Headwater catchments are important sources of water in many rivers. While headwater catchments are commonly developed on indurated rocks without extensive groundwater systems, the observation that many headwater streams are perennial implies that they are sustained by water in fractures, soils, or the regolith. Understanding the sources and transit times of water in headwater streams is important for understanding catchment functioning and predicting the impacts of changing climate or land use. This study uses major ion geochemistry and tritium (3H) to determine water sources and transit times in first-order streams in the Otway Ranges, southeast Australia.
Comparison of the geochemistry of soil water, water from soil pipes (macropores), and riparian groundwater indicates that macropore flow is the major contributor to streamflow. The streams are gaining and the lack of riparian zone groundwater inputs may be due to the presence of low hydraulic-conductivity organic-rich streambed sediments or compartmentalisation of shallow groundwater by clays in the weathered rocks. Similarly, much of the soil water exists in isolated pockets of isolated water that are not connected to the soil pipes. The stream water has tritium (3H) activities of 1.80 to 2.06 TU. These are significantly lower than the 3H activities of modern rainfall (2.6 to 3.0 TU), even during the higher winter flows. The water from the soil pipes has 3H activities of 1.80 to 2.25 TU, the riparian zone groundwater has 3H activities of 1.35 to 2.39 TU, and one sample of soil water has a 3H activity of 2.22 TU. Mean transit times calculated using a range of lumped parameter models are between 3 and 57 years. Relatively long mean transit times are consistent with the major ion geochemistry that implies that waters are resident for sufficient time for weathering reactions and evapotranspiration to have occurred.
While the discharge from the soil pipes increases following periods of high rainfall, the long mean transit times implies that this water is stored for several years within the regolith before discharge, with storage volumes estimated as >108 m3. Thus the increase in streamflow is not the simple transmission of recent rainfall through the macropores but mobilisation of existing catchment stores. The streams will be buffered against year-on-year variations in rainfall but are vulnerable to longer-term variations in rainfall or land use. Management of these catchments needs to consider the impacts on the macropores and the delayed responses caused by the large storage volumes.