Integrated shrub management in semi-arid woodlands of eastern Australia: effects of chemical defoliants applied after an initial disturbance

Abstract

Previous fire experiments using artificial fuel have shown that annual fires, especially those applied in the autumn, can effectively control coppicing understorey shrubs in semi-arid woodlands. Such frequent fire is impossible to apply under natural conditions given the limited time available for sufficient herbage fuel to accumulate. Preliminary screening studies were therefore undertaken to test the hypothesis that chemical sprays applied at concentrations less than those normally recommended could be used to mimic high-frequency experimental fires. The effectiveness of 11 chemicals (7 arboricides and 4 dessicants) applied at a range of concentrations was assessed on one site by spot-spraying 5-year-old coppice regeneration of Eremophila mitchellii (budda or false sandalwood) and E. sturtii (turpentine). Chemical activity was assessed by regularly monitoring leaf effect, i.e. by rating the degree of leaf discolouration, scorching, blackening and ultimately leaf fall, over the ensuing 12 months following treatment. Arsenal� and Roundup CT� induced the highest shrub mortalities across all size classes while mortality rates were consistently higher for E. mitchellii than for E. sturtii. A second experiment involved 5 chemicals (4 arboricides and 1 dessicant) applied in a similar manner to 7-year-old seedlings of Cassia nemophila (syn. Senna nemophila) (punty bush). Significant damage to foliage (> 80% leaf effect) of all 3 shrub species was recorded 2 months after treatment with either Roundup CT� or Roundup� (i.e. either 450 or 360 g/L glyphosate respectively), as well as with Arsenal� (250 g/L imazapur + 60 g/L isopropylamine) but only at the highest concentration (i.e. 100% of the 'recommended' rate). In some lower concentration treatments, leaf effect was still increasing 6 months after treatment.In a second series of screening experiments involving 1- and 2-year-old coppices sprayed in autumn and spring, significant interaction occurred between coppice age and season of spraying when averaged over both Eremophila species. At lower concentrations (i.e. 12.5 and 25% of maximum recommended rate), autumn application of Roundup CT� was more effective than spring application, especially once regeneration was 2 years old. Gramoxone� was also most effective at all rates above 12.5% of the maximum when applied in the autumn to two-year-old coppice. However, Garlon� (600 g/L triclopyr) and Tordon 50-DA(r) (50 g/L picloram + 200 g/L 2,4-D) were more effective when applied to 1-year-old coppice in the spring. Overall, the most effective low-concentration treatment was Roundup CT� applied in the autumn to two-year-old coppice. Low-concentration treatment of one-year-old coppice with Roundup CT� and Arsenal� was also consistently more effective when carried out in the autumn (80-90% leaf effect). The probability of shrub mortality was inversely related to coppice biomass with smaller coppices clearly more vulnerable to the added pressure imposed by secondary chemical treatment, independent of application rate.