Conclusion

This study shows that no individual suspended sediment monitoring method is suitable for providing the necessary data to understand sediment transport dynamics in a flowing gully. Rather, the application of a combination of complimentary methods is necessary to reliably provide representative data from these challenging aquatic environments. For example, autosamplers provide multiple samples over a flow event but fail to sample coarser particles accurately; the calibration of samples collected by autosamplers with the time-integrated data from a PASS sampler, which does sample coarser particles relatively accurately, would result in a more reliable dataset than the use of either method alone. Other configurations could be used to improve the spatial scale of monitoring effort, for example, low cost methods (i.e., PASS and RS samplers) could be deployed at various locations throughout a network of gullies, whereas multiple methods (i.e., PASS and RS samplers, turbidity logger, and an autosampler) would be deployed at the gully network outlet.
The modified PASS sampler performed well in both laboratory and field trials. The modification of the PASS sampler to operate in gullies is a good example of how existing techniques can be customised to operate in the harsh environments typical of gully systems. We aim to further modify the PASS sampler by interfacing a flow meter and pump controller so that its sample rate can be matched to stream velocity, thus allowing the collection of a flow proportional (i.e., isokinetic) sample. Further comparisons using the PASS sampler and other methods in different gullies with varied suspended sediment dynamics are required to confirm its validity as an automated sampling method for gully systems.