Distinguishing effects of natural disturbances such as droughts on natural vegetation from potential industrial influence in the Port Curtis area using regional reference sites and the differential response of closed and open canopy vegetation

The advent of a shale oil processing industry in Port Curtis region (approximately 15 Ian northwest of Gladstone) in addition to existing industries provided the impetus to establishment of 15 'Sentinel' monitoring sites to assess potential effects of industrially-sourced gaseous outfalls from the Gladstone airshed on natural vegetation. To provide a baseline for widespread regional natural disturbances a further 5 sites ('Reference') were established some 40 - 50 Ian southwest of Gladstone in the Kroombit Tops area. In addition to using passive gas gels to monitor gaseous outfalls of S02, N02, 03, bioindicators monitored included foliar health indices (leaves per I m of branch, average leaf weight and a chlorosis - necrosis score), soil and litter associated invertebrate assemblages and lichen cover. Gaseous outfalls of S02 and N02 at Sentinel sites were 2 to 15 times greater than concentrations at Reference sites but 0 3 was similar. Significant changes in average leaf weight, mostly with a declining trend as typical of effects of gaseous outfalls on vegetation and an increase in variability of invertebrate assemblages in 2001 were found. Significant changes in numbers of leaves per branch (no consistent increasing or decreasing trend) and the chlorosis - necrosis score were also found. The latter bioindicator actually declined and this is the reverse of the predicted response to effects of gaseous outfalls on vegetation and suggests an improvement in vegetation health. There was no clear linkage between these patterns in change in leaf weight and gaseous outfalls within the Sentinel area. Two lines of evidence point to the significant declines in average leaf weight and increases in vertebrate assemblage variability within Sentinel sites being a response to a widespread regional drought in 2001 and 2002. Firstly, similar patterns of declining leaf weight were observed in the regional Reference area relatively remote from Port Curtis where levels of S02 and N02 concentration are substantially lower. Secondly, response of two types of vegetation ('closed', > 70% canopy cover typical of dry rainforest and 'open', 30 - 70% canopy cover typical of sclerophyll woodlands and forests) differed with closed canopy vegetation consistently showing significant declines in average leaf weight irrespective of location whereas most open canopy vegetation sites were non-significant except those associated with ridge-tops in the Sentinel area. Within the Sentinel area only closed vegetation showed an increase in invertebrate assemblage variability in 2001. Dry rainforest vegetation is known to suffer seasonal water stress in the late dry season and a greater vulnerability to drought effects could explain the patterns of change in the bioindicators monitored. The usefulness of dry rainforest vegetation as a way of internally calibrating for drought effects by acting as a sentinel vegetation type for natural disturbance of this kind needs to be more widely tested. If this ability is confirmed for this vegetation type, then it may offer a more cost-effective protocol for testing for natural drought effects compared with regional reference areas, or they may provide an opportunity to have some ability to calibrate for regional effects of droughts where regional reference sites are lacking.