Runoff of Terrestrial Sediment and Nutrients into the Great Barrier Reef World Heritage Area Miles Furnas and Alan Mitchell CONTENTS Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37 Catchments Adjoining the GBR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38 Suspended Sediments in North Queensland Rivers . . . . . . . . . . . . . . . . . . . . . . . . 41 Dissolved and Particulate Nutrients in North Queensland Rivers. . . . . . . . . . . . . . 42 Nutrient Transformations within Catchments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 Acknowledgments. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 INTRODUCTION Nutrients (e.g., C, N, P, trace metals) are the elemental building blocks of all marine ecosystems, including coral reefs. The quantity, form, and availability of nutrients directly affect ecosystem productivity and the quality of the water upon which the health of corals and other reef organisms ultimately depend. While corals can grow in a variety of shallow-water tropical habitats and in waters with a range of suspended sediment loads and nutrient concentrations, well-developed reef ecosystems only occur where the waters are typically characterised by low suspended particulate and dissolved nutrient concentrations. Extensive coral reefs do not typically develop or persist where nutrient and suspended sediment concentrations are acutely or chroni- cally very high. Nutrients which support the productivity of ecosystems of the Great Barrier Reef (GBR) come from a variety of sources. These include shelfbreak upwelling of sub- surface waters from the Coral Sea (Furnas & Mitchell, 1996), biological nitrogen fixation (Larkum et al., 1988; Bell et al., 1999), benthic (Alongi, 1989) and pelagic (Ikeda et al., 1982; Hopkinson et al., 1987) mineralisation of organic matter, 4 37 © 2001 by CRC Press LLC 38 Oceanographic Processes of Coral Reefs disturbances of shelf sediments (Ullman & Sandstrom, 1987; Chongprasith, 1992), sewage discharges (Brodie, 1992), and terrestrial runoff (Moss et al., 1993; Furnas et al., 1995; Mitchell et al., 1996 and 1997). Terrestrial runoff is the largest source of nutrients directly influenced by human activities, largely through the runoff of agricultural fertilisers and land use practices which accelerate soil erosion. Neither the magnitude of terrestrial inputs of sediment and nutrients to the GBR nor the effects of terrestrially derived nutrients on ecosys- tems within the GBR are well constrained at this time. Based upon spatial estimates of agricultural land use and associated soil loss, Moss et al. (1993) and Rayment and Neil (1997) calculated that sediment and nutrient runoff have increased severalfold since the beginning of European cropping and pastoral activities in the catchments adjoining the GBR. CATCHMENTS ADJOINING THE GBR Continental drainage basins with an aggregate area of 425,000 km 2 drain into the Great Barrier Reef World Heritage Area (GBRWHA) between the tip of Cape York and Fraser Island (Figure 1, Table 1). A variety of vegetation types and land uses occur in these basins. In areas of high rainfall, particularly between 15 and 18°S, rain- forest and wet coastal forests originally predominated. A considerable proportion of the GBR catchment has been logged or cleared to some degree, particularly in the coastal lowlands or flatter terrains. In terms of economic return and nutrient (fer- tiliser) inputs, sugarcane cultivation is the principal agricultural activity in the GBR catchment though the total area planted is Ͻ3% of the total catchment area (Gilbert, 2000). Most of the remaining area of the GBR catchment is or was covered by wood- lands or savanna woodlands. Open range cattle grazing is the principal land use in these areas. The most complete record of gauged runoff into the GBRWHA (1968 to 1994) averaged close to 40 km 3 of freshwater per year, ranging between 11 km 3 (1987) and 125 km 3 (1974). Taking account of basin areas, rainfall distributions in each basin, and basin rainfall–runoff relationships, average annual freshwater runoff is estimated to be 66 km 3 , with extremes of 20 and 180 km 3 . The average annual freshwater dis- charge from the two largest river systems, the Burdekin and Fitzroy (16.4 km 3 year Ϫ1 ), accounts for 25% of the average total freshwater runoff. Total annual gauged discharge into the GBRWHA is strongly correlated with total annual discharge from these two rivers (r 2 ϭ 0.8). Within catchments, runoff of freshwater is strongly seasonal (Figure 2), and within seasons, also closely coupled to rainfall inputs. Several seasonal patterns in runoff from regional river systems are evident. Runoff from the relatively small, wet, tropical catchments of the central GBR (Daintree, 16°S to Tully Rivers, 18°S) is char- acterised by several events per wet season. Inter-annual variability in total runoff from these catchments is relatively small. Although wet tropical catchments comprise only 3% of the total GBR catchment area, 22% of the total freshwater runoff is derived from them. Discharge from the monsoonal catchments of Cape York (e.g., Normanby River) and the dry catchments of the southern half of the GBR (e.g., Burdekin and Burnett © 2001 by CRC Press LLC Runoff of Terrestrial Sediment and Nutrients into the Great Barrier Reef 39 TABLE 1 Summary Statistics for Major Catchment Basins Draining into the Great Barrier Reef Region Area Percent Annual Runoff (km 3 ) * Average (mm) Adj. Runoff Basin Name (km 2 ) Gauged Mean Maximum Minimum Rainfall Runoff % Runoff Volume (km 3 ) Jacky-Jacky Creek 2,963 0 635 2.63 Olive-Pascoe Rivers 4,179 30.7 3.71 7.14 0.27 1,187 888 0.75 3.71 Lockhart River 2,883 0 1,225 1.92 Stewart River 2,743 17.2 1.21 2.30 0.01 1,222 441 0.36 1.21 Normanby River 24,408 32.9 4.95 17.49 0.60 1,185 203 0.17 4.95 Jeannie River 3,637 18.9 1.54 4.69 0.13 1,344 423 0.32 1.54 Endeavour River 2,104 28.0 1.82 4.92 0.44 1,939 865 0.45 1.82 Daintree River 2,192 39.1 1.26 3.52 0.11 2,492 575 0.23 1.26 Mossman River 466 11.6 0.59 1.21 0.18 2,208 1,265 0.57 0.59 Barron River 2,902 89.2 0.81 2.66 0.16 1,453 279 0.19 0.81 Mulgrave-Russell Rivers 1,983 48.1 3.64 7.21 1.32 3,016 1,836 0.61 3.64 Johnstone River 2,325 59.4 4.67 9.12 1.65 2,996 2,009 0.67 4.67 Tully River 1,683 87.5 3.29 5.37 1.24 2,855 1,954 0.68 3.29 Murray River 1,107 13.6 1.06 2.60 0.38 2,098 958 0.46 1.06 Herbert River 9,843 86.9 4.01 11.99 0.53 1,506 407 0.27 4.01 Black River 1,057 32.8 0.38 1.54 0 1,530 360 0.23 0.38 Ross River 1,707 55.8 0.49 3.37 0.01 1,027 287 0.28 0.49 Haughton River 4,044 68.1 0.74 3.18 0.02 888 183 0.21 0.74 Burdekin River 130,126 99.9 10.29 54.46 0.52 727 79 0.11 10.29 Don River 3,695 16.0 0.75 3.66 0 1,045 203 0.19 0.75 Proserpine River 2,535 13.2 1.08 3.95 0.02 1,360 426 0.31 1.08 O’Connell River 2,387 30.1 1.54 4.19 0.07 1,469 645 0.44 1.54 Pioneer River 1,570 92.3 1.19 5.15 0 1,385 758 0.55 1.19 continued © 2001 by CRC Press LLC 40 Oceanographic Processes of Coral Reefs TABLE 1 (continued) Area Percent Annual Runoff (km 3 ) * Average (mm) Adj. Runoff Basin Name (km 2 ) Gauged Mean Maximum Minimum Rainfall Runoff % Runoff Volume (km 3 ) Plane Creek 2,539 19.4 1.49 4.97 0.05 1,125 587 0.52 1.49 Styx River 3,012 0 1,010 0.59 Shoalwater 3,605 0 975 0.60 Waterpark Creek 1,835 13.0 1.11 2.68 0.23 860 605 0.70 1.11 Fitzroy River 142,537 95.3 6.08 23.22 0.18 735 43 0.06 6.08 Calliope River 2,236 58.1 0.30 1.08 0.02 790 134 0.17 0.30 Boyne River 2,590 88.2 0.29 2.40 0 968 112 0.12 0.29 Baffle Creek 3,996 36.9 0.78 3.03 0.08 893 195 0.22 0.78 Kolan River 2,901 79.7 0.41 2.10 0.02 1,065 141 0.13 0.41 Burnett River 33,248 97.9 1.15 6.37 0.12 763 35 0.05 1.15 Burrum River 3,358 13.6 0.55 2.36 0.03 766 164 0.21 0.55 Mary River 9,440 50.3 2.72 9.27 0.26 1,174 288 0.25 2.72 Note: Average rainfall calculated by overlaying basin boundaries on the long-term isohyet map. Basin areas, gauged runof f (1968–1994), and rainfall data obtained from the Queensland Department of Natural Resources. *10 9 m 3 ϭ 1 km 3 ϭ 1 million megalitres (10 6 Ml). © 2001 by CRC Press LLC Runoff of Terrestrial Sediment and Nutrients into the Great Barrier Reef 41 Rivers) are typically characterised by a single major event per year. The magnitude of this event can vary dramatically between years. In the drier southern catchments, several years may pass between significant runoff events. For example, no major floods have occurred in the Fitzroy River since 1991. The largest flows in all catchment types are most frequently associated with intense rainfall derived from tropical cyclones coming from the Coral Sea or rain depressions derived from cyclones moving inland, either southward from the Gulf of Carpentaria or westward from the Coral Sea. These storms can drop massive amounts of water at rates that greatly exceed the capacity of catchments to absorb the input and that have a high erosive potential as the water runs off the land (e.g., Ciesiolka, 1987). SUSPENDED SEDIMENTS IN NORTH QUEENSLAND RIVERS Suspended sediment loads transported by regional rivers vary directly with the amount of water moving through the system. Considerable between-catchment differences in sediment loads also occur. Fine soil particles reaching larger channels of river systems are rapidly transported downstream. Figure 3 illustrates relationships between dis- charge and suspended sediment concentrations measured over single wet seasons in the Tully River (wet tropics) and the Herbert River (a largely dry catchment). Wet season hydrographs in the Tully River (Figure 3, top) typically consist of multiple flow events. Maximum concentrations of fine (Ͻ10 m) suspended sedi- ments in the Tully River are 0.5 g l Ϫ1 or less. Suspended sediment concentrations rise rapidly on the leading edge of the flood hydrograph and decay exponentially there- after. At most times, suspended loads are 0.1 g l Ϫ1 or less. Flood events rarely persist for more than a few days. Wet, tropical catchments such as that of the Tully River are naturally characterised by a high level of vegetation cover. Freshly cultivated fields or newly cleared land are the only significant areas of bare soil. Soil loss rates from rainforest in the wet tropics can be quite low (Ͻ0.1 tonnes ha Ϫ1 year Ϫ1 : Douglas, 1967; Capelin & Prove, 1983). In contrast, soil loss rates from cultivated fields or freshly cleared land can range from Ͻ5 to Ͼ500 tonnes ha Ϫ1 year Ϫ1 depending upon rainfall intensity, slope, vegetation cover, and land management practises (Sallaway, 1979; Prove et al., 1986; Prove & Hicks, 1992). Filtered particulate matter from the Tully River has an average N and P content of 2825 and 564 ppm (g g Ϫ1 ). These values are enriched compared to the mean N (1560 Ϯ 1770 ppm) and P (352 Ϯ 340 ppm) composition of catchment soils (M. Grundy, QDNR, personal communication). During the 1996 –1997 wet season (total freshwater discharge ϭ 2.3 km 3 ), we estimate that the Tully River transported 1.45 ϫ 10 5 tonnes of fine sediments. Very little sediment is transported during the dry season when water levels and suspended solids concentrations are low. For the sedi- ment nutrient compositions given above, we estimate annual sediment-associated N and P exports of 410 and 82 tonnes, respectively. In contrast to the Tully River, the bulk of freshwater and fine sediment exports from the mainly dry catchment rivers of the central and southern GBR such as the © 2001 by CRC Press LLC 42 Oceanographic Processes of Coral Reefs Herbert most frequently comes as a single large event (Figure 3, bottom). Suspended sediment concentrations measured in dry tropical rivers are generally higher than those measured in wet tropical rivers. Maximum suspended solids concentrations measured in the Burdekin River during flood peaks can be on the order of 2 to 3 g l Ϫ1 and often remain at concentrations Ͼ0.2 g l Ϫ1 for weeks. These higher suspended par- ticle loads reflect both differences in soil types between wet and dry catchments and higher area-specific erosion rates due to much lower levels of vegetative cover. Maximum suspended loads in the Herbert River during flood events are lower than the Burdekin River, peaking at circa 1 g l Ϫ1 . Suspended sediments collected in the Herbert River have mean N and P contents of 1470 and 350 ppm, respectively. These are close to the average of measured soil N (1160 Ϯ 960 ppm) and P (350 Ϯ 320 ppm) compositions (M. Grundy, QDNR, personal communication). For the 1997–1998 wet season (total freshwater discharge ϭ 3.7 km 3 ), the time series of sus- pended load measurements give an estimated fine sediment export of 8.93 ϫ 10 5 tonnes. Using the sediment composition ratios given above, this translates to sedi- ment-bound N and P exports of 1314 and 315 tonnes, respectively. DISSOLVED AND PARTICULATE NUTRIENTS IN NORTH QUEENSLAND RIVERS Dissolved and particulate nutrient (N, P) concentrations in north Queensland rivers vary with time in a variety of ways. Figure 4 presents an example from the wet, trop- ical South Johnstone River where dissolved and particulate nutrient concentrations were measured intensively over 2 years. A large proportion of the Johnstone River catchment area (26%) remains covered by rainforest. The two major agricultural land uses are sugarcane cultivation in the lower part of the catchment (16%) and grazing (circa 40%). Total gauged discharge of freshwater from the catchment in the 2 years sampled was similar (0.83 and 0.89 km 3 ), though the temporal pattern of discharge differed between years. Concentrations of particulate N and P were highly variable and generally increased with river discharge, reflecting concurrent increases in catchment erosion and suspended sediment loads. In contrast, concentrations of dissolved inorganic (NH 4 ϩ , NO 2 Ϫ , NO 3 Ϫ , PO 4 ϭ ) and dissolved organic N and P (DON, DOP) fluctuated within smaller concentration ranges, largely independent of both the discharge rate and the suspended sediment load. Nitrate is the principal inorganic N species in river waters. The highest nitrate concentrations (to 765 g NO 3 ϪN l Ϫ1 ) were typically measured during the initial flow event of the wet season (first flush), when soluble nitrate ions produced and stored in aerobic soils over the dry season are washed into the river systems. Lower nitrate concentration peaks occur during late-season flood peaks. Short-lived decreases in nitrate concentrations may occur during major cyclonic flood events, suggesting a dilution of catchment sources. Dissolved organic N concentrations are somewhat higher, but again, relatively constant with time. When the dissolved N concentrations are flow-weighed and integrated over the course of the wet seasons, nitrate exports from the South Johnstone catchment account for 37% of the total N exports and 60% of the dissolved N exports. © 2001 by CRC Press LLC Runoff of Terrestrial Sediment and Nutrients into the Great Barrier Reef 43 Average annual N and P exports from the South Johnstone River catchment for the 1989–1991 hydrological years are estimated to be 314 and 30 tonnes, respectively, with volume-weighted annual mean N and P concentrations of 364 and 35 mg m Ϫ3 . A large proportion of the total export occurs in the relatively short peri- ods of high discharge. Approximately 78% of the P was exported in particulate form. Dissolved inorganic and organic P accounted for approximately 15 and 9% of exports. In comparison, DIN, DON, and particulate N accounted for 43, 16, and 41% of the total N exports. For the entire Johnstone River catchment, Hunter and Walton (1997) estimated total average N and P exports of 750 and 132 tonnes/year. Integrations of annual nutrient fluxes from the Herbert River catchment indicate that the proportions of total N and P export in dissolved and particulate forms were of similar order to those observed in the South Johnstone River. If the bulk of particu- late N and P are associated with fine soil particles suspended in river waters, total N exports from the Tully River (1996–1997) and Herbert River (1997–1998) catch- ments are estimated to be 1000 and 3200 tonnes, respectively. Concurrent total P exports derived from sediment fluxes would be 105 and 404 tonnes. NUTRIENT TRANSFORMATIONS WITHIN CATCHMENTS Time series sampling at two sites located longitudinally on the Herbert River flood- plain over the 1989–1990 and 1990–1991 hydrological years illustrate nutrient trans- formations occurring within catchments and the potential influence of agricultural land use on river nutrient exports (Figure 5). The Herbert River catchment contains a variety of vegetation communities and land uses. Cropping, principally of sugarcane, is largely restricted to the floodplain which comprises circa 30% of the total catch- ment area (D. Walker, CSIRO personal communication). A significant proportion of the lower catchment has been cleared and virtually all of the current inputs of N and P fertilisers (9800 and 1330 tonnes of N and P, respectively, in 1990; Pulsford, 1993) are applied there. Much of the upper catchment is covered by dry savanna woodland where the principle agricultural land use is cattle grazing. Figure 5 illustrates differences between concentrations of major nitrogen and phosphorus species in Herbert River waters measured contemporaneously at the upper (Yamani Falls National Park) and lower (Ingham) ends of the floodplain. Concentrations of particulate nitrogen (PN) at the lower floodplain site are not sig- nificantly different (circa 2 g l Ϫ1 ) from contemporaneous PN concentrations at the upstream site over much of the year. Inorganic nitrogen concentrations, chiefly of nitrate, were consistently higher (mean difference ϭ 79 g N l Ϫ1 ) at the downstream site. Differences between upstream and downstream DIN concentrations were small- est during flood events due to dilution. Similar longitudinal trends in DIN concentra- tions have also been observed in sub-catchments draining rainforest into the Herbert River floodplain (BSES, Furnas, unpublished data). In contrast, concentrations of DON at the upstream site were consistently higher (mean difference ϭ 46 N g l Ϫ1 ) than downstream concentrations. If the river flow at Yamani Falls is assumed to be identical to that measured at Ingham (making the overly conservative assumption that there is negligible net freshwater input to the lower catchment), then integrated fluxes © 2001 by CRC Press LLC 44 Oceanographic Processes of Coral Reefs of DIN, chiefly nitrate, leaving the lower catchment in runoff were four to eight times the DIN flux entering at the upper end. In contrast, the integrated flux of DON at Ingham is circa 87% of the nominal input flux. Particulate N fluxes at the down- stream site are approximately 1.8 times the input fluxes, though suspended sediment concentrations generally did not differ greatly between the two sites. Overall, total N inputs at the upper end of the floodplain (1045 and 1600 tonnes for 1989–1990 and 1990–1991, respectively) are 59 and 62% of the calculated discharge fluxes (1784 and 2590 tonnes N) at the lower floodplain site. In the case of phosphorus, upstream–downstream differences between concen- trations of dissolved P species were relatively small (mean pairwise differences for DIP and DOP ϭ 2.2 and 0.1 g P l Ϫ1 , respectively). Particulate P concentrations at the lower catchment site averaged 5.6 g l Ϫ1 greater than the upper catchment site. Total P inputs at the upper end of the floodplain (89 and 74 tonnes for 1989–1990 and 1990–1991, respectively) were 50 and 17% of the calculated downstream exit fluxes (180 and 430 tonnes P). The observed upstream–downstream differences between concentrations of nitrogen species in the lower Herbert River indicate that the floodplain is the major source of the DIN, chiefly nitrate, exported from the Herbert River catchment. A por- tion of the DON generated in the upper catchment is consumed or oxidised in reaches on the floodplain. Absolute differences between DIN (largely nitrate) concentrations at Yamani Falls and Ingham are greatest during low-flow periods, suggesting that lat- eral inputs of nitrate-rich groundwater are a significant source of this nitrate. Similar longitudinal increases in nitrate concentrations have been observed in the Tully River (Furnas, unpublished data). High nitrate groundwaters have been observed in other lowland sugar-growing regions adjoining the GBRWHA (Keating et al., 1996). The net increments of N and P exported from the lower catchment for 1989–1990 and 1990–1991 (740 and 990 tonnes for N; 91 and 356 tonnes for P) can be compared to fertiliser N and P inputs to the lower catchment (9800 and 1330 tonnes, respectively, in 1990: Pulsford, 1993). DISCUSSION Rivers discharging into the GBRWHA make a significant contribution to the nutrient budgets of the region’s ecosystems (Furnas et al., 1995). At the shelf scale, annual inputs of freshwater from rivers (22 to 180 km 3 ) are small relative to freshwater inputs from rainfall (circa 170 to 440 km 3 year Ϫ1 ). However, as the nutrient content of regional rainwaters is very low (Furnas et al., 1995), rain contributes considerably less than rivers to shelf-scale nutrient inputs. Overall, annual volumes of runoff com- prise approximately 4% of the volume of seawater inshore of the 20-m isobath. Because of lateral resistance to flow, Coriolis steering of buoyancy currents, and the event-based nature of river discharge, the most direct effects of freshwater, sedi- ments, nutrients, and contaminants in runoff are concentrated in the nearshore zone (Gagan et al., 1987; Sandstrom, 1988). Despite considerable effort over the last decade to monitor concentrations of dis- solved and particulate nutrients in a number of NE Queensland river systems (e.g., Mitchell et al., 1991 and 1996) and make volume-weighted estimates of nutrient © 2001 by CRC Press LLC Runoff of Terrestrial Sediment and Nutrients into the Great Barrier Reef 45 exports, considerable uncertainties remain in our estimates of nutrient and sediment exports to the shelf. This uncertainty is due to the small number of datasets with both the temporal resolution necessary to accurately integrate exports during flood events and duration to resolve inter-annual variability between and within catchments. As an example, the two largest river systems discharging into the GBR (Burdekin, Fitzroy: combined watershed area Ͼ270 ϫ 10 3 km 2 ) account for 25% of the long-term aver- age discharge to the shelf. At present, detailed time series of nutrient concentrations during significant flood events are only available for the Burdekin River. No signifi- cant flood events have occurred in the Fitzroy catchment since 1991. Likewise, though catchments on Cape York (11 to 15°S) comprise 10% of the total catchment area and receive 13% of the estimated rainfall, no useful time series of nutrient con- centrations in these rivers exist. Despite the paucity of data, several approaches can be taken to make first-order estimates of sediment, and nutrient inputs to the GBR. The simplest approach involves extrapolation from the quantity of sediment delivered annually by rivers to the GBR shelf, the nominal N and P content of that sediment, and the proportion of total nutrients transported on sediment particles. Table 2 presents estimates of aver- age nutrient delivery based upon estimates of shelf-scale sediment inputs. The sedi- ment inputs have been calculated from the estimated accumulation of sediment in the coastal sediment wedge (Belperio, 1983) or weighted discharge-transport rela- tionships derived from a small number of rivers (Neil & Yu, 1995 and 1996). The accuracy and precision of these estimates and the reliability with which they can be extrapolated are difficult to assess. The N and P inputs are calculated assuming that the soils eroding into the GBRWHA have a mean sediment N and P content of 0.13 and 0.04%, respectively, and that sediment-associated N and P constitute 40 and 80% of the total N and P delivery fluxes. Volume-specific fine sediment fluxes mea- sured to date in wet and dry catchment rivers using turbidity time series (Figure 3) vary approximately 20-fold (5 ϫ 10 4 tonnes km Ϫ3 in wet tropical rivers to 1 ϫ 10 6 tonnes km Ϫ3 in the Burdekin River). These end member volume-specific loads give TABLE 2 Estimated Total Terrestrial Nitrogen and Phosphorus Inputs to the Great Barrier Reef Region Calculated from Estimates of Sediment Delivery to the GBR Lagoon Sediment Input N Inputs a P Inputs b Source (tonnes ϫ 10 6 ) (tonnes ϫ 10 3 ) (tonnes ϫ 10 3 ) Belperio, 1983 27.4 89.0 13.7 Moss et al., 1993 15.3 49.6 7.6 Neil & Yu, 1996 (pre-European) 7.4 23.9 3.7 Neil & Yu, 1996 (modern) 28.0 91.1 14.0 Furnas & Mitchell (this study) 3.3–66 11–215 1.6–33 a PN ϭ 0.13% of sediment weight and PN ϭ 40% of ⌺N. b PP ϭ 0.04% of sediment weight and PP ϭ 80% of ⌺P. © 2001 by CRC Press LLC 46 Oceanographic Processes of Coral Reefs potential total N exports ranging between 11 and 215 ϫ 10 3 tonnes and P exports ranging between 1.65 and 35 ϫ 10 3 tonnes (Table 2). These ranges encompass sediment derived nutrient inputs calculated from literature sources. Total volume- specific loads in most individual rivers have not been weighted as yet between these end members, but it seems most likely that the upper ends of the ranges are too high. Based upon detailed time series measurements of dissolved and particulate nutri- ents in the South Johnstone River (Furnas et al., 1995), we estimate annual volume- weighted N and P export fluxes to be 364 and 35 tonnes km Ϫ3 of discharge, respectively. Volume-weighted total N and P concentrations derived for post-cyclone flood events in the Johnstone River (3060 tonnes N km Ϫ3 , 930 tonnes P km Ϫ3 : Hunter & Walton, 1997) and Herbert River (574 tonnes N km Ϫ3 , 121 tonnes P km Ϫ3 : Mitchell et al., 1997) are considerably higher. Particulate matter mobilised during flood events is enriched in both N and P (Mitchell & Furnas, 1997) relative to seasonally averaged sediment N and P concentrations, reflecting higher erosion rates and scavenging of nutrients from catchment soils. Recent integrations of flow- and time-varying nutrient concentrations in the Burdekin River yield mean volume-weighted N and P concentrations of 940 tonnes N km Ϫ3 and 145 tonnes P km Ϫ3 , respectively (Furnas, unpublished data). Runoff from the drier catchments has higher suspended sediment and particulate nutrient concen- trations than water discharged from the wet, tropical rivers due to differences in soil types and lower vegetation cover. If the volume-weighted nutrient concentrations determined for the Burdekin River are taken as representative of discharge from dry catchments (32 km 3 discharge) and the South Johnstone concentrations are taken to represent wet tropical and Cape York catchments (34 km 3 discharge), mean annual river N and P inputs to the GBRWHA are estimated to be 4.7 ϫ 10 4 and 1.7 ϫ 10 3 tonnes, respectively. In comparison, shelf water stocks of N and P (dissolved and par- ticulate) inshore of the 20-m isobath are on the order of 1.4 ϫ 10 5 tonnes N and 1.4 ϫ 10 4 tonnes P, respectively. Such a crude comparison must be taken cautiously as a large portion of the N and P transported by rivers is likely denitrified or buried at or close to the river mouth (Nixon et al., 1996). The comparison does illustrate that riverine inputs of nutrients can have a significant effect on both nearshore and shelf- scale nutrient budgets. A variety of evidence clearly shows that agricultural and land-use activities in catchments adjoining the GBRWHA have led to increased erosion of soils within those catchments. Working from this type of data, simple models of runoff, land use and sediment delivery (Moss et al., 1993; Rayment & Neil, 1997) suggest that river- ine sediment (and concurrent nutrient) fluxes to the GBRWHA from adjoining catchments have increased several fold since the advent of European agricultural practises. The available empirical data from river sampling confirm the general con- clusion, but are as yet insufficient to strongly constrain the magnitude of the increase. Efforts to develop management policies to halt or reverse this trend require accurate information on the scope of the problem and the mechanisms which drive it. Even more important, such information is essential to educate land users and develop a community-based land conservation ethos that ultimately benefits the land, land users, and ecosystems of the GBRWHA. © 2001 by CRC Press LLC [...]... Restricted terrestrial carbon input to the continental shelf during cyclone Winifred: implications for terrestrial runoff to the Great Barrier Reef province Coral Reefs 6, 113 –119 Gilbert, M 2000 Population and Major Land Use in the Great Barrier Reef Catchment Area: Spatial and Temporal Trends Unpublished report to the Great Barrier Reef Marine Park Authority, Townsville Hopkinson, C.S., Sherr, B.F.,... Land-Use Patterns and Nutrient Loadings of the Great Barrier Reef Region James Cook University, Townsville Pulsford, J.S 1993 Historical Nutrient Usage in Coastal Queensland River Catchments Adjacent to the Great Barrier Reef Marine Park Research Report No 40 , Great Barrier Reef Marine Park Authority, Townsville, 63 pp © 2001 by CRC Press LLC Runoff of Terrestrial Sediment and Nutrients into the Great. .. hydrocarbons in surface sediments from the north Queensland coast and Great Barrier Reef: effects of tropical cyclone Winifred Organic Geochemistry 12, 44 5 45 4 Ullman, W.J & Sandstrom, M.W 1987 Dissolved nutrient fluxes from the nearshore sediments of Bowling Green Bay, central Great Barrier Reef lagoon (Australia) Estuarine, Coastal and Shelf Science 24, 289 –303 Wasson, R.J 1997 Runoff from the land to the. .. Barrier Reef Conference James Cook University, Townsville Brodie, J 1992 Urban impact on the Great Barrier Reef pp 18–26 in Proceedings of the Workshop on Land-Use Patterns and Nutrient Loadings of the Great Barrier Reef Region James Cook University, Townsville Capelin, M.A & Prove, B.G 1983 Soil conservation problems of the humid coastal tropics of North Queensland Proceedings Australian Society of Sugar... the land-sea margin of the North Atlantic Ocean Biogeochemistry 35, 141 –180 Prove, B.G., Truong, P.N., & Evans, D.S 1986 Strategies for controlling caneland erosion in the wet tropical coasts of Queensland Proceedings of the Australian Society of Sugar Cane Technologists 77 – 84 Prove, B.G & Hicks, W.H 1992 Soil and nutrient movements from rural lands of North Queensland pp 67 –76 in Proceedings of the. ..Runoff of Terrestrial Sediment and Nutrients into the Great Barrier Reef 47 ACKNOWLEDGMENTS We thank Michele Skuza, Marg Wright, Jane Wu Won, and Neil Johnson for assistance with nutrient analyses River discharge, rainfall, and soil nutrient data were obtained from the Queensland Department of Natural Resources This work was supported by the Australian Institute of Marine Science, the Great Barrier Reef. .. Unwin, P 1999 Nitrogen fixation by Trichodesmium spp in the central and northern Great Barrier Reef lagoon: relative importance of the fixed nitrogen load Marine Ecology Progress Series 186, 119 –126 Belperio, A.P 1983 Late quaternary terrigenous sedimentation in the Great Barrier Reef lagoon pp 71–76 in Baker, J.T., Carter, R.M., Sammarco, P.W & Stark, K.P (eds) Proceedings of the Inaugural Great Barrier. .. and nutrient fluxes from river systems draining to the Great Barrier Reef Marine Park pp 23– 34 in Hunter, H.M., Eyles, A.G & Rayment, G.E (eds) Proceedings of the Conference on Downstream Effects of Land Use, Rockhampton, April 1995 Queensland Department of Natural Resources, Brisbane Mitchell, A.W., Bramley, R.G.V., & Johnson, A.K.L 1997 Export of nutrients and suspended sediment during a cyclone-mediated... 1995 Nitrogen and Phosphorus Budgets for the Central Great Barrier Reef Research Report No 36, Great Barrier Reef Marine Park Authority, Townsville, 95 pp Furnas, M.J & Mitchell, A.W 1996 Nutrient inputs to the central Great Barrier Reef (Australia) from subsurface intrusions of Coral Sea waters: a two-dimensional displacement model Continental Shelf Research 16, 1127 –1 148 Gagan, M.K., Sandstrom, M.W.,... rivers and sea pp 23 41 in Conference Proceedings: The Great Barrier Reef: Science, Use and Management, Vol 1, November 1996 James Cook University, Townsville © 2001 by CRC Press LLC 50 Oceanographic Processes of Coral Reefs FIGURE 1 Major rivers draining into the GBRWH Area FIGURE 2 Comparative time series of daily fresh water discharge from the Normanby River (monsoonal catchment—Cape York Peninsula), . cropping and pastoral activities in the catchments adjoining the GBR. CATCHMENTS ADJOINING THE GBR Continental drainage basins with an aggregate area of 42 5,000 km 2 drain into the Great Barrier Reef. Processes of Coral Reefs of DIN, chiefly nitrate, leaving the lower catchment in runoff were four to eight times the DIN flux entering at the upper end. In contrast, the integrated flux of DON at Ingham. impact on the Great Barrier Reef. pp. 18–26 in Proceedings of the Workshop on Land-Use Patterns and Nutrient Loadings of the Great Barrier Reef Region. James Cook University, Townsville. Capelin, M.A.