Land and Water · 2005-10-14 · Land and Water CSIRO Land and Water Technical Report No 18/04...

Land and Water

CSIRO Land and Water Technical Report No 18/04

October 2004

Interpreting Airborne Geophysics as an adjunct to

Hydrogeological Investigations for

Salinity Management:

Honeysuckle Creek Catchment, Victoria

Pauline English, Peter Richardson, Mark Glover,

Hamish Cresswell & John Gallant

CSIRO Land & Water, Canberra

Part 1 of 2

Interpreting Airborne Geophysics as an adjunct to Hydrogeological Investigations for Salinity Management:

Honeysuckle Creek Catchment, Victoria

Pauline English, Peter Richardson, Mark Glover, Hamish Cresswell & John Gallant

CSIRO Land and Water October 2004

FOUNDATION FOR RURAL AND REGIONAL RENEWAL

ii

ISSN 1446-6171

© 2004 CSIRO To the extent permitted by law, all rights are reserved and no part of this publication covered by copyright may be reproduced or copied in any form or by any means except with the written permission of CSIRO Land and Water.

Important Disclaimer: CSIRO Land and Water advises that the information contained in this publication comprises general statements based on scientific research. The reader is advised and needs to be aware that such information may be incomplete or unable to be used in any specific situation. No reliance or actions must therefore be made on that information without seeking prior expert professional, scientific and technical advice. To the extent permitted by law, CSIRO Land and Water (including its employees and consultants) excludes all liability to any person for any consequences, including but not limited to all losses, damages, costs, expenses and any other compensation, arising directly or indirectly from using this publication (in part or in whole) and any information or material contained in it.


iii

Acknowledgements We wish to thank the Foundation for Rural and Regional Renewal (FRRR) for their participation in and financial commitment for this work. This research stems from the Heartlands Initiative, a collaborative project between the Murray Darling Basin Commission (MDBC) and CSIRO aimed at integrating research and on-ground action to foster sustainable land use in the Murray-Darling Basin. The MDBC is acknowledged for their foresight and co-partnership for our work in the Honeysuckle Catchment in Victoria in recent years. Our appreciation is extended to the landholders in Honeysuckle Catchment and adjoining areas for their support, access to their properties, shared wisdom and enthusiasm for our efforts in the region. We particularly value the farmers and local residents who stopped by the roadside to exchange information and good cheer with us during our weeks of drilling in the area. We thank the townsfolk of Benalla, Violet Town and other communities who provided support services during our field program. Barry Oswald and Phil Stevenson, Goulburn Broken Catchment Management Authority (GBCMA), greatly supported this project. The Mid Goulburn Broken Implementation Committee is thanked for its support and for providing a forum for reporting on our research in the region. We appreciate the high degree of cooperation with staff of the Victorian Departments of Primary Industry (DPI) and Sustainable Environments (DSE), including Xiang Cheng, Mark Reid, David Heislers, Maree Platt, Phil Cook, Mark Cotter, Anthony Christensen, Alan Willocks and Greg Robertson. Phil Dyson has provided constructive criticism and input to parts of this report. We give thanks to our CSIRO colleagues who contributed many hours and their specialist skills to this project: Lu Zhang, Linda Gregory, Jen Austin, Fred Leaney, Megan Lefournour, Ruth Palmer and Heinz Buettikofer. Finally, particular thanks to Vanessa Wong, the Australian National University (ANU), who volunteered to come out drilling with us and greatly assisted our fieldwork. Lastly, we thank Xiang Cheng and Richard Cresswell for kindly reviewing this report, for their interest in the work and for providing constructive criticism.


iv


v

Table of Contents

Acknowledgements...................................................................................................... iii Table of Contents.......................................................................................................... v List of Figures & Tables ............................................................................................. vii Executive Summary..................................................................................................... ix 1. Introduction............................................................................................................... 1

1.1 Background......................................................................................................... 1 1.2 Sub-catchment prioritisation............................................................................... 2 1.3 Objectives ........................................................................................................... 3 1.4 Study area ........................................................................................................... 4 1.5 Airborne Geophysics .......................................................................................... 8 1.6 Previous work ..................................................................................................... 9 1.7 Project scope ....................................................................................................... 9

2. Research Approach ................................................................................................. 11 2.1. Overall approach.............................................................................................. 11 2.2. Groundwater and Salinity ................................................................................ 11

2.2.1. Airborne Geophysics ................................................................................ 11 Magnetic and Gamma-ray spectrometry (‘radiometric’) data ........................ 12 Airborne Electromagnetic (AEM) data........................................................... 12

2.2.2. Field investigations ................................................................................... 15 Drilling and piezometer installation ............................................................... 15 EM-34 Traverses and individual site measurements ...................................... 17 EM-39 Profiling.............................................................................................. 17 Salinity Measurements.................................................................................... 17

Groundwater and surface water salinity measurements ............................. 17 Soil salinity measurements ......................................................................... 18

Magnetic Susceptibility .................................................................................. 18 Contouring ...................................................................................................... 18 Groundwater dating ........................................................................................ 20

Radiocarbon (14C) analysis ......................................................................... 20 Oxygen-16, Deuterium and Tritium data.................................................... 22

2.3. Surface Water Yield Modelling....................................................................... 22 2.4. Land Resource Assessment ............................................................................. 24

Multiple-resolution Valley Bottom Flatness (MrVBF) ...................................... 24 Topographic Wetness Index (TWI) .................................................................... 25 Predictions of Soil Depth.................................................................................... 25 Available Water Capacity................................................................................... 26

3. Geologic and Hydrogeologic Framework............................................................... 27 3.1. Geology............................................................................................................ 27 3.2. Geomorphology ............................................................................................... 28 3.3. Groundwater Flow Systems............................................................................. 28 3.4. Soils ................................................................................................................. 31

4. Results and Interpretation ....................................................................................... 35 4.1 Airborne Magnetic and Gamma-ray Spectrometry Interpretation.................... 35

4.1.1. Magnetic Data........................................................................................... 35 4.1.2. Radiometrics ............................................................................................. 41


vi

4.2 Airborne Electromagnetic (AEM) data & target area selection........................ 43 4.3 Miepoll.............................................................................................................. 47

4.3.1. Airborne Geophysics and Drill-holes ....................................................... 47 4.3.2. Soil map .................................................................................................... 52 4.3.3. MrVBF...................................................................................................... 53

4.4 Baddaginnie ...................................................................................................... 53 4.4.1. Airborne Geophysics and Drill-holes ....................................................... 53 4.4.2. Soil map .................................................................................................... 58 4.4.3. MrVBF...................................................................................................... 59

4.5 Sheep Pen Creek ............................................................................................... 60 4.5.1 Airborne Geophysics, Drill-holes and Soils .............................................. 60 4.5.2. MrVBF...................................................................................................... 64

4.6 Groundwater and Salinity ................................................................................. 67 4.6.1. Depth to Groundwater, Elevation of the Watertable and Groundwater Salinity ................................................................................................................ 67 4.6.2. Groundwater ages ..................................................................................... 74

4.7 Salt sources and stores, Groundwater recharge and discharge ......................... 79 4.7.1. Salt sources and stores .............................................................................. 79 4.7.2. Recharge and discharge ............................................................................ 81

4.8 Surface Water Yield.......................................................................................... 84 4.9 Land Resource Assessment............................................................................... 86

5. Summary & Recommendations .............................................................................. 91 5.1 Where are salt stores located, and how much salt is present? .......................... 91 5.2 Are all salt stores accounted for in the AEM imagery? .................................... 92 5.3 Are these salt stores being mobilised to lower landscape positions and to the catchment waterways? ............................................................................................ 94 5.4 What intervention measures are appropriate?................................................... 96

5.4.1. Strathbogie Ranges and adjacent piedmont plain ..................................... 96 5.4.2. Caniambo Hills ......................................................................................... 98 5.4.3. Riverine Plain............................................................................................ 99

5.5 Utility of the Airborne Geophysics datasets ................................................... 102 5.5.1. Airborne magnetic data........................................................................... 102 5.5.2. Airborne radiometric data ....................................................................... 103 5.5.3. Airborne Electromagnetic data ............................................................... 104

References................................................................................................................. 111 APPENDICES .............................................................................................................. 1 APPENDIX 1................................................................................................................ 3 APPENDIX 2................................................................................................................ 7 APPENDIX 3.............................................................................................................. 11 APPENDIX 4.............................................................................................................. 15 APPENDIX 5.............................................................................................................. 35


vii

List of Figures & Tables Figure 1. Map of the sub-catchments and the airborne geophysics survey area Figure 2. Digital Elevation Model of the study area Figure 3. Rainfall map for the Goulburn-Broken catchment Figure 4. Cleared Strathbogie Ranges and waterlogged piedmont plains Figure 5. Salt scalds in the Warrenbayne area Figure 6. Salt efflorescence above the exposed watertable in a gully at Boho Figure 7. Downes Swamp, Sheep Pen Creek lowlands - Broken River floodplain Figure 8. AEM CDI 0-5 m images: (a) uncalibrated and (b) calibrated Figure 9. FRRR drill-hole locations Figure 10. Drilling FRRR09 through a resistive AEM area at Faithful Creek at the

western edge of the Violet Town Plain Figure 11. Down-hole electrical conductivity profiling using the EM-39 device Figure 12. Measuring creekwater EC, Sheep Pen Creek Figure 13. Radiocarbon (14C) decay curve Figure 14. Sampling groundwater into barrels for 14C analysis Figure 15. Mean annual evapotranspiration and rainfall graph Figure 16. Relationship between mean annual water yield and rainfall Figure 17. Groundwater Flow Systems (GFS) of the Honeysuckle Creek study area Figure 18. Soil map of the study area Figure 19. Tunnel erosion in Sodosols Figure 20. Grey Vertosols from a low-lying break-of slope area Figure 21. Total Magnetic Intensity (TMI) image Figure 22. Iron-oxide coated gravels Figure 23. Outcrop of Silurian shale with differentiated iron oxide units Figure 24. Airborne gamma-ray image Figure 25. Elevation of the ‘base of conductor’ beneath Violet Town Plain Figure 26. Hydrographs for Bores 142 and 144 Figure 27. Miepoll AEM 0-5 m CDI Figure 28. Miepoll soil map Figure 29. Miepoll MrVBF plot Figure 30. Baddaginnie AEM 10-15 m CDI Figure 31. Waterlogged area at Stoney Creek, Baddaginnie Figure 32. White clay intersected 0-20 m in FRRR11 at Baddaginnie Figure 33. Jubilee Swamp Figure 34. Baddaginnie soil map Figure 35. Baddaginnie MrVBF plot Figure 36. MrVBF plot for the Warrenbayne-Boho area, Strathbogie Ranges Figure 37. Sheep Pen Creek AEM: 5-10 and 15-20 m CDIs Figure 38. Sheep Pen Creek soil map Figure 39. Sheep Pen Creek MrVBF plot Figure 40. Bore location map – all reliable bores used for groundwater contours Figure 41. Depth to groundwater: (a) contours; (b) projection Figure 42. Elevation of watertable: (a) contours and flow directions; (b) projection;

(c) interpreted regional influence on groundwater flow Figure 43. Groundwater EC contours Figure 44. Distribution of groundwater 14C data: pmc and apparent ages Figure 45. Groundwater age data contours: (a) pmc; (b) apparent age; (c) EC; (d)

δ13C; (e) depth to groundwater


viii

Figure 46. Cross-section illustrating the main groundwater processes Figure 47. Surface water yield map Figure 48. (a) Soil thickness plot; (b) Regolith depth plot derived from the AEM

‘elevation of the base of the conductor’ image Figure 49. Predicted soil water holding capacity plot for the A and B horizons Figure 50. Predicted soil water holding capacity plot to 2 m depth Figure 51. Suggestions for salinity management Figure A4.1. Location of EM-34 traverses and sites Figure A4.2. EM-34 measurements and corresponding AEM measurements Figure A5.1. BRS and FRRR drill-hole locations Table 1. Soil salinity measurements and salt store estimates Table 2. Hydraulic conductivity Table 3. Bore site data: groundwater carbon isotope analysis Table 4. Groundwater 14C data Table A5.1. BRS and CSIRO drill-hole coordinates


ix

Executive Summary This report presents an interpretation of recently-flown airborne geophysical datasets for Honeysuckle Creek catchment in north central Victoria. High-resolution airborne magnetic, gamma-ray spectrometry (radiometric) and airborne electromagnetic (AEM) data were utilised for the investigation. Prior to the interpretation stage, the AEM data was calibrated against down-hole conductivity measurements and reprocessed to provide imagery of the distribution of the range of bulk conductivities of the substrate in x, y and z-space in the survey area. The airborne geophysical datasets were made available by the Murray-Darling Basin Commission (MDBC) for the study. Acquisition and processing of the data was carried out by the Bureau of Rural Sciences (BRS), Geoscience Australia (GA), and the Geological Survey of Victoria (GSV), part of the Victorian Department of Primary Industry (DPI). This interpretation and integrated research was funded by the Foundation for Rural and Regional Renewal (FRRR) and supported by the Goulburn Broken Catchment Management Authority (GBCMA). The work reported here is directed towards providing an understanding of groundwater and salinity processes in the catchment. This understanding is required for the implementation of targeted tree planting for the long term agricultural viability and environmental sustainability of the area. The airborne geophysical survey area encompasses the edge of the Riverine Plain of the Murray-Darling Basin and fringing uplands. This includes the northern flanks of the Strathbogie Ranges, composed of the Violet Town Volcanics and the Strathbogie Granite, and outliers of ancient weathered fractured Silurian metasedimentary bedrock which makes up the arcuate range of the Caniambo Hills. The Riverine Plain within the study area is composed of Quaternary Shepparton Formation fluvial and lacustrine alluvium. These sediments onlap the gentle northern slopes of the Caniambo Hills and also infill a deep structural trough that underlies the Violet Town Plain which extends broadly between the Strathbogie Ranges and the Caniambo Hills. The distribution of Groundwater Flow Systems (GFS) in the area conforms to these geological and topographic elements. The respective bedrock GFS are local to intermediate in scale and the Riverine Plain here is part of the large regional alluvial GFS. Salinity was first recognised as an issue in the area 70 years ago. Saline outbreaks are common in the break-of-slope zone at the base of the Strathbogie Ranges and in topographically low lying areas where the Riverine Plain abuts the Caniambo Hills. Saline discharges to streams and swamps are also a concern. The research was strongly field-based and utilised the airborne geophysical data to target drill-holes and to carry out down-hole measurements and subsequent laboratory analyses to answer the following key questions:

• Where are salt stores located, and how much salt is present? • Are all salt stores accounted for in the AEM? • Are these salt stores being mobilised to lower landscape positions and to the

catchment waterways? • What intervention measures are appropriate?


x

Complementary datasets, including terrain analysis, water yield modelling and soil maps, were generated or compiled for the study. These have been used to provide a link between the groundwater and salinity component of the research and the development of guidelines for targeted on-ground action. Key findings include the following:

• Intricate networks of drainage channels are well-defined in both the Total Magnetic Intensity (TMI) and AEM Conductivity Depth Images (CDI) datasets. Concordance between magnetic channels and high AEM conductivity relates to x-y space, and less so to z, the depth dimension. Thus, it is not the magnetic gravels (which may make up only a few percent of a profile within the alluvial channels) that are the high conductor. Rather, the high conductivity revealed in the AEM relates to overlying and underlying alluvial clay deposits that encompass the subordinate magnetic gravels within the buried channels.

• The magnetically-defined channels correspond with both contemporary

drainage networks incised into the higher terrain and contiguous down-gradient reaches of the same systems that are now buried beneath onlapping Riverine Plain sediments. The latter buried ‘palaeochannels’ trend northward, contrasting with the present-day drainage in the plain, which flows westward. Clearly, a complex history has been involved in the geologic and palaeoenvironmental evolution of this area where the Murray Darling Basin meets the uplands.

• Contouring of the elevation of the watertable indicates that the groundwater

flow direction beneath the Riverine Plain is westward, not northward as previously proposed. Therefore, the impression conveyed in the northern part of the survey area by the TMI − and, to a lesser extent by the AEM signature for clay-rich alluvium for the same area − dominated by north-trending palaeochannels, does not relate to the hydrogeologic functioning of the system. Very shallow groundwater saturates all substrate of the plain, palaeochannels and palaeo-interfluves alike, regardless of the buried relief that is revealed in the imagery.

• Saturated and unsaturated zones can not be readily discriminated from each

other in the AEM data. Closer to the hill crests conductivity highs are found to correspond with dry substrate and down-gradient they relate to saturated sediments that are immersed in saline groundwater, with no distinguishing difference in actual EM measurements (mS/m) or the nature of the conductivity patterns. Only drill-hole data and watertable measurements can discriminate between saturated and unsaturated zones in the subsurface. This resolution is important to assessing the mobility of observed salt stores. It is necessary to sink as many drill-holes as possible to ascertain the depths, salinities and gradient of the groundwater system, regardless of the availability of airborne geophysics data.


xi

• The calibrated AEM datasets, used in conjunction with bore hydrographs and field observations, afforded retargeting of research within the catchment. The original focus on the Violet Town Plain − which had been designated a priority area on the basis of the original, uncalibrated AEM data − was redirected to adjacent areas where the reprocessed AEM revealed distinctive high conductivity features. The FRRR research consequently focused on the Miepoll and Baddaginnie areas in particular, with some additional work directed towards the break-of-slope areas near the Strathbogie Ranges, and in Sheep Pen Creek sub-catchment.

• The FRRR research was strongly based on hydrogeologic information, both

newly acquired groundwater data collected specifically for the project, and data from >100 existing (DPI and SKM managed) bores in the study area. Groundwater measurements were not reported from the BRS calibration drilling in the survey area (Jones, 2002; Dent 2002), and the initial interpretations of the Honeysuckle Creek airborne geophysics datasets (Dent et al., 2002; Dent, 2002; and Gibson and Wilford, 2002) did not interrogate existing groundwater data.

• Additional to utilising conventional hydrogeological techniques and the

calibrated airborne geophysics datasets, the FRRR research compiled and generated as many datasets as possible to complement the research. These include: Soil and Groundwater Flow Systems (GFS) maps; hydrological modelling (Surface Water Yield); and terrain analysis modelling: the MrVBF index (Multi-resolution Valley Bottom Flatness), Topographic Wetness Index (TWI), predictions of soil depth, and available water capacity (AWC). The hydrologic and soil models are aimed at augmenting decision-making and on-ground action stemming from the interpretation stage.

• Where both EM-39 and EC1:5 profiles have been measured, there is generally

a strong correlation between conductivity and salt (Cresswell, 2002; also Appendix 5 drilling logs, this report).

• Substantial salt stores are present in the unsaturated and saturated zones of the

weathered, fractured metasedimentary rocks of the Caniambo Hills. Groundwaters throughout the fractured bedrock GFS are very saline (typically half seawater salinity). The magnitude of stored salt in the unsaturated zone has been assessed through drilling and laboratory analysis. In the regolith of the Caniambo Hills, high salt stores are present in both clay-rich alluvium/colluvium sourced from eroded shale bedrock infilling nearby subtle valleys (e.g., FRRR02, 03, 08; Appendix 5) and saprolite (e.g., FRRR01, 04, the base of SPC02). Salt stores in fractured bedrock beneath erosionally scoured hill-crests tend to be relatively low (e.g., FRRR07). The extent and severity of groundwater salinity is revealed through construction of a groundwater salinity contour map from bore data which strongly correlates highly saline groundwater with the fractured bedrock GFS. High salt levels in the groundwater originate from high salt stores in the unsaturated or previously unsaturated zone over long periods. The resident salts in the unsaturated zone of the Caniambo Hills are variably represented in the AEM. The pervasive highly saline groundwater in the fractured bedrock GFS is


xii

poorly represented. The latter limitation is most likely attributable to a low water/rock ratio and to the fine-scale of the fracture networks and low inter-granular porosities in the bedrock where widespread ancient salt stores and ancient saline groundwater reside.

• This finding − the poor detection by the AEM of pervasive saline groundwater

in fractured bedrock aquifers − may present a similar limitation with application of the AEM technology for other fractured Palaeozoic bedrock provinces around the edges of the Murray Darling Basin. This is pertinent to future AEM surveys planned for regions that encompass fractured Palaeozoic bedrock provinces such as Wagga Wagga, Kamarooka and the South West Goulburn region and widespread exposures of this country rock elsewhere in the Lachlan Fold Belt. In such regions, drilling and groundwater monitoring may be a more appropriate and cost-effective strategy than conducting airborne surveys.

• Reconstruction of the watertable in adjacent aquifer systems (Cresswell et al.,

2003) indicates a high degree of connectivity between the bedrock aquifers of the Caniambo Hills and the main alluvial aquifer system of the Riverine Plain to the north. Although the watertable, overall, is relatively flat, the ~70 m of relief implicates particularly deep watertables beneath the hillcrests and very shallow watertables beneath the adjacent plains. The consequence is saline seepage at the break-of-slope between the hills and the alluvial plain despite the low hydraulic gradient.

• Construction of the elevation of the watertable map combined with 14C dating

of resident groundwaters reveals that the system is very sluggish, regardless of interconnection between adjacent GFS. Although palaeowaters are resident in the bedrock aquifers and low hydraulic conductivities and low hydraulic gradients are represented, pulses of recharge appear to translate to pulses of discharge in down-gradient settings. A measure of hydrologic equilibrium may, therefore, have been attained, with discharge approximately equal to recharge under current climatic conditions. This may indicate hydrologic balance and a fairly brimfull aquifer system. A longer period of monitoring is required to more realistically establish the degree of hydrologic equilibrium. Such information is wholly dependent upon field observations, conventional data and system understanding.

• Break-of-slope salinity is a relatively surficial phenomenon − albeit

deleterious to productivity at the paddock scale and to the integrity of soils, local streams and habitats − and it is possible that contributing conductivities in the near surface, such as plant roots, may interfere with the AEM signal for the uppermost CDI. Thus, even though break-of-slope salinity is linked to broader, deeper causative processes, the actual saline seepage sites may be small, scattered and very shallow, underlain by a non-saline system. Saline groundwater is likely to occur within 3 m below ground (BG) at the discharge sites, with underlying groundwater being relatively fresh. Moreover, the footprint size of the AEM data resolution which is effectively 40-100 m2, may limit remote detection of many break-of-slope salinity outbreaks at the base of the Strathbogie Ranges (and analogous regions elsewhere) since these seeps


xiii

and salt scalds are commonly smaller than this grid cell size. AEM data should not be relied upon to detect or predict the occurrence of break-of-slope salinity where its manifestation is typically shallow and patchy. Ground-based datasets are more appropriate for mapping the extents of and predicting future outbreaks or expansions of these relatively dynamic saline zones. More importantly, the tendency for shallowing groundwaters near any break-of-slope zone needs to be detected a decade or a few decades before salinity develops, i.e., long before any remote sensing technique is capable of registering conductivity changes wrought about by the build-up of efflorescent salts at the exposure points and before the ensuing cycle of seasonal flushing of those salts is set into motion.

• In the lowlands of Sheep Pen Creek and Earlston sub-catchments there is a

correlation between the occurrence of Grey Vertosols, salinity outbreaks, and AEM highs. The correlation is not universal within the catchment. Not all Grey Vertosols correspond with shallow AEM highs or with known salinity outbreaks. For example, Grey Vertosols are widespread on the plain located between Honeysuckle Creek and the northern edge of the hills north of Miepoll. These areas do not correspond with pronounced shallow AEM highs or with any mapped salinity. The eastern Violet Town Plain is dominated by Vertosols and Sodosols which may correspond with widespread moderate AEM signature in the 5-15 m CDIs; no saline discharge areas have been mapped in this area. Distinctive AEM highs in subtle valleys within the hills north of Miepoll are areas of Grey Sodosols. Some mapped salinity outbreaks in the catchment correspond with Sodosols (at Kialla East and in the eastern part of Earlston sub-catchment, where there is no shallow AEM high) or with Kandosols and Kurosols (in the Strathbogie Ranges, where the shallow AEM response is variable). Similarly there is no widespread correlation between saline outbreaks and specific radiometric signatures. An appreciation of topographic position, geology, palaeoenvironmental evolution and depth and salinity of groundwater are deemed more important to understanding the occurrence of salinity at the landscape surface than characteristics that relate to bulk conductivity patterns or specific radiometric signals.

• Discharge of saline groundwater directly to the trunk drainage channels and

subordinate waterways in the relatively flat Riverine Plain is a process involving the third dimension, namely the depth of incision (the z dimension) of the main channels down-cutting through the x-y spatial dimensions in the landscape. Deeply incised streams are geomorphic features that are not well represented in regional to local-scale airborne datasets, given that the incision is of the order of 12 m depth. Despite low hydraulic gradients, small volumes of saline groundwater from shallow alluvial floodplain aquifers seep into stream beds by virtue of the considerable depth of incision of the channels and the shallowness of the watertables, not necessarily because of any substantial driving hydraulic pressures. Understanding this manifestation of salinity is best attained through geomorphologic knowledge, an appreciation of pre- and post clearing watertable levels, and monitoring of streamflow and stream salt load trends.


xiv

• Major salinity sites in the catchment, including salinised swamps and the location of an increase in the salt load in the Broken River, are associated with bedrock/topographic constrictions. Saline groundwater is evidently forced upwards to the landscape surface because of constricted flow paths, to emerge in vulnerable topographic low points. The constrictions are best revealed in topographic and geologic maps and digital elevation models (DEMs) and the MrVBF terrain analysis.

• The understanding of salt stores, salt mobility and salinity risk and hazard in

Honeysuckle Creek catchment presented in this report has mainly been derived from conventional field-based techniques with support from remotely sensed datasets. Recommendations to avert salinity or to deal with already existing saline areas are derived largely from an appreciation of physical processes operating over a range of spatial and temporal scales, and a general understanding of catchment behaviour and climatic criteria.

The present study underscores the importance of integrating the use of all remotely sensed datasets with drilling and field and laboratory analysis, and with as many supporting datasets as can be acquired for a given area under investigation. System understanding, of given catchments, and process understanding − of the underlying causes and mechanisms of landscape and waterway salinisation in wider regions − are equally important. The present study particularly emphasises the need to establish the relationships between landscape and hydrology in three dimensions. This framework then needs to incorporate knowledge about temporal behaviour (the fourth dimension) and variations that relate to climatic drivers, albeit with the tacit understanding of time lags inherent between surface hydrologic processes and responses in the groundwater system. Secondly, the approximate status quo with respect to hydrologic equilibrium or disequilibrium and salt equilibrium or disequilibrium needs to be established for any system under investigation. Understanding the balances between recharge and discharge and between salt input and salt output, and scrutiny of salt load dynamics and salinity trends in our streams are here regarded as more important than static mapping of the distribution of substrate conductivities in a given system. Recommended intervention measures stemming from the hydrogeological datasets and interpretation of the airborne geophysical data are based on our current experience in the multi-disciplinary Heartlands Initiative. The latter include low to medium rainfall agroforestry trials being conducted within the Honeysuckle catchment. These are aimed at optimising recharge reduction, commercially viable farm forestry, and the preservation or restoration of threatened environments within the region. The experience gained through the Heartlands work in recent years in both Honeysuckle and Billabong catchments has emphasised the imperative of field-based hydrogeological investigations towards understanding key features in given systems and the dynamics of salinity. Drilling and measuring watertable depths and groundwater salinities, in particular, are obligatory. Airborne geophysics datasets have been found to be useful but not essential, particularly if the cost of research and on-ground action is a limiting factor. Comprehensive salinity investigations leading to appropriate measures to ameliorate salinity can be conducted using conventional hydrogeological techniques in the absence of airborne geophysics datasets, but not vice versa. Sensible application of airborne geophysics is wholly dependent upon follow-up drilling, calibration, substantial groundwater data and system understanding.


1

1. Introduction The Goulburn Broken Catchment is one of 21 designated Priority Regions of the National Action Plan for Salinity and Water Quality drawn up by the Council of Australian Governments (NAPSWQ, 2000), and one of the four Priority Regions in Victoria. The National Action Plan identifies high priority, immediate actions to deal with salinity – particularly dryland salinity – in key regions across Australia. The NAP is aimed at helping regional communities to prevent, stabilise and reverse trends in dryland salinity where they affect the sustainability of production, biological diversity and/or infrastructure. The broad salinity issues in the Goulburn Broken region include: increased salinity due to rising watertables, known dryland salinity outbreaks of the order of 3000 hectares (Allan, 1994), significant land degradation and farm production losses, particularly in the highland front areas, and substantial salt export into the Goulburn River and thence to the northern irrigation areas and the River Murray. A Salt Loads Study of Victoria (Sinclair Knight Merz, 1999) calculated that the salt load leaving the Goulburn-Broken catchment was 6700 tonnes in 1998 and would double to 12,500 tonnes by 2020 at present rates of increase. Salt exports from the Honeysuckle Creek catchment through groundwater discharge to the Goulburn and Broken rivers are large although the catchment is not the highest salt exporter in the overall Goulburn-Broken region. Water quality in many of the tributaries in the Honeysuckle area has declined. It is assumed that declining water quality will continue within the region if there are no effective land management interventions, and stream and wetland environments would become seriously degraded as a consequence. The impacts of increased land and stream salinisation on the social, economic and environmental values of the region are incalculable.

1.1 Background This is a collaborative project, between the Foundation for Rural and Regional Renewal (FRRR), the Goulburn Broken Catchment Management Authority (GBCMA), the Commonwealth Scientific and Industrial Research Organisation (CSIRO), the Victorian Department of Primary Industry (DPI), local communities and landholders. The FRRR project relates closely to a larger program called the Heartlands Initiative, (http://www.clw.csiro.au/heartlands/publications/general_publications.html) a collaborative R&D project, in which the Murray-Darling Basin Commission (MDBC) is a major partner along with CSIRO, the GBCMA, the Natural Heritage Trust (NHT) and a number of other organisations. The Heartlands Initiative is aimed at designing management strategies and implementing on-ground works towards sustainable land use in the Murray Darling Basin (Heartlands Core Group, 2000; Cresswell et al., 2002; Cresswell et al., 2003). Two Victorian catchments are part of the current Heartlands effort, Sheep Pen Creek, in the Honeysuckle catchment (GBCMA, 2000), and the Mid-Ovens Basin. In addition, there are two NSW Heartlands catchments, Billabong Creek and Kyeamba Creek in the eastern Riverine Plain and adjacent uplands. These four catchments were selected for applied research and on-ground works by the respective catchment management groups and local communities on the basis of pressing environmental and land management issues. The overarching objective is to design and achieve socially acceptable land use change in focus


2

catchments to alleviate environmental degradation, and to transfer the knowledge gained to comparable catchments elsewhere in the basin. Two reports have been produced from the FRRR project. This report focuses on interpretations and analytical results. A complementary Client Report (English et al, 2004) summarises our main findings and the implications and recommendations stemming from our integrated research in the catchment and also includes additional information with respect to communication and capacity building within the context of Heartlands efforts in the region.

1.2 Sub-catchment prioritisation Assessment of 53 upland sub-catchments in the Goulburn Broken dryland catchment was carried out in 1999 by the then Department of Natural Resources and Environment (DNRE) with the aim of ranking the sub-catchments in terms of severity of both existing and potential salinity (Cheng, 1999). The assessment was based on six parameters: area of discharge to land, discharge to stream, area of high watertable, rising watertable, groundwater salinity and percentage of land clearance. Twelve high priority sub-catchments were highlighted. Sheep Pen Creek, including Earlston sub-catchment (Figure 1), received the highest ranking of the 53 sub-catchments. Cornella Creek, in the South West Goulburn catchment, received the second highest ranking. Honeysuckle Creek sub-catchment ― including the break-of-slope areas in the Boho locality ― received the third highest ranking. In the case of Cornella Creek in the South West Goulburn catchment, salinity relates to the distinctive geologic configuration of the Heathcote Greenstone Belt and the associated topography of the Mt Camel Range (Centre for Land Protection Research, 2003) and, as such, is atypical for the Goulburn Broken catchment overall. The Broken River sub-catchment received eighth highest ranking of the 12 high priority sub-catchments. This designation relates mainly to break-of slope salinity in the Warrenbayne locality and stream discharge to the Broken River at Nalinga. Although the median salinity of the Broken River within the study area is 172 µS/cm EC (Electrical Conductivity), this reflects the dilution effect of high streamflow sourced from the Strathbogie Ranges and the Great Dividing Range. North of Caniambo, the Gowangardie stream gauging station on the Broken River has shown a long-term upward trend in stream salinity of 10.6 µS/cm/year (Cheng, 1999). In the Kialla East Plain (Figure 1) − part of the Broken River floodplain − groundwater monitoring data show significant rising trends of 3-7 cm/year in watertables, particularly along the Broken River, Honeysuckle Creek and the boundary between the dryland agricultural area in the plains and the Shepparton Irrigation District (CLPR, 2001). In Sheep Pen Creek, 16 discharge sites were mapped, representing a total area of 714 ha; widespread high watertables were mapped on the plain and at plain-hillslope interfaces, representing 40% of the total area. Groundwater salinity measurements range from 2000 to 25000 µS/cm EC, with a median (from 45 bores), of 17800 µS/cm. Watertable rises of 11-18 cm/yr were identified in selected bores, and 97% of the land had been cleared, leaving only 3% remnant forest (Cheng, 1999).


3

In Honeysuckle sub-catchment, 56 discharge sites were mapped, with a total area of 589 ha, the majority of discharge sites being at the break-of-slope areas near Violet Town; high groundwater levels were estimated for 10% of the area, occurring at break-of-slopes and along drainage lines. Groundwater salinities ranged from 300 to 15000 µS/cm, with a median of 1450 µS/cm (43 bores). Watertables in selected bores were observed to rise 7-26 cm/yr, and 90% of the total area had been cleared, leaving only 10% remnant forest (Cheng, 1999). This prioritisation of Sheep Pen Creek and Honeysuckle Creek sub-catchments of the Mid Goulburn Broken Catchment was followed up by a study of salinity processes in the Broken and North Goulburn Plain (Dahlhaus et al., 2000) and an investigation of farm forestry options for dryland salinity control in the Broken riverine plain and surrounding upland areas (CLPR, 2001). These studies were an impetus for the CSIRO-MDBC Heartlands involvement in Honeysuckle catchment (Heartlands Core Group, 2000; Cresswell et al., 2002; Cresswell et al., 2003), and more recently, for the FRRR project and an Honours research project in the Kialla East and Sheep Pen Creek area (Fisher, 2003). Honeysuckle catchment is, thus, an area of concern for the Goulburn Broken Catchment Management Authority which has the role of coordinating environmental management of the overall region. It was selected for research due to the pressing need to find solutions to the obvious environmental issues such as dryland salinity and to the perhaps less conspicuous decline in biodiversity. Central to the required research is the recognition that natural resource management must be approached in an integrated way to ensure issues are not examined in isolation but on a catchment-wide scale.

1.3 Objectives The overall aim of the FRRR project is to help determine appropriate land use change to prevent or mitigate salinity in the Honeysuckle catchment by using conventional hydrogeological methods and airborne geophysical data. The project is divided into the following components:

• Groundwater and salinity processes • Water yield contribution to rivers streams and wetlands • Land resource assessment • Designing revegetation options • Communication and capacity building

Utilisation of the airborne geophysics ― and assessment of its applicability with respect to land use change in saline regions ― is an imperative of the project. Investigation of groundwater and salinity processes is, therefore, the central component of the study. The land resource assessment and water yield components are aimed at complementing our understanding of hydrologic and landscape processes by utilising new techniques that have not previously been applied in the Honeysuckle catchment. Designing revegetation and communication and capacity building components are not aimed at being exhaustive or conclusive, but rather, are part of on ongoing iterative process that the researchers are already committed to in the Goulburn Broken region.


4

Specific scientific objectives of the Groundwater and Salinity component of the study are directed towards addressing the following key questions:

• Where are salt stores located, and how much salt is present? • Are these salt stores being mobilised to lower landscape positions and to the


Any land use changes suggested from this study are aimed at multiple objectives, including consideration of:

• improving water quality • improving aquatic ecosystems • restoring biodiversity • long-term social and economic imperatives

1.4 Study area Honeysuckle catchment covers an area of approximately 80000 ha within the Goulburn Broken CMA region. Honeysuckle Creek is a second-order catchment which feeds the Goulburn and Broken rivers. The catchment lies on the southern edge of the Murray-Darling Basin, between the Goulburn and Ovens rivers, and between the Victorian Highlands to the south and the Broken River to the north. Violet Town lies within the Honeysuckle catchment. Figure 1 shows the boundaries of six of the sub-catchments of Honeysuckle catchment (Honeysuckle, Sheep Pen, Earlston, Kialla East, Riggs and Woolpress creeks). These sub-catchments are third-order catchments within the Goulburn-Broken region. The airborne survey extent is outlined in Figure 1; the survey area lies within the triangle formed by the northeastern Victorian towns of Shepparton, Benalla and Euroa. The need for level terrain clearance by low-flying aircraft precluded southward continuation of the survey over the Strathbogie Ranges. Figure 2 shows the topography of the area; the geomorphology is described in Section 3.2. The climate of the area is warm temperate, with dry summers and most of the rainfall occurring in the winter months. Figure 3 shows the regional rainfall gradient for the whole Goulburn Broken catchment. The average annual rainfall within the study area ranges from less than 550 mm at Kialla East in the northwest to 1000 mm on the hills south of Warrenbayne-Boho. The mean daily maximum and minimum temperatures are around 30oC and 4oC; frosts are fairly common. Annual potential evaporation is over 1000 mm. The area is more than 90% cleared and is used for dryland grazing and cropping. Salinity was first recognised as an issue in the area 70 years ago. Now, saline outbreaks are common in the break-of-slope zone at the base of the Strathbogie Ranges (e.g., Figures 4 to 6) and in topographically low lying areas where the Riverine Plain abuts the Caniambo Hills. Saline discharges to swamps (e.g., Figure 7) and to the Broken River are also a concern.


5

Figure 1. Map of sub-catchments making up Honeysuckle Creek catchment and the extent of the 2001 Airborne Geophysics survey. The survey area lies between Benalla in the east and Shepparton in the northwest. The Goulburn River is to the immediate west of the study area.

Figure 2. Digital Elevation Model (DEM) of survey area.


6

Figure 3. Rainfall gradient of the Goulburn-Broken catchment. The study area straddles the 550 to 1000 mm rainfall zone.

Figure 4. Cleared Strathbogie Ranges and waterlogged piedmont plain in the Boho area. The ranges are high rainfall/runoff areas, located in the 1000 mm/year rainfall zone, and generate excessive recharge to adjacent/subjacent aquifers. This results in shallow watertables, waterlogging, and, in places, tree mortality and replacement of grasses with sedges.


7

Figure 5. Severe salt scalding and tree mortality in the Warrenbayne piedmont area (Whites Road) near the break-of-slope, where the plain meet the ranges, and where there is excess recharge from the cleared slopes.

Figure 6. Salt efflorescing above the exposed watertable in a gully at Boho where Spiny Rush (Juncus acutus) and Sea Barley Grass (Hordeum marinum) are colonising saline groundwater discharge sites in the break-of-slope area.


8

Figure 7. Downes Swamp between Sheep Pen Creek and the Broken River, at the down-gradient end of the Honeysuckle Creek catchment. The swamp is located at a topographic constriction where saline groundwater is evidently forced upward resulting in complete salinisation of the swamp and severe eucalypt mortality.

1.5 Airborne Geophysics The MDBC Airborne Geophysics Project was set up and flown in 2001 to investigate the usefulness of airborne geophysics for the management of salinity at a catchment scale. The intention was to conduct the airborne geophysics surveys over the Heartlands project areas, namely the Upper Billabong Creek, NSW, and Honeysuckle Creek catchments. The aim was to provide high-resolution datasets − airborne magnetic, gamma-ray spectrometric, and electromagnetic data − to help understand how the landscapes work and to develop management options to alleviate the salinity. The airborne electromagnetic (AEM) data was flown with the TEMPEST system (Lane et al., 2000) at close line spacing (200 m) aimed at assisting in the identification of the distribution of salt stores and saline groundwaters at depth. The high-resolution magnetic and gamma-ray spectrometric data was aimed at assisting with the delineation of important features of the soil and regolith that may influence the hydrology of the catchments. Such information has the potential to be an aid in determining and prioritising actions to combat salinity. Calibrated reprocessed airborne geophysics datasets were released by the MDBC in early 2003 for use in the Heartlands work. For the study area to benefit from this new knowledge of salinity there are other steps to be completed: (a) local investigations need to be undertaken to assess the groundwater systems and the likelihood of groundwater mobilising stored salt either into waterways or to the land surface as saline seeps, (b) land resource assessment needs to be undertaken to determine landscape function and land suitability, (c) where revegetation works are part of the solution for managing water and salt then appropriate spatial vegetation mosaics need to be identified such that revegetation is


9

well targeted for environmental outcomes and commodity production, and (d) recommended on-ground actions need to be implemented with support through an effective local communication and knowledge exchange program. Considering these aspects is the objective of the four parts of the current FRRR project, Interpreting Airborne Geophysics as an adjunct to Hydrogeological Investigations for Salinity Management: Honeysuckle Creek Catchment, Victoria. The outcome will be enhanced community capacity to deal with the management of water and salt in the local region.

1.6 Previous work Early work on salinity within the area was carried out by George (1984), Burns (1988), Day et al. (1990), Day and Harvey (1994), Kelly (1994), Clarke (1999), Cheng (1999), Dahlhaus et al. (2000), and the Centre for Land Protection Research (CLPR, 2001). Initial interpretation of the Honeysuckle airborne geophysics survey data has been documented by Dent et al. (2002), Dent (2002), and Gibson and Wilford (2002). These interpretations did not include interrogation of groundwater data from the >100 existing bores in the area. Collaborative interpretation of the airborne geophysics datasets was subsequently carried out during several workshops hosted by the GBCMA in Benalla during 2002-2003 as part of the MDBC Airborne Geophysics Project. Technical aspects of the airborne geophysics survey are described in the MDBC Airborne Geophysics Project Final Report (edited by Dent, 2002). Interpretation of the revised, calibrated airborne geophysics data, carried out concurrently with this FRRR project, is reported by Christensen (2003; 2004a). Heartlands work in the Honeysuckle catchment − namely in Sheep Pen Creek sub-catchment − is reported in Cresswell et al. (2002), Myers et al. (2002), and Cresswell et al. (2003).

1.7 Project scope The FRRR project was an 18 month project that stemmed from and overlapped with Phase II (2001-2003) of the Heartlands project. Final reporting for Phase II of the Heartlands project was completed in August 2003 (Cresswell et al., 2003). Although the FRRR project commenced in November 2002, the core component of the project ― utilisation of the airborne geophysics data ― awaited the availability, in February 2003, of the calibrated reprocessed AEM data. This report summarises the work conducted towards the FRRR project objectives. The adopted methodologies are outlined. Some of these approaches and techniques are newly developed or have not previously been employed within the study area. Our main results and interpretations are presented here. Some aspects of our FRRR research endeavour are ongoing and have not yet been drawn to closure. Relatively static aspects of the study, such as verifying conductivity highs as salt and analysing the actual volume of salt in the observed stores, are reported. More dynamic aspects relating to recharge and discharge rates and extents


10

and future risks to assets − environmental and agricultural − may need more involved analysis over longer periods than the time allocated for the present investigation. Similarly, the communication and capacity building component of the study is ongoing and, in fact, with respect to our FRRR research, can only begin in earnest upon completion of the research component and release of documentation of our findings and recommendations. It is anticipated that further work will be conducted in later 2004 and through 2005 to incorporate and develop various aspects of the FRRR project. Ideally, this could include further planning with respect to management options in respective groundwater flow systems in the catchment. Supplementary work following on from the present project is expected to incorporate communication components that are aimed at knowledge transfer and community capacity-building. We aspire to further our commitment in Honeysuckle catchment and the Goulburn Broken region generally and to build upon the foundations laid by the Heartlands and FRRR investments in the area. Our relatively preliminary research in the FRRR study needs now to be developed according to the full process developed in the Heartlands project (Cresswell et al., 2003). Any land use allocation must be based on multiple considerations, not piecemeal applications, and will sometimes involve tradeoffs between competing land uses. Land use allocation can be driven by a series of relatively simple rules or guidelines that reflect knowledge of catchment process and environmental, social and economic objectives. Sets of guidelines for use within the FRRR study area now need to be formulated to reflect our knowledge of salinity, water yield, biodiversity, and associated considerations. This is a process of synthesising information and is drawing on the expertise of researchers involved in the wider Heartlands program.


11

2. Research Approach

2.1. Overall approach The FRRR project adopted an approach similar to that utilised for the Heartlands Initiative insofar as the following aspects were studied to achieve the objectives outlined above: • Groundwater and salinity processes • Land resource assessment • Water yield contribution to rivers and streams Emphasis was placed on the interpretation of the airborne geophysical data for the present project, and to assess the potential of these datasets to guide catchment management recommendations in the Honeysuckle catchment. The above component studies generated an understanding of how the Honeysuckle Creek system behaves and information pertinent to prioritisation of management options and to the prediction of impacts from changed land use. The methodology and conclusions can be appropriately transferred or extrapolated to other areas in the Goulburn Broken catchment. It also holds relevance to the application of airborne geophysical data to other regions, particularly the edges of the Murray-Darling Basin abutting fringing uplands.

2.2. Groundwater and Salinity Substantial groundwater investigations were undertaken in the Honeysuckle catchment for the FRRR project. This work was directed towards understanding the hydrogeology and the key groundwater dynamics operating that are relevant to the development of observed salinity and waterlogging, the potential for future rising watertables and consequent soil and stream salinisation. This effort utilised the recently-acquired MDBC airborne geophysics datasets and complementary data and involved intensive fieldwork and laboratory analytical programs, outlined below.

2.2.1. Airborne Geophysics As part of the National Action Plan for Salinity and Water Quality (NAPSWQ, 2000), the MDBC commissioned airborne geophysical surveys over the Billabong Creek catchment in NSW and the Honeysuckle Creek catchment in mid-2001. The MDBC Airborne Geophysics Project was aimed at supporting the Heartlands work in both these focus catchments. The surveys were contracted through the Bureau of Rural Sciences (BRS) and flown by Fugro Airborne Surveys Pty Ltd and Kevron Geophysics Pty Ltd, with Geoscience Australia (GA) managing the data acquisition. Survey specifications are documented by Brodie (2002). Two airborne geophysical surveys were flown over the respective catchments, using different aircraft and instrumentation:


12

• Magnetic, gamma-ray spectrometric and elevation survey, the ‘MAGSPEC’ survey, flown at 100 m flight line spacing with 60 m terrain clearance.

• Electromagnetic survey utilising high-resolution TEMPEST time domain airborne electromagnetic system, the ‘AEM’ survey, flown at 200 m flight line spacing with 115 m terrain clearance.

The flown survey area is outlined in Figure 1. The specifications of the Honeysuckle MAGSPEC and AEM surveys and basic processing are detailed by Brodie (2002). Digital terrain data were collected during the airborne MAGSPEC survey using the aircraft radar altimeter and a Global Positioning System (GPS). Magnetic and Gamma-ray spectrometry (‘radiometric’) data Airborne magnetic measurements detect local variations in the earth’s magnetic field that are typically associated with different rock types. Derived maps of the Total Magnetic Intensity (TMI), principally the 1st Vertical Derivative data, enable interpretation of the spatial distribution of magnetic patterns. The resolution, or size of the grid cells in output products is one fifth of the flight line spacing, around 20 m. Airborne gamma-ray spectrometry, or ‘radiometrics’, detects variation in the natural radioactivity of the uppermost 20-30 cm of soil, regolith and rock, potentially providing information about different types of sediments and soil materials and about surface drainage patterns in a survey area. Gamma emissions from the radioisotopes of potassium (K), thorium (Th), and uranium (U) are measured using an on-board spectrometer. The relative concentrations are displayed as red-green-blue (K-Th-U) in output imagery derived from the airborne measurements. In this scheme, a low gamma signal in all channels registers as black, commonly corresponding with quartz (SiO2) or water bodies, i.e., materials that do not emit K, Th, or U radiation. A high signal in all channels is registered as white. Airborne Electromagnetic (AEM) data The TEMPEST electromagnetic system used for the airborne survey records the apparent conductivity of the earth by measuring the induced response to an electromagnetic pulse generated on board the aircraft. For each pulse a time series of the response is recorded by a receiver towed behind the aircraft. Additional processing of the electromagnetic signals is required to present the data range as images. One type of analysis produces Conductivity Depth Images (CDIs) or depth slices. This involves complex mathematical procedures. The resolution or size of the grid cells in output products is one fifth of the flight line spacing, i.e., 40 m2 (‘footprint size’). The ground response, however, generally averages over an area 2-3 times this size, limiting on-ground resolution to ~100 m2. The generated depth slices are highly sensitive to a number of factors, including water content, salinity, porosity and mineralogy, roughly in that order of importance. Calibration is strongly time dependent, requiring significant calibration using borehole data, preferably from bores >100 m deep. The TEMPEST AEM data were calibrated on a drill-hole to drill-hole basis against conductivity logs recorded in 13 calibration drill-holes with an electromagnetic induction conductivity EM-39 device (see also Section 2.2.2, below). This phase of drilling, calibration and reprocessing of the 2001 airborne data with respect to down-hole measurements was funded by the MDBC and GBCMA and carried out by the BRS and the Victorian DPI in 2002. Eleven drill-holes were sunk by the BRS for the


13

calibration exercise and two existing CLPR drill-holes were additionally logged. This BRS calibration drilling program is documented by Jones (2002) and summarised by Dent (2002). Where reference is made to the BRS drill-holes in this report, the 2001 BRS numbering sequence is used; these drill-holes were renumbered in 2004 (both numbering sequences are given in Table A5.1, Appendix 5). The DPI calibration procedure and conversion algorithms used to produce the CDIs are described by Christensen (2002; 2003, 2004a). The 13 BRS calibration drill-holes were used to anchor the CDI processing for the given footprint for each of the sites and parameters were subsequently applied to the intervening raw AEM data. The down-hole conductivity logs for the 13 drill-holes were edited to remove extraneous data and the recorded conductivities were averaged over 5 m intervals within the range of 0.1 mS/m to 400 mS/m (Christensen, 2003, 2004a). Examples of the initial and calibrated data for the 0-5 m CDI for the Honeysuckle Survey area are compared in Figure 8 (a) and (b). Relatively high conductivities are displayed in red and low conductivities in blue. Figure 8 (a), to some extent, may represent vertical variations within the root zone. The revised CDI slices generally indicate higher conductivity areas to exist at greater depths than the initial product. The impact is greatest on the shallow CDI depth slices. This is borne out in Figure A4.2 (Appendix 4) where a comparison of the initial and reprocessed measurements for the 0-5 m CDI for 92 grid cells are graphed along with EM-34 measurements for the 92 sites (described below, Section 2.2.2) where the reprocessing has subdued the high conductivities of the shallowest CDI. The revised AEM data has tended to relocate high conductivity areas in the vertical plane rather than the horizontal (Christensen, 2002; 2003, 2004a). The reprocessed product shows an improvement in both the vertical and lateral resolution of conductivity highs compared to the initial dataset. In particular, the conductivity distribution more accurately reflects some of the known localised salinity outbreaks. The latter have been mapped by DPI and are shown as red outlines on Figures 8 (a) and (b), described further in Section 3, and are also included as overlays in various other maps and images. More importantly, the reprocessing rectifies the extent of shallow conductivity seen in the initial data, Figure 8 (a), which is inconsistent with the absence of salinity and the agricultural productivity of these extensive areas. The reprocessed data also improves the elevation data that was originally derived from the geometrical parameters relating to the transmitter-receiver separation and the transmitter elevation during the survey. The conductivities obtained by down-hole geophysical logging were supplemented by geochemical analysis to confirm that the electromagnetic measurements in the Honeysuckle survey were a good proxy for salt (R. Cresswell, 2002). The reprocessed CDI products were made available to CSIRO for the present project in February 2003. All reprocessed CDI slices for the Honeysuckle, Miepoll and Baddaginnie areas are shown in Appendices 1, 2 and 3.


14

Figure 8. Airborne Electromagnetic images: (a) Initial conductivity map for the 0-5 m depth slice, uncalibrated (2001); and (b) Revised, calibrated conductivity map 0-5 m depth slice (2003). Both images are draped on a greyscale DEM of the survey area. High conductivity = >350 mS/m (red); moderate conductivity = 350-150 mS/m (orange-yellow); low conductivity = <150 mS/m (green-blue).


15

2.2.2. Field investigations Drilling and piezometer installation CSIRO commenced drilling in the Honeysuckle catchment in January 2003 as part of the Heartlands effort, prior to the availability of the reprocessed AEM data. Our initial drilling was in Sheep Pen Creek sub-catchment, a Heartlands focus catchment, designated the highest priority of 53 upland sub-catchments in the Goulburn Broken Dryland Catchment in terms of severity of existing and potential salinity. Two Heartlands drill-holes and piezometer installations, SPC01 and SPC02, were sited in highly saline areas characterised by topographic constrictions at the base of the northern slopes of the Caniambo Hills. Onlapping alluvium at both sites is typically Grey Vertosols, and shallow watertables also characterise both sites. Additional analysis of these two SPC drill-holes and extended investigations in Sheep Pen Creek sub-catchment were enabled through the FRRR commitment which broadened our overall involvement in Honeysuckle catchment geographically and scientifically. Twelve drill-holes were sunk in May 2003 specifically for the FRRR project. Drilling sites were based on the airborne magnetic, gamma-ray spectrometry and reprocessed AEM datasets. Site selection was also aimed at infilling gaps in the distribution of existing bores (Figure 40) and eleven BRS calibration drill-holes (Figure A5.1). All drill-hole locations, FRRR01 to FRRR12 and SPC01 and SPC02, are shown in Figure 9 (red dots). The geographic locations, landscape settings, distinguishing airborne geophysical features, and rationale for selection of the sites are outlined in Appendix 5, accompanying the drilling logs for all 14 holes (SPC and FRRR). Only those holes that intercepted groundwater were installed as piezometers. An exception was FRRR09 which, although dry, was installed because the site presented itself as a potentially significant aquifer in the future if watertables rise much higher than present levels in the Honeysuckle Graben. Thus, installations were made at the following sites: FRRR02, 03, 06, 09, 10, 12, and SPC 01 and 02; other drill-holes were back-filled after down-hole measurements and drilling samples were taken. A Proline solid flight auger drill-rig and a Jarrett hand auger were used for drilling (Figure 10). The depth at which saturated sediments – i.e., aquifers, or free waters – were initially intersected during drilling of each piezometer was recorded for comparison with final water level measurements (Standing or Static Water Levels, SWLs). Aquifer confinement was indicated where the groundwater level rose after completion of drilling. Piezometers made from 50 mm diameter PVC were slotted over a 1 m long section at the base, open to the aquifer. The bottom 1.0 m of the drill-hole, around the PVC pipe, was backfilled with coarse river sand to ensure good movement of groundwater into the piezometer. The basal sandy back-fill was overlain by approximately 0.5 m of bentonite, and the remaining space around the pipe was backfilled with excavated soil and sediment. The uppermost 0.5 m was cemented, with a steel cap installed.


16

Figure 9. FRRR and SPC drill-hole locations (red dots) and DPI-managed bores sampled for 14C groundwater dating (black dots).

Figure 10. Drilling FRRR9 in the western edge of Violet Town Plain through a resistive AEM (low conductivity) area near the junction of Faithful Creek and Seven Creeks.


17

Co-ordinates and elevations of all piezometers were accurately surveyed using a Trimble Differential GPS to enable measurement of the hydraulic gradients in the catchment and for terrain analysis. The drill-holes were logged on site. The logs and descriptions are presented, along with an indication of the watertable depth − where intersected (blue line on drilling logs) − in Appendix 5. Laboratory analyses of samples collected from approximately every 0.5 m during drilling are presented along with the lithologic descriptions in Appendix 5.

EM-34 Traverses and individual site measurements Electromagnetic Induction (EMI) traverses were carried out across eight sections of the catchment, both bedrock and alluvial areas. The transect locations were aimed at following up the airborne geophysical information, particularly AEM features, both conductive and resistive patterns. These transects, with EM readings taken every 10 metres, were aimed at providing higher resolution data in specific localities than afforded by the AEM for which a 40-100 m2 ground response is represented in each grid cell (pixel). In addition to the traverses, 92 individual sites were selected for EM-34 measurements. These were typically located at road junctions, or at distinctive landmarks for ease of relocating the sites if needed for future follow-up work. Figure A4.1 (Appendix 4), shows the locations of the eight transects and the 92 individual sites. The EM-34 instrument (McNeil 1980; GEONICS EM-34/3 Technical Notes, undated), with coils aligned vertically (horizontal dipole), measures the aggregate or apparent electrical conductivity (ECa) of the uppermost metres of substrate beneath the ground surface. In the case of the eight traverses, 10 m coil spacing was used, to measure the ECa within approximately the top 7.5 m. For each of the 92 individual sites, three intercoil separations were used: 10 m, 20 m, and 40 m, to measure respective exploration depths of 7.5 m, 15 m and 30 m. The units of measurement are conductivity units, milli Siemens per metre (mS/m). The EM-34 data are provided in Appendix 4. This information, along with the AEM data, assisted our site selection for drilling (rationale for all drill-hole sites provided in Appendix 5). EM-39 Profiling Down-hole EMI measurements were taken in the FRRR drill-holes as well as down some pre-existing bores in the catchment. The GEONICS EM-39 probe (McNeill, 1986), Figure 11, measures the aggregate electrical conductivity of soil, sediment, rock and moisture immediately surrounding a given depth. The radial distance of the measurement at each sample point is approximately 0.9 m; the peak response is at 0.28 m. Electrical conductivity measurements (ECa) were taken at 20 cm intervals down each borehole, measured in mS/m. These profiles are plotted alongside the drilling logs in Appendix 5.

Salinity Measurements Groundwater and surface water salinity measurements Groundwater and surface water salinities were determined by Electrical Conductivity (EC) measurements (Figure 12), in µS/cm (also known as ‘EC units’). Piezometers were bailed dry and allowed to refill prior to measuring to ensure that aquifer water


18

was represented. The following thresholds provide a baseline for gauging the severity of groundwater salinity:

Groundwater salinity (EC) µS/cm

<800 good drinking water <2500 potable >2500 not potable 10000 saline 50000 seawater

Where groundwater was intersected in the SPC and FRRR drill-holes, the EC measurements are annotated on the drill-logs in Appendix 5. Soil salinity measurements Soil salinity was assessed by 1:5 soil:water extracts on sediment samples retrieved at 0.5 metre intervals down the drill-holes. Assessment of the severity of soil salinity, EC1:5 can be based on thresholds:

Soil Salinity (EC1:5) µS/cm

<600 low 600 - 1400 moderate

>1400 high

Soil EC1:5 measurements for the SPC and FRRR drill-holes are presented as profiles alongside the drill-logs in Appendix 5. Using average bulk densities of the samples, the EC1:5 measurements were converted to salt stores in kg/m3 and tonne/ha (Table 1). Magnetic Susceptibility Magnetic susceptibility was measured on samples from the drill holes, using a Bartington Instruments Magnetic Susceptibility Meter, model M.S.2. The instrument’s probe was placed on the surface of the plastic sample bag at five random places, and the bulk magnetic susceptibility recorded and averaged. The units of bulk magnetic susceptibility are dimensionless and the values on the plots need to be multiplied by 10-8. The data are plotted with the drilling logs in Appendix 5, along with the gravel content to give an indication of the correlation between larger-sized grains and magnetism, where applicable (although it should be borne in mind that intersected gravels sometimes represent magnetically inert quartzite fragments). Contouring Depth to Watertable (Standing Water Level, SWL), Groundwater elevation, and Groundwater Salinity (EC) have been contoured from piezometer and bore-hole data and Differential GPS survey measurements using SURFER Version 7 based on a Kriging griding method.


19

Figure 11. Down-hole conductivity profiling using the EM-39 device (Bore 126 in upper Sheep Pen Creek catchment).

Figure 12. Measuring creekwater salinity, Sheep Pen Creek.

The depth to groundwater is the measured depth, in metres below ground (BG), at each bore or piezometer. The elevation of the watertable is calculated from the surveyed elevation of the ground surface at each bore or piezometer from which the


20

measured depth to groundwater is subtracted. This provides the elevation of the watertable above the Australian Height Datum (AHD, which is approximately sea-level), at each measurement point. In the case of groundwater elevation, it is noted that SURFER disregards the influence of geology and topography and, therefore, the contouring may not be representative of the continuity of groundwater systems or of actual groundwater flow directions where bore density is low or variation in topography and geology is great. Groundwater dating Radiocarbon (14C) analysis Radiocarbon (14C) dating of groundwater was conducted to help address one of the key questions that we have focussed on for the FRRR study, namely, ‘Are the salt stores being mobilised?’ The residence times of groundwater in an aquifer system can provide valuable information about its mobility and about the mobility of any dissolved salt or the potential mobilisation of any salt that may be intercepted along the groundwater flow path. Once rainwater percolates through the soil profile from the unsaturated zone to a confined aquifer, the radioactive carbon isotope, 14C, is isolated from the atmosphere with which it is equilibrated. In the atmosphere, most 14C is in the form of 14CO2. During recharge it is incorporated into groundwater where most dissolved inorganic carbon occurs as HCO3

-. Initial 14C activity is taken to be 100 percent of modern carbon (pmc) and this then radioactively decays at a known rate; the 14C half-life is 5730 years. After 10,000 years, only 27% of the initial amount of 14C would be present. (from Figure 13), and negligible amounts (1 pmc) would remain after 30-40,000 years, which is the approximate limit of measurements. Thus, the percentage of 14C in sampled groundwaters gives its approximate age. The Honeysuckle Catchment is suitable for the application of 14C groundwater dating because there are no known ancient limestones (CaCO3) which would contribute “dead carbon” to the groundwater and thereby skew the age of resident groundwaters downwards to unrealistically old ages that would not reflect the time elapsed since recharge. Bores were carefully selected to target deeper aquifers, since these are expected to be the most confined and isolated from modern atmospheric carbon. Shallow bores and those that show high degrees of responsiveness to rainfall events were excluded from the sampling program since these are likely to be frequently recharged with modern carbon. The locations of six bores selected for 14C analysis, bores 124, 126, 130, 133, 142, 144, are shown in Figure 9 (black dots). The sampling procedure involved purging the bores at the outset to expel all groundwater from the casings that would have equilibrated to the atmosphere and thereby been contaminated with respect to carbon. At each bore, 60 litre samples of aquifer water were pumped into large barrels, Figure 14. The sampled groundwater was treated with a flocculant to facilitate capture of all available dissolved carbon, and with sodium hydroxide (NaOH) to raise the pH to enable bicarbonate to be converted to a solid carbonate. Barium Chloride (BaCl2) was used to convert the


21

aqueous carbon into barium carbonate (BaCO3). Five litres of the dense carbonate slurry were separated from the supernatant and despatched for laboratory analysis.

Figure 13. Carbon-14 decay curve.

Figure 14. Pumping groundwater into barrels for 14C analysis, Bore 130, Caniambo Hills (localities shown in Figures 9 and 44).


22

The extent by which 14C in groundwater is altered by exchange with the local environment (rocks and soil organic and inorganic material) may influence the radiocarbon age. Therefore, the stable carbon isotope 13C is also analysed in the groundwater samples, and reported as per mille (‰) deviation of the 13C/12C (ordinary carbon) ratio with respect to the PDB laboratory standard. Measurement of δ13C can allow for variations in 14C activities due to chemical processes such as reaction with radioactively dead limestones in the aquifer matrix, if present, or oxidation of excessive soil organic matter. For a given system, the range of δ13C ratios provides a measure of the robustness to the 14C data. Radiocarbon analysis was carried out at the CSIRO Isotope Laboratory in Adelaide by the direct CO2 absorption method via a Liquid Scintillation Counter (Leaney et al., 1994). Sample Number 124 contained insufficient carbon for the direct absorption technique and was analysed using Accelerator Mass Spectrometry (AMS) at the Department of Nuclear Physics, the Australian National University. Oxygen-16, Deuterium and Tritium data Additional isotope chemistry for groundwaters in the study area is incorporated from recent work of Fisher (2003) and Cartwright et al (2004). Signatures in water molecules of the heavy isotopes, oxygen-18 (18O) and deuterium (2H) with respect to the lighter 16O and 1H isotopes are related to a laboratory standard to give an indication of the climatic conditions under which recharge has taken place and retention times of these stable environmental isotopes within the hydrosphere. The more enriched a sample is with heavier isotopes the more positive its δ value. Tritium (3H), the radioisotope of hydrogen, has a half life of 12.43 years and is useful for tracking the movement of groundwater in short time spans. The detection levels, expressed in Tritium Units (TU), are very low because of attenuation in recent decades since the thermonuclear bomb testing that had enriched the atmosphere with radioelements in the 1960s. Tritium background levels appear to be stabilising at around 3-4 TU in Australia.

2.3. Surface Water Yield Modelling The FRRR project is aimed at current and future consideration of revegetation (planting of trees and shrubs) as the major option towards greater hydrologic balance and to prevent or manage dryland salinity in the study area. It is tacit that biodiversity enhancement would also be sought as an additional benefit from such investment. Greater areas of trees and shrubs use more water than the agricultural plants that they replace, as illustrated in Figure 15. Consumption of relatively large volumes of water by trees confers advantage through preventing or reducing groundwater recharge. A consequence however, is that rainwater runoff is also reduced because of this interception and the consumption of available water by trees, which can reduce streamflow accordingly, as illustrated in Figure 16.


23

0

200

400

600

800

1000

1200

1400

0 200 400 600 800 1000 1200 1400 1600 1800 2000

Rainfall (mm)

Eva

potr

ansp

iratio

n (m

m)

ET for trees

ET for grass

Figure 15. Relationships between mean annual evapotranspiration and rainfall. Solid lines indicate average annual evapotranspiration and dotted lines represent ±2STD, where STD is standard deviation (Zhang et al, 2001).

00 500

400

800

1200

Mean annual rainfall (mm)

Me

an a

nnu

al w

ate

r yi

eld

(m

m)

1000

Grass

Forest

1500 2000

Figure 16. Relationships between mean annual water yield and rainfall (Zhang et al, 2001).

If salt mobilisation is unaffected (in the short term) by tree planting then less water in streams can mean increased stream salinity and potential deterioration in the in-stream ecosystems. Thus, tree planting may not necessarily lead to immediate decreases in river salinity (Vertessy et al., 2003). In some cases, whilst young trees are growing and using large amounts of water, downstream creek and river salinity might worsen before ultimately improving over longer time scales. Thus, any revegetation for recharge reduction and salinity management must be balanced against possible (undesirable) reductions in runoff and streamflow in each given catchment. A hydrological modelling approach (Zhang et al., 1999, 2001, 2003) was applied to predict the mean contribution of surface water from constituent sub-catchments that make up Honeysuckle catchment. The model calculates mean annual


24

evapotranspiration from mean annual rainfall and potential evapotranspiration. In estimating catchment average water yield, it is assumed that there is no net change in catchment water storage over a long period of time. As a result, catchment water yield can be calculated as the difference between long-term average rainfall and evapotranspiration. The water balance model was implemented in an ArcInfo Geographic Information System (ARCGIS). Three key datasets were required to calculate evapotranspiration: catchment boundaries, long term mean annual rainfall and the percentage of forest-cover in the catchment. Estimated evapotranspiration was then subtracted from the long-term mean annual rainfall to provide the calculation of water yield. To investigate the change in water yield, each scenario required a different forest-cover input, while the catchment boundaries and rainfall surface data remained constant. The model output is presented and described in Section 4.8.

2.4. Land Resource Assessment The land resource assessment strategy involved the use of existing information, field investigations, and the application of new technologies for spatial prediction. Maximum use was made of existing soil map data for the study area that had been compiled by the Centre for Land Protection Research (CLPR, now named Primary Industries Research Victoria, PIRVic) of the Victorian Department of Primary Industries (DPI); which includes early detailed CSIRO soil mapping by Downes (1949). Terrain analysis utilised 20 m resolution DEM elevation data acquired from DPI. The terrain analysis additionally used vector contours and drainage layers from available topographic maps. Terrain variables used for the FRRR project include the Multiple-resolution Valley Bottom Flatness index (MrVBF), the Terrain Wetness Index (TWI), predictions of soil depth, and available water capacity estimations. Landscapes vary in their capacity to store water. Estimates of water storage capacities in different soil types are desired to allow better analysis of interactions between vegetation and streamflow in a given landscape. Such analyses can be conducted at local to regional scales. For the FRRR study, water storage capacities were carried out at the catchment scale. This is a preliminary analysis aimed at providing a scenario for future simulation studies to relate dryland salinity, farm forestry and water security in the region.

Multiple-resolution Valley Bottom Flatness (MrVBF) While landforms can be qualitatively described using visual observations of geomorphic features and topographic maps, potentially useful quantitative methods have been developed using digital terrain analysis. A wide variety of terrain analysis methods are available; one that has proven application in depositional and low relief areas is MrVBF (Gallant and Dowling, 2003). The MrVBF index is based on both the position of parts of the landscape and their degree of flatness. It is particularly useful for identifying flat valley bottoms at a range of scales from small hillside hollows and rises to the broad riverine plains.


25

MrVBF has been tested in a number of areas in south-eastern Australia although it had not previously been field verified in the Goulburn Broken region. The MrVBF maps of the Honeysuckle catchment generated for the present study were checked using field observations at 26 sites. MrVBF, based on 20 m grid cell elevation data, was found to reliably identify the broad valley bottoms and to discriminate between different degrees of flatness. It mistakenly identifies some low rises as low-order valley bottoms. In the present study, interpretation of the MrVBF data is carried out in conjunction with interpretation of the airborne geophysics and soil data for the target areas, Miepoll, Baddaginnie and Sheep Pen Creek (Sections 4.3, 4.4. and 4.5).

Topographic Wetness Index (TWI) The Topographic Wetness Index (TWI) is based on landform information derived from land resource survey data to provide an estimate of material balances between erosion and deposition. High resolution digital elevation models and soil-landscape maps have been successfully used to predict soil depth in areas of net erosion for landscapes where rates of soil production limit sediment transport. The index is based on catchment area, slope gradient and landscape position and works well in steeper landscapes (McKenzie et al., 2003). The TWI is used to scale soil depth only in those parts of the landscape where erosion processes dominate. In the lower parts of the landscape (depositional zones) soil depth is better estimated using the MrVBF technique. The output and discussion are provided in Section 4.9.

Predictions of Soil Depth In the Australian landscape the depth of soil and regolith (weathered material on top of bedrock) is frequently related to topography. Steep high areas have little or no regolith, commonly being eroded to bedrock with soil cover tending to have been washed down-slope as it was generated. An exception probably exists in the friable and faulted ancient bedrock of the Caniambo Hills where the thickness of the regolith may be up to 70 m. Low flat areas more generally have substantial regolith depth, including alluvial deposits (fluvial and lake sediments), aeolian deposits and soil accumulations. Other factors such as the different rock materials, climate, palaeoenvironmental evolution and the history of soil development also contribute to variations in soil and regolith depth. Much of the broader scale variation in soil properties can be mapped using soil-landscape mapping techniques, while the finer scale variation due to topography can be derived from terrain attributes such as the MrVBF and TWI analyses. Two predictions of soil depth were made for the Honeysuckle catchment using these techniques. The first assumes that MrVBF only identifies where deep soil is located and uses a nominal 5 m depth in those areas. This assumption is appropriate for soil depth although not always for other types of regolith, such as weathered fractured bedrock. Soil depth in the remaining areas is predicted using TWI in combination with estimates of total soil depth from available soil landscape maps. Values from this prediction range from 0 to 5 m. The output and discussion are provided in Section 4.9.


26

Available Water Capacity From the prediction of soil depth, the available water holding capacity for the soil profile can be gauged using estimates of water holding capacity in the A and B horizons. This is a function of variables such as percentages of sand or clay and other physio-chemical properties of the soil. A second prediction of available water holding capacity to a depth of 2 m was also produced for the Honeysuckle catchment to indicate the water available to annual plants that do not grow roots to any great depths. This was based on the profile available water capacity and a scaling factor that reflects root distribution. The output and discussion are provided in Section 4.9.


27

3. Geologic and Hydrogeologic Framework

3.1. Geology The oldest rocks in the study area are Silurian-Early Devonian in age (440 - 350 million years) and comprise a thick sequence of marine sediments. These ancient claystone, siltstone and subordinate sandstone deposits were folded, faulted and uplifted soon after deposition and now make up the highly fractured Caniambo Hills, outliers of the Eastern Victorian Uplands. In the later Devonian extensive igneous activity resulted in intrusion of the Strathbogie Granite pluton and emplacement of the Violet Town Volcanics which are crystal-rich rhyodacitic ignimbrites. These crystalline rocks now form the Strathbogie Ranges in the south of the Honeysuckle study area. The ancient sediments and the igneous rocks, together, make up the Melbourne Zone in terms of the geologic framework of Victoria. The province is bounded to the east and north by a major regional suture, the arcuate Governor Fault (Vandenberg et al., 2000), illustrated in Figure 42(c). The course of the Broken River may have been influenced by this fault line which is bounded to the north by the Cambrian greenstones and basalts of the Dookie Hills, some of the oldest rocks in Victoria. The continent was extensively glaciated during the Permian (280 - 225 million years ago). Much of the landscape morphology seen today is a legacy of both this early widespread glacial activity and an ensuing prolonged period of tectonic stability and passive erosion. It is plausible that some of the basal sediments overlying scoured bedrock − particularly in structural troughs adjacent to highlands − are glaciogenic detritus (Hughes, 2002). These sediments are expected to be difficult to distinguish from clay-rich Tertiary alluvium in the absence of glacial striations on nearby bedrock or of Permian plant fossils. Rifting between Australia and Antarctica with the break-up of the Gondwana supercontinent around 100 million years ago resulted in uplift of the Victorian highlands and rejuvenation of rivers and streams. At some stage during or after this period of tectonism, down-faulting occurred to form the Honeysuckle Graben which underlies the present-day Violet Town Plain. This structural trough, also referred to as the ‘Violet Town Sump’ (Dent, 2002), is located between the Strathbogie Ranges and the Caniambo Hills. Down-faulting of the graben may have occurred much earlier, in which case some of the clay-rich infill may be of glacial origin dating back to the Permian. Wet climatic conditions through most of the Tertiary (the last 65 million years) promoted very deep weathering of the landscape. In the case of the fine-grained Silurian sediments of the Caniambo Hills, intensively weathered bedrock has become a thick kaolinised horizon that extends to depths of 40-70 m in places. During this period too, the Murray Darling Basin formed to the adjacent north of the uplands. Victorian rivers initially flowed northwards from the uplifted ranges then subsequently westwards as the basin depocenter shifted to the west with accelerated drifting of the Australian continent away from Antarctica.


28

Sedimentation of the Murray Basin was via fluvial action, with abundant alluvium sourced from the adjacent emerging uplands. The Calivil Formation of coarse river sediments infills the ‘deep lead’ channels that underlie the main north-flowing rivers of Victoria. This includes the buried Goulburn River deep lead immediately west of the Honeysuckle study area. From the Late Tertiary to the Pleistocene, renewed uplift of the highlands resulted in rejuvenation of steams and deposition of the thick Shepparton Formation. This younger terrestrial alluvial blanket of fluvial and lacustrine sediments overlies Early Tertiary basin sediments and onlaps the northern slopes of Palaeozoic bedrock to form the Riverine Plain. Contemporaneous sedimentation filled the Honeysuckle Graben to its brimfull capacity to form the relatively flat Violet Town Plain between the Caniambo Hills and the Strathbogie Ranges.

3.2. Geomorphology The Honeysuckle study area straddles the Central Victorian Highlands and the adjacent Riverine Plain of the southern edge of the Murray-Darling Basin. The physiography reflects the multiple periods of uplift, noted above, and successive periods of erosion and valley incision and eventual backfilling of the main drainage channels. The major landscape units in the study area are the:

• uplifted and intensely jointed volcanic and granite bedrock of the Strathbogie Ranges and flanking colluvial aprons. Mt Strathbogie, south of the study area is at 1007 m (AHD). The base of the slopes, at the southern edge of the study area is at 250 m, presenting a vertical relief of several hundred metres.

• uplifted, highly dissected and largely down-worn Caniambo Hills of ancient sediments, where the vertical relief is no more than 30-90 m (rising to just over 200 m AHD). The Caniambo Hills are drained by Honeysuckle, Sheep Pen and Irish creeks.

• Violet Town Plain located between the Strathbogie Ranges and the Caniambo Hills, contiguous with the Euroa Plain to the west.

• Riverine Plain of the southern edge of the Murray-Darling Basin, dominated by floodplains of the Broken River and lower reaches of Honeysuckle and Seven creeks. The far northwestern corner of the study area, near Kialla East, is at an elevation of 120 m (AHD).

The geomorphology and underlying depositional systems of the Riverine Plain are described in detail by Butler (1958).

3.3. Groundwater Flow Systems The Groundwater Flow Systems (GFS) in the study area are shown in Figure 17. These comprise:

• Local and intermediate GFS in weathered sediments in the Caniambo Hills; • Local and intermediate GFS in colluvial and alluvial fans in the volcanic

bedrock of the Strathbogie Ranges;


29

• Local and fractured GFS in sedimentary and metamorphic bedrock in the Caniambo Hills;

• Local upland alluvial GFS in backfilled valleys in the Strathbogies Ranges; • Local weathered granite GFS in the Strathbogie Ranges south of Euroa; • Regional GFS of the riverine plain of the Violet Town Plain (between the

Strathbogie Ranges and the Caniambo Hills) and at the southern edge of the Murray-Darling Basin onlapping the northern slopes of the Caniambo Hills. The regional alluvial GFS of the Riverine Plain comprise aquifers in the largely Quaternary fluvial deposits that make up the Shepparton Formation.

The Shepparton Formation of the Regional Riverine Plain GFS extends to depths of several tens of metres in the northern plains. The Formation comprises Quaternary fluvial and lacustrine sediments deposited in a low hydraulic gradient for the last 1-2 million years. In the Riverine Plain in the northern part of the study area, particularly in the Caniambo-Kialla East plains, the Shepparton Formation contains sinuous shoestring sand deposits that were deposited in prior streams. These relict stream courses are now disaggregated levees and underlying sand lenses, 0-6 m thick (Day and Harvey, 1994). They are important in terms of the hydrodynamics of the area, particularly groundwater recharge (Section 4.7.2). Sediments along the Broken River represent the Coonambidgal Formation channel sands inset into older sediments, plus relatively recent overbank flood deposits. The Broken River is actively incising the uppermost Shepparton Formation and the Coonambidgal Formation. The incised nature of the present-day trunk drainages is, in places, significant with respect to discharge of shallow saline groundwater directly to streams, both here (Section 4.7.2) and elsewhere in the Riverine Plain, e.g., Billabong Creek, NSW (English et al., 2002). An important aquifer within the region is the Tertiary Calivil ‘deep lead’ underlying the Goulburn River to the immediate west of the survey area. Our Heartlands work in Sheep Pen Creek sub-catchment included detailed analysis of watertables in the fractured bedrock aquifers of the Caniambo Hills and of watertables in the adjacent, onlapping Riverine Plain alluvial aquifers and also scrutiny of all available hydrographs and trends for the network of bores (http://www.clw.csiro.au/heartlands/publications/general_publications.html: pp. 154-156). It was established that there is considerable connectivity between the two main GFS, with a fairly flat watertable continuing from beneath the hills to the adjacent plains. This connectivity between the aquifers holds despite the great contrast in aquifer and GFS properties and a contrast in some of the behavioural patterns over time. Accordingly, our scientific understanding of the hydrogeology of the Honeysuckle catchment − and management implications stemming from this understanding − cannot view the GFS as highly partitioned, but rather, as a fairly well-interconnected system overall. Active saline areas, mapped by DPI (Allan, 1994), are shown as red outlines in Figure 17. These groundwater discharge zones are seen to occur near the contact of distinct GFS, particularly where there is a topographic break, where plains meet slopes. Numerous saline areas occur at the break-of-slope where local upland alluvial aprons abut the volcanic rocks and associated colluvium of the Strathbogie Ranges


30

(Figure 17). Noteworthy outliers of the latter GFS, associated with the Violet Town Volcanics, occur along the Hume Freeway east of Violet Town (Figure 17). These outcrops have bearing on the hydrology of the Baddaginnie area near the Freeway, as described in Section 4.4, although salinity is less of a concern here. Further north, the main saline areas are associated with discharge of shallow groundwater directly to creeklines and swamps, particularly where there are topographic/bedrock constrictions (Figures 2 and 17). The various GFS in the Caniambo Hills and the Strathbogie Ranges are amongst the highest salt generators in both the Goulburn-Broken region and other upland catchments at the edges of the Murray-Darling Basin. Estimated salt export based on Rural Water Commission of Victoria measurements (RWC, 1988) and broad-scale modelling outputs based on saltfall, catchment water yield and in-valley EC indicate salt generation in the range 20-100 tonne/km2/year in these GFS in Honeysuckle Creek catchment (Dowling et al., 2004).

Figure 17. Groundwater Flow Systems (GFS) of the Honeysuckle Creek study area (CLPR, 2002). Mapped saline discharge areas are shown as red outlines.


31

3.4. Soils The distribution of soil types in the study area is shown in Figure 18. Clearly, geology and landscape position govern the distribution of soil types.

• Sodosols (‘texture contrast soils’) dominate the study area, particularly on the broad alluvial plains and in subsidiary valleys of alluvium derived from weathered sedimentary rock in the Caniambo Hills. A characteristic of Sodosols is the propensity for tunnel erosion, as noted for the area by Downes (1949). This type of erosion (Figure 19) is a result of disturbance of the hydrologic balance through removal of vegetative cover and overgrazing and occurs particularly on hardset soil surfaces where there are stump holes and cracks that concentrate runoff.

• Rudosols (‘lithosols’) occur on eroded hilltops of sedimentary rocks in the

Caniambo Hills and at the edges of the Violet Town Volcanics; they have negligible pedological development and differ little from their parent rock types.

• Kandosols and Kurosols have formed on the Violet Town Volcanic rocks and

associated colluvium/alluvium in the Strathbogie Ranges. • Chromosols are widespread on weathered sedimentary bedrock of the

Caniambo Hills and also on outliers of the sedimentary country rock exposed at the base of the Strathbogie Ranges.

• Vertosols characterise low-lying swampy areas and are locally known as

‘Upotipotpon Clay’. These clayey (>35% clay, by definition), commonly waterlogged soils grade to Calcarosols in the lower reach of Sheep Pen Creek. An example of mottled ‘Upotipotpon Clay’ drilled in a waterlogged area in Caniambo is shown in Figure 20.

• Scattered Ferrosols are associated with iron oxide and kaolinitic clay sourced

originally from differentiated iron-oxide rich layers in outcrops of the ancient sedimentary beds of the Caniambo Hills. In places, “buckshot” gravels (‘ironstone’ or ‘ferruginous nodules’, e.g., Figure 23) have been locally transported by past and recent colluvial and alluvial activity and become incorporated into the soil profile.

Enlarged soil maps of the focus study areas, Miepoll, Baddaginnie and Sheep Pen Creek, are provided in Sections 4.3, 4.4 and 4.5.


32

Figure 18. Soil map of the study area.

Figure 19. Tunnel erosion in shallow Yellow Sodosols on shale bedrock, Caniambo Hills.


33

Figure 20. Grey Vertosols, locally known as ‘Upotipotpon Clay’, in SPC01 drill-hole in a low-lying saline area at Caniambo. The grey chemically reduced clay is the result of prolonged waterlogging in shallow groundwaters and the reddish-brown mottles indicate strongly oxidizing recharge waters percolating from overlying episodic swamp waters. Black manganese oxide patches further suggest influxes of dissolved oxygen down into the groundwater system.


34


35

4. Results and Interpretation

4.1 Airborne Magnetic and Gamma-ray Spectrometry Interpretation

4.1.1. Magnetic Data Magnetic patterns representing the subsurface geology of the study area have previously been interpreted by Dahlhaus et al. (2000), based on regional-scale airborne magnetic data of the Australian Geological Survey Organisation (AGSO, now Geoscience Australia, GA) and the Geological Survey of Victoria (GSV). The higher resolution magnetic map from the Honeysuckle MAGSPEC survey (MDBC data, 2001) has been interpreted by Dent et al. (2002), Dent (2003), Gibson and Wilford (2002), Christensen (2004b), and collaboratively in several multi-agency workshops held during 2002-2003 as part of the MDBC Airborne Geophysics Project. Total Magnetic Intensity (TMI) image The Total Magnetic Intensity (TMI) image from the Honeysuckle airborne survey area is shown in Figure 21. The most distinctive magnetic feature is the dendritic pattern of ‘magnetic channels’ that dominates most of the study area, defined by short wavelength (high frequency) low amplitude, magnetic field anomalies.

Figure 21. Total Magnetic Intensity (TMI) image, 1st vertical derivative.


36

Magnetism in the survey area Magnetic susceptibility measurements taken in the field for the present study combined with lithological information from our drilling and laboratory measurements on the FRRR drill-hole samples (Appendix 5) ascertain that the source of magnetic anomalies in the study area are detrital iron-oxide coated gravels such as illustrated in Figure 22. Such gravels are abundant on the present-day ground surface in the Caniambo Hills. The fragments have originated from differentiated ferruginous zones in the sedimentary bedrock of the Caniambo Hills such as illustrated in Figure 23, and would have accumulated in the landscape as lag deposits as former encompassing clay units were preferentially weathered and eroded away. The magnetic gravels must have been washed downstream periodically from the hills into the channels and been subsequently buried by other alluvial material, mainly clays, also sourced originally from local Silurian bedrock. The magnetic valley fill deposits are likely to be Tertiary-Quaternary in age, transported and deposited during wet periods when stream discharges were high. Field observations suggest that goethitic-hematitic iron-oxides (relatively non-magnetic) in the bedrock have been chemically transformed to a magnetic iron-oxide species, such as maghemite (Fe2O3), at some later stage. The latter may have occurred due to the effects of very long periods of exposure to sunlight and lightening strikes, or heat from bushfires, or biological activity resulting in transformation of the structure of iron and oxygen ions making up the iron-oxide. Magnetic susceptibility measurements here range exponentially from: 0.03 x 10 -3 (SI units) for white claystone, 0.39 x 10 -3 for red iron-oxide (goethitic-hematitic) units, to 5.32 x 10 -3 for black ironstone gravels at the base of the cutting (Figure 23).

Figure 22. Iron-oxide coated gravels (“buckshot” or “ferruginous” gravels) accumulated as lag deposits at the landscape surface in the Caniambo Hills, at the headwaters of Sheep Pen Creek.


37

Figure 23. Outcrop of steeply dipping Silurian shale in a road cutting on the Violet Town - Shepparton Rd, showing pronounced differentiation between white leached and red iron-oxide impregnated units. Black magnetic “buckshot” gravel lag deposits are strewn at the base of the cutting and on the landscape surface. The steep dip of the original sedimentary beds and fracture networks are important with respect to the propensity for the weathered bedrock to facilitate recharge to the underlying groundwater system. The clay-rich composition, in places, promotes salt storage.

Magnetic channels Attenuation of the definition of the ‘magnetic channels’ away from the Caniambo Hills most likely relates to depletion of supply of iron-rich fragments with distance from the main hills from which they have been sourced. This applies to magnetic channels flowing both northwards beneath the edge of the Riverine Plain of the Murray-Darling Basin and southwards beneath the Violet Town Plain from the hill crests. An alternative theory with respect to the observed petering out of the magnetic pattern approaching the northern edge of the study area is considered further below. The trend of the magnetic channels beneath the riverine plain is northerly and the channels originate in the Caniambo Hills, where they are commonly coincident with present-day ephemerally active drainage lines that are tributaries of Honeysuckle and Sheep Pen creeks. Thus, only the buried lower northern reaches of the channels can be termed ‘palaeochannels’ sensu stricto. The magnetic drainage pattern for Sheep Pen Creek for example, Figure 21 and enlarged in Figure 37, shows concordance with present-day drainage lines in the upper tributaries and discordance in the lower reaches of the creek. The magnetic channels commonly coincide with linear trends of various orientations that must relate to faults, joints and other fundamental structural features. In the lower Sheep Pen Creek area, the maghemite gravels are concentrated within the upper 10 m of alluvial sediments (e.g., drill-hole log for SPC02, Appendix 5, also BRS drill-hole HC03, Jones, 2002), deepening to 20-30 m BG further north, near the Broken River (BRS drill-hole HC04), and to 50 m BG beneath


38

Kialla East (BRS drill-hole HC09). The magnetic gravels may make up only about 5% of the buried valley fill which is otherwise dominated by clay. Beneath the riverine plain there is a notable disparity between the general northerly trend of the buried dendritic ‘magnetic channels’ and the present-day westerly drainage direction, namely, Broken River, Sheep Pen Creek and the lower reach of Honeysuckle Creek which converge westward. This disparity may have a tectonic origin, related to activity on the Governor Fault Zone, the major arcuate fault zone to the immediate east and north of the study area, Figure 42(c), that is evident on broader scale regional magnetic data (presented in VandenBerg et al., 2000; Dahlhaus et al., 2000). The Broken River appears to have formerly continued northward from north of Benalla along the present-day course of Broken Creek, which follows the eastern side of the Dookie Hills (see Figure 42c). The present-day lower reach of Broken River that flows westward along with its accompanying broad floodplain appears to be a relatively recent development. The magnetic channels in the study area give the impression that they continue northward well beyond the Broken River (i.e., towards the River Murray). However, the regional airborne magnetic data reveal their attenuation little further north than the Broken River and immediately south of a nearby large magnetic anomaly that represents an extensive subsurface continuation of the greenstones-basalt rocks that outcrop in the Dookie Hills, as interpreted in Figure 42(c). Groundwaters become shallow here, to <5 m BG beneath both the northern and southern sides of Broken River and in the whole broad area westward to Shepparton, suggesting that this large shallow magnetic body is a major influence to regional groundwater flow. This near-surface obstruction may also help explain the apparent attenuation of the northward-trending magnetically-defined palaeochannels as they approach the Broken River as a consequence of tectonic activity on the intervening Govenor Fault involving uplift of the northern fault block (the Dookie Hills) and disruption of the southern fault block. An alternative theory with respect to attenuation of the magnetic signature in the northern part of the survey area relates to changes in sedimentation patterns that may have occurred in response to either changing fluvial activity on the plain and/or the postulated tectonism. Past fluvial activity of the Broken River, once its course was diverted from northward (coincident with the present-day Broken Creek course) to westward may have physically reworked and dispersed the magnetic gravel deposits. Regardless of which scenario occurred, younger alluvium was subsequently deposited on the disrupted landscape of the Broken River floodplain. This younger sedimentary package overlying the attenuated − or disrupted − magnetic channels is potassium-rich, as shown in the radiometrics image (Figure 24) and described in Section 4.1.2. This sedimentary package contrasts with underlying alluvium in terms of conductivity, as indicated in the shallow AEM CDI slices for the area (0-15 m CDIs, Appendix 1). Magnetic patterns around the base of the Strathbogie Ranges In the southeastern corner of the survey area, pronounced high amplitude, short-wavelength magnetic patterns follow the crenulated topographic contours that mark the base of outcropping Violet Town Volcanics of the Strathbogie Ranges. The same magnetic signature is also observed around outlying exposures of the weathered


39

volcanic rock further north along the Hume Freeway near Baddaginnie. Clearly, this is a surficial or very near-surface magnetic signature. It is noted that the Violet Town Volcanic rock itself is non-magnetic although there are residuals of magnetic basalt in the area and also a very large magnetic body at considerable depth in the Earths’ crust beneath the Strathbogie igneous complex (English, 1988). Accordingly, the conspicuous shallow magnetic pattern can not be attributable to concentrations of detrital magnetite (Fe3O4) or maghemite (Fe2O3) eroded from outcrop of the volcanic rock. The formation of magnetic minerals at or near the landscape surface must, therefore, relate to regolith or soil processes. Butler (1992) proposes three alternative low-temperature oxidation processes that can yield maghemite:

a) the formation of maghemite from iron oxides by repeated oxidation-reduction cycles during soil formation;

b) natural burning in the presence of organic matter at temperatures above ~200oC aiding the conversion of paramagnetic Fe-bearing minerals to maghemite;

c) dehydration of lepidocrocite (FeOOH), a common iron-hydroxide weathering product of iron silicates.

The Violet Town Volcanics contain substantial iron-silicates, both biotite and garnet, and iron-oxides weathered from same. Any of the above processes, or a combination of all three over considerable periods of time, may be accountable for the observed magnetism around the base of the ranges and outlying exposures of the weathered bedrock. Conversion of iron oxides to maghemite can be readily induced in the laboratory although the rates of such conversions in the natural environment are not well known. Process (b) may be quite rapid in the case of lightening strikes or bushfires. It is conjectured that if process (a) is found to be dominant in given regions of the Australian regolith, the combination of the distinctive magnetic patterns and topographic position might be utilised diagnostically to remotely identify and rapidly map break-of slope zones that are seasonally or cyclically waterlogged. If it can be constrained that conversion of the iron oxides in particular geologic settings has occurred since European settlement as a consequence of rising watertables after land clearing, the magnetic patterns may, potentially, be useful in mapping broad-scale break-of-slope regions where waterlogging is an issue or potential issue. Palaeochannels as Groundwater Conduits? The northerly trends of the buried magnetic channels beneath the Riverine Plain revealed in the airborne magnetic data and the impression that these trends may serve as conduits for northward groundwater flow directly towards the River Murray are deceptive. The regional geologic and hydrogeologic data do not support the proposition for continued northward flow of groundwater north of the study area. Instead, groundwater flow in the riverine plain aquifers is westerly towards the Goulburn River south of Shepparton, essentially parallel with surface water flow in the lower reaches of the main streams in the study area. This westerly trend is revealed in groundwater elevation contours constructed from existing bore data and new data from the FRRR piezometers (Figure 42, below).


40

Gravel layers, regardless of whether the gravels are magnetic iron oxide or quartzose, are no doubt preferred pathways for lateral groundwater flow within a given locality. They can only be important conduits for throughflow within large distances if there is continuity in the gravel layers. The airborne magnetic image (Figure 21) indicates continuity of the gravels in x-y space, as viewed from the survey aircraft. The drill-hole data, however, tends to reveal a much more haphazard and sparse distribution of the gravels in the z-plane, i.e., in the vertical profiles (Appendix 5; also BRS drill-holes HC03, HC04, HC05). Therefore the magnetic channels may not be as well-integrated as perceived in the airborne image. Where gravels constitute only a very minor percentage of alluvial valley fill, as in the observed Honeysuckle Creek profiles, their role as preferential conduits for groundwater flow can not be major. It is noted that at the edge of the Riverine Plain across the northern part of the study area, watertables are only 2-5 m BG, and gravel layers have been logged at 10-15 m BG in the palaeochannels (e.g., SPC01 and 02, Appendix 5). Therefore, all sediments in the buried channels are saturated with groundwater as are all sediments in the intervening buried interfluves. Groundwater is not spatially restricted to the palaeochannels but forms a fairly flat shallow watertable, regardless of the underlying buried topography and sedimentary packages that make up the edges of the basin. Thus, although the magnetically defined channels are very distinct and noteworthy features of the area, their importance as coherent groundwater conduits should not be over-emphasised, particularly given the westward, not northward, flow of groundwater revealed in Figure 42(a) and the regional-scale obstruction to northward flow portrayed in Figure 42(c). On the southern slopes of the Caniambo Hills the trend of magnetic channels is southerly, towards and/or beneath the Violet Town Plain. These are ephemerally active tributaries that feed the middle reaches of the major perennial east-west flowing creeks that traverse the plain, Honeysuckle and Riggs creeks. This information enables extrapolation of a picture of the evolution of the Honeysuckle Graben being filled to brimfull capacity with sediments derived not only from the eroding Strathbogie Ranges to the south but also from the Caniambo Hills drainage system to the north. Once sedimentation had approached the present level of the Violet Town Plain, the major creeks that rise in the Strathbogie Ranges were able to find exit points through the Caniambo Hills. The role of magnetically-defined channels with respect to groundwater flow may be more important in the southern half of the study area, beneath the southern slopes of the Caniambo Hills, than in the northern plains. This seems particularly the case where magnetic features are strongly linear and likely to correspond with fault zones that, potentially, direct some flow southwards into the Honeysuckle Graben from fractured rock aquifers beneath the hillcrests. Our Heartlands groundwater investigations in Sheep Pen Creek catchment indicate the importance of fault zones in groundwater hydrology in the Caniambo Hills. For example, groundwater flow beneath the upper reach of Sheep Pen Creek was found to be bi-directional, flowing southwesterly to the Honeysuckle Graben, in addition to the dominant northwesterly flow path beneath the middle and lower reaches of the creek flowing to the Riverine Plain (Cresswell et al., 2003).


41

Conclusion The value of the airborne magnetic data lies in its usefulness for interpreting the geology and reconstructing the palaeoenvironmental evolution of the study area, particularly the disposition of fluvial activity in the recent past and some regolith processes. Some of the revealed structures and magnetic features have bearing on present-day hydrologic processes. However, misinterpretation (such as presupposing that groundwater flow beneath the riverine plain must continue northwards via the magnetically-defined palaeochannels beyond the survey area) is likely without a broader perspective and substantiating hydrogeologic data.

4.1.2. Radiometrics The gamma-ray spectrometric (radiometric) image from the Honeysuckle MAGSPEC survey (MDBC data, 2001) is shown in Figure 24. Considerable partitioning in the distribution of the radiogenic isotopes, K, Th and U, in the surface landscape is evident.

Figure 24. Airborne gamma-ray spectrometric image: K% = red, Th = green, U = blue, white = high total gamma count, black = no gamma radiation in K, Th, U radioisotopes (e.g., quartz, surface water bodies or waterlogged soils).

Violet Town and Euroa Plains Relatively fresh alluvium sourced from the Violet Town Volcanics and the Strathbogie Granite is seen to emit high gamma-radiation in multiple channels


42

(white), indicative of relatively limited leaching of mineral constituents in the eroded sediments. In the southwestern part of the study area potassium-rich alluvium is clearly sourced from the Strathbogie Granite near Euroa, relating to detrital potassium feldspar eroded from the granite and transported by fluvial action across the Euroa Plain and the western part of the Violet Town Plain. This distinctive sedimentary package is seen to breach a low point in the Caniambo Hills east of Miepoll to form a low-angle alluvial fan between outliers in the Caniambo Hills from where it spreads out onto the Riverine Plain in the northwestern part of the study area. Further west across the Euroa Plain, mixed alluvium is low gamma-emitting, suggesting that the sediments are relatively old and derived from both the granite and from Silurian bedrock outcrops. Sediment sourced from the far eastern range of the Caniambo Hills blankets much of the Violet Town Plain with alluvium that is relatively Th-rich (green) or K-poor (no pink signal). Across the middle part of the Violet Town Plain, this older alluvium is overlain by high-gamma emitting alluvium that is clearly transported by tributaries of the upper Honeysuckle Creek system that rises in the high rainfall zone of the Strathbogie Ranges. Where Honeysuckle Creek breaches a low point in the Caniambo Hills, the relatively fresh sediment is seen to form a narrow alluvial fan that spreads out northward to the edge of the Broken River floodplain west of Caniambo. A mix of transported sediment, sourced from both the ancient Caniambo Hills bedrock and from the high-gamma emitting Violet Town Volcanics, appears to make up this low-angle alluvial fan. Caniambo Hills Weathered sedimentary rock (‘saprolite’) of the Caniambo Hills is, for the most part, depicted by high Th radiogenic emissions (green). This indicates the relatively high clay content of the saprolite and adsorption of Th by constituent clay minerals and/or the absence of K due to prolonged weathering and leaching. The presence of magnetic gravels ― as described above ― in these local lower parts of the landscape within the hills is also suggested by the radiogenic Th-signature since iron oxides have the propensity to sequester thorium. In contrast, sedimentary bedrock on nearby hilltops has been scoured and retains less of its saprolite layer. The crests are thereby seen to be more potassic (high K gamma emissions) than weathered counterparts of the surrounding slopes and small side valleys that are cut into the bedrock where transported saprolite has now become proximal alluvial deposits. Field observations reveal that there is a positive correlation between high Th and the degree of weathering in the study area overall. The potassium signal and its association with either fresh crystalline bedrock or with local alluvium derived from same − in the case of the Strathbogie Granite and Violet Town Volcanics − is clear-cut and is more directly associated with landscape evolution processes than Th associations which are more inferential. Riverine Plain The broad Broken River floodplain across the northern part of the survey area is characterised by highly radiogenic alluvium, including K and U-bearing minerals. The regional geology indicates that this alluvial blanket is sourced from outside the


43

study area, from granites and volcanic rocks in the highlands to the far southeast in the headwaters of the Broken River catchment. Conclusion With respect to the utility of the radiometric data, this is a surficial data layer, and it is potentially useful for mapping the distribution of rocks, sediments and soil types in the study area. The radiometric dataset does not have direct bearing on salt stores or salinity risk. It does, however, provide excellent discrimination of the main geologic components, distinctive bedrock types and transport pathways of sediments from source areas. Contrasting alluvial packages in down-gradient landscapes are clearly related to respective parent rocks up-catchment. These ‘radiogenic packages’ have bearing on the mineral composition and textures of associated soils and could be combined with both existing soil information and high-resolution DEMs to generate useful soil maps of the area. This approach, in turn, may contribute to our understanding of the distribution of potential salt stores but only with respect to the surface environment whereas, commonly, deeper hidden salt stores present the main salinity hazard. Whether the distinctive “radiogenic packages” revealed in the radiometrics image (e.g., high-K floodplains, high-Th weathered Silurian bedrock) can be related directly to the productivity of given soil types that are spatially associated with these remotely sensed gamma-emissions is not known. Likewise, the propensity for the radiometric dataset, with its clear partitioning of the surface landscape, for guiding the planting of specific tree or crop species is not ascertained but would present an avenue for future research.

4.2 Airborne Electromagnetic (AEM) data & target area selection Preliminary interpretations of the Honeysuckle AEM data were made by Dent et al. (2002) on the original 2001 datasets. The uncalibrated 2001 CDI slices indicated widespread conductive materials located near the ground surface in the Violet Town Plain and within the northwestern quadrant of the survey area, such as illustrated in Figure 8(a). These first interpretations suggested that hitherto unknown salinity might be associated with the large patterns of high conductivity (Dent et al., 2002). ‘Violet Town Sump’ Our attention for the FRRR project was initially drawn to the Violet Town Plain and the underlying Honeysuckle Graben, or ‘Violet Town Sump’ in the early stage of our research, based on the uncalibrated data and initial interpretations of Dent et al (2002). Using processed AEM data the Violet Town Sump is represented in the ‘Elevation of the base of the conductive unit’ in Figure 25. In this derivative image, white areas represent areas of no conductive material, i.e., below a designated conductivity threshold of 150 mS/m. These areas are typically outcrops of relatively fresh bedrock or well-flushed sediments. Subcropping bedrock and well-flushed areas are represented by relatively thin conductive cover (red-yellow areas). Deeper depocentres of conductive sediments and/or thick layers of weathered bedrock are


44

blue in the image. This processed AEM data is also shown in Figure 48(b), where it is portrayed as the thickness of the regolith (in m BG). The processed AEM data provides a perspective of the configuration of the ‘Violet Town Sump’ at depth, although the DEM (Figure 2) and GFS map (Figure 17) better illustrate linear features that no doubt strongly dominate the architecture of the graben. It is noted that Figure 25 aids conceptualisation of the sub-surface, based on the distribution of conductive to non-conductive materials alone. The fact that the AEM cannot distinguish fresh bedrock and well-flushed sediments is significant because the two non-conductive units play completely different roles in terms of hydrogeology. The latter concern is addressed below.

Figure 25. Elevation of the base of the conductive unit (m AHD) derived from inversion of the original TEMPEST AEM data (using a conductivity threshold of 150 mS/m, see text). Known salinity outbreaks are shown as red outlines. This derivative AEM data is reproduced in Figure 48(b) where it is conveyed as the thickness of the regolith.

Groundwater trends within the sump Existing relatively sparse bore data for the Violet Town Plain area were interrogated at the outset for the present research. Bore 142 located north of Violet Town and on the northern side of Honeysuckle Creek, and Bore 144 further northeast, adjacent to Jubilee Swamp (locations shown in Figure 9), presented interesting groundwater trends as shown in Figure 26. These two bores were constructed in deep fractured bedrock (>70 m) underlying very thick, heavy clay deposits and deeply weathered material at the base of the graben. The position of Bore 144 is marginal, within the graben but located at the southern edge of low outliers of the Caniambo Hills. The hydrographs reveal steadily rising groundwater for the whole eleven year observation period, 1992 to the present. The trends do not reflect rainfall patterns and have continued to rise during the recent 2-3 year drought of 2001-2003. Groundwater salinity (EC) has also increased progressively in the two bores with a doubling of EC seen in recent years (Figure 26). This salinity trend suggests mixing with more saline groundwater, the provenance of which is not known (see Section 4.7.1).


45

Although rising at a steady rate of 15-18 cm/year, the current depth to the watertable is considerable: at 23 m below ground (BG) in Bore 142, and 30 m BG in Bore 144. At the present rates of rise, the watertables would not be expected to intersect with the ground surface for another 100-200 years. In fact, based on the evidence garnered for the present study (drill-hole FRRR09, described below) rising groundwater may ultimately exit the sump westward to aquifers beneath the Euroa Plain and flow towards the Goulburn River, rather than continue to rise much higher than present levels.

0 500 1000 1500 2000 2500 3000 3500 4000 4500

-24.4

-24.2

-24

-23.8

-23.6

-23.4

-23.2

-23

5

6

7

8

9

10

11

12

13

1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004

Th

ou

san

ds

Days since 1/1/1992

Met

res

belo

w g

roun

d su

rfac

e

Ele

ctric

al C

ondu

ctiv

ity (

uS/c

m)

Depth to water EC

142

0 500 1000 1500 2000 2500 3000 3500 4000 4500

-31.8

-31.6

-31.4

-31.2

-31

-30.8

-30.6

-30.4

-30.2

5

7

9

11

13

15

17

19

21

1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004

Th

ou

san

ds

Days since 1/1/1992

Met

res

belo

w g

roun

d su

rfac

e

Ele

ctric

al C

ondu

ctiv

ity (

uS/c

m)

Depth to water EC

144

Figure 26. Watertable and groundwater EC trends since 1992 for bores 142 and 144 in the eastern part of Violet Town Plain. Bore locations shown in Figure 9.

This observation, that the Violet Town Plain does not appear to be in any immediate danger of being overtaken by rising saline groundwater, coupled with the availability in 2003 of the reprocessed AEM data, shifted our focus away from the central part of


46

the Violet Town Plain for the FRRR research effort. This shift of focus does not ignore the fact that there are environmental, agronomic and natural resource management issues needing to be addressed in the Violet Town Plain. Rather, in terms of the objectives of the present study, the airborne geophysics directed our attention to some adjacent areas where explanations for observed features were demanded.

Honeysuckle Creek Survey CDIs and target area selection The full array of calibrated reprocessed AEM CDI slices for the 0 – 100 m depth range are presented in Appendix 1, and the 0-5 m CDI image is enlarged in Figure 8(b). Moderate conductivity beneath the Violet Town Plain in approximately the 10 to 40 m CDIs is unresolved but does not appear to relate to saline groundwater except perhaps below 20 m depth BG. This moderate conductivity may relate to ancient swamp clays in the thick sedimentary pile within the graben, inferred from the geomorphic setting. The presence of thick periglacial clay deposits accumulated in the graben, relict from Permian glaciation, is also conjectured. It is possible that such layers are preferentially storing salts that once accumulated in the root zones of former forests over many thousands of years of transpiration processes, with such salts having been gradually flushed down to deeper clay layers within the sump. The legacies of earlier long-term transpiration processes, rather than evaporation processes − given that saline discharge sites are not known in the axis of the Violet Town Plain − may have provided a major source of salt for groundwater within the sump for prolonged periods (the groundwater salinities for the area are documented in Section 4.6.1). The reprocessed CDIs revealed two areas adjacent to the Violet Town Plain that warranted attention with respect to conductivity anomalies: the Miepoll area, in the western part of the Caniambo Hills, and the Baddaginnie area in the southeastern part of the survey area, adjacent to the Hume Freeway. In the case of the Baddaginnie area, the conductivity highs of concern are at 5 – 20 m depth, whereas the Miepoll anomalies are observed from the surface down to 25 m depth. Enlarged AEM images for all CDI slices of these two target areas, Miepoll and Baddaginnie, are provided in Appendix 2 and 3. The release of the reprocessed AEM data in 2003 enabled us to additionally utilise the AEM imagery during Phase II of our Heartlands commitment in Sheep Pen Creek sub-catchment. Sheep Pen Creek sub-catchment had previously been designated the highest priority of 53 sub-catchments in the Goulburn-Broken region in terms of severity of both existing and potential salinity, on the basis of several parameters (Cheng, 1999), as outlined in Section 1 of the present report. Much of our Heartlands work in Sheep Pen Creek sub-catchment preceded the availability of the reprocessed AEM data. The imagery was however ultimately utilised in our 2003 Heartlands Phase II reporting (Cresswell et al., 2003) as well as in the FRRR research project. The Miepoll and Baddaginnie areas, thus, were selected for the main research focus for the FRRR project upon the release of the reprocessed AEM data. Some attention, additional to our main Heartlands effort, was given to Sheep Pen Creek sub-catchment, as noted above, and the Warrenbayne-Boho break-of-slope areas as part of the FRRR project to link our research in the heretofore uninvestigated areas − Miepoll and Baddaginnie − with better known saline areas in the Honeysuckle


47

Catchment. Specifically, we contribute some new interpretative aspects about Sheep Pen Creek sub-catchment and the Warrenbayne-Boho area in the present report that result from our recent years of commitment in the Honeysuckle catchment overall and which add to the existing information about these relatively well-studied areas.

4.3 Miepoll

4.3.1. Airborne Geophysics and Drill-holes Localised conductivity highs The reprocessed AEM images reveal a series of pronounced conductivity highs in the depth range of 0-25 m (Figure 27 and Appendix 2) north of Miepoll. These conductivity anomalies are coincident with a series of SW-NE trending magnetically-defined channels that correspond with very subtle valleys in weathered bedrock hills in the present-day landscape of the western part of the Caniambo Hills. These hills are bounded to the west by Seven Creeks, to the south by Faithful Creek and, to the north by anabranching channels of Honeysuckle Creek. Drainage here in the target area north of Miepoll is towards the northeast to a series of swamps that meet Honeysuckle Creek in the middle of the Kialla East sub-catchment (labelled in Figure 1). This outlier of Caniambo Hills is designated a Local to Intermediate Weathered Sedimentary GFS which is surrounded by the extensive Regional Riverine Plain GFS.

Figure 27. Miepoll AEM 0-5 m depth slice (calibrated, reprocessed) overlain on the airborne magnetic image. Red = high bulk conductivity; Blue = low bulk conductivity (resistive) substrate. See also Appendix 2 for CDI slices for the Miepoll area, 0-100 m.


48

This bedrock area is clearly distinguished in the radiometric image (Figure 24) by a strong Th signal. This is indicative of a pronounced degree of weathering and advanced leaching of any former potassium, and sequestering of thorium by clays and iron oxides. The very dark tone of the radiometric signal for the bedrock hills here may also relate to the quartzitic, or secondary silicified, nature of the bedrock since SiO2 emits no gamma rays (black in the radiometric image) These exposed weathered hills of highly weathered Silurian shale form an ‘island’ that is surrounded by surface material that is a low conductor in the AEM imagery and high in potassium in the radiometric image. The latter corresponds with relatively ‘fresh’ alluvium sourced from the Strathbogie Granite to the south. Drill-holes and Salt stores Nine drill-holes were sunk in the Miepoll area, FRRR01 - 09, to target both AEM highs and lows and the phenomenon of ‘fresh’ alluvium, as revealed by high-gamma radiation and low conductivity patterns that blanket high conductivity units. Drilling logs and accompanying field and laboratory analyses are presented in Appendix 5. Estimated salt stores and hydraulic conductivities are given in Tables 1 and 2, below (analytical procedures described in Section 2.2.2). The latter were aimed, respectively, at providing some quantitative data to correlate with the remotely-sensed AEM data and to link groundwater dynamics with observed static features. Within the Miepoll target area, piezometers were installed in four drill-holes: FRRR02, 03, 06 and 09. Those holes where groundwater was not intersected − or which are unlikely to ever become aquifers − were backfilled. Drill-holes FRRR01-03 targeted very high AEM anomalies (0 to >25 m CDIs) in subtle valleys north of Miepoll. The extant valleys coincide with ‘magnetically-defined’ channels in the airborne magnetic data. FRRR01 and 02 intersected weathered, fractured bedrock at 18.5 – 19.5 m depth, at the base of the repective gentle valleys. The drilling confirmed that the conductive material is dominated by locally-sourced alluvial clay and underlying saprolite and that there is a strong correspondence between high down-hole conductivity in the EM-39 profile and high EC1:5, substantiating that salt is the main conductor. Low conductivity and low EC1:5 measurements are associated with the magnetic gravels intersected during drilling. Given the high correlation between down-hole conductivity measurements and actual salt concentrations, the clay is not the main conductor but rather, is a highly appropriate material to sequester and retain salt stores. Kaolinite tends to sequester sodium chloride because of its high cation exchange capacity, so salt storage is fostered because of chemical bonds as well as because of the physical capacities of fracture networks and poorly consolidated saprolite. High salt stores of over 5 kg/m3 or 600-1100 tonne/ha are represented (Table 1). Hydraulic conductivity in the clays is low, 2-7 m/year (Table 2). FRRR01, drilled to 20 m BG was dry. FRRR02, drilled to 20 m depth intersected groundwater which settled to a SWL at around 11 m BG, EC 26700 µS/cm. FRRR03, drilled to 13.5 m depth intersected a clayey gravel aquifer at the base; the SWL is at around 9 m BG, EC 26800 µS/cm. There were no pre-existing bore holes in this outlier of the Caniambo Hills and our FRRR01-03 drill-holes specifically targeted the alluvial clay valleys in the bedrock, rather than the bedrock per se although this was intersected at the base of the valley fill. The weathered, fractured bedrock itself contains considerable salt stores (e.g.,


49

FRRR01 below 7 m depth BG, EC1:5 ~ 1000 mS/m) and, at depth, highly saline groundwater, similar to that encountered in Sheep Pen Creek catchment. The saline groundwater is pervasive throughout the Caniambo Hills, although salt stores in the regolith are expected to be highly variable. Non-detection in the AEM of pervasive saline groundwater and, where present, high salt stores in the regolith associated with the fractured bedrock GFS, is discussed further below. It is important to keep in perspective the fact that concentrations of detrital magnetic gravels provide spatial definition of the channels that also happen to be the local depocentres for alluvial clay deposits which, in turn, attract and retain salt. The gravels themselves are not necessarily significant in terms of salt storage. Elsewhere, clay deposits in the magnetically-defined channels are no more important in terms of salt stores and saline groundwater than intervening interfluves, as in the lowlands of Sheep Pen Creek catchment (Section 4.5.1, below). Further east, drill-holes FRRR04 and 05 were aimed at investigating discordance between AEM and EM-34 measurements. The lack of accord can most likely be attributed to the contrasting ‘footprint’ size, 40-100 m2 for the AEM and 10 m for the EM-34 measurements. Drill-hole FRRR04 comprised saprolitic shale containing moderately high salt stores in the 7 m thick weathered zone, >400 tonne/ha (Table 1). FRRR05 represents a mix of local clays from nearby weathered shale and more distally sourced alluvium from the Violet Town Volcanics, which contains much less stored salt than FRRR04 (Table 1). Both these shallow drill-holes were dry. Drill-hole FRRR06 established that the layer of ‘fresh’ alluvium down-catchment from nick points in the Caniambo Hills (low conductivity in the AEM; high K in the radiometric image) is <5 m thick and overlies much more conductive material, most likely clays sourced from weathered shale of the Caniambo Hills. The drill-hole ascertained three diagnostic factors: low EM-39 profile for the upper few meters, correspondingly low EC1:5 and low salt stores (Table 1), and high gravel content in these upper few metres (FRRR06, Appendix 5). Hydraulic conductivity in the older alluvial clay aquifer material is negligible (Table 2). The CDI shows the underlying highly conductive material to be present somewhere above 10 m depth (Appendix 2), indicating reasonable accord between the AEM and the drill-hole data. FRRR06 intersected groundwater which settled to a SWL at around 6 m BG, EC 19300 µS/cm. FRRR07 was drilled to ascertain the nature of a typical site where the airborne geophysics registered low AEM conductivity and high magnetic signatures. The low AEM conductivity relates to relatively fresh shale bedrock and the absence of both saprolite and groundwater, and the high magnetism relates to surficial ironstone gravels at the head of a subtle ‘magnetically defined channel’. FRRR08 was drilled to determine the nature of a typical conductivity high where both AEM and ground EM-34 measurements are in accord. The substrate comprises dark red Sodosols developed on colluvial clay at the base of a hillslope and contains a higher than average salt store. Both these shallow drill-holes, FRRR07 and 08, were dry. Hydrologic closure of the ‘Violet Town Sump’ – buried bedrock high? FRRR09 was drilled to ascertain the nature of resistive substrate at the western edge of the Violet Town Plain immediately south of Miepoll as represented by the ‘depth to the base of the conductor’ portrayal of the AEM data, Figure 25. The image


50

indicates shallow conductive cover in the area. Two alternative explanations presented themselves: subsurface closure of the western end of the sump by buried resistive bedrock or, the presence of a thick sequence of well-flushed, poorly conductive sediment in the subsurface. In the case of the former alternative, such a buried bedrock high, if present, may represent hydrologic closure of the sump, and future implications if groundwaters continued to steadily rise. The log for FRRR09 (Appendix 5) shows alluvial clays grading to sandy clays below 9 m and gravel below 13 m to the bottom of the hole at 20 m. These data substantiate that there is no subsurface basement high south of Miepoll that could represent closure of the sump. In this regard, Figure 25 may be deceptive. The resistive unit is clearly well-flushed alluvium. Relatively low down-hole conductivity and low EC1:5 measurements are shown in the FRRR09 log (Appendix 5) and moderately low salt store estimates are given in Table 1. The confluence of two perennial creeks at this locality, Seven Creeks and Faithful Creek, would account for the well-flushed condition and low salt stores in this alluvial pile. Although the watertable was not intersected in the 20 m drill-hole, FRRR09, a piezometer was installed because of the likelihood that these gravels would becomes a significant aquifer if groundwater tables rise above 20 m BG within the sump to the east. Such an aquifer would be an obvious outlet for groundwaters to exit from the sump to the west. Drilling here has revealed alluvium to be greater than 20 m thick. It is thus evident that the AEM data does not distinguish between resistive bedrock and well-flushed sediments. Saturated and unsaturated zones – limitation with AEM data The most important concern with respect to the AEM data and the three Miepoll drill-holes, FRRR01-03, is that it is not possible to differentiate whether high bulk conductivity at these sites relates to the saturated zone, i.e., saline groundwater, or to the unsaturated zone, without drill-hole data. FRRR01 was found to be dry, whereas in FRRR02 and 03, saline groundwater was intersected below 10 m BG. The AEM patterns and CDI conductivity measurements are the same for unsaturated and saturated zones at each site. This factor has bearing on assessing salinity risk from remotely sensed AEM data alone, since salt in the unsaturated zone is mainly only relevant to the immediate soil profile and the potential for these unsaturated zone stores to be leached vertically downwards. In contrast, the presence of saline groundwater at any given site has more serious implications involving lateral hydrologic processes that may deleteriously affect extensive down-gradient areas. Non-detection by the AEM of high salt stores and saline groundwater in bedrock Weathered fractured bedrock in the Miepoll area contains substantial salt stores and, at depth, highly saline groundwater, similar to that encountered in Sheep Pen Creek catchment (Cresswell et al, 2004; and Section 4.5, below). Non-detection of these bedrock salt stores in the AEM imagery may be because of characteristically low water to rock ratios and because the bulk conductivity of the overlying alluvial clay-filled valleys is greater than that of the surrounding bedrock (Section 4.5.1). The failure of the AEM technique to detect saline groundwater in fractured bedrock is discussed by Christensen (2004b). Clearly, there is a stronger correlation between the AEM and the drill data in alluvium than in fractured rocks.


51

This finding is similar to that of Edwards and Webb (2003) in the Kamarooka area near Bendigo, a catchment that is badly affected by salinity. Kamarooka catchment straddles both the basement of Ordovician metasedimentary rocks of the Victorian uplands and overlying Quaternary alluvium of the southern Murray Basin (the Shepparton Formation), very analogous to the Honeysuckle Catchment. The distribution of TEMPEST AEM conductivity was found to have little in common with the distribution of saline groundwater. This was attributed to low inter-granular porosity of the bedrock and the inability of TEMPEST to detect pore water salinity in such material. Onlapping sediments with higher porosity than the bedrock gave higher conductivities generally although the AEM failed do enormous variation in groundwater salinities, 5000 to 20,000 µS/cm EC (Edwards and Webb, 2003). Down-hole EM measurements and salt stores An important conclusion with respect to field and laboratory data is the high correlation in the study area between conductivity, as measured with the EM-39, and actual salt, as measured in the laboratory EC1:5 analyses on down-hole samples (Table 1 and Appendix 5).

Site I.D. Average Total salt down to Salt store salt store store bottom of (tonne/ha) (kg/m3) (tonne/ha) hole down to (metres) 6 metres

FRRR 01 5.7 1128 20 304 FRRR 02 5.5 1086 20 284 FRRR 03 4.9 637 13 199 FRRR 04 5.3 426 8 324 FRRR 05 2.6 193 7.5 164 FRRR 06 4.1 652 15 161 FRRR 07 2.6 158 6 158 FRRR 08 5.5 553 10 316 FRRR 09 2.4 455 20 145 FRRR 10 2.4 450 19 120 FRRR 11 3.8 759 20 208 FRRR 12 2.1 99 5.75 108 SPC 01 6.4 1151 18 332 SPC 02 5.5 638 11.5 313

Table 1. Salt stores estimated from EC1:5 measurements and bulk densities on samples

collected at 0.5 m depth intervals during drilling.

Site Ksat Ksat Ksat Ksat (m/day) (m/year) (mm/day) (mm/hour)

FRRR 02 0.0056 2 6 0.23 FRRR 03 0.018 7 18 0.75 FRRR 06 0.00004 0.015 0.04 0.002 FRRR 10 0.088 32 88 3.66 SPC 01 0.408 149 408 17 SPC 02 0.009 3 9 0.38

Table 2. Hydraulic conductivities for selected bore holes, based on recovery rates following bailing and relating K to time.


52

This result concurs with the work of R. Cresswell (2002) in the BRS calibration drilling program in the Honeysuckle catchment where down-hole conductivity was found to be a good proxy for salt. This correlation may not always hold for all regions, however, it may reasonably be anticipated where salt stores are high and overwhelmingly dominate the aggregate conductivity of a given substrate.

4.3.2. Soil map The soil map for the Miepoll area is provided in Figure 28. Vertosols and Sodosols dominate with Chromosols corresponding closely to outcropping Silurian bedrock of the Caniambo Hills. There is a fairly good correspondence between the patterns of soil distribution and the patterns of gamma-radiation packages in the landscape. In particular, the alluvial plain sediments across the northern part of the area −the Broken River and lower Honeysuckle Creek floodplains, mapped as Brown Sodosols − correspond well with high-K in the radiometric image. High AEM conductivity in the 0-5 m CDI slice in the hills north of Miepoll correspond with the distribution of Grey Sodosols, although the distribution of Sodosols is far more widespread than the AEM highs. An area of Ferrosols near FRRR07 corresponds with observed ironstone lag on weathered bedrock and with low AEM conductivity and a high-Th radiometric signature in respective airborne geophysics datasets.

Figure 28. Miepoll area soil map draped over a greyscale DEM.


53

Figure 29. Miepoll MrVBF plot.

4.3.3. MrVBF The MrVBF plot for the Miepoll area is shown in Figure 29. This partitioning of the landscape shows the Violet Town Plain and the Riverine Plain as the lowest flattest topographic class. The high AEM conductivity in the 0-5 m CDI slice corresponds with the intermediate topographic classes (5.5 to 3.5 MrVBF index).

4.4 Baddaginnie

4.4.1. Airborne Geophysics and Drill-holes AEM highs and drilling rationale The reprocessed AEM images for the Baddaginnie to Violet Town area are shown in Appendix 3 (0-100 m CDIs). The targeting of this area as part of the FRRR research was based on high AEM anomalies in the ~5-20 m CDIs immediately south of the Hume Freeway. This is well-illustrated in the 10-15 m CDI for the area, Figure 30. The rationale for drilling here was to assess the potential danger to the infrastructure of the Freeway and the Sydney-Melbourne railway line if the remotely sensed anomalies relate to saline groundwater, particularly given that the area is already predisposed to waterlogging. This area is in Woolpress sub-catchment (Figure 1) which is fed by numerous creeks that rise in the Strathbogie Ranges in the Warrenbayne area to the south and which feed Jubilee Swamp to the north, in the eastern part of the Violet Town Plain. Jubilee Swamp is a very large ephemeral swamp (labelled in Figures 1 and 30; illustrated in Figure 33) northwest of Baddaginnie. Jubilee Swamp is also fed by streams that rise in the easternmost arm of the Caniambo Hills, located northeast of Baddaginnie. An adjunct objective for FRRR work in the Baddaginnie area was to establish whether Jubilee Swamp might be in danger from rising saline groundwater such as has


54

occurred at Downes Swamp (Figure 7) and swamps in the Earlston area (labelled in Figure 1).

Figure 30. Baddaginnie AEM 10-15 m depth slice (calibrated, reprocessed) overlain on the airborne TMI. Red = high bulk conductivity; Blue = low bulk conductivity (resistive) substrate. See also Appendix 3 for all CDI slices for the Baddaginnie area, 0-100 m.

The extensive AEM high dominating the eastern Violet Town Plain (Figure 30 and Appendix 3) relates to near-surface Silurian bedrock and clayey alluvium derived from same. This is corroborated in the radiometric image (Figure 24) which reveals the relative abundance of thorium (green) and weathered material that is leached of potassium. Drill-holes Two drill-holes, FRRR10 and 11 were sunk in locations corresponding to subsurface AEM highs observed in the 5-20 m CDIs (Appendix 3 and Figure 30) immediately south of the Freeway near Baddaginnie. The area is mapped as a Local Upland Alluvial GFS although is closely bounded to the south and north by GFS associated with outliers of the Violet Town Volcanics (Figure 17). FRRR10 is located in a waterlogged area (187 m AHD), illustrated in Figure 31, near Turnip and Stoney creeks which rise in the Strathbogie Ranges some 8 km to the south. The log for FRRR10 shows alluvial clay dominating the 19.5 m drill-hole, with variable down-hole EM-39 conductivity and EC1:5 measurements. Stored salt is around 450 tonne/ha. The watertable is at around 16.5 m BG and the groundwater EC around 13000 µS/cm. Salt stores in the profile are moderate (Table 1) and the saturated hydraulic conductivity is estimated at around 32 m/year (Table 2). A piezometer was installed at the FRRR10 site.


55

Figure 31. Waterlogged area at Stoney Creek near Baddaginnie, immediately south of the Hume Freeway, characterised by wetlands vegetated with sedges. View is southwards. Waterlogging here is caused by natural obstruction to northward surface water flow by bedrock hills and exacerbated by the embankment of the Hume Freeway. Paddocks in the middle distance seasonally have crops on raised beds, further indication of the magnitude of waterlogging in the area. Stream salinity in the creek exceeds 700 µS/cm EC, indicating that some surface salt is being transported from the Boho salinity sites at the break-of-slope to the south via Stoney Creek.

Figure 32. White clay (‘saprolite’) weathered from in situ rhyodacite of the Violet Town Volcanics in FRRR11, where AEM highs are located near Baddaginnie.


56

FRRR11, close to the Hume Freeway, was drilled to 20 m depth with the whole profile composed of dry white clay, representing in situ weathered rhyodacite of the Violet Town Volcanics (Figure 32). The hole did not intersect groundwater and was backfilled. Stored salt is high, at 760 tonne/ha. Waterlogging near Baddaginnie The substantial depth to groundwater in FRRR10, at around 16.5 m BG, and the fact that FRRR11 was dry to 20 m BG, indicates that the AEM highs observed in the 5-20 m CDIs are not attributable to saline groundwater close to the ground surface but relate to the unsaturated zone, i.e., to intensely weathered bedrock and dense alluvial clays. The observed waterlogging is surface water that is slightly saline to fresh, receiving abundant replenishment from the 1000 mm rainfall zone of the cleared Strathbogie Ranges. These surface waters are being forced to pond up-gradient from (i.e., south of) obstructions caused by outcrops of weathered volcanic bedrock. These rock outcrops are exposed in road cuttings along the Hume Freeway. Waterlogging here is no doubt further exacerbated by the embankments of the freeway and the railway line impeding northward flow to the Violet Town Plain. It is noted that waterlogging is apparent along the southern side of the Hume Freeway further to the east as well as in the Baddaginnie area. Excessive clearance of the Strathbogie Ranges to the south evidently results in high runoff and high streamflow in the numerous creeks in this area, and substantial overbank flooding. These excess waters are forced to pond in the area because of the combination of natural and anthropogenic obstructions to efficient overland flow towards Jubliee Swamp and the main creeks that flow through the Caniambo Hills to the Riverine Plain. The natural bedrock obstructions may also impede groundwater flow in the area to some degree. Waterlogging south of the Hume Freeway in the Baddaginnie area has a deleterious effect on the agronomic productivity of the area and potentially also on the infrastructure of the freeway and railway line. Notwithstanding, excess water in the landscape in this particular vicinity is not directly attributable to saline groundwater being discharged to the ground surface, unlike the situation at the break-of-slopes sites at Warrenbayne and Boho to the south and Sheep Pen Creek to the north. Jubilee Swamp Jubilee Swamp, Figure 33, in the eastern part of Violet Town Plain is the largest wetland in the survey area, covering (507 ha). It is fed by several converging creeks, Folly Creek being the main one, joined by Woolpress and Turnip creeks. The swamp flows into Stoney Creek which crosses the Violet Town Plain close to the southern edge of the Caniambo Hills, to join Honeysuckle Creek some 12 km to the west, before the latter cuts northward through the hills (Figure 1). Jubilee Swamp is a shallow ephemeral wetland; it is floored with Grey Vertosols. It supports an important remnant vegetation ecosystem dominated by E. camaldulensis (Kelley, 1994). FRRR12 was hand augered through the swamp floor to establish whether a shallow saline watertable is present. This objective was prompted by the occurrence further north of swamps becoming major targets for severe salinisation; examples include Downes Swamp and lowlands in the Earlston sub-catchment.


57

Augering at site FRRR12 extended to 6 m depth BG, the log is given in Appendix 5. The profile comprises alluvial clay with subordinate sand and silt and is mottled below the uppermost couple of metres, indicative of subsurface retention of infiltrated swamp waters, at least on a seasonal or periodic basis. No groundwater was intersected in the 6 m interval. The EM-39 profile shows progressively increasing conductivity with depth. The EC1:5 profile reveals a pronounced salt spike at 5 m BG. This is taken to represent root zone accumulation of salt that has progressively accumulated in the clayey substrate as water has been withdrawn by tree roots. Estimated salt stores are the lowest of those calculated from the study area overall, 2 kg/m3 or 100 tonne/ha. This indicates a relatively well-flushed substrate in spite of the mottling that represents at least some prolonged periods of oxidation and reduction in the presence of standing water and the salt spike that represents a zone from which water has been commonly been extracted by tree roots.

Figure 33. Jubilee Swamp, a large shallow ephemeral swamp that supports a dense woodland of River Red Gum northwest of Baddaginnie in the eastern part of the Violet Town Plain. The swamp was augered as part of the FRRR project to ascertain whether the swamp is at risk of salinisation.

The depth to the base of conductor image, Figure 25, shows thin conductive cover beneath the area. This is analogous to the conductivity profile for the area south of Miepoll, at FRRR09, as revealed in Figure 25, and described above. At Miepoll, the resistive character of the substrate is a consequence of the sedimentary pile being well-flushed beneath the confluence of the main creeks; it does not represent a buried basement high of unweathered bedrock. In the Jubilee Swamp area, the conductive layer is particularly thin at the confluence of Folly, Swamp and Stoney creeks where there is also a cluster of very large dams that correspond with a white area that represents no conductive material (<150 mS/m). The AEM patterns are thus well-correlated with well-flushed substrate although there is no way to discriminate from


58

the CDIs alone whether the conductivity distribution relates to saturated or unsaturated materials.

4.4.2. Soil map The soil map for the Baddaginnie-Violet Town area is shown in Figure 34. The Violet Town Volcanics of the Strathbogie Ranges are associated with Kandosols, including outliers of weathered volcanics adjacent to the Hume Freeway. Colluvial-alluvial fans fringing the ranges are associated with Chromosols and Rudosols. Outliers of weathered Silurian shale of the Caniambo Hills correspond with Chromosols (as in the Miepoll area). Between the Strathbogie Ranges and the Hume Freeway − where drill-holes FRRR10 and 11 are located − Brown Sodosols are mapped. The Violet Town Plain to the north and west is dominated by Vertosols and Sodosols which may correspond with a widespread moderate AEM signature in the 5-10 m CDI (Appendix 3); no saline discharge areas have been mapped in this area. Given that the widespread AEM high beneath the eastern Violet Town Plain continues to at least 15 m BG (Figure 30), a layer of salt accumulated through prior transpiration processes, such as portray by the salt spike in the log for FRRR12 (Appendix 5), may be represented. The broad-scale AEM high in this area is seen to be interrupted where presumably waterways have flushed the underlying substrate to produce resistive patterns in the uppermost CDIs. Bedrock areas to the south and their associated Kandosols and Chromosols correspond well with resistive AEM patterns because these soils types are developed on bedrock.

Figure 34. Baddaginnie area soil map draped over a greyscale DEM.


59

4.4.3. MrVBF The MrVBF plot for the Baddaginnie to Violet Town area is shown in Figure 35 and for the Warrenbayne-Boho area in Figure 36. In Figure 35, the Violet Town Plain corresponds with the lowest, most flat terrain unit (MrVBF Class 5.5-6.5 – blue). Alluvial aprons surrounding the Strathbogie Ranges and the Caniambo Hills are classified as moderate terrain classes in the index (5.5-3.5 – greens). The most noteworthy features are isolated pockets of local low, flat terrain in the piedmont zone that are segregated from the Violet Town Plain by intervening topographic features. Our Baddaginnie drill-holes, FRRR10 and 11, are located in two of these topographic ‘pockets’, immediately south of the Hume Freeway. Both these areas are bounded to the north by bedrock outcrops of weathered volcanic rock and both areas are characterised by nearby swampy low-lying areas. The MrVBF terrain analysis thus corresponds well with observations in the landscape in this area. There is also a measure of correlation between these subtle topographic features depicted by the MrVBF analysis and areas of high AEM conductivity revealed in the shallow CDI slice.

Figure 35. Baddaginnie MrVBF plot.

The significance of these localised topographic features with respect to near-surface hydrology is even more substantial in the Warrenbayne-Boho area, Figure 36. Some of the most severe break-of-slope salinity occurs in the Warrenbayne and Boho areas adjacent to the ranges and these occurrences were amongst the earliest salt outbreaks to emerge in the district, some 70 years ago. It is apparent from the terrain analysis that isolated low flat valley floors are present in both the Warrenbayne and Boho areas (MrVBF class 5.5-4.5 – green areas in Figure 36). These pockets are occluded downslope to the north by enclosing steeper topographic features. Given the high rainfall in the ranges and the particularly high runoff rates that have occurred since the slopes were cleared (Section 1; e.g., Figure 4), such semi-enclosed valleys would have difficulty transmitting excess water northwards down-gradient beyond the


60

obstructions. Excessive recharge of local aquifers, high watertables, waterlogging and break-of-slope salinity ensue, such as illustrated in Figures 5 and 6.

Figure 36. MrVBF plot for the Warrenbayne-Boho area, Strathbogie Ranges.

Clearly, the discharge capacities of isolated sub-valleys and topographic pockets such as these are exceeded and hydrologic imbalance can be substantial in such systems. Local down-gradient topographic/bedrock constrictions limit the amount of excess recharge water that can be transmitted to the broader plains. The propensity applies to both the surface drainage systems and underlying groundwater bodies.

4.5 Sheep Pen Creek

4.5.1 Airborne Geophysics, Drill-holes and Soils Sheep Pen Creek sub-catchment was the focus of CSIRO-MDBC Heartlands work in 2001-2003 which work has been documented by Cresswell et al (2003). Aspects of this work are incorporated in the FRRR project with respect to the objective of interpreting the airborne geophysics data for the Honeysuckle Creek catchment overall. Additionally, our drilling data and laboratory analyses for Sheep Pen Creek are included here because these detailed data were not documented in the Phase II Heartlands report and represent our main work in the Riverine Plain GFS for the study area for the present research phase. Figure 37 shows two CDI slices of the AEM survey data for the Sheep Pen Creek sub-catchment: 5-10 m and 15-20 m CDIs, overlain on the airborne magnetic data. These particular depth slices are selected to illustrate the range of conductivities with respect to the bedrock and alluvial geology of the sub-catchment and its known hydrogeology (Cresswell et al, 2003).


61

Figure 37. Sheep Pen Creek AEM imagery: (a) 5-10 m depth slice; (b) 15-20 m depth slice, overlain on the airborne magnetic image. Red = high bulk conductivity; Blue = low bulk conductivity (resistive) substrate.


62

Logs for the two drill-holes, SPC01 and SPC02, are given in Appendix 5. The conductivity highs in the catchment lowlands relate to the aggregate effects of saline groundwater, high clay content and high salt stores in the unsaturated zone. Clay strata in the magnetically-defined channels, in particular, show high bulk conductivities. Salt stores estimated from EC1:5 measurements in SPC01 and 02 are high: 5.5-6.5 kg/m3 or 640-1150 tonne/ha (Table 1). Variable hydraulic conductivities for the two drill-holes given in Table 2 suggest that the location of SPC01 is within a higher than normal zone of hydraulic conductivity (Ksat 150 m/year) possibly because of the presence of channel gravels (at 10-17 m BG, SPC01, Appendix 5). A fault or fracture-line in the adjacent bedrock system here may also be functioning as a preferred pathway for groundwater flow to the alluvial aquifer (such as the structure illustrated in Figure 42c). In contrast, the recovery rate and low conductivity of SPC02 (Ksat 3 m/year), is probably more typical for Grey Vertosols (‘Upotipotpon Clay’), Figure 38, both here and where they occur elsewhere. It should be emphasised that the concordance between magnetic channels and high AEM conductivity relates to x-y space, and less so to z, depth. The EM-39 conductivity profile for SPC02 in Appendix 5, for example, shows high conductivity corresponding with clays and lower conductivity with magnetic gravels. Thus, it is not the magnetic gravels (which may make up only a few percent of a profile within the alluvial channels) that are the high conductor. Rather, the high conductivity revealed in the AEM relates to overlying and underlying alluvial clay deposits within the buried channels which happen to be well represented in the magnetic data. The pattern of high conductivity in northward-flowing palaeochannels continues to considerable depth within the sub-catchment, to 50 m (Appendix 1), beneath the Caniambo – Kialla East plains. There is good correspondence between mapped saline discharge areas at the ground surface (red outlines) and the shallow AEM data (Figure 37) and also with the distribution of Grey Vertosols and the base of the hills (Figure 38) in alluvium that onlaps the hills. Non-detection by the AEM of saline groundwater in bedrock The AEM CDIs do not reveal the presence of saline groundwater in the fractured bedrock aquifers of the Caniambo Hills. This applies to the selected depth slices shown in Figure 37 and deeper CDIs, to the full survey depth of 100 m (Appendix 1). Salt stores in the bedrock hills are high, particularly where kaolinitic clays are involved, either in situ saprolite or locally transported alluvium sourced from same. Up to 1100 tonne/ha (Table 1) are present, and far greater if the estimates are extrapolated to a depth of 50-60 m down to the deepest watertable depth. In the present study, high salinity extending to 80 m BG was measured. For example, in the upper reach of Sheep Pen Creek, Bore 126 (Figure 11; Total Depth 78 m; SWL at 8 m BG, Groundwater EC 25000 µS/cm), EM-39 measurements of 350-500 mS/m were typical for tens of vertical metres. When correlated with EM-39 profiles for FRRR drill-holes for which EC1:5 measurements are available, salt stores of approximately 450-750 tonne/ha are indicated. Substantial variability in salt stores is to be expected given the vast area and the heterogeneity of the Silurian bedrock itself, the fault and fracture networks and the regolith of the Caniambo Hills.


63

Such high salt stores are comparable to other areas of deeply weathered fractured weathered bedrock in Victoria, potentially to 3000 tonne/ha (P. Dyson, written comm., 2004). This implies that beneath parts of the Caniambo Hills, more than 1000 tonne/ha of stored salt could be leached vertically downwards with infiltrating rainfall via the fracture network to the groundwater system. Given that the Caniambo Hills cover one third of the survey area, about 50,000 ha, which, at 1000 tonne/ha salt storage, equates to over 50 million tonnes of unmobilised salt resident in the hills. Additionally, the weathered bedrock continues beyond the area of outcrop, where it is onlapped by alluvial sediments of the adjacent plains, substantially increasing the size of the salt store accommodated in buried bedrock and its regolith blanket. This very high salt store is evident from the groundwater salinity throughout the Caniambo Hills, which typically exceeds half seawater salinity. Investigations over the past 30 years indicate that these high salt levels in the groundwater originate from high salt stores in the unsaturated or previously unsaturated zone (P. Dyson, written comm., 2004). Thus there is generally a relationship between high salt stores in the regolith and high groundwater EC. Groundwater salinities approaching 30000 µS/cm are at odds with the AEM data for the fractured bedrock GFS (Figure 37; Appendices 1 and 2). This suggests that the volumes of water relative to rock matrix in the hills is low and that, in the lowlands, the bulk conductivity of salinised clays and saline groundwaters is greater than the bulk conductivity of lesser quantities of saline groundwater in the bedrock aquifers. High salt stores in deeply weathered terrain − saprolite which provides the perfect medium for salt storage − are almost ubiquitous in southern Australia, particularly in the 500-600 mm temperate rainfall belt. They explain very strong contrasts in salinity between groundwater in fresh fractured rocks, and groundwater in adjacent deeply weathered fractured rocks. They are established through vegetative concentration (transpiration) of salts introduced in rainfall over millennia in a process that involves ion exclusion in the root zone of native vegetation (P. Dyson, written comm., 2004). This consideration is discussed further in Section 4.7.1 with respect to isotopic data. In addition to low water to rock ratios giving a lack of agreement between observed salinity and the AEM data, very low water content in the unsaturated regolith is probably contributing as well, particularly where the groundwater is deep as in the Caniambo Hills. This limitation of the AEM technique to register high salt stores in deeply weathered fractured sedimentary rock requires resolution, particularly with respect to the need to integrate the GFS framework with airborne geophysics (P. Dyson, written comm., 2004). Our groundwater salinity contour map, Figure 43, provides a more accurate representation of the distribution of saline groundwater in the catchment than does the AEM imagery because of the limitations of the latter in fractured bedrock terrain.


64

Figure 38. Sheep Pen Creek soil map draped over a greyscale DEM.

Airborne Geophysics and Soils An association is observed between Grey Vertosols, salinity, and a thorium-rich, or potassium-poor, gamma-radiation signal (Figure 24) in the lower reach of Sheep Pen Creek, in Downes Swamp and the centre of Earlston sub-catchment, all at the edge of the Riverine Plain. The mapping of these sites as ‘Upotipotpon Clay’, the equivalent to Grey Vertosols, was conducted by conventional soil mapping techniques some 50 years ago (Downes, 1949). The Vertosols here lie in topographically low areas, interpreted in the present study to be thick accumulations of palaeoswamp clays, deposited in standing water. The sites are seen as Th-rich, or K-poor, in the radiometric image. Other areas in the catchment, however, contain Grey Vertosols and have high-Th, low-K gamma signatures but are not associated with salinity. In the case of the eastern Violet Town Plain, an extensive such area is elevated >20 m above the watertable. Thus, in the lowlands of Sheep Pen Creek and Earlston sub-catchments, the salinity association relates more to palaeoenvironmental setting and topographic position than to the distribution of radioelements or specific soil types. Surficial data, such as the radiometric image, can belie the full extent, importance and antiquity of salt stores. For example, in the present study, we measured high salinity extending to at least 80 m depth BG in Sheep Pen Creek, noted above. An appreciation of the third dimension and of the evolution of given landscapes needs to accompany surficial datasets such as radiometric and soil maps for interpretative purposes.

4.5.2. MrVBF The MrVBF plot for Sheep Pen Creek and the area to the adjacent north is shown in Figure 39. Significant is the array of topographic constrictions in both the lower reach


65

of Sheep Pen Creek and along Broken River. These constrictions are formed by outliers of weathered fractured bedrock that happen to contain high salt stores. Downes Swamp is completely salinised (Figure 7). Similarly, highly saline groundwater actively discharges into Sheep Pen Creek at the pronounced topographic constriction to the south west. Both these sites are very low and flat according to the MrVBF index (Class 5.5-6.5 – blue), Figure 39. Very high salt stores are present in soils associated with fractured bedrock at Nalinga and the Dookie area, adjacent to Broken River (Dahlhaus, et al., 2000); these soils include Sodosols and Chromosols. The Gowangardie stream gauging station on the Broken River (labelled in Figure 39) has shown a long-term upward trend in stream salinity of 10.6 µS/cm/year. The array of bedrock constrictions along the river is seen in Figure 17 (outliers of fractured metasedimentary outcrops close to the Broken River, green GFS unit in Figure 17). The configuration of constricting bedrock outcrops here is analogous to that documented along the upper reach of Billabong Creek in our Heartlands work in NSW (Cresswell et al., 2003). Groundwater is shallow in both the Broken River floodplain and the analogous floodplain in Billabong Creek where discharge capacities are repeatedly exceeded. Waterlogging is common upstream from each constriction. In some sub-catchments saline groundwater is released directly into the trunk drainage lines as a function of the natural depth of incision of the main channels that may be 12 m below the surrounding landscape surface. Thus, the shallow depth of the floodplain aquifers is unnatural − the result of land clearing and watertable rise in recent decades − whereas the deep incision of the main streams across the Riverine Plain is a natural geologic phenomenon, and groundwater flow into the streams is a consequence of the two physical factors in spite of low hydraulic gradients in the floodplains (English et al., 2002).

Figure 39. Sheep Pen Creek MrVBF plot.


66

The MrVBF plot, used in conjunction with geologic maps and field measurements, is useful for pinpointing where there are juxtapositions between topographic constrictions and bedrock types that contain high salt stores. In such locations, the capacity of the system to transmit groundwater down-gradient may to be limited and an increased potential for groundwater discharge, waterlogging and salinisation can be predicted.


67

4.6 Groundwater and Salinity

4.6.1. Depth to Groundwater, Elevation of the Watertable and Groundwater Salinity The distribution of DPI and SKM managed water bores and our FRRR piezometer sites is shown in Figure 40. Groundwater data from >100 bores were utilised in the investigation. These existing groundwater datasets were not interrogated in the initial interpretations of the Honeysuckle Creek airborne geophysics datasets (Dent et al., 2002; Dent, 2002; and Gibson and Wilford, 2002). Some bores in the area were not incorporated in our analysis because of unreliable data, either because the sites have not been surveyed or measurements for depth to groundwater or EC are not available. Groundwater measurements were not reported from the eleven BRS drill-holes in the survey area (Jones, 2002; Dent 2002) so these drill-holes are not included in Figure 40 (the sites, however, are shown in Figure A5.1). The methodology and parameters used to generate the groundwater contour maps: Depth to Groundwater (Figure 41), Elevation of the Watertable (Figure 42), and Groundwater Salinity (Figure 43), and the inherent limitations with contouring, are outlined in Section 2.2.2. The difference between the Depth to Groundwater BG and the Elevation of the Watertable (AHD) measurements is also explained in Section 2.2.2.

Figure 40. Bore location map for all reliable data points used for groundwater contours. Red dots = bores monitored by DPI or, near the Shepparton Irrigation Area, by Sinclair Knight Merz (SKM); orange dots = CSIRO Land & Water (CLW) drill-holes sunk for the FRRR and Heartlands projects.


68

Depth to Groundwater A map of the depth to water below the ground surface is shown in Figure 41(a) and the corresponding data are projected in Figure 41(b).

Figure 41. Depth to groundwater: (a) contours; (b) projection of the depth to watertable below ground (BG), viewed from the southwest. These projections need to be compared with Figure 42 where the depths are transformed to watertable elevation (m AHD) and the hydraulic gradient is represented.


69

It is emphasised that the Depth to the Watertable measurements (m BG) ignore the landscape elevation with respect to sea level (or, AHD) and local relief. They therefore do not provide an indication of the hydraulic gradient or groundwater flow directions within the catchment; the latter are gained from the derived, or reduced, Elevation of Watertable measurements as shown in Figure 42. The depth measurements range from zero to 50 m (BG). Downes Swamp (Figure 7) in the northeast corner of the study area represents a ‘groundwater window’, where the watertable is exposed (zero depth). Shallow groundwater, 2-5 m BG, is present beneath the Broken River- Kialla East floodplain across the whole northern part of the survey area, including the Shepparton Irrigation Area (SIA) in the northwest corner. Shallow groundwater is also present beneath low-lying parts of Earlston sub-catchment and near the Honeysuckle – Violet Ponds anabranch channels to the nearby west (localities labelled in Figure 1; the corresponding topography is shown in Figure 2). The latter location, at the debouchment point where Honeysuckle Creek exits the hills and begins to anabranch on the adjacent plain, is a likely recharge mound. These two isolated watertables that are <5 m BG and which are alluvial aquifers adjacent to the northern flanks of the Caniambo Hills may represent perched groundwater bodies that are largely hydrologically closed by the surrounding − and possibly underlying − Silurian bedrock. Shallow groundwater of <5 m BG is present at the break-of-slope in the Warrenbayne-Boho area adjacent to the Strathbogie Ranges in the southeastern part of the study area. Here, the capillary fringe of the watertable is, in widespread patches, within approximately 2 m of the ground surface from where capillary rise of saline groundwater in response to solar radiation occurs, resulting in the formation of efflorescent salt at the ground surface and bare salt scalded ground. From the base of the ranges in the south of the survey area there is a steady increase in the depth to groundwater northwards beneath Violet Town Plain. A groundwater depression extending to 40 m depth BG is present beneath the central part of the plain, some 15 m deeper than the surrounding watertable. Given that the overlying plain here is relatively flat, these concentric contours can be viewed as representing a ‘groundwater sink’ (more appropriately portrayed in Figure 42). This feature is located beneath the confluence of Honeysuckle and Riggs creeks west of Violet Town. This suggests the presence of a structural trough that is part of the geologic architecture of the graben, the basal sediments and basement of which serve as a deep aquifer. A similar large, relatively deep groundwater depression is indicated further west, beneath the Euroa Plain. Beneath eastern parts of the Caniambo Hills the depth to groundwater is substantial, 35-50 m BG. Similarly, beneath the topographically highest hills of the Caniambo Hills − that coincide with the divide between Sheep Pen and Earlston sub-catchments − the water table is at >40 depth BG. This is a reflection of the height of the hills rather than the depth to underlying groundwater which forms a relatively flat surface in this area. The greatest depth to groundwater, >50 m, in the eastern part of the study area appears to correspond with a major NW-SE striking fault zone. This interpreted fault zone is followed by the upper reach of Sheep Pen Creek where it is incised into the hills. The fault also bounds the sharply-defined southern edge of the easternmost range of the Caniambo Hills at the eastern end of the Violet Town Plain. Particularly deep penetration of groundwater within the fault zone is indicated here.


70

Elevation of the Watertable Figures 42 (a) and (b), showing the elevation of the watertable, illustrate the actual groundwater surface beneath the study area, ignoring the relationship between the landscape topography and the depth to watertable that tends to restrict interpretation of Figure 41. All water levels plotted in Figure 42 have been surveyed with respect to sea level, or AHD. This shows a representation of the groundwater surface, including its slope and the presence of groundwater mounds and depressions, with all the overlying topography of the hills, valleys and plains peeled away. The hydraulic gradient is steepest (close contour spacing) in the area between the Strathbogie Ranges and Violet Town Plain, indicating that groundwater flow is driven by gravity. Clearly, the ranges are a major recharge zone and the adjacent alluvial aquifers are the most hydrologically dynamic parts of the whole Honeysuckle groundwater system, facilitated not only by slope and high runoff but also by higher hydraulic conductivity in colluvial aprons around the ranges. The groundwater depression in the Violet Town Sump, suggested above with respect to Figure 41 (a) and (b), is well represented; at 120 m AHD, 10-20 m deeper than the surrounding groundwater surface. Another groundwater depression is present in the eastern part of the study area, beneath the Baddaginnie-Jubilee Swamp area. The latter suggests a second deep pocket in the Violet Town Sump and, when viewed in conjunction with the linear landforms in Figure 2, indicates the likely presence of fault-bounded structures that may indicate a sub-basin within the graben here, northeast of Jubilee Swamp. It was noted in our Heartlands report on groundwater processes in Sheep Pen Creek sub-catchment (Cresswell, 2003) that, beneath the upper reach of Sheep Pen Creek, groundwater flows towards the southeast to the Honeysuckle Graben whereas beneath the lower reaches of the creek it flows towards the northwest to the Caniambo Plain. Faults appear to govern the hydrogeology in this area in particular. The groundwater depression at 140 m AHD in Figure 42(a), coinciding with this interpreted fault is probably hydrologically closed and isolated from the surrounding groundwater system. A groundwater mound at 140 m AHD in the central west of the study area is a shallow recharge mound that is only around 5 m below the ground surface. The location suggests two possible scenarios: (a) that there is excess water in Honeysuckle Creek at its debouchment point and permeable sediments in the anabranching channels that facilitate recharge to the subjacent alluvial aquifer; (b) that this may be a perched watertable, possibly underlain by buried Silurian bedrock. From all major parts of the landscape, the Caniambo Hills, the Riverine Plain, the Strathbogie Ranges and Violet Town Plain, the regional hydraulic gradient slopes very gently to the west. Thus, groundwater flow from the whole catchment appears to be consistently directed towards the Goulburn River although bore data from the west of the survey area is required to substantiate this interpretation.


71

Figure 42. Elevation of the groundwater surface: (a) contours and groundwater flow directions; (b) projection of the groundwater surface, viewed from the southwest.


72

Figure 42 (c) Magnetic high interpreted from regional airborne magnetic (TMI) image representing both exposed and near-surface mafic rocks, “Cambrian Greenstones” of the Dookie Igneous Complex which is interpreted to influence regional groundwater flow. The magnetic high is interpreted from Geoscience Australia and GSV TMI imagery, reproduced in English, 1988; Vandenberg et al, 2000, Dahlhaus et al., 2000, and Meyers, 2002. Structures (red) after Vandenberg et al, 2000.

It is important to emphasise that in the Riverine Plain groundwater flow is westward, not northward as might be interpreted from the local airborne magnetic image (Figure 21). The dominant northward trend of magnetic channels in the plains does not define the groundwater flow direction. This fact, along with the indicated groundwater flow paths, is reinforced by the regional airborne magnetic image (reproduced in English, 1988; VandenBerg et al., 2000; Dahlhaus et al., 2000) and noted in Section 4.1.1, above. The extensive buried and outcropping magnetic body, the ‘Dookie Igneous Complex’ north of the Broken River and outside the present survey area, must obstruct northward groundwater flow, as indicated in Figure 42(c). This large bedrock body may play a role in both the widespread presence of very shallow watertables in the whole area to the west, and in predisposing the regional flow direction westward towards the Goulburn River near Shepparton. Regardless of groundwater flow directions interpreted from the watertable surface, it should be noted that, for the most part, hydraulic conductivities are variable and


73

generally very low. Hydraulic conductivities estimated from the FRRR and SPC drill-holes range from 0.006 to 0.4 m/day (Table 2) and from four of the BRS drill-holes, the range is from 0.4 to 1.0 m/day (Jones et al., 2002). Flow rates are accordingly interpreted to be slow. Groundwater Salinity (EC) Groundwater salinity measurements (µS/cm EC) are plotted in Figure 43. The contours do not embody vertical variation in groundwater salinity for given points but, rather, horizontal variation within the overall survey area. Such data can sometimes be deceptive when viewed in this format because the measurement of a single freshwater sample − such as rainwater that has percolated into a shallow piezometer − will skew the contouring and not represent the underlying groundwater body. This situation was observed in nested piezometers at Downes Swamp in the northeast of the area where the groundwater EC is known to be very high but is not necessarily detected in some uncapped piezometers. Furthermore, such representation of the distribution of salinities in x-y space belies the possible existence of stacked aquifers of potentially differing salinities in z-space. Notwithstanding, the general trends revealed are consistent with our understanding of the system as a whole. The lowest salinity groundwater occurs in the recharge areas of the Strathbogie Ranges, in the southeast. The inferred salinity gradient from the ranges − where rainfall and runoff must be fresh − to the 6000 µS/cm contour is very steep. This indicates that salt is being picked up in the recharge zone at the break-of-slope saline areas at the base of the ranges. Figure 43 is more representative of the occurrence of salinity here than is the AEM image. Saline groundwater close to the ranges occurs within the uppermost 20 m BG (George, 1984; Cartwright et al, 2004). Deeper aquifers here contain fresh groundwater. George (1984) has noted that the volume of groundwater transported by the shallow aquifer is 200 times greater than that of the deep groundwater system. This reflects that salinity processes operating at the break-of-slope are relatively surficial and dynamic. Some salts from the break-of-slope area are probably being slowly transported northward, down-gradient, via either the piedmont zone shallow aquifer (e.g., FRRR10 groundwater EC 10000 µS/cm; SWL 16 m BG) or surface waters (e.g., Stoney Creek EC 750 µS/cm), possibly as far north as the Violet Town Sump aquifer (EC 12000-19000 µS/cm; Section 4.2; Figure 26). Recycling of salts is likely to be exacerbated during and following drought periods, possibly reflecting the high salt/water ratio in the near-surface environment during these times. Notwithstanding, the areas of break-of-slope salinity are small and, accordingly, may not be major contributors to groundwater salinity within the regional context, particularly given that salts have been accumulating for perhaps millennia through transpiration processes (Section 4.7.1.) and the break-of-slope phenomenon is recently developed. Low salinity groundwater is present beneath the Broken River floodplain, where there is direct recharge from rainfall through shoestring sand levees and lenses of the upper Shepparton Formation (Section 4.7.2). The plain may also receive periodic overbank floodwaters that infiltrate the shallow floodplain aquifers, helping to keep them flushed. Similarly, the influence of fresh irrigation waters is indicated in shallow aquifers underlying the Shepparton Irrigation Area in the northwest. Thus, the whole northern part of the survey area is relatively well-flushed by fresh surface waters,


74

either natural episodic floodwaters or applied irrigation waters. This is borne out in the shallow AEM slices where resistive sediments concur with the groundwater data (Appendix 1). In contrast, the CDIs for depths >10 m in the northern area show high bulk conductivity that suggest the maximum depth of influence of infiltrated fresh surface waters at around 10 m BG, below which saline groundwaters are likely to be present.

Figure 43. Groundwater EC contours.

Saline groundwater is present beneath the Caniambo Hills, representing the major salt store in the region, where groundwater ECs approach or exceed half seawater salinity. This very saline groundwater is seen to extend northwards to alluvial aquifers in the Kialla East areas. The propensity for the accumulation of saline groundwater in weathered fractured bedrock aquifers is described in Section 4.5.1. Importantly, the highly saline weathered fractured bedrock aquifers in the Caniambo Hills are connected to the adjacent alluvial aquifers of the plains, as noted in our Heartlands reporting from Sheep Pen Creek (Cresswell et al., 2003). This saline groundwater is particularly exposed at the break-of-slope at the intersection of the plains and the hills. Figure 43 is more representative of salinity with respect to the Caniambo Hills than is the AEM image.

4.6.2. Groundwater ages The groundwater dating component focused on the most saline reservoirs, i.e., the weathered fractured bedrock aquifers beneath the Caniambo Hills, and the groundwater sump within the Honeysuckle Graben. Shallow aquifers, as in the Riverine Plain, (<5 m BG), are taken to be unconfined and frequently recharged, and thereby in contact with modern carbon, so these floodplain areas were excluded from the 14C sampling exercise. Similarly, where the hydraulic gradient is steep as well as


75

shallow, as beneath the piedmont zone near the Strathbogie Ranges, groundwater flow is taken to be relatively rapid and involving recent recharge waters, hence these aquifers were also excluded from 14C analysis. The objective was to ascertain the mobility of the most saline groundwaters with respect to the potential risk to down-gradient agricultural and environmental assets and off-site export to the Goulburn River, rather than comprehensively date groundwaters in all GFS in the catchment. Bore ID Easting Northing Elevation Temp pH HCO3

m (AHD) deg. C mg/L

124 380656.7 5961614.3 161.55 21.3 6.48 315 126 384754.7 5960242.3 151.77 20.5 6.90 448 130 377932.7 5960118.3 163.15 21.6 247 133 375902.7 5955594.3 138.78 20.5 342 142 382872.7 5949444.3 159.63 21.1 6.50 363 144 387852.7 5952814.3 171.45 20.5 6.48 347

Bore ID Ground- Depth Depth

Screen depth below Elev. of EC

water to of G'water bottom of µS/cm elev. G'water hole surface hole

124 139.21 22.34 81.45 59.11 80.10 22700 126 143.51 8.27 58.43 50.17 93.34 24400 130 127.83 35.32 94.50 59.18 68.65 25200 133 125.94 12.85 76.00 63.16 62.78 27400 142 136.03 23.60 82.16 58.56 77.47 11970 144 141.32 30.14 76.00 45.87 95.45 18360

Table 3. Bore site and groundwater data for the six samples collected for carbon analysis.

Bore No

Lab No

δ 13C ‰ PDB

14C pmc + σ

Years BP (uncorrected)

124 CS 2547 -16.7 65.6 ± 0.3 3500

126 CS 2568 -15.4 17.7 ± 1.4 14200

130 CS 2569 -17.3 74.5 ± 1.5 2500

133 CS 2572 -18.0 40.0 ± 1.6 7500

142 CS 2573 -17.7 11.5 ± 1.1 17800

144 CS 2571 -16.4 14.5 ± 1.1 15900

Table 4. Groundwater carbon isotope data and indicated ages. Analytical techniques described in Section 2. Sample No. 124 analysed via AMS, all other samples via direct CO2 absorption. δ 13C = the deviation of the 13C/12C (ordinary carbon) ratio per mil (‰) with respect to the PDB laboratory standard. pmc = percent modern carbon. Uncorrected ages estimated from the 14C half-life decay curve (Figure 13) and pmc contents, assuming an initial activity of 100 percent.


76

Data for six bores sampled for carbon isotope analysis are provided in Table 3, carbon isotope data are summarised in Table 4, and the distribution of the sampling points is shown in Figure 44, along with pmc and indicated 14C ages for each site. Figure 45 shows plots of the distribution of selected data from Tables 3 and 4 within the sampled area: pmc, 14C age, EC, δ13C, and depth to groundwater.

Figure 44. Distribution of 14C concentrations in groundwater within the study area: pmc and apparent ages in years before present (BP), indicative of the time since recharge for the groundwater sample.

Carbon-14 activities range between 11.5 and 74.5 pmc, indicating an apparent 14C groundwater age range of 2500 to 17800 years BP. This represents the time elapsed since the water entered the aquifer after movement through the soil zone, or, to a lesser extent, since stream water may have infiltrated directly to the underlying aquifer. The 14C data for the sampled bores indicates no major recharge in modern times, at least not within these deeper aquifers, i.e., within the last few decades or more. These data are compatible with the 14C groundwater age range for bedrock aquifers in Sheep Pen Creek, 5 - 75 pmc, 2270 – 24,000 years BP, determined by Fisher (2003). The δ13C values of the six FRRR samples fall within a tight range between - 15.4 and - 18.0 ‰. This indicates that neither dead carbon (from marine carbonate dissolution from ancient limestones) nor oxidation of organic material is contaminating the


77

system. The δ13C values are distributed in a very consistent pattern across the area, Figure 45 (c), further reinforcing the robustness of the dataset. There is no obvious trend of increasing 14C age with depth, Figure 45 (e). Old groundwater (lowest pmc) resides in the Violet Town Graben and beneath the Caniambo Hills where there is obviously the greatest inertia in the Honeysuckle system, as suggested by the low hydraulic gradient (Figure 42a), where groundwater sumps are indicated. The presence of younger groundwaters down-gradient is probably a function of greater access to recharge waters. It is noted that the younger groundwaters are more saline than the older groundwaters. Thus, the sites beneath the mid-slopes of Sheep Pen Creek and in Earlston sub-catchment have probably received recharge waters more readily and have also either received, generated or stored more salt than upper sub-catchment sites and the Honeysuckle Graben groundwater system. This suggests greater accumulation of salt in the past beneath the mid- to lower slopes due to transpiration processes, as described in Section 4.7.1. It may also, or alternatively, suggest some mixing of groundwaters at these mid- to lower-slope locations with recycled salt generated at nearby break-of-slope saline sites at the northern edges of the hills. That palaeowaters from the Caniambo Hills bedrock aquifers are seeping to the adjacent alluvial aquifers of the Riverine Plain is indicated by the presence of several bores in the Kialla East area analysed by Fisher (2003) that gave very low pmc values (3-20 pmc) even though the majority of bores in the plain here give high pmc values (modern groundwater ages). The latter indicates modern recharge on the Riverine Plain, as described in Section 4.7.2, but with lateral input of ancient groundwaters flowing from beneath the hills. Both lateral and vertical processes appear to be represented by the 14C age data for the groundwater system. Notwithstanding, groundwater flow rates are evidently very slow where lateral processes are invoked, given that waters several thousand years old remain in residence. Where the hydraulic gradient is steep, particularly beneath the piedmont zone near the Strathbogie Ranges, groundwater flow is taken to be relatively rapid and involving fresh recharge waters. Where the hydraulic gradient is flat, beneath the plains, groundwater flow is slowest. Within the deep groundwater depressions in the Honeysuckle Graben, groundwaters are ancient and probably not flowing at all. The 14C data indicate very long groundwater residence times in much of the catchment. The antiquity of groundwaters in the Honeysuckle Graben and beneath the Caniambo Hills, of many thousands of years of age, implies considerable inertia in large parts of the system, as represented by the sampled area. The edges of the Riverine Plain in the north of the study area, where watertables are shallow and there appears to be greater direct influence from rainfall and surface water processes in the floodplains, the groundwater system appears to be more dynamic both hydrologically and in terms of salt recycling (Section 4.6.3, below). The long to very long residence times in the Caniambo Hills and Honeysuckle Graben imply that mobilisation of stored salt is slow down-catchment from these major storage areas.


78

Figure 45. Groundwater age data contours: (a) pmc, (b) apparent 14C age, (c) EC, (d) δ13C, (e) depth to groundwater.


79

4.7 Salt sources and stores, Groundwater recharge and discharge

4.7.1. Salt sources and stores Salt accumulations residing today within eastern Riverine Plain landscapes and waterways have, for the most part, been derived over many thousands of years to tens of thousands of years. The salt accumulations have two main origins: rainfall deposited ‘cyclic salts’ of marine origin, and aeolian accessions (‘dry dustfall’) sourced from more saline Late Quaternary environments in western parts of the Riverine Plain and the Mallee region further west. Blanketing loessic aeolian desposits, termed ‘parna’, are believed to be present in the study area although overlain by fresh floodplain alluvium associated with the Broken River (Butler 1956). The dust accessions would have once brought entrained salt into the eastern Riverine Plain and − whether or not original salts are still resident in the system, the extant parna deposits have the propensity to sequester latter-day salt influxes. Based on solute concentrations in rainwater in the Murray-Darling Basin for 1974-75, Blackburn and McLeod (1983) calculated present-day salt accession volumes from rainwater of 38 kg/ha/year (5.1 mgL-1) for Albury area. Accessions of these cyclic salts have been greater during arid periods of the past, particularly during the glacial periods when vast salt stores were exposed on the continental shelf during periods of low sea levels. Large quantities of earlier salt acquisitions have remained stored in the regolith – in weathered bedrock and fine-grained alluvium in particular and groundwater – of the region. Additional to cyclic salts from rainfall and subordinate windborne salt, a third potential salt source is through mineral weathering in the local environment (‘indigenous solutes’). Salt sources in the eastern Riverine Plain catchments and fringing uplands are further discussed by English et al (2002). Cartwright et al (2004) have identified three main relationships in the Honeysuckle catchment based on chloride/bromide (Cl/Br) ratios in groundwaters. The low salinity of groundwaters associated with the Violet Town Volcanics, their major ion chemistry, low Cl/Br ratios with respect to chloride, and variable δ 18O and 2H (Deuterium) isotopic composition serve to indicate that the chemistry is controlled by dissolution of silicate minerals and by rainfall-recharge waters. Saline groundwaters in the Riverine Plain in the northern part of the study area with variable Cl/Br ratios and homogenised δ 18O and 2H values indicate salt concentration through evaporation. This is understandable given the conspicuous presence of break-of-slope salinity adjacent to the ranges. Early salt accumulation here by means of transpiration processes, as described below, prior to clearing of native forests on the Strathbogie slopes, is not precluded in this instance. Intermediate salinities and intermediate Cl/Br ratios indicate mixed groundwaters (Cartwright et al, 2004). Earlier work by George (1984) in the Boho area showed that the shallow groundwater system is in chemical equilibrium with albite (NaAlSi3O8). This suggests a role for active weathering of feldspars in the Violet Town Volcanics contributing to solutes in the near-surface system. In contrast, the deeper, sluggish groundwaters are in


80

equilibrium with kaolinite (Al4Si4O10·OH8) and are probably old, stagnant and in chemical disequilibrium with the near-surface environment. Stable isotope data show that groundwaters in the Caniambo Hills and Kialla East plain are slightly enriched in the heavy isotopes 18O and 3H and have undergone some evaporation prior to infiltration through the unsaturated zone (Fisher, 2003). However, the isotopic shift is not large enough to attribute the very high salinity of the groundwaters to evaporative processes alone. Transpiration processes, whereby deep-rooted vegetation has excluded salts from water taken up at the root zone, is the most likely explanation for observed 18O and 2H values because transpiration will not produce an appreciable shift in stable isotope composition (Fisher, 2003) although it will concentrate salts and increase salinity. This suggests that much of the salt stored in Caniambo Hills in particular accumulated prior to the clearing of trees. The deep kaolinitic clay composition of the bedrock aquifers and the antiquity of groundwaters therein revealed by the 14C ages concur to explain long-term retention of salt stores, strong sequestering because of the cation exchange capacities of the clay promoting bonding between clay and salt, and extremely sluggish groundwater flow. These data and interpretations are also consistent with the noted importance of large salt stores in the unsaturated zone of the Caniambo Hills, Section 4.5.1. The stable isotope data attributes much of the salt accumulation in the Caniambo Hills to millennia of transpiration processes prior to clearing of the forests (Fisher, 2003). This can explain the high salinity of groundwaters therein. These ancient saline groundwaters may be the main culprits accounting for the high average salt export estimates for the Honeysuckle Creek catchment of 20-100 tonne/km2/year (RWC, 1988; Dowling et al., 2004). Shallow groundwaters within the Riverine Plain show more isotopic enrichment along evaporative lines suggesting that evaporative processes are a more dominant process for concentrating salts there (Fisher, 2003). Given that the watertable here is 0-5 m BG, commonly <2 m BG (e.g., logs for SPC02 and 02, Appendix 5, this report), evaporation of water molecules directly from the watertable in response to solar radiation, and retention of salts in the soil profile agrees with the isotopic signatures. This does not exclude a past history of transpiration processes accumulating salt in the Shepparton Formation prior to removal of trees from the plain and prior to elevation of the watertable to its present shallow level. It is consistent that ancient salt stores accumulated in this manner are susceptible to recycling under present-day conditions, particularly given the responsiveness of parts of the plain to modern recharge processes (Section 4.7.2). Groundwater salinities here are less than would be otherwise expected (<6000 µS/cm EC, Section 4.6.1), on the strength of the dominant evaporative processes alone. This is because of the presence of sand levees that promote the infiltration of fresh water − rainfall or episodic overbank floodwaters from the Broken River − into the shallow Shepparton Formation aquifers, as described below, plus the influence of applied irrigation waters further west in the SIA. Clearly, the accumulation of abundant salts in the Honeysuckle catchment aquifers is not exclusively a phenomenon of the last 150 years or so since European settlement of the region. The proportion of resident salts that can be ascribed to the two processes − ancient long-term transpiration processes beneath former Eucalypt forests versus modern salt stores that have accumulated through the combination of


81

continued transpiration processes in parts of the catchment and evaporative concentration from very shallow watertables − is not known.

4.7.2. Recharge and discharge The interpreted main recharge and discharge dynamics in the study area are illustrated in Figure 46, drawn largely from the work of X. Cheng and M. Reid, DPI, and corroborated in the present study. The hydraulic gradient and groundwater salinity (EC) contours (Figures 42 and 43) reveal the base of the Strathbogie Ranges as the main recharge area. George (1984) estimated that recharge in the Boho area has increased by 20 mm/year in response to reduction in transpiration following clearing of the native forest. It is this 20 mm excess recharge water sourced from the Strathbogie slopes that seasonally evaporate at the break-of-slope adjacent to the ranges. Additional recharge waters sourced from the runoff zones − facilitated by relatively high hydraulic conductivity in the colluvial aprons − are slowly flowing northwards. The Honeysuckle Graben is filling at a rate of 15-20 cm/year (Figure 26), possibly from a range of sources: infiltration from perennial streams and ephemeral swamps, diffuse recharge (“deep soil water loss”) from extensive cropping lands, and lateral through flow from the south and north. Ancient groundwaters, >15000 years old, reside in fractured rocks at the base of the Violet Town Sump. In places, the latter may have closer hydraulic connection with fractured rock aquifers beneath the Caniambo Hills to the north than with the overlying alluvial deposits within the graben.

Figure 46. Groundwater recharge and discharge processes in the study area (Modified after Cheng and Reid, 2001, unpublished).

In the Riverine Plain across the northern part of the survey area, shoestring sand levees and underlying sand lenses, which are the disaggregated legacies of prior streams, are important to groundwater recharge. The levees are 0-6 m thick, and form gentle rises on the Broken River floodplain around Kialla East and Caniambo (Day and Harvey, 1994). These highly permeable surface and near-surface sand units drain fresh water (150-1500 µS/cm EC) both laterally and vertically. Annual recharge is


82

100 mm/year through the levees, whereas the average for the overall landscape of the plains, levees and the surrounding clayey soils, combined, is 20 mm/year (Day and Harvey, 1994). Four lines of evidence substantiate the widespread occurrence of direct vertical recharge to shallow aquifers in the Shepparton Formation in the Kialla East and Caniambo plains: (a) freshening of groundwater, from 14000 to 4000 µS/cm EC approaching the northwest part of the study area (Figure 43); (b) numerous hydrographs showing marked responsiveness to rainfall events that are above average in magnitude (Figures 17 (a) to (c) in the Heartlands report (Cresswell et al., 2003: http://www.clw.csiro.au/heartlands/publications/general_publications.html, p. 154-156); (c) low conductivity in the 0-10 m CDIs of the AEM data that indicates a well-flushed upper substrate, possibly to a maximum depth of 15 m BG (Appendix 1); and, (d) radiogenic isotope data, summarised below. Several groundwater samples in the Kialla East area analysed for tritium by Fisher (2003) gave detection levels of 1.5 to 2.6 TU. This indicates modern recharge to the Riverine Plain aquifers. Of twelve groundwater samples from the Caniambo Hills aquifers, only two gave tritium levels >1.5 TU, indicating negligible modern recharge in the bedrock aquifers. Cross-correlation of tritium and 14C values generally indicates a bimodal distribution, i.e., most palaeowaters identified by low pmc have negligible tritium, and most samples with detectable tritium contain high pmc values. Where there is divergence, mixing of older groundwaters with some modern infiltrating waters can be assumed. In the skeletal soils (‘lithosols’ or Rudosols, Figure 18) on the crests of the Caniambo Hills, recharge has been estimated at 30 mm/year (Craig, 1988). Recharge via stony ground, e.g., Figure 22, and associated gradational soils is at least ten times greater than through the duplex soils of the gentle-sloping lower landscape (Craig, 1988). The vertical dip of the folded Silurian metasedimentary rocks (Figure 23) facilitates infiltration of rainfall to the underlying groundwater system (although this may not be the most efficient recharge mechanism in the catchment overall). Recharge, however, is reduced on the crests and ridges of the Caniambo Hills where there is an overstorey of trees (Kelly, 1994). At the Heartlands agroforestry trial blocks at ‘Coomalong’ in the headwaters of Sheep Pen Creek, red ironbark (E. tricarpa) have been observed to utilise all incident rainfall, with no percolation beyond the root zone in recent years. High salinity, subdued hydraulic gradients and low pmc concentrations indicate long residence times of groundwaters in the bedrock aquifers and very sluggish lateral flow. Where there are faults and fractures, such as illustrated in Figure 42(c), recharge and discharge processes are both expected to be accelerated. Saline discharge occurs, as noted, at the break-of-slope where the Violet Town and Euroa plains meet the Strathbogie Ranges and likewise where the Riverine Plain meets the base of the Caniambo Hills (red outlines in Figure 8b). This is due to excessive recharge caused by clearing of deep-rooted native trees from the hills and to the abrupt changes in hydraulic gradient and, in places, to a change in hydraulic conductivity as well. In the Riverine Plain, discharge occurs directly to streams and levee banks. Watertables in the Shepparton Formation have risen many metres in the past 100 years and have intersected the already naturally deeply-incised drainage channels such that slow inflow of groundwater to the creeks is inevitable.


83

Thirdly, discharge is promoted at bedrock/topographic constrictions because natural discharge capacities of enclosed valley floors and their underlying aquifers are exceeded. At such loci, groundwaters are forced upwards to intersect the ground surface, streams, and swamps. Examples of serious salinisation caused by this topographic and hydrologic juxtaposition, combined with the co-occurrence of high salt stores in the weathered, fractured bedrock at the constrictions are conspicuous. Most noteworthy are the lower reach of Sheep Pen Creek and Downes Swamp (Figure 7), Earlston sub-catchment (large low-lying areas outlined in red in Figure 8b) and the Broken River at Nalinga (Figure 39). The imposition of limited discharge capacities may also apply in the piedmont zone near the Strathbogie Ranges where numerous sub-valleys are virtually enclosed by outlying outcrops of bedrock. This is particularly applicable at both Warrenbayne and Boho and subordinate surrounding sub-valleys, as illustrated in Figure 36. It was first noted by George (1984) that the discharge sites at Boho had remained stable for 50 years. Similarly, it is noted that other saline discharge sites tend to expand during periods of higher than average rainfall and to later contract to pre-existing extents (M. Reid, Victorian DPI, pers. comm., 2003). This indicates that the system may be at, or closely approaching, a new state of hydrologic equilibrium, albeit with watertables that are very much higher than their pre-clearing levels. These observations suggest that the respective groundwater systems are virtually full to capacity. Notwithstanding this apparent balance between recharge and discharge, it is evident from most of the data collected for the present study and our interpretations that major salt disequilibrium characterises the catchment. Salt has been resident in the various GFS for millenia, in weathered bedrock in the Caniambo Hills, particularly in subtle valleys incised into the hills that are infilled with locally-sourced kaolinitic clay, and in clayey alluvium in the Shepparton Formation, such as palaeo-swamp settings where Grey Vertosols are now present. These ancient salts are currently being flushed from the system − most likely to the Goulburn River − albeit at a modest rate compared to the volume of extant salt stores in the catchment. If the volumes of groundwater currently discharging are in at least quasi-equilibrium with the recharge volumes that are excess from pre-clearing days, it holds that reduction of recharge will result in a commensurate reduction in discharge which, in turn, would lower the watertables. Thus recharge reduction is desirable for the whole catchment even if purging the system of ancient salt stores may be a more intractable proposition. Salt exports from the Honeysuckle Creek catchment through groundwater discharge to the Goulburn and Broken rivers are large although the catchment is not the highest salt exporter in the overall Goulburn-Broken region. Certainly salt exports are orders of magnitude greater than contemporary salt inputs to the Honeysuckle catchment. The present-day salt accession volumes from rainwater are around 38 kg/ha/year (3.8 tonnes/km2/year) or 5.1 mgL-1 (data for Albury, from Blackburn and McLeod, 1983). Our data in Table 1 indicate a salt store of at least 20 - 100 tonne/km2/year. This suggests that, depending upon conditions, potentially 5 to 25 times as much salt can be exported from the catchment as is arriving in rainfall. Actual export volumes are difficult to quantify because of dilution by high streamflows inherited from the headwaters in the Victorian Alps and because of the influence of applied irrigation


84

waters, sourced from Eilden Weir, nearer the Shepparton Irrigation Area, because of a limited array of stream gauging stations, and because the rainfall of recent years may not be typical over decadal timescales. Relatively little salt is being exported to the Goulburn River directly from Honeysuckle Creek itself. Most is being transported via the shallow aquifer system that ultimately drains the whole catchment. This discharge is likely to be to both the Broken River and the Goulburn River by relatively slow discharge directly to the riverbed or banks. In the case of the Broken River, the salt load acquired along its course west of Nalinga ultimately ends up in the Goulburn River at the confluence immediately south of Shepparton although it is diluted here by mixing with infiltrated irrigation waters.

4.8 Surface Water Yield The FRRR project is considering revegetation (planting of trees and shrubs and encouraging the planting of perennial grasses) as an option to reduce recharge and lower the watertables and thereby reduce discharge wherever possible in the catchment. Greater areas of trees and shrubs use more water than the agricultural plants that they replace, thus conferring advantage through preventing groundwater recharge. A consequence however, is that rainwater runoff is also reduced because of consumption of available water by trees. This can reduce streamflow accordingly. Therefore, any revegetation for recharge reduction and salinity management must be balanced against possible undesirable reductions in runoff and streamflow. A hydrological modelling approach based on Zhang et al. (1999, 2001), Figures 15 and 16 (Section 2.3), has been applied to predict the mean contribution of water from different parts of the Honeysuckle Creek catchment. This indicates the parts of the catchment where revegetation activity such as tree planting would have the greatest impact on runoff and the amount of water finding its way to the river system. The model has been parameterised with rainfall data and calculated for the entire catchment. The predicted mean annual water yield (mm) for Honeysuckle Creek catchment is shown in Figure 47. It should be borne in mind that ancient groundwaters, as determined by 14C dating (Section 4.6.2), contribute to streamflow although the volumes involved are probably not great compared to surface water. The role of palaeowaters seeping to the streams is not accommodated in the modelling which relates to present rainfall and evapotranspiration rates. The higher rainfall parts of the catchment, particularly the Strathbogie Ranges, are the biggest contributors of water into the Honeysuckle Creek system. Other parts of the Goulburn-Broken region − outside Honeysuckle Catchment in the Great Dividing Range − generate substantially more flow to the main rivers, the Broken and Goulburn, than the areas shown here. The information from this analysis can be used when looking at appropriate land use options for salinity management in the study area. Seasonal variability in water yield needs be taken into account for application of the water yield modelling. Also, the model is based on evapotranspiration of well-established plantations and does not take forest age, tree species and soil suitability into account. The latter consideration with respect to soils in the catchment is broached below, Section 4.9. Thus the water yield model provides a broad brush


85

guide for initial planning phases towards attaining a healthier water balance in the catchment. It is mainly the impact of tree planting on local stream flow, Honeysuckle Creek itself and its tributaries that need to be considered in this regard, not the impact on riverflow in the Goulburn and Broken rivers. Reduction of streamflow within the catchment through afforestation programs therefore mainly needs to be weighted against the beneficial effects of lowering the watertable − particularly where the groundwater is saline − and reducing the potential extent of saline discharge areas.

Figure 47. Predicted mean annual surface water yield (mm), based on the hydrological modelling approach of Zhang et al. (1999, 2001).


86

4.9 Land Resource Assessment Outputs from the MrVBF modelling for the target study areas are presented in respective sub-sections, above: in Figure 29 for the Miepoll area, Figure 35 for the Baddaginnie area, Figure 36 for the Warrenbayne-Boho area, and Figure 39 for Sheep Pen Creek sub-catchment. In the report sub-sections, these plots were interpreted with respect to the AEM imagery and soil maps for the individual study areas, and were also utilised at the commencement of the project to help target our FRRR drill-hole sites. The MrVBF plots are also useful for planning land use change and designating revegetation areas in light of the interpretations derived from our integrated research. Hillcrests and relatively steep slopes ― which are potentially high runoff and high recharge areas ― are represented by low MRVBF orders 0 to 3.5 (grey, orange, pink, yellow). Lateral, down-gradient flow is the dominant hydrologic process in these areas. They are suitable for farm forestry aimed at recharge reduction and runoff interception, contingent upon the soil being sufficiently thick to sustain mature trees. Dense stands of remnant mature eucalypt and acacia forest are present on the hillcrests and upper slopes of the Caniambo Hills which not only suggests sequestering of all rainfall by the trees but possibly also the roots finding water in deep fractures in the bedrock. Depositional parts of the landscape correspond with the higher value MRVBF indices, orders 7.5 to 4.5 (magenta, purple, dark green). These areas tend to contain deep soils. These areas are sometimes the loci for swamp environments, or were once swamps in the geologic past when they accumulated dense, clay-rich soils. They also coincide with floodplains along the main stream courses. Vertical hydrologic processes are dominant here. Commonly, these low flat areas are prone to waterlogging and salinisation. The MrVBF index might therefore have some application to anticipating the areas most likely at risk of concentrating excess water in the surface to near-surface hydrologic system. The index is particularly useful for depicting topographic constrictions that almost certainly impinge upon surface and groundwater hydrology, particularly in this landscape where so much of native vegetation has been cleared and recharge rates can be excessive. Such topographic constrictions are commonly associated with existing saline areas in the catchment, e.g., the lower reach of Sheep Pen Creek and Downes Swamp. This partitioning of the landscape is also potentially useful for guiding tree planting above break-of-slope areas to offset groundwater discharge and lower the watertable in these dynamic areas where exposure of the watertable and salt efflorescence are already well-established processes.


87

Figure 48 (a) Soil thickness plot for the survey area based on broad-scale terrain analysis with the MrVBF and Topographic Wetness Index, (b) Regolith depth plot derived from the AEM Elevation of the base of the conductor (Figure 25), included for comparison.


88

A soil thickness plot for the study area is shown in Figure 48(a). This shows considerable detail derived from the MrVBF and Topographic Wetness Indices (outlined in Section 2) and available soil information. In large part, this plot relates to erosional and depositional domains. For comparison, Figure 48(b) is a ‘regolith thickness’ plot derived from the AEM elevation of the base of the conductor image (Figure 25), included here to illustrate the appreciable correspondence between the two very different plots. Figure 48(b) may not represent true ‘regolith thickness’ because the AEM does not distinguish between fresh bedrock and well-flushed sediment (Section 4.3.1). In the AEM-derived plot, Figure 48(b), white areas are exposed bedrock and areas with negligible soil development, on the crests of the Caniambo Hills and outliers of the Violet Town Volcanics north of the Strathbogie Ranges. These outcrops and subcrops are fringed with regolith that is 5-20 m thick (variable blue areas). These comprise colluvial aprons around the Stathbogies and saprolite associated with the weathered Silurian shales of the Caniambo Hills, near situ upland valleys that have been infilled with locally sourced clays, and deeper soils distal to the hills. Two main depocentres, containing regolith >20 m thick, are depicted in dark blue, Figures 48(a) and (b): the Honeysuckle Graben underlying the Violet Town Plain, and the Murray Darling Basin in the northwest part of the study area, beneath the Riverine Plain in Figure 48(a). There is a correspondence between the latter deep terrain classes and the AEM CDIs which represent conductivity for the survey area. Deeper regolith modelled on the basis of terrain analysis and available soil data, Figure 48(a), is shown to correlate well with high conductivity areas in the AEM images (large red areas in CDIs, Appendix 1). Likewise, there is an appreciable correspondence between the modelled areas of shallow soil, yellow-green in Figure 48(a), and resistive areas (which have AEM conductivities of <150 mS/cm) and are shown in light blue or white in Figure 48(b). Figure 48(a) has potential usefulness for selecting areas for tree planting aimed at mitigating break-of-slope salinity as well as for more general recharge reduction on the hill crests and upper slopes. In the case of appropriate break-of-slope tree planting, the GFS and local depth to watertable should also be taken into consideration. Additional to topographic constraints, soil thickness is important with respect to the viability of planting perennial vegetation − trees or deep-rooted grasses − as well as to understanding deep drainage processes in difference landscapes. Figure 49 is a model output of the predicted available water capacity for the catchment, based on soil thickness and physio-chemical soil properties for the A and B horizons that relate to water retention potential. Figure 50 is a modelled prediction of available water holding capacity to a depth of 2 m based on a scaling factor that reflects root distribution (McKenzie et al., 2003).


89

Figure 49. Predicted water holding capacity plot for Honeysuckle catchment based on the soil depth and soil parameters (for the A and B horizons).

Figure 50. Predicted available water holding capacity for to a depth of 2 m.


90

These plots have potential application for broad-scale planning of revegetation − trees in deeper soils that can sustain growth to maturity in terms of water availability, and perennial pastures and crops which do not have the same depth and water requirements as farm forestry. The catchment-scale models need to be supported by detailed local information for farm-scale designs for revegetation towards abatement of waterlogging and offsetting salinity in down-gradient positions. Application of the soil thickness and available water holding capacity maps, ideally, should be used in conjunction with the water yield plot (Figure 47) and associated information about the hydrologic impacts of afforestation on streamflow. The water yield plot and associated information (e.g., Figure 15 and 16) suggest that trees should be planted in regions of higher rainfall to offset loss of down-gradient streamflow. It should be born in mind that maintenance of streamflow is mainly only important in this regard for the within-catchment creeks, Honeysuckle, Riggs, Irish, Sheep Pen, etc. The Broken and Goulburn rivers are sustained by high surface water discharges from maximum rainfall zones in the Victorian Alps (and by releases from Eildon Weir in the case of the latter) and are not very dependent upon contributions from Honeysuckle catchment. Mosaics, phase plantings and farm forestry belts located away from drainage lines are appropriate options where maintenance of local streamflow and preservation of wetland habitats are deemed more important than runoff and recharge reduction. In this regard, the soil thickness and water capacity outputs may be appropriately used in conjunction with water yield models to optimise revegetation for whole-of-catchment health, agroforestry productivity and sustainable water balance.


91

5. Summary & Recommendations The key findings of this interpretation of the airborne geophysical data and additional datasets compiled or generated specifically for the FRRR project are summarised here. Some implications relating to the scientific research are also outlined below, along with general recommendations for management options stemming from our investigation. Management recommendations are also provided in a concurrent client report (English et al., 2004). The latter report deals not only with salinity management but also with consideration of restoration of biodiversity in the catchment, information about farm forestry trials we have conducted in the area, and communication and capacity building components of the project. The client report is aimed at providing a link with the Heartlands program and the next phase of implementing on-ground action. Much of the fieldwork and research for the FRRR project has been aimed at answering the following questions:

• Where are salt stores located, and how much salt is present? • Are all salt stores accounted for in the AEM? • Are these salt stores being mobilised to lower landscape positions and to the


In this section we bring together the findings with respect to these questions. We also present some concluding comments regarding the application of the airborne geophysics datasets and our complementary research.

5.1 Where are salt stores located, and how much salt is present? The main salt store in the Honeysuckle catchment is pervasive highly saline groundwater in the fractured weathered sedimentary bedrock of the ancient Caniambo Hills and in locally accumulated kaolinitic clay derived from the weathering of this bedrock. The latter commonly infills subtle valleys incised into the hills. Up to 1000 tonne/ha of salt is stored above the watertable, loosely bonded chemically to kaolinite clay and sequestered in fracture networks and poorly consolidated saprolite. Groundwater in the bedrock aquifers is very saline, typically around 20000 µS/cm EC or more (i.e., approaching half seawater salinity). These salt stores have accumulated through many thousands of years, possibly many tens of thousands of years, via transpiration processes whereby former eucalypt-dominated forests progressively left salt behind in the root zone. These salt accumulations gradually built up in the underlying kaolinitic rock in the thick unsaturated zone whilst a proportion has been flushed vertically down to the aquifers, as evinced by the groundwater salinities. These are palaeowaters, aged 2500 to >10,000 years old. Salt stores of 600-1000 tonne/ha have accumulated in dense alluvial clays at the edges of the Riverine Plain, namely in the lowlands of Sheep Pen Creek and Earlston


92

sub-catchments. These areas are mapped as Grey Vertosols (‘Upotipotpon Clay’) and may represent former swamp deposits in topographically constricted pockets. Watertables are shallow, typically <5 m, and commonly <2 m BG. Groundwater salinities here are similar to those in the surrounding hills, >20000 µS/cm EC. Generally the system is more dynamic at the edge of the Riverine Plain compared to that in the hills. This is because of influxes of fresh recharge waters in shoestring sand lenses in the plain and possibly also to episodic influence of river water and overbank flood waters from the Broken River. It is noted that not all Grey Vertosols in the area contain high salt stores and that other soil types may preferentially store salt (see Section 5.2, below), and that factors other than soil type come into play with respect to the propensity to store salt. In the south of the study area, small salt stores are located at the break-of-slope at the base of the Strathbogie Ranges. The Violet Town Volcanics themselves are not a major store of ancient salts but are merely generating a relatively modest quantity of solutes through present-day weathering and erosion in the near-surface layers of rock and colluvial aprons, along with some meteoric accessions. The break-of-slope saline outbreaks are the combined effects of high runoff and excessive recharge from the slopes, the sharp changes in hydraulic conductivity and hydraulic gradient at the base of the range, exposure of the watertable or of the capillary fringe, and consequent evaporative concentration of available solutes at the localised groundwater ‘outcrops’. The resultant salt scalds can be seasonal or episodic, with salt efflorescing during high evaporation rates in summer and subsequent leaching by winter rainfall and high runoff from the slopes. They can be regarded as cyclically perennial salt stores. The dissolved salts are flushed northwards into subjacent aquifers in the piedmont zone. Over much longer timeframes, some of these excess saline waters may contribute to the more stagnant groundwater system within the Honeysuckle Graben further north although it is likely that these deep reservoirs were saline long before the break-of-slope discharge sites developed.

5.2 Are all salt stores accounted for in the AEM imagery? Many, but not all salt stores are accounted for in the AEM imagery. Subtle valleys incised into the Caniambo Hills which are seen to extend down-gradient beneath the Riverine Plain, are filled with saline sediments that are dominated by clay and are well-depicted in the AEM CDI depth slices. The surrounding and underlying metasedimentary bedrock itself and the groundwaters therein, in the upper reaches of these valley systems, are not well-represented in the AEM despite the widespread presence of highly saline groundwater in the weathered profiles and fracture networks. For example, in drill-holes FRRR02 and SPC02, groundwater with salinities exceeding 25000 µS/cm EC were intersected in saprolitic aquifers which, themselves, contain moderate salt stores (>1000 mS/cm EC1:5 and correspondingly moderate EM-39 measurements, Appendix 5). For the most part, bedrock in the Caniambo Hills registers low bulk conductivity in the CDIs, <150 mS/m (Appendix 1), apparently failing to detect highly saline groundwater that is pervasive in the GFS. Drill-hole SPC02 is very close to the site of the BRS drill-hole HC09 (Figure A5.1) within which saprolite below 20 m gave lower EM-39 and EC1:5 measurements than the proximal SPC02 drill-hole. The differing salinity measurements for these two sites remains to be explained although calibration of the EM-39 device and the methodologies used for EC1:5 analyses may need cross-referencing. No depth to


93

groundwater nor groundwater salinity measurements were documented from the BRS calibration drilling (Jones, 2002; Dent, 2002) to fully constrain the nature of bulk conductivities in the survey area. The inability of the AEM technique to detect saline groundwater and salt stores in the bedrock and regolith of the Caniambo Hills is most likely because of low water/rock ratios, with salts and waters confined to often narrow fractures and joints, possibly to micro-fractures, that have a heterogeneous distribution in the rock matrix. Our groundwater salinity contour map, Figure 43, better represents the distribution of saline groundwater in the catchment than the AEM imagery. An understanding of the hydrogeology − the hydraulic gradient in addition to groundwater salinity − is perhaps more pertinent than the AEM with respect to appreciating the overall salinity situation of the catchment. This represents a serious limitation with the AEM technology given that the Caniambo Hills are host to massive salt stores, as are possibly many other fractured Palaeozoic bedrock provinces around the edges of the Murray Darling Basin. This fact is pertinent to any future AEM surveys planned for regions that encompass fractured Palaeozoic bedrock terrain such as Wagga Wagga, Kamarooka and the South West Goulburn region. In such areas, drilling and measuring groundwater ECs may prove to be a more appropriate and cost-effective strategy than conducting airborne surveys. Even if detailed physical analysis of the density, size and interconnectedness of fracture systems in bedrock aquifers were to be conducted, these need to be linked to hydrogeologic data − watertable depth, flow directions and salinity of the contained groundwater system − before AEM can be usefully applied. Salt stores in the Shepparton Formation of the Riverine Plain in the northwestern quadrant of the study area are conveyed well in the AEM, extending laterally down-gradient from subtle clayey upland alluvial valleys in the Caniambo Hills to the plains. A similar finding from TEMPEST AEM data for the Kamarooka area near Bendigo is documented by Lane et al. (2001). In the lower reaches of the Honeysuckle Creek catchment, at depths of 10 – 30 m the AEM highs merge within a broad area in plan view where the conductivity pattern corresponds roughly to the configuration of coalesced creeks (the lower reaches of Honeysuckle Creek and Seven Creeks and their tributaries). In x-y space, the high conductivity area is very extensive in the 10-20 m CDIs, i.e., the upper Shepparton Formation. Below 20 m, the area of high conductivity is seen to contract, and below 40 m depth only resistive substrate is detected. In the shallow CDIs, 0 to 15 m depth, there is clear demarcation where the Shepparton Formation sediments have been flushed by episodic fresh floodwaters from the Broken River floodplain or by irrigation waters in the Shepparton area. An alternative explanation is that the floodplain comprises non-saline sediments, sourced from far up-catchment in the Great Dividing Range rather than from the Caniambo Hills. Regardless of the provenance and/or degree of flushing of this sedimentary package, the AEM is valuable in its capacity to depict this demarcation in the alluvial occurrences of salt and saline groundwaters. The footprint size of the AEM data resolution, which is effectively 40-100 m2, may limit detection of some break-of-slope salinity outbreaks at the base of the Strathbogie Ranges since these seeps and scalds are commonly scattered and smaller than the grid cell size. Thus, even though break-of-slope salinity is linked to broader,


94

deeper causative processes, the actual manifestation can tend to result in small scattered seepage sites within a narrow zone along the topographic contour. The latter are relatively surficial features − albeit deleterious to productivity at the paddock scale and to the integrity of soils, local streams and habitats − and it is possible that contributing conductivities in the near surface, such as plant roots, may interfere with the signal for each given AEM footprint in the uppermost CDIs. The groundwater salinity contour map, Figure 43, better represents the sharp increase in salinity at the base of the slopes than does the AEM imagery. The vertical salinity gradient − additional to lateral variations such as presented in a contour map − is also potentially diagnostic in such areas. Sub-sampling of the shallow AEM CDI to focus on small break-of-slope areas and applying histogram stretches to the data may ameliorate the limitation with the technology. In the case of break-of-slope salinity, the AEM imagery, for the most part, can only detect what is already known and already conspicuous at the landscape surface and has limited capacity to predict potential future outbreaks in such zones. Other information − the appearance of indicator species and bare ground in the landscape where plains meet hills, aerial photographs, bore data and ground geophysical traverses − should be relied upon in to predict the occurrence and/or expansion of break-of-slope salinity. Terrain analysis, such as application of the MrVBF index, also has potential application here where AEM may be inadequate. A limitation with the AEM is that there is no discrimination between salt in the unsaturated zone and saline groundwater. In the case of the eleven calibration drill-holes suck by the BRS in 2002, no depth to groundwater nor groundwater salinity measurements were documented; this limits the understanding of the AEM signatures and the nature of salinity in the overall catchment. In effect, it is necessary to sink as many drill-holes as possible to ascertain the depth to groundwater, the severity of salt accumulation and the disposition and dynamics of salt stores in the landscape and groundwater systems. This fact is relevant to the following key question regarding the mobility of salt stores.

5.3 Are these salt stores being mobilised to lower landscape positions and to the catchment waterways? The AEM provides a static ‘snapshot’ of the aggregated conductivity of each given grid cell area (effectively 40-100 m2) in the landscape as it is being overflown by the survey aircraft. The AEM dataset, alone, gives no indication of the mobility of salt or saline groundwater, or of temporal fluxes, be they vertical or lateral. Thus it is imperative that the AEM data is combined with other datasets, particularly groundwater data, for sensible interpretation. Once the distribution of major salt stores are located in the catchment − through the combination of the AEM datasets, drilling, groundwater EC measurements and laboratory analysis − conventional hydrogeology techniques can provide a strong indication of the mobility of saline groundwater, and its directions and rates of flow. Tacit here is that salts below the saturated zone are the most mobile, depending upon hydraulic conductivity, that salt in the unsaturated zone (soils and regolith above the watertable) is less of a hazard to down-gradient assets, except in the case of actual surface efflorescence which may wash downstream during rainfall events.


95

Construction of a map of the elevation of the groundwater surface has enabled us to establish where the hydraulic gradient is steepest, where the system is most dynamic in terms of lateral throughflow, and where these flow paths are with respect to both identified salt stores that they intersect and to lower landscape positions and streams. Groundwater samples from carefully selected sites were dated using the 14C technique to give an indication of residence times in areas where salt stores are verified. Used in conjunction, the hydrogeologic information and the distribution of 14C ages give an indication of the rates of groundwater flow. Palaeowaters, aged several to many thousands of years, are stored in the deep fractured bedrock aquifers beneath the Caniambo Hills. The system is very sluggish. Notwithstanding, groundwaters are slowly seeping out at break-of-slope areas at the base of the gentle hills and also travelling further down-gradient to Shepparton Formation aquifers lying shallowly beneath the lowlands. The lateral flow rate may be of the order of 1 km/1000 years, except perhaps where there are preferential flow paths afforded by permeable fault zones where the rate would be accelerated. The simple fact that the aquifers give the impression of being at their maximum storage capacity may be the driving force behind discharge rates being roughly equal to recharge rates, even though the actual discharging waters may have had a prior long residence time in the up-gradient system. This quasi-hydrologic balance is more important than the time that a given molecule of water takes to travel from the recharge zone to the discharge zone. In this scenario, any inputs of fresh recharging waters serve to force old groundwaters out of the system at the discharge end because the system is full rather than because of a steep hydraulic gradient or high hydraulic pressures or high conductivities. At present, available hydrographs (presented in Cresswell et al., 2003) span the past two decades. A longer period of monitoring, over a fuller range of climatic conditions, is required to more firmly establish the concept of hydrologic equilibrium for these GFS. The hydraulic gradient from the base of the Strathbogie Ranges northward is moderately steep, indicating a relatively fast rate of groundwater flow. These flow paths are mobilising salts that accumulate seasonally at the break-of-slope salinity sites. Most of this salt ends up in the shallow piedmont zone aquifers and some is probably also added to ancient salt stores in the Honeysuckle Graben. Ancient groundwaters in deep sumps at the base of the graben, beneath the Violet Town Plain, appear to be immobile. This system is not full to capacity although it is gradually filling up with constant or seasonal influxes from diverse recharge mechanisms. The hydraulic gradient elsewhere in the study area, in the Riverine Plain and in the western part of the Violet Town Plain and contiguous Euroa Plain is low, indicating slow rates of salt mobilisation from Shepparton Formation aquifers. Despite the low gradient, saline groundwater outcrops in some swamps and low-lying areas because of the shallow depth of the watertable. Importantly, discharge from shallow saline watertables is likely to be promoted to the rivers because of the deep incision of the Goulburn and Broken river channels although the actual volumes of such seepage may be small. The research has supported the notion that a new hydrologic equilibrium has been or is close to becoming established in the catchment in recent decades, with recharge approximately equaling discharge. Thus, in pre-clearing times, 150 years or so ago, there would have been negligible recharge to the groundwater system because most


96

rainfall and runoff would have been used up by native forests. Discharge too would have been negligible as a consequence, particularly given that the watertable was probably many metres deeper than at present and seldom intersecting the landscape surface or streambeds. Ancient salt stores and ancient groundwaters were probably well contained at depth in the system and barely or only poorly mobilised to the landscape and rivers prior to the historic period. With perturbation of this hydrologic balance upon clearing of the forests, the ancient salt stores have been mobilised, albeit at a fairly sluggish rate. Salt outputs from the catchment now exceed contemporary salt inputs by a factor of between 5 and 25. Given that discharge approximates recharge under the current hydrologic regime, the only way of stemming salt discharge is to curtail recharge and thereby lower the watertable.

5.4 What intervention measures are appropriate? This project has sought to identify and quantify salt stores in the Honeysuckle Catchment − utilising high resolution airborne geophysics and existing data in the first instance and conventional hydrogeological techniques where the remotely sensed data proved insufficient − and to then ascertain the mobility of the salts. The next step, given this understanding, is to consider the most appropriate intervention measures to protect agricultural and environmental assets from mobilised salts. General implications from our findings are considered here and given in a complementary client report (English et al., 2004). The aim has been to generate well-founded guidelines rather than provide a blueprint for land management in the catchment. Three general options are available to manage or adapt to saline landscapes: (a) revegetation to reduce recharge and lower the watertable, and to stabilise and rehabilitate already salinised areas, (b) engineering works to intercept saline groundwaters before they reach valuable assets and, (c) adopting land use − and concomitant social and economic imperatives − to living with salt and possibly utilise salt land and saline waters profitably. Suggested measures are summarised in Figure 51. It is emphasised, with respect to the findings of the present study and to assessing salinity risk, that where the rates of salt mobility are low (metres/year) and where salt is more than a few hundred meters from an asset, the salinity hazard can be considered low. Specific recommendations for farm forestry included below are based on a CLPR (2001) assessment (compiled by X. Cheng and M. Reid, DPI), and on our Heartlands experience in the catchment. Suitability criteria for farm forestry adopted in the CLPR (2001) assessment were weighed on the following critical factors that affect tree growth and water use: climate (particularly rainfall), depth to watertable BG, salinity of groundwater, soil (type, depth and chemistry) and landscape position.

5.4.1. Strathbogie Ranges and adjacent piedmont plain The main recharge in the Honeysuckle catchment occurs in the high rainfall zone of the Strathbogie Ranges, particularly where the slopes have been cleared of trees.


97

Large volumes of excess runoff at the base of the ranges exceed the discharge capacities of the surface and sub-surface drainage systems. This is particularly the case at the break-of-slope and within an aureole extending a few kilometers out from the slopes where outliers of bedrock occlude both overland flow and groundwater throughflow. Waterlogging occurs in the surface environment and, likewise, excessive recharge occurs in subjacent aquifers in the piedmont plain. Reafforestation of the cleared slopes is required to utilise much of the ~1000 mm of annual rainfall where it falls, whilst at the same time allowing sufficient runoff to maintain flow in creeks that rise in the ranges and to sustain down-catchment wetlands. Reforestation of at least 50% of the area of the cleared slopes may be required to balance the water budget here. Soil depth may be a limiting factor to suitability for tree planting, and in this regard, the soil thickness and available water capacity modelling conducted for the present study, Section 4.9, may be a useful guide for initial planning. Concentrated tree planting, in belts and blocks, in particular, is required towards the base of the slopes to harvest as much runoff as possible and/or to mine interflow water before it reaches the break-of-slope zone. The MrVBF index (Figure 36) may provide a useful guideline for planning the locations of tree belts at appropriate topographic contours. Selection of tree species, along with consideration of density, thickness of belts and spacing between trees, is crucial because of the possibility of growth not being sustainable during dry years. Modelling by Daamen et al (2002) has shown that the large recharge occurring on the slopes at Boho can not be offset by a narrow band of trees at the break-of-slope. Establishment of 100 m wide plantations of trees, for example, mainly only produces positive effects in the immediate area and close surroundings (Daamen et al., 2002). The CLPR (2001) assessment of the suitability of areas in the Goulburn Broken region for farm forestry indicates that the Strathbogie Ranges and adjacent piedmont zone have a high to moderate suitability for plantations of Sugar Gum (Eucalyptus cladocalyx), Eurabbie (E. globulus ssp. bicostata) and Mugga (E. sideroxylon). The area has a low to moderate suitability for Southern Blue Gum (E. globulus ssp. globulus) and Red Ironbark (E. tricarpa). Earlier trials of blue gum belts above the break-of-slope at Warrenbayne and Boho have, in places, resulted in high mortality rates prior to maturation of the trees (field observation). This heralds the need to strike a balance with respect to species selection and tree densities. Currently, as part of the Heartlands Initiative, agroforestry trials include a 7 year old shining gum (E. nitens) and Tasmanian Blue Gum (E. globulus) stands on ‘Rotherlea’ in the 700 mm rainfall zone in the Warrenbayne piedmont zone (Landowner: Angus Howell). An understanding of plantation water requirements during different stages of growth and variable climatic conditions, such as prolonged drought and potential drying up of the watertable in shallow soils over bedrock, is also needed in these hydrologically dynamic and strategic zones. Saline discharge areas at the break-of-slope need to be stabilised against expansion of bare ground and soil erosion through fencing to exclude stock and through establishment of salt tolerant trees and grasses: e.g., River Red Gum (E. camaldulensis), tall wheat grass (Agropyron elongatum), phalaris (Phalaris aquatica) and tall fescue (Festuca arundinacea).


98

In the piedmont zone and adjacent Violet Town Plain general best practice is aimed at leakage control by native woodland and perennial pasture. Given that soils here are relatively acidic − the result of weathering of the volcanic rock and of alluvium sourced from same − lucerne (Medicago sativa) may not be an appropriate perennial pasture. Mixed pastures of tall wheatgrass, clover and chicory may be more suitable if weeds can be controlled, interspersed with appropriate eucalypt species. Alley farms, with rows of trees alternated with grassed swathes may be an appropriate farm forestry option on the plain. The CLPR (2001) report indicates that the Violet Town Plain is suited to Red Ironbark (E. tricarpa) plantations and parts of the plain are suited to Sugar Gum plantations. Some parts of the plain are also moderately suited to Eurabbie and Mugga plantations. The medium rainfall Spotted Gum (E. maculata), which produces good timber and has less of a tendency to sacrifice pasture beneath the canopy, thereby permitting grazing up to the trunks, may be an appropriate species for the area (W. Hill, Warrenbayne landholder, pers. comm., 2003). At Burke’s Flat in the 600 mm rainfall zone in the Victorian uplands, conversion of the catchment to perennial pasture combined with the use of trees along rocky ridges has successfully controlled salinity although it is regarded as unlikely that the use of trees alone would have reduced salinisation of the lower landscape (Reid, 1995; Reid et al, 1997).

5.4.2. Caniambo Hills The Caniambo Hills represent a second major recharge zone in the study area, albeit a far less dynamic one than the Strathbogie Ranges. Both recharge and discharge processes here are slow because of the substantial depth and nature of the aquifers and the low hydraulic gradient. The fact that widespread ancient groundwaters are resident, as identified by application of the 14C dating technique, gives an indication of the sluggishness of the Caniambo system. Notwithstanding these distinctive characteristics of the bedrock aquifers, the watertable beneath the hills appears to be well-connected to the watertable in the adjacent Shepparton Formation aquifers of the surrounding plains and recharge through porous lithosols on the hill crests and via the deep fracture networks in the bedrock ultimately translates to discharge at the break-of-slope. It is emphasised that only the water balance can be manipulated here: reducing recharge in order to reduce discharge. The salt imbalance is a much more intractable proposition. Ancient salt stores bound up in weathered fractured bedrock can not be flushed from the system except in geologic time-scales. The best bet is to acknowledge its presence and its intractability and to limit the quantities of recharge waters infiltrating through the thick salt-laden unsaturated zone, i.e., to keep water and the salt stores separate from each other. Possibly >50% of the area of the Caniambo Hills may need to be forested to reduce recharge to near pre-clearing levels and to have large effects on salt export, to lower the watertable and avert saline discharge north of the hills, at the break-of-slopes and in down-gradient swamps and stream channels in the Riverine Plain. This estimate, of >50% reafforestation with appropriate species, is based on comparable studies in analogous areas where modelling has been conducted to simulate the effects of land management scenarios on the water and salt balance of investigated catchments. For


99

example, recent modelling results in the South West Goulburn region (CLPR, 2003), where weathered fractured Silurian bedrock is dominant, indicate that >50% of the catchments need to be forested to control recharge. Modelling of a local GFS of weathered fractured bedrock at Kamarooka near Bendigo where there are low to moderate gradients and break-of-slope salinity (Hekmeijer et al., 2001; National Land and Water Resources Audit, 2001) may be relevant to Honeysuckle catchment. At Kamarooka, land use change to reduce recharge by 50% is predicted to lower the watertable within 100 years, resulting in a slight (10%) reduction of salt-affected areas. More extensive revegetation and improved farming practices to reduce recharge by 90% is predicted to result in a 40% decline in the area affected by shallow watertables in 50 years, with elimination within 100 years (Hekmeijer et al., 2001; National Land and Water Resources Audit, 2001). Increased soil depth allows a greater range of vegetation options to control leakage. The soil thickness and available water capacity plots provided in Section 4.9 are aimed at guiding site selection for revegetation. In all cases in the upper parts of the Caniambo Hills, trees will be harvesting rain for growth rather than drawing upon the watertable which is generally too deep and too saline. According to the CLPR (2001) assessment, the Caniambo Hills have a high suitability for Red Ironbark (E. tricarpa), a high to moderate suitability for plantations of Sugar Gum, and a moderate suitability for Mugga plantations. Eleven-year old Red Ironbark (E. tricarpa) plantations at ‘Coomalong’ in the upper Sheep Pen Creek sub-catchment in the Caniambo Hills are being monitored as part of the Heartlands Initiative (Landowner: Jack Freewin). These agroforestry trials are aimed at optimising tree densities for commercial return in low rainfall (<700 mm/year) catchments of the Murray-Darling Basin (Falkiner et al., 2004). The productivity of significant areas of trees in the Caniambo Hills may be curtailed by progressive concentration of salt in the root zone because of the typical high salt stores in the fractured bedrock of the Caniambo Hills. In this regard, the Heartlands trials may prove crucial to future farm forestry in the Caniambo Hills and analogous areas, given that these are now in their twelfth year. Ground geophysics may elucidate the presence of salt stores in prospective sites for future establishment of farm plantations and tree belts so that lower conductivity − i.e., either better flushed or less kaolinitic − areas are selected before trees are planted. Loss of water yield from the medium rainfall zone of the Caniambo Hills through establishment of new plantations and tree belts will mainly affect Sheep Pen and Irish creeks. These and their associated swamps are, in any event, ephemeral waterways under the present climatic and land use conditions. The impact of reafforestation here would be far less than potential water yield compromises in the Strathbogie Ranges sub-catchments with respect to the perennial creeks and associated wetlands in the catchment.

5.4.3. Riverine Plain The Riverine Plain section of the study area is part of a large regional GFS which makes it very difficult to manage. Shallow saline watertables limit the revegetation


100

options in the plain. Remnant trees, in places, are already showing signs of salt toxicity although along the Broken River and main creeks River Red Gum thrive because of access to fresh streamflow derived largely from distant uplands. The Albacutya type of River Red Gum, which originates from Mallee salt lakes, has a particularly high salt tolerance and may be appropriately established in some parts of the Riverine Plain here. According to the CLPR (2001) report, parts of the Kialla East – Caniambo plains are moderately suited to the establishment of Mugga (E. sideroxylon) and Broombush (Melaleuca uncinata) plantations. Recharge in the plain is mainly via direct or diffuse processes to the shallow watertable, although some lateral flow of ancient groundwaters is received from adjacent bedrock aquifers beneath the Caniambo Hills. Revegetation in the plain needs to be aimed at mining the watertable and drying out the soil profile rather than intercepting rainfall or runoff. It should be noted, however, that only local improvement can be expected given the broad-scale link with a vast alluvial groundwater system. Mitigation measures here are more a case of ‘buying time’ to enable continued cropping and grazing of the plain from one decade to the next, rather than influencing the hydrologic behaviour of the system. Lucerne and salt tolerant perennial grasses, phalaris, tall wheat grass, tall fescue and puccinellia, and salt tolerant annual clover may be appropriate towards drying out the upper metres of soil and improving overall soil condition. Lucerne may be best suited to shoestring sand lenses if local recharge control is desired. The suggested salt tolerant species are better suited to discharge control in vulnerable low-lying parts of the landscape. The productive capacity, high water use and watertable control potential of perennial pastures, particularly lucerne, has been well documented over the past decade. Lucerne trials in the Kialla East plain have demonstrated recharge reductions of 20-100 mm/year, with the potential to use up to 395 mm/year (Day and Harvey, 1994) although this appears to be a temporary response (M. Reid, Victorian DPI, pers. comm., 2003). Recent CSIRO trials in the Riverine Plain at Wagga Wagga, NSW, have shown that lucerne is effective at drying up the soil profile to at least 3 metres and in some cases more than 5 m depth, even in soils where annual crops only root to less than 1 m. The lucerne phase removed 168 mm of water from below the rooting depth of annual crops, water that would have otherwise been lost to production and contributed to groundwater recharge (Verburg and Bond, 2003; W. Bond, CSIRO Land & Water, unpublished data, 2004). It is not assumed here that this capacity for lucerne to dry up the soil profile can impact on groundwater level within the broader-context; it might best be regarded as a paddock-scale ‘band-aid’ treatment wherever the GFS is regional in extent. Maintenance of soil structure and soil health are important in the plains, particularly given the preponderance of Vertosols and Sodosols. Conservation farming practices involving perennial phases in cropping, minimum tillage and stubble retention are appropriate. Engineering strategies involving groundwater pumping or drains and disposal basins, to intercept saline groundwater before it reaches valuable agricultural and environmental assets, may be appropriate for parts of the plain. Engineering options for the management of dryland salinity are described in detail in a report by the Land


101

and Water Resources Research and Development Corporation (2001). These measures may, however, be prohibitive because of high costs and potential off-site environmental impacts.

Figure 51. Generalised suggestions for recharge reduction, watertable control, salinity management, and improved catchment health.


102

5.5 Utility of the Airborne Geophysics datasets The FRRR project in the Honeysuckle Creek catchment has utilised high resolution airborne geophysical data − magnetic, radiometric and electromagnetic − in conjunction with conventional datasets and fieldwork towards the project aim of understanding and ameliorating the problem of salinity and to aid planning for sustainable land use in the catchment. Concluding comments here relate to the advantages and limitations of these datasets with respect to the FRRR project objectives, not to cost effectiveness in terms of the project nor in terms of potential similar applications in study areas elsewhere.

5.5.1. Airborne magnetic data The TMI coverage of the study area provides detailed information about the sub-surface geology relevant to reconstruction of the catchment architecture and buried features such as prior drainage networks where they are defined by the presence of magnetic gravels. Magnetic data does not confer any direct information about the presence or absence of water or salt. Major magnetic features in the survey area clearly relate to different basement blocks and associated structural elements − the upthrust block of Silurian metasedimentary bedrock that make up the Caniambo Hills, the adjacent Honeysuckle Graben which underlies the Violet Town Plain, and linear structures representing faults and fractures within the various bedrock units. These distinctive features, and their bearing on the disposition of present-day major landscape elements, are best interpreted when the TMI image and DEM for the area are viewed in conjunction. Linear magnetic features both within the Caniambo Hills and extending beneath the onlapping Riverine Plain are taken to be faults and fractures. These are potentially significant zones where both groundwater recharge and discharge may be more readily promoted compared to surrounding rock and sediment, depending upon where the linear features intersect the ground surface. On hillcrests and upper slopes such intersections may represent high recharge zones; this assumes that the inferred faults and fractures are not in-filled with occluding secondary materials. Down-slope, such intersections are potential discharge sites, particularly where the base of the hills, the edge of the plain and the watertable are juxtaposed. Many of the magnetically-defined channels in the study area are coincident with the tributary systems of contemporary drainage systems. This is particularly the case on the upper slopes of the Caniambo Hills where the gravels are found to be sourced from exposures of the ancient, chemically differentiated sedimentary bedrock. Elsewhere, some of the magnetically-defined channels are palaeochannels in that they are buried and do not relate to contemporary courses of the drainage network. In fact, the difference is offset some 90º in the Riverine Plain where the magnetic palaeochannels trend northwards and the present-day creeks and Broken River flow westward. Initially, when this high resolution TMI image was first released, in 2001, various workers in multi-agency workshops suggested that the conspicuous northward-


103

flowing palaeochannels may function as preferential conduits for northward saline groundwater flow, potentially directly towards the River Murray. This is a fairly overt first impression gained from the image. We countered this preliminary interpretation by using detailed fieldwork and plotting up available groundwater elevations, and by bringing in broader regional-scale airborne magnetic data. Groundwater flow within the Riverine Plain part of the study area is westward, not northward. Further, the palaeochannels beneath the Riverine Plain are not necessarily preferential flow paths for saline groundwater because our drill-hole data has revealed that all substrate beneath the flat landscape − buried channels and intervening buried interfluves alike − is saturated and the watertable is consistently flat and very shallow (<5 m, commonly <2 m BG) within the overall plain. The high resolution TMI is a revealing dataset, however, caution must be exercised to translate this detailed information directly to an interpretation of the hydrogeologic behaviour of a given system in the absence of substantiating data.

5.5.2. Airborne radiometric data The radiometric image of the Honeysuckle study area reveals pronounced partitioning of the surface landscape in terms of its geology, geomorphology, mineralogy and texture. This dataset detects only the uppermost 20-30 cm of the ground surface, so its application is limited to the distribution of exposed bedrock, weathered regolith and alluvial blankets and the uppermost soil layer. Radiometric data do not confer any direct information about the presence or absence of water or salt. The radiometric image has useful application for mapping the near-surface geology, regolith and soils, and the distribution of alluvium between source areas and down-catchment depositional positions. Thus, contemporary surface hydrologic patterns can be inferred. This mapping is best conducted in conjunction with a DEM and geology map of the area. Substantial field observations are required to translate the gamma-ray data to specific soil types because numerous factors other than the presence or absence of the radioisotopes, K, Th and U, distinguish different soil types. One correlation has been noted above, the association between Grey Vertosols at the edge of the Riverine Plain, and salinity, in the lower reach of Sheep Pen Creek, Downes Swamp and the centre of Earlston sub-catchment. These sites were mapped by conventional soil mapping techniques some 50 years ago as ‘Upotipotpon Clay’ (Downes, 1949), the equivalent to Grey Vertosols. The present study interprets the soils in these particular low-lying areas as thick deposits of palaeoswamp clay. The sites are seen as thorium-rich, or potassium-poor, in the radiometric image. However, not all Th-rich, K-poor areas mapped as Grey Vertosols are associated with salinity, other such areas in the radiometric image are known to be well-flushed or are elevated many metres to tens of metres above the watertable. The mentioned saline Grey Vertosol areas occur in topographic low points where the watertable is very shallow, thus, it is the topographic position and the depth to watertable driving salinisation rather than the predisposition of a given soil type. For the most part within the Honeysuckle study area there is no strong correlation between salinity and a specific soil type or the distribution of radioelements. It must be reiterated that the radiometric data do not reflect salt or water in the landscape; the usefulness of the imagery lies in the information conveyed about the surficial


104

landscape of the catchment and, potentially, to agricultural land use. An excellent investigation of the application of radiometric data for mapping soils and land degradation is documented by Bierwirth (1996).

5.5.3. Airborne Electromagnetic data Following are the most significant aspects of and findings from this research in the Honeysuckle catchment, with emphasis on the AEM dataset which was used in conjunction with substantial fieldwork and several other datasets. Both advantages and limitations with the utility of the AEM stemming from our experience are highlighted. Included are comments regarding whether or not the AEM was used to reveal a key result or interpretation relevant to the objectives of the project. The research approach has been a highly integrated one directed towards a specified set of objectives in the catchment. It has not been aimed at assessing the AEM technology per se or towards predicting its future utility or potential adoption in new geographic regions.

• The reprocessed 2002 CDI datasets, generated after calibration of the AEM data with down-hole EM-39 conductivity measurements, provide a much more realistic representation of the conductivity distribution in x, y and z-space within the survey area than the original uncalibrated 2001 CDI datasets.

• An extensive highly conductive area occupying the near-surface of the Violet

Town Plain that was conveyed in the uncalibrated shallow CDI (2001) has been greatly subdued through the calibration and reprocessing exercise.

• Intricate networks of drainage channels are well-defined in both the TMI and

AEM datasets. Co-occurrence of high magnetism and high conductivity relate to the nature of sediments infilling the channels, magnetic gravels and saline kaolinitic clays, respectively. The channels (as depicted by high TMI and high AEM) correspond with both contemporary drainage networks incised into the higher terrain and contiguous down-gradient reaches of the same systems that are now buried beneath onlapping Riverine Plain sediments of the Murray Darling Basin. The latter buried ‘palaeochannels’ trend northward, contrasting with the present-day drainage in the plain, which flows westward.

• The excellent spatial correlation of the two datasets, TMI and AEM, in plan

view (x-y space) does not hold in the z-dimension. Magnetic gravels represent <5% of the channel infill which is dominated by saline kaolinitic clays. Down-hole conductivity profiles (EM-39 measurements) reveal low conductivity and low salt stores (EC1:5 measurements) associated with concentrations of the magnetic gravels, and high conductivity and high salt stores in underlying and overlying alluvial clay deposits.

• The highly conductive saline clay material that corresponds with the magnetic

drainage network in x-y space is apparent in the 5 to 40 m CDIs. In plan view this anomaly coalesces beneath the Riverine Plain in the 10 to 25 m CDIs, and contracts substantially between 25 to 40 m CDIs. Resistive substrate is present at depths greater than 40 m.


105

• It is emphasised that the respective saturated and unsaturated zones are not discriminated in the AEM data for the highly conductive saline clays in the magnetic channel networks. Closer to the hill crests the conductivity highs are found to correspond with dry substrate and down-gradient they relate to saturated sediments immersed in saline groundwater. Only drill-hole data and groundwater measurements can discriminate between the saturated and unsaturated zones within the substrate.

• This limitation, an identical AEM signature for saline clays above the

watertable and saline groundwater, is relevant to assessing the mobility of observed salt stores. It is necessary to sink as many drill-holes as possible to ascertain the depths and gradient of the groundwater system before the AEM data can be sensibly interpreted.

• In the northern part of the survey area, beneath the Riverine Plain, the

watertable is shallow and almost flat, < 5 m and commonly < 2 m BG. Thus, sediments in both the buried channels and intervening buried interfluves that make up the ‘palaeotopography’ beneath the flat, wholly infilled landscape are all saturated. The conspicuous magnetically-defined palaeochannels therefore do not necessarily represent preferential groundwater flow paths.

• Contouring of the elevation of the watertable indicates that the groundwater

flow direction beneath the Riverine Plain is westward, not northward. Therefore, the impression conveyed by the AEM and TMI imagery, dominated by north-trending palaeochannels, is illusionary and does not relate to the hydrogeologic behaviour of the system.

• That northward throughflow of saline groundwater does not occur in the

Riverine Plain is reinforced by broader-scale regional TMI imagery. The latter reveals the presence of a very large near-surface bedrock body contiguous with the outcropping Dookie Hills which would occlude northward groundwater flow much beyond the latitude of the Broken River and would promote westward groundwater flow.

• The shallow CDIs, 0 to 10 m depth, reveal a blanket of ‘fresh’ alluvium along

the northern edge of the survey area. This low conductivity package corresponds well with the distribution of potassic-rich alluvium in the Broken River floodplain that is revealed in the radiometric image. The correlation suggests that the blanketing fresh alluvium is sourced from outside the immediate catchment. The AEM data indicates the extent and depth of flushing of floodplain sediments by episodic floodwaters. Below 15 m depth, the distribution of highly conductive sediment relates to coalesced deposits of clays derived originally from ancient weathered, eroded sediments of the Caniambo Hills.

• Moderate conductivity beneath the Violet Town Plain in approximately the 10

to 30 m CDIs is not fully resolved but does not appear to relate to saline groundwater except perhaps between 20 - 40 m BG.


106

• The ‘Elevation to the Base of the Conductor’ image derived from the original AEM datasets reveals resistive substrate at the western end of the Violet Town Plain that gives the impression of possible hydrologic closure of the Honeysuckle Graben by a buried ridge of resistive bedrock. Drilling here refuted this interpretation and revealed well-flushed granitic gravels as the source of the resistive AEM signature. This indicates that the ‘Violet Town Sump’ to the east is not wholly enclosed.

• Hydrographs for this groundwater system in the Honeysuckle Graben reveal

steadily rising watertables and increasing salinities. These watertables are presently at 20-30 m BG, so there is no immediate threat to the landscape. Drilling based on the AEM depth of the conductor indicates that the Violet Town Plain is unlikely to be overtaken by a rising saline watertable. Once the watertable rises to a threshold approaching 20 m BG, groundwaters are expected to exit the sump westward to aquifers beneath the Euroa Plain and, potentially, towards the Goulburn River.

• Drilling in the area immediately south of the Hume Freeway near Baddaginnie

has established that conductivity highs in the 5 to 20 m CDIs relate to saprolitic outliers of Violet Town Volcanics, not to saline groundwater. Observed waterlogging here relates to surface water, not shallow groundwater. Waterlogging in the piedmont plain is the consequence of excessive runoff from the cleared Strathbogie Ranges and occluded throughflow northwards because of bedrock outcrops. Embankments of the Freeway and railway line may also be contributing factors, impeding northward flow.

• Saline groundwaters in weathered, fractured metasedimentary rocks and salt

stores in the overlying unsaturated zone of the Caniambo Hills are not detected well in the AEM datasets. This is regarded as the most serious limitation with the AEM for the survey area because the salt stores are considerable and saline groundwater is pervasive. The presence of salt stores in the unsaturated zone was assessed through drilling and laboratory analysis. The extent and severity of groundwater salinity were revealed through construction of a groundwater salinity contour map from bore data which strongly correlates highly saline groundwater with the fractured bedrock GFS. The lack of registration of these resident salts − in both the saturated and unsaturated zones in the Caniambo Hills bedrock system − in the AEM is attributed to the low water/rock ratio, the fine-scale of the fracture networks where salt and saline groundwater are stored, and generally low inter-granular porosity. This situation corroborates that of Edwards and Webb (2003) in the analogous Kamarooka area where the distribution of TEMPEST AEM conductivity was found to have little in common with the distribution of saline groundwater. The latter finding was attributed to low inter-granular porosity of the bedrock and the inability of the AEM to detect pore water salinity in such material. Onlapping sediments with higher porosity than the bedrock gave higher conductivities generally although the AEM failed do enormous variation in groundwater salinities (Edwards and Webb, 2003).

• The AEM data could not be relied upon to elucidate the most important

hydrogeologic aspects relating to groundwater and salinity in the survey area.


107

Our earlier reconstruction of the watertable in adjacent aquifer systems (Cresswell et al., 2003) indicates a high degree of connectivity between the bedrock aquifers of the Caniambo Hills and the alluvial aquifers of the Riverine Plain to the north. This translates to saline seepage at the break-of-slope between the hills and the Riverine Plain despite the low hydraulic gradient.

• Construction of a watertable map and inferred groundwater flow directions,

combined with 14C dating of resident groundwaters, reveal that the system − although connected with adjacent GFS − is very sluggish. The bedrock aquifer system appears to be at its full capacity and may be approaching a measure of hydrologic equilibrium, with discharge approximately equal to recharge under current conditions. Although palaeowaters are resident in the bedrock aquifers and low hydraulic conductivities and low hydraulic gradients are represented, discharge occurs at the base of the hillslopes because of the current watertable levels. Pulses of recharge appear to translate to pulses of discharge in down-gradient settings, indicative of hydrologic balance within the system. This vital information is not supplemented by the airborne geophysical data but, rather, is wholly dependent upon field observations, conventional data and system understanding.

• That recharge and discharge pulses between the Caniambo Hills bedrock

aquifers and the onlapping alluvium of the Riverine Plain is linked to preferential flow paths afforded by major faults and fracture zones is inferred but not proven.

• Construction of the groundwater salinity contour map has proven more useful

than the AEM at representing the overall distribution of fresh to saline groundwater in the catchment.

• Break-of-slope salinity is a relatively surficial phenomenon and it is possible

that contributing conductivities in the near surface, such as plant roots, may interfere with the AEM signal for the uppermost CDI. Thus, even though break-of-slope salinity is linked to broader, deeper causative processes, the actual saline seepage sites may be small, scattered and very shallow, underlain by a non-saline system. AEM data should not be relied upon to detect or predict the occurrence of break-of-slope salinity where its manifestation is shallow and patchy. Ground-based datasets are most appropriate for mapping the extents of and predicting future outbreaks or expansions of these relatively dynamic saline zones. More importantly, the tendency for shallowing groundwaters near any break-of-slope zone needs to be detected a decade or a few decades before salinity develops, i.e., long before any remote sensing technique is capable of registering conductivity changes wrought about by the build-up of efflorescent salts at the exposure points and before the ensuing cycle of seasonal flushing of those salts is set into motion.

• Discharge of saline groundwater directly to the trunk drainage channels and

subordinate waterways in the Riverine Plain is a process involving the third dimension, namely the depth of incision of the main channels, a feature that is not represented in any remotely-sensed dataset. Despite low hydraulic gradients,


108

saline groundwater seeps from shallow alluvial floodplain aquifers into stream beds by virtue of the considerable depth of incision of the channels, not necessarily because of any substantial driving hydraulic pressures. Understanding this crucial manifestation of salinity is best attained through geomorphologic knowledge and an appreciation of pre- and post clearing watertable levels.

• Major salinity sites in the catchment, including salinised swamps and the

location of a slight increase in the salt load in the Broken River, are associated with bedrock/topographic constrictions. Saline groundwater is forced upwards to the landscape surface because of constricted flow paths, to emerge in vulnerable topographic low points. The constrictions are best revealed in topographic and geologic maps.

• Shallow AEM highs correspond with Grey Vertosols and known salinity at the

landscape surface in the lowlands of Sheep Pen Creek and Earlston sub-catchments. However, the correspondence does not hold elsewhere. Grey Vertosols are widespread in the eastern Violet Town Plain and also on the plain located between Honeysuckle Creek and the hills north of Miepoll. These areas do not correspond with pronounced shallow AEM highs or with mapped salinity. AEM highs correspond with soil types other than Vertosols, and known salinity sites correspond with a variety of soil types. For example, the AEM highs within the hills north of Miepoll are in areas of Sodosols, and mapped salinity areas elsewhere in the catchment correspond with Sodosols (at Kialla East and in the eastern part of Earlston sub-catchment) or with Kandosols and Kurosols (in the Strathbogie Ranges). An appreciation of topographic position, geology, palaeoenvironmental evolution and depth and salinity of groundwater are deemed more important to understanding the occurrence of salinity at the landscape surface than characteristics that relate to bulk conductivity patterns or specific radiometric signals.

In conclusion, it is emphasised that our use of the AEM data has conferred some advantages and benefits to our understanding of the Honeysuckle Creek system. However, the AEM data has failed to assist in the recognition of: (a) the most extensive occurrence of highly saline groundwaters in the catchment, i.e., in the weathered fractured bedrock GFS of the Caniambo Hills; (b) the most dynamic saline zone, i.e., the Strathbogie Ranges break-of-slope sites; and (c) landscape positions that are actively being salinised which relate to very straightforward hydrologic-geomorphic relationships. Moreover, temporal variations and fluxes that are relevant to the hydrogeologic behaviour of the catchment are not accommodated in the AEM data which presents a static ‘snapshot’ of the distribution of the represented range of conductivities. In the case of break-of-slope salinity, even if the AEM technology is refined to accurately map these shallow outbreaks, such detection is likely to be too late because the processes causing discharge and salt efflorescence were probably set into motion decades before their manifestation. Therefore, drilling and monitoring groundwater levels and salinities long before the emergence of break-of-slope salinity is more crucial than the capacity to map their appearance after the fact.


109

Our recent work in Simmons Creek sub-catchment, in the Upper Billabong System, NSW (English et al., 2002) provides a counterpoint to the above assessment. The Simmons Creek groundwater and salinity investigation was conducted and completed in 2002 before the TEMPEST AEM survey data became available for our Heartlands effort in this catchment. We utilised available airborne magnetic and radiometric data to advantage in Simmons Creek catchment. However, our conceptual understanding of the key groundwater and salinity processes was constructed without reference to AEM data. Ultimately, with the subsequent release of the AEM coverage for the area, our groundwater salinity contours for the upper aquifers were found to closely match the distribution of a shallow conductivity high that dominates the 0-20 m CDIs of the catchment. This high degree of correlation between the AEM data and groundwater contour maps generated from drilling and field measurements is salutary, at least for the Simmons Creek case study, and potentially for analogous areas. On the one hand it endorses the veracity of the AEM data in some terrains, i.e., high bulk conductivity can represent shallow saline groundwater and overlying saline substrate. On the other hand the correlation endorses the credibility of conventional hydrogeologic techniques for providing a ‘stand alone’ approach to groundwater and salinity investigations. Only through drilling and field measurements and contouring of the latter data could the nature of the conductivity high −- unsaturated versus saturated layers − and the dynamics of the groundwater and salinity processes be established. It is reiterated that AEM data represent aggregated conductivity and there is no differentiation between high conductivity in the unsaturated zone and high conductivity related to saline groundwater for a given data point in x-y-z space. The general impression conveyed from our three years of work in the Billabong and Honeysuckle catchments is that reliable groundwater and salinity research leading to appropriate on-ground action can be conducted in the absence of AEM data, but not vice versa. This does not dismiss the potentially positive and beneficial application of the AEM technique in diverse hydrogeologic provinces nor does it ignore future refinements of the technology that should bring wider applicability to dryland salinity investigations. The present study emphasises the importance of integrating the use of all remotely sensed datasets with drilling and field and laboratory analysis, and with as many supporting datasets as can be acquired for a given area under investigation. This is a further call for a combined approach to salinity hazard mapping if airborne geophysics are adopted for management of dryland salinity. It endorses earlier reviews relating to the diagnosis of causes, the prognosis of trends, and the delineation of treatments of salinity, e.g., George et al. (1998); George and Woodgate (2002); and Spies and Woodgate (2004a, 2004b). System understanding, of given catchments, and process understanding − of the underlying causes and mechanisms of landscape and waterway salinisation in widespread regions − are important. The present study particularly emphasises the need to establish the relationships between landscape and hydrology in three dimensions, and to then incorporate temporal fluxes and variations that relate to climatic drivers. Secondly, the status quo with respect to hydrologic equilibrium or disequilibrium and salt equilibrium or disequilibrium needs to be established for any system under investigation. Understanding the balances between recharge and discharge and between salt input and salt output and scrutiny of salt load and salinity trends in our streams are here regarded as more important than static mapping of the distribution of substrate conductivities in a given system.


110


111

References Allan MJ, 1994. Method for assessing dryland salinity in Victoria. Centre for Land Protection Research Technical Report 34. Department of Natural Resources and Environment, Victoria. Bierwirth P, 1996. Investigation of Airborne Gamma-ray Images as a Rapid Mapping Tool for Soil and Land Degradation − Wagga Wagga, NSW. Australian Geological Survey Organisation Record 1996/22. Blackburn G and McLeod S, 1983. Salinity of atmospheric precipitation in the Murray-Darling Basin drainage division, Australia. Australian Journal of Soil Research 21: 411-434. Brodie RC, 2002. Honeysuckle Creek Airborne Geophysical Surveys, Metadata. In Dent DL (Editor). MDBC Airborne Geophysics Project - Draft Final Report, Consultancy D2018, December 2002. Appendix 2: 14-25. Burns PR, 1988. Land capability study in the Shire of Violet Town. Victorian Department of Conservation, Forests and Lands. Butler BE, 1956. Parna in the Riverine Plain of South-Eastern Australia and the soils thereon. Australian Jounal of Agricultural Research. 7: 536-553. Butler BE, 1958. Depositional systems of the Riverine Plain of south-eastern Australia in relation to soils. Soil Publication No. 10, CSIRO Australia. Butler, RF, 1992. Paleomagnetism: Magnetic Domains to Geologic Terranes. Blackwell Scientific Publications, Boston. pp. 319. Cartwright I, Fulton S, Nichol C, Weaver T, Reid M and Cheng X, 2002. Hydrogeology, hydrogeochemistry and environmental isotopes as tracers of dryland salinity in the Honeysuckle Creek region, near Benalla, Victoria. In Phillips GN and Ely KS, Victoria Undercover, Collaborative Geoscience in Northern Victoria, Benalla 2002, Conference Proceedings and Field Guide, pp. 249-259. Cheng X, 1999. Goulburn-Broken Dryland Catchment Prioritisation - Upland sub-catchments. Centre for Land Protection Research, Victorian Department of Natural Resources and Environment, Technical Report 58. Christensen A, 2002. Calibration of Electromagnetic Data. In Dent DL (Editor). MDBC Airborne Geophysics Project - Draft Final Report, Consultancy D2018, December 2002: 20-38. Christensen A, 2003. Calibration of Honeysuckle Creek Conductivity Depth Imaging. Australian Society of Exploration Geophysicists Preview: 27-30. Christensen A, 2004a. Calibration of Honeysuckle Creek conductivity depth imaging. Geological Survey of Victoria Unpublished Report 2004/1.


112

Christensen A, 2004b. Interpretation of the Honeysuckle Creek airborne geophysics – Interim Report. Geological Survey of Victoria Unpublished Report 2004/4. Clarke J, 1999. Local Area Plan for Mid Goulburn Area – Sheep Pen Creek Section. Draft Report. pp. 24. CLPR, 2001 (Compiled by X. Cheng and M. Reid). Targeting suitable areas for farm forestry in the Broken and North Goulburn Plains and surrounding uplands. Centre for Land Protection Research, Victorian Department of Natural Resources and Environment, Technical Report 65. CLPR, 2002. Goulburn Broken Groundwater Flow Systems 1:750 000 map. Centre for Land Protection Research, Victorian Department of Natural Resources and Environment (18/3/2002). CLPR, 2003. South West Goulburn Salinity Study. Department of Primary Industries. Centre for Land Protection Research Report No. 28. Craig AJ, 1988. Identifying preferential recharge areas in the Earlston sub-catchment, Victoria. Thesis. La Trobe University (unpublished). Cresswell H, English P, Zhang L, Gallant J, McKenzie N, Austin J, Gregory L, Glover M, Richardson P, Paydar Z, Green T, Hickel K, Palmer R, Best A, Ive J, Abel N, Hill P, Langston A, Cawsey M, Langridge J, Ryan P, Plant J, Jovanovic T, Booth T, Marcar N, Myers B, 2003. Heartlands – Planning for Sustainable Land Use and Catchment Health: A report of the Heartlands Initiative. CSIRO – Murray-Darling Basin Commission Technical Report No. 1. http://www.clw.csiro.au/heartlands/publications/general_publications.html Cresswell HP, Ive JR and Myers BJ, 2002. Heartlands − supporting targeted land use change and evaluating environmental and commercial outcomes. In Phillips GN and Ely KS, Victoria Undercover, Collaborative Geoscience in Northern Victoria, Benalla 2002, Conference Proceedings and Field Guide, pp. 205-206. Cresswell R, 2002. Laboratory procedure and geochemistry notes. In Dent DL (Editor). MDBC Airborne Geophysics Project - Draft Final Report, Consultancy D2018, December 2002: 39-45. Dahlhaus PG, Reid M, Cheng X, 2000. Broken and North Goulburn Plain Salinity Study - Salinity Processes Summary. Centre for Land Protection Research, Victorian Department of Natural Resources and Environment, Technical Report 66. Daamen CC, Hoxley GP, Collett KO, Patrick KJ and Shepard P, 2002. Evaluation of land-management options for dryland salinity: the next step after airborne geophysical interpretation. Exploration Geophysics 33: 127-135. Day C, Sinclair J, Dyson P, 1990. Dryland Salinity Processes at Sheep Pen Creek (unpublished thesis draft).


113

Day C and Harvey W, 1994. Groundwater Processes and Salinity Risk at Kialla East, North East Victoria. Centre for Land Protection Research, Victorian Department of Conservation and Natural Resources, Technical Report 15. Dent DL, 2002. Interpretation: Honeysuckle Creek. In Dent DL (Editor), 2002. MDBC Airborne Geophysics Project - Draft Final Report, Consultancy D2018, December 2002: 60-71. Dent DL (Editor), 2002. MDBC Airborne Geophysics Project - Draft Final Report, Consultancy D2018, December 2002. Dent DL, Munday TJ, Brodie RC and Lawrie KC, 2002. Implications for salinity and land management - Honeysuckle Creek, Victoria: a preliminary interpretation of high-resolution airborne geophysical data. In Phillips GN and Ely KS, Victoria Undercover, Collaborative Geoscience in Northern Victoria, Benalla 2002, Conference Proceedings and Field Guide, pp. 223-233. Dowling T, Dawes W, Evans R, Dyson P, Walker G, 2004. Prioritising upland catchments in the Murray-Darling Basin with respect to salinity benefits from afforestation. CSIRO Land and Water Technical Report No. 15/04. Downes RG, 1949. A Soil, Land-use, and Erosion Survey of Parts of the Counties of Moira and Delatite, Victoria. CSIRO Bulletin 243. Edwards MD and Webb JA, 2003. Ground-truthing of a TEMPEST airborne electromagnetic survey in the salinised Kamarooka Catchment, near Bendigo in central Victoria. In: Roach IC (Ed.). Advances in Regolith. CRC Landscape Evolution and Mineral Exploration Symposia: Extended Abstracts, pp. 110-114. English P, 1988. Magnetic patterns in granites of the Lachlan Fold Belt. The Australian National University. Unpublished report. English P, Richardson P, Glover M, Cresswell H, Gallant G and Oswald B, 2004. Interpreting and Implementing Airborne Geophysics. CSIRO Land & Water Client Report for the Foundation for Rural and Regional Renewal. English P, Richardson P, Stauffacher M, 2002. Groundwater and Salinity Processes in Simmons Creek sub-catchment, Billabong Creek, NSW. CSIRO Land and Water Technical Report 24/02 http://www.clw.csiro.au/publications/technical2002/. Falkiner R, Theiveyanathan T, Marcar N, Myers B, Stewart L, (2004). Low to medium rainfall farm forestry. A report of the Heartlands Initiative. CSIRO – Murray-Darling Basin Commission Technical Report No. 2. Fisher EM 2003. An investigation of salt mobilisation potential at Shepparton, Victoria, using an integrated approach of geophysics, electrokinetics, and hydrogeochemistry. BSc. Hons thesis (unpubl.) University of Wollongong. Gallant JC and Dowling TI, 2003. A multi-resolution index of valley bottom flatness for mapping depositional areas. Water Resources Research 39/12: 1347.


114

GEONICS Limited (undated). EM-34-3 Survey Interpretation Techniques. Technical Note TN-8. George RJ, 1984. Groundwater systems and dryland salinity in a catchment near Boho, NE Victoria. Proceedings of the Royal Society of Victoria 96: 45-53. George RJ, Beasley R, Gordon I, Heislers D, Speed R, Brodie R, McConnell C, Woodgate P, 1998. The National Airborne Geophysics Project: National Report: Evaluation of Airborne Geophysics for catchment management. Report for the Australian Department of Agriculture, Fisheries and Forestry and the National Dryland Salinity Program. George, RJ and Woodgate P, 2002. Critical factors affecting the adoption of airborne geophysics for management of dryland salinity. Exploration Geophysics 33: 84-89. Gibson D and Wilford J, 2002. Aspects of regolith and landscape of the Strathbogie-Caniambo-Dookie area: the need for interpretation of detailed geophysical datasets in the light of regional data and models. In Phillips GN and Ely KS, Victoria Undercover, Collaborative Geoscience in Northern Victoria, Benalla 2002, Conference Proceedings and Field Guide, pp. 235-247. Heartlands Core Group, 2000. Draft Heartlands Five Year Plan for 2001-2005. A report of the Heartlands Initiative, December 2000, Publication No. HL-01. Hekmeijer P, Dawes D, Bond W, Gilfedder M, Stauffacher M, Probert M, Huth N, Gaydon D, Keating B, Moore A, Simpson R, Salmon L, Stefanski A, 2001. Assessment of Salinity Management Options for Kamarooka, Victoria: Groundwater and Crop Water Balance Modelling. CSIRO Land and Water, Canberra. Hughes, MJ, 2002. Permian sediments of the Honeysuckle Plain Graben, Benalla, Victoria. In Phillips GN and Ely KS, Victoria Undercover, Collaborative Geoscience in Northern Victoria, Benalla 2002, Conference Proceedings and Field Guide, pp. 61-627. Jones G, 2002. Honeysuckle Creek Drilling Program: A Bureau of Rural Sciences Report to Goulburn Broken Catchment Management Authority. Jones G, Baker P, Cresswell R, Macaulay S, Dent D, Jaycock J, Spring J, Plazinska A, 2002. Honeysuckle Creek: Bore construction, sampling and pump testing. Bureau of Rural Sciences Report to Goulburn Murray Water. Kelly MJ, 1994. Sheep Pen Creek High Priority Area: Interim Management Statement. Goulburn Dryland Salinity Management Plan. Victorian Department of Conservation and Natural Resources Report. pp. 45. Land and Water Resources Research and Development Corporation, 2001. Assessment of the Efficacy of Engineering Options for the Management of Dryland Salinity. Final Research Report by the National Dryland Salinity Program and Sinclair Knight Merz.


115

Lane R, Green A, Golding C, Owers M, Pik P, Plunkett C, Sattel D and Thorn B, 2000. An example of 3D conductivity mapping using the TEMPEST airborne electromagnetic system. Exploration Geophysics 31: 162-172. Lane R, Heislers D and McDonald P, 2001. Filling in the gaps − validation and integration of airborne EM data with surface and subsurface observations for catchment management − an example from Bendigo, Victoria, Australia. Exploration Geophysics 32: 225-235. Leaney FW, Herczeg AL and Dighton JC, 1994. New developments for the direct CO2 absorption method for radiocarbon analysis. Quaternary Geochronology (Quaternary Science Reviews) 13: 171-178. McKenzie N, Gallant J, Gregory L, 2003. Estimating Water Storage Capacities in Soil at Catchment Scales. Centre for Catchment Hydrology Technical Report 03/03. http://www.catchment.crc.org.au/pdfs/technical200303.pdf McNeill JD, 1980. Electromagnetic terrain conductivity measurement at low induction numbers. GEONICS Ltd. Technical Note TN-6. McNeil JD, 1986. GEONICS EM-39 borehole conductivity meter – theory of operation. GEONICS Ltd. Technical Note TN-20. Meyers J, 2002. Goephysical overview of the Euroa region, northern Victoria. In Phillips GN and Ely KS, Victoria Undercover, Collaborative Geoscience in Northern Victoria, Benalla 2002, Conference Proceedings. pp. 17-23. Myers BJ, 2002. Agroforestry research in the Heartlands initiative – trials at the Coomalong tree farm, Violet Town, Victoria. In Phillips GN and Ely KS, Victoria Undercover, Collaborative Geoscience in Northern Victoria, Benalla 2002, Conference Proceedings and Field Guide, pp. 213-216. National Land and Water Resources Audit, 2001. Australian Dryland Salinity Assessment 2000: Extent, impacts, processes, monitoring and management options. National Land and Water Resources Audit. NAPSWQ, 2000. National Action Plan for Salinity and Water Quality. Council of Australian Governments. Reid M, 1995. Burke’s Flat − a salinity treatment success story. Proceedings of the Murray-Darling 1995 Workshop, Wagga Wagga. Reid M, Clifton C, and Heislers D, 1997. Dryland Salinity Management in the Victorian Uplands. Groundwater in the balance. Murray-Darling 1997 Workshop, Toowoomba, Queensland 26-28 August 1997, Extended Abstracts: 66-71. Rural Water Commission of Victoria, 1988. Stream salinities and salt loads in the Goulburn-Broken catchments. RWC Investigations Branch Report No. 1988/4. Sinclair Knight Merz Pty Ltd, 1999. Projections of the ultimate salt loads from the Victorian Dryland Catchments to the Murray River. Summary of Method and Results


116

Draft C. Natural Resources and Environment, Agricultural Resources Conservation Land Management. Spies B and Woodgate P, 2004a. Salinity Mapping Methods in the Australian Context. Programs Committee of Natural Resource Management Council: Land and Water Australia and the National Dryland Salinity Program. Technical Report, pp. 146. Spies B and Woodgate P, 2004b. Salinity Mapping Methods in the Australian Context. Programs Committee of Natural Resource Management Council: Land and Water Australia and the National Dryland Salinity Program. User Guide, pp. 33. VandenBerg AHM, Willman CE, Maher S, Simons BA, Cayley RA, Taylor DH, Morand VJ, Moore DH, Radojkovic A., 2000. The Tasman Fold Belt System in Victoria: Geology and mineralisation of Proterozoic to Carboniferous rocks. Geological Society of Victoria Special Publication. Verburg K and Bond W, 2003. Use of APSIM to simulate water balances of dryland farming systems in south eastern Australia. CSIRO Land and Water Technical Report 50/03. Vertessy R, Zhang L and Dawes, WR, 2003. Plantations, river flows and river salinity. Australian Forestry 66/1:55-61. Zhang L, Dawes WR, Walker GR. 1999. Predicting the Effect of Vegetation Changes on Catchment Average Water Balance. Technical Report No. 99/12, Cooperative Research Centre for Catchment Hydrology. Zhang L, Dawes WR, Walker GR., 2001. Response of mean annual evapotranspiration to vegetation changes at catchment scale. Water Resources Research, 37:701-708. Zhang L, Dowling T, Hocking M, Morris J, Adams G, Hickel K, Best A, Vertessy R, 2003. Predicting the effects of large-scale afforestation on annual flow regime and water allocation: an example for the Goulburn-Broken Catchments. Cooperative Research Centre for Catchment Hydrology Technical Report 03/5. http://www.catchment.crc.org.au/pdfs/technical200305.pdf.

Date post:	04-Jul-2020
Category:	Documents
Upload:	others
View:	0 times
Download:	0 times

Land and Water · 2005-10-14 · Land and Water CSIRO Land and Water Technical Report No 18/04...

Documents