Much of the land disturbed by open-cut coal mining in the Bowen Basin, Central Queensland, has been rehabilitated using improved pasture species. It has been assumed in the past that these areas will be capable of sustaining regular grazing pressure and that ultimately, these areas will be returned to commercial grazing. However, little research has been done on the productivity of these pastures and what, if any, should be considered a sustainable stocking rate. The aims of this project were: to investigate the productivity of pastures established on rehabilitated open-cut coal mines in the Bowen Basin; and, to recommend long-term sustainable stocking rates. This work continues on from the study of Bisrat (2002).
There were 18 primary productivity sites established on buffel grass (Cenchrus ciliaris) dominant pastures on three coal mines in the Bowen Basin as a part of a previous project. Eleven of these sites were monitored over the 2001-2002 growing season. This number was expanded to 51 sites monitored during the 2002-2003 growing season, with four of these sites placed on unmined areas surrounding the mines. These expanded sites were located selectively in an attempt to maximise the variation between sites in terms of certain variables crucial to pasture growth on rehabilitated open-cut coal mines identified by previous workers (slope, nitrogen, salinity, Exchangeable Sodium Percentage (ESP) and Cation Exchange Capacity (CEC) (Bisrat 2002)). Pasture parameters measured within each site included: total ground cover; pasture basal area; and, pasture dry matter yield. Estimates of the Rainfall Use Efficiency (RUE) values for each of these sites were made using the peak pasture yield and cumulative rainfall over the buffel grass growing season. For the 2001-2002 growing season, total cover ranged from 57-83%, pasture dry matter yield ranged from 1.6-4.3 t/ha and RUE values ranged from 5.8-15.6 kg/ha/mm. For the 2002-2003 growing season at the sites established on rehabilitated coal mines, total cover ranged from 22-78%, pasture basal area ranged from 0.8-9%, pasture dry matter yield ranged from 0.3-7.5 t/ha and RUE values ranged from 1.4-26.1 kg/ha/mm. At the sites on unmined land surrounding the mines over the 2002-2003 growing season, pasture dry matter yield ranged from 0.8-3.5 t/ha and RUE values ranged from 2.6-16.1 kg/ha/mm.
An investigation of the distribution of dry matter yield data showed that it was skewed and would need to be log transformed prior to statistical analysis. To further improve the likelihood of .successful predictions of yield, a covariance approach was adopted using pasture basal area as a surrogate for the level of success of pasture establishment at each location. An analysis of covariance approach was then used to determine which site variables had the greatest effect on the log transformed pasture dry matter yield. The variables included in this analysis were based on previous experience and their individual relationships with the log transformed dry matter yield variable. Once this analysis was performed, a model was produced using pasture basal area as the covariate and included total rainfall (mm), slope (%), surface roughness ratio, total soil nitrogen (%), ESP, Electrical Conductivity (EC) and clay (%>) as explanatory variables. This model predicted 74%> of the variation in log dry matter yield and 62%> of the variation in the untransformed dry matter yield data.
An intensive, rotational grazing trial previously established at the Goonyella Riverside mine in Central Queensland was used in an attempt to determine the effect of a number of long-term stocking pressures on these pasture systems. This trial was grazed once over the 2001-2002 growing season and once over the 2002-2003 growing season. Pasture assessments were performed before and after each grazing cycle. The stocking pressures imposed were designed to mimic pasture utilisation rates ranging from 9-85% of end of season growth. During the 2001-2002 grazing cycle, the actual seasonal pasture utilisation rates achieved varied from 10-98% and during the 2002-2003 grazing cycle, the actual pasture utilisation rates varied from 12-114%.
The method used for calculating long-term sustainable stocking rates for this trial was based on the assertion of McKeon et al.(1990) that a pasture utilisation level of 30% is achievable in 80% of years on improved pastures in Central Queensland. Using the RAIN MAN computer program, the amount of rainfall in 80% of years can be estimated from historical records. Applying this information, a pasture growth figure in 80% of years was estimated, and a long-term sustainable stocking rate was suggested, based on 30% utilisation of that growth. Using this method, a long-term sustainable stocking rate of approximately 5 ha/hd is suggested for this site.
Two extensive grazing trials, previously established at the Blackwater and Norwich Park mines to observe the effects of high, medium and low stocking rates were redesigned. The Norwich Park trial was redesigned to increase the grazing pressure on the pasture, whilst the Blackwater trial was abandoned for reasons beyond our control. The separate stocking rates were essentially abandoned at the Blackwater trial during April 2002, leaving a single stocking rate of 2.9 ha/hd over the entire site. The Norwich Park trial was continued with the three set stocking rates increased from 3.4, 4.6 and 7.5 ha/hd to 2.4, 3.8 and 5.2 ha/hd during August 2002, resulting in average stocking rates over the monitoring period of 3.2, 4.4 and 7 ha/hd.
The same method of estimating 30% pasture utilisation in 80% of years was applied to the data from the Norwich Park trial. Using this method, a stocking rate of approximately 3.1 ha/hd was suggested as being sustainable in the long term. There were no signs of degradation in any of the paddocks monitored at the Norwich Park trial, confirming that the sustainable stocking rate for this site is higher than the highest average stocking rate imposed over the period of 3.2 ha/hd. The stocking pressure of 2.9 ha/hd at the Blackwater grazing trial was shown to be too high, as it resulted in a loss of dry matter yield and ground cover.
It was concluded that pastures established on rehabilitated open-cut coal mines have the potential to be just as productive as unmined pastures. However, because of the variability in pasture productivity shown it is recommended that not all areas rehabilitated to improved pasture species be set aside for grazing, due to their fragility. The stocking rates recommended for each site monitored is slightly conservative when compared to the 3 ha/hd recommended for improved buffel grass pastures in the region, but this conservative approach is warranted given the fragility of these sites and their propensity to erode.