Forgot your password?
Document Actions
Biodiversity of the Wetlands of the Kakadu Region, Northern Australia (Finlayson et al, 2006)
Aquat. Sci. 68 (2006) 374–399
1015-1621/06/030374-26
Aquatic Sciences
DOI 10.1007/s00027-006-0852-3
© Eawag, Dübendorf, 2006
Overview Article
Biodiversity of the wetlands of the Kakadu Region,
northern Australia
C. Max Finlayson1,2,*, John Lowry1, Maria Grazia Bellio1,2, Suthidha Nou1,3, Robert Pidgeon1, Dave Walden1,
Chris Humphrey1 and Gary Fox1
1
Environmental Research Institute of the Supervising Scientist, G.P.O Box 461, Darwin, NT 0801, Australia
2
Present address: International Water Management Institute, P.O. Box 2075, Colombo, Sri Lanka
3
Present address: Parks Australia, Post Office, Jabiru, NT 0886, Australia
Received: 15 October 2005; revised manuscript accepted: 12 December 2005
Abstract. The biodiversity values of the wetlands in the ing high in some habitats. Most fauna analyses have fo-
Kakadu Region of northern Australia have been recog- cussed on vertebrates with a large amount of information
nised as being of national and international significance, on waterbirds and fish in particular. However, despite
as demonstrated through their listing by the Ramsar Con- extensive effort over the past two decades much is still
vention on Wetlands. Analyses of the wetland biodiver- unknown about the biota. While the invertebrate fauna in
sity have resulted in the production of species list for the streams has received some attention, a large taxo-
many taxa, and some population and community-level nomic classification effort is required. The functional in-
analyses of biomass and abundance, and the mapping of ter-relationships between habitats and species have
habitats at multiple scales. Wetland habitats include in- largely not been assessed. Further, the ecology of many
ter-tidal mud-flats, mangroves, hyper-saline flats, fresh- species is only cursorily known. At the same time there
water flood plains and streams. The tidal influence on the has been increased attention to pressures on the wetlands,
saline wetlands is pronounced, as is the influence of the such as weeds and feral animals, water pollution, and the
annual wet-dry cycle of the monsoonal climate on the potential impact of climate change and salinisation of
flood plains and streams. The vegetation is diverse and freshwater habitats. Importantly, given the social context
highly dynamic with rapid turnover of organic material of the region, increased attention is being directed to-
and nutrients. The fauna is abundant with endemism be- wards traditional use and management of the wetlands.
Key words. Wetlands; wetland fauna; wetland vegetation.
Introduction groups in the wetlands within Kakadu National Park and
the broader Alligator Rivers Region in northern Australia.
In this paper we summarize data on species numbers and The Region is located to the east of the city of Darwin in
describe the general ecology of major plant and animal the Northern Territory of Australia (Fig. 1) and is part of
a larger bio-geographical region known loosely as the
wet-dry tropics that extends across northern Australia
(Finlayson et al., 1997; Finlayson, 2005). Isolated and
* Corresponding author phone: +94-11 2787404;
lowly populated, the wetlands are of immeasurable value
fax: +94-11 2786854; e-mail: m.finlayson@cgiar.org
for the world‘s natural and cultural heritage and have at-
Published Online First: September 29, 2006
Overview Article
Aquat. Sci. Vol. 68, 2006 375
tracted great interest, particularly over the past 20–30 specific. As these species are integral parts of the wetlands
years, for their conservation values (Finlayson and von and may contribute considerably to wetland functions and
Oertzen, 1996a). The Alligator Rivers Region comprises processes (e.g. energy and nutrient cycles) they are in-
the ~20,000 km2 Kakadu National Park and a further cluded in this description, although often there is scant
~8,000 km2 of land within the western portion of Arnhem information about their biology or ecological role/s.
Land (a large tract of Aboriginal owned land to the east of The concept of a wetland species is based around that
Kakadu National Park) that comprises part of the catch- used for wetland plants by Finlayson et al. (1989), name-
ment of the East Alligator River. Lowry and Knox (2002, ly, a plant or animal that is adapted to withstand periods
2006) further extended the areal region to include parts of of waterlogging or flooding for a period of time. Under
the adjacent catchments. Given the recognition of the this general definition species found in the seasonally
name “Kakadu” and its areal dominance within the Re- inundated areas (woodlands and grasslands) that fringe
gion the general name “Kakadu Region” is adopted here the billabongs and flood plains are included along with
and taken to include the wider region, as done by Finlay- the wholly aquatic species, as well as species that use the
son and von Oertzen (1996b) when reporting on the wetland less frequently but are fully dependent on it. This
vegetation of the Region. is similar to the approach proposed by Gopal and Junk
Wetlands in the Kakadu Region include mangroves, (2000) who also provided further subdivisions based on
freshwater flood plains, salt flats and small permanent the extent of the reliance of the species on the wetland.
lakes (Finlayson and Woodroffe, 1996). The wetlands of This more detailed categorization has not been adopted
Kakadu National Park have been designated as interna- given a lack of information on the taxonomy and ecology
tionally important under criteria established by the Ram- of many taxa.
sar Wetlands Convention, due in part, to their importance As there is little specific information from the riparian
in a biogeographical context, the outstanding diversity of fringes of the streams, except from generic biological
their plant communities, and their role in conserving the surveys, this habitat is not emphasised. The monsoonal
large numbers of waterfowl that congregate on the flood forest habitats that fringe or abut the streams and flood
plains during the Dry season (Finlayson et al., 1990; plains are not included; Wilson et al. (1996) provide a
Finlayson and von Oertzen, 1996a). World Heritage listing detailed description of the vegetation of the Region.
of Kakadu National Park also refers to the natural heritage
value of the tidal flats and flood plains, and the floristic
Ecological characterization of the
diversity and endemism of the wetland vegetation.
Kakadu Region
Descriptions of the diversity and ecology of wetland
species in the Region reflect large differences in the
amount of available information for the taxonomic Landforms
entities. The information is uneven and incomplete. The The geological evolution of the Region was summarised
descriptive information presented here is followed by an by East (1996) who described three major landforms.
outline of the pressures on the wetlands, as ascertained in Lowry and Knox (2002, 2006) provided more detailed
recent years using structured and integrated approaches mapping of six geomorphic landscape classes across the
and risk assessment in particular. Region; the distribution of these classes is shown in Fig-
ure 3, whilst the area of each is listed in Table 1.
1. Arnhem Land Plateau – in the eastern and southern
Definition and classification of wetland species parts of the Region; an average height of 300 m above
the adjacent plains; composed predominantly of
Wetlands in the Kakadu Region comprise a number of quartzose sandstone known as the Mamadawerre For-
hydrological patterns – tidally influenced as well as annu- mation (previously known as the Kombolgie Forma-
ally flooded. Whilst the tidal influence extends some 70– tion – Carson et al., 1999); highly dissected with a
90 km along the major rivers, most attention has been di- high proportion of bare rock; bounded in the north and
rected towards the freshwater flooding of the rivers and west by a steep escarpment.
adjacent floodplain wetlands. The freshwater wetlands are 2. Dissected Foothills of Igneous Origin – a variety of
subject to a reasonably predictable mono-modal flood landforms, including rocky hills, boulder-covered
pulse (Finlayson et al., 1990, 1989; Finlayson and strike ridges, stony hillocks and occasional granite
Woodroffe, 1996) that has been related to the rainfall data pillars; formed through igneous activity and subjected
and the more complex seasonal climatic pattern recog- to exposure and/or erosional processes over the last
nized by the indigenous inhabitants of the Region (Fig. 2; 2.5 billion years.
Finlayson, 2005). During the dry months the wetlands 3. Mamadawerre (Koolpinyah) Surface – gently undulat-
may dry completely and can be colonized by terrestrial ing plains and rises of the lowland landscape; com-
plant and animal species that may or may not be wetland posed of sedimentary plains, and found predominantly
376 C. M. Finlayson et al. Wetland biodiversity of the Kakadu region
Figure 1. Location of Kakadu National Park and the Alligator Rivers Region in northern Australia’s wet-dry tropics (from Finlayson et al.,
1997 and Lowry and Finlayson, 2004).
Overview Article
Aquat. Sci. Vol. 68, 2006 377
Table 1. Geomorphic land classes in the Kakadu Region (from
Lowry and Knox, 2002, 2006).
Area km2
Geomorphic landscape class
Alluvial floodplains 2960
Arnhem Land plateau and escarpment 9918
Coastal floodplains 3923
Deeply weather, eroded Mamadawerre Surface 13006
Dissected foothills of igneous origin 3435
Mamadawerre Surface – sedimentary plains 7045
Table 2. Area (km2) of catchments within the Kakadu Region (from
Lowry and Knox, 2006).
Catchment Within Kakadu
Within the
National Park
Kakadu Region
Mary River 8142 1269
Wildman River 3336 1708
West Alligator River 1444 1444
South Alligator River 11945 10900
East Alligator River 10340 2205
Katherine River* 5168 1601
* most of Katherine river catchment is outside of the Kakadu Region
tion of alluvial sediments by the rivers has resulted in
the development of flood plains along the rivers.
6. Coastal floodplains – bound the coastal margins;
poorly drained plains of very low relief; resulted from
Figure 2. Generalised representation of: a) the climate of the Aus-
sea level rise which reached its present level approxi-
tralian wet-dry tropics; b) hydrological change on the floodplains
mately 6000 years ago and drowned the coastal river
(variability is represented by dashed lines); and c) an Aboriginal
calendar. A 14 month period is used to illustrate the extension of valleys; infilling of the valleys led to the development
seasons across the calendar year. The hydrological information was
of broad coastal flood plains; large meandering rivers
adapted from Sanderson et al. (1983) by Finlayson et al. (1990) and
form the upper reaches of the coastal flood plain and
the Aboriginal calendar from Ovington (1986) and Morris (1996).
grade into saline mudflats near the coast.
The Aboriginal seasons are described in the following manner: Gu-
numeleng pre-monsoon season; Gudjeweg monsoon season; Bang- The Region contains the entire catchments of the Wild-
gerreng harvest time; Yegge cool weather time; Wurreng early dry
man, West Alligator, South Alligator and East Alligator
season; and Gurrung hot dry season.
Rivers, as well as part of the Mary and Katherine catch-
ments that are within Kakadu National Park. Lowry and
in the south-west of the Kakadu Region; formed by the Knox (2002, 2006) also include the adjacent parts of the
gradual erosion and retreat of the Arnhem Land pla- Katherine and Mary River catchments (Table 2, Fig. 3).
teau exposing the underlying, more resistant substrata More accurate mapping has been prepared recently, con-
and the development of undulating rises and plains. firming that the original map excluded part of the catch-
4. Deeply Weathered Mamadawerre (Koolpinyah) Sur- ment of the East Alligator River and corrected the mis-
face – a deeply weathered surface composed of plains, conception that Kakadu National Park contained the
broad valleys, very low-gradient slopes, and isolated entire catchment of the South Alligator River (Lowry and
hills and ridges of resistant rock; located mainly be- Knox, 2006).
tween the Plateau and the coastal flood plains, al-
though inliers occur in the Plateau.
5. Alluvial floodplains – along the middle upper reaches Climate and hydrology
of the major rivers; in the upper reaches of the rivers The climate is generally taken to comprise two broad
the floodplains are typically confined within narrow seasons – the Wet season, which commences late in the
river valleys. The plains have a gentle slope and year (November-December) and lasts for 3−4 months,
gradually merge into the coastal flood plains; deposi- and the Dry season (Taylor and Tulloch, 1985), as shown
378 C. M. Finlayson et al. Wetland biodiversity of the Kakadu region
Figure 3. Geomorphic land classes in the Kakadu Region (from Lowry and Knox, 2002; 2006).
schematically in Figure 2. The most significant features logical data, and a generalized hydrological cycle that
of the Wet season are thunderstorms, tropical cyclones was developed separately from field observations at
and rain depressions. As cyclones move inland they form Magela Creek and flood plain in Kakadu National Park
rain depressions and are an important source of rain. (developed by Sanderson et al., 1983, and modified by
Rainfall is also associated with monsoonal troughs, with Finlayson et al., 1990; Fig. 2). At the start of the Wet
2–3 occurring each year, that usually produce widespread season intermittent rain storms saturate the soils of the
cloud and rainfall, regional convection that provides lo- lowland Mamadawerre (Koolpinyah) Surface and as
calized showers and easterly disturbances that, in some more consistent rain occurs water collects in the creeks
years, extend the rainy season. The Dry season is charac- and thence in the large tidal rivers. Heavy rainfall over
terized by south-east trade winds and very little rain fall. the plateau can send water cascading into the streams on
In general, temperatures are warm to hot throughout the lowland surface. Once the creeks and rivers are full
the year with a mean monthly maxima of 37 °C in No- the freshwater spills across the flood plains and can cover
vember and 31.5 °C in July. In the Wet season warm them to a depth of several metres. The base flow in the
creeks across the lowlands is less than 5 m3 s–1 with peak
temperatures are accompanied by relative humidity of
about 80 %. Cloud cover is greatest during the Wet sea- flows late in the Wet season, reaching and exceeding on
occasions, 1000 m3 s–1. The large range in both total an-
son.
Local Aboriginal people have a refined perception of nual and mean monthly discharge along the Magela
the climate and recognize six seasons based on the rela- Creek is representative of the water regimes experienced
tionship between changes in the weather and the availa- in the lowlands across the Region (Fig. 4); there is little
bility of food items (Ovington, 1986; Morris, 1996). The information on the flow patterns in the plateau.
calendar they recognize is usually presented in a circular Flooding occurs once the catchment is saturated;
manner, but when presented in a linear manner and com- heavy falls of rain later in the season generate more wide
pared with the meteorological data the patterns outlined spread flooding than equivalent rainfall earlier in the sea-
by the Aboriginal calendar are readily identifiable from son. The water regimes of the creeks and rivers reflect the
the meteorological data (Finlayson, 2005; Fig. 2). climatic variability of the region (as shown in Fig. 2).
Finlayson (2005) also depicted the relationship be- While flow may be maintained in the upper reaches of the
tween the knowledge base developed over millennia by watercourses by springs or seeps, freshwater flow in the
Aboriginal people, the more recently obtained meteoro- mid-lower sections of creeks and rivers generally ceases
Overview Article
Aquat. Sci. Vol. 68, 2006 379
Figure 4. Total annual and mean monthly flow in Magela creek (data from Northern Territory Department of Natural Resources, Environ-
within a few months of the end of the rains. The creeks and from a 1:1,000,000 map compiled as part of a vegeta-
flood plains then dry out except for a few permanent tion survey of the Northern Territory to 1:250,000 and
swamps and lagoons, known locally as billabongs (Finlay- 1:100,000 surveys of specific communities. Vegetation
son et al., 1990). in the region can be broadly categorised into 8 broad
The spring tidal range in van Diemen Gulf is 5–6 m classes, which reflect the influence of the soils, geology
and estuarine water can extend more than 100 km up- and landform of the area. These include mangrove for-
stream and generally remains within the stream channel ests, sedges and herbaceous plants, paperbark forests,
(Woodroffe et al., 1989). The groundwater level in the grassland and savanna, tall open forest and woodland,
surrounding landscape is recharged by the Wet season monsoon rainforest and woodland, spinifex grassland,
rains, but can fall 2–4 m during the Dry season. and heath shrubland.
Wilson et al. (1996) provide an overview of the veg-
etation and land cover. Around 55 % of the terrestrial
Vegetation/Landcover vegetation is described as tropical tall grass savanna,
The Region contains in excess of 1800 species of higher composed of eucalypt-dominated open forest and wood-
plants (macro-algae and vascular plants; Brennan, 1996; land, with a 1–2 m grassy understorey. A further 30 % of
Wilson et al., 1996). The floral diversity reflects the the Region is covered by heaths, and open woodlands
range of landscapes and habitats (Hoatson et al., 2000). with sparse grass understorey. Closed canopy monsoon
Many vegetation surveys have been conducted, ranging rainforests are restricted to lowland springs, rock out-
380 C. M. Finlayson et al. Wetland biodiversity of the Kakadu region
crops, flood plain and beach levees, and sandstone. Vege- including Rhizophora slowly gaining dominance over
tation communities on the freshwater flood plains exhibit Bruguiera/Ceriops species, a brief period of dominance
marked variations in their composition, due to the ex- by Sonneratia sp. and subsequent succession from Son-
tremes of the Wet and Dry seasons (Finlayson et al., neratia through Rhizoporaceous forest to stands of Avi-
1989, 1990; Finlayson and Woodroffe, 1996; Finlayson, cennia. The sinuous/cuspate phase began about 5300
2005). In the Wet season, flowering plants are profuse years ago and represents a period during which the South
and widely distributed across the floodplain with high Alligator River re-established a distinct channel mean-
dry weight production. However, during the Dry season dering across the estuarine plains, changing during the
the vegetation is sparse and far less diverse. Mangrove last 2500 years from an entirely sinuous form to a pre-
communities are restricted to a narrow coastal strip and dominantly cuspate form in the central part of the tidal
to the reaches of the major rivers subject to tidal influ- section. Associated with the transition from the ‘big
ence (Finlayson and Woodroffe, 1996). swamp’ to the ‘sinuous’ phase there was widespread re-
Fire and invasive plant and animal species have a placement of mangrove species with the grasses and
significant impact on the extent and distribution of plant sedges characteristic of the present day freshwater wet-
species and the land cover (Anderson, 1996; Cowie, lands. The timing of the transition from mangroves to
1996). Saline intrusion, caused by either climate change, freshwater wetlands probably varied throughout the Re-
or damage to the natural levees separating freshwater and gion.
saline wetland communities, by feral animals (water buf-
falo) has the potential to cause significant change to the
Biodiversity of different plant and
vegetation communities of the floodplains (Bayliss et al.,
animal groups
1997; Eliot et al., 1999). Significant change in the flood-
plain vegetation has already occurred with the removal of
feral buffalo (Skeat et al., 1996). Wetlands in the Kakadu Region comprise tidally influ-
enced mangroves and salt flats, seasonally inundated
flood plains and billabongs, and small permanent lakes
Wetland evolution (Finlayson et al., 1990). The latter are small and not com-
Stratigraphic studies, particularly those on the wetlands mon. The flood plains (and associated lagoons) cover
of the South Alligator River have demonstrated that sub- about 195,000 ha and extend along the major rivers (West
stantial environmental change has occurred over the last Alligator, South Alligator and East Alligator) and creeks
20,000 years, attributed largely to changes in sea-level (Magela, Nourlangie, Jim Jim and Cooper) throughout
(see overview in Finlayson and Woodroffe, 1996). About the Region. Published studies of the species-level diver-
18,000 years ago, the sea levels across northern Australia sity of the wetlands have centred on the flood plain of
were some 150 m lower than today and the coastline was Magela Creek, a tributary of the East Alligator River.
several hundred kilometres north of the present location. It is noted that whilst the role of the seasonal wetting
The sea has since risen and around 8000 years ago tidal and drying has been highlighted the effect of stochastic
water flooded into the valleys, as illustrated by data from events on the composition and distribution of species
the South Alligator River valley which is now filled to a within the wetlands has not been considered in detail -
depth of 14-16 m with fine muds and alluvium. ecological investigations have principally considered
The stratigraphy of the deltaic-estuarine plain of the species presence and distribution across the wetlands
South Alligator River, and probably the other river sys- with limited analysis and modelling of ecological intere-
tems in the Region, indicates that the wetlands developed lationships and processes. While Finlayson (1988) pro-
in three major phases: i) transgressive: ii) big swamp; vide an initial summary of the high standing biomass
and iii) sinuous/cuspate. Finlayson and Woodroffe (1996) present on the flood plains the inter-relationship between
summarised the many studies which generally support the chemical, physical and biological components of the
the notion of wide-scale change in the vegetation of the wetlands are not elaborated, except for pointing out that
plains and rivers. The transgressive phase (8000-6800 for some vertebrates much of their food supply comes
years ago) was characterised by marine incursion into the from the adjacent terrestrial environment.
prior valley and terrestrial ecosystems were displaced by
mangrove forests as the sea level rose. By 6800 years ago
a mangrove forest dominated by Rhizophoraceous spe- Phytoplankton
cies had become established. The onset of the ‘big Taxonomic studies were undertaken mainly in the early
swamp’ phase (6800-5300 years ago) followed stabilisa- 1980s. Thomas (1983) provided a listing of 160 diatom
tion of the sea level around its present level. Mangrove taxa from 32 genera and considered the flora rich and
forests established over most of the present-day flood allied to that of south-east Asia with annual changes in
plains for around 6000 years with successional changes water quality and the variable environment contributing to
Overview Article
Aquat. Sci. Vol. 68, 2006 381
the diversity. Bedell (2001) confirmed that the distribution al. (1989); Finlayson and Woodroffe (1996) provided a
and relative abundance of diatom species in artificial and list of mangrove species; Brennan (1996) a list of all
natural wetlands was more strongly related to variations in higher plants; and Cowie et al. (2000a) a flora of the
electrical conductance (EC) and pH than temperature or coastal flood plains of the Northern Territory. The follow-
dissolved oxygen. Out of the 125 species observed, 7 were ing information is taken mainly from Cowie et al.
identified as indicators of pH and/or EC, as shown below: (2000b). The wetland flora consists mostly of widely
– acidic conditions: Aulacoseira granulata, Brachysira distributed taxa with only 3 genera, Hygrochloa, Omeg-
exilis var. lanceolata, Brachysira serians var. brachy- andra and Maidenia, being endemic to Australia. Some
sira (A), and Navicula radiosa; 24 % (57 species) of the species that occurred primarily
– alkaline conditions: Achnanthidium minutissimum and on floodplains were endemic to Australia; 57 % (132 spe-
Nitzschia palea; cies) were widespread in the Old World; 13 % (30) were
– EC < 100 µS/cm: Aulacoseira granulata and Brachy- pantropical; and 6 % (19 species) restricted to Australia.
sira serians var. brachysira (A); and Of the Australian endemics, only 4 species that were not
– EC > 100 µS/cm: Achnanthidium minutissimum, restricted to the floodplains, Bambusa arnhemica, Hy-
Nitzschia gracilis (A) and Nitzschia palea. grochloa aquatica, Nymphoides spongiosa and Nym-
phoides subacuta, were restricted to the Northern Terri-
A high level of interconnectedness existed between pH, tory. The non-endemic nature of the flora was considered
EC, wetland type and species diversity. Artificial sites to reflect the many opportunities for exchange of species
generally differed from more natural sites with a higher with other landmasses during recent geological times.
pH and a higher EC than the creek and billabong sites. Within the Kakadu Region, Finlayson et al. (1989)
Species diversity was lower in artificial sites than in natu- built on information provided by other researchers and
ral sites. Further, water quality was a key determinant prepared a generalized vegetation map for the Magela
governing the diatom assemblages present in different flood plain based on observations from 5–6 years and, in
wetland types. McBride (1983) also determined that the particular, the Wet seasons, when many of the aquatic
biomass of diatoms attached to macrophytes was equiva- plants reach their peak biomass and are easier to differen-
lent to about 30 % of the macrophyte biomass of 16,500 t tiate and map. The analysis took into account the asser-
dry weight in the area investigated. tion by Sanderson et al. (1983) that an over-emphasis on
Ling and Tyler (1986) recorded over 530 taxa from detailed and short-term sampling did not account suffi-
groups other than diatoms. The Desmidiaceae (desmids) ciently for the high natural variability. It is not known
were paricularly rich floristically and, as with the dia- how well the description provided for the Magela reflects
toms, were similar to those of south-east Asia. Broady the situation on other flood plains, especially given
(1984) investigated the widespread desiccated crusts and changes that have occurred with the removal of feral buf-
felts on the floodplain surface during the Dry season and faloes (Bubalus bubalis) (Skeat et al., 1996). Ten com-
recorded Microchaete and Scytonema species not previ- munities were identified (Fig. 5) and are briefly described
ously recorded. Heterogeneous genera, e.g. Scytonema, below.
Calothrix, Hapalosiphon and Stigonema, were present i) Melaleuca open forest and woodland (tree canopy
and secreted mucilaginous sheaths that combined with cover of 10-70 %): areas dominated by one or more
the intertwining filaments were assumed to stablilize the Melaleuca species - M. viridiflora and M. cajaputi
soil surface and reduce moisture loss. A characteristic around the edges and at the northern end of the
feature of the flood plain during the Dry season is the flood plain, and M. leucadendra in back-swamps
development of large patches of green and red surface inundated for 6-8 months. The understorey varies.
scums comprising phytoflagellates, in particular Pyrami- ii) Melaleuca open woodland (tree canopy cover <
monas, Chlamydomonas, Chlorogonium and Euglena 10 %): Melaleuca leucadendra in areas inundated for
species. Euglena sanguinea is responsible for the com- over 6 months. Understorey species are usually the
monly observed red colour, but changes in species com- same as those in adjacent areas of the flood plain.
position can cause a change to green. iii) Nelumbo-Nymphoides herbland: a mixed commu-
nity dominated by the water lilies Nelumbo nucifera
and Nymphoides indica that occur in permanently
Macrophytes and semi-permanently wet areas.
iv) Orzya grassland: dominated by Oryza meridionalis
Distribution. The summary provided by Finlayson (2005) towards the end of the Wet season. In the Dry sea-
of the information available on the vegetation of the son it consists of bare ground and dead Oryza me-
Magela flood plain has been adapted for use here. A list- ridionalis stems with persistent Phyla nodiflora and
ing of freshwater macrophytes (macro-algae and vascular Ludwigia adscendens as xerophytic forms, and
plants) found in the region was produced by Finlayson et Pseudoraphis spinescens.
382 C. M. Finlayson et al. Wetland biodiversity of the Kakadu region
v) Hymenachne grassland: dominated by Hymenachne achne acutigluma and Urochloa mutica along with
acutigluma throughout the year. the herb Ludwigia adscendens, occur along the
vi) Pseudoraphis grassland: dominated by the perennial banks of the billabongs.
emergent grass Pseudoraphis spinescens which has
a turf-like habit during the Dry season and grows up More recent analyses have shown that these communities
through the water during the Wet season. still predominate (J. Lowry unublished information)
vii) Hymenachne-Eleocharis grass-sedgeland: swampy along with several invasive species, in particular Salvinia
areas that dry out seasonally and are dominated by molesta and Urochloa mutica, and the native species Ses-
Hymenachne acutigluma or Eleocharis spp., which bania sesbans which intermittently dominates some areas
are slower to establish. of the plain, assumedly in responses to particular flooding
viii) Mixed grass-sedge-herbland: a variety of species and germination conditions that have not been identified.
with the dominant species depending on the topo- In addition to the flood plain analyses, Finlayson et al.
graphic situation. Oryza meridionalis occurs on the (1993a) reported a semi-quantitative analysis of vegeta-
drier sites with Pseudoraphis spinescens in slightly tion dominance in 5 billabongs from the Magela creek
wetter places, while Eleocharis spp. and Hymen- and flood plain – three located on small tributaries and
achne acutigluma occur in the deeper sites. On sites initially filled by water from the main creek rather than
that remain flooded for 10-11 months Nymphoides from the tributary, and two that represented remnants of
indica and Nymphaea macrosperma may be present. deep channels on the flood plain. The three located on the
ix) Eleocharis sedgeland: Eleocharis spp. dominate tributaries had a generalised vegetation zonation consist-
during the Wet season, but are replaced by annual ing of:
herbs during the Dry season; i) fringing Melaleuca spp. woodland in seasonally in-
x) Open-water community: permanent billabongs, undated areas;
flow channels, shallow waterholes contain Nym- ii) a mix of grasses and sedges in seasonally inundated
phaea macrosperma and Nymphaea pubescens and areas shaded by woodland;
a number of submerged plant species. Floating iii) a belt of Eleocharis sp. in water that is usually <1.5
grass mats comprising Leersia hexandra, Hymen- m deep during the Wet season;
Overview Article
Aquat. Sci. Vol. 68, 2006 383
Table 3. Number of plant species found in four broad habitats areas
iv) a small area of open water usually 1.5-2.0 m deep in
on the Magela flood plain (from Finlayson et al., 1989).
the Wet season; and
v) patches of waterlilies and submerged plants along the Habitat Total Annuals Perennials Geophytic
boundary between the sedges and the open water. species Perennials
Permanent billabongs 46 19 2 6
The dominant plant species, based on ‘visual biomass’
Seasonally inundated 94 57 29 8
and ‘ground cover’, are the Melaleuca spp. trees and the
flood plain
geophytic, perennial Eleocharis spp. sedges. In contrast,
Fringe zone 158 100 50 8
the two billabongs located on the flood plain had a gener-
Permanent swamps 21 5 11 5
alised vegetation zonation comprising:
i) fringing Melaleuca spp., Pandanus aquaticus, and
Barringtonia acutangula trees along a levee that
comprised the western bank;
ii) a mix of grasses and sedges and a few trees interfac- flood plain, except for the permanently wet areas. The
ing with the flood plain grass communities along seasonally inundated plain and the fringe zone contained
the gently sloping eastern bank; around 40 % and 70 % respectively, of the species, com-
iii) a mix of grasses, herbs and sedge overlaying a float- pared with 20 % in the billabongs and 10 % in the perma-
ing mat of Salvinia molesta extending from the nent swamps (Table 3).
banks towards the middle of the billabong; and In total there were 139 annual species with 102 terres-
iv) a discontinuous fringe of submerged plants and trial and 37 aquatic species. Eighty-nine of the terrestrial
waterlilies along the edge of the floating mat. species occurred in the fringe zone with 27 only on the
plain which is seasonally inundated for a longer period
As noted above, in recent decades the floodplain vegeta- than the fringe zone. Finalyson et al. (1989) provide a list-
tion has undergone considerable change as a consequence ing of growth strategy and form for each species. This in-
of invasion by feral animals, e.g. buffalo and pigs (Sus cluded 4 growth strategies (annual, perennial, or geophytic
scrofa), as well as weeds, e.g. mimosa (Mimosa pigra), perennial) and 2 primary (terrestrial or aquatic) and 7 sec-
salvinia (Salvinia molesta) and paragrass (Urochloa mu- ondary growth forms (tree, shrub, grass, sedge, vine, palm
tica), changes in fire regimes, and saline intrusion. How- or herb). Terrestrial annuals represented a diverse group of
ever, there have been few attempts to predict changes or species with 60 of them classified as herbs, 18 as sedges
project future scenarios for vegetation management de- and 17 as grasses. Twenty-seven of the aquatic annuals are
spite the extent of change and importance of the flood herbs and 6 are shrubs. There were 68 perennial species,
plains. 50 in the fringe zone. 34 of the perennials are terrestrial,
Finlayson et al. (1989) determined that both the dura- 26 aquatic, with 8 others difficult to classify. There are 12
tion and depth of inundation greatly influenced the distri- terrestrial trees including Eucalyptus spp., Pandanus spi-
bution of plant species, and attempted to explain this ralis, Lophostemon lactifluus and Syzygium suborbiculare.
though an empirical model of plant succession in wet- The aquatic perennial species are dominated by 12 herbs,
lands, developed by Van der Valk (1981), to assess chang- including Hydrilla verticillata, Ludwigia adscendens, Ne-
es in vegetation that occur as a result of changes in the lumbo nucifera and Nymphoides indica, and by 5 grasses,
annual hydrological pattern (Finlayson, 1990; Finlayson including the widespread Hymenachne acutigluma and
and Woodroffe, 1996). Despite limitations the model pro- Pseudoraphis spinescens. There were 14 geophytic spe-
vided a framework for predicting changes in vegetation cies, the more widespread include the Nymphaea and
patterns. Eleocharis species.
Within the broad categorisation of growth strategies
Growth strategies and forms. Finlayson et al. (1989) as- and forms there are many morphological and physiologi-
sessed the growth forms of the 222 plant species found cal adaptations that enable particular species to occupy
on the Magela flood plain across four broad habitat cate- various niches within the cycles of dry and wet condi-
gories, seasonally inundated plain, seasonally inundated tions. Coiwe et al. (2000a) provide an overview of the
fringe zone, billabong and permanent swamp. (The taxo- many specific adaptations that enable plants to survive
nomic listing and hence the number of species recorded the variability due to changes in the hydrological cycle.
by Finlayson et al. (1989) have not been updated in ac-
cordance with the taxonomy presented in the flora pro- Productivity. The productivity of the floodplain vegeta-
vided more recently by Cowie et al. (2000b). The fringe tion on the Magela flood plain has been investigated,
zone covered the edges of the flood plain and included covering seasonal changes in the dry weight of aquatic
the Melaleuca forests/woodlands, with the seasonally grasses and litterfall from Melaleuca trees (Finlayson,
inundated plain habitat covering the remainder of the 1988, 1991; Finlayson et al., 1993b). Changes in the
384 C. M. Finlayson et al. Wetland biodiversity of the Kakadu region
higher, although an analysis of density across the entire
floodplain was not undertaken.
Aquatic invertebrates
Available taxonomic information, zoogeographical rela-
tions and the general ecology of various groups of aquatic
invertebrates of the Region were reviewed by Humphrey
and Dostine (1994). Summary taxonomic data are pre-
sented in Table 4 based on Humphrey and Dostine (1994)
with updated information from Corbett (1996) and an
unpublished database on voucher specimens (Environ-
mental Research Institute of the Supervising Scientist).
This represents approximately 300 micro-invertebrate
(Copepoda, Cladocera, Rotifera) and over 600 macro-in-
vertebrate (insects, worms, mites, larger crustacean
groups, molluscs and sponges) species from freshwater
habitats in the Region, including sandstone plateau
streams and springs, upland (but below-escarpment) sec-
tions of streams with permanent flow, mid-reaches of
streams with seasonal flow, permanent waterbodies on or
adjacent to the main stream channels, and seasonally-in-
undated flood plains near the terminus of major streams.
The composition of the freshwater invertebrate fauna
Figure 6. Above-ground dry weight biomass of three aquatic grass of the Region is both Australian (Williams, 1980) and/or
species (from Finlayson, 1991). The least significant difference (P =
south-east Asian at generic and higher taxonomic levels,
depending upon the group. Across all taxonomic groups,
more data are required before making accurate compari-
sons with the invertebrate fauna in other climatic zones,
above-ground biomass (dry weight/unit area) or the both in Australia and elsewhere (Shiel and Williams,
widespread aquatic grasses Pseudoraphis spinescens, 1990; Humphrey and Dostine, 1994). Nevertheless, may-
Hymenachne acutigluma and Oryza meridionalis were fly (Ephemeroptera) and caddisfly (Trichoptera) data
determined during 1983-84 (Fig. 6; Finlayson, 1991). have been used in comparative global zoogeographical
The dry above-ground biomass of the grass species was studies (Vinson and Hawkins, 2003), while Cranston
influenced by water depth which in itself is directly re- (2000) reported high diversity of chironomid flies in both
lated to the rainfall pattern and surface water flow. The the wet and wet-dry tropics of Australia compared with
annual dry weight production for these species was 0.51 temperate Australia. Cranston et al. (1997) also observed
kg m–2 for Orzya meridionalis, 1.91 kg m–2 for Pseudora- high chironomid species richness in natural, as well as
phis spinescens, and 2.09 kg m–2 for Hymenachne acu- mining-induced, acidic waters of the Region, contrary to
tigluma. findings in temperate studies. They explained this by the
Dry weight production of the widespread Melaleuca large tropical (Australian and south-east Asian) pool of
woodlands and forests on the Magela flood plain were species tolerant of naturally-occurring acidic waters
estimated through an analysis of litterfall data (Finlays- common to these regions. Despite this, the Australian
on, 1988; Finlayson et al., 1993b). In an intensively sam- chironomid fauna is impoverished compared with
pled forest the total litterfall was approximately 0.7 kg northern hemispheric faunas (Cranston, 2000).
m–2 y–1 compared to 1.5 kg m–2 y–1 at a less intensively Kay et al. (1999) remarked upon the similarity of
sampled site (Finlayson, 1988). The value of 0.7 to 1.5 kg macro-invertebrate families in fresh waters of north-
m–2 y–1 is within the range recorded for other forests at the western Australia, compared with elsewhere across
same latitude (Lonsdale, 1988). The above ground bio- northern Australia. Cranston (2000) noted little additional
mass of Melaleuca species on the Magela flood plain was richness and only modest novelty of chironomid species
also calculated using an algorithm relating diameter at in waters of north-western Australia compared to the
breast height to tree height and fresh weight (Finlayson et Kakadu Region, a pattern that appears to hold generally
al., 1993b). This resulted in a calculated tree weight of for most other invertebrate faunal groups (C. Humphrey,
260 ± 0.3 t ha-1 based on a tree density of 294 trees ha–1; unpublished observations). Humphrey (1999) postulated
elsewhere on the floodplain tree densities were much that high seasonality of the extensive lowland aquatic
Overview Article
Aquat. Sci. Vol. 68, 2006 385
Table 4. Number of families, genera and species of aquatic invertebrates recorded in the Kakadu Region. Numbers in italics are
(under)estimates only, while ‘–’ indicates data not available.
Phylum/Class Order Families Genera Species
Mollusca/Gastropoda Pulmonata 4 6 7
5A
Prosobranchiata 2 5
Mollusca/Bivalvia Unionoida and Veneroida 2 2 2
Arthropoda/Insecta Diptera Chironomidae 43 122
7 (non- Chironomidae) – –
Trichoptera 9 21 105
Ephemeroptera 3 14 25
Coleoptera 16 50 100
Hemiptera 11 – –
Odonata 8 54 77
Lepidoptera 1 – 10
Arthropoda/Crustacea Isopoda 1 1 20
Decapoda 5 8 20
Ostracoda – – 6
Conchostraca 2 – 6
Arthropoda/Crustacea/Copepoda Calanoida and Cyclopoida – 7 14
Cladocera 6 30 48
Arthropoda/Arachnida Hydracarina 15 29 –
Annelida Oligochaeta 5 – –
Hirudinea 2 – –
Porifera – 4 7
Rotifera – – 227
A: Includes one exotic species
environments of northern Australia would select animals In contrast to the aquatic insects and the lowland
and plants that were readily dispersed. Moreover, lowland fauna generally of the Region, components of the
freshwater ecosystems of large parts of northern Australia crustacean groups the Decapoda and Isopoda, occurring
are relatively young in geological terms – a feature which, in freshwaters of the ancient ‘stone country’ of the
together with high seasonality and species vagility, has eastern part of the Region contain a substantial and sig-
probably mitigated against endemism at regional and nificant endemic component. This fauna includes an en-
smaller catchment scales (Humphrey, 1999). Some ele- demic family of shrimps, the Kakaducarididae, compris-
ments of the insect fauna from mainly upland sections of ing two endemic genera, Leptopalaemon and Kakaducaris
streams, however, may be more restricted in distribution. (Bruce, 1993; Bruce and Short, 1993), as well as an en-
This includes the mayfly family, the Leptophlebiidae; 7 of demic genus of phreatoicidean isopod (Eophreatoicus),
the 9 species of Leptophlebiids found in the Kakadu Re- of Gondwanic origin, that has exceptional species-level
gion appear to be restricted to the Northern Territory (P diversity (G. Wilson personal communication). Most of
Suter, unpublished data). Indeed, this insect family con- these macro-crustacean species have very restricted dis-
tains the only species in the Region (Tillyardophlebia tributions, often limited to single streams, seeps or
dostinei) so far found to be restricted to a single freshwater springs. This diversity and endemism was attributed by
stream (C. Humphrey unpublished data). Humphrey (1999) to the antiquity and persistence of the
Not only are major invertebrate faunal groups plateau/escarpment and associated perennial streams,
widespread across northern Australia, but at family-level springs and seeps, and isolating mechanisms including
at least, the fauna (including that of the Kakadu Region) fragmentation of habitat (long-term climate changes, ero-
has a high year-to-year persistence (constancy) compared sion) and the generally poor dispersal characteristics of
with other regions of Australia, a feature that is strongly these crustacean groups.
related to the relatively low degree of inter-annual The invertebrate ecology, including seasonal dynam-
variability of stream flow (Humphrey et al., 2000). ics, of seasonally-flowing portions of streams near an
386 C. M. Finlayson et al. Wetland biodiversity of the Kakadu region
Table 5. Diversity of freshwater fishes recorded from the Kakadu Region.
Order Family Number of genera Number of species Marine vagrant (m) or catadromous (c)
Pleurotremata Carcharhinidae 2 2 m
Pristidae 1 1 m
Hypotremata Dasyatidae 1 1 m
Osteoglossiformes Osteoglossidae 1 1
Elopiformes Megalopidae 1 1 c
Anguilliformes Anguillidae 1 1 c
Clupeiformes Clupeidae 2 3 m
Siluriformes Ariidae 1 3
Plotosidae 3 5
Atheriniformes Atherinidae 1 2
Synbranchiformes Synbranchidae 1 1
Pleuronectiformes Cynoglossidae 1 1
Soleidae 2 2 ?
Perciformes Belonidae 1 1
Melanotaeniidae 1 5
Pseudomugilidae 1 2
Ambassidae 2 3
Centropomidae 1 1 c
Terapontidae 5 6
Apogonidae 1 1
Toxotidae 1 2
Mugilidae 1 2 c
Gobiidae 1 3 c
Eleotrididae 5 8
Scatophagidae 2 2 m
Kurtidae 1 1
Gerreidae 1 1 m
Total 27 42 62
operating uranium mine (Ranger mine) and a uranium and Cranston, 1995; Hardwick et al., 1995; Humphrey et
ore deposit (Jabiluka) has been investigated for all phases al., 2002). The responses of aquatic invertebrates to a
of the hydrological cycle, including recolonisation after range of natural and human-related disturbances in the
early Wet season re-wetting, early-mid Wet season flows, Region have also been investigated, including fire re-
recessional flow period and pool-formation phase. Pal- gimes, invasive wetland grasses and tourism (Douglas,
tridge et al. (1997) found that upon re-wetting with the 1999; Douglas and O’Connor, 1999; M. Stowar unpub-
first stream flows in Magela Creek, most recolonising lished information).
taxa were derived from perennial upper reaches through
drift, though contributions were also derived from adja-
cent billabongs and resting stages (especially microcrus- Fishes
taceans) in the sandy substratum. The seasonal dynamics Freshwater fishes are among the most comprehensively
of aquatic invertebrates from permanent waterbodies on studied aquatic organisms in the Region (Taylor, 1964;
or adjacent to the main stream channels have also been Midgley, 1973; Pollard, 1974; Bishop et al., 1986, 1990,
studied (Marchant, 1982a, b, c; Outridge, 1988). 2001). The ecology, community structure and
Much effort has been directed at the use of aquatic biogeographic relationships of the fish fauna are
invertebrates for detecting and assessing mining impact reasonably well known and have been described and
(and recovery) in the Region (Humphrey, 1990; Hum- reviewed by Bishop and Forbes (1991). Sixty two
phrey and Dostine, 1994; Faith et al., 1991, 1995; Smith freshwater fish species have been recorded, represented
Overview Article
Aquat. Sci. Vol. 68, 2006 387
by 44 entirely freshwater species, 4 species that repro- tidal waters (H. Larson personal communication). The
duce in estuarine or marine waters (catadromous) and 14 two other shark species, the Bizant River Shark (Glyphis
marine or estuarine species that commonly enter non- sp. A) and the Northern River Shark (Glyphis sp. C) are
tidal freshwaters (marine vagrants) (Table 5). listed respectively as critically endangered and en-
All except three Australian freshwater fish (Neocera- dangered under IUCN criteria. Their distribution is
todus forsteri, Scleropages leichardti and S. jardinii) are limited to only a few river systems in northern Australia,
comparatively recent descendents from marine families. although it is acknowledged that further specimens and
When this recent colonisation of freshwater habitats is taxonomic work are required to properly determine the
combined with the small area of freshwater habitats on distribution and status of these species (Larson, 2000;
the Australian continent it is not surprising that the spe- Pogonoski et al., 2002).
cies richness of the continent (302 species) is very low by Longitudinal zonation of fish species is not very pro-
global standards (Allen et al., 2002).The diversity of nounced along the major stream channels. Marine va-
freshwater fish is higher in the tropics compared to tem- grants are more likely to be found in freshwater bodies
perate regions of Australia. Shiel and Williams (1990) close to the limit of tidal influence. Upstream there is lit-
suggested that periods of extreme aridity in the southern tle variation in the species composition until the first
temperate areas and ease of migration of new species dispersal barriers occur near the escarpment of the sand-
from marine to freshwaters as a result of greater length stone plateau. In this lowland zone on most major streams
and low-gradient of tropical rivers were most likely to 20 to 30 species are likely to be encountered at a single
determine species richness of freshwater fish. However, location. Upstream from dispersal barriers, such as wa-
Bishop and Forbes (1991) and more recently Unmack terfalls and large cascades, species diversity declines
(2001) have suggested an evolutionary explanation argu- greatly and pools on the plateau may contain only one or
ing that freshwater fish diversity in Australia is a reflec- two species. Most species utilise both riverine and asso-
tion of the fish diversity in adjacent seas and this is ciated lentic wetlands when available. There are 6 spe-
greater in the tropics. The fish diversity in the Kakadu cies that are restricted to riverine habitats and another 2
Region is relatively high by comparison to other regions that rarely enter lentic wetlands. At least 3 species occur
within the continent - this may be related to the rich di- only in small tributaries within the plateau and escarp-
versity of marine fish in the Indo-Pacific archipelago. ment zone.
Catchment size is a major factor determining fish di- Many of the fish species in the Region are common
versity in different drainage systems. Bishop and Forbes across northern Australia. Of 114 species known from
(1991) compared the diversity of fish in streams in the rivers of the wet-dry topics of northern Australia 32 oc-
Region with a generalised relationship between catch- cur throughout the entire zone; 11 are restricted to the
ment area and diversity developed for tropical rivers by Arafura bio-region of the Northern Territory; and 3 are
Welcomme (1979). By separating the drainage basin into endemic to the Region and nearby western Arnhemland,
sub-catchments they found that streams in the Region including Mariana’s Hardy Head (Craterocephalus mari-
had higher diversity than many other streams of similar anae), the Sharp-nose Grunter (Syncomistes butleri) and
size elsewhere; however, this seems spurious as it as- Midgley’s Grunter (Pingalla midgleyi). (Recent surveys
sumes that fish diversity in a sub-catchment is independ- have extended the known range of the hardyhead and the
ent of the rest of the catchment. Pusey and Kennard grunter.) Another 3 species have very disjunct distribu-
(1996) in comparing the fish diversity of 10 tropical riv- tions across northern Australia and Papua New Guinea,
ers along the far northern east coast of Queensland found including the Coal Grunter (Hephaestus carbo), Banded
that while catchment size is an important factor deter- Rainbowfish (Melanotaenia trifasciata) and Black-
mining species diversity, differences in fish community striped Rainbowfish (Melanotaenia nigrans). The latter
structure were related to latitude. reflect an historic freshwater connection between Aus-
About half the fish species encountered in the Kakadu tralia and Papua New Guinea that was severed by sea-
Region are small to medium in size, being usually less level rises within the past 30,000 years.
than 30 cm in length. The small-bodied fish fauna of the Upstream escarpment habitats are nutrient poor and
Region is dominated in overall abundance by are dominated by herbivore/detritivores and omnivores.
centropomids (perchlets), melanotaeniids (rainbow fish) The Sharp-nose Grunter has specially adapted teeth, jaws
and atherinids (hardyheads); the larger bodied fish fauna and intestines for grazing epiphytes from rocks and logs,
is dominated by ariids and plotosids (Fork and Eel tailed whilst the archerfish (Toxotes chatareus and Toxotes
Catfish), clupeids (Boney Bream), megalopids (tarpon) lorentzi) can capture terrestrial prey by emitting a jet of
and to a lesser extent, terapontids (grunters). Some other water from the mouth to dislodge the prey from
larger species include Barramundi (Lates calcarifer), overhanging vegetation (Bishop et al., 2001). A range of
Saratoga and three estuarine shark species. Only one trophic categories is represented amongst the fish com-
shark, the Bull Shark Carcharinus leucas, enters non- munities of lowland waterbodies, that are generally nutri-
388 C. M. Finlayson et al. Wetland biodiversity of the Kakadu region
ent rich. Fish in the Region exhibit a number of breeding lowing commercial exploitation except for limited
strategies, with a high proportion of species exhibiting catches for the aquarium fish trade.
some form of parental care, including live bearing, buccal
incubation and nest guarding (Bishop et al., 2001).
A characteristic feature of the ecology of the fish Amphibians
communities is the movement of fish between habitats Australia’s amphibia are represented by five families in
with the re-linking of Dry season refuge areas of the the order Anura (the frogs and toads) with four families
escarpment, corridor and floodplain zone once Wet of native frogs (the Microhylidae, Myobatrachidae (or
season flows commence (Bishop and Forbes, 1991). Leptodactylidae), Hylidae (or Pelodryadidae), and the
Feeding, breeding and recruitment of juveniles mostly Ranidae) and the introduced cane or marine toad, Bufo
occurs during the Wet season months with most species marinus from the Bufonidae family (Robert and Watson,
spawning early in the Wet season, taking advantage of 1993). The Kimberley region and the northern part of the
the increased area of habitat available to supply resources Northern Territory which contains the Kakadu Region
for growth and reproduction (Bishop et al., 2001). Many are considered to be an area of significant amphibian di-
species breed in off-river water bodies of lowland versity. Despite this general information there are many
streams and in the vast areas of inundated flood plain. In gaps in the knowledge about the ecology and biology of
some seasonal tributaries of the larger Alligator Rivers, many frog species of this area, with new species being
including Magela and Nourlangie Creeks, there are found as recently as 1997 and 2001 (Northern Territory
massive upstream migrations in the late Wet season of Frogs Database, 2003).
mainly small bodied fishes (dominated by ambassids and The first comprehensive study of frogs in the Kakadu
melanotaeniids) from floodplain nursery areas to Dry Region was undertaken in the Wet seasons of 1978–1979
season refuges (Humphrey and Dostine, 1994). These and 1979–1980 (Tyler et al., 1983). Woinarksi and Gam-
spectacular migrations often consist of several hundred bold (1992) undertook a further survey to examine distri-
thousand fish per hour moving past a single observation bution patterns of herpeto-fauna in relation to environ-
point (Bishop et al., 1995). mental gradients (substrate, moisture availability,
Large natural fish mortalities, or ‘fish kills’ are a vegetation); frogs were found to be most strongly associ-
common occurrence in the Region (Pidgeon, 2001). ated with substrate and moisture availability and oc-
These are caused by a number of factors with low curred in 4 distinct assemblages – a sandstone assem-
dissolved oxygen levels the most common cause. Influx blage, lowland clay-flat assemblage, wet forest
of toxic chemicals by natural processes (weakly acidic assemblage and two species with idiosyncratic ranges
waters containing toxic levels of aluminium, and leached (Woinarski and Gambold, 1992). These surveys have re-
ichthyocidal compounds such as saponins from trees and corded 25 frog species from 3 families – the Hylidae with
shrubs), and disease have also been recorded. Disease 2 genera and 16 species; the Myobatrachidae with 5 gen-
rarely results in large-scale mortalities at one time. Most era and 8 species); and the Microhylidae with 1 genera
fish kills occur during the early Wet season when the first and 1 species (Table 7). The cane toad is a recent invader
flows introduce oxygen depleting organic matter from (van Dam et al., 2002a).
the catchment and concentrated contaminants (described Most of the frog species display a clear seasonal di-
above). Storms and flows can mix anoxic water from the chotomy of behaviour, being inactive during the Dry
bottom of billabongs and stir up sediment and detritus season with at least 9 species aestivating underground
which can further reduce oxygen levels. Fish kills often (Tyler and Crook, 1987). Some of these species are fos-
result in the deaths of many thousands of fish and usually sorial (Cyclorana australis, C. longipes, Limnodynastes
involve the larger bodied fish such as Barramundi, Fork ornatus), while others (such as Litoria nasuta, L. inermis,
and Eel-tailed Catfish, Tarpon and Mullet (Bishop, 1980; L. wotjulemensis and L. tornieri) are known to descend
Bishop et al., 1982; Brown et al., 1981; Noller, 1983; down deep cracks in the hardened mud at the edge of bil-
Pidgeon, 2001). labongs (Tyler and Crook, 1987). In contrast, during the
There is no commercial fishery in freshwaters of the Wet season, frogs are very obvious and abundant (Tyler
Region, but there are important artisanal fisheries by in- and Crook, 1987) and are an important food source for a
digenous people and recreational fishery. The latter is variety of birds, reptiles and fish (Morris, 1996).
largely a sport fishery targeting only one species, the Bar- The sandstone assemblage of frogs is endemic to
ramundi. This fishery is highly managed with minimum northern Australia with species occupying crevices near
size limits, small bag limits and seasonal closures that small perennial streams in the sandstone escarpments and
vary with measured fishing pressures. Upland sections of plateaux (Braithwaite et al., 1991; Morris, 1996; Woinar-
streams in Kakadu are generally off-limits to (non-indig- ski and Gambold, 1992). These species cope with ex-
enous) recreational fishing. Otherwise, there is no man- treme environmental conditions with the tadpoles tolerat-
agement for other freshwater species other than not al- ing temperatures of around 42 °C, and the adults in the
Overview Article
Aquat. Sci. Vol. 68, 2006 389
Table 6. Frog species recorded from the Kakadu Region (from Press et al., 1995, taxonomy as per Northern Territory Frogs Database,
2003).
Family Species Common name(s)
Myobatrachidae Crinia bilingual Bilingual Frog, Ratchet Frog
Hylidae Cyclorana australis Burrowing Frog, Giant Frog
Hylidae Cyclorana longipes Long-footed Frog
Hylidae Litoria bicolour Northern Dwarf Tree Frog
Hylidae Litoria caerulea Green Tree Frog
Hylidae Litoria coplandi Saxicoline Tree Frog, Copland‘s Rock Frog
Hylidae Litoria dahlia Dahl‘s Aquatic Frog
Hylidae Litoria inermis Peter‘s Frog
Hylidae Litoria meiriana Rockhole Frog
Hylidae Litoria microbelos Dwarf Rocket Frog, Javelin Frog
(formerly Litoria dorsalis)
Hylidae Litoria nasuta Rocket Frog
Hylidae Litoria pallida Pale Frog
Hylidae Litoria personata Masked Cave Frog, Masked Rock Frog
Hylidae Litoria rothii Roth‘s Tree Frog
Hylidae Litoria rubella Desert Tree Frog, Red Tree Frog
Hylidae Litoria tornieri Tornier‘s Frog
Hylidae Litoria wotjumulensis Wotjulum Frog
Microhylidae Austrochaperina aldelphe Northern Territory Frog
(formerly Sphenophyrne adelphe)
Myobatrachidae Limnodynastes convexiusculus Marbled Frog
Myobatrachida Limnodynastes ornatus Ornate Burrowing Frog
Myobatrachidae Megistolotis lignarius Capenter Frog
Myobatrachidae Notoden melanoscaphus Northern Spadefoot Toad, Golfball Frog
Myobatrachidae Uperoleia arenicola Jabiru Toadlet
Myobatrachidae Uperoleia inundata Floodplain Toadlet
Myobatrachidae Uperoleia lithomoda Stonemason Toadlet
Dry season with limited surface water, low humidity and and Gambold, 1992). Most of these species are burrowing
high daytime temperatures (Morris, 1996). The sand- frogs (Uperoleia spp., Limnodynastes convexiusculus and
stone specialist assemblage consists of Megistolotis liga- Cyclorana australis). One of the burrowing frog species,
narius, Litoria personata, Litoria meiriana and Litoria Uperoloeia arenicola, is endemic to the Region. Another
coplandi. During the Dry season these species seek ref- group of frogs is most abundant in riparian or monsoon
uge in sandstone crevices and caves in the escarpment. forests, comprising mainly tree frogs (Litoria spp., and
M. lignarius and L. personata are endemic to the Kakadu Crinia bilingua) (Woinarski and Gambold, 1992).
sandstone escarpment.
Two species are found throughout the Region and
around Australia, without any strong association to envi- Reptiles
ronmental variables. Litoria caerulea has a widespread The reptile fauna of the Region is one of the most diverse
distribution throughout northern Australia and along the in Australia, with 127 species recorded (Table 7) and in-
east coast (Woinarski and Gambold, 1992). The burrowing cluding only one introduced species, the Asian House
frog Limnodynastes ornatus also has a widespread distri- Gecko (Hemidactylus frenatus). The herpeto-fauna com-
bution throughout Australia, but in the Kakadu Region is prise 15 families and 62 genera. From a conservation
most abundant on sandy flats with tall forests (Woinarski viewpoint, using the IUCN criteria (2000), the Logger-
and Gambold, 1992). The highest diversity of frogs within head Turtle (Caretta caretta) is considered endangered
the Region occurs on lowland wet clay flats, with open whereas there is little concern for the Green Turtle (Che-
vegetation and extensive tussock grass cover (Woinarski lonia mydas); both species of sea turtle have been re-
390 C. M. Finlayson et al. Wetland biodiversity of the Kakadu region
Table 7. Reptile fauna of the Region (from Press et al., 1995).
Order Suborder Family Number of Genera Number of Species
Crocodilia Crocodylidae 1 2
Testudines Carettochelydidae 1 1
Chelidae 3 5
Cheloniidae 4 4
Squamata Sauria Agamidae 5 9
Gekkonidae 7 16
Pygopodidae 3 4
Scincidae 11 36
Varanidae 1 11
Squamata Serpentes Acrochordidae 1 2
Boidae 3 6
Colubridae 8 8
Elapidae 10 13
Hydrophiidae 3 3
Typhlopidae 1 7
Total 15 62 127
ported nesting (Miles, 1998) on an island that is free of Of 11 species of goanna found in the region 5 are
most of the mainland egg eaters, such as dingos, feral found in aquatic or semi-aquatic habitats. In contrast to
pigs and humans (Miles, 2000). The Flatback (Nator de- the Mangrove Monitor, Merten’s (Varanus mertensi) and
pressa) and Pacific Ridley (Lepidochelys olivacea) Tur- Mitchells (Varanus mitchelli) Water Monitors are usually
tles have also been recorded nesting (Vanderlely, 1995). associated with freshwater streams and billabongs, ob-
The Oenpelli Python (Morelia oenpelliensis) is listed as taining approximately 40 % of their food (biomass) from
vulnerable. aquatic systems (Shine, 1986a). The sand goannas (Vara-
Eighteen species of large reptiles are listed as using nus gouldii and Varanus panoptes) occur over a range of
riparian and floodplain habitats of the Magela Creek sys- habitats, but frequently use dry flood plains and the edges
tem within Kakadu (Finlayson et al., 1990) with a further of watercourses to forage for food. The proportion of
6 water-dependent species. The latter includes: the Pig- prey from aquatic habitats that is eaten by V. panoptes is
nosed Turtle (Carettochelys insculpta); the Mangrove high, up to 30 % by weight, but negligible in V. gouldii
Monitor (Varanus indicus), restricted to mangrove com- (Shine, 1986a).
munities along the estuaries; the Little File Snake Six species of turtle are found in the freshwater
(Achrocordus granulatus); and three species of aquatic reaches of the rivers, creeks and billabongs in the Region.
colubrid snakes (Cerberus rynchops, Fordonia leuco- The Pig-nosed Turtle, although common where it occurs
balia and Myron richardsonii), usually found in estuarine in both Australia and Papua New Guinea, has a fairly
and mangrove habitats. limited distribution in the Region and, hence, the South
Two species of crocodile have a broad distribution Alligator River in Kakadu represents a significant refuge
across northern Australia and are plentiful in the region. (Press et al., 1995). Chelodina burrungandjii and Elseya
The Estuarine or Saltwater Crocodile (Crocodylus poro- latisternum are found in pools on the sandstone plateau,
sus) inhabits coastal rivers extending well inland via while Elseya dentata inhabits sandy or rocky stretches
major rivers and billabongs on the river floodplains. The between the escarpment and floodplains. (Press et al.,
Freshwater Crocodile (Crocodylus johnstoni) inhabits 1995). The characteristic floodplain turtle is the Northern
permanent freshwater rivers, lagoons and billabongs Long-necked Turtle (Chelodina rugosa) which is carniv-
across northern Australia (Cogger, 2000). There are no orous and during the Dry season survives by burying its-
reliable population estimates for these species in the Re- self in the mud and aestivating.
gion although it is generally considered that they are The Arafura File Snake (Achrochordus arafurae) is
populous and have recovered from the decimation that mainly restricted to the larger billabongs in the Dry sea-
occurred as a consequence of hunting activities that were son with many more in lower mainstream billabongs than
formally stopped in 1971. in the upper reaches of creeks (Shine, 1986b). In the Wet
Overview Article
Aquat. Sci. Vol. 68, 2006 391
Table 8. Waterbirds families listed in the Asia-Pacific Migratory
season most are found in shallow, inundated grasslands
Waterbird Strategy 2001–2005 and found in the Kakadu Region.
(Shine, 1986a). They are a popular food item for Abo-
riginal people in the Region (Shine, 1986b). The Little Taxonomic Family Common Name Number
File Snake inhabits estuarine habitats and feeds on crabs of species
and small fish. Other species of snake that occur on the
Podicipedidae Grebes 2
floodplains include the Water Python (Liasis fuscus) and
Phalacrocoracidae Cormorants and Darter 5
the King Brown Snake (Pseudechis australis). The Water and Anhingidae
Python has been recorded at the extraordinary high den-
Pelecanidae Pelicans 1
sity of 714 km2 on the nearby Adelaide River floodplain
Ardeidae Herons, Egrets and Bitterns 12
(Madsen and Shine, 1996). As water levels drop with the
Ciconidae Storks 1
onset of the Dry season a number of species, including
Threskiornithidae Ibises and Spoonbills 5
the Olive Python (Liasis olivaceus) and the Northern
Death Adder (Acanthophis praelongus) move onto the Anatidae Swans, Geese and Ducks 12
drying floodplain. The Keelback or Freshwater Snake Gruidae Cranes 2
(Tropidonophus mairii) is found in freshwater habitats, Rallidae Rails, Gallinules and Coots 7
whilst several species of colubrid snakes inhabit man- Jacanidae Jacanas 1
grove areas where they feed on crustaceans and fish.
Haematopodidae Oystercatchers 2
Three species of sea-snakes have been recorded in the
Recurvirostridae Stilts and Avocet 2
estuarine and tidal reaches of the rivers – Hardwick’s
Glareolidae Pratincoles 2
Sea-snake (Lapemis hardwickii), Darwin Sea-snake (Hy-
Chradriidae Plovers 11
drelaps darwiniensis) and Stoke’s Sea-snake (Astrotia
stokesii) (Press et al., 1995). Scolopacidae Sandpipers 22
The rate of endemism within the area is high. Sadlier Laridae Gulls, Terns and Skimmers 12
(1990) found 74 reptile species in the Magela Creek re-
gion, with 13 % appearing to be endemic to the wider
sandstone plateau.
systems of Kakadu National Park for waterbirds has
been recognised in the listing of the park under the Ram-
Waterbirds sar Wetlands Convention. The high diversity and abun-
The coastal floodplains and the alluvial floodplains of dance of waterbirds species are also a contributing factor
the Region are of immense importance for waterbirds to the listing of Kakadu National Park as a World Herit-
(Morton et al., 1991). Of the 20 families of waterbirds age area.
accepted under the Ramsar definition and included in the Within the Kakadu Region numerous roosts of shore-
Asia-Pacific Migratory Waterbird Strategy (2001–2005), birds, containing 2,000 or more birds, are spread along
16 are found in the Region (Table 8). These families the coastline with almost 9,000 individuals recorded at
comprise 99 species of waterbirds, of which 52 are shore- Finke Bay in a survey in September 1993 (Chatto, 2003).
birds (or waders) and 32 migratory (Morton et al., 1990; Ground surveys have also been used with around 12,500
Chatto, 2003). individuals being recorded in late April 1992 and late
The importance of the Region to waterbirds has been March 1992, along the coast between the South Alligator
extensively documented over many years. The wetlands River and Minimini Creek. The more numeros species
support many species of national and international sig- included Whimbrel (Numenius phaeopus), Eastern Cur-
nificance, including those protected under the Japan lew (Numenius madagascariensis), Common Green-
Australian Migratory Bird Agreement (JAMBA) and the shank (Tringa nebularia) and Black-tailed Godwit
China Australia Migratory Bird Agreement (CAMBA). (Limosa limosa). The mangroves fringing the river banks
The flood plains in particular provide important refuge provide suitable habitat for nesting colonial waterbirds
areas critical to the conservation of waterbirds through- with large multi-species colonies of Intermediate Egret
out northern Australia (Morton et al., 1991). During the (Egretta intermedia), Cattle Egret (Ardea ibis), Great
Dry season months of August-October the floodplains Egret (Ardea alba), Little Egret (Egretta garzetta), Pied
are used intensively by up to 2 million waterbirds, in- Heron (Ardea picata), Little Pied Cormorant (Phalacroc-
cluding large concentrations of geese and ducks (Morton orax melanoleucos), Little Black Cormorant (Phalacroc-
et al., 1991; Bayliss and Yeomans, 1990), such as the orax sulcirostris) and Australian White Ibis (Threskiornis
Magpie Goose (Anseranas semipalmata), the Wandering molucca) being recorded between 1990 and 1999 (Chat-
Whistling Duck (Dendrocygna arcuata) and Plumed to, 2000). In all 18 species of waterbirds have been re-
Whistling Duck (Dendrocygna eytoni). The importance corded in the mangroves and 27 on the inter-tidal flats.
of the mosaic of contiguous wetlands across the river Waterbird usage of the flood plains varies seasonally with
392 C. M. Finlayson et al. Wetland biodiversity of the Kakadu region
internationally important (i.e. contain more than the
1 %% of the population in the East Asian-Australasian
Flyway; Chatto, 2003).
Many other bird species that are not regarded as
waterbirds, as described above, depend on wetlands in
the Region for at least part of their life cycle (Table 9).
Whilst the Region is recognised as providing important
habitats for these species their ecology and population
levels have not on the whole been been investigated. The
number of wetland-dependent bird species present in the
habitat categories used in Table 9 are: coastal habitats 16;
mangroves 67; monsoon forests 99; and floodplain and
riparian areas 137. The riparian habitats in particular sup-
port many wetland-dependent species, such as Kingfish-
ers and the White-breasted Sea-eagle, and small passeri-
Figure 7. Seasonal use by waterbirds of four floodplains (Nour-
formes, such as Honeyeaters and Flycatchers, with about
langie, Magela and Cooper Creeks and East Alligator River) in the
Kakadu Region, as determined by aerial surveys (from Morton et half of them being endemic to northern Australia (Mor-
al., 1984).
ton and Brennan, 1991).
some species moving between the floodplains as habitat Mammals
conditions change in response to rainfall and flooding Many mammals may be considered wetland-dependent
and become more or less suitable for breeding, roosting in the sense that they access the water, although very few
and/or feeding (Morton et al., 1984). Waterbird usage of species spend their entire life cycle in the wetlands. A
four floodplains in the early 1980s is shown in Figure 7. notable exception is the Dugong (Sirenia artiodactyla).
The diets of aquatic herbivourous species, such as the A number of mammals are known to make some use of
Magpie Goose, Wandering Whistling Duck, Plumed the floodplain environment in the Northern Territory,
Whistling Duck, and the Green Pygmy Goose (Nettapus including 2 species of rodents, 2 species of dasyurids
pulchellus) are linked to the phenological state of the (carnivorous marsupials), 3 macropods, a number of in-
floodplain plants many which flower and seed towards sectivorous bats, 2 species of flying foxes, and the dingo
the end of the Wet season (Morton et al., 1990a, b; Fin- (Cowie et al., 2000a). The Dusky Rat (Rattus colletti)
layson et al., 1990). As the water recedes during the Dry seems to be a characteristic species that shelters in
season, a number of waterfowl species from southern cracks in the floodplain soils during the Dry season,
Australia, such as Hardhead (Aythya australis), Grey Teal feeds on seeds, and stems and corms of sedges and
grasses, and reaches high densities (around 150 ha–1 be-
(Anas gracilis), Pink-eared Duck (Malacorhynchus
membranaceus) arrive. The Dry season grasslands host ing recorded on a nearby floodplain; Madsen and Shine,
migratory species such as the Little Curlew (Numenius 1996). In the Wet season they move to adjacent higher
minutus) (Garnett and Minton, 1985; Bamford, 1988a, ground from which it recolonises the floodplain as they
1990; Morton et al., 1991; Schulz, 1989). The Little Cur- dry. The False Water Rat (Xeromys myoides) is seem-
lew, a species that breeds in Siberia and over-winters in ingly not as common as the Dusky Rat and may in fact
Australia, arrive in the Region in the latter part of Sep- be vulnerable. The fringing vegetation, in particular the
tember; they build up in numbers until the onset of the Melaleuca trees, provide roosting sites for colonies of
Wet season, at which time they move to the sub-coastal flying foxes (Pteropus scapulatus and Pteropus alecto)
floodplains some distance to the east and west of the Re- and when flowering are a source of nectar which also
gion (Bamford, 1990). Morton et al. (1991) estimated attracts the Northern Blossom Bat (Macroglossus min-
approximately 300,000 Little Curlew staging through the imus).
Region during October in the early 1980s; Bamford Introduced mammals that have invaded or make ex-
(1990) estimated 50,000 were present in the late dry sea- tensive use of the wetlands include the Asian water buf-
sons of 1987–1989. Other migratory shorebirds occur- falo (Bubalus bubalis), European domestic pig (Sus sco-
ring in large numbers during the migratory periods and fa), cattle (Bos taurus) and horses (Equs callabus). All
using the freshwater wetlands of the Region as well as are considered pest species with the water buffalo having
neighbourhood floodplains include the Sharp-tailed overgrazed and degraded the wetlands and except for a
Sandpiper (Calidris acuminata) and Marsh Sandpiper small domesticated herd have been removed from the
(Tringa stagnatilis), both species, as with the Little Cur- Park (Skeat et al., 1996).
lew, being present in numbers that make these wetlands
Overview Article
Aquat. Sci. Vol. 68, 2006 393
Table 9. Other wetland-dependent birds (i.e. those not classed as waterbirds) present in the Kakadu Region.
Family group Coastal wetland Mangrove Moonsoon forest Floodplain and
Riparian areas
Mound-builders 1 1
Quails 3
Kites-Hawks-Eagles 5 7 10 15
Falcons 1 6
Rails-Crakes-Swamphens-Coots 3 1 5
Bustards 1
Pigeons-Doves 3 6 4
Cockatoos 1 2 4 5
Lorikeets-Parrots 1 3 4
Cuckoos-Koels-Coucals 3 6 8
Owls 3 3
Barn Owls 2
Frogmouth 1 1
Nightjars 1 1 2
Owlet-Nightjars 1 1
Swifts 2 2 2 2
Kingfishers 2 5 5 6
Bee-eaters 1 1 1 1
Rollers 1 1 1 1
Pittas 1 1
Australian treecreepers 1
Fairy-wrens 1 1
Australian warblers 2 4 4
Honeyeaters-Chats 8 13 15
Robins 2 2 3
Babblers 1 1
Sitellas 1
Whistlers-Shrike-thrushes 5 4 4
Monarchs-Fantails-Drongos and Mag- 9 9 9
pie Larks
Cuckoo-shrikes 4 5 5
Orioles and Figbirds 2 3 3
Woodswallows and Butcherbirds 1 2 3 5
Crows 1 1 1 1
Bowebirds 1 1
Larks 1
Pipits and Wagtails 1 2
Weavers-Waxbills and allies 3 9
Flowerpeckers 1 1 1 1
Threats issues faced by floodplain wetlands in northern Australia
(Storrs and Finlayson, 1997; Finlayson et al., 2005).
A number of general analyses and reviews over the past Many of these affect the Kakadu Region, although the
few decades have identified the threats and management relative importance of each varies across the flood plains.
394 C. M. Finlayson et al. Wetland biodiversity of the Kakadu region
As elsewhere in northern Australia these issues have buthiuron (van Dam et al., 2004), uranium mining (van
been assessed in site-based analyses within the Region Dam et al., 2002b) and climate change (Bayliss et al.,
and some comprehensive databases now exist (e.g. man- 1997; Eliot et al., 1999). It is evident that information
agement of specific invasive species – Cook et al., 1996, collation and analysis is not only underway, but is ongo-
Rea and Storrs, 1999 - and uranium mining – Humphrey ing; the challenge is to translate this information into
et al., 1999, Gardner et al., 2002). Other issues are being practical advice for managers and for managers to heed
assessed, but not necessarily in a coherent manner (e.g. the warnings.
coastal and climate change – Eliot et al., 1999). The The complexity of managing wetlands within the
likely impacts of global climate change are perhaps the Kakadu Region is illustrated by two examples raised by
most insidious and least understood by local managers Finlayson (2005). The first considers changes in burning
(Bayliss et al., 1997). It is anticipated that climate change patterns on the floodplains. Recent investigations within
will not only affect the biological, chemical and physical Kakadu National Park have been trialing the use of fire to
features of wetlands, but will also interact strongly with establish floodplain vegetation mosaics more suited for
other existing or emerging pressures on wetlands with access by traditional Aboriginals for hunting and food
synergistic, antagonistic or cumulative effects. collection. In this instance access had been restricted by
Information on many threats is increasingly being an overabundance of the native grass Hymenachne acu-
collected through the adoption of structured risk assess- tigluma that had spread since the removal of buffalo from
ment procedures, such as that shown in Figure 8 (Finlay- the plains in the mid-1980s. Thus, we have a scenario
son and van Dam, 2004). For many of the main pest where the removal of a feral animal has been accompa-
species the extent of their invasion of the wetlands has nied by the expansion of a native plant that restricted
been assessed to some extent, although often incom- Aboriginal access to traditional food resources on the
pletely. In many instances the biology of the species may floodplains. Importantly, these investigations are being
also be known or is being studied; however, vital infor- led by local people with assistance from park managers
mation on the ecological changes wrought by these spe- and researchers (P. Christopherson and P. Bayliss per-
cies is often confined to a few isolated studies, if any, sonal communication). The second example covers the
and/or anecdotal evidence. Further information is be- complex biotic inter-relationships that occur on the
coming available through risk assessments for Mimosa floodplains, such as those between the large populations
(Mimosa pigra; Walden et al., 2004), cane toads (Bufo of magpie geese and the vegetation resources that they
marinus; van Dam et al., 2002a), the herbicide te- require for feeding and nesting. Finlayson (2005) specu-
lated that the number of geese could affect the future ex-
tent of at least some of these plant species either through
direct consumption or when nesting by damaging the
stems with subsequent physiological effects and changes
in growth conditions.
On the whole there has been little investigation of the
interactions between the ecological components of the
floodplain wetlands; some exceptions occur, but such
investigations do not seem to be considered de rigeur for
management planning or indeed for guiding responses to
the many threats faced by the wetlands. The recent focus
on risk assessments has gone someway towards provid-
ing information that can be more readily used for man-
agement purposes. There is however, an evident gap be-
tween collecting information on the biodiversity of the
wetlands and providing science-based guidance for man-
agement purposes, or for this guidance to be accepted
and used. The scenario shown through the assessment of
the likely impact of climate change and sea level rise
(Bayliss et al., 1997; Eliot et al., 1999) illustrates this
disjuncture. Climate change and sea level rise is project-
ed to result in the replacement of many freshwater wet-
lands with saline wetlands and with consequent displace-
ment or loss of the freshwater species diversity, and to
exacerbate the pressure from existing threats and pres-
Figure 8. Generalised framework for ecological risk assessment
sures. Yet, management steps have not been agreed, pos-
(modified from US EPA, 1998).
Overview Article
Aquat. Sci. Vol. 68, 2006 395
sibly due to limited awareness or understanding of the Region is sought it will be necessary to revisit past sur-
issues, or even resignation that adaptation rather than veys etc. and ascertain how much has changed – it is not
mitigation is a better response. At the same time full ad- fully known how well the information referred to in this
vantage has not been undertaken of monitoring opportu- paper reflects the current biodiversity status. It may also
nities, in this instance the existence of transects estab- be useful to address the priorities for further biodiversity
lished in the mangroves along the East Alligator River survey – should there be more effort seeking information
some 20 years ago as well as new technologies using GIS on endemic or vulnerable species, or are more general
and remotely sensed data, as shown by the mangrove surveys required first? Is there sufficient baseline or ref-
change detection undertaken along the West Alligator erence data to assess the effect of existing let alone fur-
River (Lucas et al., 2002). Making better use of such op- ther invasive species? Does this data correlate with and/
portunities is a challenge for researchers and managers or incorporate appropriate local information?
alike. The wetland landscape of the Kakadu Region has re-
ceived wide recognition and the available data supports
this. It is as widely recognized that these wetlands have
Conclusions undergone major change and face major pressures now
and in the near-future. Addressing these pressures is pos-
The species diversity of the wetlands of the Kakadu Re- sible – the tools are available and more recent manage-
gion is diverse and profuse. Many taxa have been identi- ment approaches have sought to involve local people and
fied and species distributions and seasonal and annual increase transparency in decision making about manage-
variations investigated. The information base though is ment actions as well as research priorities. The next steps
incomplete and uneven. Much data have been collected in ensuring the maintenance of the biodiversity values of
in specific locations with few region-wide investigations the wetlands of the Kakadu Region are complex, espe-
and with a few notable exceptions, e.g. Russell Smith et cially if they are subject to increasing pressure from both
al. (1997), based mainly on surveys undertaken by con- existing and emerging drivers of change. It is further an-
temporary scientific means with limited input from the ticipated that more sophisticated ecological understanding
local aboriginal community. The latter is changing and of the processes that support the food webs in the wetlands
increasingly local people are participating in joint activi- (Douglas et al., 2005) will assist in future management.
ties or initiating their own analyses in relation to changes
due to invasive species and fire management. Obtaining
Acknowledgments
region-wide information is being facilitated through GIS
and remote sensing (e.g. Lucas et al., 2002; Lowry and
Finlayson, 2004; Lowry and Knox, 2006) although con- Most of the information reported in this paper was col-
tinuing on-ground data collection may be difficult and lected through funding supplied through the Environ-
expensive to achieve, especially given the changing na- mental Research Institute of the Supervising Scientist
ture of the wetlands within and between years. Further, in over a period extending back to 1978. Collecting this in-
recent decades the wetlands have changed in response to formation involved a lot of work and dedication from
deliberate management actions – those addressing inva- many people and partner organizations, including the lo-
sive species being better known. The outcomes of these cal inhabitants of the Park and surrounding land. This
management actions have been complex with the re- effort is recognized through the production of this review
moval of large numbers of feral buffalo having been – thanks to everyone involved – it was not easy and it was
widely supported, but with less understanding of the not always fun but it was done.
likely consequences. It may or may not have been possi-
ble at the time to project a scenario where the native grass
References
Hymenachne acutigluma spread to dominate large parts
of the floodplains. Allen, G. R., S. H. Midgley and M. Allen, 2002. Field Guide to the
The advent of new technologies provides an opportu- Freshwater Fishes of Australia, Western Australian Museum,
nity for more effective inventory, assessment and moni- Perth, Australia, 394 pp.
Andersen, A., 1996. Fire ecology and management. In: C. M. Fin-
toring of the wetland environments; this needs to be ex-
layson and I. von Oertzen (eds.), Landscape and Vegetation
plored and suitable investments made. It may also be Ecology of the Kakadu Region, Northern Australia, KluwerA-
worthwhile taking a step back and more thoroughly ana- cademic Publishers, Dordrecht, The Netherlands, pp. 179–195.
Bayliss, P. and K. M. Yeomans, 1990. Seasonal distribution and
lyzing the information bases available and complement-
abundance of magpie geese, Anseranas semipalmata Latham,
ing the mainly qualitative risk assessments with more
in the Northern Territory, and their relationship to habitat,
quantitative and landscape-wide analyses based on statis- 1983–86. Australian Wildlife Research 17: 15–38.
tical probabilities and model scenarios. At the same time Bayliss, B., K. Brennan, I. Eliot, C. M. Finlayson, R. Hall, T. House,
R. Pidgeon, D. Walden and P. Waterman, 1997. Vulnerability
if a clear picture of the biodiversity value of the Kakadu
396 C. M. Finlayson et al. Wetland biodiversity of the Kakadu region
Assessment of Predicted Climate Change and Sea Level Rise in re-evaluation of Palaemontes holthuisi Strenth, with a designa-
the Alligator Rivers Region, Northern Territory, Australia, Su- tion of a new genus, Calathaemon. Hydrobiologia 257: 73–94.
pervising Scientist Report 123, Canberra, Australia, 134 pp. Carson, L. J., P. W. Haines, A. Brakel, B. A. Pietsch and P. A. Fer-
Bamford, M. J., 1988a. Kakadu National Park: a preliminary survey enczi, 1999. 1:250,000 Geological Map Series Explanatory
of migratory waders October/November 1987. RAOU Report Notes Minlingimbi SD53-2. Department of Mines and Engery,
No. 41. Northern Territory Geological Survey and Australian Geologi-
Bamford, M. J., 1990. RAOU Survey of Migratory Waders in cal Survey Organisation, Darwin, Australia.
Kakadu National Park: Phase III.RAOU Report 70. Royal Aus- Chatto, R, 2000. Waterbird breeding colonies in the Top End of the
tralasian Ornithologists Union, Melbourne. Northern Territory. Technical Report No 69. Parks and Wildlife
Bedell, E., 2001. Diatom Assemblages and Water Quality in the Commission of the Northern Territory, Palmerston, Australia,
Freshwater Wetlands of Kakadu National Park. Unpublished 156 pp.
Honours Thesis, Curtin University of Technology, Perth, Aus- Chatto, R., 2003. The distribution and status of shorebirds around
tralia, 77 pp. the coast and coastal wetlands of the Northern Territory. Techni-
Bishop, K. A. and Forbes, M. A., 1991. The freshwater fishes of cal Report No 73. Parks and Wildlife Commission of the North-
northern Australia. In: C. D. Haynes, M. G. Ridpath and M. A. ern Territory, Palmerston, Australia, 206 pp.
J. Williams (eds.), Monsoonal Australia: Landscape, Ecology Cogger, H. G., 2000. Reptiles and Amphibians of Australia. Reed
and Man in the Northern Lowlands, A.A. Balkema, Rotterdam, Books Australia Pty Ltd, Melbourne, Australia, 788 pp.
pp. 79–107. Cook, G. D., S. A. Setterfield and J. Maddison, 1996. Scrub invasion
Bishop, K. A., 1980. Fish kills in relation to physical and chemical of a tropical wetland: Implications for weed management. Eco-
changes in Magela Creek (East Alligator River system, North- logical Applications 6: 531−537.
ern Territory) at the beginning of the tropical Wet season. Aus- Corbett, L. 1996. Aquatic Studies at Ranger Mine: A Whole Ecosys-
tralian Zoologist 20: 485–500. tem Approach. CSIRO Division of Wildlife and Ecology, Dar-
Bishop, K. A., S. A. Allen, D. A. Pollard and M. G. Cook, 1986. win, Australia, 152 pp.
Ecological Studies on the Freshwater Fishes of the Alligator Cowie, I. D., 1996. Weed ecology. In: C. M. Finlayson and I. von
Rivers Region, Northern Territory. Volume I: Outline of the Oertzen (eds.), Landscape and Vegetation Ecology of the Kakadu
Study, Summary, Conclusions and Recommendations. Research Region, Northern Australia, Kluwer Academic Publishers, Dor-
Report 4 (i), Supervising Scientist for the Alligator Rivers Re- drecht, The Netherlands, pp. 113–135.
gion, AGPS, Canberra, Australia, 63 pp. Cowie, I. D., M. D. Armstrong, J. C. Z. Woinarski, P. S. Brockle-
Bishop, K. A., S. A. Allen, D. A. Pollard and M. G. Cook, 1990. hurs, P. S. Short and C. R. Dunlop, 2000a. An overview of the
Ecological Studies on the Freshwater Fishes of the Alligator floodplains. In: I. D. Cowie, P. S. Short, P. Osterkamp and M.
Rivers Region, Northern Territory. Volume II: Synecology. Re- Madsen (eds.), Floodplain Flora: A Flora of the Coastal Flood-
search report 4 (ii), Supervising Scientist for the Alligator Riv- plains of the Northern Territory, Australia, ABRS, Canberra,
ers Region, AGPS, Canberra, Australia. Australia, pp. 1–33.
Bishop, K. A., S. A. Allen, D. A. Pollard and M. G. Cook, 2001. Eco- Cowie, I. D., P. S. Short, P. Osterkamp and M. Madsen, (eds.)
logical Studies on the Freshwater Fishes of the Alligator Rivers 2000b, Floodplain Flora: A Flora of the Coastal Floodplains of
Region, Northern Territory: Autecology. Supervising Scientist the Northern Territory, Australia, ABRS, Canberra, Australia,
Report 145, Supervising Scientist, Darwin, Australia, 570 pp. 382 pp.
Bishop, K. A., L. M. Baker and B. N. Noller, 1982. Naturally-occur- Cranston, P., 2000. August Thienemann’s influence on modern
ring ichthyocides and a report on Owenia vernicosa F. Muell. chironomidology – an Australian perspective. Verh. Interna-
(Family Meliaceae), from the Magela Creek system, Northern tionalis Vereinigung Limnologiae. 27: 278–283.
Territory. Search 13: 150–153. Cranston, P. S., P. D. Cooper, F. R. A. Hardwick, C. L. Humphrey
Bishop, K. A., R. W. J. Pidgeon and D. J. Walden, 1995. Studies of and P. L. Dostine, 1997. Tropical acid streams–the chironomid
fish movement dynamics in a tropical floodplain river: Prereq- (Diptera) response in northern Australia. Freshwater Biology
37: 473–483.
uisites for a procedure to monitor the impacts of mining, in the
Douglas, M. M., 1999. Tropical Savanna Streams: Their Seasonal
use of biota to assess water quality, Australian Journal of Ecol-
Dynamics and Response to Catchment Disturbance, Doctor of
ogy 20: 81–107.
Philosophy thesis, Monash University, Melbourne, Australia.
Braithwaite, R. W., Friend, G. R. and J. C. Wombey, 1991. Reptiles
Douglas, M. M. and R. A. O’Connor, 1999. Effects of the exotic
and amphibians. In: C. D. Haynes, M. G. Ridpath and M. A. J.
macrophyte, para grass (Urochloa mutica), on benthic and epi-
Williams (eds.), Monsoonal Australia: Landscape, Ecology and
phytic macroinvertebrates of a tropical floodplain. Freshwater
Man in the Northern Lowlands, A.A. Balkema, Rotterdam, The
Biology 48: 962–971.
Netherlands, pp. 109–124.
Douglas, M. M., S. E. Bunn and P. M. Davies, 2005. River and wet-
Brennan, K., 1996. An Annotated Checklist of Vascular Plants of the
land food webs in Australia’s wet-dry tropics: general principles
Alligator Rivers Region, Northern Territory. Darwin, Australia:
and implications for management. Marine and Freshwater Re-
Supervising Scientist Report 109, Supervising Scientist, Dar-
search 56: 329–342.
win, Australia, 136 pp.
East, T. J., 1996. Landform evolution. In: C. M. Finlayson and I. von
Broady, P. A., 1984. A Preliminary Investigation of the Cyanobacte-
Oertzen (eds.), Landscape and Vegetation Ecology of the
ria (“Blue-green Algae”) 0f the Magela Creek System and
Kakadu Region, Northern Australia, KluwerAcademic Publish-
Nearby Sites, Northern Territory, Australia. Unpublished report
ers, Dordrecht, The Netherlands, pp. 37–55.
to the Supervising Scientist for the Alligator Rivers Region,
Eliot, I., C. M. Finlayson and P. Waterman, 1999. Predicted climate
Northern Territory, Australia.
change, sea level rise and wetland management in the Austral-
Brown, T. E., A. W. Morley, N. T. Sanderson and R. D. Tait, 1981.
Report of a large fish kill resulting from natural acid water ian wet-dry tropics. Wetlands Ecology and Management 7:
conditions in Australia. Journal of Fish Biology 22: 335–350. 63–81.
Bruce, A. J., 1993. Kakaducaris glabra gen. nov., sp. nov., a new Faith, D. P., P. L. Dostine and C. L. Humphrey, 1995. Detection of
freshwater shrimp from the Kakadu National Park, Northern mining impacts on aquatic macroinvertebrate communities: re-
Territory, Australia, Crustacea: Decapoda: Palaemonidae with sults of a disturbance experiment and the design of a multivari-
the designation of a new subfamily Kakaducaridinae. Hydrobi- ate BACIP monitoring programme at Coronation Hill, Northern
ologia 268: 27–44. Territory. Australian Journal of Ecology 20: 167–180.
Bruce, A. J. and J. W. Short, 1993. Leptopalaemon gagadjui gen. nov., Faith, D. P., C. L. Humphrey and P. L. Dostine, 1991. Statistical
a new freshwater palaemonid shrimp from Arnhem land, and a power and BACI designs in biological monitoring: comparative
Overview Article
Aquat. Sci. Vol. 68, 2006 397
evaluation of measures of community dissimilarity based on Australia. Supervising Scientist Report 169, Supervising Scien-
benthic macroinvertebrate communities in Rockhole Mine tist, Darwin, Australia, 145 pp.
Creek, Northern Territory, Australia. Australian Journal of Ma- Garnett, S. and C. Minton, 1985. Notes on the movements and dis-
rine and Freshwater Research 42: 589–602. tribution of Little Curlew Numenius minutes in northern Aus-
Finlayson, C. M., 1988. Productivity and nutrient dynamics of sea- tralia. Australia Bird Watcher 11: 69–73.
sonally inundated flood plains in the Northern Territory. In: D. Gopal, B. and W. J. Junk, 2000. Biodiversity in wetlands: an introduc-
Wade-Marshall and P. Loveday (eds.), North Australia: Progress tion. In: B. Gopal, W. J. Junk and J. A. Davis (eds..), Biodiversity
and Prospects, vol. 2 Floodplain Research, ANU Press, Darwin, in wetlands: assessment, function and conservation, Vol. 1, Back-
Australia, pp. 58–83. huys Publishers b.V., Leiden, The Netherlands, pp. 1–10.
Finlayson, C. M., 1990. Plant ecology and management of an inter- Hardwick, R. A., P. D. Cooper, P. S. Cranston, C. L. Humphrey and
nationally important wetland in monsoonal Australia. In: J. A. P. L. Dostine, 1995. Spatial and temporal distribution patterns
Kusler and S. Day (eds.), Wetlands and River Corridor Manage- of drifting pupal exuviae of Chironomidae (Diptera) in streams
ment, The Association of Wetland Managers, New York, USA, of tropical northern Australia. Freshwater Biology 34:
pp. 90–98. 569−578.
Finlayson, C. M., 1991. Primary production and major nutrients in Hoatson, D., D. Blake, A. Wygralak, S. Needham, Allen, B., Miles,
three grass species on a tropical flood plain in northern Aus- G., P. Hauser and I. Oswald-Jacobs, 2000. Kakadu and Nitmiluk
tralia. Aquatic Botany 41: 263–280. National Parks, Northern Territory: A Guide to the Rocks,
Finlayson, C. M., 2005. Plant ecology of Australia’s tropical flood- Landforms, Plants, Animals, Aboriginal Culture and Human
plain wetlands: a review. Annals of Botany 96: 541–555. Impact. Australian Geological Survey Organisation, Canberra,
Finlayson, C. M. and R. A. van Dam, 2004. Developing local capac- Australia, 110 pp.
ity and ensuring relevance in risk assessment for tropical wet- Humphrey, C., 1999. Potential effects of the Jabiluka Mine upon the
lands. SETAC Globe Jan–Feb, 36–38. biodiversity of Kakadu National Park. Appendix 3 in Response
Finlayson, C. M. and I. von Oertzen, 1996a. The Kakadu region. In: to the ICSU Review of the Supervising Scientist’s Report to the
C. M. Finlayson and I. von Oertzen (eds.), Landscape and Veg- World Heritage Committee. Supervising Scientist, Environment
etation Ecology of the Kakadu Region, Northern Australia, Australia, Darwin, Australia.
KluwerAcademic Publishers, Dordrecht, The Netherlands, pp. Humphrey, C. L., A. W. Storey and L. Thurtell, 2000. AUSRIVAS:
1–15. operator sample processing errors and temporal variability-
Finlayson, C. M. and I. von Oertzen, (eds.) 1996b. Landscape and implications for model sensitivity. In: J.F. Wright, D. W. Sutcliffe
Vegetation Ecology of the Kakadu Region, Northern Australia, and M. T. Furse (eds.), Assessing the Biological Quality of Fresh
KluwerAcademic Publishers, Dordrecht, The Netherlands, 202 Waters: RIVPACS and Other Techniques, Freshwater Biological
pp. Association, Ambleside, U.K., pp. 143–163.
Finlayson, C. M. and C. D. Woodroffe, 1996. Wetland vegetation. Humphrey, C. L., L. Thurtell, R. W. J. Pidgeon, R. A. van Dam and
In: C. M. Finlayson and I. von Oertzen (eds.), Landscape and C. M. Finlayson, 1999. A model for assessing the health of
Vegetation Ecology of the Kakadu Region, Northern Australia, Kakadu’s streams. Australian Biologist 12: 33–42.
KluwerAcademic Publishers, Dordrecht, The Netherlands, pp. Humphrey, C. L., P. Martin and C. leGras, 2002. Use of the freshwa-
81–112. ter mussel, Velesunio angasi, in the monitoring and assessment
Finlayson, C. M., B. J. Bailey and I. D. Cowie, 1989. Macrophytic of mining impact in Top End streams. In: S. J. Markich and
Vegetation of the Magela Flood Plain, Northern Australia. Re- R.A. Jeffree (eds.), The Finniss River: A Natural Laboratory of
search Report No. 5, Office of the Supervising Scientist, Syd- Mining Impacts – Past, Present and Future, Australian Nuclear
ney, Australia, 38 pp. Science and Technology Organisation, report E/748, Sydney,
Finlayson, C. M., M. G. Bellio and J. B. Lowry, 2005. A conceptual Australia, pp. 43–48.
basis for the wise use of wetlands in northern Australia – linking Humphrey, C. L. and P. Dostine, 1994. Development of biological
information needs, integrated analyses, drivers of change and hu- monitoring programs to detect mining waste impacts upon
man well-being. Marine and Freshwater Research 56: 269–277. aquatic ecosystems of the Alligator Rivers Region, Northern
Finlayson, C. M., I. D. Cowie and B. J. Bailey, 1990. Characteristics Territory, Australia. Proceedings of Conservation and Manage-
of a seasonally flooded freshwater system in monsoonal Aus- ment of tropical inland waters: Problems, solutions and pros-
tralia. In: D. F. Whigham, R. E. Good and J. Kvet (eds.), Wet- pects, Conference. University of Hong Kong. Mitteilungen In-
land Ecology and Management: Case studies, Kluwer Aca- ternationalis Vereinigung Limnologiae 24: 293–314.
demic Publishers, Dordrecht, The Netherlands, pp. 141–162. Kay, W. R., M. J. Smith, A. M. Pinder, J. M. McRae, J. A. Davis and
Finlayson, C. M., I. D. Cowie, and B. J. Bailey, 1993b. Litterfall in S. A. Halse, 1999. Patterns of distribution of macroinvertebrate
a Melaleuca forest on a seasonally inundated flood plain in families in waters of north-western Australia. Freshwater
tropical northern Australia. Wetland Ecology and Management Ecology 41: 299–316.
2: 177–188. Larson, H., 2000. Threatened fish profiles – northern speartooth
Finlayson, C. M., R. N. Hall and B. L. Bayliss, 1997. Regional Re- shark Glyphis sp. C. Australian Society for Fish Biology
view of Wetland Management Issues: Wet-dry Tropics of Newsletter 30 (1): 30.
Northern Australia, LWRRDC Occasional Paper 03/97, Can- Ling, H. U. and Tyler, P. A., 1986. A Limnological Survey of the
berra, Australia, 35 pp. Alligator Rivers Region II. Freshwater Algae Exclusive of
Finlayson, C. M., K. Thompson, I. A. M. L von Oertzen and I. D. Diatoms. Research Report 3, Supervising Scientist for the
Cowie, 1993a. Vegetation of Five Magela Creek Billabongs, Alligator Rivers Region, AGPS, Canberra, Australia, 176 pp.
Alligator Rivers Region, Northern Australia. Technical Memo- Lonsdale, W. M., 1988. An analysis of litterfall in forests of the
randum 46, Supervising Scientist for the Alligator Rivers Re- world. Annals of Botany 61: 319–324.
gion, AGPS, Canberra, Australia, 26 pp. Lowry, J. B. and C. M. Finlayson, 2004. A Review of Spatial Data-
Finlayson, C. M., B. J. Bailey, W. J. Freeland and M. R. Fleming, sets for Wetland Inventory in Northern Australia. Supervising
1988. Wetlands of the Northern Territory. In: A. J. McComb and Scientist Report 178, Supervising Scientist, Darwin, Australia,
P. S. Lake (eds.), The Conservation of Australian Wetlands, Sur- 18 pp.
rey Beatty and Sons Pty Limited, Sydney, Australia, pp. 103– Lowry, J. B. and M. Knox, 2002. Geomorphic landscapes of the
126. Kakadu region, Northern Territory, Australia. Supervising Sci-
Gardner, S., C. M. Finlayson and R. W. J. Pidgeon, (eds.) 2002. entist poster, Supervising Scientist, Darwin, Australia.
Description and Literature Review of the Flora and Vertebrate Lowry, J. B. and M. Knox, 2006. A description of the geomorphic
Fauna of Magela Creek, Alligator Rivers Region, Northern landscapes of the Kakadu region, Northern Territory, Australia.
398 C. M. Finlayson et al. Wetland biodiversity of the Kakadu region
Supervising Scientist Report 186, Supervising Scientist, Dar- Pogonoski, J. J., D. A. Pollard and J. R. Paxton, 2002. Conservation
win, Australia. Overview and Action Plan for Australian Threatened and
Lucas, R. M., J. C. Ellison, A. Mitchell, B. Donnelly, M. Finlayson, Potentially Threatened Marine and Estuarine Fishes 2002.
A. K. Milne, 2002. Use of stereo aerial photography for quanti- Environment Australia, Canberra, Australia.
fying changes in the extent and height of mangroves in tropical Pollard, D. A., 1974. The freshwater fishes of the Alligator Rivers Re-
Australia. Wetlands Ecology and Management 10: 161–175. gion ‘Uranium Province’ area (Top End Northern Territory) with
Madsen, T. and Shine, R., 1996. Seasonal migration of predators particular reference to the Magela Creek catchment (East Alligator
and prey – a study of pythons and rats in tropical Australia. River System). Report 1. In: P. J. F. Newton (ed.), Alligator Rivers
Ecology 77: 149–156. Area Fact Finding Study, Report E305, Australian Atomic Energy
Marchant, R., 1982a. The macroinvertebrates of Magela Creek, North- Commission, AAEC, Sydney, Australia, pp. 1–71.
ern Territory. Research Report No. 1, Supervising Scientist for the Press, T., J. Brock and A. Anderson, 1995. Fauna. In: T. Press, D.
Alligator Rivers Region, AGPS, Canberra, Australia, 46 pp. Lea., A. Webb and A. Graham (eds.), Kakadu: Natural and Cul-
Marchant, R., 1982b. Life spans of two species of tropical mayfly tural Heritage and Management, Australian Nature Conserva-
nymph (Ephemeroptera) from Magela Creek, Northern Territo- tion Agency and North Australia Research Unit, The Australian
ry. Australian Journal of Marine and Freshwater Research 33: National University, Darwin, Australia, pp. 197–214.
173–179. Pusey, B. J. and M. J. Kennard, 1996. Species richness and geo-
Marchant, R., 1982c. Seasonal variation in the macroinvertebrate graphical variation in assemblage structure of the freshwater
fauna of billabongs along Magela Creek, Northern Territory. fish fauna of the Wet Tropics Region of northern Queensland.
Australian Journal of Marine and Freshwater Research 33: Marine and Freshwater Research 47: 563–573.
329–342. Rea, N. and M. J. Storrs, 1999. Weed invasions in wetlands of Aus-
McBride, T. P., 1983. Diatom Communities of the Mudginberri Cor- tralia’s Top End: reasons and solutions. Wetland Ecology and
ridor, Northern Territory, Australia. Doctor of Philosopy thesis, Management 7: 47–62.
Macquarie University, Sydney, Australia. Roberts, J. D. and G. F. Watson, 1993. Biogeography and Phylogeny
Midgley, S. H., 1973. Freshwater fish inventory and associated of the Anura. In: C. J. Glasby, G. J. B. Ross and P. L. Beesley
habitat information. Alligator Rivers Region Environmental (eds.), Fauna of Australia, Vol. 2A, Australian Government
Fact Finding Study. Department of the Northern Territory, Can- Publishing Service, Canberra, Australia.
berra, Australia, 30 pp. Russell Smith, J., D. Lucas, M. Gapindi, B. Gunbunuka, N. Kapiri-
Miles, G., 1998. Wildlife of Kakadu and the Top End of the Northern gi, G. Namingam, K. Lucas, P. Giuliani and G. Chaloupka,
Territory, Australia, Barker Souvenirs, Alice Springs, Australia. 1997. Aboriginal resource utilisation and fire management prac-
Miles, G., 2000. Kakadu Wildlife: Kakadu: A World Heritage Na- tices in western Arnhem land, monsoonal northern Australia:
tional Park, Northern Territory, Australia. Barker Souvenirs, notes for prehistory, lessons for the future. Human Ecology 25:
Alice Springs, Australia. 159–195.
Morris, I., 1996. Kakadu National Park, Australia. Steve Parish Sadlier, R. A., 1990. The Terrestrial and Semiaquatic Reptiles (Lac-
Publishing, Bowen Hills, Australia, 225 pp. ertilia, Serpentes) of the Magela Creek Region, Northern Terri-
Morton, S. R., K. G. Brennan and P. Dostine, 1984. Diversity, abun- tory. Technical Memorandum 32, Supervising Scientist for the
dance and importance of avian food links in food chains. In: Alligator Rivers Region, AGPS, Canberra, Australia, 86 pp.
Alligator Rivers Region Research Report 1983-1984. Austral- Sanderson, N. T., D. V. Koontz and A. W. Morley, 1983. The ecol-
ian Government Publishing Service, Canberra. ogy of the vegetation of the Magela Creek flood plain: upper
Morton, S. R., K. G. Brennan and M. D. Armstrong, 1990a. Distri- section from Oenpelli road crossing to Nankeen Billabong.
bution and abundance of Magpie Geese, Anseranas semipal- Unpublished report in Scientific Workshop, Environmental
mata, in the Alligator Rivers Region, Northern Territory. Aus- Protection in the Alligator Rivers Region, Jabiru, Australia, pp.
tralian Journal of Ecology 15: 307-320. 33.1–33.9.
Morton, S. R., K. G. Brennan and M. D. Armstrong, 1990b. Distri- Schulz, M., 1989. The importance of wetlands in Kakadu National
bution and abundance of ducks in the Alligator Rivers Region, Park to selected waterbirds. A Report to the Asutralian National
Northern Territory. Australian Wildlife Research 17: 573-590. Parks and Wildlife Service, May 1989.
Morton, S. R. and K. G. Brennan, 1991. Birds. In: C. D. Haynes, M. Shiel, R. J. and W. D. Williams, 1990. Species richness in tropical
G. Ridpath and M. A. J. Williams (eds), Monsoonal Australia: freshwaters of Australia. Hydrobiologia 202: 175–183.
Landscape, ecology and man in the northern lowlands, AA Shine, R., 1986a. Diets and abundances of aquatic and semi-aquatic
Balkema, Rotterdam, Netherlands, pp. 133–149. reptiles in Alligator Rivers Region. Technical Memorandum 16,
Noller, B. N., 1983. Physico-chemical aspects of natural fish kills on Supervising Scientist for the Alligator Rivers Region, AGPS,
the Magela Creek system. In: Proceedings of the Scientific Canberra, Australia, 57 pp.
Workshop on Environment Protection in the Alligator Rivers Shine, R., 1986b. Predation upon filesnakes (Acrochordus arafurae)
Region, Paper no. 61. Supervising Scientist for the Alligator by aboriginal hunters: selectivity with respect to size, sex and
Rivers Region, Sydney, Australia. reproductive condition. Copeia 1986: 238–239.
Northern Territory Frogs Database, 2003. CD-Rom, Frogwatch, Skeat, A. J., T. J. East and L. K. Corbett, 1996. Impact of feral water
Noonamah, Australia. buffalo. In: C. M. Finlayson and I. von Oertzen (eds.), Landscape
Outridge, P. M., 1988. Seasonal and spatial variations in benthic and Vegetation Ecology of the Kakadu Region, Northern Aus-
macroinvertebrate communities of Magela Creek, Northern tralia, Kluwer Academic Publishers, Dordrecht, The Netherlands,
Territory. Australian Journal of Marine and Freshwater Re- pp. 155–177.
search 39: 211–223. Smith, M. and P. Cranston, 1995. Recovery of an acid mine-waste
Ovington, D., 1986. Kakadu: A World Heritage of Unsurpassed impacted tropical stream: The chironimid story. In: Chirono-
Beauty. Canberra, Australia: Australian Government Publishing mids, from Genes to Ecosystems, 12th International Conference
Service, 116 pp. on Chironomidae, 161–173.
Paltridge R. M., P. L. Dostine, C. L. Humphrey and A. J. Boulton, Storrs, M. J. and C. M. Finlayson, 1997. Overview of the Conserva-
1997. Macroinvertebrate recolonization after re-wetting of a tion Status of Wetlands of the Northern Territory. Supervising
tropical seasonally-flowing stream (Magela Creek, Northern Scientist Report 116, Supervising Scientist, Canberra, Austral-
Territory, Australia). Marine and Freshwater Research 48: ia, 90 pp.
633−645. Taylor, W. R., 1964. Fisheries of Arnhem Land. In: Records of the
Pidgeon, B., 2001. Natural causes of fish kills in Top End waterways. American–Australian Scientific Expedition to Arnhem Land,
eriss note, Supervising Scientist, December, Australia, 4 pp. Vol. IV: Zoology 45–307.
Overview Article
Aquat. Sci. Vol. 68, 2006 399
Taylor, J. A. and D. Tulloch, 1985. Rainfall in the wet-dry tropics: Van der Valk, A. G., 1981. Succession in wetlands: a Gleasonian
extreme events at Darwin and similarities between years during approach. Ecology 62: 688–696.
the period 1870–1983. Australian Journal of Ecology 10: Vanderlely, R., 1995. An Ecological Survey of the Sea Turtles of
281−295. West Alligator Head and Field Island, Kakadu National Park.
Thomas, D. P., 1983. A Limnological Survey of the Alligator Rivers Master of Science thesis, Northern Territory University, Dar-
Region I. Diatoms (bacillariophyceae) of the Region. Research win, Australia.
Report 3, Supervising Scientist for the Alligator Rivers Region, Vinson, M. R. and C. P. Hawkins, 2003. Broad-scale geographical
AGPS, Canberra, Australia, 160 pp. patterns in local stream insect genera richness. Ecography 26:
Tyler, M. J. and G. A. Crook, 1987. Frogs of the Magela Creek Sys- 751–767.
tem. Technical memorandum 19, Supervising Scientist for the Walden, D., R. van Dam R, M. Finlayson, M. Storrs, J. Lowry and
Alligator Rivers Region, AGPS, Canberra, Australia, 46 pp. D. Kriticos, 2004. A Risk Assessment of the Tropical Wetland
Tyler, M. J., G. A. Crook and M. Davies, 1983. Reproductive biol- Weed Mimosa pigra in Northern Australia. Supervising Scien-
ogy of the frogs of the Magela Creek system, Northern Territo- tist Report 177, Supervising Scientist, Darwin, Australia, 66
ry. Records of the South Australian Museum 18: 415–460. pp.
Unmack, P., 2001. Biogeography of Australian freshwater fishes. Welcomme, R. L., 1979. Fisheries ecology of floodplain rivers.
Journal of Biogeography 28: 1053–1089. Longman, London, U.K.
US EPA (U.S. Environmental Protection Agency), 1998. Guidelines Williams, W. D., 1980. Australian Freshwater Life. Macmillan,
for Ecological Risk Assessment. EPA/630/R-95/002F. Risk As- Melbourne, Australia.
sessment Forum, Washington, DC, USA. Wilson, B. A., J. Russell-Smith and R. Williams, 1996. Terrsetrial
van Dam, R. A., D. J. Walden and G. W. Begg, 2002a. A Preliminary vegetation. In: C. M. Finlayson and I. von Oertzen (eds.), Land-
Risk Assessment of Cane Toads in Kakadu National Park. Su- scape and Vegetation Ecology of the Kakadu Region, Northern
pervising Scientist Report 164, Supervising Scientist, Darwin, Australia, KluwerAcademic Publishers, Dordrecht, The Nether-
Australia, 92 pp. lands, pp. 57–79.
van Dam, R. A., C. L. Humphrey and P. Martin, 2002b. Mining in Woinarski, J. C. Z. and N. Gambold, 1992. Gradient analysis of a
the Alligator Rivers Region, northern Australia: Assessing po- tropical herpetofauna: distribution patterns of terrestrial reptiles
tential and actual impacts on ecosystem and human health. and amphibians in Stage III of Kakadu National Park, Australia.
Toxicology 181/182: 505–515. Wildlife Research, 19: 105–127.
van Dam, R. A., C. Camilleri, C. Turley and S. J. Markich, 2004. Woodroffe, C. D., J. Chappell, B. G. Thom and E. Wallensky, 1989.
Ecological risk assessment of tebuthiuron following application Depositional model of a macrotidal estuary and flood plain,
on Australian tropical wetlands. Environmental Toxicology and South Alligator River, northern Australia. Sediment. 36: 737–
Chemistry 10: 1069–1097. 756.
1015-1621/06/030374-26
Aquatic Sciences
DOI 10.1007/s00027-006-0852-3
© Eawag, Dübendorf, 2006
Overview Article
Biodiversity of the wetlands of the Kakadu Region,
northern Australia
C. Max Finlayson1,2,*, John Lowry1, Maria Grazia Bellio1,2, Suthidha Nou1,3, Robert Pidgeon1, Dave Walden1,
Chris Humphrey1 and Gary Fox1
1
Environmental Research Institute of the Supervising Scientist, G.P.O Box 461, Darwin, NT 0801, Australia
2
Present address: International Water Management Institute, P.O. Box 2075, Colombo, Sri Lanka
3
Present address: Parks Australia, Post Office, Jabiru, NT 0886, Australia
Received: 15 October 2005; revised manuscript accepted: 12 December 2005
Abstract. The biodiversity values of the wetlands in the ing high in some habitats. Most fauna analyses have fo-
Kakadu Region of northern Australia have been recog- cussed on vertebrates with a large amount of information
nised as being of national and international significance, on waterbirds and fish in particular. However, despite
as demonstrated through their listing by the Ramsar Con- extensive effort over the past two decades much is still
vention on Wetlands. Analyses of the wetland biodiver- unknown about the biota. While the invertebrate fauna in
sity have resulted in the production of species list for the streams has received some attention, a large taxo-
many taxa, and some population and community-level nomic classification effort is required. The functional in-
analyses of biomass and abundance, and the mapping of ter-relationships between habitats and species have
habitats at multiple scales. Wetland habitats include in- largely not been assessed. Further, the ecology of many
ter-tidal mud-flats, mangroves, hyper-saline flats, fresh- species is only cursorily known. At the same time there
water flood plains and streams. The tidal influence on the has been increased attention to pressures on the wetlands,
saline wetlands is pronounced, as is the influence of the such as weeds and feral animals, water pollution, and the
annual wet-dry cycle of the monsoonal climate on the potential impact of climate change and salinisation of
flood plains and streams. The vegetation is diverse and freshwater habitats. Importantly, given the social context
highly dynamic with rapid turnover of organic material of the region, increased attention is being directed to-
and nutrients. The fauna is abundant with endemism be- wards traditional use and management of the wetlands.
Key words. Wetlands; wetland fauna; wetland vegetation.
Introduction groups in the wetlands within Kakadu National Park and
the broader Alligator Rivers Region in northern Australia.
In this paper we summarize data on species numbers and The Region is located to the east of the city of Darwin in
describe the general ecology of major plant and animal the Northern Territory of Australia (Fig. 1) and is part of
a larger bio-geographical region known loosely as the
wet-dry tropics that extends across northern Australia
(Finlayson et al., 1997; Finlayson, 2005). Isolated and
* Corresponding author phone: +94-11 2787404;
lowly populated, the wetlands are of immeasurable value
fax: +94-11 2786854; e-mail: m.finlayson@cgiar.org
for the world‘s natural and cultural heritage and have at-
Published Online First: September 29, 2006
Overview Article
Aquat. Sci. Vol. 68, 2006 375
tracted great interest, particularly over the past 20–30 specific. As these species are integral parts of the wetlands
years, for their conservation values (Finlayson and von and may contribute considerably to wetland functions and
Oertzen, 1996a). The Alligator Rivers Region comprises processes (e.g. energy and nutrient cycles) they are in-
the ~20,000 km2 Kakadu National Park and a further cluded in this description, although often there is scant
~8,000 km2 of land within the western portion of Arnhem information about their biology or ecological role/s.
Land (a large tract of Aboriginal owned land to the east of The concept of a wetland species is based around that
Kakadu National Park) that comprises part of the catch- used for wetland plants by Finlayson et al. (1989), name-
ment of the East Alligator River. Lowry and Knox (2002, ly, a plant or animal that is adapted to withstand periods
2006) further extended the areal region to include parts of of waterlogging or flooding for a period of time. Under
the adjacent catchments. Given the recognition of the this general definition species found in the seasonally
name “Kakadu” and its areal dominance within the Re- inundated areas (woodlands and grasslands) that fringe
gion the general name “Kakadu Region” is adopted here the billabongs and flood plains are included along with
and taken to include the wider region, as done by Finlay- the wholly aquatic species, as well as species that use the
son and von Oertzen (1996b) when reporting on the wetland less frequently but are fully dependent on it. This
vegetation of the Region. is similar to the approach proposed by Gopal and Junk
Wetlands in the Kakadu Region include mangroves, (2000) who also provided further subdivisions based on
freshwater flood plains, salt flats and small permanent the extent of the reliance of the species on the wetland.
lakes (Finlayson and Woodroffe, 1996). The wetlands of This more detailed categorization has not been adopted
Kakadu National Park have been designated as interna- given a lack of information on the taxonomy and ecology
tionally important under criteria established by the Ram- of many taxa.
sar Wetlands Convention, due in part, to their importance As there is little specific information from the riparian
in a biogeographical context, the outstanding diversity of fringes of the streams, except from generic biological
their plant communities, and their role in conserving the surveys, this habitat is not emphasised. The monsoonal
large numbers of waterfowl that congregate on the flood forest habitats that fringe or abut the streams and flood
plains during the Dry season (Finlayson et al., 1990; plains are not included; Wilson et al. (1996) provide a
Finlayson and von Oertzen, 1996a). World Heritage listing detailed description of the vegetation of the Region.
of Kakadu National Park also refers to the natural heritage
value of the tidal flats and flood plains, and the floristic
Ecological characterization of the
diversity and endemism of the wetland vegetation.
Kakadu Region
Descriptions of the diversity and ecology of wetland
species in the Region reflect large differences in the
amount of available information for the taxonomic Landforms
entities. The information is uneven and incomplete. The The geological evolution of the Region was summarised
descriptive information presented here is followed by an by East (1996) who described three major landforms.
outline of the pressures on the wetlands, as ascertained in Lowry and Knox (2002, 2006) provided more detailed
recent years using structured and integrated approaches mapping of six geomorphic landscape classes across the
and risk assessment in particular. Region; the distribution of these classes is shown in Fig-
ure 3, whilst the area of each is listed in Table 1.
1. Arnhem Land Plateau – in the eastern and southern
Definition and classification of wetland species parts of the Region; an average height of 300 m above
the adjacent plains; composed predominantly of
Wetlands in the Kakadu Region comprise a number of quartzose sandstone known as the Mamadawerre For-
hydrological patterns – tidally influenced as well as annu- mation (previously known as the Kombolgie Forma-
ally flooded. Whilst the tidal influence extends some 70– tion – Carson et al., 1999); highly dissected with a
90 km along the major rivers, most attention has been di- high proportion of bare rock; bounded in the north and
rected towards the freshwater flooding of the rivers and west by a steep escarpment.
adjacent floodplain wetlands. The freshwater wetlands are 2. Dissected Foothills of Igneous Origin – a variety of
subject to a reasonably predictable mono-modal flood landforms, including rocky hills, boulder-covered
pulse (Finlayson et al., 1990, 1989; Finlayson and strike ridges, stony hillocks and occasional granite
Woodroffe, 1996) that has been related to the rainfall data pillars; formed through igneous activity and subjected
and the more complex seasonal climatic pattern recog- to exposure and/or erosional processes over the last
nized by the indigenous inhabitants of the Region (Fig. 2; 2.5 billion years.
Finlayson, 2005). During the dry months the wetlands 3. Mamadawerre (Koolpinyah) Surface – gently undulat-
may dry completely and can be colonized by terrestrial ing plains and rises of the lowland landscape; com-
plant and animal species that may or may not be wetland posed of sedimentary plains, and found predominantly
376 C. M. Finlayson et al. Wetland biodiversity of the Kakadu region
Figure 1. Location of Kakadu National Park and the Alligator Rivers Region in northern Australia’s wet-dry tropics (from Finlayson et al.,
1997 and Lowry and Finlayson, 2004).
Overview Article
Aquat. Sci. Vol. 68, 2006 377
Table 1. Geomorphic land classes in the Kakadu Region (from
Lowry and Knox, 2002, 2006).
Area km2
Geomorphic landscape class
Alluvial floodplains 2960
Arnhem Land plateau and escarpment 9918
Coastal floodplains 3923
Deeply weather, eroded Mamadawerre Surface 13006
Dissected foothills of igneous origin 3435
Mamadawerre Surface – sedimentary plains 7045
Table 2. Area (km2) of catchments within the Kakadu Region (from
Lowry and Knox, 2006).
Catchment Within Kakadu
Within the
National Park
Kakadu Region
Mary River 8142 1269
Wildman River 3336 1708
West Alligator River 1444 1444
South Alligator River 11945 10900
East Alligator River 10340 2205
Katherine River* 5168 1601
* most of Katherine river catchment is outside of the Kakadu Region
tion of alluvial sediments by the rivers has resulted in
the development of flood plains along the rivers.
6. Coastal floodplains – bound the coastal margins;
poorly drained plains of very low relief; resulted from
Figure 2. Generalised representation of: a) the climate of the Aus-
sea level rise which reached its present level approxi-
tralian wet-dry tropics; b) hydrological change on the floodplains
mately 6000 years ago and drowned the coastal river
(variability is represented by dashed lines); and c) an Aboriginal
calendar. A 14 month period is used to illustrate the extension of valleys; infilling of the valleys led to the development
seasons across the calendar year. The hydrological information was
of broad coastal flood plains; large meandering rivers
adapted from Sanderson et al. (1983) by Finlayson et al. (1990) and
form the upper reaches of the coastal flood plain and
the Aboriginal calendar from Ovington (1986) and Morris (1996).
grade into saline mudflats near the coast.
The Aboriginal seasons are described in the following manner: Gu-
numeleng pre-monsoon season; Gudjeweg monsoon season; Bang- The Region contains the entire catchments of the Wild-
gerreng harvest time; Yegge cool weather time; Wurreng early dry
man, West Alligator, South Alligator and East Alligator
season; and Gurrung hot dry season.
Rivers, as well as part of the Mary and Katherine catch-
ments that are within Kakadu National Park. Lowry and
in the south-west of the Kakadu Region; formed by the Knox (2002, 2006) also include the adjacent parts of the
gradual erosion and retreat of the Arnhem Land pla- Katherine and Mary River catchments (Table 2, Fig. 3).
teau exposing the underlying, more resistant substrata More accurate mapping has been prepared recently, con-
and the development of undulating rises and plains. firming that the original map excluded part of the catch-
4. Deeply Weathered Mamadawerre (Koolpinyah) Sur- ment of the East Alligator River and corrected the mis-
face – a deeply weathered surface composed of plains, conception that Kakadu National Park contained the
broad valleys, very low-gradient slopes, and isolated entire catchment of the South Alligator River (Lowry and
hills and ridges of resistant rock; located mainly be- Knox, 2006).
tween the Plateau and the coastal flood plains, al-
though inliers occur in the Plateau.
5. Alluvial floodplains – along the middle upper reaches Climate and hydrology
of the major rivers; in the upper reaches of the rivers The climate is generally taken to comprise two broad
the floodplains are typically confined within narrow seasons – the Wet season, which commences late in the
river valleys. The plains have a gentle slope and year (November-December) and lasts for 3−4 months,
gradually merge into the coastal flood plains; deposi- and the Dry season (Taylor and Tulloch, 1985), as shown
378 C. M. Finlayson et al. Wetland biodiversity of the Kakadu region
Figure 3. Geomorphic land classes in the Kakadu Region (from Lowry and Knox, 2002; 2006).
schematically in Figure 2. The most significant features logical data, and a generalized hydrological cycle that
of the Wet season are thunderstorms, tropical cyclones was developed separately from field observations at
and rain depressions. As cyclones move inland they form Magela Creek and flood plain in Kakadu National Park
rain depressions and are an important source of rain. (developed by Sanderson et al., 1983, and modified by
Rainfall is also associated with monsoonal troughs, with Finlayson et al., 1990; Fig. 2). At the start of the Wet
2–3 occurring each year, that usually produce widespread season intermittent rain storms saturate the soils of the
cloud and rainfall, regional convection that provides lo- lowland Mamadawerre (Koolpinyah) Surface and as
calized showers and easterly disturbances that, in some more consistent rain occurs water collects in the creeks
years, extend the rainy season. The Dry season is charac- and thence in the large tidal rivers. Heavy rainfall over
terized by south-east trade winds and very little rain fall. the plateau can send water cascading into the streams on
In general, temperatures are warm to hot throughout the lowland surface. Once the creeks and rivers are full
the year with a mean monthly maxima of 37 °C in No- the freshwater spills across the flood plains and can cover
vember and 31.5 °C in July. In the Wet season warm them to a depth of several metres. The base flow in the
creeks across the lowlands is less than 5 m3 s–1 with peak
temperatures are accompanied by relative humidity of
about 80 %. Cloud cover is greatest during the Wet sea- flows late in the Wet season, reaching and exceeding on
occasions, 1000 m3 s–1. The large range in both total an-
son.
Local Aboriginal people have a refined perception of nual and mean monthly discharge along the Magela
the climate and recognize six seasons based on the rela- Creek is representative of the water regimes experienced
tionship between changes in the weather and the availa- in the lowlands across the Region (Fig. 4); there is little
bility of food items (Ovington, 1986; Morris, 1996). The information on the flow patterns in the plateau.
calendar they recognize is usually presented in a circular Flooding occurs once the catchment is saturated;
manner, but when presented in a linear manner and com- heavy falls of rain later in the season generate more wide
pared with the meteorological data the patterns outlined spread flooding than equivalent rainfall earlier in the sea-
by the Aboriginal calendar are readily identifiable from son. The water regimes of the creeks and rivers reflect the
the meteorological data (Finlayson, 2005; Fig. 2). climatic variability of the region (as shown in Fig. 2).
Finlayson (2005) also depicted the relationship be- While flow may be maintained in the upper reaches of the
tween the knowledge base developed over millennia by watercourses by springs or seeps, freshwater flow in the
Aboriginal people, the more recently obtained meteoro- mid-lower sections of creeks and rivers generally ceases
Overview Article
Aquat. Sci. Vol. 68, 2006 379
Figure 4. Total annual and mean monthly flow in Magela creek (data from Northern Territory Department of Natural Resources, Environ-
within a few months of the end of the rains. The creeks and from a 1:1,000,000 map compiled as part of a vegeta-
flood plains then dry out except for a few permanent tion survey of the Northern Territory to 1:250,000 and
swamps and lagoons, known locally as billabongs (Finlay- 1:100,000 surveys of specific communities. Vegetation
son et al., 1990). in the region can be broadly categorised into 8 broad
The spring tidal range in van Diemen Gulf is 5–6 m classes, which reflect the influence of the soils, geology
and estuarine water can extend more than 100 km up- and landform of the area. These include mangrove for-
stream and generally remains within the stream channel ests, sedges and herbaceous plants, paperbark forests,
(Woodroffe et al., 1989). The groundwater level in the grassland and savanna, tall open forest and woodland,
surrounding landscape is recharged by the Wet season monsoon rainforest and woodland, spinifex grassland,
rains, but can fall 2–4 m during the Dry season. and heath shrubland.
Wilson et al. (1996) provide an overview of the veg-
etation and land cover. Around 55 % of the terrestrial
Vegetation/Landcover vegetation is described as tropical tall grass savanna,
The Region contains in excess of 1800 species of higher composed of eucalypt-dominated open forest and wood-
plants (macro-algae and vascular plants; Brennan, 1996; land, with a 1–2 m grassy understorey. A further 30 % of
Wilson et al., 1996). The floral diversity reflects the the Region is covered by heaths, and open woodlands
range of landscapes and habitats (Hoatson et al., 2000). with sparse grass understorey. Closed canopy monsoon
Many vegetation surveys have been conducted, ranging rainforests are restricted to lowland springs, rock out-
380 C. M. Finlayson et al. Wetland biodiversity of the Kakadu region
crops, flood plain and beach levees, and sandstone. Vege- including Rhizophora slowly gaining dominance over
tation communities on the freshwater flood plains exhibit Bruguiera/Ceriops species, a brief period of dominance
marked variations in their composition, due to the ex- by Sonneratia sp. and subsequent succession from Son-
tremes of the Wet and Dry seasons (Finlayson et al., neratia through Rhizoporaceous forest to stands of Avi-
1989, 1990; Finlayson and Woodroffe, 1996; Finlayson, cennia. The sinuous/cuspate phase began about 5300
2005). In the Wet season, flowering plants are profuse years ago and represents a period during which the South
and widely distributed across the floodplain with high Alligator River re-established a distinct channel mean-
dry weight production. However, during the Dry season dering across the estuarine plains, changing during the
the vegetation is sparse and far less diverse. Mangrove last 2500 years from an entirely sinuous form to a pre-
communities are restricted to a narrow coastal strip and dominantly cuspate form in the central part of the tidal
to the reaches of the major rivers subject to tidal influ- section. Associated with the transition from the ‘big
ence (Finlayson and Woodroffe, 1996). swamp’ to the ‘sinuous’ phase there was widespread re-
Fire and invasive plant and animal species have a placement of mangrove species with the grasses and
significant impact on the extent and distribution of plant sedges characteristic of the present day freshwater wet-
species and the land cover (Anderson, 1996; Cowie, lands. The timing of the transition from mangroves to
1996). Saline intrusion, caused by either climate change, freshwater wetlands probably varied throughout the Re-
or damage to the natural levees separating freshwater and gion.
saline wetland communities, by feral animals (water buf-
falo) has the potential to cause significant change to the
Biodiversity of different plant and
vegetation communities of the floodplains (Bayliss et al.,
animal groups
1997; Eliot et al., 1999). Significant change in the flood-
plain vegetation has already occurred with the removal of
feral buffalo (Skeat et al., 1996). Wetlands in the Kakadu Region comprise tidally influ-
enced mangroves and salt flats, seasonally inundated
flood plains and billabongs, and small permanent lakes
Wetland evolution (Finlayson et al., 1990). The latter are small and not com-
Stratigraphic studies, particularly those on the wetlands mon. The flood plains (and associated lagoons) cover
of the South Alligator River have demonstrated that sub- about 195,000 ha and extend along the major rivers (West
stantial environmental change has occurred over the last Alligator, South Alligator and East Alligator) and creeks
20,000 years, attributed largely to changes in sea-level (Magela, Nourlangie, Jim Jim and Cooper) throughout
(see overview in Finlayson and Woodroffe, 1996). About the Region. Published studies of the species-level diver-
18,000 years ago, the sea levels across northern Australia sity of the wetlands have centred on the flood plain of
were some 150 m lower than today and the coastline was Magela Creek, a tributary of the East Alligator River.
several hundred kilometres north of the present location. It is noted that whilst the role of the seasonal wetting
The sea has since risen and around 8000 years ago tidal and drying has been highlighted the effect of stochastic
water flooded into the valleys, as illustrated by data from events on the composition and distribution of species
the South Alligator River valley which is now filled to a within the wetlands has not been considered in detail -
depth of 14-16 m with fine muds and alluvium. ecological investigations have principally considered
The stratigraphy of the deltaic-estuarine plain of the species presence and distribution across the wetlands
South Alligator River, and probably the other river sys- with limited analysis and modelling of ecological intere-
tems in the Region, indicates that the wetlands developed lationships and processes. While Finlayson (1988) pro-
in three major phases: i) transgressive: ii) big swamp; vide an initial summary of the high standing biomass
and iii) sinuous/cuspate. Finlayson and Woodroffe (1996) present on the flood plains the inter-relationship between
summarised the many studies which generally support the chemical, physical and biological components of the
the notion of wide-scale change in the vegetation of the wetlands are not elaborated, except for pointing out that
plains and rivers. The transgressive phase (8000-6800 for some vertebrates much of their food supply comes
years ago) was characterised by marine incursion into the from the adjacent terrestrial environment.
prior valley and terrestrial ecosystems were displaced by
mangrove forests as the sea level rose. By 6800 years ago
a mangrove forest dominated by Rhizophoraceous spe- Phytoplankton
cies had become established. The onset of the ‘big Taxonomic studies were undertaken mainly in the early
swamp’ phase (6800-5300 years ago) followed stabilisa- 1980s. Thomas (1983) provided a listing of 160 diatom
tion of the sea level around its present level. Mangrove taxa from 32 genera and considered the flora rich and
forests established over most of the present-day flood allied to that of south-east Asia with annual changes in
plains for around 6000 years with successional changes water quality and the variable environment contributing to
Overview Article
Aquat. Sci. Vol. 68, 2006 381
the diversity. Bedell (2001) confirmed that the distribution al. (1989); Finlayson and Woodroffe (1996) provided a
and relative abundance of diatom species in artificial and list of mangrove species; Brennan (1996) a list of all
natural wetlands was more strongly related to variations in higher plants; and Cowie et al. (2000a) a flora of the
electrical conductance (EC) and pH than temperature or coastal flood plains of the Northern Territory. The follow-
dissolved oxygen. Out of the 125 species observed, 7 were ing information is taken mainly from Cowie et al.
identified as indicators of pH and/or EC, as shown below: (2000b). The wetland flora consists mostly of widely
– acidic conditions: Aulacoseira granulata, Brachysira distributed taxa with only 3 genera, Hygrochloa, Omeg-
exilis var. lanceolata, Brachysira serians var. brachy- andra and Maidenia, being endemic to Australia. Some
sira (A), and Navicula radiosa; 24 % (57 species) of the species that occurred primarily
– alkaline conditions: Achnanthidium minutissimum and on floodplains were endemic to Australia; 57 % (132 spe-
Nitzschia palea; cies) were widespread in the Old World; 13 % (30) were
– EC < 100 µS/cm: Aulacoseira granulata and Brachy- pantropical; and 6 % (19 species) restricted to Australia.
sira serians var. brachysira (A); and Of the Australian endemics, only 4 species that were not
– EC > 100 µS/cm: Achnanthidium minutissimum, restricted to the floodplains, Bambusa arnhemica, Hy-
Nitzschia gracilis (A) and Nitzschia palea. grochloa aquatica, Nymphoides spongiosa and Nym-
phoides subacuta, were restricted to the Northern Terri-
A high level of interconnectedness existed between pH, tory. The non-endemic nature of the flora was considered
EC, wetland type and species diversity. Artificial sites to reflect the many opportunities for exchange of species
generally differed from more natural sites with a higher with other landmasses during recent geological times.
pH and a higher EC than the creek and billabong sites. Within the Kakadu Region, Finlayson et al. (1989)
Species diversity was lower in artificial sites than in natu- built on information provided by other researchers and
ral sites. Further, water quality was a key determinant prepared a generalized vegetation map for the Magela
governing the diatom assemblages present in different flood plain based on observations from 5–6 years and, in
wetland types. McBride (1983) also determined that the particular, the Wet seasons, when many of the aquatic
biomass of diatoms attached to macrophytes was equiva- plants reach their peak biomass and are easier to differen-
lent to about 30 % of the macrophyte biomass of 16,500 t tiate and map. The analysis took into account the asser-
dry weight in the area investigated. tion by Sanderson et al. (1983) that an over-emphasis on
Ling and Tyler (1986) recorded over 530 taxa from detailed and short-term sampling did not account suffi-
groups other than diatoms. The Desmidiaceae (desmids) ciently for the high natural variability. It is not known
were paricularly rich floristically and, as with the dia- how well the description provided for the Magela reflects
toms, were similar to those of south-east Asia. Broady the situation on other flood plains, especially given
(1984) investigated the widespread desiccated crusts and changes that have occurred with the removal of feral buf-
felts on the floodplain surface during the Dry season and faloes (Bubalus bubalis) (Skeat et al., 1996). Ten com-
recorded Microchaete and Scytonema species not previ- munities were identified (Fig. 5) and are briefly described
ously recorded. Heterogeneous genera, e.g. Scytonema, below.
Calothrix, Hapalosiphon and Stigonema, were present i) Melaleuca open forest and woodland (tree canopy
and secreted mucilaginous sheaths that combined with cover of 10-70 %): areas dominated by one or more
the intertwining filaments were assumed to stablilize the Melaleuca species - M. viridiflora and M. cajaputi
soil surface and reduce moisture loss. A characteristic around the edges and at the northern end of the
feature of the flood plain during the Dry season is the flood plain, and M. leucadendra in back-swamps
development of large patches of green and red surface inundated for 6-8 months. The understorey varies.
scums comprising phytoflagellates, in particular Pyrami- ii) Melaleuca open woodland (tree canopy cover <
monas, Chlamydomonas, Chlorogonium and Euglena 10 %): Melaleuca leucadendra in areas inundated for
species. Euglena sanguinea is responsible for the com- over 6 months. Understorey species are usually the
monly observed red colour, but changes in species com- same as those in adjacent areas of the flood plain.
position can cause a change to green. iii) Nelumbo-Nymphoides herbland: a mixed commu-
nity dominated by the water lilies Nelumbo nucifera
and Nymphoides indica that occur in permanently
Macrophytes and semi-permanently wet areas.
iv) Orzya grassland: dominated by Oryza meridionalis
Distribution. The summary provided by Finlayson (2005) towards the end of the Wet season. In the Dry sea-
of the information available on the vegetation of the son it consists of bare ground and dead Oryza me-
Magela flood plain has been adapted for use here. A list- ridionalis stems with persistent Phyla nodiflora and
ing of freshwater macrophytes (macro-algae and vascular Ludwigia adscendens as xerophytic forms, and
plants) found in the region was produced by Finlayson et Pseudoraphis spinescens.
382 C. M. Finlayson et al. Wetland biodiversity of the Kakadu region
v) Hymenachne grassland: dominated by Hymenachne achne acutigluma and Urochloa mutica along with
acutigluma throughout the year. the herb Ludwigia adscendens, occur along the
vi) Pseudoraphis grassland: dominated by the perennial banks of the billabongs.
emergent grass Pseudoraphis spinescens which has
a turf-like habit during the Dry season and grows up More recent analyses have shown that these communities
through the water during the Wet season. still predominate (J. Lowry unublished information)
vii) Hymenachne-Eleocharis grass-sedgeland: swampy along with several invasive species, in particular Salvinia
areas that dry out seasonally and are dominated by molesta and Urochloa mutica, and the native species Ses-
Hymenachne acutigluma or Eleocharis spp., which bania sesbans which intermittently dominates some areas
are slower to establish. of the plain, assumedly in responses to particular flooding
viii) Mixed grass-sedge-herbland: a variety of species and germination conditions that have not been identified.
with the dominant species depending on the topo- In addition to the flood plain analyses, Finlayson et al.
graphic situation. Oryza meridionalis occurs on the (1993a) reported a semi-quantitative analysis of vegeta-
drier sites with Pseudoraphis spinescens in slightly tion dominance in 5 billabongs from the Magela creek
wetter places, while Eleocharis spp. and Hymen- and flood plain – three located on small tributaries and
achne acutigluma occur in the deeper sites. On sites initially filled by water from the main creek rather than
that remain flooded for 10-11 months Nymphoides from the tributary, and two that represented remnants of
indica and Nymphaea macrosperma may be present. deep channels on the flood plain. The three located on the
ix) Eleocharis sedgeland: Eleocharis spp. dominate tributaries had a generalised vegetation zonation consist-
during the Wet season, but are replaced by annual ing of:
herbs during the Dry season; i) fringing Melaleuca spp. woodland in seasonally in-
x) Open-water community: permanent billabongs, undated areas;
flow channels, shallow waterholes contain Nym- ii) a mix of grasses and sedges in seasonally inundated
phaea macrosperma and Nymphaea pubescens and areas shaded by woodland;
a number of submerged plant species. Floating iii) a belt of Eleocharis sp. in water that is usually <1.5
grass mats comprising Leersia hexandra, Hymen- m deep during the Wet season;
Overview Article
Aquat. Sci. Vol. 68, 2006 383
Table 3. Number of plant species found in four broad habitats areas
iv) a small area of open water usually 1.5-2.0 m deep in
on the Magela flood plain (from Finlayson et al., 1989).
the Wet season; and
v) patches of waterlilies and submerged plants along the Habitat Total Annuals Perennials Geophytic
boundary between the sedges and the open water. species Perennials
Permanent billabongs 46 19 2 6
The dominant plant species, based on ‘visual biomass’
Seasonally inundated 94 57 29 8
and ‘ground cover’, are the Melaleuca spp. trees and the
flood plain
geophytic, perennial Eleocharis spp. sedges. In contrast,
Fringe zone 158 100 50 8
the two billabongs located on the flood plain had a gener-
Permanent swamps 21 5 11 5
alised vegetation zonation comprising:
i) fringing Melaleuca spp., Pandanus aquaticus, and
Barringtonia acutangula trees along a levee that
comprised the western bank;
ii) a mix of grasses and sedges and a few trees interfac- flood plain, except for the permanently wet areas. The
ing with the flood plain grass communities along seasonally inundated plain and the fringe zone contained
the gently sloping eastern bank; around 40 % and 70 % respectively, of the species, com-
iii) a mix of grasses, herbs and sedge overlaying a float- pared with 20 % in the billabongs and 10 % in the perma-
ing mat of Salvinia molesta extending from the nent swamps (Table 3).
banks towards the middle of the billabong; and In total there were 139 annual species with 102 terres-
iv) a discontinuous fringe of submerged plants and trial and 37 aquatic species. Eighty-nine of the terrestrial
waterlilies along the edge of the floating mat. species occurred in the fringe zone with 27 only on the
plain which is seasonally inundated for a longer period
As noted above, in recent decades the floodplain vegeta- than the fringe zone. Finalyson et al. (1989) provide a list-
tion has undergone considerable change as a consequence ing of growth strategy and form for each species. This in-
of invasion by feral animals, e.g. buffalo and pigs (Sus cluded 4 growth strategies (annual, perennial, or geophytic
scrofa), as well as weeds, e.g. mimosa (Mimosa pigra), perennial) and 2 primary (terrestrial or aquatic) and 7 sec-
salvinia (Salvinia molesta) and paragrass (Urochloa mu- ondary growth forms (tree, shrub, grass, sedge, vine, palm
tica), changes in fire regimes, and saline intrusion. How- or herb). Terrestrial annuals represented a diverse group of
ever, there have been few attempts to predict changes or species with 60 of them classified as herbs, 18 as sedges
project future scenarios for vegetation management de- and 17 as grasses. Twenty-seven of the aquatic annuals are
spite the extent of change and importance of the flood herbs and 6 are shrubs. There were 68 perennial species,
plains. 50 in the fringe zone. 34 of the perennials are terrestrial,
Finlayson et al. (1989) determined that both the dura- 26 aquatic, with 8 others difficult to classify. There are 12
tion and depth of inundation greatly influenced the distri- terrestrial trees including Eucalyptus spp., Pandanus spi-
bution of plant species, and attempted to explain this ralis, Lophostemon lactifluus and Syzygium suborbiculare.
though an empirical model of plant succession in wet- The aquatic perennial species are dominated by 12 herbs,
lands, developed by Van der Valk (1981), to assess chang- including Hydrilla verticillata, Ludwigia adscendens, Ne-
es in vegetation that occur as a result of changes in the lumbo nucifera and Nymphoides indica, and by 5 grasses,
annual hydrological pattern (Finlayson, 1990; Finlayson including the widespread Hymenachne acutigluma and
and Woodroffe, 1996). Despite limitations the model pro- Pseudoraphis spinescens. There were 14 geophytic spe-
vided a framework for predicting changes in vegetation cies, the more widespread include the Nymphaea and
patterns. Eleocharis species.
Within the broad categorisation of growth strategies
Growth strategies and forms. Finlayson et al. (1989) as- and forms there are many morphological and physiologi-
sessed the growth forms of the 222 plant species found cal adaptations that enable particular species to occupy
on the Magela flood plain across four broad habitat cate- various niches within the cycles of dry and wet condi-
gories, seasonally inundated plain, seasonally inundated tions. Coiwe et al. (2000a) provide an overview of the
fringe zone, billabong and permanent swamp. (The taxo- many specific adaptations that enable plants to survive
nomic listing and hence the number of species recorded the variability due to changes in the hydrological cycle.
by Finlayson et al. (1989) have not been updated in ac-
cordance with the taxonomy presented in the flora pro- Productivity. The productivity of the floodplain vegeta-
vided more recently by Cowie et al. (2000b). The fringe tion on the Magela flood plain has been investigated,
zone covered the edges of the flood plain and included covering seasonal changes in the dry weight of aquatic
the Melaleuca forests/woodlands, with the seasonally grasses and litterfall from Melaleuca trees (Finlayson,
inundated plain habitat covering the remainder of the 1988, 1991; Finlayson et al., 1993b). Changes in the
384 C. M. Finlayson et al. Wetland biodiversity of the Kakadu region
higher, although an analysis of density across the entire
floodplain was not undertaken.
Aquatic invertebrates
Available taxonomic information, zoogeographical rela-
tions and the general ecology of various groups of aquatic
invertebrates of the Region were reviewed by Humphrey
and Dostine (1994). Summary taxonomic data are pre-
sented in Table 4 based on Humphrey and Dostine (1994)
with updated information from Corbett (1996) and an
unpublished database on voucher specimens (Environ-
mental Research Institute of the Supervising Scientist).
This represents approximately 300 micro-invertebrate
(Copepoda, Cladocera, Rotifera) and over 600 macro-in-
vertebrate (insects, worms, mites, larger crustacean
groups, molluscs and sponges) species from freshwater
habitats in the Region, including sandstone plateau
streams and springs, upland (but below-escarpment) sec-
tions of streams with permanent flow, mid-reaches of
streams with seasonal flow, permanent waterbodies on or
adjacent to the main stream channels, and seasonally-in-
undated flood plains near the terminus of major streams.
The composition of the freshwater invertebrate fauna
Figure 6. Above-ground dry weight biomass of three aquatic grass of the Region is both Australian (Williams, 1980) and/or
species (from Finlayson, 1991). The least significant difference (P =
south-east Asian at generic and higher taxonomic levels,
depending upon the group. Across all taxonomic groups,
more data are required before making accurate compari-
sons with the invertebrate fauna in other climatic zones,
above-ground biomass (dry weight/unit area) or the both in Australia and elsewhere (Shiel and Williams,
widespread aquatic grasses Pseudoraphis spinescens, 1990; Humphrey and Dostine, 1994). Nevertheless, may-
Hymenachne acutigluma and Oryza meridionalis were fly (Ephemeroptera) and caddisfly (Trichoptera) data
determined during 1983-84 (Fig. 6; Finlayson, 1991). have been used in comparative global zoogeographical
The dry above-ground biomass of the grass species was studies (Vinson and Hawkins, 2003), while Cranston
influenced by water depth which in itself is directly re- (2000) reported high diversity of chironomid flies in both
lated to the rainfall pattern and surface water flow. The the wet and wet-dry tropics of Australia compared with
annual dry weight production for these species was 0.51 temperate Australia. Cranston et al. (1997) also observed
kg m–2 for Orzya meridionalis, 1.91 kg m–2 for Pseudora- high chironomid species richness in natural, as well as
phis spinescens, and 2.09 kg m–2 for Hymenachne acu- mining-induced, acidic waters of the Region, contrary to
tigluma. findings in temperate studies. They explained this by the
Dry weight production of the widespread Melaleuca large tropical (Australian and south-east Asian) pool of
woodlands and forests on the Magela flood plain were species tolerant of naturally-occurring acidic waters
estimated through an analysis of litterfall data (Finlays- common to these regions. Despite this, the Australian
on, 1988; Finlayson et al., 1993b). In an intensively sam- chironomid fauna is impoverished compared with
pled forest the total litterfall was approximately 0.7 kg northern hemispheric faunas (Cranston, 2000).
m–2 y–1 compared to 1.5 kg m–2 y–1 at a less intensively Kay et al. (1999) remarked upon the similarity of
sampled site (Finlayson, 1988). The value of 0.7 to 1.5 kg macro-invertebrate families in fresh waters of north-
m–2 y–1 is within the range recorded for other forests at the western Australia, compared with elsewhere across
same latitude (Lonsdale, 1988). The above ground bio- northern Australia. Cranston (2000) noted little additional
mass of Melaleuca species on the Magela flood plain was richness and only modest novelty of chironomid species
also calculated using an algorithm relating diameter at in waters of north-western Australia compared to the
breast height to tree height and fresh weight (Finlayson et Kakadu Region, a pattern that appears to hold generally
al., 1993b). This resulted in a calculated tree weight of for most other invertebrate faunal groups (C. Humphrey,
260 ± 0.3 t ha-1 based on a tree density of 294 trees ha–1; unpublished observations). Humphrey (1999) postulated
elsewhere on the floodplain tree densities were much that high seasonality of the extensive lowland aquatic
Overview Article
Aquat. Sci. Vol. 68, 2006 385
Table 4. Number of families, genera and species of aquatic invertebrates recorded in the Kakadu Region. Numbers in italics are
(under)estimates only, while ‘–’ indicates data not available.
Phylum/Class Order Families Genera Species
Mollusca/Gastropoda Pulmonata 4 6 7
5A
Prosobranchiata 2 5
Mollusca/Bivalvia Unionoida and Veneroida 2 2 2
Arthropoda/Insecta Diptera Chironomidae 43 122
7 (non- Chironomidae) – –
Trichoptera 9 21 105
Ephemeroptera 3 14 25
Coleoptera 16 50 100
Hemiptera 11 – –
Odonata 8 54 77
Lepidoptera 1 – 10
Arthropoda/Crustacea Isopoda 1 1 20
Decapoda 5 8 20
Ostracoda – – 6
Conchostraca 2 – 6
Arthropoda/Crustacea/Copepoda Calanoida and Cyclopoida – 7 14
Cladocera 6 30 48
Arthropoda/Arachnida Hydracarina 15 29 –
Annelida Oligochaeta 5 – –
Hirudinea 2 – –
Porifera – 4 7
Rotifera – – 227
A: Includes one exotic species
environments of northern Australia would select animals In contrast to the aquatic insects and the lowland
and plants that were readily dispersed. Moreover, lowland fauna generally of the Region, components of the
freshwater ecosystems of large parts of northern Australia crustacean groups the Decapoda and Isopoda, occurring
are relatively young in geological terms – a feature which, in freshwaters of the ancient ‘stone country’ of the
together with high seasonality and species vagility, has eastern part of the Region contain a substantial and sig-
probably mitigated against endemism at regional and nificant endemic component. This fauna includes an en-
smaller catchment scales (Humphrey, 1999). Some ele- demic family of shrimps, the Kakaducarididae, compris-
ments of the insect fauna from mainly upland sections of ing two endemic genera, Leptopalaemon and Kakaducaris
streams, however, may be more restricted in distribution. (Bruce, 1993; Bruce and Short, 1993), as well as an en-
This includes the mayfly family, the Leptophlebiidae; 7 of demic genus of phreatoicidean isopod (Eophreatoicus),
the 9 species of Leptophlebiids found in the Kakadu Re- of Gondwanic origin, that has exceptional species-level
gion appear to be restricted to the Northern Territory (P diversity (G. Wilson personal communication). Most of
Suter, unpublished data). Indeed, this insect family con- these macro-crustacean species have very restricted dis-
tains the only species in the Region (Tillyardophlebia tributions, often limited to single streams, seeps or
dostinei) so far found to be restricted to a single freshwater springs. This diversity and endemism was attributed by
stream (C. Humphrey unpublished data). Humphrey (1999) to the antiquity and persistence of the
Not only are major invertebrate faunal groups plateau/escarpment and associated perennial streams,
widespread across northern Australia, but at family-level springs and seeps, and isolating mechanisms including
at least, the fauna (including that of the Kakadu Region) fragmentation of habitat (long-term climate changes, ero-
has a high year-to-year persistence (constancy) compared sion) and the generally poor dispersal characteristics of
with other regions of Australia, a feature that is strongly these crustacean groups.
related to the relatively low degree of inter-annual The invertebrate ecology, including seasonal dynam-
variability of stream flow (Humphrey et al., 2000). ics, of seasonally-flowing portions of streams near an
386 C. M. Finlayson et al. Wetland biodiversity of the Kakadu region
Table 5. Diversity of freshwater fishes recorded from the Kakadu Region.
Order Family Number of genera Number of species Marine vagrant (m) or catadromous (c)
Pleurotremata Carcharhinidae 2 2 m
Pristidae 1 1 m
Hypotremata Dasyatidae 1 1 m
Osteoglossiformes Osteoglossidae 1 1
Elopiformes Megalopidae 1 1 c
Anguilliformes Anguillidae 1 1 c
Clupeiformes Clupeidae 2 3 m
Siluriformes Ariidae 1 3
Plotosidae 3 5
Atheriniformes Atherinidae 1 2
Synbranchiformes Synbranchidae 1 1
Pleuronectiformes Cynoglossidae 1 1
Soleidae 2 2 ?
Perciformes Belonidae 1 1
Melanotaeniidae 1 5
Pseudomugilidae 1 2
Ambassidae 2 3
Centropomidae 1 1 c
Terapontidae 5 6
Apogonidae 1 1
Toxotidae 1 2
Mugilidae 1 2 c
Gobiidae 1 3 c
Eleotrididae 5 8
Scatophagidae 2 2 m
Kurtidae 1 1
Gerreidae 1 1 m
Total 27 42 62
operating uranium mine (Ranger mine) and a uranium and Cranston, 1995; Hardwick et al., 1995; Humphrey et
ore deposit (Jabiluka) has been investigated for all phases al., 2002). The responses of aquatic invertebrates to a
of the hydrological cycle, including recolonisation after range of natural and human-related disturbances in the
early Wet season re-wetting, early-mid Wet season flows, Region have also been investigated, including fire re-
recessional flow period and pool-formation phase. Pal- gimes, invasive wetland grasses and tourism (Douglas,
tridge et al. (1997) found that upon re-wetting with the 1999; Douglas and O’Connor, 1999; M. Stowar unpub-
first stream flows in Magela Creek, most recolonising lished information).
taxa were derived from perennial upper reaches through
drift, though contributions were also derived from adja-
cent billabongs and resting stages (especially microcrus- Fishes
taceans) in the sandy substratum. The seasonal dynamics Freshwater fishes are among the most comprehensively
of aquatic invertebrates from permanent waterbodies on studied aquatic organisms in the Region (Taylor, 1964;
or adjacent to the main stream channels have also been Midgley, 1973; Pollard, 1974; Bishop et al., 1986, 1990,
studied (Marchant, 1982a, b, c; Outridge, 1988). 2001). The ecology, community structure and
Much effort has been directed at the use of aquatic biogeographic relationships of the fish fauna are
invertebrates for detecting and assessing mining impact reasonably well known and have been described and
(and recovery) in the Region (Humphrey, 1990; Hum- reviewed by Bishop and Forbes (1991). Sixty two
phrey and Dostine, 1994; Faith et al., 1991, 1995; Smith freshwater fish species have been recorded, represented
Overview Article
Aquat. Sci. Vol. 68, 2006 387
by 44 entirely freshwater species, 4 species that repro- tidal waters (H. Larson personal communication). The
duce in estuarine or marine waters (catadromous) and 14 two other shark species, the Bizant River Shark (Glyphis
marine or estuarine species that commonly enter non- sp. A) and the Northern River Shark (Glyphis sp. C) are
tidal freshwaters (marine vagrants) (Table 5). listed respectively as critically endangered and en-
All except three Australian freshwater fish (Neocera- dangered under IUCN criteria. Their distribution is
todus forsteri, Scleropages leichardti and S. jardinii) are limited to only a few river systems in northern Australia,
comparatively recent descendents from marine families. although it is acknowledged that further specimens and
When this recent colonisation of freshwater habitats is taxonomic work are required to properly determine the
combined with the small area of freshwater habitats on distribution and status of these species (Larson, 2000;
the Australian continent it is not surprising that the spe- Pogonoski et al., 2002).
cies richness of the continent (302 species) is very low by Longitudinal zonation of fish species is not very pro-
global standards (Allen et al., 2002).The diversity of nounced along the major stream channels. Marine va-
freshwater fish is higher in the tropics compared to tem- grants are more likely to be found in freshwater bodies
perate regions of Australia. Shiel and Williams (1990) close to the limit of tidal influence. Upstream there is lit-
suggested that periods of extreme aridity in the southern tle variation in the species composition until the first
temperate areas and ease of migration of new species dispersal barriers occur near the escarpment of the sand-
from marine to freshwaters as a result of greater length stone plateau. In this lowland zone on most major streams
and low-gradient of tropical rivers were most likely to 20 to 30 species are likely to be encountered at a single
determine species richness of freshwater fish. However, location. Upstream from dispersal barriers, such as wa-
Bishop and Forbes (1991) and more recently Unmack terfalls and large cascades, species diversity declines
(2001) have suggested an evolutionary explanation argu- greatly and pools on the plateau may contain only one or
ing that freshwater fish diversity in Australia is a reflec- two species. Most species utilise both riverine and asso-
tion of the fish diversity in adjacent seas and this is ciated lentic wetlands when available. There are 6 spe-
greater in the tropics. The fish diversity in the Kakadu cies that are restricted to riverine habitats and another 2
Region is relatively high by comparison to other regions that rarely enter lentic wetlands. At least 3 species occur
within the continent - this may be related to the rich di- only in small tributaries within the plateau and escarp-
versity of marine fish in the Indo-Pacific archipelago. ment zone.
Catchment size is a major factor determining fish di- Many of the fish species in the Region are common
versity in different drainage systems. Bishop and Forbes across northern Australia. Of 114 species known from
(1991) compared the diversity of fish in streams in the rivers of the wet-dry topics of northern Australia 32 oc-
Region with a generalised relationship between catch- cur throughout the entire zone; 11 are restricted to the
ment area and diversity developed for tropical rivers by Arafura bio-region of the Northern Territory; and 3 are
Welcomme (1979). By separating the drainage basin into endemic to the Region and nearby western Arnhemland,
sub-catchments they found that streams in the Region including Mariana’s Hardy Head (Craterocephalus mari-
had higher diversity than many other streams of similar anae), the Sharp-nose Grunter (Syncomistes butleri) and
size elsewhere; however, this seems spurious as it as- Midgley’s Grunter (Pingalla midgleyi). (Recent surveys
sumes that fish diversity in a sub-catchment is independ- have extended the known range of the hardyhead and the
ent of the rest of the catchment. Pusey and Kennard grunter.) Another 3 species have very disjunct distribu-
(1996) in comparing the fish diversity of 10 tropical riv- tions across northern Australia and Papua New Guinea,
ers along the far northern east coast of Queensland found including the Coal Grunter (Hephaestus carbo), Banded
that while catchment size is an important factor deter- Rainbowfish (Melanotaenia trifasciata) and Black-
mining species diversity, differences in fish community striped Rainbowfish (Melanotaenia nigrans). The latter
structure were related to latitude. reflect an historic freshwater connection between Aus-
About half the fish species encountered in the Kakadu tralia and Papua New Guinea that was severed by sea-
Region are small to medium in size, being usually less level rises within the past 30,000 years.
than 30 cm in length. The small-bodied fish fauna of the Upstream escarpment habitats are nutrient poor and
Region is dominated in overall abundance by are dominated by herbivore/detritivores and omnivores.
centropomids (perchlets), melanotaeniids (rainbow fish) The Sharp-nose Grunter has specially adapted teeth, jaws
and atherinids (hardyheads); the larger bodied fish fauna and intestines for grazing epiphytes from rocks and logs,
is dominated by ariids and plotosids (Fork and Eel tailed whilst the archerfish (Toxotes chatareus and Toxotes
Catfish), clupeids (Boney Bream), megalopids (tarpon) lorentzi) can capture terrestrial prey by emitting a jet of
and to a lesser extent, terapontids (grunters). Some other water from the mouth to dislodge the prey from
larger species include Barramundi (Lates calcarifer), overhanging vegetation (Bishop et al., 2001). A range of
Saratoga and three estuarine shark species. Only one trophic categories is represented amongst the fish com-
shark, the Bull Shark Carcharinus leucas, enters non- munities of lowland waterbodies, that are generally nutri-
388 C. M. Finlayson et al. Wetland biodiversity of the Kakadu region
ent rich. Fish in the Region exhibit a number of breeding lowing commercial exploitation except for limited
strategies, with a high proportion of species exhibiting catches for the aquarium fish trade.
some form of parental care, including live bearing, buccal
incubation and nest guarding (Bishop et al., 2001).
A characteristic feature of the ecology of the fish Amphibians
communities is the movement of fish between habitats Australia’s amphibia are represented by five families in
with the re-linking of Dry season refuge areas of the the order Anura (the frogs and toads) with four families
escarpment, corridor and floodplain zone once Wet of native frogs (the Microhylidae, Myobatrachidae (or
season flows commence (Bishop and Forbes, 1991). Leptodactylidae), Hylidae (or Pelodryadidae), and the
Feeding, breeding and recruitment of juveniles mostly Ranidae) and the introduced cane or marine toad, Bufo
occurs during the Wet season months with most species marinus from the Bufonidae family (Robert and Watson,
spawning early in the Wet season, taking advantage of 1993). The Kimberley region and the northern part of the
the increased area of habitat available to supply resources Northern Territory which contains the Kakadu Region
for growth and reproduction (Bishop et al., 2001). Many are considered to be an area of significant amphibian di-
species breed in off-river water bodies of lowland versity. Despite this general information there are many
streams and in the vast areas of inundated flood plain. In gaps in the knowledge about the ecology and biology of
some seasonal tributaries of the larger Alligator Rivers, many frog species of this area, with new species being
including Magela and Nourlangie Creeks, there are found as recently as 1997 and 2001 (Northern Territory
massive upstream migrations in the late Wet season of Frogs Database, 2003).
mainly small bodied fishes (dominated by ambassids and The first comprehensive study of frogs in the Kakadu
melanotaeniids) from floodplain nursery areas to Dry Region was undertaken in the Wet seasons of 1978–1979
season refuges (Humphrey and Dostine, 1994). These and 1979–1980 (Tyler et al., 1983). Woinarksi and Gam-
spectacular migrations often consist of several hundred bold (1992) undertook a further survey to examine distri-
thousand fish per hour moving past a single observation bution patterns of herpeto-fauna in relation to environ-
point (Bishop et al., 1995). mental gradients (substrate, moisture availability,
Large natural fish mortalities, or ‘fish kills’ are a vegetation); frogs were found to be most strongly associ-
common occurrence in the Region (Pidgeon, 2001). ated with substrate and moisture availability and oc-
These are caused by a number of factors with low curred in 4 distinct assemblages – a sandstone assem-
dissolved oxygen levels the most common cause. Influx blage, lowland clay-flat assemblage, wet forest
of toxic chemicals by natural processes (weakly acidic assemblage and two species with idiosyncratic ranges
waters containing toxic levels of aluminium, and leached (Woinarski and Gambold, 1992). These surveys have re-
ichthyocidal compounds such as saponins from trees and corded 25 frog species from 3 families – the Hylidae with
shrubs), and disease have also been recorded. Disease 2 genera and 16 species; the Myobatrachidae with 5 gen-
rarely results in large-scale mortalities at one time. Most era and 8 species); and the Microhylidae with 1 genera
fish kills occur during the early Wet season when the first and 1 species (Table 7). The cane toad is a recent invader
flows introduce oxygen depleting organic matter from (van Dam et al., 2002a).
the catchment and concentrated contaminants (described Most of the frog species display a clear seasonal di-
above). Storms and flows can mix anoxic water from the chotomy of behaviour, being inactive during the Dry
bottom of billabongs and stir up sediment and detritus season with at least 9 species aestivating underground
which can further reduce oxygen levels. Fish kills often (Tyler and Crook, 1987). Some of these species are fos-
result in the deaths of many thousands of fish and usually sorial (Cyclorana australis, C. longipes, Limnodynastes
involve the larger bodied fish such as Barramundi, Fork ornatus), while others (such as Litoria nasuta, L. inermis,
and Eel-tailed Catfish, Tarpon and Mullet (Bishop, 1980; L. wotjulemensis and L. tornieri) are known to descend
Bishop et al., 1982; Brown et al., 1981; Noller, 1983; down deep cracks in the hardened mud at the edge of bil-
Pidgeon, 2001). labongs (Tyler and Crook, 1987). In contrast, during the
There is no commercial fishery in freshwaters of the Wet season, frogs are very obvious and abundant (Tyler
Region, but there are important artisanal fisheries by in- and Crook, 1987) and are an important food source for a
digenous people and recreational fishery. The latter is variety of birds, reptiles and fish (Morris, 1996).
largely a sport fishery targeting only one species, the Bar- The sandstone assemblage of frogs is endemic to
ramundi. This fishery is highly managed with minimum northern Australia with species occupying crevices near
size limits, small bag limits and seasonal closures that small perennial streams in the sandstone escarpments and
vary with measured fishing pressures. Upland sections of plateaux (Braithwaite et al., 1991; Morris, 1996; Woinar-
streams in Kakadu are generally off-limits to (non-indig- ski and Gambold, 1992). These species cope with ex-
enous) recreational fishing. Otherwise, there is no man- treme environmental conditions with the tadpoles tolerat-
agement for other freshwater species other than not al- ing temperatures of around 42 °C, and the adults in the
Overview Article
Aquat. Sci. Vol. 68, 2006 389
Table 6. Frog species recorded from the Kakadu Region (from Press et al., 1995, taxonomy as per Northern Territory Frogs Database,
2003).
Family Species Common name(s)
Myobatrachidae Crinia bilingual Bilingual Frog, Ratchet Frog
Hylidae Cyclorana australis Burrowing Frog, Giant Frog
Hylidae Cyclorana longipes Long-footed Frog
Hylidae Litoria bicolour Northern Dwarf Tree Frog
Hylidae Litoria caerulea Green Tree Frog
Hylidae Litoria coplandi Saxicoline Tree Frog, Copland‘s Rock Frog
Hylidae Litoria dahlia Dahl‘s Aquatic Frog
Hylidae Litoria inermis Peter‘s Frog
Hylidae Litoria meiriana Rockhole Frog
Hylidae Litoria microbelos Dwarf Rocket Frog, Javelin Frog
(formerly Litoria dorsalis)
Hylidae Litoria nasuta Rocket Frog
Hylidae Litoria pallida Pale Frog
Hylidae Litoria personata Masked Cave Frog, Masked Rock Frog
Hylidae Litoria rothii Roth‘s Tree Frog
Hylidae Litoria rubella Desert Tree Frog, Red Tree Frog
Hylidae Litoria tornieri Tornier‘s Frog
Hylidae Litoria wotjumulensis Wotjulum Frog
Microhylidae Austrochaperina aldelphe Northern Territory Frog
(formerly Sphenophyrne adelphe)
Myobatrachidae Limnodynastes convexiusculus Marbled Frog
Myobatrachida Limnodynastes ornatus Ornate Burrowing Frog
Myobatrachidae Megistolotis lignarius Capenter Frog
Myobatrachidae Notoden melanoscaphus Northern Spadefoot Toad, Golfball Frog
Myobatrachidae Uperoleia arenicola Jabiru Toadlet
Myobatrachidae Uperoleia inundata Floodplain Toadlet
Myobatrachidae Uperoleia lithomoda Stonemason Toadlet
Dry season with limited surface water, low humidity and and Gambold, 1992). Most of these species are burrowing
high daytime temperatures (Morris, 1996). The sand- frogs (Uperoleia spp., Limnodynastes convexiusculus and
stone specialist assemblage consists of Megistolotis liga- Cyclorana australis). One of the burrowing frog species,
narius, Litoria personata, Litoria meiriana and Litoria Uperoloeia arenicola, is endemic to the Region. Another
coplandi. During the Dry season these species seek ref- group of frogs is most abundant in riparian or monsoon
uge in sandstone crevices and caves in the escarpment. forests, comprising mainly tree frogs (Litoria spp., and
M. lignarius and L. personata are endemic to the Kakadu Crinia bilingua) (Woinarski and Gambold, 1992).
sandstone escarpment.
Two species are found throughout the Region and
around Australia, without any strong association to envi- Reptiles
ronmental variables. Litoria caerulea has a widespread The reptile fauna of the Region is one of the most diverse
distribution throughout northern Australia and along the in Australia, with 127 species recorded (Table 7) and in-
east coast (Woinarski and Gambold, 1992). The burrowing cluding only one introduced species, the Asian House
frog Limnodynastes ornatus also has a widespread distri- Gecko (Hemidactylus frenatus). The herpeto-fauna com-
bution throughout Australia, but in the Kakadu Region is prise 15 families and 62 genera. From a conservation
most abundant on sandy flats with tall forests (Woinarski viewpoint, using the IUCN criteria (2000), the Logger-
and Gambold, 1992). The highest diversity of frogs within head Turtle (Caretta caretta) is considered endangered
the Region occurs on lowland wet clay flats, with open whereas there is little concern for the Green Turtle (Che-
vegetation and extensive tussock grass cover (Woinarski lonia mydas); both species of sea turtle have been re-
390 C. M. Finlayson et al. Wetland biodiversity of the Kakadu region
Table 7. Reptile fauna of the Region (from Press et al., 1995).
Order Suborder Family Number of Genera Number of Species
Crocodilia Crocodylidae 1 2
Testudines Carettochelydidae 1 1
Chelidae 3 5
Cheloniidae 4 4
Squamata Sauria Agamidae 5 9
Gekkonidae 7 16
Pygopodidae 3 4
Scincidae 11 36
Varanidae 1 11
Squamata Serpentes Acrochordidae 1 2
Boidae 3 6
Colubridae 8 8
Elapidae 10 13
Hydrophiidae 3 3
Typhlopidae 1 7
Total 15 62 127
ported nesting (Miles, 1998) on an island that is free of Of 11 species of goanna found in the region 5 are
most of the mainland egg eaters, such as dingos, feral found in aquatic or semi-aquatic habitats. In contrast to
pigs and humans (Miles, 2000). The Flatback (Nator de- the Mangrove Monitor, Merten’s (Varanus mertensi) and
pressa) and Pacific Ridley (Lepidochelys olivacea) Tur- Mitchells (Varanus mitchelli) Water Monitors are usually
tles have also been recorded nesting (Vanderlely, 1995). associated with freshwater streams and billabongs, ob-
The Oenpelli Python (Morelia oenpelliensis) is listed as taining approximately 40 % of their food (biomass) from
vulnerable. aquatic systems (Shine, 1986a). The sand goannas (Vara-
Eighteen species of large reptiles are listed as using nus gouldii and Varanus panoptes) occur over a range of
riparian and floodplain habitats of the Magela Creek sys- habitats, but frequently use dry flood plains and the edges
tem within Kakadu (Finlayson et al., 1990) with a further of watercourses to forage for food. The proportion of
6 water-dependent species. The latter includes: the Pig- prey from aquatic habitats that is eaten by V. panoptes is
nosed Turtle (Carettochelys insculpta); the Mangrove high, up to 30 % by weight, but negligible in V. gouldii
Monitor (Varanus indicus), restricted to mangrove com- (Shine, 1986a).
munities along the estuaries; the Little File Snake Six species of turtle are found in the freshwater
(Achrocordus granulatus); and three species of aquatic reaches of the rivers, creeks and billabongs in the Region.
colubrid snakes (Cerberus rynchops, Fordonia leuco- The Pig-nosed Turtle, although common where it occurs
balia and Myron richardsonii), usually found in estuarine in both Australia and Papua New Guinea, has a fairly
and mangrove habitats. limited distribution in the Region and, hence, the South
Two species of crocodile have a broad distribution Alligator River in Kakadu represents a significant refuge
across northern Australia and are plentiful in the region. (Press et al., 1995). Chelodina burrungandjii and Elseya
The Estuarine or Saltwater Crocodile (Crocodylus poro- latisternum are found in pools on the sandstone plateau,
sus) inhabits coastal rivers extending well inland via while Elseya dentata inhabits sandy or rocky stretches
major rivers and billabongs on the river floodplains. The between the escarpment and floodplains. (Press et al.,
Freshwater Crocodile (Crocodylus johnstoni) inhabits 1995). The characteristic floodplain turtle is the Northern
permanent freshwater rivers, lagoons and billabongs Long-necked Turtle (Chelodina rugosa) which is carniv-
across northern Australia (Cogger, 2000). There are no orous and during the Dry season survives by burying its-
reliable population estimates for these species in the Re- self in the mud and aestivating.
gion although it is generally considered that they are The Arafura File Snake (Achrochordus arafurae) is
populous and have recovered from the decimation that mainly restricted to the larger billabongs in the Dry sea-
occurred as a consequence of hunting activities that were son with many more in lower mainstream billabongs than
formally stopped in 1971. in the upper reaches of creeks (Shine, 1986b). In the Wet
Overview Article
Aquat. Sci. Vol. 68, 2006 391
Table 8. Waterbirds families listed in the Asia-Pacific Migratory
season most are found in shallow, inundated grasslands
Waterbird Strategy 2001–2005 and found in the Kakadu Region.
(Shine, 1986a). They are a popular food item for Abo-
riginal people in the Region (Shine, 1986b). The Little Taxonomic Family Common Name Number
File Snake inhabits estuarine habitats and feeds on crabs of species
and small fish. Other species of snake that occur on the
Podicipedidae Grebes 2
floodplains include the Water Python (Liasis fuscus) and
Phalacrocoracidae Cormorants and Darter 5
the King Brown Snake (Pseudechis australis). The Water and Anhingidae
Python has been recorded at the extraordinary high den-
Pelecanidae Pelicans 1
sity of 714 km2 on the nearby Adelaide River floodplain
Ardeidae Herons, Egrets and Bitterns 12
(Madsen and Shine, 1996). As water levels drop with the
Ciconidae Storks 1
onset of the Dry season a number of species, including
Threskiornithidae Ibises and Spoonbills 5
the Olive Python (Liasis olivaceus) and the Northern
Death Adder (Acanthophis praelongus) move onto the Anatidae Swans, Geese and Ducks 12
drying floodplain. The Keelback or Freshwater Snake Gruidae Cranes 2
(Tropidonophus mairii) is found in freshwater habitats, Rallidae Rails, Gallinules and Coots 7
whilst several species of colubrid snakes inhabit man- Jacanidae Jacanas 1
grove areas where they feed on crustaceans and fish.
Haematopodidae Oystercatchers 2
Three species of sea-snakes have been recorded in the
Recurvirostridae Stilts and Avocet 2
estuarine and tidal reaches of the rivers – Hardwick’s
Glareolidae Pratincoles 2
Sea-snake (Lapemis hardwickii), Darwin Sea-snake (Hy-
Chradriidae Plovers 11
drelaps darwiniensis) and Stoke’s Sea-snake (Astrotia
stokesii) (Press et al., 1995). Scolopacidae Sandpipers 22
The rate of endemism within the area is high. Sadlier Laridae Gulls, Terns and Skimmers 12
(1990) found 74 reptile species in the Magela Creek re-
gion, with 13 % appearing to be endemic to the wider
sandstone plateau.
systems of Kakadu National Park for waterbirds has
been recognised in the listing of the park under the Ram-
Waterbirds sar Wetlands Convention. The high diversity and abun-
The coastal floodplains and the alluvial floodplains of dance of waterbirds species are also a contributing factor
the Region are of immense importance for waterbirds to the listing of Kakadu National Park as a World Herit-
(Morton et al., 1991). Of the 20 families of waterbirds age area.
accepted under the Ramsar definition and included in the Within the Kakadu Region numerous roosts of shore-
Asia-Pacific Migratory Waterbird Strategy (2001–2005), birds, containing 2,000 or more birds, are spread along
16 are found in the Region (Table 8). These families the coastline with almost 9,000 individuals recorded at
comprise 99 species of waterbirds, of which 52 are shore- Finke Bay in a survey in September 1993 (Chatto, 2003).
birds (or waders) and 32 migratory (Morton et al., 1990; Ground surveys have also been used with around 12,500
Chatto, 2003). individuals being recorded in late April 1992 and late
The importance of the Region to waterbirds has been March 1992, along the coast between the South Alligator
extensively documented over many years. The wetlands River and Minimini Creek. The more numeros species
support many species of national and international sig- included Whimbrel (Numenius phaeopus), Eastern Cur-
nificance, including those protected under the Japan lew (Numenius madagascariensis), Common Green-
Australian Migratory Bird Agreement (JAMBA) and the shank (Tringa nebularia) and Black-tailed Godwit
China Australia Migratory Bird Agreement (CAMBA). (Limosa limosa). The mangroves fringing the river banks
The flood plains in particular provide important refuge provide suitable habitat for nesting colonial waterbirds
areas critical to the conservation of waterbirds through- with large multi-species colonies of Intermediate Egret
out northern Australia (Morton et al., 1991). During the (Egretta intermedia), Cattle Egret (Ardea ibis), Great
Dry season months of August-October the floodplains Egret (Ardea alba), Little Egret (Egretta garzetta), Pied
are used intensively by up to 2 million waterbirds, in- Heron (Ardea picata), Little Pied Cormorant (Phalacroc-
cluding large concentrations of geese and ducks (Morton orax melanoleucos), Little Black Cormorant (Phalacroc-
et al., 1991; Bayliss and Yeomans, 1990), such as the orax sulcirostris) and Australian White Ibis (Threskiornis
Magpie Goose (Anseranas semipalmata), the Wandering molucca) being recorded between 1990 and 1999 (Chat-
Whistling Duck (Dendrocygna arcuata) and Plumed to, 2000). In all 18 species of waterbirds have been re-
Whistling Duck (Dendrocygna eytoni). The importance corded in the mangroves and 27 on the inter-tidal flats.
of the mosaic of contiguous wetlands across the river Waterbird usage of the flood plains varies seasonally with
392 C. M. Finlayson et al. Wetland biodiversity of the Kakadu region
internationally important (i.e. contain more than the
1 %% of the population in the East Asian-Australasian
Flyway; Chatto, 2003).
Many other bird species that are not regarded as
waterbirds, as described above, depend on wetlands in
the Region for at least part of their life cycle (Table 9).
Whilst the Region is recognised as providing important
habitats for these species their ecology and population
levels have not on the whole been been investigated. The
number of wetland-dependent bird species present in the
habitat categories used in Table 9 are: coastal habitats 16;
mangroves 67; monsoon forests 99; and floodplain and
riparian areas 137. The riparian habitats in particular sup-
port many wetland-dependent species, such as Kingfish-
ers and the White-breasted Sea-eagle, and small passeri-
Figure 7. Seasonal use by waterbirds of four floodplains (Nour-
formes, such as Honeyeaters and Flycatchers, with about
langie, Magela and Cooper Creeks and East Alligator River) in the
Kakadu Region, as determined by aerial surveys (from Morton et half of them being endemic to northern Australia (Mor-
al., 1984).
ton and Brennan, 1991).
some species moving between the floodplains as habitat Mammals
conditions change in response to rainfall and flooding Many mammals may be considered wetland-dependent
and become more or less suitable for breeding, roosting in the sense that they access the water, although very few
and/or feeding (Morton et al., 1984). Waterbird usage of species spend their entire life cycle in the wetlands. A
four floodplains in the early 1980s is shown in Figure 7. notable exception is the Dugong (Sirenia artiodactyla).
The diets of aquatic herbivourous species, such as the A number of mammals are known to make some use of
Magpie Goose, Wandering Whistling Duck, Plumed the floodplain environment in the Northern Territory,
Whistling Duck, and the Green Pygmy Goose (Nettapus including 2 species of rodents, 2 species of dasyurids
pulchellus) are linked to the phenological state of the (carnivorous marsupials), 3 macropods, a number of in-
floodplain plants many which flower and seed towards sectivorous bats, 2 species of flying foxes, and the dingo
the end of the Wet season (Morton et al., 1990a, b; Fin- (Cowie et al., 2000a). The Dusky Rat (Rattus colletti)
layson et al., 1990). As the water recedes during the Dry seems to be a characteristic species that shelters in
season, a number of waterfowl species from southern cracks in the floodplain soils during the Dry season,
Australia, such as Hardhead (Aythya australis), Grey Teal feeds on seeds, and stems and corms of sedges and
grasses, and reaches high densities (around 150 ha–1 be-
(Anas gracilis), Pink-eared Duck (Malacorhynchus
membranaceus) arrive. The Dry season grasslands host ing recorded on a nearby floodplain; Madsen and Shine,
migratory species such as the Little Curlew (Numenius 1996). In the Wet season they move to adjacent higher
minutus) (Garnett and Minton, 1985; Bamford, 1988a, ground from which it recolonises the floodplain as they
1990; Morton et al., 1991; Schulz, 1989). The Little Cur- dry. The False Water Rat (Xeromys myoides) is seem-
lew, a species that breeds in Siberia and over-winters in ingly not as common as the Dusky Rat and may in fact
Australia, arrive in the Region in the latter part of Sep- be vulnerable. The fringing vegetation, in particular the
tember; they build up in numbers until the onset of the Melaleuca trees, provide roosting sites for colonies of
Wet season, at which time they move to the sub-coastal flying foxes (Pteropus scapulatus and Pteropus alecto)
floodplains some distance to the east and west of the Re- and when flowering are a source of nectar which also
gion (Bamford, 1990). Morton et al. (1991) estimated attracts the Northern Blossom Bat (Macroglossus min-
approximately 300,000 Little Curlew staging through the imus).
Region during October in the early 1980s; Bamford Introduced mammals that have invaded or make ex-
(1990) estimated 50,000 were present in the late dry sea- tensive use of the wetlands include the Asian water buf-
sons of 1987–1989. Other migratory shorebirds occur- falo (Bubalus bubalis), European domestic pig (Sus sco-
ring in large numbers during the migratory periods and fa), cattle (Bos taurus) and horses (Equs callabus). All
using the freshwater wetlands of the Region as well as are considered pest species with the water buffalo having
neighbourhood floodplains include the Sharp-tailed overgrazed and degraded the wetlands and except for a
Sandpiper (Calidris acuminata) and Marsh Sandpiper small domesticated herd have been removed from the
(Tringa stagnatilis), both species, as with the Little Cur- Park (Skeat et al., 1996).
lew, being present in numbers that make these wetlands
Overview Article
Aquat. Sci. Vol. 68, 2006 393
Table 9. Other wetland-dependent birds (i.e. those not classed as waterbirds) present in the Kakadu Region.
Family group Coastal wetland Mangrove Moonsoon forest Floodplain and
Riparian areas
Mound-builders 1 1
Quails 3
Kites-Hawks-Eagles 5 7 10 15
Falcons 1 6
Rails-Crakes-Swamphens-Coots 3 1 5
Bustards 1
Pigeons-Doves 3 6 4
Cockatoos 1 2 4 5
Lorikeets-Parrots 1 3 4
Cuckoos-Koels-Coucals 3 6 8
Owls 3 3
Barn Owls 2
Frogmouth 1 1
Nightjars 1 1 2
Owlet-Nightjars 1 1
Swifts 2 2 2 2
Kingfishers 2 5 5 6
Bee-eaters 1 1 1 1
Rollers 1 1 1 1
Pittas 1 1
Australian treecreepers 1
Fairy-wrens 1 1
Australian warblers 2 4 4
Honeyeaters-Chats 8 13 15
Robins 2 2 3
Babblers 1 1
Sitellas 1
Whistlers-Shrike-thrushes 5 4 4
Monarchs-Fantails-Drongos and Mag- 9 9 9
pie Larks
Cuckoo-shrikes 4 5 5
Orioles and Figbirds 2 3 3
Woodswallows and Butcherbirds 1 2 3 5
Crows 1 1 1 1
Bowebirds 1 1
Larks 1
Pipits and Wagtails 1 2
Weavers-Waxbills and allies 3 9
Flowerpeckers 1 1 1 1
Threats issues faced by floodplain wetlands in northern Australia
(Storrs and Finlayson, 1997; Finlayson et al., 2005).
A number of general analyses and reviews over the past Many of these affect the Kakadu Region, although the
few decades have identified the threats and management relative importance of each varies across the flood plains.
394 C. M. Finlayson et al. Wetland biodiversity of the Kakadu region
As elsewhere in northern Australia these issues have buthiuron (van Dam et al., 2004), uranium mining (van
been assessed in site-based analyses within the Region Dam et al., 2002b) and climate change (Bayliss et al.,
and some comprehensive databases now exist (e.g. man- 1997; Eliot et al., 1999). It is evident that information
agement of specific invasive species – Cook et al., 1996, collation and analysis is not only underway, but is ongo-
Rea and Storrs, 1999 - and uranium mining – Humphrey ing; the challenge is to translate this information into
et al., 1999, Gardner et al., 2002). Other issues are being practical advice for managers and for managers to heed
assessed, but not necessarily in a coherent manner (e.g. the warnings.
coastal and climate change – Eliot et al., 1999). The The complexity of managing wetlands within the
likely impacts of global climate change are perhaps the Kakadu Region is illustrated by two examples raised by
most insidious and least understood by local managers Finlayson (2005). The first considers changes in burning
(Bayliss et al., 1997). It is anticipated that climate change patterns on the floodplains. Recent investigations within
will not only affect the biological, chemical and physical Kakadu National Park have been trialing the use of fire to
features of wetlands, but will also interact strongly with establish floodplain vegetation mosaics more suited for
other existing or emerging pressures on wetlands with access by traditional Aboriginals for hunting and food
synergistic, antagonistic or cumulative effects. collection. In this instance access had been restricted by
Information on many threats is increasingly being an overabundance of the native grass Hymenachne acu-
collected through the adoption of structured risk assess- tigluma that had spread since the removal of buffalo from
ment procedures, such as that shown in Figure 8 (Finlay- the plains in the mid-1980s. Thus, we have a scenario
son and van Dam, 2004). For many of the main pest where the removal of a feral animal has been accompa-
species the extent of their invasion of the wetlands has nied by the expansion of a native plant that restricted
been assessed to some extent, although often incom- Aboriginal access to traditional food resources on the
pletely. In many instances the biology of the species may floodplains. Importantly, these investigations are being
also be known or is being studied; however, vital infor- led by local people with assistance from park managers
mation on the ecological changes wrought by these spe- and researchers (P. Christopherson and P. Bayliss per-
cies is often confined to a few isolated studies, if any, sonal communication). The second example covers the
and/or anecdotal evidence. Further information is be- complex biotic inter-relationships that occur on the
coming available through risk assessments for Mimosa floodplains, such as those between the large populations
(Mimosa pigra; Walden et al., 2004), cane toads (Bufo of magpie geese and the vegetation resources that they
marinus; van Dam et al., 2002a), the herbicide te- require for feeding and nesting. Finlayson (2005) specu-
lated that the number of geese could affect the future ex-
tent of at least some of these plant species either through
direct consumption or when nesting by damaging the
stems with subsequent physiological effects and changes
in growth conditions.
On the whole there has been little investigation of the
interactions between the ecological components of the
floodplain wetlands; some exceptions occur, but such
investigations do not seem to be considered de rigeur for
management planning or indeed for guiding responses to
the many threats faced by the wetlands. The recent focus
on risk assessments has gone someway towards provid-
ing information that can be more readily used for man-
agement purposes. There is however, an evident gap be-
tween collecting information on the biodiversity of the
wetlands and providing science-based guidance for man-
agement purposes, or for this guidance to be accepted
and used. The scenario shown through the assessment of
the likely impact of climate change and sea level rise
(Bayliss et al., 1997; Eliot et al., 1999) illustrates this
disjuncture. Climate change and sea level rise is project-
ed to result in the replacement of many freshwater wet-
lands with saline wetlands and with consequent displace-
ment or loss of the freshwater species diversity, and to
exacerbate the pressure from existing threats and pres-
Figure 8. Generalised framework for ecological risk assessment
sures. Yet, management steps have not been agreed, pos-
(modified from US EPA, 1998).
Overview Article
Aquat. Sci. Vol. 68, 2006 395
sibly due to limited awareness or understanding of the Region is sought it will be necessary to revisit past sur-
issues, or even resignation that adaptation rather than veys etc. and ascertain how much has changed – it is not
mitigation is a better response. At the same time full ad- fully known how well the information referred to in this
vantage has not been undertaken of monitoring opportu- paper reflects the current biodiversity status. It may also
nities, in this instance the existence of transects estab- be useful to address the priorities for further biodiversity
lished in the mangroves along the East Alligator River survey – should there be more effort seeking information
some 20 years ago as well as new technologies using GIS on endemic or vulnerable species, or are more general
and remotely sensed data, as shown by the mangrove surveys required first? Is there sufficient baseline or ref-
change detection undertaken along the West Alligator erence data to assess the effect of existing let alone fur-
River (Lucas et al., 2002). Making better use of such op- ther invasive species? Does this data correlate with and/
portunities is a challenge for researchers and managers or incorporate appropriate local information?
alike. The wetland landscape of the Kakadu Region has re-
ceived wide recognition and the available data supports
this. It is as widely recognized that these wetlands have
Conclusions undergone major change and face major pressures now
and in the near-future. Addressing these pressures is pos-
The species diversity of the wetlands of the Kakadu Re- sible – the tools are available and more recent manage-
gion is diverse and profuse. Many taxa have been identi- ment approaches have sought to involve local people and
fied and species distributions and seasonal and annual increase transparency in decision making about manage-
variations investigated. The information base though is ment actions as well as research priorities. The next steps
incomplete and uneven. Much data have been collected in ensuring the maintenance of the biodiversity values of
in specific locations with few region-wide investigations the wetlands of the Kakadu Region are complex, espe-
and with a few notable exceptions, e.g. Russell Smith et cially if they are subject to increasing pressure from both
al. (1997), based mainly on surveys undertaken by con- existing and emerging drivers of change. It is further an-
temporary scientific means with limited input from the ticipated that more sophisticated ecological understanding
local aboriginal community. The latter is changing and of the processes that support the food webs in the wetlands
increasingly local people are participating in joint activi- (Douglas et al., 2005) will assist in future management.
ties or initiating their own analyses in relation to changes
due to invasive species and fire management. Obtaining
Acknowledgments
region-wide information is being facilitated through GIS
and remote sensing (e.g. Lucas et al., 2002; Lowry and
Finlayson, 2004; Lowry and Knox, 2006) although con- Most of the information reported in this paper was col-
tinuing on-ground data collection may be difficult and lected through funding supplied through the Environ-
expensive to achieve, especially given the changing na- mental Research Institute of the Supervising Scientist
ture of the wetlands within and between years. Further, in over a period extending back to 1978. Collecting this in-
recent decades the wetlands have changed in response to formation involved a lot of work and dedication from
deliberate management actions – those addressing inva- many people and partner organizations, including the lo-
sive species being better known. The outcomes of these cal inhabitants of the Park and surrounding land. This
management actions have been complex with the re- effort is recognized through the production of this review
moval of large numbers of feral buffalo having been – thanks to everyone involved – it was not easy and it was
widely supported, but with less understanding of the not always fun but it was done.
likely consequences. It may or may not have been possi-
ble at the time to project a scenario where the native grass
References
Hymenachne acutigluma spread to dominate large parts
of the floodplains. Allen, G. R., S. H. Midgley and M. Allen, 2002. Field Guide to the
The advent of new technologies provides an opportu- Freshwater Fishes of Australia, Western Australian Museum,
nity for more effective inventory, assessment and moni- Perth, Australia, 394 pp.
Andersen, A., 1996. Fire ecology and management. In: C. M. Fin-
toring of the wetland environments; this needs to be ex-
layson and I. von Oertzen (eds.), Landscape and Vegetation
plored and suitable investments made. It may also be Ecology of the Kakadu Region, Northern Australia, KluwerA-
worthwhile taking a step back and more thoroughly ana- cademic Publishers, Dordrecht, The Netherlands, pp. 179–195.
Bayliss, P. and K. M. Yeomans, 1990. Seasonal distribution and
lyzing the information bases available and complement-
abundance of magpie geese, Anseranas semipalmata Latham,
ing the mainly qualitative risk assessments with more
in the Northern Territory, and their relationship to habitat,
quantitative and landscape-wide analyses based on statis- 1983–86. Australian Wildlife Research 17: 15–38.
tical probabilities and model scenarios. At the same time Bayliss, B., K. Brennan, I. Eliot, C. M. Finlayson, R. Hall, T. House,
R. Pidgeon, D. Walden and P. Waterman, 1997. Vulnerability
if a clear picture of the biodiversity value of the Kakadu
396 C. M. Finlayson et al. Wetland biodiversity of the Kakadu region
Assessment of Predicted Climate Change and Sea Level Rise in re-evaluation of Palaemontes holthuisi Strenth, with a designa-
the Alligator Rivers Region, Northern Territory, Australia, Su- tion of a new genus, Calathaemon. Hydrobiologia 257: 73–94.
pervising Scientist Report 123, Canberra, Australia, 134 pp. Carson, L. J., P. W. Haines, A. Brakel, B. A. Pietsch and P. A. Fer-
Bamford, M. J., 1988a. Kakadu National Park: a preliminary survey enczi, 1999. 1:250,000 Geological Map Series Explanatory
of migratory waders October/November 1987. RAOU Report Notes Minlingimbi SD53-2. Department of Mines and Engery,
No. 41. Northern Territory Geological Survey and Australian Geologi-
Bamford, M. J., 1990. RAOU Survey of Migratory Waders in cal Survey Organisation, Darwin, Australia.
Kakadu National Park: Phase III.RAOU Report 70. Royal Aus- Chatto, R, 2000. Waterbird breeding colonies in the Top End of the
tralasian Ornithologists Union, Melbourne. Northern Territory. Technical Report No 69. Parks and Wildlife
Bedell, E., 2001. Diatom Assemblages and Water Quality in the Commission of the Northern Territory, Palmerston, Australia,
Freshwater Wetlands of Kakadu National Park. Unpublished 156 pp.
Honours Thesis, Curtin University of Technology, Perth, Aus- Chatto, R., 2003. The distribution and status of shorebirds around
tralia, 77 pp. the coast and coastal wetlands of the Northern Territory. Techni-
Bishop, K. A. and Forbes, M. A., 1991. The freshwater fishes of cal Report No 73. Parks and Wildlife Commission of the North-
northern Australia. In: C. D. Haynes, M. G. Ridpath and M. A. ern Territory, Palmerston, Australia, 206 pp.
J. Williams (eds.), Monsoonal Australia: Landscape, Ecology Cogger, H. G., 2000. Reptiles and Amphibians of Australia. Reed
and Man in the Northern Lowlands, A.A. Balkema, Rotterdam, Books Australia Pty Ltd, Melbourne, Australia, 788 pp.
pp. 79–107. Cook, G. D., S. A. Setterfield and J. Maddison, 1996. Scrub invasion
Bishop, K. A., 1980. Fish kills in relation to physical and chemical of a tropical wetland: Implications for weed management. Eco-
changes in Magela Creek (East Alligator River system, North- logical Applications 6: 531−537.
ern Territory) at the beginning of the tropical Wet season. Aus- Corbett, L. 1996. Aquatic Studies at Ranger Mine: A Whole Ecosys-
tralian Zoologist 20: 485–500. tem Approach. CSIRO Division of Wildlife and Ecology, Dar-
Bishop, K. A., S. A. Allen, D. A. Pollard and M. G. Cook, 1986. win, Australia, 152 pp.
Ecological Studies on the Freshwater Fishes of the Alligator Cowie, I. D., 1996. Weed ecology. In: C. M. Finlayson and I. von
Rivers Region, Northern Territory. Volume I: Outline of the Oertzen (eds.), Landscape and Vegetation Ecology of the Kakadu
Study, Summary, Conclusions and Recommendations. Research Region, Northern Australia, Kluwer Academic Publishers, Dor-
Report 4 (i), Supervising Scientist for the Alligator Rivers Re- drecht, The Netherlands, pp. 113–135.
gion, AGPS, Canberra, Australia, 63 pp. Cowie, I. D., M. D. Armstrong, J. C. Z. Woinarski, P. S. Brockle-
Bishop, K. A., S. A. Allen, D. A. Pollard and M. G. Cook, 1990. hurs, P. S. Short and C. R. Dunlop, 2000a. An overview of the
Ecological Studies on the Freshwater Fishes of the Alligator floodplains. In: I. D. Cowie, P. S. Short, P. Osterkamp and M.
Rivers Region, Northern Territory. Volume II: Synecology. Re- Madsen (eds.), Floodplain Flora: A Flora of the Coastal Flood-
search report 4 (ii), Supervising Scientist for the Alligator Riv- plains of the Northern Territory, Australia, ABRS, Canberra,
ers Region, AGPS, Canberra, Australia. Australia, pp. 1–33.
Bishop, K. A., S. A. Allen, D. A. Pollard and M. G. Cook, 2001. Eco- Cowie, I. D., P. S. Short, P. Osterkamp and M. Madsen, (eds.)
logical Studies on the Freshwater Fishes of the Alligator Rivers 2000b, Floodplain Flora: A Flora of the Coastal Floodplains of
Region, Northern Territory: Autecology. Supervising Scientist the Northern Territory, Australia, ABRS, Canberra, Australia,
Report 145, Supervising Scientist, Darwin, Australia, 570 pp. 382 pp.
Bishop, K. A., L. M. Baker and B. N. Noller, 1982. Naturally-occur- Cranston, P., 2000. August Thienemann’s influence on modern
ring ichthyocides and a report on Owenia vernicosa F. Muell. chironomidology – an Australian perspective. Verh. Interna-
(Family Meliaceae), from the Magela Creek system, Northern tionalis Vereinigung Limnologiae. 27: 278–283.
Territory. Search 13: 150–153. Cranston, P. S., P. D. Cooper, F. R. A. Hardwick, C. L. Humphrey
Bishop, K. A., R. W. J. Pidgeon and D. J. Walden, 1995. Studies of and P. L. Dostine, 1997. Tropical acid streams–the chironomid
fish movement dynamics in a tropical floodplain river: Prereq- (Diptera) response in northern Australia. Freshwater Biology
37: 473–483.
uisites for a procedure to monitor the impacts of mining, in the
Douglas, M. M., 1999. Tropical Savanna Streams: Their Seasonal
use of biota to assess water quality, Australian Journal of Ecol-
Dynamics and Response to Catchment Disturbance, Doctor of
ogy 20: 81–107.
Philosophy thesis, Monash University, Melbourne, Australia.
Braithwaite, R. W., Friend, G. R. and J. C. Wombey, 1991. Reptiles
Douglas, M. M. and R. A. O’Connor, 1999. Effects of the exotic
and amphibians. In: C. D. Haynes, M. G. Ridpath and M. A. J.
macrophyte, para grass (Urochloa mutica), on benthic and epi-
Williams (eds.), Monsoonal Australia: Landscape, Ecology and
phytic macroinvertebrates of a tropical floodplain. Freshwater
Man in the Northern Lowlands, A.A. Balkema, Rotterdam, The
Biology 48: 962–971.
Netherlands, pp. 109–124.
Douglas, M. M., S. E. Bunn and P. M. Davies, 2005. River and wet-
Brennan, K., 1996. An Annotated Checklist of Vascular Plants of the
land food webs in Australia’s wet-dry tropics: general principles
Alligator Rivers Region, Northern Territory. Darwin, Australia:
and implications for management. Marine and Freshwater Re-
Supervising Scientist Report 109, Supervising Scientist, Dar-
search 56: 329–342.
win, Australia, 136 pp.
East, T. J., 1996. Landform evolution. In: C. M. Finlayson and I. von
Broady, P. A., 1984. A Preliminary Investigation of the Cyanobacte-
Oertzen (eds.), Landscape and Vegetation Ecology of the
ria (“Blue-green Algae”) 0f the Magela Creek System and
Kakadu Region, Northern Australia, KluwerAcademic Publish-
Nearby Sites, Northern Territory, Australia. Unpublished report
ers, Dordrecht, The Netherlands, pp. 37–55.
to the Supervising Scientist for the Alligator Rivers Region,
Eliot, I., C. M. Finlayson and P. Waterman, 1999. Predicted climate
Northern Territory, Australia.
change, sea level rise and wetland management in the Austral-
Brown, T. E., A. W. Morley, N. T. Sanderson and R. D. Tait, 1981.
Report of a large fish kill resulting from natural acid water ian wet-dry tropics. Wetlands Ecology and Management 7:
conditions in Australia. Journal of Fish Biology 22: 335–350. 63–81.
Bruce, A. J., 1993. Kakaducaris glabra gen. nov., sp. nov., a new Faith, D. P., P. L. Dostine and C. L. Humphrey, 1995. Detection of
freshwater shrimp from the Kakadu National Park, Northern mining impacts on aquatic macroinvertebrate communities: re-
Territory, Australia, Crustacea: Decapoda: Palaemonidae with sults of a disturbance experiment and the design of a multivari-
the designation of a new subfamily Kakaducaridinae. Hydrobi- ate BACIP monitoring programme at Coronation Hill, Northern
ologia 268: 27–44. Territory. Australian Journal of Ecology 20: 167–180.
Bruce, A. J. and J. W. Short, 1993. Leptopalaemon gagadjui gen. nov., Faith, D. P., C. L. Humphrey and P. L. Dostine, 1991. Statistical
a new freshwater palaemonid shrimp from Arnhem land, and a power and BACI designs in biological monitoring: comparative
Overview Article
Aquat. Sci. Vol. 68, 2006 397
evaluation of measures of community dissimilarity based on Australia. Supervising Scientist Report 169, Supervising Scien-
benthic macroinvertebrate communities in Rockhole Mine tist, Darwin, Australia, 145 pp.
Creek, Northern Territory, Australia. Australian Journal of Ma- Garnett, S. and C. Minton, 1985. Notes on the movements and dis-
rine and Freshwater Research 42: 589–602. tribution of Little Curlew Numenius minutes in northern Aus-
Finlayson, C. M., 1988. Productivity and nutrient dynamics of sea- tralia. Australia Bird Watcher 11: 69–73.
sonally inundated flood plains in the Northern Territory. In: D. Gopal, B. and W. J. Junk, 2000. Biodiversity in wetlands: an introduc-
Wade-Marshall and P. Loveday (eds.), North Australia: Progress tion. In: B. Gopal, W. J. Junk and J. A. Davis (eds..), Biodiversity
and Prospects, vol. 2 Floodplain Research, ANU Press, Darwin, in wetlands: assessment, function and conservation, Vol. 1, Back-
Australia, pp. 58–83. huys Publishers b.V., Leiden, The Netherlands, pp. 1–10.
Finlayson, C. M., 1990. Plant ecology and management of an inter- Hardwick, R. A., P. D. Cooper, P. S. Cranston, C. L. Humphrey and
nationally important wetland in monsoonal Australia. In: J. A. P. L. Dostine, 1995. Spatial and temporal distribution patterns
Kusler and S. Day (eds.), Wetlands and River Corridor Manage- of drifting pupal exuviae of Chironomidae (Diptera) in streams
ment, The Association of Wetland Managers, New York, USA, of tropical northern Australia. Freshwater Biology 34:
pp. 90–98. 569−578.
Finlayson, C. M., 1991. Primary production and major nutrients in Hoatson, D., D. Blake, A. Wygralak, S. Needham, Allen, B., Miles,
three grass species on a tropical flood plain in northern Aus- G., P. Hauser and I. Oswald-Jacobs, 2000. Kakadu and Nitmiluk
tralia. Aquatic Botany 41: 263–280. National Parks, Northern Territory: A Guide to the Rocks,
Finlayson, C. M., 2005. Plant ecology of Australia’s tropical flood- Landforms, Plants, Animals, Aboriginal Culture and Human
plain wetlands: a review. Annals of Botany 96: 541–555. Impact. Australian Geological Survey Organisation, Canberra,
Finlayson, C. M. and R. A. van Dam, 2004. Developing local capac- Australia, 110 pp.
ity and ensuring relevance in risk assessment for tropical wet- Humphrey, C., 1999. Potential effects of the Jabiluka Mine upon the
lands. SETAC Globe Jan–Feb, 36–38. biodiversity of Kakadu National Park. Appendix 3 in Response
Finlayson, C. M. and I. von Oertzen, 1996a. The Kakadu region. In: to the ICSU Review of the Supervising Scientist’s Report to the
C. M. Finlayson and I. von Oertzen (eds.), Landscape and Veg- World Heritage Committee. Supervising Scientist, Environment
etation Ecology of the Kakadu Region, Northern Australia, Australia, Darwin, Australia.
KluwerAcademic Publishers, Dordrecht, The Netherlands, pp. Humphrey, C. L., A. W. Storey and L. Thurtell, 2000. AUSRIVAS:
1–15. operator sample processing errors and temporal variability-
Finlayson, C. M. and I. von Oertzen, (eds.) 1996b. Landscape and implications for model sensitivity. In: J.F. Wright, D. W. Sutcliffe
Vegetation Ecology of the Kakadu Region, Northern Australia, and M. T. Furse (eds.), Assessing the Biological Quality of Fresh
KluwerAcademic Publishers, Dordrecht, The Netherlands, 202 Waters: RIVPACS and Other Techniques, Freshwater Biological
pp. Association, Ambleside, U.K., pp. 143–163.
Finlayson, C. M. and C. D. Woodroffe, 1996. Wetland vegetation. Humphrey, C. L., L. Thurtell, R. W. J. Pidgeon, R. A. van Dam and
In: C. M. Finlayson and I. von Oertzen (eds.), Landscape and C. M. Finlayson, 1999. A model for assessing the health of
Vegetation Ecology of the Kakadu Region, Northern Australia, Kakadu’s streams. Australian Biologist 12: 33–42.
KluwerAcademic Publishers, Dordrecht, The Netherlands, pp. Humphrey, C. L., P. Martin and C. leGras, 2002. Use of the freshwa-
81–112. ter mussel, Velesunio angasi, in the monitoring and assessment
Finlayson, C. M., B. J. Bailey and I. D. Cowie, 1989. Macrophytic of mining impact in Top End streams. In: S. J. Markich and
Vegetation of the Magela Flood Plain, Northern Australia. Re- R.A. Jeffree (eds.), The Finniss River: A Natural Laboratory of
search Report No. 5, Office of the Supervising Scientist, Syd- Mining Impacts – Past, Present and Future, Australian Nuclear
ney, Australia, 38 pp. Science and Technology Organisation, report E/748, Sydney,
Finlayson, C. M., M. G. Bellio and J. B. Lowry, 2005. A conceptual Australia, pp. 43–48.
basis for the wise use of wetlands in northern Australia – linking Humphrey, C. L. and P. Dostine, 1994. Development of biological
information needs, integrated analyses, drivers of change and hu- monitoring programs to detect mining waste impacts upon
man well-being. Marine and Freshwater Research 56: 269–277. aquatic ecosystems of the Alligator Rivers Region, Northern
Finlayson, C. M., I. D. Cowie and B. J. Bailey, 1990. Characteristics Territory, Australia. Proceedings of Conservation and Manage-
of a seasonally flooded freshwater system in monsoonal Aus- ment of tropical inland waters: Problems, solutions and pros-
tralia. In: D. F. Whigham, R. E. Good and J. Kvet (eds.), Wet- pects, Conference. University of Hong Kong. Mitteilungen In-
land Ecology and Management: Case studies, Kluwer Aca- ternationalis Vereinigung Limnologiae 24: 293–314.
demic Publishers, Dordrecht, The Netherlands, pp. 141–162. Kay, W. R., M. J. Smith, A. M. Pinder, J. M. McRae, J. A. Davis and
Finlayson, C. M., I. D. Cowie, and B. J. Bailey, 1993b. Litterfall in S. A. Halse, 1999. Patterns of distribution of macroinvertebrate
a Melaleuca forest on a seasonally inundated flood plain in families in waters of north-western Australia. Freshwater
tropical northern Australia. Wetland Ecology and Management Ecology 41: 299–316.
2: 177–188. Larson, H., 2000. Threatened fish profiles – northern speartooth
Finlayson, C. M., R. N. Hall and B. L. Bayliss, 1997. Regional Re- shark Glyphis sp. C. Australian Society for Fish Biology
view of Wetland Management Issues: Wet-dry Tropics of Newsletter 30 (1): 30.
Northern Australia, LWRRDC Occasional Paper 03/97, Can- Ling, H. U. and Tyler, P. A., 1986. A Limnological Survey of the
berra, Australia, 35 pp. Alligator Rivers Region II. Freshwater Algae Exclusive of
Finlayson, C. M., K. Thompson, I. A. M. L von Oertzen and I. D. Diatoms. Research Report 3, Supervising Scientist for the
Cowie, 1993a. Vegetation of Five Magela Creek Billabongs, Alligator Rivers Region, AGPS, Canberra, Australia, 176 pp.
Alligator Rivers Region, Northern Australia. Technical Memo- Lonsdale, W. M., 1988. An analysis of litterfall in forests of the
randum 46, Supervising Scientist for the Alligator Rivers Re- world. Annals of Botany 61: 319–324.
gion, AGPS, Canberra, Australia, 26 pp. Lowry, J. B. and C. M. Finlayson, 2004. A Review of Spatial Data-
Finlayson, C. M., B. J. Bailey, W. J. Freeland and M. R. Fleming, sets for Wetland Inventory in Northern Australia. Supervising
1988. Wetlands of the Northern Territory. In: A. J. McComb and Scientist Report 178, Supervising Scientist, Darwin, Australia,
P. S. Lake (eds.), The Conservation of Australian Wetlands, Sur- 18 pp.
rey Beatty and Sons Pty Limited, Sydney, Australia, pp. 103– Lowry, J. B. and M. Knox, 2002. Geomorphic landscapes of the
126. Kakadu region, Northern Territory, Australia. Supervising Sci-
Gardner, S., C. M. Finlayson and R. W. J. Pidgeon, (eds.) 2002. entist poster, Supervising Scientist, Darwin, Australia.
Description and Literature Review of the Flora and Vertebrate Lowry, J. B. and M. Knox, 2006. A description of the geomorphic
Fauna of Magela Creek, Alligator Rivers Region, Northern landscapes of the Kakadu region, Northern Territory, Australia.
398 C. M. Finlayson et al. Wetland biodiversity of the Kakadu region
Supervising Scientist Report 186, Supervising Scientist, Dar- Pogonoski, J. J., D. A. Pollard and J. R. Paxton, 2002. Conservation
win, Australia. Overview and Action Plan for Australian Threatened and
Lucas, R. M., J. C. Ellison, A. Mitchell, B. Donnelly, M. Finlayson, Potentially Threatened Marine and Estuarine Fishes 2002.
A. K. Milne, 2002. Use of stereo aerial photography for quanti- Environment Australia, Canberra, Australia.
fying changes in the extent and height of mangroves in tropical Pollard, D. A., 1974. The freshwater fishes of the Alligator Rivers Re-
Australia. Wetlands Ecology and Management 10: 161–175. gion ‘Uranium Province’ area (Top End Northern Territory) with
Madsen, T. and Shine, R., 1996. Seasonal migration of predators particular reference to the Magela Creek catchment (East Alligator
and prey – a study of pythons and rats in tropical Australia. River System). Report 1. In: P. J. F. Newton (ed.), Alligator Rivers
Ecology 77: 149–156. Area Fact Finding Study, Report E305, Australian Atomic Energy
Marchant, R., 1982a. The macroinvertebrates of Magela Creek, North- Commission, AAEC, Sydney, Australia, pp. 1–71.
ern Territory. Research Report No. 1, Supervising Scientist for the Press, T., J. Brock and A. Anderson, 1995. Fauna. In: T. Press, D.
Alligator Rivers Region, AGPS, Canberra, Australia, 46 pp. Lea., A. Webb and A. Graham (eds.), Kakadu: Natural and Cul-
Marchant, R., 1982b. Life spans of two species of tropical mayfly tural Heritage and Management, Australian Nature Conserva-
nymph (Ephemeroptera) from Magela Creek, Northern Territo- tion Agency and North Australia Research Unit, The Australian
ry. Australian Journal of Marine and Freshwater Research 33: National University, Darwin, Australia, pp. 197–214.
173–179. Pusey, B. J. and M. J. Kennard, 1996. Species richness and geo-
Marchant, R., 1982c. Seasonal variation in the macroinvertebrate graphical variation in assemblage structure of the freshwater
fauna of billabongs along Magela Creek, Northern Territory. fish fauna of the Wet Tropics Region of northern Queensland.
Australian Journal of Marine and Freshwater Research 33: Marine and Freshwater Research 47: 563–573.
329–342. Rea, N. and M. J. Storrs, 1999. Weed invasions in wetlands of Aus-
McBride, T. P., 1983. Diatom Communities of the Mudginberri Cor- tralia’s Top End: reasons and solutions. Wetland Ecology and
ridor, Northern Territory, Australia. Doctor of Philosopy thesis, Management 7: 47–62.
Macquarie University, Sydney, Australia. Roberts, J. D. and G. F. Watson, 1993. Biogeography and Phylogeny
Midgley, S. H., 1973. Freshwater fish inventory and associated of the Anura. In: C. J. Glasby, G. J. B. Ross and P. L. Beesley
habitat information. Alligator Rivers Region Environmental (eds.), Fauna of Australia, Vol. 2A, Australian Government
Fact Finding Study. Department of the Northern Territory, Can- Publishing Service, Canberra, Australia.
berra, Australia, 30 pp. Russell Smith, J., D. Lucas, M. Gapindi, B. Gunbunuka, N. Kapiri-
Miles, G., 1998. Wildlife of Kakadu and the Top End of the Northern gi, G. Namingam, K. Lucas, P. Giuliani and G. Chaloupka,
Territory, Australia, Barker Souvenirs, Alice Springs, Australia. 1997. Aboriginal resource utilisation and fire management prac-
Miles, G., 2000. Kakadu Wildlife: Kakadu: A World Heritage Na- tices in western Arnhem land, monsoonal northern Australia:
tional Park, Northern Territory, Australia. Barker Souvenirs, notes for prehistory, lessons for the future. Human Ecology 25:
Alice Springs, Australia. 159–195.
Morris, I., 1996. Kakadu National Park, Australia. Steve Parish Sadlier, R. A., 1990. The Terrestrial and Semiaquatic Reptiles (Lac-
Publishing, Bowen Hills, Australia, 225 pp. ertilia, Serpentes) of the Magela Creek Region, Northern Terri-
Morton, S. R., K. G. Brennan and P. Dostine, 1984. Diversity, abun- tory. Technical Memorandum 32, Supervising Scientist for the
dance and importance of avian food links in food chains. In: Alligator Rivers Region, AGPS, Canberra, Australia, 86 pp.
Alligator Rivers Region Research Report 1983-1984. Austral- Sanderson, N. T., D. V. Koontz and A. W. Morley, 1983. The ecol-
ian Government Publishing Service, Canberra. ogy of the vegetation of the Magela Creek flood plain: upper
Morton, S. R., K. G. Brennan and M. D. Armstrong, 1990a. Distri- section from Oenpelli road crossing to Nankeen Billabong.
bution and abundance of Magpie Geese, Anseranas semipal- Unpublished report in Scientific Workshop, Environmental
mata, in the Alligator Rivers Region, Northern Territory. Aus- Protection in the Alligator Rivers Region, Jabiru, Australia, pp.
tralian Journal of Ecology 15: 307-320. 33.1–33.9.
Morton, S. R., K. G. Brennan and M. D. Armstrong, 1990b. Distri- Schulz, M., 1989. The importance of wetlands in Kakadu National
bution and abundance of ducks in the Alligator Rivers Region, Park to selected waterbirds. A Report to the Asutralian National
Northern Territory. Australian Wildlife Research 17: 573-590. Parks and Wildlife Service, May 1989.
Morton, S. R. and K. G. Brennan, 1991. Birds. In: C. D. Haynes, M. Shiel, R. J. and W. D. Williams, 1990. Species richness in tropical
G. Ridpath and M. A. J. Williams (eds), Monsoonal Australia: freshwaters of Australia. Hydrobiologia 202: 175–183.
Landscape, ecology and man in the northern lowlands, AA Shine, R., 1986a. Diets and abundances of aquatic and semi-aquatic
Balkema, Rotterdam, Netherlands, pp. 133–149. reptiles in Alligator Rivers Region. Technical Memorandum 16,
Noller, B. N., 1983. Physico-chemical aspects of natural fish kills on Supervising Scientist for the Alligator Rivers Region, AGPS,
the Magela Creek system. In: Proceedings of the Scientific Canberra, Australia, 57 pp.
Workshop on Environment Protection in the Alligator Rivers Shine, R., 1986b. Predation upon filesnakes (Acrochordus arafurae)
Region, Paper no. 61. Supervising Scientist for the Alligator by aboriginal hunters: selectivity with respect to size, sex and
Rivers Region, Sydney, Australia. reproductive condition. Copeia 1986: 238–239.
Northern Territory Frogs Database, 2003. CD-Rom, Frogwatch, Skeat, A. J., T. J. East and L. K. Corbett, 1996. Impact of feral water
Noonamah, Australia. buffalo. In: C. M. Finlayson and I. von Oertzen (eds.), Landscape
Outridge, P. M., 1988. Seasonal and spatial variations in benthic and Vegetation Ecology of the Kakadu Region, Northern Aus-
macroinvertebrate communities of Magela Creek, Northern tralia, Kluwer Academic Publishers, Dordrecht, The Netherlands,
Territory. Australian Journal of Marine and Freshwater Re- pp. 155–177.
search 39: 211–223. Smith, M. and P. Cranston, 1995. Recovery of an acid mine-waste
Ovington, D., 1986. Kakadu: A World Heritage of Unsurpassed impacted tropical stream: The chironimid story. In: Chirono-
Beauty. Canberra, Australia: Australian Government Publishing mids, from Genes to Ecosystems, 12th International Conference
Service, 116 pp. on Chironomidae, 161–173.
Paltridge R. M., P. L. Dostine, C. L. Humphrey and A. J. Boulton, Storrs, M. J. and C. M. Finlayson, 1997. Overview of the Conserva-
1997. Macroinvertebrate recolonization after re-wetting of a tion Status of Wetlands of the Northern Territory. Supervising
tropical seasonally-flowing stream (Magela Creek, Northern Scientist Report 116, Supervising Scientist, Canberra, Austral-
Territory, Australia). Marine and Freshwater Research 48: ia, 90 pp.
633−645. Taylor, W. R., 1964. Fisheries of Arnhem Land. In: Records of the
Pidgeon, B., 2001. Natural causes of fish kills in Top End waterways. American–Australian Scientific Expedition to Arnhem Land,
eriss note, Supervising Scientist, December, Australia, 4 pp. Vol. IV: Zoology 45–307.
Overview Article
Aquat. Sci. Vol. 68, 2006 399
Taylor, J. A. and D. Tulloch, 1985. Rainfall in the wet-dry tropics: Van der Valk, A. G., 1981. Succession in wetlands: a Gleasonian
extreme events at Darwin and similarities between years during approach. Ecology 62: 688–696.
the period 1870–1983. Australian Journal of Ecology 10: Vanderlely, R., 1995. An Ecological Survey of the Sea Turtles of
281−295. West Alligator Head and Field Island, Kakadu National Park.
Thomas, D. P., 1983. A Limnological Survey of the Alligator Rivers Master of Science thesis, Northern Territory University, Dar-
Region I. Diatoms (bacillariophyceae) of the Region. Research win, Australia.
Report 3, Supervising Scientist for the Alligator Rivers Region, Vinson, M. R. and C. P. Hawkins, 2003. Broad-scale geographical
AGPS, Canberra, Australia, 160 pp. patterns in local stream insect genera richness. Ecography 26:
Tyler, M. J. and G. A. Crook, 1987. Frogs of the Magela Creek Sys- 751–767.
tem. Technical memorandum 19, Supervising Scientist for the Walden, D., R. van Dam R, M. Finlayson, M. Storrs, J. Lowry and
Alligator Rivers Region, AGPS, Canberra, Australia, 46 pp. D. Kriticos, 2004. A Risk Assessment of the Tropical Wetland
Tyler, M. J., G. A. Crook and M. Davies, 1983. Reproductive biol- Weed Mimosa pigra in Northern Australia. Supervising Scien-
ogy of the frogs of the Magela Creek system, Northern Territo- tist Report 177, Supervising Scientist, Darwin, Australia, 66
ry. Records of the South Australian Museum 18: 415–460. pp.
Unmack, P., 2001. Biogeography of Australian freshwater fishes. Welcomme, R. L., 1979. Fisheries ecology of floodplain rivers.
Journal of Biogeography 28: 1053–1089. Longman, London, U.K.
US EPA (U.S. Environmental Protection Agency), 1998. Guidelines Williams, W. D., 1980. Australian Freshwater Life. Macmillan,
for Ecological Risk Assessment. EPA/630/R-95/002F. Risk As- Melbourne, Australia.
sessment Forum, Washington, DC, USA. Wilson, B. A., J. Russell-Smith and R. Williams, 1996. Terrsetrial
van Dam, R. A., D. J. Walden and G. W. Begg, 2002a. A Preliminary vegetation. In: C. M. Finlayson and I. von Oertzen (eds.), Land-
Risk Assessment of Cane Toads in Kakadu National Park. Su- scape and Vegetation Ecology of the Kakadu Region, Northern
pervising Scientist Report 164, Supervising Scientist, Darwin, Australia, KluwerAcademic Publishers, Dordrecht, The Nether-
Australia, 92 pp. lands, pp. 57–79.
van Dam, R. A., C. L. Humphrey and P. Martin, 2002b. Mining in Woinarski, J. C. Z. and N. Gambold, 1992. Gradient analysis of a
the Alligator Rivers Region, northern Australia: Assessing po- tropical herpetofauna: distribution patterns of terrestrial reptiles
tential and actual impacts on ecosystem and human health. and amphibians in Stage III of Kakadu National Park, Australia.
Toxicology 181/182: 505–515. Wildlife Research, 19: 105–127.
van Dam, R. A., C. Camilleri, C. Turley and S. J. Markich, 2004. Woodroffe, C. D., J. Chappell, B. G. Thom and E. Wallensky, 1989.
Ecological risk assessment of tebuthiuron following application Depositional model of a macrotidal estuary and flood plain,
on Australian tropical wetlands. Environmental Toxicology and South Alligator River, northern Australia. Sediment. 36: 737–
Chemistry 10: 1069–1097. 756.