A COMPARISON OF 2 METHODS FOR SAMPLING ASSEMBLAGES OF SUBTERRANEAN, WOOD-EATING TERMITES (ISOPTERA)

Within a 50 X 50 m area of wandoo Eucalyptus capillosa woodland in the Western Australian wheatbelt, the diversity and frequency of occurren ce of wood-eating termite species was assessed at two food types. Over a 12 month period, monthly termite activity was determined: (i) at so und/undecayed artificial baits (seasoned wooden stakes of Jarrah, Karr i, Pine, Batu, Oregon; Jarrah sawdust; paper rolls); and (ii) at natur ally occurring timber, fallen logs and branches of wandoo, in varying stages of decay. Termite diversity was 11 species at baits, 18 species at wandoo out of an overall site richness of 21 species. Karri attrac ted the most species (9); sawdust attracted none. At wandoo, Nasutiter mes exitiosus, Coptotermes acinaciformis and Occasitermes occasus acco unted for 59% of samples where termites were recorded. At baits, Heter otermes occiduus accounted for a mean of 80% of samples across bait ty pes, but was rarely sampled at wandoo (5% of samples). Only H. occiduu s, C. acinaciformis and Amitermes neo-germanus ate bait. Pine, Oregon and paper rolls were most effective in attracting foraging termites in terms of highest per cent of replicates showing bait consumption and highest consumption rates. Jarrah and Batu were least attractive to fo raging termites. Samples from wandoo underestimated the relative frequ ency of occurrence of H. occiduus within the study site. Coptotermes a cinaciformis, which attack large food items, and certain species of Am itermes, which forage on subterranean food, may have been underestimat ed by both sampling methods. These findings indicate that a proper und erstanding of the structure of wood-eating termite assemblages within a given area requires a composite sampling strategy which addresses te rmites that eat sound or decayed wood, as well as surface and subsurfa ce foragers.