Abstract
1. I investigate the ecology and physiology of arthropods living in extreme environments, particularly the alpine zone of the Rock and Pillar Range, Central Otago, New Zealand (45°25'S, 170°04'E), and at Cape Bird, Ross Island, Antarctica (77°13/S, 166°26'E).
2. Strategies of insect cold tolerance have been traditionally divided into freeze tolerance and freeze intolerance. Data from the literature are used to extend preexisting categories to encompass the range of freezing tolerance. I conclude that freeze tolerance is not part of a continuum with freeze intolerant strategies, but represents a parallel, alternative strategy of insect cold tolerance.
3. Microhabitat temperature recordings from the Rock and Pillar Range are combined with laboratory freezing and mortality data to make inferences about the thermal ecology of the freeze-tolerant alpine cockroach Celatoblatta quinquemaculata (Dictyoptera: Blattidae). Cockroaches experience repeated freeze- thaw cycles and potentially lethal minimum temperatures in the field. Decreased snow cover as a result of climate change may increase thermal stress by increasing the frequency of freeze-thaw events and resulting in more extreme minimum temperatures.
4. Field water contents of C. quinquemaculata were significantly lower in winter than in summer. Cockroaches could survive losing up to 82% of their body water, and recovered rapidly when water was available. Experimentally dehydrating cockroaches did not affect cold tolerance, suggesting that the seasonal change in field water content is not directly related to cold hardiness.
5. Both abiotic (eg: rock slab size) and biotic factors affected rock refuge choice in C. quinquemaculata, Hemideina maori (Orthoptera: Anostostomatidae) and Neoramia childi (Aranea: Agelenidae) on the Rock and Pillar Range. Fewer rocks were occupied by C quinquemaculata in winter than summer, probably as a result of the selection of alternative winter microhabitats.
6. The Antarctic springtail Gomphiocephalus hodgsoni (Collembola: Hypogastruridae) utilises a freeze avoiding strategy to survive tempera tines below -35°C. The mean low group supercooling point (LGSCP) in October was -35.4°C/ and G. hodgsoni accumulated massive amounts (>80 Mg/mg dry weight) of glycerol. In spring, haemolymph osmolality was high (up to 2518 mOsm/kg) and thermal hysteresis proteins were present. By midsummer, LGSCP had increased to - 28.3°C, glycerol content had decreased to <10 jig/mg dry weight, haemolymph osmolality was c. 650 mOsm/kg, and thermal hysteresis activity was undetectable.
7. Cloches were used to passively increase temperatures for 27 months at Cape Bird and investigate invertebrate responses to climate change. Invertebrate populations showed no consistent changes, which suggests that temperature may not be the main factor limiting invertebrate populations in this environment.
8. The abundance and distribution of terrestrial invertebrates was investigated across a stream at Cape Bird. The moisture environment became more homogeneous after the stream dried up, but only Stereotydeus mollis (Acari: Prostigmata) changed its distribution in response to the changing conditions. The environs of the stream had higher water availability, algal biomass, organic content and decreased salinity relative to areas away from the stream, and nematodes, rotifers, Protozoa and mites were associated with this environment, whereas G. hodgsoni was associated with the riparian margins.
9. The distribution of terrestrial invertebrates was investigated over 12 km2 at Cape Bird. Invertebrate communities were not dependent on omithogenic input. Arthropod presence was strongly related to soil algae, but collembolan distribution appeared decoupled from the presence of macroscopic vegetation, which has important implications for vegetation-based selection of terrestrial Antarctic areas for protection.