Quantifying the effects of land management interventions on water quality in the Coal River catchment

This study examined spatial variation in water quality and its relationship to riparian land management in the Coal River Valley, SE Tasmania. Historical water quality data from stations at Baden, downstream of the Craigbourne Dam, Richmond and White Kangaroo Rivulet collected between 1999 and 2008 were obtained from DPIPWE. The water quality variables selected for study were water temperature, electrical conductivity (EC), dissolved oxygen (DO), water pH, turbidity, nitrate, total nitrogen (TN), dissolved reactive phosphorus (DRP), total phosphorus (TP) and stream flow. Riparian land use within one kilometre of the river was assessed and digitised using 2005/7 colour aerial photographs and water quality data for that period were examined and possible linkages investigated. The historical data demonstrates complex spatial patterns of in-stream water quality parameters in the Coal River Valley. The Craigbourne reservoir and apparent differences in geology between subcatchments explained some of the differences observed in water quality parameters better than land use. There was no significant variation in water temperature but higher values were recorded at Richmond. Similarly, higher EC was observed at the base of the catchment (Richmond) than at the top of the catchment (Baden). The influence of the Craigbourne Dam was reflected in higher pH, DO, DRP, TP and water flow at the station downstream of this large water reservoir in the middle of the catchment. In the White Kangaroo tributary higher nitrate and TN were observed but at Richmond, below its confluence with the Coal River, lower TP and TN were recorded. There was a significant negative correlation between DO and water temperature observed in the Coal River. However, positive correlations were found between stream flow and rainfall with turbidity at all stations except downstream of the reservoir. Stream nitrogen and phosphorus showed a significant relationship with rainfall at Richmond. Positive correlations of turbidity with nitrate, TN, DRP and TP show nutrients bound to sediment are a likely source of many nutrients in the river. Consequently, riparian vegetation could play a vital role in reducing sediment load and nutrient concentration in the river system. Subcatchment riparian land use and water quality data from 2005/7 suggests that lower turbidity at Richmond (2.82 NTU) compared to White Kangaroo Rivulet (4.25 NTU) and Baden (4.70 NTU) may be due to the impact of higher percentages of willow trees in the Richmond subcatchment (1.92 %) followed by White Kangaroo subcatchment (0.26%) and Baden (0%). Riparian land management works such as planting native vegetation and fencing (approx 4 km) on the river banks could have reduced the sediment load and nutrient in the river by preventing erosion caused by stock access to river water. This is supported by 2005/7 nutrient data at Richmond where TP (0.015 mg/L) and TN (0.625 mg/L) were observed as compared to Baden (TP = 0.020 mg/L, TN = 0.69 mg/L) and Downstream Craigbourne Dam (TP = 0.022 mg/L, TN = 0.71 mg/L). However this lower part of the river still had the highest amount of willows in the riparian strip, despite willow removal programmes, and so the results are thus confounded. A higher percentage of native pastures in the riparian strip were found to be associated with lower turbidity in the river. Finally, forest cover was found to be positively correlated with nitrate nitrogen which is likely to be caused by nitrogen fixing acacia dominated forests