Vegetation changes after fire on two land systems in arid North-west New South Wales.

Abstract

Following a sequence of favourable years in which pasture growth over much of the arid zone of Australia reached very high ievels, controlled burns were carried out on two contrasting vegetation types in the extreme north-west of New South Wales. A wheei-point apparatus was used to measure subse- quent changes in botanical composition and foliage cover over a four year period. On a pasture periodically dominated by Mitchell grass (Astrebla spp.) burning while growing conditions were favourable resulted in only a small long- term decrease in the cover of Mitchell grass. In the short-term all chenopod species were eliminated and a wider range and greater abundance of annual forbs were promoted in the following spring. On a similar area burned by wildfire in a year of low summer rainfall the response from Mitchell grass was much poorer and botanical composition of the pasture present in the following spring differed from that which developed in the spring following the controlled burn. It also differed from that of the unburnt pasture. The major differences were due to the response of forb species and are attributed to variation in seasonal rainfall. On a dune-system pasture the dominant grasses were species of Aristida and Enneapogon. These are relatively short-lived and appear to have little ability to regrow from the butt after fire. Their slow regeneration after the burn was reflected in the substantial increase in relative abundance of perennial forbs in the following autumn, and of annual forbs the next spring. Although fire appeared to have no long-term effect on the pasture it dramatically reduced tree and shrub numbers. It is suggested that during years in which abnormal quantities of Mitchell grass are present in this region, controlled burning could be a useful form of management. A mosaic of patches burnt at different times would reduce the potential for wide-scale wildfires, provide refuge areas for stock and wildlife in the event of wildfire, and promote a wider choice of plant material for grazing animals. However, in dune-systems vegetation, removal of the pasture cover and reduction of the tree and shrub density would constitute an erosion risk.