DESICCATION RESISTANCE IN 6 SUB-ANTARCTIC WEEVILS (COLEOPTERA, CURCULIONIDAE) - HUMIDITY AS AN ABIOTIC FACTOR INFLUENCING ASSEMBLAGE STRUCTURE

1. On the usually wet, sub-Antarctic Marion Island, interspecific competition for refugia during low humidity or dry periods is thought to have been responsible for patterns of body-size and microhabitat distribution of the ectemnorhinine weevils inhabiting the cryptogam-dominated epilithic biotope, whereas this process is thought to be unimportant in the angiosperm-dominated vegetated biotope. 2. Adults of the six weevil species indigenous to the island were tested in the laboratory for their ability to resist desiccation at 10-degrees-C and 5% rh., and to replenish lost water by drinking, and these results were compared with the body water content of field-collected individuals. 3. Once corrected for body mass, no meaningful differences in maximum tolerable water loss were found between the species. However, time to maximum water loss and rate of water loss were inversely related and showed two clear trends. Species inhabiting dry, rockface habitats had relatively low rates of water loss, took longer to reach the maximum tolerable water loss, and had a relatively high body water content, whereas species from moist, angiosperm-dominated habitats had a relatively high rate of water loss, reached maximum tolerable water loss faster, and had a relatively low body water content. 4. The very small (1.55 mg), fellfield-inhabiting Antarctonesiotes elongatus differed from the other epilithic species by having rates of water loss similar to species from the moist lowland habitats and larvae of the supra-littoral Palirhoeus eatoni are known to have a greater tolerance to salt-water than any of the other epilithic species. 5. These results suggest that in at least two of the syntopic species pairs showing body-size differences, these differences are unlikely to have promoted coexistence via the elimination of competition for refugia due to similarities in desiccation resistance, although this may have been the case in a third pair. The overriding difference in physiology between Palirhoeus eatoni and the remaining epilithic species suggests that Palirhoeus eatoni should be omitted from the epilithic biotope species pool when generating species pairs for null distributions.