Alan S. Palmer

Loess up to 10 m thick covers marine terraces, old river aggradation terraces, and gentle hill country slopes in Wairarapa, New Zealand. In the drier eastern and southern parts of Wairarapa, the loess is subject to dry summer and autumn and wet winter and spring in most years. Under these conditions, and in association with strongly seasonal evapotranspiration, the loess has formed Pallic Soils. The objective of this study is to define physical and geotechnical properties to explain the field behaviour of the loess in order to support applications in agriculture and engineering works. From the outset, Wairarapa loess may have been already more rich in clay than many other deposits around the world, because of the Mesozoic greywacke–argillite parent rock from which it was derived. Wairarapa loess is characterized by being very compacted, with subsoil fragipans with densities up to 1.8 Mg m ⁻³ . In winter, water perches on the impermeable fragipan, saturating the soil above, often to the surface. Wetting and drying in this zone promotes weathering, clay formation and translocation. Ohakean and Ratan loess deposits are more silty than the underlying clay‐rich Porewan loess. Whereas the former loesses blew from extensive low angle aggrading fans, at least some of the Porewan loess was derived from lacustrine deposits on the floor of Wairarapa Valley. Textures of all Wairarapa loess, and particularly the Porewan loess, are clay‐rich. The weathering regime and textures result in extremely low hydraulic conductivity for the loess. Atterberg limit tests indicate that the loesses, particularly the Porewan loess and underlying lacustrine clays, are poor to very poor construction materials. Atterberg limits also help to understand the behaviour of wet Pallic Soils under stocking and traffic. Pallic Soils compact readily when moist and are very weak when saturated, acting as liquids when disturbed, resulting in extreme pugging, dispersion, and soil loss. Pallic Soils in Wairarapa are prone to shallow landslides, even on gentle slopes because of their loss of strength when saturated.

Loess in Aotearoa New Zealand (ANZ) has been studied since its first documented recognition (on Banks Peninsula) in 1878 by Julius von Haast. A decade later, John Hardcastle revealed that southern ANZ loess was both glacial in origin and contained signals of past climates. As a fine‐grained aeolian deposit dominated by quartz and feldspar (± mica), it is derived mainly from greywacke and schist rocks shattered by frost cracking at high elevations, fluvially comminuted then deflated from aggrading floodplains, and from exposed continental shelves during cold periods. Such quartzo‐feldspathic loess predominates in eastern and southern parts of both South and North islands and in Westland and Tasman. In addition, subsurface tephric loess prevails in central‐western North Island. Unlike many deposits overseas, ANZ loess generally lacks secondary calcium carbonate, is denser (with lower macro‐ and mesoporosity, higher clay content) and, although less prone to collapse, is susceptible to accelerated erosion including tunnel gullying and landsliding. Mean rates of accumulation in the Last Glacial Maximum were ~3−25 mm/century; mass accumulation rates were generally 70−150 g cm ⁻² yr ⁻¹ but locally could be as low as 20 g cm ⁻² yr ⁻¹ and as high as 360 g cm ⁻² yr ⁻¹ . Studies mapping and characterizing loess have been driven in part by cognizance of its importance as a widespread parent material (~60%−70% of ANZ's soils contain loess) for arable/pastoral soils underpinning ANZ's predominantly agronomy‐based economy. A key feature of ANZ loess studies has been the recognition of developmental upbuilding pedogenesis involving syn‐depositional alteration of loess as it accumulates in cold and/or cool (stadial) periods, and stronger alteration (forming more‐developed soils, which subsequently become recognizable as buried soil stratigraphic units or paleosols) during minimal accumulation in interglacial and/or warm (interstadial) periods. Multisequal loess‐paleosol successions are the result. ANZ loess‐paleosol successions have been dated and correlated via a range of dating and age‐equivalent techniques including tephrochronology and paleomagnetism. Most loess is ≤500 kyrs old, but occurrences as old as Pliocene (Otago) and c. 1.7–1.0 Ma (Waikato, Wairarapa) are recognized. Loess chronostratigraphy has enabled loess and associated buried soils/paleosols to be correlated to the Quaternary marine oxygen isotope records, which, alongside mapping efforts, have revealed landscape responses to climate and tectonic controls particularly along the eastern regions of ANZ bordering the Hikurangi Subduction Margin.

Effects of global warming on soil organic carbon (C) can be investigated by comparing sites experiencing different temperatures. However, observations can be affected by covariance of temperature with other environmental properties. Here, we studied a thermal gradient in forest soils derived from volcanic materials on Mount Taranaki (New Zealand) to disentangle the effects of temperature and reactive minerals on soil organic C quantity and composition. We collected soils at four depths and four elevations with mean annual temperatures ranging from 7.3 to 10.5 °C. Soil C stocks were not significantly different across sites (average 162 MgC ha⁻¹ to 85 cm depth, P > .05). Neither aluminium (Al)-complexed C, nor mineral-associated C changed significantly (P > .05) with temperature. The molecular characterisation of soil organic matter showed that plant-derived C declined with increasing temperature, while microbial-processed C increased. Accompanying these changes, soil short-range order (SRO) constituents (including allophane) generally increased with temperature. Results from structural equation modelling revealed that, although a warmer temperature tended to accelerate soil organic C decomposition as inferred from molecular fingerprints, it also exerted a positive effect on soil total C presumably by enhancing plant C input. Despite a close linkage between mineral-associated C and soil organic C, the increased abundance of reactive minerals at 30–85 cm depth with temperature did not increase soil organic C concentration at that depth. We therefore propose that fresh C inputs, rather than reactive minerals, mediate soil C responses to temperature across the thermal gradient of volcanic soils under humid-temperate climatic conditions.

Eolian mineral dust is an active agent in the global climate system. It affects planetary albedo and can influence marine biological productivity and ocean‐atmosphere carbon dynamics. This makes the understanding of the global dust cycle crucial for constraining the dust/climate relationship, which requires long‐term dust emission records for all major dust sources. Despite their importance, the sources of atmospheric dust deposited in the Southern Ocean remain poorly constrained. Eolian dust in the Pacific sector of the Southern Ocean is generally assumed to originate from Australia, with minor contributions from New Zealand. Here, we present a high‐resolution elemental record of terrestrial inputs for the past ∼410 kyr from marine sediment core PS75/100‐4 recovered from east of South Island, New Zealand. Sediment grain size is slightly finer than that of loess deposits from South Island, New Zealand, and is coarser than that of marine sediments in the Tasman Sea to the west of New Zealand, which indicates that the dust originated mainly from New Zealand and not only from Australia. Core PS75/100‐4 records lithogenic mass accumulation rates ranging from ∼0.01 to 0.69 g/cm²/kyr (∼0.20 g/cm²/kyr average), with variations over a factor of ∼3–4 over glacial versus interglacial timescales for the past 410 kyr. Our geochemical data correlate well with the Southern Ocean and Antarctic eolian dust records and suggest a westerly wind‐supplied dust signal from New Zealand. Our findings, therefore, suggest that New Zealand should be considered an important long‐term regional dust source in global dust cycle models.