Patterns in the distribution and abundance of terrestrial arthropods on sub-Antartic Marion Island

Abstract

Quantitative studies of complete soil arthropod communities are considered essential if a thorough understanding of the structure and dynamics of soil faunas, and their likely response to anthropogenic environmental change, is to be gained. In practice, however, high species richness and poor taxonomic knowledge for most systems often hampers the acquisition of such information. In consequence, many studies resort to the use of higher taxa or more qualitative data. Sub-Antarctic terrestrial systems which are moderately diverse with well-known faunas allow us to bridge some of these problems. Two major terrestrial habitat types (or biotopes) are distinguished on sub-Antarctic islands, namely the epilithic and vegetated biotopes. These two habitat types differ in terms of current extent of vegetational cover and age relative to previous glacial events. The epilithic biotope includes the older habitats such as rocky shores and fellfields that have escaped glaciation, have little vascular plant cover, and is cryptogam¬dominated. Quantitative data on the soil fauna of a fellfield habitat on sub-Antarctic Marion Island is used to illustrate the pitfalls associated with the qualitative approach. Fellfield is an important vegetation complex in the broad Antarctic region, and on Marion Island consists of both bare rocky areas and cushion plants. Soil arthropod communities in these two distinct habitat components were quantified over one year. Species richness was relatively high (42 spp.) and consisted almost exclusively of indigenous species, although abundances were lower compared with less extreme vegetation types in the sub-Antarctic. In general, arthropods either showed no pronounced seasonal peak in abundance, or a summer peak, although these patterns differed between habitat components within species, and between species. Quantitative analyses highlighted prominent differences between the two habitat components in arthropod community structure, despite the fact that most species were common to both of them. Qualitative analyses, in contrast, obscured these differences, while the use of higher taxa for analyses of seasonality resulted in abundant species generally masking the phenologies of less abundant ones. It is concluded that quantitative analyses of soil arthropod communities are essential if natural and anthropogenic changes in their structure are to be detected, monitored and understood. Studies on sub-Antarctic insects have suggested that species inhabiting the epilithic habitats exhibit higher habitat preference or specificity and also a higher incidence of biotic interactions, such as competition, than those in vegetated habitats. The Acari, a more speciose but less well studied group than the insects, is used to independently test the aspect of habitat specificity. Seventeen different habitats or zones belonging to both the epilithic and vegetated biotopes were quantitatively sampled for mites over a one year period. These included a rocky shore, mire and non-mire lowland vegetation, and a mid-altitude fellfield. Species richness across all habitats was 39 spp., while rocky shore habitats showed higher abundances, but lower species richness, and a distinct fauna, to strictly terrestrial habitats. Multivariate analyses indicated that mite assemblage structure differed significantly between all different habitat types. However, most species in the terrestrial habitats (both epilithic and vegetated) were shared, suggesting that the epilithic fellfield, and especially its vegetated component (Azorella selago cushion plants) would have been suitable refugia during glaciation. The shoreline, due to its distinct fauna, seems a less likely refuge from which recolonisation of vegetated habitats could have taken place. Most species could be considered habitat generalists, although the epilithic habitats (shore and fellfield) had more habitat specific species than the lowland vegetated habitats. Bothrometopus elongatus is one of four Ectemnorhinus-group species restricted to the epilithic biotope on the Prince Edward Islands. The biology of this species was examined over a full year at Kerguelen Rise, a mid-altitude fellfield site on Marion Island. B. elongatus adults eclose from April onwards, reaching maximum densities (c. 17 individuals.m-2) in September. Females mature approximately three eggs at a time and these commence hatching in July. Larval eclosion reaches a peak in November, during which time larval densities are also highest (c. 153 individuals.m-2). The larvae develop through six instars, which is within the range found for other Ectemnorhinus-group species. The high densities of B. elongatus in fellfield habitats, and its single, virtually discrete annual generation, make this species unusual among insects indigenous to the sub-Antarctic. The latter generally have low densities compared to other micro-arthropods, prolonged life cycles, and flexible life histories. It is suggested that the diversity of life histories found amongst the indigenous insects at Marion Island presents considerable potential for testing environmental effects on insect life histories. An overview of sub-Antarctic insect life history data suggests that the indigenous species, with generally prolonged life cycles, are at a disadvantage relative to introduced ones that have more rapid cycles and often complete several generations per year. This is reason for concern given rapid climate change at these islands.