Compound trace fossils formed by plant and animal interactions: Quaternary of northern New Zealand and Sapelo Island, Georgia (USA)

Quaternary coastal-terrestrial deposits from Aupouri and Karikari Peninsulas, northern North Island (New Zealand), and from Sapelo Island, Georgia (USA), reveal sedimentary structures produced by plant roots that provided habitats suitable for root-sucking, deposit-feeding, burrowing, brooding and other terrestrial invertebrate activities. The structures are typically large (<1 to >2 m in length, with widths from a few to >45 cm), and consist of white sand-filled tubes that are circular to subcircular in cross-section, and cylindrical to downwardly tapering in longitudinal view. Dark brown (humic and ferruginous) haloes demarcate the white sand fill. These structures open upwards, often connecting into overlying palaeosols, and are inferred to reflect the root architecture of several large forest trees that are morphologically comparable with those of some modern conifers, e.g. kauri (Agathis australis) in New Zealand and loblolly pine (Pinus taeda) in Sapelo Island. The white sand cores of these structures are a passive fill originating from a podzolised horizon, and/or from drifting aeolian sands. The root structures were temporarily open to the Quaternary surface and represent cavities remaining after tree death and/or toppling. The yellowish host sands are frequently mottled by trace fossils, as are boundaries between outer dark brown haloes and inner white sand fill to the root structures. Many of these traces are small, of simple form, and cannot readily be ascribed to any ichnotaxon. Meniscate burrows (Taenidium) that have been identified from both localities were probably produced by cicada nymphs. The moist and sheltered tree root-protected environment persisted for some time after tree death and was a desirable microhabitat for a number of invertebrates. Around the margins of these root casts, trace fossils and tiering fabrics may cross one another irregularly, develop oblique to primary bedding surfaces, or can even be inverted. Such stratigraphically disjunct relationships could be misleading in structural and palaeoenvironmental assessment of older or tectonically deformed strata that include palaeosols. These observations add an additional dimension to reconstructions of ancient forest cover and terrestrial continental environments.