The Darling River: Algal growth and the cycling and sources of nutrients Project M386 - SB/1/34
CRCFE, CSIRO Land & Water, MDFRC Client Report
A key aim of the project was to identify environmental conditions responsible for the occurrence of blooms of blue-green algae (cyanobacteria) in the Darling River. Emphasis was given to describing the role of phosphorus and identifying its major sources of supply. Specifically we sought to: develop an empirical model of phytoplankton growth identify the main environmental conditions that trigger blooms of cyanobacteria identify the amounts and the physico-chemical forms of nutrients transported in the river determine the factors that control ì bioavailabilityî (ie. availability to phytoplankton) of the various forms of major nutrients assess whether the present phosphorus concentrations in the river are higher than they were before European settlement identify and quantify the major sources and types of phosphorus reaching the Darling River The influence of riverine conditions on the growth of phytoplakton was investigated in a section of the Darling River near the township of Bourke. Sediment transport was investigated from samples of suspended particles and bottom sediments collected from sites distributed along the length of the river. Catchment sources of particles were investigated using soil samples collected from transects of 100km length distributed across different land types. 2. Major findings Concentrations of cyanobacteria in Bourke Weir Pool are inversely related to discharge rate. Populations greater than 1,000 cells mL-1 (cell per millilitre) do not occur when discharge rates exceed 800-1,000 ML d-1 (megalitres per day). At discharge rates lower than 500 ML d-1 large blooms occur more frequently. The growth of cyanobacteria is enhanced by increased temperature stratification that develops in summer when discharge rates decline below 800 ML day-1. Vertical attenuation of light was linearly related to turbidity in the weir pool at Bourke. This relationship was used to estimate the depth of light penetration from turbidity monitoring data and to calculate an index of light suitability based on the ratio of light penetration to mixing depth. During the 44 months of the study the ratio indicated that light was limiting for at least 50% of the time and was adequate for 32% of the time. Improving light conditions initiated the growth of Cyanobacteria and significant populations were always associated with periods of adequate light. Reductions in turbidity during periods of reduced flow were associated with increases in salinity. An inverse correlation was found between salinity and turbidity with turbidity declining significantly when conductivity exceeded 300 μS cm-1. Increased salinity was shown to enhance particle aggregation leading to increased sedimentation rates. From the relationships between salinity and turbidity, and turbidity and light penetration, it was calculated that light was limiting in Bourke Weir Pool when turbidity exceeded 100 NTU and that turbidity fell below this level when conductivity increased above ca. 300 μS cm-1. Median concentrations of total phosphorus (TP) and total nitrogen (TN) are relatively high in the river. Despite this, phytoplankton numbers remain low for much of the year due to light limitation. Cyanobacterial blooms dominated by nitrogen fixing species occur during low flow periods when TN:TP ratios are high. This is contrary to the general view that nitrogen fixers are selected for by low TN:TP ratios indicative of reduced nitrogen supplies. By comparing methods for assessing nutrient limitation of phytoplankton it was demonstrated that the TN:TP ratio is an unreliable indicator of nutrient supply. Potential nutrient limitation was better identified using measurements of the available forms of nutrients. Nitrogen is generally the nutrient in lowest supply for supporting phytoplankton growth and this creates conditions that are advantageous to nitrogen fixing blue-green algae. Reductions of nitrogen loads without concomitant phosphorus reduction will exacerbate this situation. Under the current conditions nitrogen loads should be managed so that either they do not increase or they are reduced appropriately in conjunction with phosphorus reductions. During blooms of nitrogen fixing cyanobacteria the forms of phosphorus available for uptake are reduced to low concentrations. This suggests that efforts to reduce the supply of available forms of phosphorus could lead to a reduction in the intensity of cyanobacterial blooms. However, this view is tempered by the observation that phosphorus limitation was alleviated by a re-supply of nutrient from the bottom sediments when conditions were suitable. An additional benefit of reducing phosphorus loads to the river is that it should reduce the likelihood of nitrogen limitation and provide less of an advantage to nitrogen fixing cyanobacteria. Most of the phosphorus load to the Darling River is bound to particles but only a small fraction is required to stimulate and support algal growth. This makes it is difficult to unequivocally identify the sources of dissolved orthophosphate utilised by the phytoplankton. Total Catchment Management plans aimed at reducing nutrient loads should address both diffuse and point sources as both are playing a role in contributing phosphorus to the system. Diffuse sources are difficult to mange but must be addressed. A comparison of Darling River particles with those from seven other Australian rivers indicated that the maximum adsorption capacity for phosphorus was not substantially different between the rivers. In contrast the affinity of binding sites appeared to be lower for particles from the Darling River compared with particles from the other rivers. It is possible that this lower affinity is responsible for the relatively high ratios of dissolved to total phosphorus that occur in the Darling River. Water contributions from tributaries were analysed as a first step to identifying sources of sediments and associated phosphorus. An analysis of cumulative flows showed the Namoi and Upper Barwon to be the highest flow contributors to the Darling-Barwon, each being responsible for 21% of the total tributary inflow upstream of Wilcannia and 35% of the flow upstream of Walgett. The Culgoa contributed 16% of tributary inflow upstream of Wilcannia, and the Macquarie and Boomi followed with each contributing about 10%. Large errors in the water balance appear to be due to difficulties inherent in gauging river flows during periods of floods when over-bank losses of water and sediment occur. The data show that these gauging problems occur in the lower part of the Border Rivers Basin and also downstream from Bourke or Brewarrina. The existing flow and turbidity data are not adequate to derive a reliable model of cumulative sediment load. Suitable data for modelling sediment transport requires high frequency turbidity data collected over several years and including a range of major flow events. Particle composition was used to identify sources of sediment from amongst the various catchments. The transported sediments are derived predominantly from lowland weathered granite and sedimentary rock sources with upland basalts contributing less than 5% of the load. An analysis of the contemporary (1991-1994) sediment load from each tributary to the load at Bourke gave estimates for their contributions that contrasted sharply with the long-term water loads. Sediment cores indicated that the long-term contribution of the Namoi River has been much less than is indicated by contemporary measurements obtained under recent atypical flows. For typical flow conditions the Namoi would make a much smaller sediment contribution, estimated at 20%, which matches that indicated by sediment cores and long term water loads. This illustrates that transport of sediment is highly variable in river systems such as the Barwon-Darling that have a highly variable flow regime. Sediment core samples were collected from the Darling-Barwon channel upstream and downstream of the Namoi River junction. The oldest dates in sub-samples from the cores were 132 ± 12 and 287 ± 30 years. There is little variation in the phosphorus concentration down through the cores and the values are consistent with concentrations in contemporary sediment. The lack of variation indicates that phosphorus concentrations in the Darling-Barwon channel have remained constant over this period. The implication is that European land management practices have not affected the concentrations of sediment-associated phosphorus. In contrast, phosphorus concentrations in a sediment core from the Namoi River increased by 30% in the top layer indicating a change about 30 years ago. Prior to this the phosphorus concentrations had remained constant for over 100 years. Trace element geochemistry suggested no contribution from fertiliser phosphorus in the top layer of the core. Cores collected from the middle reaches of the Namoi River showed that after European settlement sources of sediment changed from mainly sedimentary and granite rock to an increased contribution from upland basaltic rock. In the upper catchment this increase occurred 150 years ago and it has taken 120 years for the effects of changes to sediment supply to impact on the lower reaches. This highlights the inefficiency of sediment transport in these lowland river systems. Fallout radionuclide tracers were used to determine the land-use types or land-forms that are the predominant sources of sediment delivered to the waterways of the Darling-Barwon river system. The calculations are sensitive to the residence time of particles transported through the system. If sediment residence time is assumed to be <30 years, then the minimum relative sediment contribution from channel bank and gully sources is 69+11%, while the maximum contribution from cultivated land is 31+13%, and the maximum contribution from uncultivated regions is 10+3%. If sediment residence times are longer than 40 years the input from cultivated lands can be neglected. As residence time increase above 40 years there is a reduction in the relative sediment contribution derived from gully bank erosion (decreasing from a maximum of 75%) and a continual increase in sediment contribution derived from uncultivated land (rising from a minimum of 25%). If residence times are in excess of 80 years then the model suggests that most of the sediment is sourced from surface eroded material. Confidence in the predictions provided by these scenarios requires clarification of the sediment transport processes and improved estimates of the residence time of fine sediments in the river system. 3. Further management implications Increased flows within the river will reduce the probability of cyanobacterial blooms but have some disadvantages. Low flows are frequently the result of extensive water abstractions and so any increase in flow reduces the availability of water for extractive purposes. Increased flows also present problems from the ecological perspective as flows for algal bloom suppression during summer could disadvantage aquatic organisms that have evolved to low summer flow conditions. It is expected that salinities of major tributaries will increase significantly in the future. It is possible that the river will become clearer as salinity levels increase and this may enhance the occurrence of algal blooms unless counteracted by high flows. Further information is required on sources of salinity and their influx to the river and especially important is the relationship between river discharge and salinity. A large quantity of phosphorus originates from lowland catchment soils and comes from sub-surface sources. Thus efforts to minimise soil loss through bank slumping and channel erosion will have benefits. This will be assisted by development of riverbank (riparian) vegetation strips to intercept nutrients and to stabilise banks against erosion. Flow management is also important in reducing bank erosion of the main river channel, as is the provision of off-river stock watering facilities. Although fertilisers have not been identified as contributing significantly to nutrient loads it will be necessary to confirm and maintain this situation by utilising best management practices to minimise fertiliser loss from fields. Techniques to minimise nutrient supplies from point sources includes tertiary treatment or land disposal of sewage effluent and the interception and treatment of irrigation drains, stormwater returns and run-off from animal sheds. Improved management of some of these point and diffuse sources is already occurring but overall success will require adoption of a Total Catchment Management (TCM) program for nutrient reduction and water quality monitoring. Improved river clarity as a result of catchment management strategies reducing suspended sediment loads may improve the growth conditions for algal blooms, although counteracted to some extent by associated declines in nutrient delivery. These conditions may also support more extensive communities of macrophytes and attached forms of microalgae and cyanobacteria. Attached microalgae and cyanobacteria can compete for nutrients with the phytoplankton community and reduce the likelihood of large phytoplankton populations occurring. Further research is required on these interactions and the role of the benthic community. Similar outcomes might result from improvements in water clarity due to salinity increases but will be tempered by the sensitivity of the organisms to salinity. Information is required on salinity sensitivities and changes in the structure of the aquatic community due to increased salinity. The supply of nutrients from the bottom sediments is a difficult process to manage. Supply from this source can be minimised if depleted oxygen conditions are not allowed to develop near the water sediment interface. Increased flow will assist in transferring oxygen to the bottom sediments and reduce the occurrence of low oxygen conditions. Control of organic carbon loads can also help alleviate oxygen depletion but the conditions required for this to occur need further investigation. Under sustained low flow de-oxygenation of the bottom waters is likely and intervention techniques are required to minimise nutrient release. The results of the project demonstrate the hierarchical and complex interplay of factors that influence water quality and lead to the development of phytoplankton blooms in the Darling River. The project has described key processes responsible for the degradation of water quality and identified where management efforts need to be focussed to improve conditions. It is hoped that further studies will build on this research as we struggle to maintain our river systems into the future.