The lower Broken Creek: aspects of water quality and growth of Azolla species

"December 2007".

Project Number: Broken Creek Azolla Research (phase 2) - M/BUS/211.

MDFRC item.

38 pages.

Background and project outline: Historically, the lower Broken Creek was an ephemeral creek with water derived from natural catchment flows. The construction of a series of small weirs and changes to flow regulations have altered the creek, with the lower section now comprising a series of weir pools. Rices Weir is the last weir before the creek enters the Murray River and is noted for having reasonably poor water quality. Poor water quality (low dissolved oxygen - DO) was considered a major factor that contributed to a fish death event in 2002. In addition, other consequences of anaerobic waters, namely, potential increase in hydrogen sulfide, ammonia and other products of microbial metabolism, also were considered as possible factors compounding the problems of low dissolved oxygen. A further characteristic of the Broken Creek is the extensive growth of the floating fern Azolla. The Azolla blooms can occur over many 100s of meters, up to kilometres and will have the capacity to greatly alter the ecological condition of the creek and to have a major impact on the overall water quality of the creek. A conceptual model to explain aspects of Azolla growth and water quality involves the fixation of carbon and nutrients by Azolla, but these are deposited into sediments on death, only to be recycled into water column, thus stimulating phytoplankton growth and subsequent growth of Azolla. This study addresses the following key questions to test the validity of the conceptual model: What standing biomass of Azolla in present in the Broken Creek? When does Azolla grow, at what rate does Azolla grow, and what factors drive the bloom and bust cycle? To what extent are phytoplankton important in the Broken Creek? How important are decomposition processes in the sediments? Do sediments play a role in determining water quality in the Broken Creek? What are the key water quality parameters and sediment conditions in the Broken Creek? Are water quality and sediment characteristics similar across different weir pools of the lower Broken Creek?.

Key findings: The amount of Azolla biomass per square metre was highly variable and depended on whether multiple layers of plants were present. Estimates derived for Rices Weir showed between 27 and 277 g (dry weight)/m2 with single layers of plants, which corresponded to 0.3 - 4.2 kg (wet weight)/m2. Azolla growth rates were approximately 8 g (dry mass)/m2/d, which is equivalent to approximately 120 g (wet weight)/m2/d. Phytoplankton abundance, (measured as chlorophyll-a concentration) was generally between 40-70 μg/L, but a major bloom (240 μg/L chlorophyll-a) occurred toward the end of the death phase of the Azolla. Oxygen profiles showed an oxygen gradient from surface to the sediment was present on almost all sampling occasions of the study and that on occasion, the DO at depth could be less than 1 mg/L. High sediment oxygen demand (SOD) were found throughout Rices Weir but the sediment oxygen demand increased up to 3 fold following the decline in Azolla biomass. Increases in SOD would then explain the decline in oxygen in the overlying water column. Immediately following the crash of the Azolla bloom (about 7th December 2007), oxygen levels in the surface layer remained relatively high, but there was clear decline in oxygen levels at depth (to less than about 2 mg O2/L). On the next sampling occasion there had been a significant decline in the oxygen throughout the water column again with the lowest levels observed at the sediment water interface. The oxygen levels found in Rices period at this time would be toxic to many native fish. Daytime oxygen levels on the 19th of December had returned to relatively high concentrations and this may be a consequence of the formation of a significant algal bloom. In addition to changes in oxygen levels, following the decline in Azolla biomass, there also an immediate increase in the dissolved phosphorus in the water column, followed by a slight delay by spikes in both ammonia and, to a lesser extent, nitrate. The pulse of nutrients corresponded to a massive algal bloom with the weir pool (with chlorophyll-a levels up to about 230 -240 μg Chl-a /L). A second, smaller bloom was observed in late January - early February. The December algal bloom also corresponded to the only time in the study period when gross primary production in the water column exceeded community respiration. Confirmation of the source of nutrients is required as persistent stratification would mean that diffusion of nutrients from sediment would be rather slow. It should be confirmed that nutrients were not from an upstream source. The overall picture that emerged from the cross-weir pool study was that sediment oxygen demand, carbon loads in the sediments, and water quality are generally similar across the different weir pools. These data combine to show that the overall ecological condition, and linkages between Azolla and water quality (particularly DO) are the same throughout the whole of the lower section of the Broken Creek. The sediments have a major role in determining the nutrient status of the Broken Creek. The system now undergoes regular growth and death cycles of Azolla and a large reserve of carbon and nutrient exists within the sediments. On occasions, the DO becomes critically low, and even when this does not occur, sediment respiration is sufficiently high to drive down the DO at depth –to a point where DO can be close to zero. Such an event occurred during and immediately after the death phase of the Azolla.

Ammonia levels in the creek were not high enough to be considered toxic. This study did not measure free sulfide in the water but sedimentary sulfur is relatively high and a role for sulfur in the poor water quality of Broken Creek can not be discounted. Be that as it may, O2 is probably the main factor likely to cause stress to fish. The ability for fish to survive low DO events will be determined by that rate that the decrease in O2 occurs e.g, if the rate of decline is sufficiently slow then the fish potentially have some capacity to move to other sections of the creek (upstream or down stream). The latter will also depend on the extent of the low DO in the creek as low DO is likely to be present over much of the weir pool. It is difficult to quantify the carbon loads exported to the Murray. There was some evidence of increase sediment carbon down stream following Azolla crash, however movement of Azolla is also linked to wind direction as much as flow. Wind events that essentially push the Azolla upstream are also likely to lead to increased levels of Azolla within Rices Weir, as the Azolla will continue to grow. The overall effect of the wind will be that Azolla export is likely to be very variable over short time frames, but when it does happen, large “slugs” of plant material may be moved down stream. Recommendations: Predicting low DO events. Regular monitoring of DO and flow manipulation still appears to be the most practical approach to maintain suitable DO in Broken Creek. In situ probes should continue to be the main monitoring tool, but the possible introduction of a photographic system to monitor the spread and condition of the Azolla could be considered. Daily visual evaluation of the Azolla will assist in determining extent and spread of Azolla and may help in predicting critical low DO events. The key factor responsible for the death of Azolla still remains to be determined. Although temperatures do get relatively high, they still do not reach levels considered to be catastrophic for Azolla. Ongoing temperature monitoring at fine scales may provide an answer as to why Azolla declines. Azolla removal. Simple removal of the Azolla may not necessarily represent an ideal short-term management strategy for the Broken Creek as large algal blooms will almost certainly follow any major manipulation of the Azolla. Long-term removal may be of some value, but the carbon and nutrient loads within the sediments are likely to lead to poor water quality for a long period of time. Azolla has very high growth rates and any removal will have to be in excess of the rates presented in this report.