VALLEY FLOOR GULLIES IN WESTERN NSW, AUSTRALIA: PROCESSES AND HISTORIES

Considerable landscape change followed settlement of European pastoralists in semi-arid and arid rangelands of NSW, Australia in the mid to late 19th century, and particularly extensive incision of valley floors. The cause was the change from indigenous hunter-gathering to sheep and cattle grazing. Here, we review evidence from sites widely distributed across the western NSW region, and monitored over varied time-scales, which confirms a post-European settlement time frame for these changes. Age control is provided by radiocarbon dating of charcoal from Aboriginal heat retainer hearths, OSL dating of sediments, burial of fences and other European artefacts, and direct measurement of geomorphic processes over days, years and decades.

Stratigraphic analysis of upland valley fills exposed in the walls of entrenched channels shows that sediments derived from erosion of topsoils on the slopes were initially deposited across valley floors forming a distinctive unit termed post-European material (PEM). This was followed by a shift from relatively shallow, single or multi-thread channels to the wide, flat-bottomed valley floor gullies, equivalent to North American arroyos, characteristic of the region today. Downstream, the systems are characterised by rapid sedimentation, channel avulsion and floodout formation.

Monitoring over the last 20 y indicates that erosion rates outside stream channels may exceed 200 t ha-1 y-1, and that channel enlargement and knickpoint retreat through the upland stream systems are continuing to destabilise valley floors. Mean overall rates are approximately 0.01 m y-1, but more likely much higher following intense rainfall. Downstream, buried fences and abandoned dams indicate that the channels of ephemeral creeks have migrated laterally at rates of 1-2 m y-1 since 1883, and are associated with channel down-cutting of 0.06 m y-1 over the same period.

The range of rates presents challenges to integrating localised data into a more coherent regional picture. Mean long-term rates conflate brief periods of intense activity with far longer periods of little change. However, the overall end result is a redistribution of soil from the more susceptible parts of the landscape to sediment sinks.