Aquatic ecosystem responses to flooding of the Menindee Lakes in 2004

"December 2004".

Project Number: Flood Monitoring of Deflation Basin Lakes MDBC Project R4013 M/BUS/59 AC155B.

MDFRC item.

This MDBC funded project (Project R4013), an adjunct to MDBC Project R1011 ('The ecology of Ephemeral Deflation Basin Lakes (EDBL)' examined aquatic responses to flooding of the Menindee Lakes during 2004. This enabled us to refine our conceptual model of EDBL function through wetting and drying cycles developed as part of Project R1011, thereby providing a more informed ecological basis for the development of water management strategies for this and other ephemeral ecosystems. The objectives of this investigation were firstly to examine post-flood responses of four deflation basin lakes within the Menindee Lakes complex, each differing in their hydraulic history, and secondly, to test key responses predicted by our conceptual model. The conceptual model predicted that flooding would; -lead to an increase in productivity stimulated by sediment and vegetation nutrient release, -increase invertebrate abundance through the provision of submerged vegetation that would provide important habitat, and -initiate an increase in the productivity of the fish community. The fact that the 4 lakes studied had different hydraulic histories provided an opportunity to examine the impact of drying history on lake's response to inundation. It was predicted that historical reductions in both the frequency and severity of episodes of lake drying would; -reduce the potential for nutrient mobilization from the sediments following inundation, thereby reducing the potential for primary production. and -affect the amount and type of organic matter on the lake bed which would have impacts on the type of habitat being provided for invertebrates. We also hypothesized that submerged terrestrial vegetation would constitute an important source of nutrients, and that this would stimulate aquatic primary production. The leeching of nutrients from inundated vegetation in Lake Malta was inferred by the development of thick epiphytic biofilms over the supporting vegetation. However, only net increases in water column phosphorus concentrations were observed, suggesting firstly that the epiphytes had a higher demand for nitrogen than for phosphorus, and secondly that less nitrogen was available to support phytoplankton growth. Whereas phytoplankton biomass in the absence of inundated vegetation decreased once sediment derived labile nitrogen and phosphorus pools became depleted, substantially more epiphyte biomass developed and persisted in response to the slower and more protracted release of nutrients from the inundated vegetation. In addition to its role as nutrient source, the presence of inundated vegetation was thought to provide potentially important aquatic habitat structure for a range of organisms, increasing the system's capacity to support greater organism biomass. Whilst no habitat effect was identified for zooplankton, some evidence was available to indicate that it did support greater macroinvertebrate abundances. As predicted, the magnitudes of the initial post-flood pulse in zooplankton abundance differed between lakes, with the lowest occurring in the most permanent lake, Lake Tandure. However, it was not possible to determine whether these reflected ‘real' differences in zooplankton productivity driven by hydraulically mediated differences in both food resource availability and re-colonization potential from the sediments, or ‘apparent' differences in zooplankton productivity driven by differences in predation pressure. Littoral benthic substrate composition changed along a gradient of decreasing lake permanence, with that of Lake Tandure consisting almost entirely of woody debris derived from fringing black box trees and that of the smaller more ephemeral lakes primarily of fibrous plant debris derived from inundated lakebed grasses and herbs. These differences in substrate composition accounted for much of the differences in benthic macro-invertebrate community structure encountered between the lakes, suggesting that hydrology had the capacity to indirectly modify lake trophic structure through its effect on particulate organic matter inputs. We predicted that the inundation of lakes would stimulate increases in lake fish populations, both as a consequence of immigration and recruitment following flood induced spawning, and that these increases would be linked with concomitant increases in the availability of potential food resources and floodplain spawning/nursery habitat. Fish catches in each of the four lakes investigated increased during the first month of inundation before declining towards the end of the study period. These increases coincided with observed increases in zooplankton abundances. Emigration of fish following subsidence of the post-inundation zooplankton pulse and avian predation likely accounted for the subsequent declines in fish catches. The ability of carp to adapt their spawning seasonality to local flow conditions is a key characteristic underpinning the success of this invasive species. Flooding during February stimulated a significant spawning event in carp and was followed by successively smaller spawning events at intervals of 5-6 weeks. Whilst the initial spawning event occurred with the rising limb of the flood pulse within the main channel, subsequent spawning also occurred within the lakes (Lake Malta). Lowest larval and juvenile abundances were recorded in Lake Tandure. This was a likely consequence of greater initial predation pressure by macroinvertebrates rather than a response to differences in lake drying history and food resource productivity. Although differences in growth rates between lakes were not examined, comparisons with data available for both mainstream and floodplain environments support our expectations of floodplains being more productive environments for fish. Differences in hydraulic regime and/or the severity of the most recent lake drying episode between lakes had demonstrable effects on aquatic primary and secondary productivity and on invertebrate community structure. Most of the post-inundation responses and differences between lakes were observed during the first month of inundation. However, within six months many of these differences among lakes had disappeared as seasonality began to play an increasingly important role. Our results provide evidence for the existence of links between the frequency and severity of lake drying episodes, nutrient releases from both sediments and inundated vegetation and post-inundation aquatic primary and secondary production. These links have important implications not only for management actions that reduce the frequency and duration of lake drying episodes, and thus also the time available for lakebed vegetation to develop, but also for the management of lakebed vegetation itself, which may be compromised by activities such as grazing or cropping. EDBL occur throughout the drier regions of the Murray-Darling Basin. Despite certain geomorphological similarities, key determinants of ecosystem function such as the prevailing hydrologic regime naturally vary greatly between lakes and have been modified through water resource development initiatives. Whilst reductions in the frequency and severity of lake drying events is widely considered to be a major degrading force (e.g. Bunn et al. 1997, Briggs 1998, Kingsford 2000a,b) and the re-imposition of drying phases has been recommended for a number of regulated wetlands within the Murray-Darling Basin (e.g. Briggs, 1988; Hydrotechnology, 1995), consideration needs be given to maintaining, or indeed increasing, the diversity of hydrologic regimes at the landscape scale. Such an approach is most likely to facilitate the persistence of species at that scale, by providing both temporal and spatial heterogeneity in the timing of developmental cues and refuges during dry periods.