Investigations of plant subfossil cuticles at a Holocene raised bog complex, northern New Zealand
Subfossil plant cuticles, the very resistant waxy layer covering vascular land plants, are a neglected source of information in peat studies, despite their high preservation and identification potential. A lack of standardised methods and reference material are major contributing factors. In this thesis, a new method is introduced to test if subfossil plant cuticles from Moanatuatua Bog in the northern North Island of New Zealand can give a robust reconstruction of local bog surface vegetation changes during the Holocene. The method was successfully established and applied at coarse sampling resolution to show vegetation changes across the full length of the core and at fine sampling resolution around charcoal layers to reconstruct the post-fire response pattern of the main plant species on the bog. Additionally, bulk density and organic matter analyses were carried out to provide further insight into these changes. At the core site, towards the southern margins of Moanatuatua Bog, swamp forest had developed by 15000 cal yr BP. Until ca. 10500 cal yr BP, the vegetation assemblage was sedge-dominated, indicating swamp and/or fen conditions. A significant increase in macroscopic charcoal particles coincided with the transition to a more diversified vegetation composition. At around 4500 cal yr BP, the vegetation became restiad-dominated, indicating full raised bog conditions. The coarse resolution cuticle results were further compared to a pollen record from the same sequence, which was established independently. This comparison showed that plant subfossil cuticles can provide additional information to pollen analysis in cases where pollen is hard to identify or poorly preserved. Specifically, restiad pollen is hard to differentiate, yet cuticles of Empodisma and Sporadanthus have very distinct features. Also, Cyperaceae pollen is very poorly preserved at Moanatuatua Bog and the Cyperaceae pollen curve shows a poor match with the Cyperaceae cuticle record. It is suggested therefore that Cyperaceae pollen at this site – and potentially other peat sites – is a less reliable indicator of local sedge communities than a Cyperaceae cuticle record. At fine resolution, results were blurred across a time interval that was marginal for reconstructing response patterns due to the constraints imposed by sampling resolution and peat accumulation rate of Moanatuatua Bog. Nevertheless, two out of three charcoal layers recorded a local fire on the bog surface, with one layer displaying the expected vegetation response. After the fire, Empodisma, as a mid-successional species, re-established on the bog surface before Sporadanthus, a late-successional species. The other layer was dominated by sedges and showed no response pattern, as is to be expected due to the very fast recovery of sedges. In general, sample preparation for cuticle analysis proved to be fast with relatively little equipment or chemicals needed. With detailed reference material, identification to species level is possible due to distinctive and pronounced cuticle features. Plant cuticle analysis is therefore proposed to be a reliable tool to reconstruct long-term and short-term vegetation changes from peat sequences.