Phytoliths from subantarctic Campbell Island: plant production and soil surface spectra

Abstract

Phytolith (plant opal) production, and its preservation within the soil surface, is described for the first time in the subantarctic region from Campbell Island, ca. 600 km south of New Zealand, forming the basis for a new modern reference collection for the region. Plant samples (many from species endemic to the island and vegetation community dominants) were sourced from Campbell Island and from gardens in New Zealand. Short and long cell phytoliths are described from four graminoids ( Chionochloa antarctica, Poa litorosa and Poa sp. (Poaceae), Carex trifida (Cyperaceae)), four forbs ( Acaena minor var. antarctica and Acaena anserinifolia (Rosaceae), Myosotis capitata (Boraginaceae), Anaphalioides bellidioides (Asteraceae)) and a shrub ( Hebe elliptica (Plantaginaceae)) from 23 analysed taxa. The remainder were found to be non-producers, including the distinctive macrophyllous forbs or ‘megaherbs’. Soil surface samples collected with 11 of the plants from Campbell Island contain abundant, predominantly graminoid-type phytoliths. Two-sample Kolmogorov-Smirnov tests indicate the soil surface assemblages beneath different vegetation community types are nevertheless similar to each other. This is interpreted to be a direct result of widespread grasses in the island's vegetation that produce abundant phytoliths of robust morphology. The over-representation of grass phytoliths in the soils may also be explained by the absence of other non-grass phytolith producer taxa in the vicinity of the collection locality, or the dissolution or fragmentation in the soil of fragile forms. Localised concentrations of sedge ‘hat-shaped’ phytoliths in two soil surface samples result in significantly different assemblages from the other collection sites, suggesting that despite the high proportion of grass-type phytoliths, there is still potential for differentiating vegetation community types on Campbell Island on the basis of the dispersed soil surface assemblages. With further knowledge about modern phytolith production in the subantarctic, this technique may then be applied to regional fossil assemblages. The predominantly grass-type phytoliths described during this pilot study differ significantly from the Oligocene and Miocene tree/shrub-dominated assemblage of the Cape Roberts Project cores from the Ross Sea region of Antarctica, interpreted to have been sourced from a vegetation growing in similar, but less oceanic, climatic conditions. The reasons for this disparity, reflected by the terrestrial palynomorph records of previous studies undertaken on the Cape Roberts Project cores and present day Campbell Island pollen rain, remain unclear, but may involve differences in the growing environment or different continental origins of elements of the vegetation.