indicating the strength of the activity (and a si ngle spike). When no activity was detected, the graph was empty. Activity levels were measured by how far each extending bar (referred to as a spike) reached on the graph, from > 0 to 10 notches, with 10 being the highest possible activity level. Any time the blue activity bar receded (towards 0) and then registered a positive gain from that spot was considered a spike of activity. The maximum level a single spike reached during the 30 sec reading was recorded for each spot. 3) T he Term atrac PDA h as an accelerometer reading that functions much like the activity indicating graph but relates
93 to the movement of the Termatrac sensing device. Any movement displayed in the accelerometer rendered activity readings invalid as the difference between termite m ovement and movement of the sensing device cannot be separated. All readings used in this experiment occurred without any movement in the accelerometer bar. Calibration of A ctivity Readings Preliminary experimentation with the Termatrac was conducted in order to determine how the device would respond to termite movement in a controlled setting. The eastern subterranean termite, Reticulitermes flavipes (Kollar), was used as the test termite due to the lack of an available drywood termite colony. The Ter matrac sensing device was lodged in a block of Styrofoam to keep the detector facing upwards and parallel to the floor. Three substrates of different thicknesses were used to separate the sensor from the termite(s) and termites were constrained on top of the substrate with a small inverted souffl cup (30 ml, 4 cm diameter, Dart Container Co., Mason, MI) to remain in the detected area. The smallest distance was a single sheet of white copy paper (Office Depot, Boca Raton, FL) placed directly on the sensor while the medium substrate consisted of a single white pine square (12.7 cm length and width, 2.5 cm thick) which kept the termite(s) 2.5 cm from the sensor. The greatest distance between termite(s) and the sensing device was achieved by stacking two wo oden squares on top of each other with the termite(s) 5 cm above the sensor. Groups of one, two and five termites were tested ten times each for every substrate. PDA activity readings (number of spikes, maximum spike and level per 30 sec) were conducted and evaluat ed in the manner described below
94 Location and D escription of I nfestations Apartment buildings and houses in Jacksonville, FL, Middleburg, FL, Orange Park, FL and Gainesville, FL were used in this experiment. Several research sites were locat ed in areas with heavy drywood termite infestations. Each site was unique and the infestation histories were largely unknown but had shown signs of a drywood termite infestation prior to use in this experiment (Table 5 1). There were a total of six struc tures used in this experiment. Five structures contained one site each while the house in Gainesville, FL, contained four distinct, disconnected sites. Site 1: Apartment staircase; Jacksonville, Florida. This site consisted of a wooden staircase in a de cades old house that had been converted into apartments. Termite damage and fecal pellets were present on steps and a beam of wood running alongside the steps and live termites were found when a damaged spot was pressured. Site 2: Wood flooring; Jack sonville, Florida. This site was located in an apartment and the infestation was only detected in a single floor beam next a baseboard. Site 3: Windowsill; Jacksonville, Florida. Termite activity was detected in a single window frame in this house. Site 4: Door frame; Orange Park, Florida Termite activity at this site was associated with a doorframe at the front of the house. Site 5: Kitchen Cabinets; Middleburg, Florida. This site consisted of a wall mounted kitchen cabinet inside a house. Sites 6 9: Gainesville, FL. This house was found to have several disconnected pockets of termite activity, and each pocket was considered a separate site for this study. Site 6 & 7: House posts: Two, 30 cm diameter wooden posts located on the exteri or of the house. Site 8: Outside wooden wall panels; Gainesville, Florida. An exterior site consisting of a large (1.25 m x 2 m) section of wood paneling on the outside wall of the home.
95 Site 9: Wood flooring and baseboard; Gainesville, Florida. Th is site was comprised of a corner section of a bedroom (1.5 m by 1 m), including wood floorboards and a baseboard. Inspection Protocol A termite detection device, Termatrac T3i, was used to locate drywood termite infestations. The radar function of the Te rmatrac T3i was used to obtain activity readings starting at the place where evidence of a drywood termite infestation (swarm, visible damage, or fecal pellets) was observed. Readings were taken every 10 cm forming a grid of measured spots surrounding the location with observed termite evidence. The T3i device was placed on a tripod or other supporting device whenever possible. Areas where termite movement was detected were marked with vinyl electrical tape (3M Corp., Maplewood, MN) and assigned a number A 30 sec video recording of the Termatrac PDA display showing movement was made for each numbered spot to allow for visual and quantitative comparisons of the same spot over 1 year. Chemical Treatment Holes were drilled, without the use of a resistog raph, using a 1.0 mm circular drill bit to a depth of 8 cm into wood within a 5 cm radius of areas with termite activity readings using the Termatrac. Fecal pellet kick out holes were used when available. Drilled holes that appeared to contact a gallery were used during termiticide injection. Termidor Dry (BASF Corp., Ludwigshafen, Germany), a micro cellulose action bulb, was used in the experiment. The label a llowed for 0.1 g to 1.0 g per injection point, corresponding to 3 to 30 pumps of product according to the product label. Preliminary use of Termidor Dry has demonstrated that the actual weight of
96 product emitted per puff would be highly variable and there was often far less released than expected. Holes that seemed to have contacted a gallery were be given between injection site. The makeup of the activity grid and the amou nt of wood involved in the infestation were used to help determine how many treatment spots per infestation. Injection sites were be labeled and covered with tape (3M Corp., Maplewood, MN) to minimize disturbance to the colony. Post Treatment Inspection s Infestations were revisited periodically and all sites were evaluated 12 months after the initial visit when a final evaluation of the control of the drywood termite infestations was conducted. At each follow up visit, 30 sec radar readings were taken again using the Termatrac T3i at each labeled site. For each reading, radar detected activity was recorded as either present or absent. When activity was present, the total number of spikes for the 30 sec reading and maximum spike during that period were recorded. Activity levels were classified and assigned a number: 0 = no activity (no activity spikes); 1 = very low (spikes reaching 0.25 notches); 2 = low (spikes reaching to 0.75, with the majority of spikes b etween 0.25 and 0.50 notches); 3 = m edium (spikes reaching to 2, with the majority of spikes between 0.75 and 1.25 notches); 4 = high (spikes reaching to 3.5, with the majority of spikes betwee n 1.25 and 2.5 notches); 5 = very high (spikes reaching to 10 with the majority of spikes above 2 .5 notches).
97 Because drywood termites are also susceptible to daily temperature fluctuations, which may inhibit the growth of a colony and its activity, all measurements were done as close to the original sampling time of the day as possible. Statistica l Analyses All data analyses were performed using JMP Pro 10 statistical software (SAS Institute 2012). Comparisons were made between the number of spikes, maximum spike, activity levels and the presence or absence of activity at each spot across visits. The number of spikes and maximum spike per 30 sec reading were square root transformed and subjected to an analysis of variance (ANOVA) of means. The level of significance was set at 0.05 for all statistical tests. Significant means were separated using test. Results Calibration The frequency of activity (d.f. = 2, F = 87.67, P < 0.0001) and mean maximum spike (d.f. = 2, F = 87.67, P < 0.0001) per 30 sec recording increased as more termites were used in preliminary activity readings (Fig ure 5 2). The thickness of the substrate was also significant as termites closer to the sensing device registered more spikes per 30 sec reading (d.f. = 2, F = 662.27, P < 0.0001) and a greater maximum spike (d.f. = 2, F = 65.71, P < 0.0001). The activit y levels recorded in the laboratory and in the field were very closely related to the number of spikes per 30 sec reading (Figures 5 2 and 5 3).
98 Field Sites Of the nine locations used in this study, there were 80 spots with termite activity measured duri ng the initial visits with an average of 48.52 1.42 spikes per 30 sec reading at each spot with a maximum spike of 1.63 0.13. The number of active spots found per site ranged from two to 23, while the number of injection points (between two to 21 per site) and treatment amount (between 161.5 mg and 2443.8 mg) reflected the large difference in the number of active spots found at each site. However, the large variability in the volume of Termidor Dry released with each puff was neither intentional nor re lated with the specifics of any particular site (Table 5 1). T he Termatrac detected a significant decline in termite activity within one year of treatment. As time increased post treatment, the mean number of spikes (d.f. = 2, F = 384.09, P < 0.0001) and the maximum spike (d.f. = 2, F = 156.49, P < 0.0001) significantly decreased. Two sites, with 9 spots between them, were i naccessible during the 6 mo evaluations so 71 spots were re measured. Nineteen out 71 of those spots showed activity when re measure d 6 mo post treatment and there were 1/8 of the average number of spikes recorded per 30 sec reading (6.79 1.92) and 1/10 the average maximum spike. At the 1 yr evaluations, 5 out of 80 spots were found with termite activity. The average number of spik es per 30 sec reading (1.99 0.95) and maximum spike (0.04 0.03) were substantially lower than the initial measurements (Figure 5 3). For locations with activity 6 mo post treatment, there was average was 34.43 5.25 spikes per 30 sec reading with a maximum of 0.86 0.21 spikes (n=14). Among the five active spots at the 1 yr mark, there was an average of 31.80 7.23 spikes per
99 30 sec reading with a maximum of 0.65 0.34 spikes (n=5). Both of these were less than the initial measurements. Along with a decrease in the number of spikes and maximum spike per 30 sec reading, activity levels significantly decreased as time increased post treatment (d.f. = 2, F = 384.09, P < 0.0001). During initial inspections, the mean activity level was medium (2.94 0.11). At 6 and 12 mo inspections, activity levels were classified as very low at 6 mo (0.45 0.12) and 12 mo (0.12 0.06) after treatment. I n i solating sites that were activ e at the 6 and 12 mo post treatment measurements there was also a significa nt decline in the level of activity as time increased post treatment (d.f. = 2, F = 8.19, P = 0.0014); there was a medium high level activity at those same sites initially (3.57 0.23) compared to low at 6 mo (2.29 0.30) and 12 mo (2.00 0.55). The fiv e active spots found at 1 yr were distributed in three out of nine locations. Locations three and six had two active spots each but there were far fewer spikes and lower maximum spikes at 1 yr than there were at the same spots during initial inspection. Location five had one spot with activity that never decreased throughout the duration of this study, and after re measuring for activity 1 yr post treatment, this site was found to contain new areas with termite activity an d this site produced a swarm 15 m o after treatment. One other site (#9) produced a swarm but it was not determined if the alates were associated with the treatment area or swarmed in from elsewhere in the structure. Discussion This research represents the first quantifiable, non destru ctive, method for evaluating localized drywood termite treatments with a microwave radiation device. Other researchers have used the Termatrac to locate and determine the absence or
100 presence of termites after treatment, but the treated wood was dissembled afterwards to count termites (Hickman and Forschler 2012); an impractical method in this study because infestations were located in houses. This study is an important step in bringing the findings of laboratory experiments to practical situations encount ered by pest control operators as the Termatrac T3i gains popularity as a drywood termite detector. The microwave detection device used in this study, Termatrac T3i, was a newly released device that performs better than the devices cited in previous studi es (Peters and Creffield 2002). In addition, the ease of use of the Termatrac device (lower sensitivity, accelerometer to distinguish between device and termite movement, high portability, and readability) along with the ability for it to detect activity through materials other than wood make it a more attractive option than an AED for use in the field. The distance between the Termatrac sensing device and termites in the gallery is not known in the field, but using the Termatrac activity readings taken in the laboratory help illustrate what may be occurr ing out of view in the field. In calibration studies, t he Termatrac was shown to reliably detect a single termite within 25 mm of the surface of wood. The distance between termites and the Termatrac sen sor is more of a determination of how well termites can be detected than the number of termites. It was much more difficult to detect five termites through 50 mm of wood than to detect one termite through 25 mm of wood. It is worthwhile to translate acti vity readings in a known setting into the unknown, for example, during the initial inspection of site 8, spot #16 had 61 spikes of activity with a maximum spike of 1.5 notches and an activity level of 3. The se values match up nearly identically with the a verage readings for five termites being detected through 25 mm of wood. There may be a linear relationship between
101 substrate thickness (and type) and the number of termites present that can be elucidated with an extensive exploration of this concept. In terms of detection of termites within wood, Hickman and Forschler (2012) reported 9.5% false negatives when using the Termatrac T3i on boards infested with drywood termites. This number may be higher in the field due to the variety in substrates and dist ance of galleries from the surface of the wood, however, the use of a more tightly gridded inspection method (every 10 cm compar ed to every 15 cm used in this study), may have minimized the likelihood that false negatives were as prevalent The Termatrac has also been reported to have decreased reliability in areas with high moisture, but moisture was not believed to be a factor in the infestations used in this study. The temperature inside drywood termite galleries is closely correlated to the tem perature of the external environment and three out of nine sites used in this study were located outdoors and thus were subjected to a large range of temperatures. Termites are sensitive to temperature extremes and colonies have been shown to move in res ponse to temperature changes, often in predictable patterns ( Minnick et al. 1973, Cabrera and Rust 1996, Schef frahn et al. 1997, Lewis et al. 2011) Sites were measured as close to the same time of day as possible at each visit, but outdoor inspections in the wi nter could not be avoided. One year inspections were conducted in the same month as the initial inspe ction so seasonal weather differences did n ot affect final evaluations. T here were many factors that may have affected the likelihood for treatment success. Most importantly, it could not be determined whether the spread of active
102 sites at a single locati on represented one or multiple colonies. E limination of termite activity would require termiticide injection into each disconnected gallery system if multiple colonies exist The presence of multiple colonies cannot be readily determined without destruct ive sampling or the use of X ray technology. Unfortunately, the likelihood of multiple colonies in close association with each other is high for drywood termites that swarm indoors as they are poor fliers and not aided in distribution by wind. Assuming there is a single colony, there are variables that may have influenced the application of product and measuring of termite activity. Penetration of the gallery system and introducing termiticide is essential to achieving any measure of control. The use o f a resistograph has been shown to increase the likelihood of contacting the gallery (Lewis and Haverty 1996, Hickman and Forschler 2012) but a resistograph was not available for use during this study. However, heavy pressure from the puffing application device was felt when a drilled hole did not intersect with a gallery. This rudimentary information provided the operator with an indication of whether a gallery was contacted. The applica tion of termiticide and amount of product released was likely influenced by the force the operator applied on the puffing device, the volume of termiticide in the vial and the angle at which the device was held. Without the use of product markers and dest ructive sampling, I was unable to determine the distribution and range of product inside the gallery system (Hickman and Forschler 2012). Even when the gallery was intersected during drilling, the gallery may be narrow and any product injected may not rea ch other sections of the gallery. Along with constricted
103 areas due to gallery excavation, the physical presence of termites may impede the spread of termiticide during injection. In a study quan tifying drywood termite control it is important to determin e the time needed to obtain control after treatments were made Despite targeted application of termiticide, termite populations declined relatively slowly; nearly 27% of spots showe d activity 6 mo after treatment and 6% after 1 yr The variability in th e amount of Termidor Dry applied to each treatment site may explain why ther e was still so much activity 6 mo after treatment. The product label allows for 0.1 g to 1.0 g per treatment spot yet only three out of nine sites received at least 0.1 g per trea tment spot and two sites (five and seven) only received a quarter of the recommended minimum amount of termiticide. Label directions a ssumed that 33.33 mg of termiticide will be released with each bulb compression. In this experiment, the actual amount r eleased ranged from 2.0 mg to 10.8 mg, with an average of 7.1 mg; over four times less than was expected. Site #7, for example, had nine treatment spots with a total of 0.22 g of Termidor Dry. According to the label, with 108 puffs (average of 12 per tre atment spot), approximately 3.6 g of termiticide should have been injected. Despite less than 10% of the recommended application volume, no termite acti vity was detected even 6 mo after treatment, suggesting that Termidor Dry can work effectively below th e label rate. However, given the complex nature of localized treatment of drywood termite colonies, it was not surprising that several locations were still fairly active si x months after treatment. This may have resulted from the reduced amount of produc t applied compared to the minimum rates suggested. Out of the three failed treatment sites, two of them received
104 one sixth the recommended minimum amount. The low weight of the product released might suggest that the spread of Termidor Dry was limited in the gallery system. Neoteny in drywood termite colonies allows for greater rebounding after the population of the colony has been reduced (Ferster et al. 2001, Korb and Lenz 2004). If a site contains one spot with very low activity levels one year after treatment, it would not be prudent to assume this colony was weake ning. A ny site that still had activity one year post treatment should be considered a control failure. Two treatments performed in this study were tentatively called failures while one sit e was an outright failure. At site five, new active sites were found and a swarm occurred 1 yr after application of Termidor Dry. The inj ection site was made within 2.5 cm of consistently highly active spot as determined by the Termatrac readings. T he treatment must have either not contacted the gallery or there may have been a constriction that did not allow the product to reach that section of the gallery. However, assuming that there is a contiguous gallery structure, and Termidor Dry retains its t ermiticidal properties long term, it is reasonable to expect that termites would eventually travel back through treated zones. According to the manufacturer, not only is this formulation of fipronil non repellent, but due to its powdered cellulose carrier termites readily will consume it. I t is difficult to determine what effect the accumulation of dead termites has on conspecifics. If areas with accumulations of dead termites are abandoned or physically blocked from future entry, even a well coated sec tion of a gallery may not provide long term protection due to an absence of termite activity in these treated areas.
105 The measure of a successful drywood termite treatment is not readily apparent. A drywood termite treatment can only responsibly be referre beca use a persistent absence of swarms, new damage or fecal pellets are only supportive evidence but not definite indications of a lack of termite activity. However, the activity readings provided by the Termatrac offer a mea sure of the effectiveness of treatment. It is unlikely that localized drywood termite treatments will ever be as reliable as fumigation, but the technology currently available is helping to close the gap (Thoms 2000, Lewis 2003, Potter 2011, Hickman and F orschler 2012). The difficulties and uncertainties of spot treatments for drywood termite infestations can be lessened by employing a radar device as a termite activity measuring tool. The ability to confirm the absence of termite movement months to year s after treatment can give the operator confidence that at least part of the drywood termite infestation has been eliminated.
106 Table 5 1. Characteristics of field sites, Termidor Dry application and summary of termite activity Termite presence Spots measured Treatment spots Treatment amount (mg) Number of puffs Site Infestation type 1 Apt Staircase (I) D, P 5 3 161.5 15 2 Wood floor (I) D, S 2 2 258.4 24 3 Windowsill area (I) P, S 4 4 419.9 39 4 Door frame (I) D, P 7 6 454.9 87 5 Kitchen ca binet (I) D, P, S 10 5 128.3 43 Wooden post (O) D, P 11 8 313.7 60 Wooden post (O) D, P 10 9 223.9 108 Wood paneling (O) D, P 23 21 2443.8 252 Wood floor (I) D, P. S 8 8 384.1 60 Average 8.9 2.1 7.3 1.9 532.1 241.9 76.4 24.0 Table 5 1. Continued Milligrams per puff Milligrams per spot Active spots at 6 months Active spots at 1 year % Reduction in activity Site Infestation type 1 Apt Staircase (I) 53.8 4 0 100% 2 Wood floor (I) 129.2 0 100% 3 Windowsill area (I) 105.0 4 2 50% 4 Door frame (I) 5.2 75.8 5** 0 100% 5 Kitchen cabinet (I) 3.0 25.7 1 1 90% Wooden post (O) 5.2 39.2 2 2 82% Wooden post (O) 2.1 24.9 0 0 100% Wood paneling (O) 9.7 116.4 1 0 100% Wood floor (I) 6.4 48.0 2 0 100% Average 7.1 1.2 68 .7 13.2 26.7% 6.3% 91.33% (O) and (I) refer to outdoor and indoor sites. D (damage), P (fecal pellets), S (past alate swarm). 9 were all located at the same structure. 3. The amount of product applied was divided by the number of puffs at each site. *Site 2 did not receive a 6 month evaluation. **Site 4 value represents a 9 month evaluation
107 Figure 5 1. Sample display from Termatrac T3i The gain setting (6) is i ndicated by the vertical gauge while the top horizontal display shows activity readings (this bottom gauge is the accelerometer which shows whether the activity displayed abo ve is due to movement detected by the sensor or by the sensor being jostled.
108 Figure 5 2. Termatrac activity readings (# of spikes (A), maximum spike (B)) for groups of 1 (blue bars) 2 (red bars), or 5 termites (green bars) at three different d istan ces from the microwave sensing Termatrac T3i
109 Figure 5 3. Initial, 6 month and 12 month termite activity readings number of spikes (A) and maximum spike (B) averaged for all sites. Sites 2 and 4 did not receive a 6 month evaluation. Locations 6 9 were located within the same structure.
110 CHAPTER 6 CONCLUSIONS The first objective in this dissertation wa s to determine the pre lethal effects associated with limited exposure of R. flavipes to soil treated with four non repellent soil termiticides: chl orantraniliprole, chlorfenapyr, imidacloprid and fipronil. A new method for evaluating termite function was used to measure the pre lethal and lethal effect s of termiticide exposure In addition to determining the effects on termites directly exposed to termiticides, donor termites were allowed to horizontally tran sfer termiticides to chemically naive recipient termites. R. flavipes donor termites successfully transferred termiticide to recipi ent termites in as little as 24 h. The relatively high percen tage of initially pre lethal and healthy donor termites, followed by high donor and recipient mortality, suggests that termites exposed to slow acting termiticides transferred termiticides from donor to recipient termites T he pre lethal effects of non rep ellent soil termiticide exposure on R. flavipes indicate that termites retain sufficient physical functionality to transfer termiticides to conspecific termites away from treated soil The second objective involved additional exploration of the pre lethal effects of chlorantraniliprole exposure in R. flavipes B rief exposure to soil treated with chlorantraniliprole was used to determine length of exposure needed to reduc e the intensity an d quality of tunneling behavior. It was determined that tunneling b ehavior of R. flavipes workers w as reduced with as little as 10 min exposure to treated soil. The results obtained with speed the termiticide chlorantraniliprole demonstrate that the loss of tunneling behavior occurs quickly and is persistent, despite a h ealthy appearance of affected termites
111 The final objective of this dissertation wa s to investigate the use of a termite detection device, the Termatrac T3i, to locate drywood termite infestations and quantify the activity of infestations. Spots with dete cted termite activity were treated with an injectable termiticide, Termidor Dry, and revisited six and twelve months post treatment. A significant decline in termite activity was detected one year after treatment, evidenced by a decline in the mean number of spikes and the maximum spike detected. The Termatrac T3i was shown to reliably measu re and quantify cryptic termite presence. I n conjunction with an injectable termiticide this device provide s for a method to control and verify localized drywood term ite infestations
112 LIST OF REFERENCES Bagnres, a G., A. Pichon, J. Hope, R. W. Davis, and J. L. Clment 2009 Contact versus feeding intoxication by fipronil in Reticulitermes termites (Isoptera: Rhinotermitidae): Laboratory evaluation of toxicity, uptake, clearance, and trans fer among individuals. J. Econ. Entomol. 102: 347 356. Bess, H. A. 1971 Control of the Drywood termite, Cryptotermes brevis in Hawaii. Univ Hawaii Agric. Exp. Stn. Tech. Bull. 87: 1 32. Bignell, B. C., R M. Hollingworth, K. I. Ahammadsahib, C. D. Kukel, and S. Donovan 2004 Insecticidal action and mitochondrial uncoupling activity of AC=303,630 and related halogenated prryoles Pestic. Biohem. Physiol. 50: 115 12 8. Bl ske, V. U., and H. Hertel 2001 Repellent and toxic effects of plant extracts on subt erranean termites (Isoptera: Rhinotermitidae). J. Econ. Entomol. 94: 1200 1208. Blske, V. U., H. Hertel, and B. T. Forschler 2003 Repellent effects of isoborneol on subterranean termites (Isoptera: Rhinotermitidae) in soils of different composition. J. Econ. Entomol. 96: 1267 1274. Brown, K. S., B. M. Kard, M. E. Payton, and D. R. Kuehl 2009 Excavation of field collected soils by Reticulitermes flavipes (Isoptera: Rhinotermitidae). J. Kansas Entomol. Soc. 82: 283 292. Buczkowski, G., C. W. Scherer, and G. W. Bennett 2006 Horizontal transfer of bait in the German cockroach: Indoxacarb causes secondary and tertiary mortality. J. Chem. Ecol. 101: 894 901. Buczkowski, G., C. W. Scherer, and G. W. Bennett 2012 Toxicity and horizontal transfer of chlorant raniliprole in the eastern subterranean termite toxicity. J. Econ. Entomol. 105: 1736 1745. Bulmer, M., and J. Traniello 2002 Foraging range expansion and colony genetic organization in the subterranean termite Reticulitermes flavipes (Isoptera: Rhinoter mitidae). Environ. Entomol. 31: 293 298. Cabrera, B., and M. Rust 1996 Responses to light and thermal gradients by the western drywood termite (Isoptera: Kalotermitidae). Environ. Entomol. 25: 436 445.
113 Cameron, S. L., N. Lo, T. Bourguignon, G. J. Svenson and T. A. Evans 2012 A mitochondrial genome phylogeny of termites (Blattodea: Termitoidae): robust support for interfamilial relationships and molecular synapomorphies define major clades. Mol. Phylogenet. Evol. 65: 163 73. Choe, D. H., and M. K. Rust 2008 Horizontal transfer of insecticides in laboratory colonies of the Argentine ant (Hymenoptera: Formicidae). J. Econ. Entomol. 101: 1397 405. Chouvenc, T., and N. Y. Su 2010 Apparent synergy among defense mechanisms in subterranean termites (Rhinote rmitidae) against epizootic events: limits and potential for biological control. J. Econ. Entomol. 103: 1327 37. Chouve nc, T., N. Y. Su, and J. K. Grace 2011 Fifty years of attempted biological control of termites Analysis of a failure. Biol. Cont rol. 59: 69 82. Cornelius, M. L. 2012 Individual behavior of workers of the Formosan subterranean termite (Isoptera: Rhinotermitidae) on co nsecutive days of tunnel Construction. Insects. 3: 367 377. Cornelius M.L., and J.K. Grace. 1997 Effect of Termite Soldiers on the Foraging Behavior of Coptotermes formosanus (Isoptera: Rhinotermitidae) in the Presence of Predatory Ants. Sociobiol. 29: 2 47 253 Cornelius, M. L., D. J. Daigle, W. J. Connick, A. Parker, and K. Wunch 2002 Responses of Coptotermes formosanus and Reticulitermes flavipes (Isoptera: Rhinotermitidae) to three types of wood rot fungi cultured on different substrates. J. Econ. En tomol. 95: 121 8. Cornelius, M. L., and W. L. A. Osbrink 2010 Effect of soil type and moisture availability on the foraging behavior of the Formosan subterranean termite (Isoptera: Rhinotermitidae). J. Econ. Entomol. 103: 799 807. Cornelius, M. L., and W L. A. Osbrink 2011 Influence of dry soil on the ability of Formosan subterranean termites, Coptotermes formosanus to locate food sources. J. Insect Sci. 11: 162. Crosland, M., C. Lok, T. Wong, M. Shakarad, and J. Traniello 1997 Division of labour in a lower termite: the majority of tasks are performed by older workers. Anim. Behav. 54: 999 1012.
114 Deheer, C. J., and E. L. Vargo 2004 Colony genetic organization and colony fusion in the termite Reticulitermes flavipes as revealed by foraging patterns o ver time and space. Mol. Ecol. 13: 431 441. Ebeling, W. 1983 Breakthrough for drywood termite control. IPM Pract. 5: 1 2. Ebeling, W. 1997 Beginnings of structural IPM. IPM Pract. 19: 1 8. Ebeling, W., and R. J. Pence. 1957 Relation of particle size to the penetration of subterranean termites through barriers of sand and cinders. J. Econ Entomol. 58: 690 692. Eger, J. E. J., M. D. Lees, P. A. Neese, T. H. Atkinson, E. M. Thoms, T. Messenger, J. J. Demark, L. C. Lee, E. L. Vargo, and M. P. Tolley 2012 Elimination of subterranean termite (Isoptera: Rhinotermitidae) colonies using a refi ned cellulose bait matrix containing noviflumuron when monitored and replenished quarterly. J. Econ. Entomol. 105: 533 539. Eggleton, P. 2000 Global patterns of termite diversity. pp. 25 52. In T. Abe, M. Higashi, and D. E. Bignell (eds.), Termites: Evolu tion, Sociality, Symbioses, Ecology. Kluwer Academic, Dordrecht, Netherlands. Emerson, A. E. 1965 A review of the Mastotermitidae (Isoptera), including a new fossil genus from Brazil. Am. Mus. Novit. 17: 1 46. Engel, M. S. 2011 Family group names for te rmites (Isoptera), redux. Zookeys. 184: 171 84. Evans, T. A B. T. Forschler, and J. K. Grace 2013 Biology of invasive termites: a worldwide review. Annu. Rev. Entomol. 58: 455 74. Evans, T. A 2006 Foraging and building in subterranean termites: task s witchers or reserve labourers? Insectes Soc. 53: 56 64. Fei, H and G. Henderson. 2004 Effects of temperature, directional aspects, light conditions, and termite species on subterranean termite zctivity (Isoptera: Rhinotermitidae). Environ. Entomol. 33: 242 248 Fei, H G Henderson and R A. Laine. 2005 Trail following behavior of Coptotermes formosanus (Isoptera: Rhinotermitidae) on co ncentra tion gradients of 2 phenoxyethanol. Sociobiol. 45: 483 494.
115 Ferreira, M. T., and R. H. Scheffrahn 2011 Light attraction and subsequent colonization behaviors of alates and dealates of the West Indian drywood termite (I soptera: Kalotermitidae). Fl a Entomol. 94: 131 136. Ferster, B., R. H. Scheffrahn, E. M. Thoms, and P. N. Scherer 2001 Transfer of toxicants fr om exposed nymphs of the drywood termite Incisitermes snyderi (Isoptera: Kalotermitidae) to unexposed nestmates. J. Econ. Entomol. 94: 215 222. Forbes, C., and W. Ebeling 1986 Update: Liquid nitrogen controls dryw ood termites. IPM Pract. 111: 1 4. Forsch ler, B. T. 1994 Suvivorship and tunneling activity of Reticulitermes flavipes (Kollar) (Isoptera: Rhinotermitidae) in response to termiticide soil barriers with and without gaps of untreated soil. J. Econ. Entomol. 29: 43 54. Gahlhoff, J. E., and P. G. Ko ehler 2001 Penetration of the eastern subterranean termite into soil treated at various thicknesses and concentrations of Dursban TC and Premise 75. J. Econ. Entomol. 94: 486 491. Gautam, B. K., and G. Henderson 2011 Effect of soil type and exposure du ration on mortality and transfer of chlorantraniliprole and fipronil on Formosan subterranean termites (Isoptera: Rhinotermitidae). J. Econ. Entomol. 104: 2025 2030. Gautam, B. K., and G. Henderson 2012 Laboratory study of the influence of substrate type and temperature on the exploratory tunneling by Formosan subterranean termite. Insects. 3: 629 639. Gautam, B. K., G. Henderson, and R. W. Davis 2012 Toxicity and horizontal transfer of 0.5% fipronil dust against Formosan subterranean termites. J. Econ. Entomol. 105: 1766 1772. Gold, R., H. J. Howell, B. Pawson, M. Wright, and J. Lutz 1996 Persistence and bioavailability of termiticides to subterranean termites (Isoptera: Rhinotermitidae) from five soil types and locations in Texas. Sociobiol. 28: 337 363. Grace, J. K., and W. W. Wilcox 1988 Isolation and trail following bioassay of a decay fungus associated with Reticulitermes hesperus Banks (Isoptera: Rhinotermitidae). Pan Pac. Entomol. 64: 243 249. Gray, H. E. 1960 Vikane: A new fumigant for contr ol o f drywood termites. Pest Contr Tech. 28: 43 46. Green, J. M., M. E. Scharf, and G. W. Bennett 2005 Impacts of soil moisture level on consumption and movement of three sympatric subterranean termites (Isoptera: Rhinotermitidae) in a laboratory assay. J. Econ. Entomol. 98: 933 937.
116 Gulmahjamad, H. 1997 Naturally occurring infestations of drywood termites in books. Pan Pac. Entomol. 73: 245 247. Hannig, G. T., M. Ziegler, and P. G. Maron 2009 Feeding cessation effects of chlorantraniliprole, a new a nthranilic diamide insecticide, in comparison with several insecticides in distinct chemical classes and mode of action groups. Pest Manag. Sci. 65: 969 9 74. Hartke, T. R., and B. Baer 2011 The mating biology of termites: a comparative review. Anim. Beha v. 82: 927 936. Hayashi, Y., N. Lo, H. Miyata, and O. Kitade. 2007 Sex linked genetic influence on caste determination in a termite. Science 318: 985 987. Henderson, G. 2003 Liquid learning. Pest Contr. Tech 31: 48 59. Hermann, H.R. 1984. Defensive Me chanisms in Social Insects. Praeger, New York. 167 180. Hertel, H., A. Hanspach, and R. Plarre 2011 Differences in Alarm Responses in Drywood and Subterranean Termites (Isoptera: Kalotermitidae and Rhinotermitidae) to Physical Stimuli. J. Insect Behav. 2 4: 106 115. Hickman, R., and B. T. Forschler 2012 Evaluation of a localized treatment technique using three ready to use products against the drywood termite Incisitermes snyderi (Kalotermitidae) in naturally infested lumber. Insects. 3: 25 40. Hill, G. F. 1942 Termites (Isoptera) from the Australian region. Council for Scientific and Industrial Research, Melbourne, Australia. Horwood, M. A P. C. Bohringer, and A. Kathuria 2012 Control of subterranean termite (Isoptera: Rhinotermitidae) infestation s in wooden power poles using bandages containing fipronil. Aust. J. Entomol. 51: 266 271. Horwood, M. A ., T. Westlake, and A. Kathuria 2010 Control of subterranean termites (Isoptera: Rhinotermitidae) infesting power poles. J. Econ. Entomol. 103: 2140 2 146. Hu, X. P. 2005 Evaluation of efficacy and nonrepellency of indoxacarb and fipronil treated soil at various concentrations and thicknesses against two subterranean termites (Isoptera: Rhinotermitidae). J. Econ. Entomol. 98: 509 517. Hu, J., and B. T. Forschler 2012 Neotenic Phenotype and Sex Ratios Provide Insight into Developmental Pathways in Reticulitermes flavipes (Isoptera: Rhinotermitidae). Insects. 3: 538 552.
117 Hu, J., K. B. Neoh, A. G. Appel, and C. Y. Lee 2012 Subterranean termite open air foraging and tolerance to desiccation: Comparative water relation of two sympatric Macrotermes spp. (Blattodea: Termitidae). Comp. Biochem. Physiol. A. Mol. Integr. Physiol. 161: 201 20 7. Hu, X. P., D. Song, and C. W. Scherer 2005 Transfer of indoxacarb among workers of Coptotermes formosanus (Isoptera: Rhinotermitidae): effects of dose, donor: recipient ratio and post exposure time. Pest Manag. Sci. 61: 1209 12 14. Huang, Q. Y., C. L. Lei, and D. Xue 2006 Field evaluation of a fipronil bait against subt erranean termite Odontotermes formosanus (Isoptera: Termitidae). J. Econ. Entomol. 99: 455 461. Huang, Q. Y., W. P. Wang, R. Y. Mo, and C. L. Lei 2008 Studies on feeding and trophallaxis in the subterranean termite Odontotermes formosanus using rubidium chloride. Entomol. Exp. Appl. 129: 210 215. Ibrahim, S. a, G. Henderson, and H. Fei 2003 Toxicity, repellency, and horizontal t ransmission of fipronil in the F ormosan subterranean termite (Isoptera: Rhinotermitidae). J. Econ. Entomol. 96: 461 467. Ibrahi m, S.A., G. Henderson, H. Fei and R. A. Laine. 2005 2 Phenoxyethanlol is a double edged sword against the Formosan subterranean termite (Isoptera: Rhinotermitidae). J. Biopest Int l 1: 25 37. Inward, D.J.G., Vogler, A.P., and P. Eggleton. 2007 A Compr ehensive Phylogenetic Analysis of Termites (Isoptera) Illuminates Key Aspects of Their Evolutionary Biology. Mol. Phylogenet. Evol. 44 : 953 967. Jaffe, K., S. Issa, and C. Sainz Borgo 2012 Chemical Recruitment for Foraging in Ants (Formicidae) and Termit es (Isoptera): A Revealing Comparison. Psyche 2012: 1 11. Jones, S. C. 1990 Effects of population density on tunneling by Formosan subterranean termite (Isoptera: Rhinotermitidae) through treated soil. J. Econ. Entomol. 83: 875 878. Jordan, B. W., B. E. Bayer, P. G. Koehler, and R. M. Pereira 201 3 Bait Evaluation Methods for Urban Pest Management, pp. 445 469. In I nsecticides: Development of s afer and more effective t echnol ogy Stanislav Trdan (ed.). InTech, Rijeka, Croatia. Jouquet, P., S. Traor, C. C hoosai, C. Hartmann, and D. Bignell 2011 Influence of termites on ecosystem functioning. Ecosystem services provided by termites. Eur. J. Soil Biol. 47: 215 222.
118 Kamhbampati, S., and P. Eggleton. 2000 Taxonomy and phylogeny of termites, pp. 1 23. In T. Abe, M. Higashi, and D. E. Bignell (eds.), Termites: Evolution, Sociality, Symbioses, Ecology. Kluwer Academic, Dordrecht, Netherlands. Korb, J and M. Lenz. 2004 Reproductive decision making in the termite, Cryptotermes secundus (Kalotermitidae), under variable food conditions. Behav. Ecol. 15: 390 395. Korb, J. 2007 Workers of a drywood termite do not work. Front. Zool. 4: 1 7. Korb, J., M. Buschmann, S. Schafberg, J. Liebig, and A. G. Bagnres 2012 Brood care and social evolution in termites. Proc. Biol. Sci. 279: 2662 2671. Lacey, M. J., E. Smon, J. Krasulov, D. Sillam Dusss, A. Robert, R. Cornette, M. 2011 Chemical communication in termites: syn 4,6 dimethylundecan 1 ol as trail following pheromone, syn 4,6 dimethylunde canal and (5E) 2,6,10 trimethylundeca 5,9 dienal as the respective male and female sex pheromones in Hodotermopsis sjoestedti (Isoptera, Arc hotermopsidae) J. Insect Physiol. 57: 1585 1591. Lahm, G. P., D. Cordova, and J. D. Barry 2009 New and selective ryanodine receptor activators for insect control. Bioorg. Med. Chem. 17: 4127 33. Lee, M., and E. K. Kang 2011 Termite tunneling feature extraction using genetic algorithm. Soft Comput. 16: 419 425. Lewis, V. R. 1993 Alternative termite detection and co ntrol techniques. Pest Contr. Technol. 21: 46, 48, 50 51. Lewis, V. R. 2003 IPM for drywood Termites (Isoptera: Kalotermitidae ). J. Entomol. Sci. 38: 181 199. Lewis, V. R., and M. L. Haverty 1996 Evaluation of six techniques for control of the western drywood termite (Isoptera: Kalotermitidae) in structures. J. Econ. Entomol. 89: 922 934.
119 Lewis, V. R. S. Leighton, R. Tabuchi, and M. Haverty 2011 Seasonal a nd daily patterns in activity of the western drywood termite, Incisitermes minor (Hagen). Insects. 2: 555 563. Lima, J. T., and A. M. Costa Leonardo 2012 Tunneling behaviour of the Asian subterranean termite in heterogeneous soils: presence of cues in th e foraging area. Anim. Behav. 83: 1269 1278. Mankin, R. W., W. L. Osbrink, F. M. Oi, and J. B. Anderson 2002 Acoustic detection of termite infestations in urban trees. J. Econ. Entomol. 95: 981 98 8. Mao, L., G. Henderson, and C. W. Scherer 2011 Toxicit y of seven termiticides on the Formosan and eastern subterranean termites. J. Econ. Entomol. 104: 1002 1008. Matsumura, F., H. C. Coppel, and A. Tai 1968 Isolation and identification of termite trail following pheromone. Nature. 219: 963 964. Matsuura, K ., C. Himuro, T. Yokoi, Y. Yamamoto, E. L. Vargo, and L. Keller 2010 Identification of a pheromone regulating caste differentiation in termites. Proc. Natl. Acad. Sci. U. S. A. 107: 12963 1296 8. Messenger, M. T., N. Y. Su, C. Husseneder, and J. K. Grace 2005 Elimination and reinvasion studies with Coptotermes formosanus (Isoptera: Rhinotermitidae) in Louisiana. J. Econ. Entomol. 98: 916 9 29. Minnick, D. R. 1970 Behavioral and toxicological studies of the drywood termite, Cryptotermes brevis (Walker) (K alotermitidae: Isoptera). Ph.D Diss. Univ. of Fl Gainesville, FL. Minnick, D. R., S. H. Kerr, and R. C. Wilkinson 1973 Humidity behavior of the drywood termite Cryptotermes brevis Environ. Entomol. 2: 597 601. Mulrooney, J. E., M. K. Davis, T. L. Wagn er, and R. L. Ingram 2006 Persistence and efficacy of termiticides used in preconstruction treatments to soil in Mississippi. J. Econ. Entomol. 99: 469 475. Mulrooney, J. E., and P. D. Gerard. 2009 Tunneling and activity of Reticulitermes flavipes (Isop tera: Rhinotermitidae) exposed to low concentrations of nonrepellen t termiticides. Sociobiol. 53 : 695 706. Mulrooney, J. E., T. L. Wagner, T. G. Shelton, C. J. Peterson, and P. D. Gerard 2007 Historical review of termite activity at forest service termi ticide test sites from 1971 to 2004. J. Econ. Entomol. 100: 488 494. Myles, T. G. 1995 New records of drywood termite introduction, interception and extirpation in Ontario. Proc. Entomol. Soc. Ontario. 126: 77 83.
120 Neoh, K. B., J. Hu, B. H. Yeoh, and C. Y. Lee 2012 a Toxicity and horizontal transfer of chlorantraniliprole against the Asian subterranean termite Coptotermes gestroi (Wasmann): effects of donor: recipient ratio, exposure duration and soil type. Pest Manag. Sci. 68: 749 756. Neoh, K. B., B. K. Yeap, K. Tsunoda, T. Yoshimura, and C. Y. Lee 2012 b Do termites avoid carcasses? Behavioral responses depend on the nature of the carcasses. PLoS One. 7: 1 11. Noirot, C., and J. Pasteels 1987 Ontogenetic development and evolution of the worker caste i n termites. Experimentia. 43: 851 860. Oi, F. M., J. L. Castner, and P. G. Koehler. 2003 The eastern subterranean termite. University of Florida, Institute of Food and Agricultural Sciences, Cooperative Extension, Florida. http://edis.ifas.ufl.edu/IN031 Osbrink, W. L. A. M. L. Cornelius, and A. R. Lax 2005 Effect of imidacloprid soil treatments on occurrence of Formosan subterranean termites (Isoptera: Rhinotermitidae) in independent monitors. J. Econ. Entomol. 98: 2160 2168. Osbrink, W. L A and A. R. Lax 2002 Effect of tolerance to insecticides on substrate penetration by Formosan subterranean termites (Isoptera: Rhinotermitidae). J. Econ. Entomol. 95: 989 1000. Osbrink, W. L. A., R. H. Scheffrahn, N. Y. Su, and M. K. Rust. 1987 Laboratory compa risons of sulfuryl fluoride toxicity and mean time of mortality among ten termite species (Isoptera: Hodotermitidae, Kalotermitidae, Rhinotermitidae). J. Econ. Entomol. 80: 1044 1047. Parman, V., and E. L. Vargo 2010 Colony level effects of imidacloprid in subterranean termites (Isoptera: Rhinotermitidae). J. Econ. Entomol. 103: 791 798. Pence, R. 1966 Tolerance of the drywood termite, Incisitermes minor (Hagen) to extreme desiccation. PCO News. 26: 13 14. Peterson, C. J. 2010 Varying termiticide applic ation rate and volume affect initial soil penetration. J. Econ. Entomol. 103: 433 436.
121 Puche, H., and N. Su 2001 Tunnel formation by Reticulitermes flavipes and Coptotermes formosanus (Isoptera: Rhinotermitidae) in response to wood in sand. J. Econ. Entomol. 94: 1398 1404. Quarcoo, F. Y. 2009 Behavioral toxicol ogy of the eastern subterranean termite, Reticulitermes flavipes (Kollar) (Isoptera: Rhinotermitidae). Ph.D Diss. Auburn Univ. Auburn, AL. Quarcoo, F. Y., A. G. Appel, and X. P. Hu 2010 Descriptive study of non repellent insecticide induced behaviors in Reticulitermes flavipes (Isoptera: Rhinotermitidae). Sociobiol. 55: 217 228. Quarles, W. 1993 Electro gun 98% effective against termites. IPM Pract. 18: 11. Quarles, W. 1997 Summary of alternative drywood termite treatments. IPM Pract. 9. Quarles, W. 19 99 Non toxic control of drywood termites. IPM Pract. 21: 1 10. Ramakrishnan, R., D. R. Suiter, C. H. Nakatsu, and G. W. Bennett 2000 Feeding inhibition and mortality in Reticulitermes flavipes (Isoptera: Rhinotermitidae) after exposure to imidacloprid t reated soils. J. Econ. Entomol. 93: 422 428. Raymond Delpech, V., K. Matsuda, B. M. Sattelle, J. J. Rauh, and D. B. Sattelle 2005 Ion channels: molecular targets of neuroactive insecticides. Invert. Neurosci. 5: 119 1 33. Reierson, D. A. 1950 Feeding pre ferences of drywood termites and termiticidal effect of boric acid. Pest Control Oper. News. 26: 13 15. Remmen, L. N., and N. Y. Su 2005a Time trends in mortality for thiamethoxam and fipronil against Formosan subterranean termites and eastern subterrane an termites (Isoptera: Rhinotermitidae). J. Econ Entomol. 98: 911 915 Remmen, L. N., and N. Y. Su 2005b Tunneling and mortality of eastern and Formosan subterranean termites (Isoptera: Rhinotermitidae) in sand treated with thiamethoxam or fipronil. J. Econ. Entomol. 98: 906 9 10. Ripa, R., P. Luppichini, N. Y. Su, and M. K. Rust 2007 Field evaluation of potential control strategies against the invasive eastern subterranean termite (Isoptera: Rhinotermitidae) in Chile. J. Econ. Entomol. 100: 1391 1399. Rust, M. K., J. K. Grace, D. L. Wood, and D. A. Reierson 1988 The search for new termite control strategies. Calif. Agric. 42: 15 18.
122 Rust, M. K., and R. K. Saran 2006 Toxicity, repellency, and transfer of chlorfenapyr against western subterranean term ites (Isoptera: Rhinotermitidae). J. Econ. Entomol. 99: 864 872. Rust, M. K., and N. Y. Su 2012 Managing social insects of urban importance. Annu. Rev. Entomol. 57: 355 3 75. Saran, R. K. 2006 Behavioral responses, utilization, and horizontal transfer of phagostimulants, attractants, and non repellent termiticides in the western subterranean termite, Reticulitermes hesperus Ban ks (Isoptera: Rhinotermitidae). Ph.D. Diss. Univ. CA Riverside. Riverside, CA. Saran, R. K., and S. T. Kamble 2008 Concentration dependent degradation of three termiticides in soil under laboratory conditions and their bioavailability to eastern subterranean termites (Isoptera: Rhinotermitidae). J. Econ. Entomol. 101: 1373 1383. Saran, R. K., J. G. Millar, and M. K. Rust 2007 Rol e of (3Z,6Z,8E) dodecatrien 1 ol in trail following, feeding, and mating behavior of Reticulitermes hesperus J. Chem. Ecol. 33: 369 3 89. Saran, R. K., and M. K. Rust 2007 Toxicity, uptake, and transfer efficiency of fipronil in western subterranean term ite (Isoptera: Rhinotermitidae). J. Econ. Entomol. 100: 495 508. Saran, R. K., and M. K. Rust 2008 Phagostimulatory sugars enhance uptake and horizontal transfer of hexaflumuron in the eastern subterranean termite (Isoptera: Rhinotermitidae). J. Econ. En tomol. 101: 873 879. Saran, R. K ., C. Scherer, and M. Coffelt 2011 In Proc. 7 th Int l Conf. Urban Pests. W. H. Robinson and A. E. de Carvalho Campos (eds.) 18 21. Scheffrahn, R. H. 2011 Distribution, diversity, mesonotal morphology, gallery architecture, and queen physogastry of the termite genus Calcaritermes (Isoptera, Kalotermitidae). Zookeys. 53: 41 53. Scheffrahn, R. H., W. P. Robbins, P. Busey, N. Y. Su, and R. K. Mueller 1993 Evaluation of a novel, hand held, acoustic emissions detector to monitor termites (Isoptera: Kalotermitidae, Rhinotermitidae) in wood. J. Econ. Entomol. 86: 1720 1729.
123 Scheffrahn, R. H., and N. Y. Su 1994 Keys to soldier and winged adult termites (Isoptera) of Florida. Florida Entomol. 77: 460 474. Scheffrahn, R. H., N. Su, J. A. N. Krecek, A. V. A. N. Liempt, and G. S. Wheeler 1998 Prevention of colony foundation by Cryptotermes brevis and remedial control of Drywood Ter mites (Isoptera: Kalotermitidae ) with selected chemical treatments. J. Econ. Entomol. 91: 1387 1396. Schwinghammer, M. a, X. Zhou, S. Kambhampati, G. W. Bennett, and M. E. Scharf 2011 A novel gene from the takeout family involved in termite trail following behavior. Gene. 474: 12 21. Shellman Reeve, J. S. 1997 The spectrum of eusociality in termites, pp. 52 93. In Social Behavior in Insects and Arachnids J. C. Choe, and B. J. Crespi (eds.) Cambridge Univ. Press. Cambridge, UK. Shelton, T. G., and J. K. Grace 2003 Effects of exposure duration on transfer of nonrepellent termiticides among workers of Coptotermes formosan us Shiraki (Isoptera: Rhinotermitidae). J. Econ. Entomol. 96: 456 460. Shelton, T. G., J. E. Mulrooney, and T. L. Wagner 2006 Transfer of chlorfenapyr among workers of Reticulitermes flavipes (Isoptera: Rhinotermitidae) in the laboratory. J. Econ. Entomo l. 99: 886 892. Smith, J. A., R. M. Pereira, and P. G. Koehler 2008 Relative repellency and lethality of the neonicotinoids thiamethoxam and acetamiprid and an acetamiprid/bifenthrin combination to Reticulitermes flavipes termites. J. Econ. Entomol. 101: 1881 1887. Smith, M. 1995 Drywood termites: The experts answ Technol. 23: 34, 36 37, 40, 88. Smith, V. K. 1982 Biology of subterranean termites and their control. Pest Manag. 1: 7 9.
124 Smythe, R., H. Coppel, S. Lipton, and F Strong 1967 Chemical studies of attractants associated with Reticulitermes flavipes and R. virginicus J. Econ. Entomol. 60: 228 233. Snyder, T E. 1948 Our enemy the termite. Comstock Publishing Co. Ithaca, NY. Soeprono, A. M., and M. K. Rust 2004 Effect of horizontal transfer of barrier insecticides to control Argentine ants (Hymenoptera: Formicidae). J. Econ. Entomol. 97: 1675 16 81. Spomer, N. a, and S. T. Kamble 2011 Temporal changes in chlorantraniliprole and indoxacarb in four midwestern soi ls and bioefficacy against the eastern subterranean termite (Isoptera: Rhinotermitidae). J. Econ. Entomol. 104: 990 1001. Spomer, N. A, S. T. Kamble, and B. D. Siegfried 2009 Bioavailability of chlorantraniliprole and indoxacarb to eastern subterranean t ermites (Isoptera: Rhinotermitidae) in various soils. J. Econ. Entomol. 102: 1922 1927. Spomer, N. A., S. T. Kamble, R. A. Warriner, and R. W. Davis 2008 Influence of temperature on rate of uptake and subseque nt horizontal transfer of C 14 fipronil by eas tern subterranean termites (Isoptera: Rhinotermitidae). J. Econ. Entomol. 101: 902 908. Stewart, D. 1957 Sulfuryl fluoride a new fumigant for control of the drywood termite Kalotermes minor Hagen. J. Econ. Entomol. 50: 7 11. Su, N. Y. 1996 Urban entomology: termites and t ermite control, pp. 451 464. In D. Rosen, F. D. Bennet, and J. L. Capinera (eds.), Pest management in the subtropics: int egrated pest management: a Florida perspective. Intercepts Ltd., Andover, UK. Su, N. Y. 2002 Novel technologies for subterranean termite control. Sociobiol. 40: 95 101. Su, N. Y. 2005a Response of the Formosan subterranean termites (Isoptera: Rhinoter mitidae) to baits or nonrepellent termiticides in extended foraging arenas. J. Econ. Entomol. 98: 2143 52. Su, N. Y. 2005b Directional change in tunneling of subterranean termites (Isoptera: Rhinotermitidae) in response to decayed wood attractants. J. Eco n. Entomol. 98: 471 5.
125 Su, N Y., M. Tamoshiro, J.R. Yates, and M.I. Haverty. 1982. Effect of behavior on the evaluation of insecticides for prevention of or remedial control of the Formosan subterranean termite. J. Econ. Entomol. 75: 188 193. Su, N. Y., a nd R. H. Scheffrahn 1986 Field comparison of sulfuryl fluoride susceptibility among three termite species (Isoptera: Kalotermitidae, Rhinotermitidae) during structural fumigation. J. Econ. Entomol. 79: 903 908. Su, N. Y., and R. H. Scheffrahn 1990 a Eco nomically important termites in the United States and their control. Sociobiol. 17: 77 94. Su, N. Y. and R. H. Scheffrahn 1990 b Comparison of eleven soil termiticides against the Formosan subterranean termite and the eastern subterranean termite (Isopte ra: Rhinotermitidae). J. Econ. Entomol. 83: 1918 1924. Su, N. Y. and R. H. Scheffrahn 1992 Penetration of sized particle barriers by field populations of subterranean termites (Isoptera: Rhinotermitidae). J. Econ. Entomol. 85: 2275 2278. Su, N. Y. R. H. Scheffrahn, and P. M. Ban 1991 Uniform size particle barrier: A physical exclusion device against subterranean termites (Isoptera: Rhinotermitidae). J. Econ. Entomol. 84: 912 916. Su, N. Y., and R. H. Scheffrahn 1998 A review of subterranean termite control practices and prospects for integrated pest management programs. Integr. Pest Manag. Rev. 3: 1 13. Su, N. Y., and H. Puche 2003 Tunneling activity of subterranean termites (Isoptera: Rhinotermitidae) in sand with moisture gradients. J. Econ. Entomol. 96: 88 93. Surez, M. E., and B. L. Thorne 2000 Rate, amount, and distribution pattern of alimentary fluid transfer via trophallaxis in three species of termites (Isoptera: Rhinotermitidae, Termopsidae). Ann. Entomol. Soc. Am. 93: 145 155. Sun, J. Z., M. E. Lockwood, J. L. Etheridge, C. Z. Hollo mon, C. E. H. Coker, P. R. Knight, J. Sun, and J. Carroll 2007 Distribution of Formosan subterranean termite (Isoptera: Rhinotermitidae) in Mississippi. J. Econ. Entomol. 100: 1400 1408. Sun, Q., K. F. Haynes, and X. Zhou 2013 Differential undertaking response of a lower termite to congeneric and conspecific corpses. Sci. Rep. 3: 1 8.
126 Swoboda, L. E., and D. M. Miller 2004 Laboratory assays evaluate the influence of physical guidelines on subterranean termite (Isoptera: Rhinotermitidae) tunneling, bait discovery, and consumption. J. Econ. Entomol. 97: 1404 1412. Tai, A., F. Matsumura, and H. Coppel 1969 Chemical identification of the trail following pheromone for a southern subterranean termite. J. Org. Chem. 34: 2180 2182. Tai, A., F. Matsumura, and H. C. Coppel 1971 Synthetic analogues of the termite trail following pheromone, structure and biological activity. J. Insect Physiol. 17: 181 188. Thoms E. M 2000 Use of an acoustic emissions detector and intragallery injection of spinosad by pest con trol operators for remedial control of drywood termites (Isoptera: Kalotermitidae). Florida Entomol. 83: 64 74. Thorne, B. L. 1993 Alternative approaches to termite detection and control. Pest Manag. 12: 8 16. Thorne, B. L. 1997 Evolution of eusociality in termites. Annu. Rev. Ecol. Syst. 28: 27 54. Thorne, B. L. 199 8 B iology of subterranean termites of the genus Reticulitermes National Pest Control Association, Dunn Loring, VA. Thorne, B. L., and N. L. Breisch 2001 Effects of sublethal exposure to im idacloprid on subsequent behavior of subterranean termite Reticulitermes virginicus (Isoptera: Rhinotermitidae). J. Econ. Entomol. 94: 492 498. Thorne, B. L., N. L. Breisch, and M. L. Muscedere 2003 Evolution of eusociality and the soldier caste in termi tes: influence of intraspecific competition and accelerated inheritance. Proc. Natl. Acad. Sci. U. S. A. 100: 12808 128 13. Todaka, N., T. Inoue, K. Saita, M. Ohkuma, C. A. Nalepa, M. Lenz, T. Kudo, and S. Moriya 2010 Phylogenetic analysis of cellulolytic enzyme genes from representative lineages of termites and a related cockroach. PLoS One. 5: 1 10. Tomalski, M., W. Leimkuehler, C. Schal, and E. L. Vargo 2010 Metabolism of imidacloprid in workers of Reticulitermes flavipes (Isoptera: Rhinotermitidae). Ann. Entomol. Soc. Am. 103: 84 95. Tucker, C. L., P. G. Koehler, and F. M. Oi 2004 Influence of soil compaction on tunnel network construction by the eastern subterranean termite (Isoptera: Rhinotermitidae). J. Econ. Entomol 97: 89 94.
127 Vargo, E. L., and V. Parman 2012 Effect of fipronil on subterranean termite colonies (Isoptera: Rhinotermitidae) in the field. J. Econ. Entomol. 105: 523 5 32. Whitman, J. G. 2006 Observations of behaviors in the worker caste of Reticuliterm es flavipes (Kollar) (Isoptera: Rhinotermitidae). M.S. Thes. Univ. GA. Athens, GA. Whitman, J. G., and B. T. Forschler 2007 Observational notes on short lived and infrequent behaviors displayed by Reticulitermes flavipes (Isoptera: Rhinotermitidae). Ann. Entomol. Soc. Am. 100: 763 771. Woodrow, R. J., J. K. Grace, and R. J. Oshiro 2006 Comparison of localized injections of spinosad and selected insecticides for the control of Cryptotermes brevis (Isoptera: Kalotermitidae) in naturally infested structura l mesocosms. J. Econ. Entomol. 99: 1354 62. Wright, M. S., W. L. A. Osbrink, and A. R. Lax 2002 Transfer of entomopathogenic fungi among Formosan subterranean termites and subsequent mortality. J. Appl. Entomol. 126: 20 23. Yang, R., N. Su, and P. Bardun ias 2009 Individual task load in tunnel excavation by the Formosan subterranean termite (Isoptera: Rhinotermitidae). Ann. Entomol. Soc. Am. 102: 906 910.
128 BIOGRAPHICAL SKETCH Bennett William Jordan, son of Bruce and Candy Jordan, was born in 1983, in Me nomonie, Wisconsin. He graduated from Menomonie High School (Menomoni e, Wisconsin) in 2002. Bennett earned a Bachelor of Arts degree in biology from Lake Forest College (Lake Forest, IL) in 2006. Between the years of 2006 and 2 009 he worked on a chains aw crew for the Montana Conservation Corps (Bozeman, MT), surveyed streams for the United States Forest Service (Estacada, OR), conducted herpetological research for the University of Central Oklahoma (Edmond, OK) and monitored invasive species for the Minnesota Department of Agriculture (St. Paul, MN). In August 2009 Bennett enrolled at the University of Florida (Gainesville, FL) and began his pursuit of a Doctor of Philosophy majoring in entomology while working in the urban laboratory. Benn ett married Tully Clark in the summer of 2012 and has relocate d to Fairfax, VA to work for the National Pest Management Association as a staff entomologist and research scientist.