Fluvial hydrology is one of the key drivers of ecosystems and species within the Ramsar site, and is therefore fundamental to determining the site’s ecological character.
Description and Patterns in Variability
It is not meaningful to discuss patterns in variability of fluid hydrology pre and post Ramsar listing for several key reasons. Firstly, any one stream can intersect up to three of the Ramsar site Stages, so different pre-listing dates (and baselines) could apply to a single stream. Secondly, patterns in rainfall and hydrology vary greatly over a wide range of temporal scales, including decadal scale variations (see Section 3.5.1). Given the limited amount of long-term data and taking into account these complex longer term cycles, and that processes operating are greater than regional scales will ultimately control hydrology (that is, weather conditions), descriptions of these patterns before and after a time of listing is not meaningful. Patterns in hydrology for the overall data record are therefore briefly described.
The main drainage systems of the Kakadu National Park region begin with run-off from the plateau or start as springs at the foot of the Arnhem Land escarpment (Press et al. 1995b). Almost the entire region is drained into Van Diemen Gulf by four main rivers including the East Alligator, South Alligator, West Alligator and Wildman Rivers. The combined catchment area of the four major rivers is about 28 000 square kilometres (Cobb et al. 2007).
The four rivers consist of distinctly different stages along their passage to the sea (Press et al. 1995b). The first stage consists of the rivers’ upper reaches, which typically follow arrow and deep clefts in the sandstone. In the second river stage, the rivers broaden out into braided alluvial channels in the low sandy plains after leaving the plateau country and adjacent hills. This region covers most of Kakadu National Park. Four or more channels are typically separated by banks of loose sands and reformed in times of high flood. These channels divide and distribute their water widely over the third river stage, the expansive floodplains. The floodplains function as a large retarding basin, storing the water up over the wet season and gradually releasing it. The final stage of the rivers is the estuary, which is typically a relatively narrow tidal channel cutting through the floodplains. The South Alligator River has the longest estuarine section, extending approximately 50 kilometres from the sea. The estuarine channel banks are formed by silty levee banks often with a narrow ribbon of mangroves and monsoon rainforest trees in the otherwise mostly treeless floodplains.
Due to the strong seasonality in rainfall in the region, catchment runoff also follows a pronounced seasonal pattern with distinctive wetting up of the catchment in the early wet season followed by large flood flows between January and March (refer Figure 3 -32). Stream flows can vary to a large extent due to the wet season rainfall patterns. Typically, wet season stream flows comprise a series of peak flows superimposed on a base flow beginning about mid-December and ceasing about end of June (Press et al. 1995b). At the mouths of the two largest rivers, the South and East Alligator Rivers, the estimated annual flows are 2730 and 2560 million cubic metres, respectively. Estimated annual flows for Magela and Nourlangie Creek are 245 and 680 million cubic metres, respectively. However, these estimates may considerably underestimate potential discharges during extreme events (Cobb et al. 2007). Petty et al. (2008) reported that within the South Alligator River system, wet season flows vary from being contained within steep banks approximately twenty metres across within the South Alligator Valley near Gimbat, to a broad ‘sheet’ of water kilometres across flowing across the floodplains north of Yellow Water.
By the end of the dry season perennial creeks still flow albeit at a much reduced level, whilst annual creeks are generally dry but may still contain patches of stagnant water as well as billabongs within the creek bed (Petty et al. 2008). Major streams in the catchment cease to flow for several months of the year at the end of the dry season (refer Figure 3 -32; Table 3 -15). High evaporation during the dry season quickly reduces the water levels in pond waterbodies.
Table 3 15 Flow statistics for three representative gauging stations in KNP (source: Australian Natural Resources Atlas website, data from 1958 to 1999)
1974 to 1999
1960 to 1999
1976 to 1999
Catchment area of station (km2)
Mean annual flow (ML/yr)
1 499 379
1 013 968
Mean annual flow (mm)
Mean monthly flow (ML)
Mean monthly flow (mm)
Standard deviation (ML)
Minimum monthly flow (ML)
Maximum monthly flow (ML)
1 271 341
1 160 421
Coefficient of variation
East Alligator – 82019
South Alligator – 820112
Wildman – 819001
Figure 3 32 Monthly flow at three representative gauging stations in KNP (source: Australian Natural Resources Atlas website, data from 1958 to 1999)
3.5.2P2 - Fire Regimes
Reasons for Selection as ‘Critical’
Fire is one of the major forces that influences dynamics of the landscape, particularly with regard to regeneration processes of vegetation. As such, fire can have significant impacts on the landscape and is important for maintaining species and habitat diversity (Russell-Smith 1995).
Description and Patterns in Variability
The traditional fire regime practised by Bininj created a mosaic of unburnt, early and late burnt patches (Russell-Smith 1995). Fire regimes have been modified since the arrival of Europeans, and occurrences of intense late dry season fires are thought to have increased (Andersen et al. 1998, Vigilante and Bowman 2004). Fires experienced at inappropriate (too high or too low) frequencies, intensities or seasonality may lead to substantial changes in vegetation community composition and/or structure.
However, conservation managers now aim to mimic traditional patch burning to encourage optimum biodiversity (Director of National Parks 2007). The approach reduces the amount of grass fuel early in the dry season to assist with preventing late dry season fires covering large areas, thereby ensuring that communities and assets vulnerable to fire are protected (for example, intense late dry season fires result in death of Melaleuca).
Fire histories for the region are an important resource for park managers in determining the success of prescribed burning practises. As such, spatial and temporal patterns in fires within Kakadu National Park have been assessed at a whole of park scale.
Russell-Smith et al. (1997) and Gill et al. (2000) examined fire data for 1980 to 1996, with observations including the following:
Only four percent of Kakadu National Park was not subject to fire during this time period, with an average of approximately 45 percent of the park burnt each year.
Lowland savannah areas typically experience more widespread burning than plateau and floodplain areas, although burnt areas of the floodplain have significantly increased over this time period (refer Figure 3 -33).
An average of 25 percent of Kakadu National Park is burnt in the early dry season and 21 percent in the late dry season each year, with a pronounced shift over time from a fire regime dominated by late dry season fires up until the mid-1980s, to a fire regime dominated by early dry season fires.
Lowland savannah areas were typically burnt three out of every five years, while plateau areas were burnt zero to four times over the 15 years and floodplain areas were burnt zero to three times over the 15 years.
It is important to note that these figures may not necessarily represent an ideal fire regime, but do reflect fire regimes at the time of listing.
The introduction of some exotic pasture grasses such as gamba grass Andropogon gayanus, mission grass Pennisetum polystachion, olive hymenachne Hymenachne amplexicaulis and para grass Urochloa mutica has resulted in changes to fire regimes within some areas of the Ramsar site (Director of National Parks 2007), as these species are able to better colonise bare areas following late dry season fire as compared to most native species, and consequently late dry season fires are avoided in certain areas. Additionally, these grasses (especially gamba grass and mission grass) increase the fuel load of fires and result in hotter burns, which can lead to the loss of tree cover (for example, NRETAS 2009).
Figure 3 33 Areas burnt per year for Kakadu National Park (a) and various landscape types (b to d) (source: Gill et al. 2000)
3.5.3P3 – Breeding of Waterbirds
Reasons for Selection as ‘Critical’
Breeding is a critical life stage of species (as reflected in Criterion 2) that is essential in order to ensure the long-term persistence of populations that are fundamental to determining the site’s ecological character.
The most notable waterbird breeding colonies within the site are located within mangal communities of the major rivers and floodplain freshwater marshes. Breeding sites within mangroves are used by a variety of colonially nesting waterbirds (up to 12 species), though these multi-species colonies are typically dominated by egrets and herons (Chatto 2000). Chatto (2000) found that the breeding period for colonially nesting waterbirds (darters, cormorants, egrets, herons, spoonbills) extended throughout the year, generally beginning in November and ending in as late as October. The highest estimated annual usage of the five largest breeding colonies collectively amount to greater than 40 500 birds (Chatto 2000). Key sites for colonially breeding birds are associated with the downstream estuarine sections of both the East and South Alligator Rivers (southern and northern sides and within 15 kilometres of river mouth) (Chatto 2000). There are no seabird breeding colonies within the Ramsar site (Chatto 2001).
Floodplain wetlands are important for nesting waterbirds, although only five species breed in large numbers in the region (magpie geese Anseranas semipalmata, plumed whistling-duck Dendrocygna eytoni, wandering whistling-duck Dendrocygna arcuata, radjah shellduck Tadorna radjah and comb-crested jacana Irediparra gallinacea) (Bayliss and Yeomans 1990; Morton et al. 1991; Finlayson et al. 2006; Chatto 2006). Kakadu National Park’s importance as waterbird breeding habitat is highlighted by the significant breeding aggregations of magpie geese throughout the floodplains of the site (up to 27 percent of the Northern Territory breeding population), with the South Alligator floodplains regarded as the third most important area of nesting habitat after the Mary-Adelaide and Daly River floodplains (Bayliss and Yeomans 1990). Waterbirds nest throughout floodplain wetlands during the wet season, and whilst variations in breeding effort (and location of higher density nesting) have been recorded between years, this is most likely to reflect local variations in rainfall (Frith and Davies 1961; Bayliss and Yeomans 1990). Important sites for nesting waterbirds include:
South Alligator River upstream floodplains, including Boggy Plains (especially magpie geese) and Leichhardt’s Lagoon (especially wandering whistling-duck).
East Alligator River downstream floodplains, including the area around the junction of East Alligator River and Coopers Creek (especially radjah shelduck).
East Alligator River upstream floodplains, including Magela and Nourlangie Plains (especially for magpie geese).
Waterbirds are more abundant as water levels drop during the dry season, with the bulk of species increasing in numbers during the dry season (Morton et al. 1991). These birds largely migrate from wetlands located to the south of the Ramsar site, including species such as the grey teal Anas gracilis, pink-eared duck Malacorhynchus membranaceus, hardhead Aythya australisand purple swamphen Porphyrio porphyrio. Magpie geese dominate the influx of waterbirds, concentrating around permanent and semi-permanent waterbodies during the dry season, and dispersing to the floodplains following significant rains at the start of the wet season (Whitehead 1998).
Patterns in Variability
There are no available data to describe nesting densities and reproductive success, either before or after declaration of the Ramsar site. Furthermore, although summary data are available from Birds Australia, raw waterbird count data are not publicly available. Refer to Section 3.2.7 for a general description of patterns in waterbird abundance.
3.5.4P4 – Flatback Turtle Nesting
Reasons for Selection as ‘Critical’
Breeding is a critical life stage of species (as reflected in Criterion 2) that is essential in order to ensure the long-term persistence of populations that are fundamental to determining the site’s ecological character. The Ramsar site is considered critical in the context of maintaining the long-term viability of the flatback turtle Natator depressus, and underpins Critical Service 1 (see Section 3.6.1).
Flatback turtle Natator depressus is listed as vulnerable under the EPBC Act. This marine turtle species is generally found feeding in subtidal coastal waters, unlike conditions found within the site boundaries. Field Island is an important nesting area (Schäuble et al. 2006). In particular, the beaches on the western side of the island are key breeding grounds, comprising the majority of suitable nesting habitat in the region (Winderlich 1998, Schäuble et al. 2006). Field Island and surrounding waters form one of six major nesting sites in Australia, identified in the Australian Government Recovery Plan for Marine Turtles as ‘habitat critical to the survival of flatback turtles and is a key marine turtle-monitoring site within a national monitoring framework’ (Environment Australia 2003).
In general, a female flatback turtle displays a strong fidelity to her chosen nesting beach, with most females returning to the same beach within a nesting season and in successive nesting seasons (Limpus 2007). In the Northern Territory, nesting density reaches a peak in July, although some nesting may occur year-round (Fry in Limpus 2007). This dry season peak of nesting activity may be adaptive to protect the eggs from the high sand temperatures that occur in the wet season (Guinea in Limpus 2007).
Patterns in Variability
The key parameters describing this component are: (i) turtle nesting intensity indicators (number of nesting attempts per night or individuals nesting per survey night) and (ii) clutch size and clutch success. Turtle nesting patterns have been (and continue to be) monitored at Field Island annually. Schäuble et al. (2006) reported monitoring program results for these (and other) indicators on an annual basis between 1990 and 2001 (refer Figure 3 -34), which is post-listing (Stage II 1989). Note that there are no data describing breeding rates prior to site listing.
Figure 3 -34 shows that the mean number of nesting attempts has remained relatively consistent over time, whereas the maximum number of nesting attempts per night has tended to increase over time. The number of nesting individuals was variable over time, which may reflect differences in sampling effort, timing of surveys relative to the peak nesting period and changes in nesting intensity. The mean number of eggs per clutch was 52.4 (± 8.6 SD). Schäuble et al. (2006) also found that breeding success was high, with a mean clutch hatchling rate of 88 percent (± 17 percent SD) and an emergence success rate of 64 percent (± 32 SD).
This component is underpinned by the following processes:
coastal geomorphological and oceanographic processes that maintain the sandy beaches
connectivity between marine and dune habitats, and
absence of disturbance by humans and feral predators.
Figure 3 34 Mean (error bars = SD) and maximum number of nesting attempts per night, and numbers of nesting individuals per survey night recorded at Field Island (1990-2001) (source: Schäuble et al. 2006)