The purpose of this paper is to place a number of dates, obtained from two palaeoecological studies on the Murray River between Albury N.S.W. and Tocumwal, and on 3 tributaries, into the context of the late Quaternary evolution of channels and floodplain (Bowler 1978; Page et al. 1991, 1996). Large palaeochannel traces flank the modern rivers of the Riverine Plain. Their stratigraphy and common association with source bordering sand dunes suggests a substantial component of bedload transport, in contrast to the predominantly suspended load transport of the smaller, modern rivers (Bowler, 1978; Page and Nanson 1996). From detailed geomorphic and stratigraphic analyses of the Goulburn River and the Murray River near Echuca, 250 km to the west of Albury, Bowler identified two phases of palaeochannel activity. Based on radiocarbon dates, the Kotupna Phase extended from about 17-29 ka BP, and the Green Gully-Tallygaroopna Phase (GG-T) from 29 ka BP to beyond the limits of radiocarbon dating (Table 1; all radiocarbon ages reported here have been corrected using Bard 1998). The Goulburn Phase of channels with modern features began at least 9 ka BP, leading Bowler to speculate that the conversion from palaeochannel activity occurred in the late Pleistocene, around 17 to 11 ka BP. On the Murrumbidgee River to the north, Page et al. found evidence of 4 phases of palaeochannel activity. Thermoluminescence dating of floodplain deposits suggests these occurred between 13-20, 25-35, 35-55 and 80-105 ka BP. The phases were named, from youngest to oldest, Yanco, Gum Creek, Kerabury and Coleambally. Moreover, Page et al. re-dated some of the Kotupna and GG-T deposits, and found Kotupna (34 ka BP) and GG-T (65-95 ka BP) coincided with the Gum Creek and Coleambally phases, respectively.

Basal sediment infill from 7 billabongs (abandoned channels) with similar features to modern rivers (i.e. Bowler’s ‘Goulburn Phase’) has been dated using radiocarbon. Comparison of radiocarbon ages of base-soluble and insoluble fractions of bulk sediments from 3 billabongs indicates contamination with younger carbon, but suggests the channels were cutoff between about 1,100 and 6,000 years BP. The ages are in agreement with relative ages estimated from proximity to the river, elevation on the floodplain, and channel intersections. Radiocarbon dates on wood embedded in basal sediments from four other billabongs on the Murray River below Yarrawonga, the lower Kiewa River, the lower Ovens River, and the upper Goulburn River (Callemondah Billabong), returned ages of 3,010 BP, 4,720 BP, 4,210 BP and 12,510 BP, respectively. Errors for the age of these and other wood samples (below) were <100 years and are not reported here. The sample from Callemondah Billabong is about 2000 years older than the oldest Goulburn Phase deposit dated by Bowler (1978) from the Murray River. Therefore, it appears that Goulburn Phase fluvial activity has occurred for at least the past 12,500 years, consistent with the conclusion of Bowler (1978) that the conversion to Goulburn Phase activity occurred in the late Pleistocene. However, it is worth noting that late Pleistocene and early Holocene evidence for the activity of small, suspended-load rivers has not been identified from around Albury. For this area, and for the Ovens and Kiewa Rivers, the conclusion relies on the assumption of a synchronous conversion from palaeochannel to Goulburn Phase activity with the Goulburn River.

Palaeochannel traces and low source bordering sand dunes occur north of the Murray River to the west of Albury, adjacent to Goulburn Phase active and cutoff channels. Two palaeochannel deposits were dated using OSL, a point bar deposit and a source bordering sand dune 44 km and 35 km west of Albury, respectively. Because of their close association with the modern river, it was expected that these would represent a terminal phase of palaeochannel activity. The palaeochannel point bar yielded an age of 13.8±2 ka BP, which correlates with late Yanco Phase activity on the Murrumbidgee River (Page et al. 1991, 1996). (Note: the OSL dating was carried out several years ago, and the ages may need to be adjusted slightly to account for recent refinements, e.g. variations in cosmic ray dose. This will be done before publication). Thus, the palaeochannel appears to be from the waning period of palaeochannel activity. In contrast, the source bordering sand dune returned an age of 23.5±2.1 ka BP, which may coincide with Gum Creek/Kotupna Phase activity on the Murrumbidgee and Goulburn Rivers. Therefore, there may be deposits from more than one phase of palaeochannel activity in this region.

Finally, a sand and gravel deposit underlying the floodplain at Albury was dated using OSL on sand and radiocarbon on cellulose from two pieces of wood embedded in the gravels. The floodplain in this area is mainly confined by terraces, but may at times in the past have received sediment inputs from colluvial fans extending from nearby hills (Bowler, pers. comm.). An overbank deposit 2.3 m thick was underlain by about 4.5m of cross-bedded sand, then gravel at about 7m. The sand sample used for OSL was from the middle of the sand unit, and the pieces wood were buried about a meter into the gravels. Both of the wood samples were aged about 12.1 ka BP. The OSL sample returned an age of 10.1±1.3 ka BP. The depth of these sediments is below where the modern river is reworking the floodplain. Moreover, sediments from the gravel pit suggest considerable bedload transport, although with regard to the gravels, it is unclear whether the channel was transporting gravels from upstream or reworking previously deposited gravels or colluvial gravels input from nearby hills. In contrast, modern channels on the Riverine Plain are suspended load rivers. Therefore, it appears that the palaeochannel that deposited these sediments may have been similar to the palaeochannels evident on the margins of the floodplain west of Albury, or at least intermediate between the modern, suspended load channels and the bed or mixed load palaeochannels.

In summary, channels with modern features appear to have been active throughout the Holocene and possibly the closing stages of the Pleistocene, around 10-13 ka BP. However, to be applied regionally, this conclusion assumes the conversion to modern drainage was synchronous in the Goulburn River, the Murray River near Albury, and other tributaries. At Albury, examination of billabongs further towards the margins of the floodplain may uncover older Goulburn Phase channels. Deposits from at least two palaeochannel phases occur adjacent to and underlying the modern channel just west of Albury. These require detailed morphostratigraphic analysis to clarify their relation to palaeochannels to the west and north, and to late Pleistocene climatic conditions.

Table 1. Chronology of sediment deposition on the Riverine Plain (ka years B.P.).