ABSTRACT

The water currents are important because they determine where water-borne particles will travel to. Particles of interest include natural ones: such as coral eggs and fish larvae; as well as pollutants such as dredge spoils, mud, pesticides and nutrients from farm runoff and sewage treatment plants. Cases of great practical importance are described, namely the impact of land use, reef recruitment of corals and fish larvae, and global change.

We have modelled the dominant hydrodynamics of the central Great Barrier Reef continental shelf, calibrating our models against our extensive tidal and current data. The currents (both speed and direction) show variability at all scales and are complex. These currents are also three-dimensional. This complexity in the currents causes patchiness in the transport of the water-borne particles.

The Burdekin River in flood generates a lens of brackish water floating over high salinity, oceanic water, and typically travels northward along the coast. Risk assessment analysis using the model results is needed to evaluate the impact of land use on the Great Barrier Reef. Further, the fate of pollutants, such as pesticides and heavy metals in river discharges is linked to chemical processes between these contaminants and the cohesive sediments. Enhanced flocculation around marine snow flocs and mangrove vegetation detritus lessens the direct impact on coral reefs.

Hydrodynamic models were merged with a huge historical data base collected over the last 10 years on coral and fish larvae around coral reefs. The results suggest that the recruitment of coral and fish larvae depends on advection from further upstream. On arrival, reef fish larvae aggregate in hydrodynamic shadow zones around coral reefs. Reef fish recruitment is controlled by both the water currents and the larval behaviour.

Introduction - large-scale oceanography of the Great Barrier Reef

The oceanography of the Great Barrier Reef (GBR) is complex due to its large expanse along the shelf waters of Northeast Australia(see a review in Wolanski, 1994). The movement of water though the reef matrix and coastal lagoon is mostly controlled by the combined forcing of the tides, the wind, the Coral Sea currents and occasional river floods. Long-term field studies of the water flows in this region show fluctuations over short and long time scales. Locally, the currents are observed to be highly influenced by the steering and blocking effects of the reefs themselves. Numerical models, using spatial scales which incorporate the influence of the topography of the reefs, have proven to be successful in describing the complex water flows of the region (King, 1995).

This paper examines the use of oceanography in multi-disciplinary studies of important environmental problems, namely the effects of land use, reef recruitment and global change. The paper also demonstrates the usefulness of data and model visualisation as a means to effect better communication between scientists and resource managers.

Methods - data and model visualisation: technological advances

As part of the AIMS-IBM project "Coral Reefs and Mangroves: Modeling and Management" at AIMS, we have acquired significant computational and visualisation resources to further scientific research. Visualisation is carried out routinely using IBM Data Explorer - DX. Visualisation offers a valuable platform for marine scientists for:

• modelling flows in complex bathymetry (coral reefs and mangroves)
• merging and interpreting physical-chemical-biological data sets with the goal of demonstrating the key processes
• by the above technique enable an interaction between oceanographers, chemists and biologists.
• presentation of the output to marine resources managers.

Computer visualisation offers a way for physicists, chemists and biologists to merge their data sets from point measurements of currents, concentration of larvae of shrimp, fish and corals, and chemical matters (eg hydrocarbons and mud). Both the merging of the physical and biological data sets and the animation of the data are necessary because the data are patchy in time and in space and practically un-assailable using standard statistical techniques. The use of visualisation has led to breakthroughs in the scientific interpretation of data showing patchiness and great spatial and temporal variability, namely in being able to understand which are the dominant physical, chemical and biological processes that must be taken into account in models. For an environmental manager, visualisation makes modelling a practical option because with it model predictions are easy to obtain, customable and simple to interpret.

Turbid river plumes

The Burdekin River drains an area of 130000 km² of North Queensland and has the largest recorded mean annual flow for the Queensland Coast, 9.7 x 10⁹ m³. The river mouth is located between Cape Upstart to the south and Cape Bowling Green to the north (see figure 1). The runoff is highly variable, limited to a few floods (1-3 per year) from January to March.

Our work involved setting up and verifying a 3-d hydrodynamic model incorporating the plume dynamics of the Burdekin River in flood. We have verified the model against the historical data of Wolanski and Van Senden (1983) for the 1981 flood (see figure 1), the only flood event for which an extensive river plume data set exists. The peak discharge of the flood in 1981 was over 12,000 tonne of water per second, and the river delivered a total of 7.5 billion tonne of freshwater into the GBR lagoon over a 3 week period. Figure 1 shows a snapshot of the predicted plume forcing the model with actual river discharge and wind data (King and Spagnol, 1996). Given that the data were collected over a 2 day period, there is a very good agreement between the observed and predicted salinity distribution (Figure 1). Comparisons at other times also show that the model is consistently reproducing the observations (see the animations).

Figure 1. Predicted surface salinity distribution in the Great Barrier Reef during the Burdekin River flood on January 27, 1981. The inset shows the observed distribution for comparison.

Figure 2. Comparison of surface salinity distribution using identical river discharges, but 3 typical wind patterns: no wind, 10 knot SE wind and 10 knot NE wind.

Initial sensitivity analysis on the model runs show that the main driving influence on the fate of the plume water is the volume of the river discharge and the local wind forcing. Thus each year, one would expect different plume trajectories depending on the time-varying nature of both the wind and the rainfall/catchment (Figure 2). Ultimately a risk assessment approach is required using 25 years of available wind and river discharge data, to quantify the likely impacts of the river floods on the GBR.

Studies of the intrusion of the Fly River plume into Torres Strait waters (Ayukai and Wolanski, 1996) provide an insight into the likely impacts of river plumes on coral reefs in the Great Barrier Reef itself. During flow reversal process, the mixing between 'old' and 'new' plume waters result in the formation of coherent, low salinity patches. Silt is not carried by the river plume, only the clay is; however pollutants are mostly attached to the clay particles. The fate of this clay is strongly dependent on the interaction between physics and biology (Figure 3). In the estuary the fine sediment is flocculated forming small (mean diameter about 30 m) flocs of fine sediment loosely bound by organic matter and containing very few plankton and their remains. Progressive dilution of riverine sediment with reefal, calcareous sediment west of Torres Strait changed the structure of the flocs. In coastal waters when a plankton bloom is present in the Fly River plume, the plankton removes the nutrients from river runoff and created huge (diameter typically several hundreds of m) marine snow flocs. The flocs were composed of colonies of dead plankton, faecal pellets and macroscopic aggregates of apparently biological origin. The riverine fine sediments were found attached to or embedded in these flocs. This high biological activity apparently functioned as a barrier limiting the offshore transport of fine sediment. This effect is likely to occur also for the GBR though it has not been investigated before. A similar effect we found recently in Hinchinbrook Channel (Figure 3) where biologically enhanced flocculation occurs on marine snow at neap tides and mangrove vegetation detritus at spring tides. For the Fly River plume some metals, eg particulate Fe, may be a conservative that can be used to study the fate of riverine sediment carried by river plumes.

Hence the impact of land-use on the Great Barrier Reef is related to the transient fate of the flood plume waters and its suspended sediment and is also linked to estuarine filtering processes. In particular, the fate of pollutants such as pesticides and heavy metals is linked to chemical processes between these contaminants and the cohesive sediment, and to the sediment particle sorting by size.

Reef recruitment

Flows around coral reefs are complex (figure 4) and include eddies, convergence and divergence zones, tidal jets and stagnation zones. These flows are three-dimensional and are biologically important because they determine the aggregation zones of plankton and coral eggs at sea. By comparing with data from the field and from models, we have discovered that the vertical circulation is controlled by viscous walls formed by the intense turbulence in the vortex sheets shed by separation points. This is a sub-grid scale phenomenon that needs to be parameterised explicitly in 3-D models (Wolanski et al., 1996). These processes vary in time and in space.

Figure 3. Microphotographs from coastal waters near Hinchinbrook Island showing: (top) macro-floc from aggregation around plankton detritus and (bottom) macro-floc from aggregation around vegetation detritus. The flocs are about 300 m in diameter.

Figure 4. Predicted current (arrows) distribution around Bowden Reef, Great Barrier Reef, at 1530 h on December 11, 1987, together with the predicted distribution (plume) of coral larvae concentration following mass coral spawning. The white pins represent collection sites and the colour bars over the white pins represent observed coral larvae concentrations at the surface (right) and at 6 m depth (left).

In collaboration with biologists, principally Drs. J. Oliver and P. Doherty, we have used DX to explore a huge historical data base collected over the last 10 years on coral and fish larvae around coral reefs. We have used these data to calibrate numerical models of the flushing of coral larvae from the reefs where they were spawned (figure 4). We found that reef recruitment depends on advection of coral larvae from further upstream (see the animations).

The same methodology for modelling the fate of coral larvae was applied to fish larvae on coral reefs. Visualising the data revealed that the fish larvae were not initially resident around a particular coral reef, instead they arrived from upstream with the water currents. On arrival, reef fish larvae aggregated in hydrodynamic shadow zones around coral reefs. We were successful in modelling reef recruitment of fish taking into account both the water currents and the larval fish behaviour (Figure 5). This has important management implications. Indeed in most marine systems, pelagic marine fish larvae are at the mercy of fluctuating, dispersing water currents, with little relationship resulting between the adult stock population and larvae recruitment, the majority of the larvae die at sea. However for the Great Barrier Reef, we demonstrated that the opposite is true, namely a high level of coral reef fish recruitment occurs in hydrodynamic shadow zones. We also show that recruitment is controlled by both the water currents and the larval behaviour. The strong link between recruitment and adult stock suggest that fish stocks in the GBR may be particularly susceptible to over-fishing.

Figure 5. Predicted distribution of damselfish around Bowden Reef taking into account currents and fish behaviour.



Implications of Global change

Assuming that global change will be so rapid that the biological system will be faced with a sudden change in the physical forcing, we have developed three scenarios and proposed models of possible oceanographic effects of global change on coral reefs. We predicted that sheltered lagoonal reefs will suffer from an increased residence time of pollutants and a decrease of the upwelling of oceanic nutrients. Also, we predict that changes in wave dynamics over a leeward coral reef following a rise in sea level will modify coral zonation. Finally we predict that global change may increase the recruitment of fish and coral larvae as can be seen by browsing on the Internet at URL address http://ibm590.aims.gov.au.



ACKNOWLEDGMENTS

This study was supported by the Australian Institute of Marine Science, the IBM International Foundation, the CRC-Reef Research and KEPCO-KEEC.

References

Ayukai, T. and E. Wolanski (1996). The importance of biologically mediated removal of fine sediment from the Fly River plume, Papua New Guinea. Estuarine, Coastal & Shelf Science, in press.

King, B. (1995). Hydrodynamic complexities of tropical coastal waters: a study through the application of numerical models. Ph.D. thesis, Department of Physics, James Cook University, Townsville, 130 pp.

King, B. and S. Spagnol (1996). Simulating the Burdekin River in flood. In: CRC Reef Research News, Vol. 3(3), 4.

Wolanski, E. (1994). Physical oceanographic processes of the Great Barrier Reef. CRC Press (Florida), Marine Science Series, pp. 194.

Wolanski, E. and D. Van Senden (1983). Mixing of Burdekin River Flood Waters in the Great Barrier Reef, Australian Journal of Marine and Freshwater Research, Vol. 34, pp 49-63.

Wolanski, E., T. Asaeda, A. Tanaka and E. Deleersnijder (1996). Three-dimensional island wakes in the field, laboratory and numerical models. Continental Shelf Research, 16, 1437-1452.

Last updated: 14 January 1997
All contents copyright (C) 1997, AIMS.All rights reserved.