Food location and discrimination by subterranean termites (Isoptera: Rhinotermitidae)
  Key Words  
  Coptotermes formosanus, subterranean termites, foraging, tunneling
  J.K. Grace and C.E. Campora
  Tunnel construction by Coptotermes formosanus Shiraki is initially random, with tunnels radiating outward in a uniform circular distribution from the point of origin. The degree of branching of these tunnels may be species- specific, and reflect the pattern of occurrence of food resources in different environments. As open spaces, wood, or inedible objects are encountered in the soil, tunnels outward from these substrate anomalies are no longer distributed randomly. Rather, multiple exit tunnels are constructed in the same general direction as the entry tunnel. In foraging arenas, these exit tunnels never reversed direction. When an open space was encountered, the directional vector of the exit tunnels was less pronounced due to the large number of tunnels fanning outward from the open area. Termite tunneling shifts from the original uniform distribution into a more localized and directional distribution as food sources are discovered. Any environmental cues (e.g., guidelines, thermal shadows) may alter the tunnel distribution. At a very short distance, it is likely that gradients from the diffusion of both gases and plant or fungal metabolites provide directional cues. Trail pheromone is likely to play a major role in recruitment of foragers to acceptable food sources. If a single component is responsible, then gradients appear necessary to maintain this behavior, since acclimation can occur with constant continued exposure. An ephemeral recruitment component may be deposited along with a more stable orientation component. Wood treated with disodium octaborate tetrahydrate was gradually avoided over a period of several days, with termite foraging activity declining in tunnels in the vicinity of the treated wood, and increasing in portions of the arena containing untreated wood. Termite mortality did not occur in the vicinity of the treated wood, so the avoidance cannot be attributed to a necrophobic response. When the positions of borate-treated and untreated wood samples were switched in these studies, termites investigated the borate-treated samples in their new locations, followed by gradual avoidance. Termite return to the sites that had been previously occupied by borate-treated wood was very slow, and ceased immediately when borate-treated samples were again placed in these same locations. Resources appear to be mapped within the established foraging network with respect to their acceptability.