-AIMS home
-About AIMS
-Research
-Facilities
-News
-Search
-Site map
-Site index
-Topics index









TROPICS (Tropical River-Ocean Processes In Coastal Settings)

SUMMARY OF ABSTRACTS, OS11/12 TROPICS SPECIAL SESSION
AGU Ocean Sciences, 9 Feb 98, San Diego, California
TC Golden West Room
CONVENORS: G. J. Brunskill and C. A. Nittrouer

Monday Morning Session (OS11I)

0830 W E Dietrich, J Chappell, G M Day, G Parker

Holocene Evolution of the Middle Fly River, Papua New Guinea

The Fly River drains the steep, rapidly eroding Southern Fold Mountains of Papua New Guinea and crosses a low relief plain to the Gulf of Papua. At its mouth, the river drains 75,000 km2, only about 30% of which is in the steep up lands. The middle Fly lies entirely in the low relief plain where it extends about 450 km from upstream of the Ok Tedi to the junction with the Strickland River. Along this reach the slope declines from about 6.6 x 10^{-5} to 2 X 10^{-5}, the median size declines from about 0.3 mm to 0.1 mm, and the duration of flooding greatly increases, contributing to the downstream shift from a floodplain dominated by forest to one covered with swamp grass. The floodplain is etched with scroll bars and oxbows and blanketed with back swamp deposits. Currently, however, channel migration rates are slow in the forested reach (about 0.0045 channel widths per year) and nearly zero in the swamp grass reach. Preserved along the eastern margin of the floodplain is a Pleistocene channel path. Radiocarbon dating accomplished at the ANU laboratory has documented several important features of the Fly River's response to the Holocene rise in sea level: 1) the earliest Holocene channel appears around 7000 to 8000 calendar years B.P.; 2) rates of sedimentation are high (several mm/year)in the early Holocene; 3) as the rate of sea level rise slowed significantly so did the floodplain deposition rate; 4) deposition rate since about 7000 B.P. varies strongly with environment, ranging from less than 0.6 mm/yr in backswamp area, to as high as 25 mm/year on levees. Deepest drilling was about 15 m where early Holocene sediments were encountered. Others have reported Pleistocene sediments exposed in deep pools of the Fly. Taken together these observations suggest that 1) extensive estuarine conditions did not develop with post glaciation sea level rise; 2) floodplain deposition rates have been modest and in the late Holocene were low; 3) deep Holocene fills may be absent; 4) since the mid-Holocene the delivery of sediment through the middle Fly River floodplain system toward the Gulf of Papua has been relatively efficient; and 5) the lower reach may have transformed to a swamp grass with only minor channel migration fairly recently.

0850 G M Day, W E Dietrich

Sediment Deposition Across the Middle Fly River Floodplain: Implications for Sediment Delivery to the Gulf of Papua

Floodplain deposition rates and their controls are generally poorly known in modern river systems. As a consequence of mining in the headwaters of the Fly River copper-enriched sediment has been released into the system and its presence is easily detected against background pristine values, hence this sediment can be used as tracer documenting rates and patterns of sediment deposition across the 3300 km2 Fly River floodplain. Sediment delivery to Ok Tedi, a tributary of the Fly River, Papua New Guinea, has increased since 1984 from 5 Mt/a to a current rate of ca. 110 Mt/a from the Ok Tedi copper mine. Approximately 74 Mt/a of mine derived sediment (30 Mt/a of tailings, 44 Mt/a of waste rock) and 32 Mt/a sediment from valley wall erosion due to mining is delivered to the fluvial system. Of the 106 Mt/a of mine-derived material, ca. 40 Mt/a is transported through the Ok Tedi to the Middle Fly River.

We have used shallow cores, collected annually since 1990, from up to 600 locations and a network of water level recorders, rain gauges and sediment traps to trace the temporal and spatial spread of copper-enriched sediment from the river across the floodplain. Based on this extensive field program, current estimates indicate that approximately 3\% of the load entering the Middle Fly River is stored within the floodplain. The spatial distribution of mine derived sediment and the study of floodplain hydrology and sediment transport processes show deposition rates are highest near the main channel banks, but overall rates of deposition is significantly higher in the forested, better drained upper middle Fly (about 1 to 2 mm/yr) than in the chronically flooded swamp grass dominated lower middle Fly (less than 0.5 mm/yr). Furthermore, significant spread of sediment away from the river occurs up tributaries and small channels connecting the main stem with off-river water bodies and backswamp areas (tie channels). We have hypothesized that the chronic flooding in the lower swamp grass reach retards the spread of sediment-rich flood water across the floodplain, and that the tributaries and tie channels are important avenues for the dispersion of sediment throughout the floodplain environment. Floodplain deposition may also be strongly affected by larger scale climatic factors. Droughts associated with El Nino concentrate draining floodplain flow through numerous transient channels, which cut through the levee several meters below bank full as river stage drops. These then become avenues of flow and sediment to the floodplain when the river subsequently floods. Despite the large increase in the delivery of sediment to the Middle Fly River system, current hydrodynamic controls on sediment transport across the floodplain impose a significant constraint on floodplain deposition and storage. Under the current regime, the floodplain environment is not a significant trap of sediment to the Gulf of Papua.

0910 M. Bohlen and J.D. Milliman

Dissolved and Sediment Fluxes from Small New Guinean Rivers

Sediment yield (load per unit area of drainage basin) from mountainous rivers increases with decreasing basin size, mostly the result of less sediment storage and the greater impact of events (particularly heavy precipitation and landslides) in smaller watersheds. The relation between dissolved flux and basin area, however, is less obvious. Scattered data, some of which are questionable, suggest that dissolved yield increases with decreasing basin but falls in basins less than about 10,000 sq km in area, perhaps the result of the short residency time of physically eroded sediment. To investigate whether dissolved yield does fall in very small drainage basins, we sampled nine small rivers in northern New Guinea, most of which range in basin area between 150 and 500 sq km in area; only the Gogol River has a watershed greater than 1000 sq km. Rivers were sampled in late May 1997 (as part of the TROPICS program) during falling stages of river flow. Dissolved carbon (1 mg/L) and particulate organic carbon were generally very low, and geochemical proxies suggest that the residence times for both DOC and POC may be less than one day. Preliminary analyses indicate that dissolved concentrations are resemble those of much larger rivers. If these data are correct (they were being reanalyzed as this abstract was being written), then groundwater, with a correspondingly long residence time, may play a much greater role in controlling dissolved concentrations in small rivers than previously envisioned. Alternatively, high dissolved contents may indicate rapid rates of chemical erosion, not an unlikely situation in such a tropical environment.

0930 Coffee Break

0950 G Cresswell and L Pender

The Sepik River outflow into the Bismarck Sea, May 1997

The data for this work came from the research vessel Franklin, surface drifters, a moored ADCP and current meter, and a Synthetic Aperture Radar scene from RADARSAT. A 150 m deep canyon starts just out from the mouth of the river and continues 6 miles down the sloping topography at least to the 800 m isobath. It appears to have been carved by the river outflow. The surface plume was only about 1-2 m thick and it slid across the marine water for 10 miles before being turned to the west by the underlying westward-flowing New Guinea Coastal Current (NGCC), which had speeds up to 0.8 m/s. The flow in the plume converged towards the curved offshore front revealed by the SAR scene. The scene also showed the NGCC to produce wakes behind the offshore volcanic islands. The waters within the canyon were comprised of stacks of as many as ten mixed layers with suspended sediment loads peaking at the interfaces between the layers. This structure could be followed for several miles out along the canyon, although the sediment layers became progressively depleted suggesting that the suspended sediment was being mixed into the ocean interior, was falling to the sea floor, or had moved horizontally away from the canyon axis.

1010 G C Kineke and R W Sternberg

Suspended Sediment Dispersal from the Sepik River, Papua New Guinea

Measurements of suspended sediments and water properties have been made in the vicinity of the Sepik River mouth in April 1996 and May 1997. The Sepik River empties directly into a steep submarine canyon that cuts across the narrow shelf from a minimal depth of 6 m at a shallow bar in the river mouth to a depth of 800 m over a distance of 15 km. 1996 observations suggest that suspended sediment delivered by the river follows distinct dispersal pathways via a vertically divergent plume and it is likely that less than half is by way of the muddy surface plume. The thin surface plume exiting the river mouth (less than 2 m thick) contains suspended-sediment concentrations (SSCs) decreasing from approximately 200 mg/l in freshwater to background SSC of 5 mg/l or less in waters 31 ppt salinity or greater. Just seaward of the shallow river mouth bar, a dense near-bottom plume forms as evidenced by increasing SSC and an anomalous low salinity signal near the bottom, which implies trapping of sediments at the plume liftoff point. Measured near-bottom SSCs are greater than or comparable to those in the surface plume (100s of mg/l). 1997 observations down the axis of the canyon, in water depths of several hundred meters, revealed distinct layers of increased turbidity several meters thick sandwiched between clearer water. Thus most of the sediment delivered by the river appears to be dispersed via hyperpycnal (negatively buoyant) plumes. Dense nearbed suspensions formed just seaward of the bar continue down the steep slope as a hyperpycnal flow, or deposit close to the bar temporarily and then flow down the canyon as a turbidity current. These flows mix with ambient water and spread out at mid-depths for miles on different density surfaces. The dispersal mechanism of the Sepik River may serve as an analogue to other rivers during extreme high discharge events and serve to explain some of the discrepancies between the apparent pathways of muddy surface plumes and bottom sediment deposits.

1030 L K Ayliffe, M.I. Bird, P.J. Isdale, M.K. Gagan, M. Nongkas, M T McCulloch

Flooding of the Sepik River and El Niño Events: a 20 Year Record From Oxygen-isotopes, Sr/Ca and Fluorescence of a Porites Coral.

The Sepik River is the largest River in Papua New Guinea (PNG) discharging 240 km3 annually into the Bismark Sea. Meteorological records suggest that periods of drought in PNG are closely linked to El Niño events. Periods of low discharge from the Sepik River should therefore be related to El Niño events, but observational records are sparse, and limited to the last three decades. Here we explore the potential of corals growing in ocean waters adjacent to the outlet of the Sepik River as recorders of flood events using geochemical signatures preserved in their aragonitic skeletons. A 35cm high Porites coral from Blup Blup Island (lat 03o 30.9' S; long 144o 35.47' E), 30 km from the mouth of the Sepik River and lying in the path of the flood plume, was collected from 3.2m water depth in September, 1996. 1133 micro-samples were taken at 200 µm thickness down the central growth axis of the coral and Sr/Ca and oxygen-isotopes determined. Fluorescence under UV light (360nm) was determined on an adjacent transect. A chronology was established for the upper section of the coral (1985-1996) by comparing the satellite sea-surface temperatures with temperatures derived from measurements of coral Sr/Ca (using T (oC) = 168.2 - [15,674(Sr/Ca)atomic], based on observations made for near-shore corals in the Great Barrier Reef, Indonesia and north-western Australia). Sr/Ca-based temperatures over the ~ 20 year long coral record range from 27oC to 30oC and is consistent with satellite data at the site. Oxygen-isotope residuals (oxygen-isotope data corrected for a temperature effect) range from -0.35 to -1.5 per mil and are found to correlate with the intensity of coral fluorescence and with the magnitude of river discharge. In accord with the limited observational data, these proxies suggest that discharge of the Sepik River was a regular quasi-biannual phenomenon during the late 1970's, mid 1980's and early 1990's, and was at its lowest during the strong El Niño events of 1982/83, 1987 and 1991-1994. These results highlight the potential of tropical corals in such a setting to provide information on the response of the terrestrial hydrological cycle to changes in SST and salinity using historical El Niño events to define an accurate chronology.

1050 K Burns and T Ayukai

DOC and fluorescent tracer studies in Sepik River water discharge to the Bismark Sea 1996-97

As part of an integrated study of Tropical River-Ocean Processes, we present a preliminary description of river mixing patterns during high and low discharge periods for the Sepik River on the north coast of New Guinea. In both periods, the river plume entered the sea as a narrow dark wedge over the coastal seawater. The freshwater layer was less than 1m thick for distances up to 100 km, and flowed in directions dependent on the prevailing wind and currents. By tracing the plume with DOC, salinity and fluorescence measurements, we observed that in both flow regimes, DOC and its fluorescent components underwent conservative mixing throughout the salinity gradient from the river to the sea. During the extreme climatically dry conditions of August 1997, we observed salt water intrusion on the river bottom to at least 32 km from the mouth. In April 1996, DOC and TOC ranged from ~3.5 mg/l in the river to ~1 mg/l in seawater. Fluorescence units at 340 nm excitation / 440 nm emission ranged from 168 units at 0 salinity to 4.5 units at 35 o/oo. These results for the Sepik contrast with those from the Fly River which showed processes of scavenging of DOC in the high turbidity zones of the estuary. The Sepik observations will be presented in the light of scavenging and sorption-desorption models for the tropical estuaries of New Guinea. They are part of the carbon mass balance models currently under construction for the region.

1110 R. Szymczak and M.Zaw

Dynamic Biogeochemical Behaviour of Trace Elements Elucidated by 234Th/238U Inventories in Tropical Coastal Waters of Northern Australia and Papua New Guinea

Interactions of chemical, biological and physical processes regulate the efficiency of elemental scavenging in the tropical coastal zone and determine the offshore flux of terrestrially derived constituents. Particulate oxyhydroxides of the trace metals iron and manganese are the major chemical scavenging phases and along with biogenic particles form a conveyor belt for the removal of elements from surface waters to coastal sediments and/or the deep sea. Light-induced reactions, may however, significantly influence the overall efficiency of elemental scavenging via photo-retardation of particle formation and particle aggregation processes as well as photo-reduction, hence dissolution, of ferromanganous particle surfaces.

In the Gulf of Papua and adjacent tropical coastal waters the behaviour of several dissolved, but particle reactive trace elements, were observed to be a function of the residence times of particles, deduced by 234-thorium/238-uranium inventories to vary from less than 1 to more than 10 days. Longer residence times of particles in photic surface waters allow saturation of adsorption sites and greater illumination of particles, resulting in higher dissolved concentrations as the rate of input (constant) exceeds removal (photoretarded).

1130 D J Sinclair, M T McCulloch

The Coral Record of Terrestrial Fluxes into the Great Barrier Reef: Analyses of Trace Elements in Inshore Corals by Laser-Ablation ICP-MS.

In order to investigate the coral records of river flood geochemistry and coastal geochemical cycles, a method has been developed for simultaneous, high-resolution, multi-element analysis of corals by ultraviolet (l = 193 nm) laser-ablation ICP-MS. The laser beam, masked to produce a 600 mm by 50 mm slit, is scanned over the surface of the coral to generate a continuous profile of trace-element variation with distance. Sensitivity drift is controlled by pre-conditioning the ICP-MS, and bracketing each 50 mm section of coral by analyses of a synthetic CaSiO3 standard glass. The method is fully quantitative for B, Mg, Sr, Ba, and U, and semi-quantitative for Mn, Y, La, Ce, Pb, Ni, Cu, and Zn.

Three near-shore Porites sp. corals from different locations along the northern east-coast of Australia were analyzed. The environmental settings differ for each coral, but all experience periodic fresh-water discharge from nearby rivers. Ten centimetre lengths of coral, spanning the late 1960s and early 1970s, were analyzed in duplicate. Fluorescence and density were obtained from scanned images.

Trace-element profiles within a coral are well reproduced in parallel tracks. B/Ca, Mg/Ca, Sr/Ca, and U/Ca display seasonal temperature cycles, however subtle environmental or physiological factors may be perturbing this signal. Sharp troughs in winter, and broader peaks in summer indicate that linear extension may decrease in colder conditions for some corals, and this could relate to the observed formation of high density bands in winter. Y/Ca, La/Ca, and Ce/Ca co-vary with fluorescence, and therefore appear to be associated with river-floods. Mn/Ca seems to show flood spikes superimposed upon seasonal temperature fluctuations. Ba/Ca displays a clear correlation with river-floods in two of the corals, although there is evidence for a 'wash-out' effect where peaks following dry spells are larger than peaks occurring during wetter times. At one location Ba/Ca does not correlate with floods, and instead peaks in late winter to spring. Weak fluorescence indicates that river floods have largely dissipated before reaching this site, and Ba/Ca spikes are therefore postulated to correlate with phytoplankton blooms, or sediment resuspension.

Monday Afternoon Session (OS12H)

1330 Woolfe, K. & Larcombe, P.

Contrasting styles of sedimentation on tropical shelves of NE Australia and Papua New Guinea

Sediment transport across the continental shelf is influenced by a number of factors including the width and gradient of the shelf, the wind/wave climate, presence or absence of barrier systems and by the volume of sediment entering the coast from rivers. We present preliminary results of sediment distribution surveys of four adjacent but contrasting shelves, to illustrate and discuss the local and regional controls upon modern and Holocene sedimentation patterns.

The central Great Barrier Reef shelf is a wide (ca. 80 km) low-gradient shelf characterized by limited sediment supply (c. 30Mt/y), a wind/wave climate favourable for the formation of a turbid coastal boundary layer, and extensive barrier reef tract. In this regime sediment is largely retained in a nearshore sediment wedge and relatively little sediment escapes to the mid-shelf or beyond.

The western Cape York Peninsula (Gulf of Carpentaria) shelf, is a broad low-gradient shelf dominated by offshore winds and monsoonal cyclones. This shelf receives sediment at a rate of c. 30 Mt/y. Waves generated by the NW monsoon approach the coastline at high angles, dispersing sediment seaward. An extensive thin veneer of sediment extends across the gulf.

The Gulf of Papua displays a shelf of variable width and gradient and a large sediment supply (c. 370 Mt/y). Large long-period swell waves prevail for much of the year (generated by the SE trade winds) and shelf clinoforms locally extend to the shelf break. However, wind- and wave-driven currents keep the area immediately seaward of the Fly River mouth largely sediment free.

The north coast of PNG receives large volumes of sediment (c. 200 Mt/y), is microtidal and situated within the doldrums. The shelf is extremely narrow and clinoforms extend to the shelf break. Moreover, extensive sediment bypassing occurs through the Sepik Canyon.

1345 J D Milliman

Fate of sediment from rivers draining northern and southern New Guinea

Rivers draining New Guinea discharge more than 1 billion tons of suspended sediment to the sea annually. Although sufficient data are lacking, the marked morphological difference between the very narrow (active) margin along the north coast and the broad (passive) margin off the south coast suggests different processes and fates of sediment discharged from northward- and southward-flowing rivers.

High-resolution seismic data (100-electrode sparker) were collected off the north and southern margins of Papua New Guinea on the Australian ship R/V Franklin in June 1997. To the north, most of the sediment discharged from the Ramu and Sepik rivers bypasses the narrow (2-4 km) shelf and is deposited in the adjacent deep sea. The only major shelf depocenter is a sediment wedge that buried a previous barrier reef, presumably after the Sepik and Ramu had filled their coastal lagoons less than 3500 years ago.

In contrast, the Gulf of Papua off the southern coast receives sediment from some of the largest rivers in the southern hemisphere (e.g., Fly, Purari, Kikori), which collectively discharge more than 300 million tons of sediment annually onto a broad (50-150 km) shelf. Seismic profiles around the upper gulf show a more or less continuous wedge of modern sediment prograding over relict (albeit muddy) strata; clinoform-like morphology is seen in coastal areas, although biogenic gas locally prevents seismic penetration. Of particular interest is the thick mantle of Holocene(?) sediment that lies conformably over the uneven topography of the Papua Trough. Although our data are scattered and preliminary, we find no evidence of the mass wasting that one normally associates with large-scale sediment input to continental margins; sediment cover appears to be as great on topographic highs as in lows. Whether the lack of mass wasting results from sediment bypassing or reflects an incomplete data base remains to be seen.

1400 S.A. Kuehl and T.M. Dellapenna

Control of Drainage Basin Characteristics on Application of Cs-137 Geochronology: The Sepik River Margin

Sedimentological and geochemical studies of grab and piston core samples from the shelf and slope off the Sepik River provide clues regarding drainage basin characteristics and sediment dispersal for this small, high-relief river system discharging into an active margin setting. Bomb-produced Cs-137, injected into the stratosphere in massive quantities in the '50's and '60's, is generally below detection limits in marine surface sediments from the study area. This observation was unexpected given that geochronological studies using Pb-210 suggest significant recent sediment accumulation on the continental slope off the Sepik. Cs-137 is rapidly and nearly irreversibly exchanged with K in illite bearing soils, and is commonly found in measurable quantities in sediments delivered to the oceans by rivers. One possible explanation for the absence of Cs-137 off the Sepik is that the residence times for soils in the drainage basin is extremely low, such that much of the fallout Cs-137 from the '50's and '60's has previously been flushed from the basin. Alternatively, extremely high denudation rates for this high-relief basin could result in dilution of the fallout signal below detectable limits. In either case, observations from this study suggest that Cs-137 is not a useful time marker in this, and perhaps other similar marine environments which receive input from small, mountainous drainage basins.

1415 J P Walsh & C.A. Nittrouer

Sedimentation processes in the Gulf of Papua and the Fly River mouth, Papua New Guinea

The wet tropical regions of the Earth contribute a substantial amount of sediment (>50%) to the World oceans. The island of New Guinea alone contributes approximately 1.74 billion tons of sediment per year, a consequence of high rainfall (up to 13 m/yr) and precipitous mountains (exceeding 4000 m). Annually, an estimated 460 million tons of sediment enter the Gulf of Papua from several large rivers, most importantly the Fly and Purari. Despite the obvious global significance, this region has received relatively little sedimentological attention.

Research cruises were conducted in conjunction with the TROPICS (Tropical River-Ocean Processes in Coastal Settings) Program in 1997. Sedimentological and radiochemical observations were completed on Kasten cores, vibracores and grab samples obtained from the Gulf of Papua and the Fly River mouth. This preliminary geological investigation documents accumulation rates exceeding 1.5 cm/yr in some portions of the dispersal system, which extends from mangrove forests, across shelf clinoforms, to the continental slope. Eastward coastal transport in the Gulf of Papua carries sediment from numerous rivers to form an amalgamated shelf deposit. Additional sediment is transported off the continental shelf into the Pandora Trough on the eastern side of the Gulf.

The dispersal system shows a distinct transition in sediment character: from silty and sandy lithogenic sediments in coastal areas, to muddy clinoforms on the mid-shelf, to mixed lithogenic and relict carbonate sands on the outer shelf and the continental slope. In addition, sedimentary structure displays notable changes. Distinct silt-sand laminations and mangrove bioturbation are most apparent at the coast; on the inner and mid-shelf, partially deformed, fine laminations are common. These sediments progressively grade into a more biologically mixed, homogenous sedimentary structure offshore. Sediment budgets are complex due to the numerous fluvial sources and the presence of biogenic sediments; however, such budget estimates will help provide an understanding of the dominant transfer processes operating in this globally important environment.

1430 M A Allison, M T Lee, A S Ogston, R C Aller

Mudbank Formation along the Northeast Coast of South America

High-resolution (CHIRP) digital seismic profiles, sediment cores, and water column profiles were taken in March 1996 to examine the shoreface and inner shelf of the northern Amapa coast of Brazil. This region of slowly accretionary mud coast separates two areas of distinct morphology in the 1,600-km-long coast downdrift of the Amazon River mouth. Adjacent to the river mouth (<250 km) the coast is erosional. Further downdrift in the Guianas, a series of shore-attached, Amazon-derived mudbanks 10-30 km long migrate downdrift. A survey grid (1,400 km) of seismic profiles on the northern Amapa provides a datum for measuring the thickness of overlying modern mud. Two mudbanks were discovered opposite the shoreline of Cabo Orange separated by about 30 km. Each mudbank is about 10 km long and extends to the 20 m isobath. The more downdrift (northern) mudbank is larger, with a thickness of up to 3.5 m on the most inshore (5m) seismic profile. Both mudbanks exhibit a low-reflectivity northern side suggestive of accretion and overall downdrift migration. Cores and water column profiles taken from small boats in nearshore (<5 m) areas of the southern mudbank show it is shore-attached and has increased near bottom turbidites and lower bulk seabed densities relative to adjacent areas. The size, geometry, orientation and composition of these mudbanks suggests they are less well-developed examples of those seen further downdrift. This indicates that northern Amapap is the genesis point for the mudbank system. Comparison of seismic profiles of the northern edge of the Amazon subaqueous delta in 1983 and 1996, indicates these features do not originate from the delta. Their orientations suggests that they are shed at decadal intervals off Cabo Cassipore; a large, spit-like mudcape that has an accretionary clinoform tip. Amazon sediment is trapped here on seasonal timescales because of a strong frontal zone at 5-8 m water depth. This "turbidity maximum" feeds sediment to the nearshore zone around the cape tip by tidal excursions and solitary-wave transport.

1445 M T Lee, M A Allison, F Baltzer

A Remote-Sensing/Geochronology Study of Coastal Morphology associated with Episodic Amazon Sediment Supply

The shoreline of the northeast coast of South America is fronted by extensive underconsolidated mudflats and a 1-10 km-wide belt of mangrove swamps created from Amazon River sediment. Shoreline accretion/erosion in this area is generated by 1) the downdrift migration of 10-30-km-long, shore-attached mudbanks and 2) the concentration of sediment at river mouths forming spit-like mudcapes. Three representative regions were selected- western French Guiana (Sinnamary), eastern French Guiana (Acoupa), and northern Amnapa, Brazil (Cassipore) and examined with radioisotopes (Pb-210, Th-234, Be-7, and Cs-137) and remote sensing techniques (geo-referenced aerial and satellite mosaics) in order to examine shoreline evolution and cross- and along-shore sediment exchange. At Sinnamary, the 25-km study area shows up to 5 km (59.6 km2) of shoreline retreat between 1951 and 1987. This erosion by wave attack in an inter-mudbank period has altered the morphology of the Sinnamary River mudcape and released 8 X 19 8 tons to the coastal zone annually. The Acoupa study area exhibits similar erosion and the beginning of an accretionary phase associated with an offshore mudbank: rapid mangrove colonization and mudcape-like morphologies at the mouth of small rivers. Nearer the Amazon mouth, Cassipore shows limited shoreline and mudcape migration due to an absence of mudbank cyclicity. This difference in styles of shoreline evolution is reflected in the Be and Cs data: inventories are at least one order of magnitude higher at Sinnamary than Cassipore. We hypothesize a stronger continental influence in French Guiana coastal mud caused mainly by the enormous export of subaerial sediments from mangrove swamps. Deep physical mixing (up to 80 cm) of Th and Be in nearshore mud at both sites is further evidence of sediment mobility in the area.

1500 Coffee Break

1530 S C Apte, G M Day

Dissolved Metal Concentrations in the Fly River Estuary, Torres Strait and Gulf of Papua

Dissolved copper, cadmium and nickel concentrations were accurately measured in surface water samples from the Fly River estuary, Torres Strait and the Gulf of Papua. The Purari River was identified as a major source of cadmium and nickel whereas the Fly River, which receives inputs from the Ok Tedi copper mine, was a major source of dissolved copper. Since the Ok Tedi copper mine began production in 1984, considerable interest has focused on the fate and impacts of mine-derived inputs on the Fly River estuary and surrounding coastal waters. To date, studies have examined metal concentrations in biota and benthic sediments. There are no accurate data for dissolved trace metal concentrations in the waters of this region. The aim of this study was to accurately measure the concentrations of dissolved trace metals in surface waters from the Fly River estuary, Torres Strait and Gulf of Papua. Results of surface water samples, collected from 50 sites throughout the estuary, Torres Strait and Gulf of Papua show a trace metal front exists off the coast of Papua New Guinea. Within the estuary, dissolved copper behaves non-conservatively as indicated by the rapid removal of copper from solution in the low salinity region. Throughout most of the estuary, dissolved copper concentrations are less than 2000 ng/L. In the Gulf and Torres Strait the highest dissolved copper concentrations were found at sites closest to the Fly River Estuary. Observed copper concentrations were above levels found in surface waters of the Pacific Ocean and coastal NSW, but do not approach levels of regulatory concern. Cadmium and nickel concentrations (which were comparable with data reported for uncontaminated environments) were highest off the Purari estuary. The mean metals data (for samples >34 ppt salinity) for the Torres Strait sites were higher than the mean data for the Gulf sites. This could not be attributed to large salinity differences between the two sites and most likely reflects differing trace metal cycles and the influence of processes such as scavenging, uptake by biota and release of metals from sediments. None of the observed metal concentrations approach levels of regulatory concern. Offshore, the concentrations of trace metals are close to background and were consistent with recent data for Australian coastal waters and the Pacific Ocean. The copper concentration of 155 ng/L in the Torres Strait compares to water quality criteria for marine waters of 2900 ng/L in the USA and 5000 ng/L in Australia. The measured metal concentrations were also considerably lower than those observed in European coastal shelf waters where copper concentrations in waters of >34 ppt salinity are typically 238 ng/L.

1545 G J Brunskill, I Zagorskis, J Pfitzner

Coastal Trapping of Copper in the Gulf of Papua

Some of the components of a TROPICS mass balance for copper in the Gulf of Papua will be given. Riverine inputs of particulate Cu (pCu) to the Fly and Purari River estuaries are conservatively diluted by seawater up to salinity 20, but then Cu concentrations increase in the particulate phase in the outer estuary. Cu concentrations are highest in surface sediments of the deltas and inner shelf muds, and are very much lower in the carbonate sediments of the outer shelf. About half of surface sedimentary Cu appears to be associated with aluminosilicate detrital phases, and the rest with organic matter. Relatively high concentrations and fluxes of pCu were measured in 3 day duration sediment traps deployed at the base of the slope, but a large fraction of this flux appears to be dissolved and released from surface sediments. A Redfield-like ratio of organic carbon/Cu of about 10,000 was found in some components of this study. Shelf and slope sedimentation of Cu (over excess Pb-210 time scales) appears to exceed the rate of supply from river inputs, and this excess Cu probably comes from advection of Coral Sea water onto the shelf.

1600 P W Swarzenski, B A McKee and D Porcelli

Aqueous Uranium Geochemistry in Tropical Environs: An Estuarine Comparison of the Fly and Amazon Rivers

The Fly River of Papua New Guinea and Amazon River discharge their intensely weathered, lateritic loads onto high energy inner-shelf environments. Here energetic tides and currents induce extensive physical reworking which can strongly affect the particle scavenging efficiency as well as aggregation-dissaggregation kinetics. Aqueous geochemical processes in such settings can thus have a strong influence on the estuarine behavior of many inorganic and organic constituents, including uranium.

Water column samples were collected across a salinity gradient on the Amazon Shelf (March 1996) as well as in the Gulf of Papua/Fly River (September 1997). Uranium was measured both by ICP-MS and TIMS (for U isotopes). Ancillary water column measurements include dissolved organic carbon (DOC), Sr, Ba and suspended particulate matter (SPM). On the Amazon Shelf, dissolved uranium revealed extensive removal at all salinity values less than 5. Such strong removal likely involves efficient scavenging of dissolved U onto fine particles (i.e., colloids) and particle surface coatings (Fe/Mn oxides) which then can be removed from the water column by aggregation and settling. 234U/238U isotope (UAR) analyses also indicated extensive fractionation during early estuarine mixing within waters closest to the river mouth (UAR = 1.210 ± 0.010) out to a salinity of 5 (UAR = 1.136 ± 0.015). DOC values in the Amazon system decreased sharply from an upper extreme of 27 mg L-1 (Rio Cassipore tributary) to 2 mg L-1 at a salinity of ~ 30. The Amazon data will be compared to recent data collected from the Fly River/Gulf of Papua region.

1615 E Sholkovitz

Sources of Rare Earth Elements to the Western Pacific Ocean with Emphasis on the Fly River (Papua New Guinea).

A special feature of the western tropical Pacific Ocean is that it's surface waters receive a large import of aeolian particles from Asia and river particles from rapidly weathering islands such as New Guinea. It is this duality of sources with their different geochemical signatures and their different chemical reactivities in seawater which makes this region an excellent one to study. The rare earth elements (REE) are explored as tracers for the sources of terrestrial matter and as indicators of chemical weathering and estuarine chemistry.

The Fly River is one of the largest rivers in Papua New Guinea (PNG) and as such, drains a region of intense chemical weathering. The dissolved phase of the Fly River water has a striking REE composition characterized by a shale normalized pattern with a large maximum in the middle REE (MREE being Sm-Er). This "MREE bulge" differs greatly from that of dust being transported from Asia to the Western Pacific and distal surface seawater of the tropical Pacific.

A simple experiment was carried out in which suspended particles collected from the Fly River were leached with filtered seawater. A large MREE bulge rapidly develops ibn the filtered seawater. Hence, there is large scale fractionation as the middle REE are preferentially released from the river particles into the dissolved phase of seawater. These results imply that the reaction of seawater with PNG river particles will enhance the MREE bulge already present in the dissolved phase of Fly River water.

Dissolved concentration data from the Fly River Estuary show that there is large scale removal of all the REE in the 0-10 salinity range and that the MREE bulge is maintained through the mixing zone. Published REE data for surface seawater from the Coral Sea Basin and the East Caroline Basin, south and north of PNG respectively, also exhibits anomalies in the middle REE. It is likely that the MREE bulge observed in these tow basins is a signature which reflects the discharge of river waters from PNG and reactions of PNG river particles with seawater. The middle REE bulge may provide a new hydrographic tracer for circulation in the region.

1630 A. Gani Ilahude and Muswery Muchtar

Indonesian Program for TROPICS

Indonesia is planning to carry out a multidisciplinary oceanographic research program in the Membramo and Digul River estuaries of New Guinea (Irian Jaya) during the period of the coming 7th PELITA (Five Year Development:1999-2004). This paper discusses the details of that program. Some information on the geography of the Membramo River will be presented.

POSTER: T Ayukai and K Burns

Distribution of suspended sediments and chlorophyll in the Fly and Sepik River plumes

Suspended sediment concentration (which limits ambient light level) is the primary factor controlling the spatial distribution of phytoplankton in river plumes. In the Fly River plume, suspended sediment concentration declined from > 50 mg/l to 10-15 mg/l in a narrow salinity range between 20-25 ppt. An area of high chlorophyll concentration develops outside this transitional zone, and limits the further offshore transport of suspended sediments through active macro-aggregate formation by plankton communities1. The Sepik River plume seems to lack these features. The data from two cruises, during high and low discharge periods, did not show a sharp, sudden decline at any mid salinity range. Rather, suspended sediments ranged from >60 mg/l at 0 ppt and linearly decreased with increasing salinity to ~1 mg/l at 32 ppt. This is consistent with the conservative behavior also noted for DOC. Chlorophyll concentration in the Sepik plume ranged between 0 and 1.3 mg/l, with higher values occurring above 30 ppt salinity. The findings of this study will be discussed in relation to the observed behaviors various elements in the Fly and Sepik River plumes.

1.T. Ayukai & E. Wolanski (1997) Estuar. Cstl. Shelf Sci. 44: 629-639.



-AIMS home page


web@aims.gov.au
Last updated - December 18, 2008

Copyright ©1996-2000 Australian Institute of Marine Science

URL http://www.aims.gov.au

[ About AIMS ] [ AIMS research ] [ AIMS facilities ] [ AIMS news ] [ AIMS search ]
[ AIMS publications ] [ Doing business with AIMS ] [ What's new ]
[ Site index ] [ Navigating this site ] [
Privacy policy ] [ Links ]