The flow of precipitation from the surface through to groundwater in karst systems is a complex process involving storage in the unsaturated zone and diffuse and preferential recharge pathways. The processes associated with this behaviour are not well understood, despite the prevalence of karst aquifers being used as freshwater supplies. As a result, uncertainty regarding the ecohydrological processes in this geological setting remains large. In response to the need to better understand the impact of woody vegetation on groundwater recharge, annual evapotranspiration (ET) rates and tree water sources were measured for two years above a shallow, fresh karst aquifer. Water use strategies of the co‐occurring Eucalyptus diversifolia subsp. diversifolia Bonpl. and Allocasuarina verticillata (Lam.) L. Johnson were investigated using a monthly water balance approach, in conjunction with measurement of the stable isotopes of water, leaf water potentials and soil matric potentials. The results suggest that it is unlikely groundwater resources are required to sustain tree transpiration, despite its shallow proximity to the soil surface, and that similarities exist between ET losses and the estimated long‐term average rainfall for this area. Irrespective of stand and morphological differences, E. diversifolia and A. verticillata ET rates showed remarkable convergence, demonstrating the ability of these co‐occurring species to maximise their use of the available precipitation, which avoids the requirement to differentiate between these species when estimating ET at a landscape scale. We conclude that the water holding capacity of porous geological substrates, such as those associated with karst systems, will play an important role in equilibrating annual rainfall variability and should be considered when assessing ecohydrological links associated with karst systems. Copyright © 2013 John Wiley & Sons, Ltd.

No abstract

The partitioning of precipitation into interception, stemflow and throughfall is an important hydrological process in forested systems, influenced heavily by climate and plant form. This study examined whether the rainfall partitioning pathways reflect the often cited influence of tree morphology, using two species in a semi‐arid karst environment. Eucalyptus diversifolia ssp. diversifolia has a multi‐stemmed habit, smooth bark and true leaves. In comparison, Allocasuarina verticillata has a single trunk, rough bark and long, thin, vertical phyllodes. We hypothesized that multiple stems and a smooth bark would be more effective at generating stemflow compared with single stems with a rough bark surface. To test this, rainfall, throughfall and stemflow were collected over two years, and stemflow funnelling ratios were calculated. The degree of similarity in overall rainfall partitioning for the two species was remarkable; although some divergence was found each month, the partitioning regressions converged. For E. diversifolia, gross rainfall partitioned into interception, throughfall and stemflow averaged 30.9%, 66.4% and 2.7%, respectively. For A. verticillata, rainfall partitioning of gross precipitation into interception, throughfall and stemflow averaged 31.4%, 65.9% and 2.7%, respectively. Maximum stemflow funnelling ratio for E. diversifolia was 74 and for A. verticillata was 147, indicating that water from stemflow is likely to play an important ecohydrological role in this environment. We further compared these findings to 31 global studies and discussed the importance of scale (individual tree vs plot) and canopy cover when reporting or interpreting rainfall partitioning results. Copyright © 2013 John Wiley & Sons, Ltd.

A groundwater risk assessment was carried out for 30 potable water supply systems under a framework of protecting drinking water quality across South Australia. A semi-quantitative Groundwater Risk Assessment Model (GRAM) was developed based on a "multi-barrier" approach using likelihood of release, contaminant pathway and consequence equation. Groundwater vulnerability and well integrity have been incorporated to the pathway component of the risk equation. The land use of the study basins varies from protected water reserves to heavily stocked grazing lands. Based on the risk assessment, 15 systems were considered as low risk, four as medium and 11 systems as at high risk. The GRAM risk levels were comparable with indicator bacteria-total coliform-detection. Most high risk systems were the result of poor well construction and casing corrosion rather than the land use. We carried out risk management actions, including changes to well designs and well operational practices, design to increase time of residence and setting the production zone below identified low permeable zones to provide additional barriers to contaminants. The highlight of the risk management element is the well integrity testing using down hole geophysical methods and camera views of the casing condition.