Implications of pumping water on the ecology of Hattah Lakes

"April 2009".

Project Number: Implications of pumping water on the ecology of Hattah Lakes - M/BUS/210.

MDFRC item.

1 of 2 reports associated with project see (Implications of pumping and ponding water on water quality and the development of diverse aquatic ecosystems: Intervention Monitoring of the Hattah Lakes Icon Site 2006/07).

This report details the intervention monitoring undertaken at Hattah Lakes by the Murray-Darling Freshwater Research Centre in 2007/08 as part of The Living Murray program. The project investigated the ecological responses of river red gum, water quality, wetland vegetation and fish to the pumping of water to Hattah Lakes. The Hattah Lakes is part of the 48,000 ha Hattah-Kulkyne National Park located in the north west of Victoria. It contains 18 freshwater lakes (12 of which are Ramsar-listed) connected by a series of floodplain channels fed by the Murray River during periods of high flow. River regulation has threatened the ecological values of Hattah Lakes due to changes in the timing, duration and frequency of wet and dry phases in the lake system. Storage and extraction of water from the Murray River upstream of Hattah Lakes has resulted in the mean discharge in the Murray River near Hattah Lakes being ≈ 50% of natural. Natural overbank flooding has occurred on only one occasion in the past 11 years. Due to the reduced frequency of flooding, pumping has become a necessary management intervention to achieve inundation of the lakes in the current regulated and drought conditions. The pumping of water into Hattah Lakes from the Murray River occurred on four occasions during the period April 2005 to December 2006. It is expected to remain a long-term management intervention to maintain ecological values at the Hattah Lakes Icon Site. River red gum The greatest increase in river red gum (RRG) crown density (and condition) occurred in the ≈ 6-month period following pumping at Lake Mournpall (treatment), Brockie (treatment) and Yelwell (experimental control that did not receive pumped water). At least part of the increase across all lakes was attributed to the well-above-average rainfall during this period. For the ≈ 6-12 month and ≈ 12-18 month periods following pumping, mean crown densities of RRG have generally been maintained or have decreased at all three lakes. This was not expected, as the environmental watering was predicted to generate increases in crown densities at Lakes Mournpall and Lake Brockie for a longer period. Statistical analysis of the crown density profiles showed significant differences between the lakes that received pumped water and the control lake. These differences were due to a significantly lesser decline in crown density at Lakes Mournpall and Brockie during the ≈ 12-18 month period than occurred at the control Lake Yelwell. The lower rate of decrease in crown density at the treatment lakes may be due to these lakes receiving pumped water. However, it may in part be attributed to the control Lake Yelwell having a RRG population in poorer condition than the treatment lakes at the commencement of the study. Analysis of RRG response with distance from the water edge revealed that the greatest increases in crown density occurred for RRG located upon the lakebeds of the treatment lakes in the 12-month period following inundation. The bases of these trees were directly inundated from the pumped water. In contrast, the RRG located at higher elevations that were not directly inundated responded to a lesser degree. The findings suggest that RRG fringing the lakebeds may require direct inundation to achieve a greater increase in crown condition from environmental watering events. Water quality Water levels at Lakes Mournpall, Hattah, Bulla and Arawak decreased over the study period and held maximum water depths of ≈ 0.3 – 0.5m in mid-September 2008. As water levels decreased, turbidity increased and became more variable due to the increased susceptibility of the lakebed sediments to re-suspension from wind and wave action. This influenced euphotic depth in the lakes which typically ranged from 0.2 – 0.8m across the lakes. Electrical conductivity increased within the lakes as dissolved salts become more concentrated through evaporation but were less than 1,500 μS.cm-1 at the end of the study period. Surface water temperatures remained within the range of 10 – 30oC. Phytoplankton blooms comprising blue-green algae species were present at all four lakes during the study period. Total chlorophyll pigment exceeded 250 μg.L-1 and blue green algae counts exceeded 1 million cells.mL-1 at all lakes for at least one sampling occasion. The blue-green algae communities of each lake remained distinct to one another over the study period with communities changing in a cyclic manner most likely due to seasonal influences. A total of 20 blue-green algae taxa were identified within Hattah Lakes, with four potentially toxic species sampled including (in decreasing abundance) Microcystis aeruginosa, Anabaena spiroides f. spiroides, Cylindrospermopsis raciborskii and Anabaena circinalis. The water quality variables of pH and dissolved oxygen were variable and influenced by the photosynthesis within the water column. Day-time dissolved oxygen concentrations were typically supersaturated (>100% saturation) and pH was elevated (peak of 10.29) when phytoplankton and submerged macrophyte biomass was high. Wetland vegetation The lakebed vegetation communities of Lakes Brockie (treatment that received pumped water) and Lake Boich (control that did not receive pumped water) were monitored over a ≈ 21 month period to determine the effects of pumping upon lakebed vegetation communities. A total of 91 plant species were recorded at Lakes Brockie and Boich during the sampling period; of these 69 were indigenous (76%) and 22 were exotic (24%). These plant species were from ≈ 21 families, with Chenopodiaceae and Asteraceae containing considerably greater numbers of taxa than the other families. Before intervention, 31 plant species were recorded at Lake Brockie and 13 plant species were recorded at Lake Boich. After intervention, the number of plant species recorded at Lakes Brockie and Boich across all four sampling times was 53 and 58, respectively. Statistically significant differences in vegetation community between the two lakes at the High and Medium elevations were measured. However, no statistically significant difference was observed between the lakes over time. This contrasts with the analysis of only the Before and After 1 survey as reported in EPA and MDFRC (2008) and is due to the influence of the three additional After surveys in the 2007/08 analysis. Whilst statistically significant results were not recorded between lakes as a result of Lake Brockie receiving pumped water, there have been notable changes in the vegetation communities at Lakes Brockie and Boich. This is particularly evident when considering the plant functional groups at the lakes. The lakebed vegetation community at Lake Brockie changed from being predominantly terrestrial before inundation to one dominated by flood-responsive species, returning to a predominantly terrestrial lakebed community ≈ 19 months after pumping. A total of 15 flood-tolerant and one aquatic/semi-aquatic species were recorded at this lake following pumping. In contrast, the vegetation community at Lake Boich consisted of predominantly terrestrial species at four or the five sampling times. A total of two flood-responsive species were sampled at this control lake following pumping. Fish The series of pumping interventions at Hattah Lakes has resulted in the development of a significant fish community within the lakes contrary to initial expectations. However, the mismatch between the fish communities of the lakes (e.g. very low abundances of common carp, bony bream and absence of eastern gambusia) and the Murray River indicates that pumping has acted as a screen upon the fish community. In March 2008, a total of 8,248 fish representing four native (golden perch, carp gudgeon, flathead gudgeon and Australian smelt) and two non-native (goldfish and common carp) species were sampled from the four lakes. In June 2008, a total of 2,994 fish representing three native (carp gudgeon, flathead gudgeon, Australian smelt) and one non-native (goldfish) species were sampled from these same lakes using the standard fyke netting technique. The additional use of two gill nets in June 2008 at Lakes Mournpall and Hattah also sampled common carp (n = 17), golden perch (n = 9) and silver perch (n = 1). Native fish comprised 98.8% and 99.9% of the sampled fish in March 2008 and June 2008, respectively. Three small-bodied native fish consisting of carp gudgeon, Australian smelt and flathead gudgeon together comprised over 98% of the total catch for each time. The results of this study were compared with those from five earlier surveys following the commencement of pumping at Hattah Lakes. Whilst relative fish abundance has consistently increased between the first five surveys, the two surveys of 2008 demonstrate a decrease in both abundance and diversity of fish within the Hattah Lakes. This decrease is attributed to the reduction in aquatic habitat as the lakes become shallow and increased vulnerability of the fish community to predation by birds. Fish currently remaining in the lakes will perish by February 2009 if the lakes dry completely as expected. This will mark the third consecutive occasion since 1996 that fish communities have developed within Hattah Lakes but have not had an opportunity to move back to the Murray River and contribute to fish populations more broadly. The near-annual reconnection of Hattah Lakes with the Murray River under natural flow scenarios highlights the changed role of Hattah Lakes for fish due to river regulation and extraction. Management recommendations Based on the results of this project, the autho