The difficulty of reliably removing an established lucerne pasture before cropping has been identified as a major problem with phase-farming systems on mixed farms. A series of experiments were undertaken on established lucerne stands at the Ginninderra Experimental Station in the Australian Capital Territory (ACT) and at the Temora Research Station in southern New South Wales (NSW) to compare the ability of grazing, either alone or in combination with herbicides, to remove a lucerne pasture. A pilot study at the Ginninderra Experimental Station in 1998–99 utilised a high stocking rate (30 dry sheep equivalents [dse]/ha), while the main study at the Ginninderra Experimental Station and the Temora Research Station in 1999–2000 used a lower stocking rate (10–12 dse/ha) considered to be closer to farmer practice in the region. Continuous grazing at the high stocking rate removed 73% of the lucerne stand over 3 months. In contrast, the stocking rates applied in the main study proved too low to substantially impact on lucerne survival at both the ACT and NSW sites (13–23% removal) under the good seasonal conditions experienced in 1999–2000 even though the pastures were continuously grazed for 8–9 months. The use of herbicides both alone or in conjunction with grazing greatly improved lucerne removal in both studies. However, herbicide efficacy was variable (53–100% removal), and seemed to be related to the time of year it was applied, the period of lucerne regrowth or the amount of rainfall before herbicide application.

An additional on-farm study was undertaken near Junee Reefs in southern NSW between 2001 and 2003 that compared the survival of a range of lucerne cultivars under simulated hay-cutting (mown) and commercial grazing regimes. Dry conditions during 2002–03 resulted in a decline in lucerne frequency at about 50% in mown control plots. Grazing increased the stand decline, particularly for many winter-active cultivars where lucerne frequency was reduced by up to 70–93%.

It was concluded that: (i) continuous grazing for prolonged periods can be effective at removing lucerne, but the rate of lucerne loss will be influenced by both stocking rate and rainfall; (ii) the application of herbicides can improve lucerne removal either in association with, or in the absence of grazing; however, herbicide efficacy appeared to be dependent upon the physiological status of the lucerne plants and/or the environmental conditions before application; and (iii) cultivar responses indicated that lucerne types could potentially be developed for phase-farming systems with increased susceptibility to grazing mismanagement.