Abstract From Dr Tim Cohen's Thesis

The partly confined valleys of south-eastern Australia provide suitable conditions for the formation of vertically accreted floodplains with laterally stable channels. Three reaches in the Bellinger catchment in the New England Fold Belt on the mid-north coast of New South Wales (NSW) provide sites to assess the attributes of confined floodplains and the impact of European settlement on otherwise highly stable systems. The nature of Late Quaternary floodplain processes in a bedrock-dominated landscape are investigated, providing the evolutionary context for contemporary channel processes.

The Bellinger catchment is characterised by an assemblage of stepped Late Quaternary alluvial units. Late Pleistocene terraces represent large more competent rivers that reworked almost entire valley floors, however, a progressive decline in discharge since the Last Glacial maximum has resulted in the abandonment of these deposits as elevated terraces or residual alluvium onlapped by contemporary floodplains. The Bellinger catchment exhibits evidence of intrinsic controls on floodplain formation superimposed over an early-mid Holocene climatic signature. A fluvially active period from 12 - 3 ka reworked late Pleistocene terraces and is termed the Nambucca Phase. In the Bellinger catchment, two floodplain surfaces, one higher than the other, both started to vertically accrete from 4 ka onwards, but with some valley locations remaining vulnerable to episodes of reworking resulting in substantial units of younger alluvium. The high floodplain is dominated by horizontally laminated, vertically accreted sequences, while the low floodplain is characterised by pronouned cut-and-fill stratigraphy. In both instances, vertical processes are the dominant mode of floodplain construction. However, an extensive AMS radiocarbon chronology supplemented with a limited OSL investigation suggests that these two surfaces are not chronologically distinct. In contrast, polycyclic terraces and floodplains can share much the same elevation but be very different in age. The assumption that the continuity of terrace or floodplain profiles along a valley represents coeval formation is shown to be frequently invalid for such confined landscapes.

European settlement from c.1840 transformed the fluvial environment, initiating phases of channel metamorphosis that do not correspond to a previously accepted model of channel change based on proposed cyclical changes in flood activity termed flood- and drought-dominated regimes. Largely in response to deforestation, the Bellinger River has undergone rapid adjustment to changing boundary conditions including measurable channel straightening, a three-fold increase in width, a five-fold increase in channel capacity, bed level incision and a re-configuration of riffle-pool units. These changes have occurred in periods of above- and below-average flood activity and are a direct result of landscape clearance for agriculture, compounded by activities such as within-channel aggregate extraction. In the latter part of the 20th century, channel capacity has continued to increase despite recent increases in reparian vegetation and a decline in flood frequency.

Wide-scale post-settlement entrenchment has produced a diverse range of in-channel depositional and erosional landforms that are not the product of particular discharge return-periods. Deposited over gravel-bar platforms, a variety of cut-and-fill benches have developed and their stratigraphy and form are controlled by their position within the channel, local sediment supply and local energy conditions. Bench processes, along with continued channel expansion, are attributes of a highly disturbed post-European system that currently displays non-equilibrium characteristics.

This significant revision of our understanding of the controlling processes and changing environment in the partly confined rivers of northern NSW has important implications for the management and future rehabilitation of these disturbed systems. The future success of river management in these valleys requires a reach-scale assessment of post-European channel responses, but framed within the context of the longer-term channel and floodplain formation processes.