concentrations were measured at Glen Bowen with a median TN of ~ 46,000 ug / L which was considerably higher in 2020 / 21 compared to the previous year for this site . Median TN values at other sites were ~ 11,000 ug / L at Mt Wickham control , ~ 6,100 ug / L at Strathbogie control , ~ 2,600 ug / L at Havilah control and 1,600 ug / L at Mt Pleasant control .
● There are differences in the sediment yields ( t / ha / y ) estimated using different measurement approaches ( loads monitoring vs terrain analysis ). In future years , a more detailed comparison of terrain versus water quality methods should be investigated to better quantify the estimates of sediment savings from these sites . This would involve doing a dedicated comparison of terrain and water quality data collected at coincident locations across a range of site .
● When sites have water quality and runoff data collected from periods before and after treatment , from both control and treatment sites , statistical analysis can be used to calculate the effectiveness of the rehabilitation on erosion control . This approach takes into account variations in response due to inter-annual rainfall variability between wet seasons .
● If water quality samples have been adequately collected across the flow event hydrograph for the main events in all seasons at both control and treatment sites ( including first flush and peak events ), then it is suitable to use concentration data to evaluate rehabilitation effectiveness . Concentration data on its own is not a substitute for loads .
● Comparison of changes in the sediment loads between sites is the most appropriate metric to use to evaluate treatment effectiveness as this represents total tonnes of fine sediment delivered or saved due to the rehabilitation . It is acknowledged that runoff and load data are notoriously challenging to obtain and highly uncertain in these systems prone to flash flooding , however , loads are considered important for ‘ gold ’ standard monitoring .
● When using control and treatment data without coincident ‘ before ’ data at each site or using ‘ before ’ and ‘ after ’ data at a treatment site without a control site to compare to , it is possible to detect changes , but it is difficult to attribute these to the treatment effects alone . The changes could be influenced by variations in rainfall , runoff , or individual gully dynamics . For sites without ‘ before ’ data , indicative trends can be identified but longer time frames ( e . g ., > 5 years ) will likely be required to detect statistically significant changes that are independent of rainfall / runoff variability .
● Most sites have shown some improvement in vegetation cover (%) or biomass following treatment ; however , gully catchment land condition remained in C or D class for most sites with long term indicators of landscape condition remaining stable or only showing minor improvement . This is because the pasture species are primarily dominated by the exotic perennial grass Indian couch , forbs , and / or exotic legumes . Indian couch is shallow rooted , produces less biomass and is more sensitive to seasonal climate variations than native perennials , resulting in low grass basal area , small root mass and relatively poor infiltration and soil quality .
● Grazing management during and following rehabilitation needs to be carefully monitored to ensure treatments are not compromised . In most cases it is important to give treated areas a spell from grazing to allow vegetation to recover from the initial disturbance associated with rehabilitation . Grazing can then be reintroduced using an adaptive management approach .
● The treatment at Strathbogie focused on hillslope runoff diversion banks which aim to effectively ‘ starve ’ the active gully . In 2019-20 , this intervention statistically improved the runoff and water quality metrics and the site was given an effectiveness value of ~ 0.95 , however , this was a relatively dry year . In 2020-21 , both the treatment and control gully experienced some runoff and the water
78