Aquatic species and communities Classification of aquatic habitats

The aquatic habitats in the Cooper subregion can be classified using the work of Jaensch (1999), Eamus et al. (2006), Kennard et al. (2010) and Fensham et al. (2011). None of these four classifications alone describes the range of water-dependent ecosystems at a consistent resolution sufficient for the BA methodology (Barrett et al., 2013), but in combination they provide robust coverage of the range of water-dependent ecosystem types.

Jaensch (1999) developed a classification of the wetlands of the south-western quadrant of Queensland, which included the Queensland parts of the Cooper subregion (Table 16). While originally developed only for the parts of the Lake Eyre drainage catchment that falls within Queensland, the classification applies equally well to the wetlands of the Lake Eyre drainage catchment in SA, that is, in the western half of the Cooper subregion. Nineteen of the wetland classes occur within the Cooper subregion (Queensland and SA combined). Artesian springs occur to the west, north and east of the Cooper subregion. Permanent freshwater lakes are restricted to the uplands of the Great Dividing Range, in the headwaters of the Lake Eyre Basin and distant from the boundary of the subregion.

Eamus et al. (2006) proposed three simple primary classes of groundwater-dependent ecosystems:

  • ecosystems dependent on the surface expression of groundwater, including baseflow rivers and streams, wetlands, some floodplains and mound springs. This class of groundwater-dependent ecosystems requires a surface expression of groundwater, which may, in many cases, then soak below the soil surface and thereby become available to plant roots
  • ecosystems dependent on the subsurface presence of groundwater, often accessed via the capillary fringe (non-saturated zone above the saturated zone of the watertable) when roots penetrate this zone. No surface expression of groundwater is required in this class of groundwater-dependent ecosystems
  • aquifer and cave ecosystems. These ecosystems include karstic, fractured rock and alluvial aquifers. The hyporheic zones of rivers and floodplains is considered in this category because these ecotones often support species that are obligate groundwater inhabitants.

The Cooper subregion lacks any ecosystems dependent on flows of groundwater at the level of the soil surface, although artesian springs occur nearby, to the west, north and east. The other two classes occur within the subregion and are associated with the major drainage lines, especially Cooper Creek. Aquifer ecosystems were not recognised in the classification of Jaensch (1999, see above) and thus there would be merit in extension of that classification to include a new class for aquifer ecosystems.

Table 16 Wetland types of south-western Queensland, as defined by Jaensch (1999)

Wetland type

Presence in subregion: Yes = present Near = present in Lake Eyre Basin, to the west or north

Waterholes and watercourses


Permanent river reaches and waterholes



Wooded watercourses



Shrubby floodplain watercourses



Watercourses without trees and shrubs


Freshwater lakes


Permanent, isolated freshwater lakes



Oxbows (cut-off river bends)



Temporary freshwater lakes without grassland



Temporary freshwater lakes with couch grassland


Saline lakes


Semi-permanent saline lakes



Temporary saline lakes




Gibber and interdunal claypan aggregations



Isolated claypans and canegrass swamps



Sedge swamps



Forb meadows on floodplains



Lignum swamps



Bluebush swamps



Cooba shrubby swamps



Eucalypt wooded swamps



Acacia/belah wooded swamps




Artesian springs


Rock holes


Rock holes in arid uplands


Through an analysis of stream gauge data from across Australia, Kennard et al. (2010) developed a classification of surface water flow regimes for creeks, rivers and river segments. Four of twelve flow regime classes are likely to occur in watercourses in the Cooper subregion:

  • predictable winter highly intermittent
  • predictable summer highly intermittent
  • unpredictable summer intermittent
  • variable summer intermittent.

Each of the four flow regime classes was found at least once in different stretches of Cooper Creek and its tributaries. There is some indication of systematic longitudinal variation of flow patterns along Cooper Creek, with south-western, downstream gauging stations indicating predictable winter intermittent flows, and north-eastern, upstream gauging stations indicating predictable or unpredictable summer intermittent flow regimes. However analytical difficulty may have led to somewhat similar flow patterns being assigned to different regime classes. With different seasonality and degrees of predictability, all observed river flow regimes are intermittent, meaning that the rivers largely dry up and retreat to waterholes between major flow events (Kennard et al., 2010). This is consistent with the findings of the Australia-wide analysis of low-flow rivers by Mackay et al. (2012); all stretches of rivers in and around the Cooper subregion were classified as highly or moderately ephemeral.

Fensham et al. (2011) developed a classification of permanent water bodies and natural wetlands, along rivers and elsewhere, for an area that substantially overlaps the subregion:

  • waterholes – enlarged segments of an ephemeral or seasonal river or watercourse which hold water after streamflow has ceased
  • rock holes – natural hollows in rocky landscapes, formed from fracturing and weathering, which store water from local runoff
  • outcrop springs (referred to as recharge springs in the earlier terminology of Fensham and Fairfax (2003)) – dependent on groundwater and occurring where sediments that form the aquifer are outcropping
  • discharge springs – dependent on groundwater that emanates through confining beds (aquitards) in areas remote from where the aquifer receives its inputs.

The ecological principles inherent in each of the regional classifications above have been incorporated into classifications of wetlands at state (Queensland) and national (across Queensland, NSW and SA) scales (Queensland WetlandInfo mapping – see Environmental Protection Agency, 2005; Interim Australian National Aquatic Ecosystem Classification Framework – see Aquatic Ecosystems Task Group, 2012). Both classifications have the advantage of completeness, with the Australian National Aquatic Ecosystem (ANAE) classification framework, in particular, including subterranean ecosystems. Diversity and ecological drivers in aquatic habitats

There is a growing, but not yet comprehensive, literature on the key ecological drivers of aquatic communities of the Cooper subregion from which important principles are summarised in this section.

Rivers and waterholes

As foreshadowed in the above description of the Cooper subregion flow regime classification of Kennard et al. (2010), temporal patterns of water and of depth of water are the primary drivers of ecological characteristics of aquatic community type, species diversity and species abundance in rivers (Silcock, 2009) – with spatial connectivity and water salinity as secondary (and often related) drivers.

In Cooper Creek and other rivers of the Lake Eyre Basin, diversity of invertebrates, fish and birds at a broad (temporal) scale follows a boom-and-bust cycle. Booms are associated with high-flow periods and better water quality (Shiel et al., 2006; Kingsford et al., 2010; Sheldon and Fellows, 2010). Recent research suggests that the full range of booms are important, including both the spectacular sequences of flows that are able to fill all ‘hydrological sponges’ in a river system and its floodplains (Leigh et al., 2010b) as well as those that flow down river channels but do not spill across floodplains nor fill all billabongs and lakes associated with a river (Balcombe and Arthington, 2009; Kerezsy et al., 2011). For invertebrates and fish, booms result from population dynamics entirely within the river systems, but for nomadic waterbirds – which Kingsford et al. (2010) describe as ‘time and space travellers’ – diversity is not determined locally or by any one river system, but instead by the probability of viable resource patches across all river systems in inland Australia.

In the inland rivers, as floods recede and water flow is no longer continuous, waterholes and riverine wetlands take on the role of refuges, habitats to which a species population retreats and persists during times of environmental stress (Marshall et al., 2006; Silcock, 2009; Sheldon et al., 2010; Puckridge et al., 2010; Davis et al., 2013). Refuges during the dry period are important areas for the survival of species and persistence of water-dependent communities. Their physical habitat characteristics (e.g. surface area, depth, underwater benches, eroding banks and snags) influence aquatic species diversity and evenness. For example, fish community and macroinvertebrate composition changes with the size, depth and structural diversity of the refuges (Balcombe et al., 2006; Arthington et al., 2010; Puckridge et al., 2010).

Connectivity within and between inland river systems is the spatial analogue of temporal variation in flow rate, and the two are interlinked as larger floods or flood sequences permit spatial connection between locations that are geographically and hydrologically more distant within a river basin. Balcombe et al. (2006), Leigh and Sheldon (2009), Fensham et al. (2011) and Kerezsy et al. (2013) have all emphasised the role of connectivity between, or conversely isolation of, waterholes and wetland refuges. Kerezsy et al. (2013) demonstrated that in the Georgina–Mulligan river basin, to the north of the Cooper subregion, connectivity interacts with life history strategies of fish, so that species can be identified as either extreme or conservative dispersers. Connectivity within river systems during times of flood has also been identified as the most probable reason for the high level of similarity between the ecological communities of the many waterholes within inland river systems (Fensham et al., 2011).

Connectivity has its limits, as determined by the maximum extent of the biggest flood events, and lack of connectivity – especially isolation between river systems – also is an important factor in patterns of endemism. For example, Australian smelt, carp gudgeon and Cooper Creek catfish occur only in the Cooper river system, while golden goby and banded grunter that occur in other Lake Eyre Basin rivers are not in the Cooper river system (Fensham et al., 2011; Kerezsy et al., 2013).

The biological consequences of the interactions between connectivity during flood periods and isolation of waterholes during low-flow periods are illustrated well by the fish communities recorded in Lake Eyre Basin waterholes during non-flood conditions in 2012 and 2013 (Cockayne et al., 2013; Sternberg et al., 2014). Twenty-one native species of fish were recorded across the entire drainage basin, but only 3 to 13 fish species were detected at any one waterhole. Just four species (bony herring (Nematalosa erebi), desert rainbow fish (Melanotaenia splendida tatei), Hyrtl’s tandan (Neosilurus hyrtlii) and Silver tandan (Porochilus argenteus)) represented more than 70% of the catch during three sampling campaigns, and were present at more than 50% of sites. The spangled perch (Leiopotherapon unicolor) was also present at more than 50% of sites, but at low abundance. The other 16 species were observed at few waterholes and in low numbers.

Despite the documented importance of connectivity along river systems, evidence for connectivity between aquatic and adjacent terrestrial ecosystems is relatively weak. Using stable isotope techniques to examine ecosystems along Cooper Creek, Bunn et al. (2003) and Fellows et al. (2007) have found that aquatic food webs in waterholes during non-flow periods depend almost exclusively on autochthonous carbon sources from within the waterhole or falling into the waterhole from the littoral zone, and very little from surrounding terrestrial sources. Likewise Arthington et al. (2010) could find little influence of medium-scale ‘watershed’ characteristics on fish assemblage structure in waterholes. Whilst major flood events might be expected to generate windows of connectivity between aquatic and floodplain ecosystems, there are as yet no data on the magnitudes of resource flows or population movements along floodplain gradients during floods.

Shiel et al. (2006) has shown the significance of water salinity in determination of zooplankton diversity and assemblage composition in the Lake Eyre Basin rivers and associated ephemeral waterbodies. The study observed generally lower species richness in sites of higher salinity, but concluded that the effects of salinity were site-specific and involved thresholds of change. The authors suggested that the threshold level for salinity at a site was probably dependent on the time span between flushing events of fresher water relative to the life history characteristics and physiological tolerances of individual species.


Cooper Creek and its major tributaries have impermanent or near-permanent lakes associated with major floods overflowing into low-lying areas of floodplains (Jaensch, 1999). The Coongie Lakes, in the western part of the Cooper subregion, is probably the best studied example (Reid and Gillen, 1988; Puckridge et al., 2010).

The lakes associated with rivers experience the same boom-and-bust dynamics, and the same sensitivity to salinity, as described above for rivers and waterholes; indeed, the lakes are frequently the largest fillable pores within the hydrological sponges of the river system (Leigh et al., 2010b). Puckridge et al. (2010) demonstrated that with increasing of permanence of lake water, fish species diversity (richness, evenness) and disease incidence increased, but fish species dominance and macroinvertebrate abundance decreased. Only the more mobile species of fish were able to utilise food resources provided by lakes that were filled for only brief periods before drying out again. The more permanent lakes also have the greatest diversity of bird species (Kingsford et al., 2010), and for this reason many have received national or international conservation protection status.

Sheldon and Puckridge (1998), Puckridge et al. (1999, 2000) and Sheldon et al. (2002) go on to observe that flooding regime alone is not sufficient to explain the characteristics of overflow lakes in the Lake Eyre Basin catchments. Lakes associated with river systems should be seen as parts of complexes rather than individual habitats, complexes that include not only the rivers that deliver the inputs of water, but also the floodplains and swamps surrounding the rivers and lakes, since locations may alternate between ecosystem or habitat type during flood-drought cycles, and local mobility of species between habitats and locations may drive species diversity.


Jaensch (1999) details a variety of types of swamps that are associated with the margins of rivers, oxbows, waterholes, lakes and springs, or the lower-lying parts of river floodplains. The dynamics and ecology of these wetlands is most likely determined primarily by surface water hydrology and seasonal or infrequent flooding from rivers and lakes, although in the wetland classes containing trees (e.g. Eucalypt wooded swamps and Acacia/belah wooded swamps), the trees are likely to be accessing groundwater. With the exclusion of the dependence of the trees on groundwater, the biodiversity and ecology of swamps are poorly studied in comparison with other wetland habitat types. Swamps are often mapped as terrestrial ecological communities and form part of the subregion’s complex mosaic of vegetation shown in Figure 34.

Rock holes

Fensham et al. (2011) observed that rock holes are poorly known in the Lake Eyre Basin. They document 18 permanent rock holes in three clusters in Paleogene sandstone ranges, including a group to the north-east of Windorah, on the north-eastern fringes of the Cooper subregion. Data were available on the biota of only five rock holes. All major groups of aquatic organisms had low diversity and abundance, and there was no evidence of endemism.

Outcrop springs

Fensham et al. (2011) mapped outcrop springs in two groups on the north-eastern fringes of the Cooper subregion: to the north-east and to the east of Windorah. More species were observed in outcrop springs than in rock holes, but all species observed were widespread. Geographically isolated populations of three species of plant, one species of mollusc (Sermyla sp.) and one species of fish (Mogurnda clivicola, the Flinders Ranges mogurnda) were observed, but no endemism was evident.

Discharge springs

Discharge springs have been studied to a much greater extent than outcrop springs. Fensham and Fairfax (2003) estimated that before 1900 there were 300 discharge spring complexes, or local clusters of springs, in the Great Artesian Basin (excluding Cape York Peninsula), of which only 36% (108 complexes) are still active. Most of these spring complexes lie around the margins of the Great Artesian Basin, including a group of spring complexes in the Bulloo catchment to the south-east of the Cooper subregion (the Eulo supergroup), a small group of springs near Lake Blanche immediately south-west of the subregion, and a substantial number of spring complexes in SA, to the west and south of Kati Thanda – Lake Eyre (see Greenslade et al., 1985; Zeidler and Ponder, 1989; Tyler et al., 1990; Harris, 1992, Lewis et al., 2013; Gotch, 2013). No discharge springs are known within the Cooper subregion, but discussion of them is included here on the basis of the precautionary principle, given possible connectivity of Great Artesian Basin aquifers beyond the bounds of the subregion.

Because the groundwater has a very long residence time in the Great Artesian Basin, discharge springs show minimal fluctuations in flow rates, and their biota are adapted to relatively stable environmental conditions (i.e. there is low resilience and low resistance to environmental change in the biota). Nevertheless these discharge springs communities do show distinct seasonal fluctuations associated with the seasonally variable balance between flow and evaporation (Lewis et al., 2013).

Isolation of discharge springs complexes means that they have the highest rates of endemism and genetic differentiation of any surface aquatic habitat type (Fensham and Price, 2004; Fensham et al., 2011; Gotch, 2013). Almost half of all species recorded from discharge springs in the Great Artesian Basin are endemic to the area. Jardinella, a genus of snail, is represented by 12 locally endemic species. Each of the three species of fish that are listed under the EPBC Act is restricted to a single springs complex. The significance of endemism led Fensham and Price (2004) to rank the discharge springs on degree of species endemism, as a basis for conservation prioritisation. The highest priority springs are not those closest to the Cooper subregion’s boundaries.

Subsurface aquatic habitats

In the Cooper subregion, there is no information on the diversity and ecology of stygobiota, the suite of organisms that inhabit subterranean groundwater habitats. Humphreys (2006) and Tomlinson and Boulton (2010) report that the biodiversity of Australia’s subsurface groundwater-dependent ecosystems is inconsistently characterised. Invertebrates (especially crustaceans) are well characterised only in the major artesian basins of WA, while microbial, fungal and protozoan diversity is virtually unexplored throughout Australia.

From the analysis of the ecological characteristics of groundwater environments by Tomlinson and Boulton (2010), it is, however, possible to identify, in principle, two sets of key drivers for subsurface ecosystems:

  • The stygobiota is adapted to relatively stable environmental conditions compared with surface aquatic environments (except Great Artesian Basin discharge springs), and thus is likely to depend strongly on the stability of the groundwater regime and other abiotic environmental factors (i.e. there is low resilience and resistance amongst stygobiota).
  • Because detritivorous microbes, rather than plants, are the basis of the food chain, ecosystem structure depends on carbon and nutrient inputs through vertical ecotones – the vadose zone in soil under terrestrial vegetation, and the hyporheic and littoral zones of springs, rivers and wetlands – and changes in resource fluxes through these ecotones could lead to change in stygobiota diversity and abundance. Recent change and threatening processes

Ecological changes to –and potentially threatening processes for – rivers, waterholes, rock holes and outcrop springs are poorly documented for the Cooper subregion. Choy et al. (2002) undertook a survey of the ecological condition of reaches of Cooper Creek (as well as the Georgina and Diamantina rivers) and concluded that they were in generally good condition except for bank damage by cattle. However, Silcock (2009) indicated that river waterholes are being slowly degraded by the cumulative impacts of cattle, feral animals, native herbivores and their consequent total grazing pressure during dry periods. Increased grazing may lead to increased rates of soil erosion and deposition in the surrounding landscape, due to the combined effects of decreased vegetation cover, increased exposure of soil surfaces and surface destabilisation by trampling – resulting in sedimentation posing a potential threat to waterholes and associated biota.

Non-native fish seem to be a lesser threat in the rivers of the subregion than in other parts of Australia. Costelloe et al. (2010) concluded that non-native fish may be disadvantaged by, and their invasions have been hampered by, the extremely variable flow regimes in Lake Eyre Basin rivers and wetlands. However, there are small populations of exotic species in these river systems. In a multi-year survey of fish in waterholes throughout the Lake Eyre Basin, Cockayne et al. (2013) and Sternberg et al. (2014) reported low catches of exotic species (<1%).

The sleepy cod (Oxyeleotris lineolatus), a species native to tropical fresh waters in northern Australia, was first reported in the Thomson River in 2008 (Kerezsy, 2010). This species has potential to colonise channels throughout the Cooper Creek river system, but little is known of its interactions with native ecological communities. Cockayne et al. (2013) and Sternberg et al. (2014) also detected sleepy cod in Cooper Creek in 2012 and 2013. Cooper Creek also supports a population of goldfish (Carassius auratus), with their abundances and spatial distribution being highly variable but strongly linked to consecutive years of high flows (Cockayne et al., 2013). The eastern mosquito fish (Gambusia holbrooki) occurs in the lower parts of Cooper Creek, in the Warburton River and in the Neales River (Cockayne et al., 2013; Sternberg et al., 2014).

Change in discharge springs have been the subject of much closer scrutiny, because of both the large number of springs that have become inactive and the sensitivity of the remaining active springs to ecological impact. The past and current threatening processes are (Fensham and Fairfax, 2003; Fensham et al., 2011; Gotch, 2013; Kerezsy and Fensham, 2013):

  • drawdown of groundwater for stock, irrigated agriculture and mining – a significant driver of inactivity and diminishment of springs
  • excavation, including dredging, conversion to dams, wells, draining, and construction of raised concrete structures that limit surface flows away from the point of discharge. At least one form of excavation affects 26% of active springs (Fensham and Fairfax, 2003)
  • impoundments
  • exotic plants (23 species recorded by Fensham and Fairfax (2003)), including date palms (Phoenix dactylifera) and Phragmites australis (Gotch, 2013), as well as plant species for ponded pastures (e.g. Brachiara mutica, Echinochloa polystachya and Hymenachne acutiglumis)
  • pests, especially the eastern mosquito fish (Gambusia holbrooki; see Kerezsy and Fensham (2013) for more details) and the cane toad (Rhinella marina)
  • feral animals, such as pigs, horses and donkeys (throughout the Great Artesian Basin area of Queensland), and camels (in western desert areas)
  • sheep grazing of aquatic and littoral vegetation, and subsequent erosion of exposed soils (in southern areas).

Fensham and Fairfax (2003) record only three discharge spring complexes in national or conservation parks throughout the Great Artesian Basin, although some additional spring complexes have been afforded conservation status since 2003. Subject to effective management plans, park status affords those spring complexes protection against some of the threatening processes above, but aquifer drawdown is a more diffuse process that would also require coordinated management actions in surrounding lands regardless of designation. Species and ecological communities of national significance

Only one unequivocally aquatic species is listed under the EPBC Act and is known to occur, or are likely to occur, in the Cooper subregion: the painted snipe (Table 12). Five other species are associated to a lesser degree with watercourses or floodplains.

The community of native species dependent on discharge springs in the Great Artesian Basin is listed under the EPBC Act for the subregion (Table 12), but occurs near the Cooper subregion rather than within the subregion. This community has become the focus of intense management concern as a result of its high levels of endemicity and the many potentially threatening processes that have been identified (e.g. Wilson, 1995; Fensham and Price, 2004; Queensland Environmental Protection Agency, 2005; Fensham et al., 2010; Fensham et al., 2011; Kerezsy and Fensham, 2013). Species of regional significance

For the 45 taxa (including some subspecies) listed under Queensland’s Nature Conservation Act 1992 but not also listed under the EPBC Act (Table 13), there has been no formal assessment of which are associated with aquatic habitat types. However, it is possible to identify two taxa with clear water dependence using record and habitat data from the Atlas of Living Australia (Atlas of Living Australia, 2014):

  • plant: Hydrocotyle dipleura (no common name)
  • amphibian: Cyclorana verrucosa (rough-collared frog).

For the 48 species listed under NSW’s Threatened Species Conservation Act 1995 (Table 14), only two species of bird could be identified as having clear water dependence using record and habitat data from the Atlas of living Australia (Atlas of Living Australia, 2014):

  • Oxyura australis (blue-billed duck)
  • Stictonella naevosa (freckled duck).

For the 82 taxa (including some subspecies) listed under SA’s National Parks and Wildlife Act 1972 but not also listed under the EPBC Act (Table 15), nine plants, two amphibians, one reptile (Emydura macquarii, the Macquarie tortoise), and 17 birds (a range of ducks and waders) have demonstrable water dependence according to record and habitat data from the Atlas of living Australia (Atlas of Living Australia, 2014).

The Cooper Creek catfish (Neosiluroides cooperensis) is endemic to the Cooper Creek catchment, and has been recorded from locations within the subregion, but it is not listed as threatened under the EPBC Act, Queensland’s Nature Conservation Act 1992 nor SA’s National Parks and Wildlife Act 1972.

Last updated:
5 January 2018
Thumbnail of the Cooper subregion

Product Finalisation date