A continuous ∼5280 calendar (cal.) yr long cryptotephrostratigraphic record of a peat core from northern New Zealand demonstrates that cryptotephra studies can enhance conventional tephra records by extending the known distribution of ash fall and enabling re-assessment of volcanic hazards. A systematic sampling strategy was used to locate peaks in glass-shard concentrations and to determine loci of individual geochemical populations, and a palynological method involving spiking samples with Lycopodium spores was adapted to facilitate accurate counting of glass-shard concentrations. Using glass shard major element compositions, and a core chronology based on eight AMS 14C ages and two visible macroscopic tephra layers, Taupo Tephra (Unit Y) (1688-1748 cal. BP) and Tuhua Tephra (6800-7230 cal. BP) (2cr-age ranges), four cryptotephras were correlated with known eruptions: Whakaipo (Unit V) (2743-2782 cal. BP), Stent (Unit Q) (4240-4510 cal. BP), and Unit K (4970-5290 cal. BP), erupted from Taupo Volcanic Centre, and Whakatane Tephra (5470-5600 cal. BP) erupted from Okataina Volcanic Centre. Mixed glass populations were found in the core, most likely an artefact of post-depositional remobilization of shards vertically (both up and down) in the peat or on its surface by wind, or a result of closely spaced eruption events, or a combination of these. A secondary glass population identified within the macroscopic Taupo Tephra was tentatively attributed to either an earlier phase within that eruption or to mixing with a slightly older Taupo-derived eruptive or (less likely) a currently unknown Okataina-derived eruptive. These results indicate that, in the absence of continuous cryptotephrostratigraphic analysis, a peak in shard concentrations may not in itself be indicative of the ‘true’ stratigraphic (ie, isochronous) level of a tephra layer. For cryptotephra studies of peat cores, we recommend (1) using a detailed sampling strategy for the analysis of distal tephra-derived glass to detect and account for any mixed populations and possible vertical spread of glass shards through the peat, and (2) analysing more shards from larger samples to help ‘capture’ sparsely represented cryptic andesitic tephra deposits.

No abstract

In comparison with temperature reconstructions, New Zealand proxy records for paleo-precipitation are rare, despite the importance of precipitation in contemporary climate variability and for projected climate impacts. In this study, records of mid-late Holocene palaeomoisture variation were derived for two hydrologically separate ombrotrophic Restionaceae bogs in northern New Zealand, based on peat humification analysis. At each site, three cores were analysed for peat humification, facilitating both intra-and inter-site comparisons. Age models for the six sequences were developed using radiocarbon dating and tephrochronology. Twelve tephras (including six cryptotephras) were recognised, four of which were used to precisely link the two sites and to define start and end points for the records at 7027 ± 170 (Tuhua tephra) and 1718 ± 10 cal. yr BP (Taupo tephra) (2σ-age ranges), respectively.We find individual differences between the six peat humification records at short-term timescales that are presumably due to local site factors, in particular changing vegetation and microtopography, or to changes in the composition of the material analysed. Stronger longer-term coherence is observed between all six records but is attributed to slow anaerobic decay over time because the implied trend towards wetter 2 summers in the late Holocene cannot be corroborated by independent climate proxies. Despite these confounding factors, centennial scale shifts in bog surface wetness are a pervasive feature of all six records with varying degrees of overlap in time that show strong correspondence with El Niño-Southern Oscillation reconstructions from the eastern equatorial Pacific. These results indicate the potential for peat humification records from New Zealand's ombrotrophic bogs to elucidate past climate variability and also demonstrate the importance of developing multiple well-dated profiles from more than one site.

No abstract