Effects of sowing treatment and landscape position on establishment of the perennial tussock grass Themeda triandra (Poaceae) in degraded Eucalyptus woodlands in south-eastern Australia

Grassy woodlands dominated by Eucalyptus species such as E. albens Benth., E. melliodora A. Cunn. ex Schauer, and E. microcarpa Maiden are poorly conserved in Australia, owing to widespread agricultural development. Understorey restoration is vitally important to enhance degraded remnants but no reliable techniques are available to restore herbaceous understorey species over large areas. Reestablishment of dominant native grasses such as Kangaroo grass (Themeda triandra Forssk.) is particularly important. This study compared Themeda establishment using a variety of sowing techniques across a range of landscape positions in degraded woodlands in central New South Wales. Four localities were sampled at three landscape positions (upper, mid-, and lower slopes). Existing vegetative cover was sprayed with herbicide and removed by mowing. A randomized block experiment was established at each site with five replicates of five seedbed treatments: all four factorial combinations of soil disturbance (disturbed, nondisturbed) and weed control (atrazine, nonatrazine), plus topsoil scalping followed by soil disturbance. After 117 days, all plots were subject to a common herbicide (atrazine) treatment to control weeds. Themeda establishment was not significantly affected by landscape position, despite significant differences in the cover of exotic species among landscape positions. The scalped/disturbed treatment resulted in significantly better establishment at 400 days (18% or 3.9 plants/m2) than the control (8% or 1.8 plants/m2). All other treatments did not differ significantly from the control suggesting that seedbed treatments are not critical to successful Themeda establishment. These results suggest that Themeda swards can be reestablished in degraded woodlands relatively simply. The similarity in outcomes across all landscape positions suggests that general recommendations for "best-bet" establishment methods may prove robust under a wide range of environmentaland seasonal conditions.