Effects of suspended solids on benthic macroinvertebrate fauna downstream of a road crossing, Jim Jim Creek, Kakadu National Park
Internal Report 256
Supervising Scientist Division
About the report
Tourist vehicle traffic using a seasonally-accessible, unformed road crossing on the upper reaches of Jim Jim Creek in Kakadu National Park, has caused a seasonal elevation in turbidity (as a result of suspended sediment) downstream of the road crossing in recent dry seasons. A modified BACIP experimental design was used to investigate the effects of this disturbance on the benthic macroinvertebrate fauna (at family-level) downstream of the road crossing. Paired sites in both Jim Jim Creek (upstream and downstream of the road crossing) and Twin Falls Creek (a control stream, with analogous but undisturbed upstream and downstream sites) were sampled for a period before and after the seasonal opening of the road crossing to public access. Turbidity levels peaked one month after the road opened, reaching an average maximum of 60 NTU (~100 mg/L suspended solids) 200 m downstream of the crossing and 30 NTU (~17 mg/L suspended solids) 1000 m downstream of the crossing. Although temporal autocorrelation of the BACIP data prevented statistical analysis by conventional inferential procedures, a technique of modelling the temporal trend in the data by introducing a covariate, stream discharge, was used. This enabled inferences to be drawn about impact-related effects in Jim Jim Creek after the onset of the disturbance.
Turbidity-related effects were evident in macroinvertebrate communities of rootmat samples collected 200 m downstream of the road crossing, whilst 1000 m downstream of the crossing, there was evidence of only some slight changes to communities. The changes observed were most evident in multivariate analyses, viz paired site dissimilarity measures and ordination. However, associated with these observed changes was a noticeable reduction in total macroinvertebrate abundance, attributed mostly to a decline in the abundance of Chironomidae at downstream Jim Jim Creek sites compared to control sites. Although the effects were limited in extent, they are cause for concern given the exceptional conservation value of the study area.
The modified approach to BACIP analysis, modelling a covariate of natural temporal change in dissimilarity data, was shown to be a valuable, cost-effective means of detecting impacts in a situation where the period of time available for monitoring was necessarily short.