Hydrochemical and isotopic characterisation of groundwaters to define aquifer type and connectivity in a subtropical coastal setting, Fraser Coast, Queensland

Abstract

Physico-chemical parameters, major ion chemistry and isotope composition of surface and groundwaters were determined in forested coastal catchments and adjacent coastal plains. Results showed obvious characterisation related to physical and hydrological setting, and highly variable spatial differences reflecting the complexities of these areas. All these coastal waters are dominated by Na–Cl and fall on a common dilution line; hydrochemical grouping is largely due to anionic differences (Cl, SO₄ and HCO₃), although Na and Mg ratios also vary. Six major hydrochemical facies were determined. For groundwaters, compositional differences are largely related to aquifer material and level of confinement; for coastal groundwaters important are tidal effects and proximity to the shoreline. Differentiation for surface waters is mainly by drainage morphology, flow regime plus proximity to the coast. Connectivity between water bodies is reflected by minor base flow to catchment streams, including with flood plain wetlands, but mostly occurs in low-lying zones where there is mixing of fresh and saline water within surface water and subterranean estuaries, or by seawater intrusion enhanced by overuse. Oxygen and hydrogen isotopic data for confined and semi-confined groundwaters along the coast indicates local recharge; fresh surface waters in the elevated catchments are shown to be sourced further inland plus have experienced evaporation.

Introduction

Globally, the impacts of anthropogenic activities on natural waters have become a major issue (de Vries and Simmers 2002; Hem 1992) and more recently, the potential impact of climate change is gaining the attention of hydrological studies especially within coastal zones. Changes in the amount and distribution of rainfall (Taschetto and England 2009; Trewin 2007), rising temperatures (Nicholls 2006) and associated increases in sea level have and will significantly alter the hydrological and geomorphological characteristics of these settings (Church 2006; Ribbe 2006; Pepper et al. 2006; Gräwe et al. 2010; Kim et al. 2009). However, of fundamental importance is the physical character of any setting (Drever 2002) and coastal settings, in particular, can be hydrologically complex. Within coastal settings, there can be substantial variations in geology, morphology and drainage system form and dynamics. Possibly the most significant feature of these settings is that they are at the interface between freshwater catchments and the marine environment, and are often also influenced by tidal processes. Typical of these settings are complex hydrological processes involving recharge, discharge, and multiple mixing environments. Due to these complexities, the links between groundwater and surface waters are usually not well understood, especially because of variable scale subsystems within an overall “system”.

The region of southeast Queensland is the fastest growing urban area in Australia, and from 2006 to 2031, its population is expected to grow from 2.8 million to 4.4 million people. The region covers 22,890 km², and stretches 240 km from the Queensland–New South Wales border in the south to Noosa in the north (Department of Infrastructure and Planning 2010). This population growth will result in substantial changes in land use and impacts to a wide range of natural systems, of note the hydrological and hydrochemical processes, which support a variety of ecosystems. In addition to this, there is the real potential of sea level rise, which in low-lying settings with tidal systems can be substantially impacted by storm surges. To obtain some understanding of these processes in such settings with comparatively low-level modification, we have chosen an area on the Fraser Coast in the southeast Queensland region to conduct an integrated investigation of hydrological systems. Although more sparsely populated than the Brisbane area to the south, the Fraser Coast has many similar geomorphological features and is also under pressure from urbanisation with around 1,500 new dwellings and a population growth of 3 to 4% each year (Department of Infrastructure and Planning 2008).

The focus of this study is the freshwater catchments and the adjoining low-lying coastal plain that hosts tidal estuaries. The estuaries discharge to the marine environment adjacent to the southern part of Fraser Island. The mainland freshwater catchments host plantation forestry, which also extends to the coastal plains. Within the coastal plains, there are also areas of remnant native vegetation and small residential communities located near estuary mouths. This Fraser Coast region is typified by subtropical rainfall, which results in highly seasonal stream flow. The area is located adjacent to the Ramsar-listed Great Sandy Strait, which is of high environmental significance. A number of threatened species (flora and fauna) exist in the marine areas and associated tidal wetlands such as dugong, dolphins, migratory shorebirds, seasonal populations of humpback whales, and rare shrubs (EPA 2005).

Here we consider the hydrological and hydrochemical processes of these coastal catchments to determine the character of the groundwaters, and the connectivity between various water bodies. The aims of this study are to (a) establish the chemistry of natural subsurface and surface waters of this typical setting as a baseline for comparison of future monitoring, (b) investigate the dominant hydrological processes contributing to the hydrochemistry of natural waters, (c) describe the hydrogeological settings in which these waters occur, and (d) determine possible solute transport pathways related to current and future land-use practices. The results of this investigation will be of value to assess the future management of such ecologically sensitive areas. A goal here is to determine the chemical character of the surface and subsurface waters within this coastal setting utilising various physico-chemical parameters, major ion and nutrient concentrations and stable water isotopes (δ¹⁸O, δ²H). The approach of the study is to group waters into hydrochemical facies, identify dominant hydrological processes, and confirm where interaction exists between water bodies.

Background

Elevated nutrients and metals in surface and groundwaters can be due to both natural and anthropogenic influences, which can be complex in coastal settings. Urbanisation, agricultural and other land modifications can have an impact on the quality of both surface and groundwater bodies; introduced solutes can infiltrate the aquifers and can be discharged in coastal waters (Burnett et al. 2006; Church 1996; Gallardo and Marui 2006; Moore 1999; Tirumalesh et al. 2007). Hydrological processes in coastal settings typically occur at highly variable scales and interaction between different water systems is common. This is especially the case when comparing upper catchment freshwater regimes with the tidal regimes of the estuaries. There are many forms of mixing between groundwaters and surface waters in these settings, which include interaction between aquifers and rivers, estuaries, and bays and ocean shorelines. In addition, there can be a hydraulic connection between different aquifers themselves. The pathways that solutes take are also frequently complex involving aquifer-to-aquifer interaction (Hodgkinson et al. 2007; Bubb and Croton 2002; Li et al. 2008) as well as exchange between ground and surface waters (Soulsby et al. 2007; Mackay et al. 2005; Bailly-Comte et al. 2009; Taniguchi 2002; Smith and Turner 2001; Reay et al. 1992; Westbrook et al. 2005) and consequently the contamination of one can lead to the contamination of the other (Pinder and Celia 2006; Hayashi et al. 1998). As an example, nutrients such as nitrogen and phosphorus, and trace metals such as copper and iron can infiltrate the aquifers as a result of agricultural practices (Shaffelke et al. 2002; Howarth et al. 2002; Church 1996; Zekster et al. 1983; Cable et al. 1997; Dowling 2002; Nakano et al. 2007; Trojan et al. 2003; Ensign and Mallin 2001) or be sourced from a wide range of impacts resulting from urbanisation and can eventually be discharged into coastal waters. These various solutes are also shown to commonly enter surface waters via direct run-off and be delivered to the ocean via estuaries (Crossland et al. 1997; Iversen et al. 1998; Stalnacke et al. 1999; Dyer 1997; Pointon et al. 2003; Soicher and Peterson 1997; Withers and Sharpley 2008).

Factors such as reduced rainfall, changes in vegetation, and groundwater over-pumping can lead to a reduction in water levels in aquifers and surface systems. In coastal settings, this reduction can result in landward movement of the seawater-freshwater interface in both ground and surface water bodies (Mulligan et al. 2007; Werner 2009) and processes in estuaries can often mirror changes in the adjacent aquifer systems. In addition, significant rainfall events typically increase discharge and run-off of dissolved and suspended nutrients and metals from anthropogenic and natural sources to adjacent water bodies. These processes and other environmental factors have been shown to contribute to the development in marine waters of Lyngbya majuscula, a toxic blue-green algae (cyanobacterium) that is a major environmental concern in southeast Queensland (BMRG 2005; Dennison and Abal 1999; Ahern et al. 2006; Albert et al. 2005; Pointon et al. 2003). Knowledge of potential solute transport mechanisms and dominant hydrological processes can assist decision-makers when planning current and future land management practices in order to avoid such environmental impacts.

The small coastal villages of Poona, Boonooroo and Tuan within the study area are all under increasing pressure from urban development. These villages are unsewered and use septic systems for effluent disposal, and the increasing number is of some concern to residents (BMRG 2005) due to potential contamination of aquifers. In addition, the potential over-exploitation of local fresh groundwaters due to increasing extraction is also of consideration. It is of note, that similar conditions to these were the case around many coastal areas of Moreton Bay in southeast Queensland until recently. There still exist areas using various combinations of groundwater, reticulated water, septics and sewage systems with local area treatment plants.

Although there has been continuous water quality monitoring in some areas within the Great Sandy Strait, a recent report (BMRG 2005) states that there has been no continuous water quality monitoring (estuarine or freshwater) in the project area prior to this study. In addition, no event monitoring water quality data exists for the coastal streams in the Great Sandy Strait study area. A major aim of the study, in addition to providing an initial assessment of water chemistry in the study area, is to provide hydrochemical and process-related information during steady-state conditions (i.e. fairly consistent climate conditions and not during high rainfall events) as a guide for future monitoring of these coastal settings which are typical of much of central and southern Queensland.

Study area

General setting

The area being studied is on the Fraser Coast of Queensland, adjacent to the Great Sandy Strait, a passage landscape between the large Fraser Island sand mass and the mainland (Fig. 1). The mainland coastal region is characterised by short, steep catchments that drain to the coastal flood plains, with lower gradients in the north of the area towards the delta of the large Mary River. For the purposes of this study the northern section of the area is referred to as the Tuan catchment and contains the subcatchments of Kalah, Maaroom, Big Tuan and Little Tuan Creek. The southern section of the area is referred to as Poona catchment and contains the subcatchments of Poona and Buttha Creeks (Fig. 2). Maryborough, the closest large town, is located on the tidal Mary River and is approximately 25 km northwest of Poona village.

Fig. 1

Map of Australia and a section of the southeast coast of Queensland. The dashed box shows the location of the study area on the Fraser Coast

Fig. 2

Fraser Coast study area showing catchments and drainage systems. Main types of land use are: plantation forestry (light grey) with forest compartments; native vegetation as buffer zones and bushland (dotted); residential settlements (dark grey); and national park (striped). Elevation decreases towards the east and northeast. Source National Resources, Mines and Energy (2004)

Climate

The climate of the area is subtropical, typical of southeast Queensland, with more than 60% of the annual rainfall occurring during the summer wet season (December to February) and comparatively dry winters (June to August). Mean annual rainfall is 1,115.7 mm. Based on 1908–2006 data for Maryborough from the Bureau of Meteorology (2009) the wettest month is February with a mean rainfall of 174.2 mm, and August, the driest, has a mean rainfall of 39.6 mm. The maximum mean monthly temperature of 30.7°C occurs in January and minimum mean monthly temperature of 8.6°C occurs in July. Table 1 lists climate conditions for data collection field trips. Rainfall distribution during the study was, however, atypical with some heavy rainfall occurring during winter months.

Table 1 Climate conditions for data collection dates

Surface water and drainage systems

In this area, the coastal bedrock was incised by river action during a period of relative low sea level (glacial period) and these channels have been partially filled with sediments during the succeeding base-level rise (interglacial period) (Neil 1998; Lang et al. 1998). According to Thom and Roy (1984) the last postglacial marine transgression (PMT) began approximately 11,000 year BP and terminated approximately 6,500 years BP. Depending on location, this was followed by a near constant relative sea level (RSL) or a falling trend from 1 to 3 m higher than present, commencing approximately 5,000 years BP. This coastal evolution is reflected by Optically Stimulated Luminescence (OSL) dating of pebbly sand beds at Beachmere, a similar setting on the northwestern margin of Moreton Bay, Queensland, approximately 150 km south of the Poona village. These ages reveal that relative sea level at Beachmere was approximately 1 m higher than present 1,700 ± 140 years BP. Shortly after (1,140 years BP) and up until approximately 140 years ago, there was a linear rate of shoreline progradation of 0.40 m per year, an apparent increase from around 0.16 m per year for the period 1700–1140 years BP (Brooke et al. 2008).

More than 70% of the study area has topographic gradients less than 1.5% while gradients in the upper catchments increase to 10%. Elevations range from −0.6 to 129 m ASL with most of the area being less than 60 m ASL. The smaller subcatchments such as Buttha, Poona, Little Tuan and Big Tuan, Maaroom and Kahlah Creek (Fig. 2) have poorly developed surface drainage systems and catchment definition.

The estuaries of this area are tidally dominated. In Poona Creek, the largest estuary, the estuary head occurs around Pappins Bridge (the Cooloola Road crossing), an adopted middle thread distance (AMTD) of 10.5 km from the mouth (see Fig. 2). Upstream from this point there is a change of slope and the catchment profile increases in gradient. At the mouth of Poona Creek, spring tidal ranges are generally slightly less than 2 m (Larsen, unpub. data, 2008).

During and after rainfall-induced high flow events, run-off from the catchment into streams and their lower order creeks can be substantial and can affect water quality in the Great Sandy Strait (Shaffelke et al. 2002; BMRG 2005; Campbell and McKenzie 2004). In addition, heavy rainfall can produce significant seepage from shallow unconfined aquifers adding to nutrients and trace metals discharged to estuarine and marine environments. Many of the tributaries are small creeks of an ephemeral nature for most of the year and consist of pools alternating with small flows.

Land-use and vegetation

Poona, Little Tuan and Boonooroo are small coastal villages with populations of approximately 200. These communities do not have town water supply and residential water is obtained from rainwater tanks and/or groundwater bores. There are short-term population increases in these coastal villages for vacation, fishing and recreation. The main land-use in the area consists of mature Pinus plantations ranging from 16 to 30 years of age, which have native vegetation buffer zones adjacent to natural waterways (see Fig. 2). Since the 1950s, extensive areas within these coastal catchments and lowlands, which were generally unsuitable or marginal for agriculture were acquired by the state government forestry department for Pinus afforestation (Costantini and Loch 2002). According to the Tuan 1:50,000 map (FPQ 2007), pine forests were first established north of Big Tuan Creek in 1974. Some small pine plots were established as early as 1948 in the study area near the Tuan Forestry Office. There were also soft-wood plantations to the east of Maryborough and at Tuan from the mid 1960s, but there were very few plots until the 1970s.

Watercourses from the current areas of exotic pine plantations drain to the coastline and the Great Sandy Strait. Riparian buffer zones have been retained throughout the plantations, in which buffer widths of the zone mostly relate to stream order. Plantation management practices tend to disrupt landscapes and drainage systems when initially established, mostly by clearing, roads and drains, and then again after 25–30 years when harvested. After clearing, typical forestry practices are as follows (M. Robinson, pers. comm. 2007): (a) herbicides to kill weeds in the plantation compartment in preparation for planting; (b) pine seedlings planted in either straight rows or mounds (the mound method helps reduce run-off); and (c) seedlings fertilised with phosphorus. Unless required, plantation compartments are not given any further fertilisation although further herbicide treatments are often carried out. After the trees are harvested, the compartment is broadcast burnt. Hardwood stumps are then heaped into windrows by dozer and burnt again (Bubb and Croton 2002).

Remnant native vegetation within the buffer zones is typically grassy forest containing various species of eucalyptus, and tufted native grasses such as Themeda australis (Kangaroo grass) and Imperata cylindrical (blady grass) and low layered forest containing paperbarks (Melaleuca quinquenervia), White Mahogany (Eucalyptus acmenioides) and a ground layer consisting largely of sedges (Coaldrake 1961). Within the tidal zone of the coastal flats saltmarsh and mangrove areas occur along the estuaries. Peripheral to the estuaries are freshwater wetlands supporting Wallum banksias (Banksia aemula), melaleucas and grasstrees. These wetlands form an important part of the overall freshwater drainage system; they often have groundwater links, and in many cases are close to saline waters.

Hydrogeological setting

This study largely addresses shallow groundwaters (<25 m depth) as those sampled come from aquifers within weathered profile of the bedrock or unconsolidated surficial formations such as alluvium or coastal deposits. The bedrock geology is mainly comprised of the Duckinwilla Group and the Elliot and Grahams Creek Formations (see Fig. 3) (Natural Resources Mines Energy 2004). These formations have experienced deep weathering since the Miocene period (24 Ma) and, as a result, much of the land surface in the study area is covered by a thick lateritic profile.

Fig. 3

Geology of the area and sample sites. Broken lines to the north indicate boundaries of the Tertiary alluvial aquifer identified by Laycock (1969) and Wang (2007) study area boundaries. Bores labelled “JL” are from Laycock (1969), those labelled “C” are from Wang (2007). Inset shows enlargement of bore transect and Poona town bores, at mouth of Poona Creek Source National Resources, Mines and Energy (2004)

Duckinwilla group (Tiaro Coal Measures)

Late Triassic to early Jurassic age feldspathic to quartzose sandstone, siltstone, shale, coal and a ferruginous oolite marker. This unit was previously defined to include both the Tiaro Coal Measures and Myrtle Creek Sandstone (Cranfield 1989). In the area of this study, the upper profile of the Tiaro Coal Measures comprises very thin sandstone–siltstone-shale facies consisting of almost equal proportions of quartz, volcanic rock fragments, feldspar (dominantly plagioclase), detrital muscovite, biotite and minor opaques set in calcite cements with almost no matrix (Cranfield 1993).

A weathered form of the Tiaro Coal Measures dominates the surficial geology (see Fig. 3) in the southern half of the study area, although in places it is unconformably overlain by the Grahams Creek Formation and Elliott Formation. These weathered materials are made up of undifferentiated coastal plain sediments consisting of sand, silt, mud and minor gravel. In many parts of coastal Queensland, catchments with low seaward gradients have developed wide alluvial deposits forming aquifers that contain large volumes of groundwater. Exploratory drilling in this area indicates that this particular formation shows little potential as a groundwater source. Unconfined surficial and semi-confined groundwaters in the Poona catchment coastal areas are generally confined to (a) small pockets of alluvium within the drainage systems and only within the coastal plain, or (b) meandering creek banks infilled as a result of channel migration. Infill materials are eroded from the Duckinwilla Formation within the low-lying coastal plain and do not occur in the elevated catchment west of Pappins Bridge. Both lithological logs and hydraulic behaviour of boreholes drilled adjacent to the Poona Creek estuary show this site is on an infilled meander bank; at this point the stream channel is migrating northward. Figure 3 shows the location of these monitoring wells.

Grahams creek formation

Late Jurassic to Early Cretaceous age intercalated volcanic and volcaniclastic sediments. The volcanic rocks are of andesitic to rhyolitic composition (Ellis 1968; Cranfield 1993). The Grahams Creek Formation is observed as weathered material adjacent to the shoreline within the southern part of the area, and also as small pockets of unweathered material within some catchments (Fig. 3). The weathered Grahams Creek Formation occurs in the form of undifferentiated coastal plain sediments of sand, silt, mud and minor gravel.

Tertiary alluvium

In an earlier report, Laycock (1969) describes an extensive Tertiary alluvium aquifer in the northern catchment. Laycock’s study focused on the Tiaro and Pialbo forestry areas in the north of the project area. A sand-gravel aquifer was indicated by lithological and hydrochemical data collected in 1969 (Laycock 1969). Further investigation was carried out in 1975 to determine the potential for the supply of good quality water from this aquifer (Laycock 1975). This aquifer occurs at a depth of 4–15 m and comprises fine to coarse grained sand and gravel, which in all cases is overlain by impermeable clays (ferricrete and fine saprolite layers of the weathered Elliot Formation profile). Potentiometric contours and isopachs of the sand-gravel aquifer derived from lithological logs indicate a palaeochannel (probably an old Mary River channel) in the north of the area, striking southeast towards Boonooroo. At the time, groundwater EC measurements indicated a variable salinity distribution, however, it was also observed that high pumping rates often resulted in saltwater intrusion particularly in bores located near the tidal reaches of streams. Laycock concluded that the alluvial aquifer, although extensive, is extremely variable in grain-size composition. As a result the success rate for site selection for high yielding bores is likely to be low and best prospects appeared to be along the direction of the palaeochannel (Laycock 1975). Data for ten sites included in the 1969 report have been included in the dataset for this project; Fig. 3 shows the indicated boundaries of this formation.

Elliot formation

Eocene to Oligocene quartzose to sublabile sandstone, conglomerate, siltstone, mudstone and shale (Natural Resources Mines Energy 2004). Sandstone forms the bulk of this formation and is generally white or yellowish, fine to medium grained, well sorted, and lithic sublabile to quartzose, with angular clasts and white clay matrix. The Elliot Formation dominates the surficial geology of the northern section of the area, and overlies the Duckinwilla Group in the southern section. This unit is comparatively thin with logs from Cranfield (1982) indicating depths of 7.5, 10.0 and 24.3 m to the Tiaro Coal Measures in boreholes within the Poona catchment. A fluviatile depositional environment is suggested for the Elliot Formation (Cranfield 1982).

The Elliot Formation occurs in two forms: a duricrusted old land surface with dominant facies being ferricrete with silcrete and indurated palaeosols at the top of a deep weathering profile and as undifferentiated coastal plain sediments of sand, silt, mud and minor gravel (Natural Resources Mines Energy 2004). The weathered Elliot Formation dominates the northern parts of the study area, occurring in its duricrust form. In the southern parts, it occurs more often as undifferentiated unconsolidated sediments overlying the Tiaro Coal Measures of the Duckinwilla Group.

Wang (2008) used a combination of XRD analysis, and aqueous and soil geochemical measurements, to develop a regolith profile typical of lithology of the weathered Elliot Formation found within the northern Tuan catchment. This type of profile is widespread at this latitude and results from intense weathering over long periods. Hydrologically, the infiltration of water is restricted at three levels: the top of a ferricrete or mottled saprolite layer, the top of a deeper fine saprolite layer containing soluble salts, and the unweathered bedrock. A thin coarse saprolite layer overlies the bedrock and contains freshwater. Data indicate that there is no vertical migration of soluble salts from the fine saprolite layer to the groundwaters within the coarse saprolite. A number of monitoring wells were established for the Wang (2008) project and four of these have been included in the dataset for this area: C2d (13 m), C3 (7 m), C4 (3.8 m) and C5 (4 m). See Fig. 3 for the locations of these bores.

Quaternary alluvium

Surficial Quaternary alluvium occurs more extensively in the northern Tuan catchment than in the Poona catchment, where it is mostly derived from weathering of the Elliot Formation and is confined to drainage systems (Natural Resources Mines Energy 2004).

Beach ridges

These are Holocene in age and consist of sand, shelly sand and minor gravel. There appears to be good quality freshwater locally within this formation, however, there is also evidence of some saline intrusion (discussed below).

Estuarine tidal flats

These Holocene estuarine tidal muds and sands occur along the coast in this section of the Great Sandy Strait, and for several hundred metres upstream of the mouth of Poona Creek. These sediments are sometimes areas of significant fresh and marine water interaction.

Soils

Much of the area is covered by deep weathering profiles overlain by shallow topsoil. This is typical of coastal lowlands along the seaboard of eastern Australia, and soils are nutrient poor and generally deficient in nitrogen and phosphorus (Coaldrake 1961). The soil profiles in the study area are highly variable. Higher topographic areas often have very shallow topsoil such as humus podzolics, red earths and yellow earths overlying deep weathering profiles while lower topography areas often contain sediments that have been eroded from the higher features. This profile is further complicated by the presence of aeolian sand dune deposits which occur throughout areas of medium elevation (A. Hammond, unpub. data, 2007).

Investigations for acid sulphate soils were conducted in 2003 within the Poona catchment and were identified in two shoreline forestry compartments. One of these (compartment 9B) had possible actual acid sulphate soils (AASS) while the other (compartment 13A) showed potential acid sulphate soils (PASS) at depth in the Tewan Creek drainage line (Malcolm et al. 2004). See Fig. 3 for these locations.

Waterlogging

Widespread waterlogging occurs in much of this region and has been of some concern to plantation management, notably the effects on catchment water balance (Bubb and Croton 2002). This is of high significance during management of harvesting. Highly variable groundwater responses were recorded within bores at different depths and these suggested that there are localised recharge areas within the forestry compartments where the aquitards causing the waterlogging may be discontinuous or have a considerably higher hydraulic conductivity than had been observed previously. This type of waterlogging is also evident in the Poona and Tuan coastal zones. Wang (2008) further investigated waterlogging in this region and confirmed that it was due to groundwater perched on layers of ferricrete and mottled saprolite within the weathered Elliot Formation.

Site locations

To test coastal zone processes, a transect of five monitoring bores (P2, P4–P6, P8) was established on a meander at the mouth of the Poona Creek estuary. An additional monitoring well, B1, was drilled just north of Buttha Creek (see Fig. 3). The holes were drilled using a hydraulic rotary drilling rig with bentonite drilling mud. Cuttings samples were collected at over 0.5 m intervals. Drillhole depths varied (Table 2) and 3 m PVC slotted screens were placed at the bottom of the borehole, gravel packed and sealed with bentonite. Wells were capped and housed with a galvanised steel casing set in a concrete surface seal. Four shallow monitoring wells (P11–P14) were later (mid-2009) installed by hand auger in muds and sands of the estuarine tidal flats between the transect and the estuary in order to further investigate tidal intrusion in this area.

Table 2 Description of sample locations

Figure 3 shows the locations of all sampling sites included in this study. Table 2 lists all sites, site descriptions, adopted middle thread distance (AMTD) from creek mouths for surface waters (SW) or depth of sampling or screening interval for groundwaters (GW), distance of groundwater sites from estuary or the Strait, topographical heights and water type computed for a representative dataset.

Poona estuary transect

This transect of bores (P2, P4–P6, P8) was established for detailed hydrochemical and water level measurements of the groundwaters adjacent to the estuary. The transect is within fluvially derived sediments of the weathered Duckinwilla Group (Tiaro Coal Measures). Bores P11–P14 lie within the muds of sands of the supratidal flats between this transect and the estuary.

Poona catchment sites

One monitoring well (B1) is located in the duricrust of the weathered Grahams Creek Formation which overlies the Duckinwilla Group formation to the southwest. Surface water site PCM is located in the tidal muds and sands at the mouth of Poona Creek. Stream sites in the Poona catchment (PC9, PC10, WP12, WP19, WP16, WP31, LRB) are within the Duckinwilla Group except for the sites at WP11 and Pappins Bridge (PB) which are within weathered Elliot Formation. The surficial geology at the residential bore sites at Poona (PCP, 134B, 204B) consists of Holocene beach ridge sands.

Tuan catchment sites

Surficial geology at the residential borehole sites at Boonooroo (RF) and Tuan (RE) consists of the duricrust of the weathered Elliot Formation while the surface water data collection site at Tuan Creek (TCA) is in the Elliot Formation. The geology at TCB, located in the Big Tuan Creek estuary, consists of mixed mafic and felsic rocks of the Grahams Creek Formation (Natural Resources Mines Energy 2004).

According to the 1969 Laycock report, the boreholes JL1, JL2, JL7, JL9 and JL10 lie within Tertiary alluvium consisting of varying amounts of clays, sands and gravels overlain by Tertiary Elliot Formation which occurs as a thin veneer of white siltstone or clay and underlain by Maryborough Formation bedrock. Bores JL13, JL20 and JL25 have a similar profile but also have a superficial layer of coastal deposits. Bore JL8 also lies within this Tertiary alluvium but is overlain by organic silts and clay, occurring east of where the sediments of the near-horizontal Elliot Formation merge with these coastal deposits. Bore JL4 occurs within the Maryborough Formation itself where it is not overlain by the Elliot Formation or Tertiary Alluvium (Laycock 1969).

Both Laycock (1969) and Wang et al. (2008) studied the northern area, however, Wang et al. (2008) focussed on vadose zone hydrology. All boreholes except C2d (13.0 m) and C3d (7 m) had depths of only around 2–4.6 m while the Laycock study focussed on the Tertiary alluvial valley occurring at variable depths and thicknesses between 4 and 15 m. Practically all of the shallow bores drilled for the Wang project occur within weathered Elliot Formation, which is reflected in the regolith profiles described. Although C2d is south of the estimated boundary for this Tertiary alluvial valley, a thin layer containing gravels and coarse sands that was intercepted at between 10.5 and 12.5 m depth within this borehole could be part of the peripheral thinning of the Tertiary alluvial valley (Fig. 3).

Methods

Data collection

Surface and groundwater samples were collected from Poona, Little Tuan, Big Tuan and Boonooroo catchments, all predominantly pine plantation areas. There are a total of 29 groundwater sites including residential boreholes and sites included from the Wang (2008) and Laycock (1969, 1975) studies and ten surface water sites. The sites from the Wang study were re-sampled in May 2008 (FT4). The selection of sites is to provide representative samples from the main geological landform settings of the area.

In total, eight sampling runs were conducted over a period of 25 months from August 2007 to August 2009; the multiple sampling approach was used to determine if seasonal variations were significant. The first field trip coincided with moderate (24–96 mm/day) to heavy (96–384 mm/day) rain resulting in flooding conditions. Of the remaining seven field trips, FT3 and FT6 had very light (<6 mm/day) to light rain (>6 mm/day, <24 mm/day) conditions (FT3) and one during very light rain (FT4). Otherwise, conditions were dry during the time of collection and the previous several days leading up to sample collection. However, there was evidence of recent rain in May of 2009 (FT6) but nil rainfall was recorded at Tuan Forestry Office for this time, the closest rain gauge station to the study area (Table 1). However, 16.6 mm was recorded at Double Island Point (an ocean coast site approximately 50 km southeast of the Tuan Forestry Office) for 31 May 2009, demonstrating the variability in rainfall over this relatively small coastal region. Also of significance to the distribution of regional rainfall is the presence of the Fraser Island landmass.

Field methods

Physico-chemical parameters (EC, pH, Eh, DO and T) were measured in situ at all sites with a TPS 90FL field meter. Samples for dissolved ion analysis were collected in 250 mL polyethylene bottles and samples for the determination of δ¹⁸O and δ²H in 35 mL McCartney bottles. Samples for cation analysis by ICP-OES were filtered in the field using 0.45 μm pore diameter polycarbonate filter papers and acidified to pH <2 using 50% nitric acid for preservation; anion bottles were not acidified and filled to eliminate air space. Collected samples were kept on ice or refrigerated for transport and storage; isotope samples were kept in containers in the dark to avoid evaporation. For all groundwater samples, bores were purged prior to physico-chemical measurements and collection of samples by use of a submersible pump or bailer. For those boreholes with continuous recharge, a limit of 10 min was set on the purging time.

Analytical methods

Inductively coupled plasma optical emission spectroscopy (Varian Liberty 200 ICP-OES and Varian Vista-MPX ICP-OES) was used to determine concentrations of the cations, Al, Mn, Fe_total, Mg, Ca, Na, Zn, Cu and K. Samples were analysed for anion concentrations of F⁻, Br⁻, Cl⁻, SO₄ ²⁻, PO₄ ³⁻ and NO₃ ⁻ by ion chromatography (IC Dionex DX300) and alkalinity by titration with hydrochloric acid within 24 h of collection. For samples collected on field trips 6, 7 and 8, an AQ2 SEAL discrete analyser was used to colorimetrically determine alkalinity (AQ2 Method No: EPA-100-A Rev.2) and concentrations of chloride (AQ2 Method No: EPA-105-A Rev. 4). For stable isotope determinations of δ²H and δ¹⁸O, samples were sent to the Isotope Analysis Service, CSIRO, Adelaide, and the results are reported in ‰ VSMOW. Isotope analysis was not included in the earlier Laycock project, and many of these bores could no longer be located.

Assessment of chemical analyses showed that some ground and surface water samples had a high ionic imbalance. The presence of significant amounts of humic and fulvic acids is the most likely cause for these charge balance errors which is a common problem for certain samples in these coastal settings. As a consequence, a relatively relaxed limit of 15% charge balance error is utilised in this study in order to obtain a good spatial coverage. Aside from P11–P14 and the Laycock sites, the dataset used here is from the fourth field trip (FT4) in May 2008 when conditions were mostly dry with occasional light showers. The shallow monitoring wells (P11–P14) were not installed until 2009 and for these August 2009 (FT8) data is used. The Laycock samples included in this study were collected and analysed in 1968; the Wang study sites were re-sampled during the May 2008 (FT4) data collection.

Graphical methods

Scatterplots, Piper and Stiff diagrams are used to assess large datasets, identify dominant processes and detect groupings of water samples. These methods are described in numerous texts, for example, Freeze and Cherry (1979); Hem (1992); Mazor (1997) and Drever (2002). Some useful examples of studies where graphical methods are also employed in order to interpret and/or characterise hydrogeochemical data are Sukhija et al. (1996); Giménez and Morell (1997); Logan et al. (1999); Sanchez Martos et al. (1999); Cruz and Silva (2000); Allen and Suchy (2001); Kim et al. (2003) and Hodgkinson et al. (2007).

Environmental stable Isotopes

There are characteristic patterns in ¹⁸O and ²H isotopes in precipitation that are related to latitude, temperature, land mass, altitude and seasonality and, as a result, these data can provide information relating to recharge sources, flow paths and mixing of natural waters (Freeze and Cherry 1979; Clark and Fritz 1997). The isotopic composition of ²H and ¹⁸O in water is expressed in per mil (‰) deviations from SMOW (Standard Mean Ocean Water) and written as δ²H and δ¹⁸O. Global and local meteoric water lines (GMWL and LMWL), equations resulting from the linear regression of δ²H and δ¹⁸O data for water samples collected globally and locally, respectively, can be used to assist in the interpretation of isotopic data. The global meteoric water line used was defined by Harmon Craig (1961) and is based on around 400 water samples of rivers, lakes, and precipitation from various countries with the equation

$$ \delta^{ 2} {\text{H}} = 8\delta^{ 1 8} {\text{O}} + 10 $$

Data sourced from the Global Network of Isotopes in Precipitation (IAEA/WMO 2006) for Brisbane (approximately 200 km south of the study area) has been used to construct the LMWL by linear regression and has the following equation.

$$ \delta^{ 2} {\text{H}} = 7. 4 9\delta^{ 1 8} {\text{O}} + 1 1. 7 4 $$

The slope of the Brisbane LMWL (BMWL), being relatively close to the GMWL slope of eight, reflects the relatively high humidity in this region. In areas where there is high humidity, kinetic evaporation effects are closer to equilibrium and the slope will be closer to the GMWL. In areas of low humidity, kinetic non-equilibrium evaporation effects will be more pronounced and the slope will become progressively lower with decreasing humidity.

Results and discussion

Major ion concentrations, physico-chemical parameters, isotope ratios and charge balance errors for all samples are summarised in Table 3.

Table 3 Physico-chemical and isotopic data

Coastal rainfall has relatively high proportions of sodium and chloride compared to inland precipitation due to prevailing southeasterly winds and cyclic salt content of the rainfall. Consequently, sodium and chloride are the dominant cations and anions in all samples in this coastal region (McNeil et al. 2005). Other major ions are low relative to sodium and chloride although in some cases bicarbonate and sulphate exceed 10% of the total ion sum. The samples have been typed based on major ions in Table 2. Excluding data collected in August 2007 during storm conditions, major cation and anion concentrations are relatively consistent over time for the majority of sites. The main cause of variability in the data is mixing of estuarine and fresh ground and surface waters. To assist in characterising samples, Na concentrations were plotted against Cl concentrations (Fig. 4) and illustrate concentration ranges and grouping of waters and various hydrological processes such as evaporation and saline-freshwater mixing in both surface and subsurface waters.

Fig. 4

Log-log plots of sodium ion versus chloride ion concentrations in mg/L for all waters sampled, relative to the saltwater dilution line. a is higher salinity waters and b is lower salinity waters. These figures display the wide range of waters in these coastal settings

Overall, groundwater is indicated to occur in three hydrogeological settings in this region, (a) zones within weathered bedrock, (b) Tertiary alluvial valley in the north, and (c) Quaternary unconsolidated materials of the tidal coastal strip which can be further subdivided into (1) coastal plain alluvial deposits, (2) tidal sands and mud, and (3) coastal beach ridge sands. Surface waters can be separated into two readily differentiable groups, (a) freshwaters within the elevated catchments and streams within the coastal plains disconnected from the main tidal channel, and (b) estuarine water within the main tidal drainage channels of the coastal plains.

(a) Weathered bedrock

Many of the unconfined groundwaters in the weathered Elliot Formation of the Tuan catchment (Wang et al. 2008) had variable salinities relating to the brackish waters within the mottled saprolite. The kaolinised pallid zone of deep weathering profiles is widely accepted as a major repository of soluble salts and the major source of these salts is most likely coastal rainfall and marine aerosols rather than bedrock weathering (Wang et al. 2008). These salinities are highly variable spatially, however, and EC values at C3, C4 and C5 were much higher than other groundwaters sampled in this area. The EC distribution suggests relict salinity or tidal influence although these sites are some distance from the coast. Certainly, C5 has a much higher EC than at any of the other sites, being close to 10 mScm⁻¹, and groundwater here may be subject to some degree of saline intrusion. The site is located at the boundary of Poona National Park, a low-lying coastal plain covered by extensive wallum heath communities and mangroves, and where the surficial geology consists mainly of sands. Some degree of saline intrusion could reasonably be expected in this type of environment.

Shallow bores C3, C4 and C5 are strongly acidic (2.7–4.2) due to input of organic acids from the soil layer whereas C2d, screened within a thin layer of coarse sands and gravels, has a much higher pH of 6.0. This appears to confirm that downward flow of water is restricted by fine (saprolite) clay layers within the weathering profile, as reported in Wang et al. (2008).

Drilling carried out for this project did not reveal any substantial unconfined alluvial aquifers in the Poona or Buttha catchments within the coastal plain, but more typically shallow soil profiles overlying semi-confining to confining weathered bedrock. There were no boreholes drilled in the elevated areas of the Poona catchment. Much of the elevated catchment is covered by bedrock of the Duckinwilla Formation and pockets of weathered Elliot Formation are likely to have a similar profile to that found in boreholes, C3, C4 and C5 described and possibly contain small amounts of fresh groundwater. Overall, however, these weathered profiles are poor water-bearing zones.

Monitoring well B1, screened in weathered materials within the Poona catchment, had a pH value of 4.2 due to organic acids leached from leaf litter at the surface or from organic materials at depth. Based on EC variation over time (assuming saline intrusion will lead to variation with tidal condition) and depth and topographical differences, it appears that B1 located 100 m from the Buttha Creek estuary could experience some saline intrusion.

(b) Tuan catchment tertiary alluvial valley

In the northern section, the Tertiary alluvial valley aquifer contains a substantial volume of good quality freshwater. All of these groundwaters from the Laycock 1969 study have pH in the near-neutral to weakly basic range. Most of the samples occurring within the Tertiary alluvium aquifer in this northern area plot closely together in the Na versus Cl plot (Fig. 4), expected for samples with similar lithology and rainfall recharge (fresh confined groundwaters). The EC and pH measurements within the Tertiary alluvium relative to the shallow unconfined waters (C3d, C4, C5) suggests that there is little to no infiltration of soluble salts through the profile to this aquifer.

(c) Quaternary unconsolidated materials of the tidal coastal strip

There are three formations along the coastal strip containing Quaternary unconsolidated materials. There are the alluvial deposits adjacent to the drainage system within the coastal plain, the supratidal tidal muds and sands adjacent to the lower estuary, and the beach ridge sands located at Poona and Boonooroo.

(1) Coastal plain alluvial deposits

In the low-lying coastal plain of the Poona catchment, there are small pockets of semi-confined groundwater adjacent to the tributaries occurring within relict fluvial channels or infilled meanders, such as the one where monitoring bores P2, P4–P6 and P8 are located. Preliminary analysis of sediment samples collected from these boreholes show layers with varying amounts of organic matter and sands alternating with clean aeolian sands. Semi-confining silt/clay layers were also present in some of these boreholes underlying the soil profile. A gravel and sand layer of high hydraulic conductivity was present in five of the boreholes at several metres depth in this transect.

All of these transect groundwaters are fresh except for P5, which is generally brackish due to tidal influence. However, during the flooding conditions of August 2007, the EC of both P2 and P5 was significantly elevated. P5 had an EC of nearly 17 mScm⁻¹ and P2 exceeded 1 mScm⁻¹. This elevated EC is most likely due to a storm surge raising water levels in the estuary, so increasing the reach of the intrusion of saline estuarine waters inland. However, the significant drop in EC between P2 and P5, which are only 20 m apart, demonstrates typical small-scale variations and the existence of a confining layer between these two sites. In addition, the fact that P8, only 40 m south of P2 did not have an elevated EC at this time shows further vertical confinement within this transect. These transect groundwaters all have acidic pH (5.1–5.6). Many are within sediments that contain few reactive minerals, being mainly composed of quartz sand, and consequently these groundwaters often have pH values related to rainfall, typically around 5.6, and, in the more acidic cases, show the input of organic acids.

Sites RE at Tuan and RF at Boonooroo have comparatively high EC values at 1,932 and 1,744 μScm⁻¹, respectively. The magnitudes and time-series variation of these EC values indicate that these aquifers are subject to saline intrusion. These two bores are shallow (<11 m) and at low elevation very close to estuarine creeks. As would be expected based on the EC measurements, the ion concentrations (Fig. 4) for these two bores also indicate saline intrusion. Samples obtained during the drilling of the RE residential borehole indicate a palaeochannel infilled with pebbles at around 10 m depth, and highly transmissive conditions for the interaction of fresh and estuarine waters. It is possible that these features represent the coastal extent of the Tertiary alluvial valley. Local residents in Boonooroo, Tuan and Poona report that many bores in the area have become increasingly saline over the last 10 years, as use of groundwater has increased.

(2) Tidal sands and muds

Unconfined groundwaters in monitoring wells P11–P14, located in the supratidal flats adjacent to Poona Creek estuary, are all saline due to tidal intrusion with EC exceeding 70 mScm⁻¹ at times. The salinities here often exceed those at PCM (Poona Creek estuary mouth) due to evaporation from the water table, which is generally less than 0.5 m below the ground surface (Fig. 4). This evaporative process leads to enrichment of Na and Cl and other ions such as sulphate relative to the saline surface waters intruding the sediments. There is also variation over time due to varying magnitudes in tidal flux and freshwater input. The shallow groundwaters within the supratidal flats adjacent to Poona Creek estuary (P11–P14) had circum-neutral values reflecting the mixing between acidic groundwaters, rainfall and basic marine waters.

(3) Coastal beach ridge sands

The beach ridge aquifer groundwaters (134B and PCP) have low EC of around 400–700 μScm⁻¹. Both are fresh and have fairly consistent measurements over time and appear to be confined. Residential borehole PCP was the only fresh groundwater with a near-neutral (pH = 6.5) in the Poona catchment. Bore 204B has an EC of 1,396 μScm⁻¹ indicating saline intrusion and is approximately 12 m deep. Borehole PCP is 19 m deep within beach ridge sands, while the depth of 134B is unknown but is thought to be of the same order. This variation in salinity may be related to depth and reflect that the beach ridge sands contain a deeper confined fresh aquifer and an upper semi-confined aquifer prone to saline intrusion.

Surface waters

All surface waters in the Tuan and Poona catchments are fresh except for the estuarine sites, PCM, PB and TCB; surface water sites PB and TCB are both brackish to saline depending on tidal conditions. Many surface waters had a pH value close to that of rainfall except for estuarine waters, PB, TCB and PCM, where pH values reflect the mixing of fresh and saline waters due to tidal flux. However, an electrical conductivity (EC) of only 59 μScm⁻¹ was measured at PB in August of 2007 when flooding occurred. Similarly, a brackish EC of 7,790 μScm⁻¹ and a much lower pH of 5.5 was measured at PCM at this time due to fresh floodwaters mixing with marine waters. At other times, EC measurements at this bore were saline and pH values were basic as would be expected for marine waters.

Site TCA has significantly higher concentrations of Na and Cl (and therefore higher TDI) than all other fresh surface waters. These higher concentrations may be a result of limited drainage from this site, due to the ephemeral nature of the creeks and flat gradients in this region. Upstream waters are being transported to this low-lying site but are not being drained from it unless there is substantial rainfall. In addition, evaporation from the water surface raises the concentration of inorganic solutes in these surface waters.

Hydrochemical facies

The Piper ternary diagram was used to support the assignment of groundwaters to different hydrochemical facies and to better define processes. The use of this plot shows partitioning between the different waters and revealed six chemical groups (Fig. 5).

Fig. 5

Piper diagram showing generalised groupings based on lithology, salinity and level of confinement. Circles are surface waters and triangles are groundwaters. See Table 4 for details of samples and groups

Representative Stiff diagrams are shown for the hydrochemical facies groups in Fig. 6. These plots show quite a lot of similarity between groups and, as a result, have been further subdivided based on water type, location and geology. These graphical methods were found to be only useful to a point and then a priori knowledge and observations were used to partition sites according to hydrochemical character and processes.

Fig. 6

Representative stiff diagrams for hydrochemical facies showing relative distributions of major ions. The predominance of Na–Cl ions in this coastal setting is evident and there is a great deal of similarity between hydrochemical facies. Due to this similarity, facies have been further subdivided based on water type, location and geology. See Table 4 for site assignments

Isotopes

Mean values for all samples analysed for δ¹⁸O and δ²H are −4.27 and −20.94‰ respectively. These values are very close to the mean isotopic composition of global precipitation of −4.0 and −22.0‰ and are typical of precipitation for which the water vapour has been sourced from seawater; the influence of other processes that alter the isotopic concentrations within these reservoirs appears to be limited. In this region, a moderating maritime effect on temperature due to the coastal location means variation is minimal compared to areas further inland. This condition is reflected in the isotope data. Excluding PCM, a marine water that is significantly enriched compared to the other surface waters, the δ¹⁸O and δ²H data have standard deviations of 0.42‰ VSMOW and 4.62‰ VSMOW and ranges of 1.43‰ (−5.03 to −3.6‰) and 13.8‰ (−29.5 to −15.7‰) respectively. Groundwater samples have an even smaller spread of 0.65‰ (−4.45 to −3.6‰) for δ¹⁸O and 6.1‰ (−21.8 to −15.7‰) for δ²H. Considering that measurement precision for these two isotopes is 0.15‰ for δ¹⁸O and 1‰ for δ²H, it is clear that the groundwater sample ratios, in particular, fall within a relatively narrow range.

To put this isotopic data into a regional perspective, a comparison is made with groundwater and surface water data from other coastal areas in southeast Queensland (Fig. 7). These locations are the Pimpama coastal plain at the southern end of Moreton Bay located approximately 250 km south of the study area (Harbison and Cox 2009) and the Bell’s Creek catchment located approximately 150 km south of the study area (Ezzy, 2005) in northern Moreton Bay (inset of Fig. 6). Similar to this study area, groundwaters within the Pimpama coastal plain were found to be of relatively local origin. In addition, the Pimpama data also showed processes such as mixing between freshwater and seawater, transpiration and evaporation (Harbison and Cox 2009). Data from the Bell’s Creek catchment study showed clear differences between waters within surficial alluvial and sand-rich coastal aquifers and bedrock groundwaters due to delayed recharge. These data also showed where bedrock discharge was contributing to groundwaters with alluvial aquifers (Ezzy 2005). Although there is some limited evidence for evaporative and saline-freshwater mixing processes reflected in the isotopic data for this project, isotopic ratios excluding PCM all group at the bottom end of these datasets. The groundwaters sampled for this study are all relatively shallow and generally within surficial sediments with local recharge. Consequently, isotopic ratios reflect those of rainfall and group closely together. In addition, as along much of the coast of Queensland winds are predominantly from the southeast, as rainfall passes over Fraser Island to reach the mainland, rain-out effects deplete ¹⁸O and ²H isotope concentrations. Within this closely spaced group, however, there is an overall trend where sites recharged from rainfall further inland are more depleted in ¹⁸O and ²H than those that are closer to the coast, i.e. there is a depletion in isotopes with distance from the shoreline.

Fig. 7

δ²H vs δ¹⁸O plot for all waters. a The range and distribution of isotope data for the study area with a regional context including data from Harbison and Cox (2002) and Ezzy (2005). The common saline water intrusion into coastal Quaternary unconsolidated aquifers is evident in these earlier studies. b The Fraser Coast isotope data in relation to PCM and VSMOW. c Detail of the samples from this study. Surface waters distribution indicates evaporation effects

Of note, nearly all groundwaters (including semi-confined and confined aquifers) appear to cluster along the Brisbane Meteoric Water Line (BMWL) (Fig. 7). The proximity of the groundwaters to the LMWL indicates that secondary processes such as evaporation are not occurring prior to infiltration. It is most likely that the shallow groundwaters (P2, P3, P6, P9, C4, C5) are recharged directly by meteoric waters and have little to no interaction with other waters, except where there is limited tidal influence. Shallow groundwaters from bores C4 and C5 in the northern catchment are slightly depleted compared to the transect groundwaters, reflecting their location further from the shoreline. Coastal residential boreholes, 204B, PCP, 134B and RE appear to have local rainfall recharge based on isotopic ratios, although bore RF is somewhat more depleted perhaps indicating recharge from further inland.

There is overall grouping of fresh surface waters (LRB, PC9, PC10, WP12, WP31) which are isotopically depleted in comparison with shallow groundwaters. This is likely a result of these waters recharging at higher elevation in the catchment (downwind). This rainfall would be isotopically lighter as compared to coastal rainfall. In addition, the fresh surface waters all plot below the LMWL. This may be the result of kinetic evaporation effects, which lead to a disproportionate enrichment in ²H and ¹⁸O, i.e. the waters are more depleted in ²H than ¹⁸O relative to the LMWL resulting in displacement below the LMWL. Isotopic ratios for these surface waters confirm that there is no interaction with marine waters. A sample of ponded water, WP11, has isotopic ratios that are strongly elevated by comparison and similar to those for semi-confined groundwater values, suggesting a significant groundwater input, but not necessarily evaporation.

Site PCM is located at the estuary mouth and has water chemistry typical of marine waters and is highly enriched in ¹⁸O and ²H compared with all other Fraser Coast samples (see Table 3). The level of enrichment indicates that there is limited fresh stream water mixed with these estuarine waters reflecting the hydrology (tidally dominated estuary) and geomorphology (flat gradients and intermittent streams) of the area. This slight enrichment δ¹⁸O (0.96‰) compared with SMOW δ¹⁸O of 0.5‰ (Clark and Fritz 1997) is due to evaporation at the ocean surface. Surface water sites PB and PCM show a correlation between δ¹⁸O and δ²H with salinity as would be expected for estuarine samples (Fry 2002; Zhang et al. 1990). Isotopic ratio data for mixed saline-freshwaters are typically placed along a line between freshwaters on the LMWL and SMOW (Standard Mean Ocean Water) or in this case, PCM. This is not the case at the surface water site TCB where there is interaction between marine and freshwaters. Depths at TCB are often greater than 3 m and water was sampled from depths less than a metre so it is possible that if the water column was stratified at the time of collection, with fresh upstream waters flowing on top of the more dense saline tidal flow, only the overlying more isotopically depleted freshwater has been collected. This would explain the proximity of TCB isotopic ratios to those of the other fresh surface sites.

Nutrients

Waters collected for this study were also analysed for nitrate and phosphate. From all 116 samples analysed, only 11 showed nitrate concentrations above 1 mg/L (1.2–3.8 mg/L) and of those only one had a concentration greater than 5 mg/L. A deficiency of nutrients such as nitrogen and phosphorus is typical of the soils and sediments in these coastal lowlands on the southeast coast of Queensland, unless there is significant artificial input due to agricultural practices or urbanisation, which is reflected in their nutrient concentrations. The maximum nitrate concentration detected was 8.3 mg/L, well below the Australian drinking water quality guideline of 50 mg NO₃ ⁻/L (National Health Medical Research Council (NHMRC) 2003), and was for a residential bore sample. This addition of nitrate is most likely a result of leaching of organic fertilizers or septic tank effluent (based on ö¹⁵N data not included here).

Dissolved phosphate was only detected in four water samples from one data collection period and ranged from 0.42 to 5.20 mg/L. These sites appear randomly distributed and phosphate concentrations were below the minimum detection limit for all other data collections. No fertilisation operations had been carried out in the adjacent plantations for at least the preceding 11 months, so it is unlikely that these results were due to forestry practices. However, as phosphorus has a strong tendency toward forming ionic complexes and compounds of low solubility with many metals, phosphate introduced to the system often precipitates out of solution and, therefore, dissolved concentrations are generally very low.

The above only applies to data collected specifically for this study. Samples from the Laycock study of the northern alluvial valley were not analysed for nitrate or phosphate concentrations.

Summary

This study has focused on the setting of coastal catchments and the adjoining flood plain. The study has shown that there is complex interaction between different water bodies and with different chemical types, and that a wide range of processes occur within this area. To clarify the occurrence of surface and groundwaters in this regional setting and the controls over them, the results from the study have been summarised into (a) hydrochemical and hydrological processes, (b) hydrochemical facies, and (c) hydrological interaction.

(a) Hydrochemical and hydrological processes

The main processes identified are summarised below:

• Coastal rainfall is a dominant influence on the water chemistry of the fresh ground and surface waters. This influence is reflected in the fact that all samples have Na and Cl as dominant ions irrespective of location or lithology.

• Saline intrusion is a dominant factor in the chemical character of many of the coastal groundwaters and some surface waters and, for these sites, elevates the concentrations of all major ions. Ionic concentrations at C5 indicate that saline intrusion may be occurring in sandy aquifers of low-lying areas some distance inland through the sandy sediments of Poona National Park.

• Another major contributor to salinity of unconfined groundwaters is the soluble salts within the saprolite layers in the Tuan catchment area, inland from the coastal zone.

• Natural deficiencies of nutrients, such as nitrate and phosphate, in the soils and sediments of this region results in low or negligible concentrations of these nutrients in all waters.

• The lithology of sediments and bedrock materials throughout the area is generally siliceous which means that little dissolved calcium or bicarbonate is contributed to the system through water/rock interaction. Any elevated bicarbonate contributed to groundwaters is most likely sourced from shell material within the coastal sediments or, in the case of surface waters, from a catchment source.

• Topography is a major influence on surface water chemistry. The lower gradient in the northern Tuan catchment reduces the flushing capacity of creeks and thereby increases the dissolved ion concentrations. In addition, the reach of tidal intrusion along the creek is limited by the elevation at or near site PB just east of Cooloola Road. This is reflected in the significant reduction in TDI between PB and LRB which, although only 2.3 km apart (AMTD) drops from a TDI of 19433 to 89 mg/L. The change in topographical height is 6 m.

• Drainage system morphology is also an important influence on the chemistry of surface waters here. Streamflow is limited during drier months when a drop in water levels results in the drainage system becoming a series of pools rather than a continuous and connected drainage network. In addition, the substantial reduction in TDI between PB and WP11 (and WP12) indicates that there is significant constriction at some point along this tributary. Although WP11 is closer to PCM than PB in relation to AMTD and the topographical height difference is only 0.7 m, TDI drops from 19433 to 81 mg/L. This indicates there is negligible interaction with marine waters at this site, a result of downstream channel infilling.

• Isotopic data indicate that the presence of the large Fraser Island land mass leads to significant rain-out effects depleting ¹⁸O and ²H in rainfall before it reaches the mainland and, as a consequence, the shallow groundwaters and surface waters here are depleted in ¹⁸O and ²H isotopes.

• Isotopic ratios of the groundwaters and surface waters for this coastal region broadly reflect the distance of rainfall recharge inland.

• Isotopic ratios at the Poona Creek estuary mouth confirm that there is little discharge from the catchment during dry conditions.

• Most surface water isotopic ratios indicate there is little to no input by groundwaters to the drainage system. The exception is WP11, located within the flood plain, which appears to have a significant groundwater contribution.

• Isotopic data indicates that most groundwaters are recharged locally, except possibly for RF, which may be recharged from further inland.

• Major ion chemistry, pH and EC of estuarine surface water samples collected during a flood event (FT1) indicate substantial discharge to the Strait from the drainage system.

• Groundwater data collected during the flood event also showed the effect of storm surges on some coastal groundwaters where elevated TDI concentrations reflected an increase in marine water input due to the forcing of marine water inland.

(b) Hydrochemical facies

Ground and surface waters within this coastal area can be broadly classified into a number of facies or groups based on the chemical character, which reflects both setting and hydrological processes. These groups are derived from the previous graphical methods, in addition to lithology, spatial distribution and time-series variation of physico-chemical parameters and ionic concentrations (Table 4). Means and ranges are shown in Table 5 for pH, EC, sulphate, bicarbonate and TDI; the other major ions, Na, Cl, Mg and Ca are not included in Table 5 as they all strongly correlate with TDI.

Table 4 Assignment of sites to hydrochemical facies based on Stiff diagrams and aquifer types and processes

Table 5 Physico-chemical character of hydrochemical facies

Group I. Evaporated estuarine groundwaters

These Na–Mg–Cl type groundwaters have the highest concentrations of ions; they are unconfined with shallow water tables within the muds and sands of the supratidal flats adjacent to Poona estuary. Tidal flux is a dominant influence on their chemical character and Na, Cl, Mg, SO₄ ²⁻, Ca, and HCO₃ ⁻ concentrations are often elevated relative to intruding estuarine surface waters due to evaporative processes. Circum-neutral pH values reflect mixing between draining acidic groundwaters and marine surface waters. Nitrate and phosphate concentrations are below detection limits.

Group II. Estuarine surface waters

These Na–Mg–Cl surface waters have variable major ion concentrations, pH (weakly acidic to basic) and EC (brackish to saline) values depending on distance from the coastline, tidal conditions and rainfall. Nitrate and phosphate concentrations are below detection limits.

Group IIIA. Saline-intruded groundwaters

These Na–Cl type waters are very similar to Group IIIB except for having lower proportions of magnesium; host lithologies are variable. The location and depth of bores is important: bore 204B lies within beach ridge sands perhaps underlain by weathered Duckinwilla Formation undifferentiated coastal plain sediments; bores JL2 and JL20 lie within the Tertiary alluvium aquifer in the Tuan catchment, JL2 is adjacent to Big Tuan Creek and JL20 is close to Boonooroo Point not far from the strait. Generally, waters within this Tertiary alluvial aquifer appear to have less magnesium content than the other coastal aquifers in the Poona catchment. The higher magnesium elsewhere is most probably sourced from the mafic volcanics of the Grahams Creek Formation. The magnitude of ionic concentrations varies significantly among these waters and is related to a combination of the intervening geology, tidal conditions, rainfall and proximity to the coast.

Group IIIB. Semi-confined saline-intruded groundwaters

These Na–Mg–Cl type groundwaters have variable major ion concentrations and pH and EC values relate to magnitude of intrusion, a function of distance from the estuary/strait, geology, tidal conditions and rainfall. Tidal processes are a major factor in determining the frequency of saline intrusion at some sites. For example, P5 has a salinity flux related to a semi-diurnal tidal cycle, whereas P2 is only intruded during highest astronomical tides or storm surges and as such may only have a major rise in TDI due to saline intrusion 2–3 times a year. Lithologies are variable: bore P2 is within the undifferentiated coastal plain materials of the weathered Duckinwilla Formation and P5 within the muds and sands of the supratidal flats adjacent to Poona Creek estuary. B1 lies within duricrusted old land surface of the Grahams Creek Formation adjacent to Buttha Creek and RE within a pebble-filled palaeochannel adjacent to the mouth of Big Tuan Creek; this may represent the coastal extent of the Tertiary valley. These sites are well distributed along the coastal strip; however, all sites except C5 are located within 450 m of an estuarine creek or the Great Sandy Strait. It is still uncertain as to whether saline intrusion is occurring within the region of C5. Stiff diagram groupings indicate that C5 is of a very similar character to the semi-confined saline-intruded groundwaters of Group IIIB; however, water isotopes (δ¹⁸O and δ²H) do not indicate saline intrusion. Nitrates are practically non-existent within these waters except at one site where the source is most likely organic fertilizer.

Group IV. Brackish shallow groundwaters

These Na–Cl waters occur within the unsaturated zone saprolite layers of the weathering profile in the Tuan catchment and are removed from any direct marine influence. These waters have relatively high Na and Cl concentrations due to the soluble salts present in these clays. Spatially, EC and ion concentrations within these shallow groundwaters are highly variable. Waters are often strongly acidic due to the input of organic acids from surficial leaf litter. Nitrates and phosphates are negligible.

Group VA. Fresh tertiary alluvium confined groundwaters

These fresh Na–Cl and Na–Cl-SO₄ type waters occur within the extensive aquifer of the Tertiary alluvium in the Tuan catchment to the north. The exception is JL4 which, according to Laycock (1969), is screened within the Maryborough Formation. For these waters, pH is neutral to weakly basic and major ion concentration, EC, TDI and pH values all indicate they are confined, and are much fresher and less acidic than the Group IV waters within the overlying saprolite layers of weathered Elliot Formation. Of note, C2d may be screened within the peripheries of the Tertiary alluvial valley.

Group VB. Fresh unconsolidated coastal sediment semi-confined to confined groundwaters

These fresh Na–Cl–HCO₃ and Na–Mg–Cl waters occur within the undifferentiated coastal plain sediments within the transect near Poona Creek estuary mouth (P4, P6, P8), the beach ridge sands adjacent to the Strait (PCP and 134B) and the Tertiary alluvium and coastal sediments at JL8 near Little Tuan Creek. Most of these waters have a higher proportion of bicarbonate than other groundwaters in this region. The bicarbonate is likely to be from minor shell material within these coastal sediments. The pH is weakly acidic to weakly basic (5.0–7.8). This range in values reflects the varying contributions of bicarbnate, organic acids and rainfall to these waters.

Group VIA. Fresh Tuan catchment surface waters

This Na–Mg–Cl water (TCA) is located in the Tuan catchment and has a much higher TDI than surface waters located in the Poona Catchment. This is most likely due to a build-up of dissolved ions in this shallow pool due to limited drainage as discussed previously. This water is acidic (pH = 5.3) due to rainwater recharge and input of organic acids.

Group VIB. Fresh Poona catchment surface waters

These Na–Mg–Cl waters are located within the drainage system of the Poona Creek catchment. There is little variability in the TDI of these waters which are fresh and have water chemistries reflecting that of coastal rainfall. Upper catchment sites have higher bicarbonate concentrations than lower catchment sites (WP11, WP12). This bicarbonate could be a result of interaction with sediments and/or rocks forming creeks, beds, and banks in the upper catchment or be sourced from further inland. Isotopic data indicates rainfall recharge is from further inland. In addition, the weathered bedrock in this area is generally siliceous containing little to no carbonate material. This suggests an inland bicarbonate source. Isotopic data also indicate that there is some degree of evaporation occurring at these surface water sites.

These creeks are generally very shallow except for where water points (artificial ponds for fire fighting—WP11, WP12, WP31) have been excavated and at times will dry out completely. These ponds may act as sinks (similar to TCA) for dissolved inorganic and organic materials stratifying the water column while any rainfall maintains a fresh water layer on top. These were all composite samples and as such do not show variation throughout the profile.

(c) Hydrological interaction

One of the goals of the study was to establish the potential connectivity between different water bodies. This aspect is summarised here.

• There is only limited contribution of alluvial groundwaters to the drainage system (i.e. base flow) in the middle and upper Poona catchment. This is largely due to the low volume of alluvial material within the upper drainage system plus the bedrock geology typically consists of a deep weathering profile containing discontinuous clay layers. These layers are very low permeability and often cause waterlogging within the shallow overlying soil profile. Overland flow and run-off through the soil profile during rainfall represent the more dominant input to the poorly developed drainage system in the upper catchment.

• The alluvial sands and gravels of meanders in the low-gradient sections of catchments provide potential pathways for solutes from the plantations to the estuary. Some of the shallow groundwaters adjacent to the Poona Creek estuary clearly show tidal influence and therefore a zone of mixing between marine and inland waters. These settings, however, are hydrogeologically complex at a local scale. Confining layers locally produce perching, and appear to limit interaction between fresh groundwaters and marine waters within the transect.

• There are several small coastal communities that use groundwater from limited freshwater aquifers within unconsolidated Quaternary sediments. The water chemistry in some residential bores indicates hydrological connectivity between the aquifers and the marine environment. This connectivity may permit nutrient and/or metal discharge into the adjacent Great Sandy Strait.

• Interaction between unconfined and confined groundwaters in the Tuan catchment appears to be negligible, with deeper confined waters maintaining lower TDS and higher pH values relative to the brackish waters contained within the weathered Elliot Formation. The Tertiary palaeovalley is semi-confined to confined and does not appear to receive recharge locally, however, chemistry indicates it is connected to marine waters near the shoreline.

Conclusions

Assessment of this coastal region and its catchments identified three fresh groundwater regimes, (a) zones within weathered bedrock, (b) Tertiary alluvial palaeovalley in the north, and (c) Quaternary unconsolidated materials of the tidal coastal strip. The study has shown variable interaction between surface waters and groundwaters, and also between different groundwaters including saline waters. This mixing is especially the case in low-gradient sections of the catchments and the coastal plain. Hydrological processes contributing to the hydrochemistry of these sites have been identified and hydrochemical facies described. Based on the sites monitored, interaction between these hydrochemical facies is shown to be minimal except in the coastal zone where saline-freshwater mixing occurs at a number of locations. The potential impacts of this are twofold. Where interaction between marine and fresh groundwaters occurs there is the potential for dissolved species to be transported to the Great Sandy Strait, especially during times of high rainfall; also, there is the potential for groundwater supplies to be degraded by saline intrusion, which is already a concern in the coastal villages of this area.

Water analysis determined only one example of anthropogenic nutrient input to the natural waters in the area, which was most likely sourced from organic fertilizers in a residential area. Other than potential issues related to planting or harvesting, plantation forestry practices appeared to have had comparatively low impact on the natural waters in this region during the time of the study.

This study confirms dynamic settings with complex hydrological systems at various scales and with multiple processes operating; and it has defined the types of surface and groundwaters that exist, their host materials and their potential connectivity. The information from this study not only provides a necessary foundation for future hydrochemical monitoring in this study area but also for much of the coastline in southeast Queensland which has similar geomorphological, hydrological and land-use characteristics. A sound understanding of regional ground and surface water hydrochemistry, hydrochemical processes and possible transport pathways for solutes to marine waters is essential for the future management of coastal environments, particularly in ecologically sensitive areas subject to increasing pressure from urbanisation.