An Assessment of the Potential Uses of Agonistic Behaviors in Termite Control

The potential use of termite-termite agonism in pest control is explored and evaluated. Intraand interspecific encounters among termites from different colonies are known to result in aggressive or avoidance behaviors in a variety of species. Recent studies suggest, however, that intraspecific confrontations only rarely evoke aggressive responses in several economically important termite genera in the United States (Coptotermes, Reticulitermes, and Zootermopsis). Thus the most promising applied use of natural agonistic behaviors among termites appears to be in research on individualand colonyrecognition systems in the Isoptera. Artificial manipulation of recognition and response cues might facilitate novel control methodologies based on masking or interfering with natural stimuli. Directions for future research are discussed. Insect behaviors are diverse, and knowledge of the behav­ iors of a species has always been a key element in pest manage­ ment. Basic knowledge of activity periods, diet preferences, and mating behaviors can be central in determining an approach toward deterring a pest species. Taking advantage of behavioral patterns or interfering with natural repertoires or both can produce very effective control methodologies. Understanding the foraging behavior of subterranean ter­ mites is crucial to prevention of damage to buildings (Beal and others 1989; Su and Scheffrahn 1990). Preventative techniques include proper construction practices to exclude or deter forag­ ing and feeding by subterranean termites and application of large volumes of chemical insecticides as soil drenches to kill or repel foraging subterranean termites. Existing and future methods for remedial control of subterranean termites will also rely on knowledge of the foraging behavior of these insects (Su and Scheffrahn 1990). Two promising future control methods involve the use of physical barriers and the use of nonrepellent, slow-acting insec­ ticides. Physical barriers will consist of a substrate that has a particle size impenetrable to subterranean termites (Su and Scheffrahn 1990). Nonrepellent, slow-acting insecticides will be used in baits or in baiting systems. Subterranean termites will recruit to the bait stations through natural foraging behaviors and feed on the chemically impregnated wood. It is then hoped that Presented at the Symposium on Current Research on Wood-Destroying Organisms and Future Prospects for Protecting Wood in Use, September 13, 1989, Bend, Oregon. Senior Research Associate and Adjunct Assistant Professor, Department of Biology, Northeastern University, Boston, Mass. 02115 and Curatorial Research Associate, Museum of Comparative Zoology, Harvard University, Cambridge, Mass. 02138; Project Leader and Principal Research Entomologist, Pacific Southwest Forest and Range Experiment Station, Forest Service, U.S. Department of Agriculture, P.O. Box 245, Berkeley, Calif. 94701 they will return to their colony and feed their insecticide-laced crop contents to colony mates before dying. Normal foraging behaviors will thus cause an entire colony to be slowly poisoned by toxic baits. Termite colony populations will become depressed, and they will ultimately be eliminated, not simply excluded, from structures. Another behavior that has recently received attention by termite biologists is agonistic behavior. Agonistic behavior involves social interactions between individuals, including sub­ mitting, fleeing and fighting, even to death. Termites show a wide range of agonistic behaviors when interacting with ter­ mites from a different colony. In all cases studied to date, termites from colonies of different sympatric species respond aggressively when they meet (Adams and Levings 1987; Clément 1982; Grassi and Sandias 1896-1897; Haverty and Thorne 1989; Ne11968; Springhetti and Amorelli 1982; Thorne 1982; Traniello and Beshers 1985). Intraspecific confrontations also result in clear aggressive or avoidance behaviors in a variety of termite species (Adams and Levings 1987; Andrews 1911; Binder 1988; Clément 1986; Dudley and Beaumont 1889; Howick and Creffield 1980; Jones 1987; Levings and Adams 1984; Nel 1968; Pearce and others 1990; Pickens 1934; Su and Haverty 1991; Thorne 1982). Since many of these termite-termite interactions result in injury or death to some or all of the insects, it is reasonable to evaluate their potential use in control of termites as structural pests. CASE STUDY: THE DAMPWOOD TERMITE GENUS Zootermopsis From an individual termite's viewpoint, the fundamental problem in encounters with other termites is the correct discrimi­ nation of colony mates from noncolony mates. We examined intraand interspecific interactions among all species of Zootermopsis Emerson in the Pacific coastal states and southwest deserts of the United States (Haverty and Thorne 1989). Those results are summarized here, and new data on response of naive soldiers from incipient colonies are reported. Zootermopsis laticeps (Banks) is morphologically distinct from other members of the genus. This species also has an isolated geographic range in Arizona, New Mexico, and northern Mexico, and lives in dead portions of live trees, usually in riparian situations. In the absence of alates, the morphological characters used to differentiate the described species Z. angusticollis (Hagen) and Z. nevadensis (Hagen) were ambigu­ ous. That imprecision made it impossible to establish valid experimental encounters between colonies. This problem was resolved when Haverty and others (1988) discovered three consistent and distinct cuticular hydrocarbon phenotypes among the two "species," Z. angusticollis and Z. nevadensis, that occur 24 USDA Forest Service Gen. Tech. Report PSW-128. 1991. along the Pacific Coast. The hydrocarbon separations enabled us to isolate a diagnostic character, shape and position of the subsidiary tooth on the right mandible of nonsoldiers, which established unambiguous discrimination of Z. angusticollis from Z. nevadensis (Thorne and Haverty 1989). Z. angusticollis shows one distinct hydrocarbon profile throughout its range (labelled phenotype II by Haverty and others [1988]). Two hydrocarbon phenotypes (I and III) are associated with the single morphological species Z. nevadensis, but each chemical phenotype has a discrete geographic distribu­ tion and may have temporal isolation in alate phenologies (Haverty and others 1988; Haverty and Thorne 1989; Thorne and Haverty 1989; Weesner 1970). In our experiments we investigated whether intraor inter-hydrocarbon phenotypes of Zootermopsis colonies show discrimination in behavioral encounters. In the laboratory we set up pairwise experimental trials in which we placed combinations of various castes from different hydrocarbon phenotypes into petri dish arenas lined with filter paper. Responses were recorded on a ranking system from no noticeable reaction to immediate lethal attack. Observed behav­ iors were typically either docile or highly aggressive, with little intermediate response. General results of these experiments are summarized here (for more detail see Haverty and Thorne 1989). In intraphenotype pairings, agonistic reactions among members of any castes are rare, except between some colonies of Z. angusticollis. Interphenotype interactions frequently stimulate aggressive re­ sponses (fig. 1), but behaviors vary depending on which colonies are involved. Z. angusticollis and Z. nevadensis phenotype III, which occur sympatrically throughout much of their range, are more aggressive towards each other than are Z. angusticollis and Z. nevadensis phenotype I, which have an allopatric distribu­ tion. The two hydrocarbon phenotypes of Z. nevadensis do show discrimination responses in behavioral encounters, and in that way act like distinct species. The behavioral discrimination was further evidence to justify designation of the Z. nevadensis complex into two subspecies, Z. nevadensis, subsp. nevadensis (Hagen) [phenotype I] and Z. nevadensis subsp. nuttingi (Haverty and Thorne) [phenotype III] (Haverty and Thorne 1989). Supplementing experiments reported in Haverty and Thorne (1989), we have recently conducted a series of comparable trials involving young, naive soldiers from experimentally bred labo­ ratory colonies. In July and August 1987, the wings of mature male and female Z. nevadensis subsp. nuttingi alates from a variety of colonies were removed and the alates were paired in small petri dishes lined with moist sand and containing a piece of Douglas-fir wood. Many of these couples successfully reared offspring. The first soldier differentiated when brood size reached approximately 15-20 individuals. In March 1989 we removed these young soldiers from incipient colonies which still con­ tained both primary reproductives, and introduced each soldier to three nymphs or larvae from a haphazardly selected colony of Z. nevadensis subsp. nevadensis [14 trials] or Z. angusticollis Figure 1―Zootermopsis angusticollis soldier attacking a Z. nevadensis nymph. USDA Forest Service Gen. Tech. Report PSW-128. 1991. 25 [10 trials]. Each soldier was used only once (experimental protocol identical to that described for Haverty and Thorne 1989, Experiment 1). In the Z. nevadensis subsp. nuttingi versus Z. nevadensis subsp. nevadensis encounters, naive Z. n. subsp. nuttingi soldiers attacked and killed Z. n. subsp. nevadensis larvae or nymphs in 5 of the 14 trials. In all 14 cases soldiers showed agonistic responses by repeatedly flaring their man­ dibles. Results of the Z. nevadensis subsp. nuttingi soldier encounters with Z. angusticollis larvae or nymphs were consis­ tent: in each of the 10 cases the soldiers attacked and killed the Z. angusticollis. These data on naive soldiers parallel results from the experiments using soldiers from mature field colonies reported in Haverty and Thorne (1989). The generally docile reactions observed in the intraphenotype trials raised the possibility that passive fusion might occur between members of two conspecific field colonies whose gallery networks met in a host log. We have demonstrated experimentally, however, that soldiers, nymphs, and pseudergates will kill and consume functional reproductives from a foreign colony, even within the same hydrocarbon phenotype. Thus, complete colony fusion in nature is unlikely (Haverty and Thorne 1989). Aside from Zootermopsis, the only cases of passive coalescence of termite colonies reported in the literature involve Reticulitermes (Rhinotermitidae) (Clement 1986; Jones 1987) and some populations of Coptotermes (Rhinotermitidae) (Su and Haverty 1991; Su and Scheffrahn 1988). The stimuli responsible for aggressive behaviors in Zootermopsis or other species of North American termites are unknown. Clearly, however, the rapidity, consistency, and effi­ ciency of these aggressive responses are effective in termitetermite interactions. In the remainder of this paper we examine realistic potential uses of these natural behaviors as elements of pest management strategies. POTENTIAL APPLIED USES OF TERMITE AGONISTIC BEHAVIORS Natural aggressive behaviors between termite colonies might be used constructively either in bioassays to learn more about the biology of a particular species or directly in termite control. As an example, in species showing consistent intraspe­ cific aggression, agonistic behaviors could be used to precisely map the foraging territory of a given colony (technique used by Pearce and others 1990). Termites from different locations could be observed in encounters with (A) individuals known to be from the focal colony and (B) members of a geographically distinct, conspecific colony. A passive response to sample A and an aggressive response to B would suggest that termites in the test sample were foragers from the same colony represented in sample A (but responses of termites to neighbors versus strang­ ers has not yet been studied [see Gordon 1989]). An agonistic response to both groups A and B would mean that the termites in the test sample were from a separate (third) colony. Precise knowledge of the foraging areas of a termite colony would be useful in studies of colony population size, growth rate, foraging range, and the density of colonies within a com­ munity. In treatment procedures, knowledge of foraging areas may become crucial as control techniques using bait toxicants are developed. Unfortunately, use of agonistic bioassays in mapping colony foraging areas will not work reliably on any of the two dominant genera of subterranean termite pests in the United States Reticulitermes and Coptotermes), because in­ traspecific mixtures of these termites are generally docile. (Broad variability in degree of agonistic response among ter­ mites from different colonies or populations or both has been noted by Haverty and Thorne 1989; Su and Haverty 1991; Su and Scheffrahn 1988; Thorne 1982.) Perhaps the most promising and intriguing outcome of the study of termite agonistic behaviors is that such work might ultimately lead to an understanding of the recognition code(s) used by termites to discriminate colony mates from noncolony mates. Recognition may involve cuticular hydrocarbon compo­ nents (Howard and Blomquist 1982; Howard and others 1980, 1982); other behavioral and chemical cues are likely involved as well. Once the discrimination process is understood, numerous applied approaches could be designed around the theme of masking or interfering with recognition stimuli. Manipulation of recognition and response cues could be used to enable chemi­ cally loaded intruders to enter a colony, to potentially instigate a civil war, or to impregnate building materials with an offen­ sive, behavioral stimulus to prevent termite colonization. Very generally, any stimulus that aggravates termites is potentially useful in termite control. The scientific goal is to find a stimulus that repels termites, or triggers an avoidance re­ sponse, but it must be a persistent avoidance or repulsion stimulus with little or no habituation. If the termites acclimate to the stimulus, it will no longer elicit the desired response. Control applications involving behavioral bioassays or live termites as an integral part of the pest management program will require trained operators and, typically, a longer treatment period than with chemical applications. Expense is predicted to be higher because of the labor involved in culturing and sam­ pling termites. Integrated approaches may be beneficial in a variety of circumstances. However, in cases where little or no pesticides are desired in the infested area, control methodologies which make use of natural behavioral repertoires may provide effective alternatives. DIRECTIONS FOR FUTURE RESEARCH Studies of Additional Taxa Further studies on patterns of agonistic behavior within and among termite species will broaden our perspective on the potential uses of such aggressive or repulsive responses in termite control. Thus far, agonistic behaviors within and between Zootermopsis species have been studied (Haverty and Thorne 1989). Intraspecific agonism has been studied in Heterotermes aureus (Snyder) (Binder 1988; Jones 1987) and Coptotermes formosanus shiraki (Su and Haverty 1991). The genus Reticulitermes is the most economically important group of wood-destroying insects in the United States, yet little behav­ ioral work has been done on the Reticulitermes species within the United States. This genus should definitely be the target of 26 USDA Forest Service Gen. Tech. Report PSW-128. 1991. future research on intraand interspecific agonistic behavior. Chemoreception The neurobiology of chemoreception has never been exam­ ined in termites, but this approach should be tractable and could help, for example, in our understanding of the role of cuticular hydrocarbons (or fractions of these lipids) in termite-termite recognition processes. Such studies would help us identify natural, biologically active compounds with potential for prac­ tical applications. Detailed Ethological Analyses Many of the behaviors between termites of different colo­ nies may be more subtle than overt aggression, avoidance, or apparent tolerance. Fine scale ethological studies of intercolony interactions must be done to investigate the spectrum between fighting and "passive" responses, and isolate the stimuli evoking each category of response. Evaluation of Experimental Methodologies Laboratory results should be confirmed in the field to validate the experimental protocols used to study dynamics between termite colonies. When intraor interspecific encoun­ ters occur between termites in their natural social and physical environment, does one observe the same passive or aggressive reactions which are seen in laboratory trials involving small subsets of the same two colonies? The methods for studying termite-termite encounters are still being devised and refined. Binder (1988) introduced some innovative approaches in his study of H. aureus. In one experi­ ment he simulated a foraging arena with a balsawood barrier separating termites from separate colonies (dyed to differen­ tiate colony affiliation). A high magnification video camera was used to document movements and behaviors over the 24 hours required for the termites to penetrate the barrier. Similar studies employing video equipment and image tracking software would expand our understanding of termite-termite behaviors. Eventu­ ally, standardization of protocols should be considered to facili­ tate comparisons between studies on related taxa.