Macroinvertebrate responses to altered low-flow hydrology in Queensland rivers Yemaya Smythe-McGuiness, Jaye Lobegeiger, Jonathan Marshall, Rajesh Prasad, Alisha Steward, Peter Negus, Glenn McGregor and Satish Choy Low flows report series, June 2012 NATIONAL WATER COMMISSION — Low Flows Report Series ii © Commonwealth of Australia 2012 This work is copyright. Apart from any use as permitted under the Copyright Act 1968, no part may be reproduced by any process without prior written permission. Requests and enquiries concerning reproduction and rights should be addressed to the Communications Director, National Water Commission, 95 Northbourne Avenue, Canberra ACT 2600 or email bookshop@nwc.gov.au. Online/print: ISBN: 978-1-921853-81-4 Published by the National Water Commission 95 Northbourne Avenue Canberra ACT 2600 Tel: 02 6102 6000 Email: enquiries@nwc.gov.au Date of publication: June 2012 An appropriate citation for this report is: Smythe-McGuinness Y et al. 2012, Macroinvertebrate responses to low-flow hydrology in Queensland rivers, National Water Commission, Canberra. Disclaimer This paper is presented by the National Water Commission for the purpose of informing discussion and does not necessarily reflect the views or opinions of the Commission or the Queensland State Government. NATIONAL WATER COMMISSION — Low flows report series iii Low flows report series This paper is part of a series of works commissioned by the National Water Commission on key water issues. This work has been undertaken by the Queensland Government on behalf of the National Water Commission. NATIONAL WATER COMMISSION — Low Flows Report Series iv Contents Executive summary Report context 1. Introduction 2. Background 2.1. Queensland low-flow characteristics 3. Pressure-Stressor-Response models 3.1. Dams and weirs 3.2. Water extraction 3.3. Interbasin water transfer 3.4. Disposal of coal seam gas industrial water into surface streams 3.5. Climate change 4. Case studies 4.1. Walla Weir case study 4.2. Wivenhoe Dam case study 5. Current monitoring relevant to low flows: issues and recommendations 5.1. Hydrology 5.2. Ecology 6. Discussion and conclusions Shortened forms References viii x 1 2 3 6 8 10 14 14 18 19 19 27 47 47 48 53 56 58 Tables Table 1: Pressure-Stressor-Response linkages of flow-related pressures identified in Queensland. The expected macroinvertebrate responses to each of the hydrological stressors are outlined, along with the case studies included in this report to test and to demonstrate the expected responses. ........................7 Table 2: Flow metrics expected for two sites in the Dawson River under current water resource development (ROP) and projected goal seam gas (CSG) flow scenarios (from McGregor et al. 2011). ......................................................................16 Table 3: Site numbers and names of macroinvertebrate sampling sites within the Burnett River catchment. ....................................................................................................21 Table 4: Time of sampling and habitats sampled within each site in the Burnett River catchment. Sampling was conducted on five occasions: before, during and after construction of the Walla Weir. Habitats sampled were E: edge, and R: riffle. ........................................................................................................................21 Table 5: Gauging station numbers and site names of macroinvertebrate sampling sites within the Brisbane River and Logan River catchments. ...........................................31 Table 6: Time of sampling and habitats sampled within each site in the Brisbane River catchment. R: riffle/run; E: edge; P: pool; and M: macrophytes. Blank spaces indicate no samples were taken (adapted from Choy et al. 2000). .......................33 Table 7: Results of pairwise comparisons between habitat types following twoway crossed ANOSIM (habitat x run, Global R = 0.191, p = 0.0043) using abundance category data. This tests the null hypothesis that there is no difference between habitats allowing for differences between runs. Shaded cells indicate pairs of habitats that were found to be significantly different (p < 0.05) ................................................................................................................................37 Table 8: Flow preferences and functional feeding group membership of significant macroinvertebrate taxa identified within results, according to the flow classification of Marshall and Marshall (in prep.). Flow preferences correlate with the flow requirements, larvae and adult habitat of taxa as outlined in Choy et al. (2000). ............................................................................................42 Table 9: The median Julian date of the one-day annual minimum flow for the Brisbane and Logan rivers for the periods before (pre-1997) and after (post1985) construction of the Wivenhoe Dam. .........................................................................43 NATIONAL WATER COMMISSION — Low flows report series v Table 10: Current DERM monitoring and research programs that may relate to low flows and ecology. Relevant programs are listed, as well as the location in which they are undertaken, the ecological indicators used, the frequency of activities, and how these activities apply to our knowledge of low-flow ecology. ..............................................................................................................................51 Figures Figure S1: Context of reports produced for the Low Flow Ecological Response and Recovery Project. Each circle represents the location of individual case studies and the size of each circle represents the spatial extent of each case study. .................................................................................................................................... x Figure 1: Flow regime is of central importance in sustaining the ecological integrity of aquatic ecosystems. The five components of the flow regime: magnitude, frequency, duration, timing and rate of change influence ecological integrity both directly and indirectly, through their effects on other primary ecological regulators. Modification of flow thus has cascading effects on the ecological integrity of rivers [source: Poff et al. 1997]. ..................................5 Figure 2: Hourly height and discharge data from GS113015A Tully River at Tully Gorge over a 10-day period, demonstrating fluctuations relating to hydroelectric water releases. ................................................................................................9 Figure 3: Surface water development status based on water diversion levels (as a proportion of the long-term mean of annual pre-development discharge) within National Land and Water Resources Audit (2002) catchment reporting units. ...................................................................................................................................11 Figure 4: Conceptual model of the ecological impacts of surface water extraction in conjunction with the impacts of dams and weirs in the case where they reduce downstream flows [source: DERM 2008]. ..............................................................13 Figure 5: Conceptual model illustrating alterations to hydraulic conditions and water quality resulting from coal seam gas wastewater disposal into surface waters [source: Takahashi et al. 2011]...............................................................................17 Figure 6: Locations of sampling sites, numbered according to gauging station, in the Burnett River catchment. The Walla Weir is situated between sites GS1360069 and GS136023A. ............................................................................................20 Figure 7: Flow-velocity-preference group structure for edge habitat samples at sites in the Burnett River in relation to the Walla Weir. Sample runs 1 and 2 refer to the pre-construction phase samples, 3 to the construction phase samples, and 4 and 5 to the operation phase samples. .....................................................23 Figure 8: Flow-velocity-preference group structure for riffle habitat samples at sites in the Burnett River in relation to the Walla Weir. Sample runs 1 and 2 refer to the pre-construction phase samples, 3 to the construction phase samples, and 4 and 5 to the operation phase samples. .....................................................24 Figure 9: Dendrograms illustrating the separation of macroinvertebrate communities based on site position from samples taken during the second operation phase of the Walla Weir within a) edge and b) riffle habitats. Samples were grouped for sites where samples were taken from each habitat type. * indicates the impounded area at Ben Anderson Barrage whereas other impounded areas were impounded by the Walla Weir. ..............................25 Figure 10: Map indicating the location of macroinvertebrate sampling sites within the Brisbane River and Logan River catchments. Site numbers are gauging station numbers. The Wivenhoe Dam is located at the base of Lake Wivenhoe. Red triangles in the Brisbane River catchment indicate regulated sites downstream of the Wivenhoe Dam, and blue triangles indicate unregulated sites. Green triangles in the Logan River catchment indicate upstream sites and yellow triangles indicate downstream sites. ........................................30 Figure 11: Ordinations showing the relationships between sampling periods (runs) within riffle/run habitat, based on a) macroinvertebrate abundance data allowing for differences between habitat types (stress=0.0000), and b) NATIONAL WATER COMMISSION — Low flows report series vi mean maximum flow velocity (stress = 0.0002) (adapted from Choy et al. 2000)...................................................................................................................................38 Figure 12: a) Ordination plot of average Faunal Abundance Category data with arrows showing direction and magnitude of differences between regulated (downstream) and unregulated (upstream) samples from riffle/run and edge habitats in the Brisbane and Logan rivers (stress = 0.1790). B – Brisbane River, L – Logan River, R – riffle/run, E – edge, r – regulated (downstream), u – unregulated (upstream). b) Selected significant faunal correlation vectors contributing to variation in macroinvertebrate community samples within the Brisbane River (HPSY = Hydropsychidae, UACA = Acarina, COEN = Coenagrionidae, VELI = Veliidae, ATYI = Atyidae, UCLA = Cladocera, CORI = Corixidae, COPE = Copepoda). c) Selected significant environmental correlation vectors explaining variance in macroinvertebrate communities within the Brisbane River (MeanSed = mean sediment sizes, logVEMA = log maximum velocity, logTOTN = log total nitrogen, logTURB = log water turbidity) (from Choy et al. 2000). .......................................................................................40 Figure 13: Schematic representations of the Logan and Brisbane rivers showing the proportions of macroinvertebrate functional feeding groups in samples from the wet season and dry season at upstream and downstream sites in each river (from Choy et al. 2000). .....................................................................................41 Figure 14: Comparison of median monthly flow volumes (ML day-1) in the Brisbane and Logan Rivers in the periods before construction of the Wivenhoe Dam on the Brisbane River (pre-1977) and after construction (post-1985). ........................................................................................................................43 Figure 15: Change in median monthly flow volumes (ML day-1) for the Brisbane and Logan rivers in the period following construction of the Wivenhoe Dam (post-1985) compared with the preceding period (pre-1977). * indicates a significant difference (P=0.05). ...........................................................................................44 Figure 16: Change in median low- and high-flow volume metrics for the Brisbane and Logan rivers in the period following construction of the Wivenhoe Dam (post-1985) compared with the preceding period (pre-1977). * indicates a significant difference (P=0.05). ...........................................................................................44 Figure 17: Ecological responses occur in reaction to the physical and chemical conditions experienced by biota, not directly to altered flow regimes. Altered flows influence ecology by interacting with various other properties of the setting to modify the conditions the biota experience. To understand ecological responses to flow modification one must first understand the relationships between physical/chemical conditions and the ecological response, and secondly understand the relationship between flow and the provision of these conditions. Stressors other than altered flow may also influence the response, so the effectiveness of flow regime restoration or protection cannot be judged directly by the occurrence or intensity of response. Rather, it must be judged by evaluating the provision of the flowrelated conditions necessary for the response to occur (Cockayne et al. 2010)...................................................................................................................................49 NATIONAL WATER COMMISSION — Low flows report series vii Executive summary This report’s purpose is to provide information on the role of low flows in shaping aquatic ecosystems, and how alteration of the naturally occurring low-flow regime can affect those ecosystems, specific to Queensland waterways. Low flows are an important influence in shaping aquatic ecosystems. As a result, aquatic biota are often adapted to the natural lowflow regime in the areas in which they occur. In many regions within Queensland low-flow or no-flow conditions are common and biota have developed strategies to enable them to survive annual dry seasons. This is in contrast with the impacts of low flows within more perennial systems where biota are not as well adapted to low flows. Uncharacteristic low-flow events within perennial systems can therefore act as a disturbance to biota and ecological processes. Changes away from natural low-flow conditions can negatively impact freshwater ecosystems – whether as an increase in the incidence of low flows within perennial systems not adapted to these conditions, or as a reduction in the frequency or duration of low flows within more ephemeral systems that are adapted to such conditions (especially when those changes exceed thresholds of biotic resistance and resilience). Development and use of water resources for urban, agricultural and industrial expansion have imposed a number of changes on natural flow regimes, including the low-flow regime itself, and these changes are often far beyond the range of variability experienced naturally. The main anthropogenic pressures within Queensland that alter the low-flow regime were identified as: barriers – dams and weirs, culverts water extraction disposal of excess industrial water interbasin transfers climate change. Using a Pressure-Stressor-Response framework approach, a conceptual understanding of the types of hydrological changes imposed by the above pressures and the likely responses from aquatic communities were outlined. Two case studies using existing departmental macroinvertebrate data in conjunction with new hydrological analyses were conducted to test conceptual understandings of macroinvertebrate responses to dams and weirs. The first case study investigated impacts of the Walla Weir on macroinvertebrate communities through flow alteration. Reduced flow velocity both within the weir impoundment where riffles were drowned out, and downstream where baseflows were reduced, led to a reduction in high-flow-preference macroinvertebrate taxa. Similarly, the second case study investigated the impacts of the Wivenhoe Dam on flow characteristics and resulting impacts on macroinvertebrate communities. The dam altered the hydrology of downstream reaches by decreasing flow variability and the duration and frequency of low-flow events and, in contrast with the Walla Weir, increased flows through downstream aquatic habitats that naturally experienced low-flow velocities. Shifts were observed in macroinvertebrate community composition, with a higher proportion of high-flow-preference taxa downstream of the dam, and altered functional feeding group composition. While these case studies gave some indication of the response of macroinvertebrate communities to alterations of the low-flow regime within Queensland aquatic ecosystems, difficulties in identifying general low-flow-ecology relationships were encountered because biota do not experience or respond directly to hydrology. Rather, flow interacts with other environmental features to produce physical, chemical and biological conditions which are perceivable to biota and which elicit or suppress particular ecological responses. Conceptual models helped show the stressors likely to be intermediating the impact of low-flow alterations NATIONAL WATER COMMISSION — Low flows report series viii on aquatic biota. However, a pervasive lack of suitable data pertaining to low lows prevented these conceptual understandings being tested further than was possible within the two case studies. A lack of data relating to biota, hydrology and the impacts of altering flow conditions therefore inhibited a thorough assessment of low-flow-ecology relationships within Queensland. An assessment of current management practices found that ecological problems relating to low flows may be detected and dealt with through existing monitoring and research in instances where they are locally relevant to anthropogenic pressures and ecological values. There are, however, several difficulties associated with assessing the effect of low flows on ecology based on existing hydrological and ecological data. To better inform flow management in future and assess the impacts of altering the low-flow regime on aquatic biota, we advocate developing conceptual understandings of the flow-related conditions required for biological responses to be identified. Ecological risk assessment frameworks, where the risk to specific values is linked to ecological responses with critical requirements for flow conditions, can be used to account for both the potential influence of confounding stressors and the complex relationships between flow, habitat and ecology. NATIONAL WATER COMMISSION — Low flows report series ix Report context This report is part of a larger series of reports produced for the National Water Commission’s Low Flow Ecological Response and Recovery Project (Figure S1). This report presents one of 11 hydro-ecological case studies. The purpose of the case studies is to test hypotheses that relate ecological process and function and biological traits to key hydrological measures that are affected by low flows. A summary of the findings in this report and the other case studies are contained in Synthesis of case studies quantifying ecological responses to low flows (Marsh et al. 2012). Guidance on ecological response and hydrological modelling for low-flow water planning Low-flow hydrological classification of Australia Review of literature quantifying ecological responses to low flows Early warning, compliance and diagnostic monitoring of ecological responses to low flows Synthesis of case studies quantifying ecological responses to low flows Figure S1: Context of reports produced for the Low Flow Ecological Response and Recovery Project. Each circle represents the location of individual case studies and the size of each circle represents the spatial extent of each case study. NATIONAL WATER COMMISSION — Low flows report series x 1. Introduction This report was prepared by the Queensland Department of Environment and Resource Management (DERM) to provide state-specific information relating to the ecological significance of low flows, as well as the potential ecological consequences of flow regulation strategies, to inform improved water planning and management. Its approach is based on the Pressure-Stressor-Response (PSR) conceptual model framework as explained in Marshall et al. (2006) and is similar to the Driving forces-Pressure-State-Impacts-Response framework used internationally (e.g. US Environmental Protection Agency and EU Water Directive) and nationally (e.g. State of the Environment reporting). The PSR framework has been used in several other ecological studies conducted by DERM to identify and explain how Queensland’s aquatic ecosystems respond to particular human activities (i.e. pressures) and the resulting physico-chemical changes (i.e. stressors) and subsequent biological changes (i.e. responses) to the aquatic environment. For the purposes of this report, only the pressures that relate to hydrology and particularly low flows have been considered. A broad definition for ‘low flows’ has been applied here, encompassing a range of conditions including baseflow, no-flow and completely dry. It is the change away from the natural flow regime that is considered important, so the pre-existing conditions in a region will determine what is ‘low flow’. Further, it is identified that dry spells are not necessarily an impact – many systems are adapted for precisely these regimes – so in some cases additional flow, and thus the loss of ‘low-flow conditions’, is investigated. Key flow regulation actions (pressures) within Queensland have been identified. Environmental alterations (stressors) caused by the pressures within freshwater ecosystems are outlined, along with the resulting ecological impacts (responses) within these systems. Where possible, case studies based on existing biotic data collected by DERM and new hydrological data are used to provide empirical illustration of these linkages and, where empirical data is lacking, future monitoring is recommended. The results of case studies are used to illustrate the ecological significance of low flows, along with the potential ecological consequences of flow regulation strategies within Queensland – in accordance with the aims of the Low Flow Ecological Response and Recovery Project. Current DERM water planning projects operating within Queensland were assessed to consider how they include low-flow requirements, and what future practices may be needed to better acquire data and incorporate understandings of low flows within current water planning and management. Findings can then be used to investigate the ecological significance of low flows within a broader context, and inform guidelines regarding water developments. NATIONAL WATER COMMISSION — Low flows report series 1 2. Background Water resource managers face the ever-increasing challenge of balancing human demand for water with environmental requirements (Naiman et al. 2002; Poff et al. 2003; Khan 2008). Human interference has altered the natural conditions of rivers and streams worldwide, and often in irreversible ways (Malmqvist & Rundle 2002; Naiman et al. 2002; Nilsson et al. 2005; Morton et al. 2009; Walters & Post 2011). Freshwater ecosystems provide goods and services of critical importance to human societies, yet they are among the most heavily altered environments (Geist 2011). Alteration of natural flow regimes can have profound effects because the natural flow regime, in synergy with water quality, are both essential attributes for healthy aquatic ecosystems (Bunn & Arthington 2002; Olden & Naiman 2010; Geist 2011). The ecological significance of flow characteristics to aquatic biota are well documented (Poff et al. 1997; Marshall et al. 2001; Bunn & Arthington 2002; Lloyd et al. 2004; Dewson et al. 2007; Leigh & Sheldon 2008; Monk et al. 2008; Naiman et al. 2008). Flow characteristics directly control in-stream water quality and habitat conditions, and mediate biological processes (Figure 1), such as spawning and the dispersal of water-dependent biota (Poff et al. 1991; Dewson et al. 2007; Naiman et al. 2008). Low flows are an important aspect of the overall flow regime, but are less studied compared with other aspects of the flow regime (Rolls et al. 2010). Biologically significant aspects of the low-flow regime are a subset of the characteristics of the wider flow regime, which include the magnitude, frequency, duration and timing of flow events; the rate of change during a flow event (Figure 1); and the degree to which these characteristics are temporally and spatially variable (Poff et al. 1997; Monk et al. 2008; Richter et al. 1998; Kennard et al. 2010; Rolls et al. 2010). The water resource planning process within Queensland is designed to balance human and environmental water requirements. The process is prescribed by the Water Act 2000, which is aligned with the requirements of the National Water Initiative and relevant state and regional strategies (Whittington 2000; DERM 2011a). Water resource plans have been undertaken for all but a few catchments within Queensland, and are required to include water requirements for future human consumptive and non-consumptive needs, as well as environmental needs. The water allocation process for each catchment involves socio-economic and environmental assessments, as well as community consultation and scientific review panels, to establish the water requirements of all sectors involved (Whittington 2000). A water resource plan is then developed which becomes the legal basis for the establishment of local water operations within the specified water planning area (DERM 2011a). Knowledge of natural flow conditions and the corresponding ecology within an area is needed to understand the consequences of altering the flow regime and inform effective management (Kennard et al. 2010; Rolls et al. 2010). Scientific assessments inform water resource plans by determining the flow requirements of ecological assets such as in-stream and riparian biota. This study aims to contribute to the understanding of low-flow ecology so that consideration of low flows may be better incorporated into water resource planning and management. This study also seeks to assess the degree to which current water resource management within Queensland includes consideration of low flows, and determine how this process can be improved. Alterations to low flows can involve changes to low-flow characteristics over space and time. Alterations to any of these flow aspects can cause changes to in-stream water quality and habitat conditions, and affect in-stream biota (Figure 1) (Poff et al. 1997; Dewson et al. 2007; Leigh & Sheldon 2008). Changes to the natural low-flow regime can be measured by comparing hydrological data before and after flow modification. Where such data is not available, a reference approach can be used whereby a nearby site with a similar flow regime, NATIONAL WATER COMMISSION — Low flows report series 2 which is in a relatively natural state, can be used to infer antecedent conditions (Choy et al. 2000; Whittington 2000; Marshall et al. 2001). Low-flow indices used to measure hydrological changes in the low-flow regime within this study are based on the flow classification indices used by Kennard et al. (2010). A preliminary low-flow classification performed by Mackay et al. (2012) built on the work of Kennard et al. (2010) using only those flow indices relevant to low flows. The authors identified six of these low-flow indices as the most useful in defining low-flow classes. Similar low-flow indices have been used within this study to measure changes in the low-flow regime, and include measures of the magnitude, timing and variability of low-flow discharge over a number of temporal scales. Details of these indices are outlined within the case studies in this report. Aquatic macroinvertebrates were used in this report as biological indicators to investigate the ecological impacts of altering the low-flow regime. Aquatic macroinvertebrates are often used to assess waterway health and gauge anthropogenic impacts on waterways (Chessman 1995; Growns et al. 1995; Metzeling et al. 2003; James & Suren 2009; Chessman et al. 2010). Aquatic macroinvertebrates are suitable biological indicators of waterway health for a number of reasons: 1) they are sensitive to changes in flow, water quality and habitat condition, and may therefore act as indicators of overall ecological health; 2) individuals tend to have low mobility and so local impacts can be more accurately detected; 3) they are important components of aquatic foodwebs, and are therefore indicative of wider aquatic ecological processes, and 4) they are ubiquitous, relatively easy to sample, and are well documented (Choy & Marshall 2000). Aquatic macroinvertebrate community composition has been demonstrated to be responsive to varying flow conditions within Queensland and is a common aquatic biological indicator used by DERM (e.g. Choy & Marshall 1999a; Marshall et al. 2000; Marshall et al. 2001). 2.1. Queensland low-flow characteristics The total land area in the Queensland jurisdiction is just under 1.74 million km 2 (ABS 2010). Queensland is under the influence of a variety of climatic conditions consisting of low rainfall and hot summers in inland western areas, a monsoon season in the north, warm temperate conditions along the eastern coastal area and low minimum temperatures inland and around the southern ranges (BOM 2011). As a result, the rainfall patterns, runoff and hydrology differ greatly between different parts of the state. Flow conditions in Queensland range between two extremes: the northern region is tropical wet while the central and central west region is arid dry. A large number of streams are characterised by variable and intermittent flow (Kennard et al. 2010; sensu Larned et al. 2010; Mackay et al.2012). Streams in Queensland have been classified according to their low-flow characteristics as predominantly ephemeral (Mackay et al. 2012). This is comparable with the flow classification by Kennard et al. (2010) which was based on a wider range of flow characteristics. Kennard et al. (2010) classified the overall natural flow regime of rivers and streams in the northern and north-western parts of Queensland as either ‘stable summer baseflow’ or ‘predictable summer highly intermittent’. Rivers and streams from the central-east to the south-east were classified as either ‘unpredictable intermittent’ or ‘unpredictable summer highly intermittent’. In the south-west, west and central west, rivers and streams were classified as ‘predictable summer highly intermittent’, ‘unpredictable summer highly intermittent’ or ‘variable summer extremely intermittent’. According to Mackay et al.'s (2012) 35 metric classification, a large number of Queensland's streams have highly ephemeral (class 6) or moderately ephemeral (class 5) low-flow regimes NATIONAL WATER COMMISSION — Low flows report series 3 - many of these streams are located in the state’s inland areas. Other low-flow classes represented include marginally ephemeral (class 3) and weakly perennial streams (class 2) with a predictable low-flow period, which tend to be located near or in coastal areas, particularly in the state’s north-east. Low-flow-related stressors are expected to impact areas in different ways, depending on the existing natural flow regime (Kennard et al. 2010; Rolls et al. 2010). Such large variability in flow conditions across Queensland presents a wide range of scenarios and implications when low-flow patterns are altered. For example, the same level of water extraction will affect aquatic ecosystems in a perennial wet tropics stream in northern Queensland differently to an ephemeral stream in the state’s west. This is due to underlying differences in hydrology, habitat conditions and ecological characteristics, which are determinants of the aquatic ecosystems’ ability to tolerate and recover from disturbances to the low-flow regime (Kennard et al. 2010; Rolls et al. 2010). Ephemeral channels in central and western Queensland often experience extended no-flow periods followed by episodic high-magnitude flows and flooding. Systems fluctuate between highly connected floodplains and highly disconnected waterholes (Mackay et al.2012). Aquatic biota within ephemeral systems are adapted to these conditions, and have developed flexible physiological and lifecycle characteristics that enable them to survive such large fluctuations in flow conditions and habitat availability (Poff et al. 1997; Bunn & Arthington 2002; Dewson et al. 2007; Takahashi et al. 2011). Many aquatic species occurring within ephemeral streams in Queensland depend on flow-related cues for critical life-history activities, many of which are associated with low-flow events. Many species use drought refugia, including waterholes and hyporheic zones; or have life-history stages that can resist desiccation in dry river beds (Marshall et al. 2005; McGregor et al. 2011; Steward et al. 2011). These characteristics, which allow biota to survive unpredictable and often harsh conditions within ephemeral streams, also mean they are sensitive to flow modifications that alter these conditions (Bunn & Arthington 2002). If flow conditions were altered so that the natural patterns of variability were lost, flow-related cues for life-history activities may not occur. For example, species with desiccation-resistant life-history stages may depend on periods of low or no flow for a part of their lifecycle, and if flows were altered to disrupt the timing and frequency of low-flow events, these species could be disadvantaged. In contrast, aquatic biota within perennial streams in the north and coastal regions of Queensland are adapted to constant and less variable flow conditions (Kennard et al. 2010; Rolls et al. 2010). If flow conditions were altered such that flow variability increased – with more low-flow events – biota within perennial streams would be less well adapted to cope with such events when compared with biota within naturally ephemeral streams. Water resource developments that alter the natural flow regime can therefore affect important hydrological cues and habitat conditions on which biota depend. This has the potential to significantly disadvantage native aquatic biota and impact the structure and function of aquatic ecosystems within Queensland (Thoms & Cullen 1998; Bunn & Arthington 2002; Richardson & Humphries 2010; McGregor et al. 2011; Takahashi et al. 2011). Disturbed conditions may also promote invasion by exotic species (Koehn 2004; Takahashi et al. 2011). The direction and magnitude of impact caused by water resource developments will vary depending on the nature of the pressure, the existing natural flow regime, and corresponding ecological characteristics of native biota. This report investigates the nature of water resource developments within Queensland and the direction of impacts expected as a result of flow alteration from these developments, with emphasis on alterations to low-flow characteristics. The case studies in this report investigate NATIONAL WATER COMMISSION — Low flows report series 4 the direction and magnitude of impact to aquatic biota caused by flow alteration from dams and weirs in the context of two streams within south-east Queensland. Figure 1: Flow regime is of central importance in sustaining the ecological integrity of aquatic ecosystems. The five components of the flow regime: magnitude, frequency, duration, timing and rate of change influence ecological integrity both directly and indirectly, through their effects on other primary ecological regulators. Modification of flow thus has cascading effects on the ecological integrity of rivers [source: Poff et al. 1997]. NATIONAL WATER COMMISSION — Low flows report series 5 3. Pressure-Stressor-Response models Water resource developments in Queensland have increased considerably during the past 50 years (DSEWPaC 2009). Medium to high levels of development are expected to significantly affect aquatic ecosystems through flow alteration and resulting habitat modification (Bunn & Arthington 2002); however, even relatively low levels of development can lead to local impacts depending on the location and nature of the alteration. The water resource planning process within Queensland has an important role in minimising the environmental impacts of water resource developments, now and in the future. Five main pressures relating to flow alteration have been identified as ecologically significant within Queensland, based on water resource planning studies conducted by DERM and the literature (e.g. Choy & Marshall 1999b; Choy & Marshall 2000; Choy et al. 2000; DSEWPaC 2009; Takahashi et al. 2011). Main flow-related pressures within Queensland include: 1) dams and weirs, 2) water extraction, 3) disposal of excess industrial wastewater into surface streams, 4) interbasin water transfer, and 5) climate change (Table 1). The PSR models for each of the five pressures are summarised in Table 1 and discussed below. The impacts of dams and weirs on in-stream biota as a result of flow alterations have been documented in a number of studies, including reports produced by DERM (e.g. Choy & Marshall 2000; Choy et al. 2000) and therefore can be demonstrated by in-depth case studies. Two case studies are presented on the impacts of Walla Weir and Wivenhoe Dam. Case studies were not conducted for the other identified pressures due to a lack of data, but the possible implications of these pressures based on the findings of other studies are discussed within PSR models. Coal seam gas water-disposal is a new emerging pressure and, as such, data relating to its effects in the environment is not yet available. Modelling and experimental approaches are being used to determine the magnitude and direction of likely impacts (see McGregor et al. 2011; Takahashi et al. 2011) and initial findings from these studies are briefly discussed in the PSR model. Similarly, with the pressure of climate change, empirical flow-ecological data was not available because the most significant impacts are likely to occur in the future. While climate change may have already had some effect on aquatic ecosystems, the relatively gradual nature of the change means very long-term datasets are required to detect trends and separate confounding effects of climatic variability. As such, the current and future effects of climate change and coal seam gas water-disposal on aquatic ecosystems require further research before PSR models can be tested. Water extraction pressure occurs frequently in Queensland at a range of scales, and is considered to be a particular threat to aquatic environments such as refugial waterholes and riffle habitats during baseflow conditions. The ecological impacts of water extraction have been studied in some areas of Australia (e.g. McKay & King 2006; Boulton et al. 2003; Nebel et al. 2008) but few studies quantify the associated flow alterations. Thus little information is available to describe the location, timing and magnitude of water extraction occurring in the state, and for this reason a case study could not be conducted. With regard to interbasin water transfer: while it is likely to cause altered flow conditions in both the source and recipient systems, the more critical – and better studied and reported – threats of interbasin transfer relate to the spread of exotic species and implications of translocation on the genetic viability of native aquatic biota populations (e.g. Davies et al. 1992; Marshall et al. 2000; Page et al. 2010). As such, case studies demonstrating the ecological impacts of water extraction and interbasin water transfer in Queensland are not provided due to a lack of data specific to the flow-related pressures and stressors caused by these pressures, however their possible implications (based on the findings of other studies) are discussed within the PSR models. NATIONAL WATER COMMISSION — Low flows report series 6 Table 1: Pressure-Stressor-Response linkages of flow-related pressures identified in Queensland. The expected macroinvertebrate responses to each of the hydrological stressors are outlined, along with the case studies included in this report to test and to demonstrate the expected responses. PRESSURE Expected macroinvertebrate RESPONSE Case study Inundation within impoundment, and low/no-flow below wall. Local loss of sensitive and riffle taxa, lower taxa richness, reduced opportunities for recolonisation (longer distances between habitats). Walla Weir, before & after construction Low flow/ baseflow Increased frequency and duration of medium/high flow. Loss of riffle taxa, change in flow-preference groups in pools, change in functional feeding group composition. Wivenhoe Dam operation Variable/seasonal flows replaced with stable or unseasonal ones Seasonality and wider flow variability Change from variable to stable flow. Increase in taxa richness and sensitive taxa, change in flow-preference groups. Wivenhoe Dam operation Hydroelectric releases interrupt baseflows with frequent, brief, high-flow pulses Baseflow Increased frequency and magnitude of high-flow events with steep rates of change. Fewer low-flow-preference taxa and a shift to those that have adaptations for high flow (e.g. hooks and suckers). Possible loss of shredder functional feeding group if leaf litter is flushed. Reduced flow leads to loss of riffles and other flowing habitats, reduced water quality Baseflow Increased frequency and duration of low and no flow. Loss of riffle taxa and sensitive taxa. Increase in tolerant or opportunistic taxa. Change in functional feeding groups, e.g. loss of filterer. Loss of surface water through extraction from pools Presence of pool refugia Change from no flow to dry. Loss of aquatic taxa that do not have desiccation-resistant stages. Some may disperse to find refuges or suitable habitat. Colonisation of the dry streambed by terrestrial taxa. Interbasin transfer Loss of surface water in source streams, increased flows in recipient streams, transfer of biota within water Low flow/ baseflow Reduced seasonality and frequency of low flow in recipient system, reduced flows in source system. Possible loss of riffle taxa and/or sensitive taxa, loss of taxa dependent on low flows in recipient streams, aquatic pests may be introduced to recipient streams. Disposal of industrial water (coal seam gas) Additional water into temporary streams No flow/ low flow Reduced frequency and duration of no and low flows or change from ephemeral to perennial. Loss of invertebrates that require desiccation or benefit from it. Possible increase in taxa richness and sensitive taxa. Increase in taxa that prefer high flow. Climate change Overall reduction in flows Baseflow Increased frequency and duration of low-flow events. Loss of high-flow-preference taxa, increase of low-flow and no-flow-preference taxa. Increased frequency of extreme events, i.e. floods and droughts Baseflow Increased frequency of no-flow and/or high-flow events. Increase in tolerant or opportunistic taxa. Dams and weirs Water extraction STRESSOR Vulnerable facet of Likely hydrological hydrology alteration Barriers and impoundments lead to loss of riffles through drowning or drying Baseflow Dam releases replace low flows with medium/high flows NATIONAL WATER COMMISSION — Low flows report series 7 3.1. Dams and weirs 3.1.1. Pressure Australia is largely dependent on water storage reservoirs given its highly variable precipitation patterns and frequent long-term droughts (Ghassemi et al. 1995). Dams and weirs secure water for irrigation, drinking water supply and hydroelectricity production (Ghassemi et al. 1995; Arthington & Pusey 2003; Richter & Thomas 2007). As of the year 2000, approximately 446 large dams stored 88 000 GL of water in Australia (Kingsford 2000). At present 183 major water storages are located in Queensland, with a total capacity of 13 389 GL, not including privately owned off-stream storages (DSEWPaC 2009). 3.1.2. Stressors A number of studies have illustrated the ecological impacts of dams and weirs (Brunke et al. 2001; Bunn & Arthington 2002; Nilsson et al. 2005; Almeida et al. 2009; Mueller et al. 2011). Dams and weirs change the natural flow regime, modify the physical properties of upstream and downstream habitats, and act as barriers to in-stream dispersal (Choy et al. 2000; Nilsson et al. 2005; Almeida et al. 2009). The impounded water upstream of dams and weirs forms large and deep lentic systems, leading to a loss of riffles, runs, shallow pools, backwater pools, shallow edges and other habitats. Impounded water also becomes stratified in terms of temperature and dissolved oxygen (Olden & Naiman 2010). Downstream of dams and weirs, controlled release of water changes the flow regime from its natural state. All aspects of the flow regime can be altered, including the magnitude, frequency, duration and timing of flow events, the rate of change during a flow event, as well as the degree to which these characteristics are temporally and spatially variable (Brunke et al. 2001; Bunn & Arthington 2002). Patterns of water release from dams and weirs vary. In some cases, water is released from impoundments only periodically (Olden & Naiman 2010). As a result, baseflow is reduced and high flows occur within a short time period, followed by long periods of low or no flows. Reducing baseflow in this way can alter in-stream habitat availability by reducing the number of riffles and runs, lowering water levels within pools, reducing the downstream movement of woody debris, and increasing the distance between in-stream habitats in areas where flow ceases (Pusey & Arthington 2003; Olden & Naiman 2010). The ecological impacts of dams and weirs in the case where they decrease downstream baseflow are summarised in a conceptual model ( Figure 4), shown in conjunction with the impacts of water extraction in Section 3.2. In other cases there is a continual release of water from dams and weirs, which increases baseflow and removes flow variability. This acts to deepen existing pools and drown out riffles, and increase the rate of nutrient movement downstream, thereby altering habitat availability. In addition, the release of stratified water can alter downstream water quality, although this is dependent on the method and timing of release (Olden & Naiman 2010). In cases where flow is regulated for hydroelectricity generation, water is released in frequent high-magnitude short-duration pulses. This significantly increases the frequency of high-flow events, with a steep gradient of change between baseflow and high flows. For example, in the Tully River in the Wet Tropics, daily releases lead to regular water level fluctuations downstream (Figure 2) (DERM 2011b). This can act as a frequent disturbance to in-stream conditions by scouring and causing siltation of the streambed, and can increase the rate of downstream movement of organic material, therefore limiting in-stream habitat availability. Reduced organic material, along with possible changes in fluvial geomorphology because of altered sediment transport patterns, can limit habitat and nutrient availability within aquatic systems below dams NATIONAL WATER COMMISSION — Low flows report series 8 and weirs (Gregory et al. 1991). Aquatic plants and animals may also have difficulty establishing or maintaining their positions under such rapidly fluctuating conditions. 1.4 Height 4000 Discharge 3500 1.2 3000 1 Height (m) 2000 0.6 1500 Discharge (ML/day) 2500 0.8 0.4 1000 0.2 500 10/06/2010 9/06/2010 8/06/2010 7/06/2010 6/06/2010 5/06/2010 4/06/2010 3/06/2010 2/06/2010 0 1/06/2010 0 Figure 2: Hourly height and discharge data from GS113015A Tully River at Tully Gorge over a 10-day period, demonstrating fluctuations relating to hydroelectric water releases. 3.1.3. Responses Modifications to flow characteristics and habitat conditions caused by dams and weirs can negatively impact riparian and in-stream biota within impounded and downstream areas, causing altered community composition and hence disrupting ecosystem function (Kingsford 2000; Brunke et al. 2001; Bunn & Arthington 2002; Döll et al. 2009). By acting as barriers to in-stream dispersal and increasing the distance between areas of suitable habitat by drowning or drying out habitats, dams and weirs can significantly affect the population viability of aquatic biota. Similarly, regular high-flow velocities from hydroelectricity releases may also impede the upstream movement of biota who swim or crawl along the stream bed, potentially leading to a loss of connectivity between populations. Lowered habitat availability and loss of specific habitat types will impact biota dependent on these habitats, and is likely to favour generalist species with wide habitat preferences. For example, loss of riffle habitat can affect biota dependent on shallow high-flow-velocity habitat such as filter-feeding aquatic macroinvertebrates that require high-flow velocity to collect food from the moving water column. Similarly, increased flows through pools can impact biota dependent on pool habitat such as aquatic macroinvertebrate ‘shredders’ that break down large organic debris, some of which are dependent on still water conditions where leaf litter can accumulate (Thoms & Cullen 1998; Bunn & Arthington 2002; Richardson & Humphries 2010). The constant disturbance of the streambed caused by the increased frequency of high flows in the case of hydroelectricity releases may cause a loss of taxa sensitive to physical disturbance, and benthic biota dependent on stable streambed habitat. NATIONAL WATER COMMISSION — Low flows report series 9 Reduced availability of coarse organic material through loss of downstream transport or increased rate of transport through a reach can also impact aquatic community composition by reducing the quantity of organic material settling onto the streambed, such as leaf litter and woody debris. For example, a shift in the functional feeding group composition of macroinvertebrate communities would be expected, with a loss of taxa belonging to the ‘shredder’ functional feeding group that consumes coarse organic material. An overall reduction in nutrient availability would significantly impact trophic structure and ecosystem function. Where baseflows are increased below dams and weirs, or where high flows from hydroelectricity releases regularly interrupt baseflow conditions, macroinvertebrate taxa with a preference for low flows may be disadvantaged. High-flow-preference taxa with physiological and/or morphological adaptations to survive such conditions (e.g. hooks or suckers for attaching to substrate) will be favoured. Similarly, in streams adapted to natural periods of low and/or no flows, increased baseflows and reduced flow variability will disadvantage biota adapted to or dependent on these conditions (e.g. species with desiccation-resistant lifecycle phases that rely on dry benthic habitat). Reduced flow variability will also lead to the loss of flow-related lifecycle cues related to dispersal and reproduction, hence disadvantaging a potentially large number of native species. Loss of variability in the flow regime could lead to an overall loss of aquatic biota with flexible life-history characteristics, with the consequence that ecosystems would lose resilience. If flow regimes were to resume their natural patterns, biota may no longer be adapted to survive natural seasonal extremes and unpredictable flow patterns (Choy et al. 2000). As a result of water stratification, impounded or released water may differ in physical and/or chemical properties to natural conditions. For example, it may be cooler or lower in dissolved oxygen. The result of water impoundment and release could therefore be to disadvantage taxa sensitive to water quality parameters, and disadvantage taxa tolerant of a wider range of water physical and chemical conditions. The impact of dams and weirs may therefore be an overall loss of sensitive taxa that depend on specific natural habitat and water quality conditions, and an increase in generalist taxa with high disturbance tolerances (which often include exotic species). The effect of changes to in-stream habitat and water quality conditions resulting from altered flow conditions caused by dams and weirs can therefore reduce the abundance of taxa dependent on specific in-stream habitats, flow conditions, or in-stream dispersal, which may lead to the loss of affected taxa in reaches impounded by and below dams and weirs. Overall aquatic community composition and diversity can therefore be affected within these systems, leading to altered competitive interactions and foodweb dynamics, and ultimately disrupting ecosystem function (Gregory et al. 1991; Bunn & Arthington 2002). 3.2. Water extraction 3.2.1. Pressure Water is extracted from surface and groundwater sources within Australia for consumption and to irrigate crops and water stock. It was estimated for the year 1996–97 that 2970 GL of water was diverted from surface water sources in Queensland and 1623 GL extracted from groundwater sources, which is likely to have significantly increased since then (DSEWPaC 2009). In dryland areas, considerable water is also harvested from floodplains during floods (ANRA 2000). Additional water extraction occurs on privately owned land for stock and domestic use, however these levels of extraction are not well documented. Surface water extraction in Queensland is highest within catchments in the state’s south-east (Figure 3). There is also considerable surface water use in the Condamine and Border Rivers basins, which flow into the Murray-Darling Drainage Division. There are eight major irrigation areas throughout the state. Areas of high water extraction in Queensland do not correspond with NATIONAL WATER COMMISSION — Low flows report series 10 high rainfall areas. For example, there is high mean annual flow into the Gulf of Carpentaria but little water extraction in this region at present (DSEWPaC 2009). This pattern of mismatched supply and demand puts considerable strain on water resources within high-use regions. Figure 3: Surface water development status based on water diversion levels (as a proportion of the long-term mean of annual pre-development discharge) within National Land and Water Resources Audit (2002) catchment reporting units. The sustainable annual yield of groundwater for Queensland is estimated to be 2784 GL (DSEWPaC 2009). Current statewide levels of groundwater use are below this sustainable level, but levels of groundwater extraction within each discrete management unit are not sustainable in all cases. Thirty-three per cent of management units experience extraction levels that are above sustainable. The greatest yields were available in the Great Artesian groundwater management unit, with 1017 GL allocated per year. Levels of groundwater development in this unit were classified as low (less than 30 per cent of sustainable yield) to moderate (40 to 70 per cent of sustainable yield). Allocations in the Tasman groundwater management unit, which spans much of the state’s east coast, were estimated at 435 GL per year, with levels of development from low to moderate. Levels of groundwater development within south-east Queensland were classified as moderate. A total of 78 GL of groundwater was extracted per year within the Clarence-Morton groundwater management unit located in the south-east (DSEWPaC 2009). Growing industries, such as mining or manufacturing, can cause short- or long-term demands for water extraction. Past experiences indicate that demands for water from urban, industrial and agricultural users will increase. Unmanaged growth in demands will place considerable strain on available water resources (ANRA 2000). 3.2.2. Stressors Surface water and groundwater extraction can significantly alter flow characteristics within affected streams (McKay & King 2006; Döll et al. 2009). Surface water extraction can reduce baseflows and increase the incidence and magnitude of low flows and alter their timing, NATIONAL WATER COMMISSION — Low flows report series 11 producing ‘artificial drought’ conditions within surface streams and connected hyporheic zones (Richter et al. 1998; Boulton 2003; Deitch et al. 2009; Finn et al. 2009). Groundwater extraction affects natural groundwater levels, which means that greater quantities of water are required for aquifer recharge. There are many linkages between groundwater, hyporheic and surface water zones, and extraction from either surface or groundwater sources will consequently impact all three zones (Boulton et al. 2003; Hancock & Boulton 2005). Groundwater extraction therefore has the potential to increase the duration and frequency of low- and no-flow periods in surface streams. Water extraction from temporary streams during the dry season can cause increased incidence and duration of no-flow periods. Extreme water extraction within perennial streams can generate no-flow periods in systems not adapted to them (Mackay et al. 2012; Marsh et al. 2012). Water extraction from surface streams reduces total water volume and water depth, and this may cause narrowing of the stream channel (Richter et al. 1998; Finn et al. 2009). Reduced water volume represents a loss of total aquatic habitat, and can also alter the types of habitats present. Narrowing of the stream channel reduces the wetted area, which lowers hyporheic habitat availability (Hancock & Boulton 2005; Stubbington et al. 2009). Loss of riffle habitats can result from decreases in water discharge, and pools are likely to become the predominant habitat type in such circumstances. This in turn may alter water quality, as stagnant pools are subject to stratification and changing temperature, nutrient and ionic conditions. Still water also promotes algal growth, which is associated with increased diurnal variability in dissolved oxygen levels (Finn et al. 2009). Extended no-flow periods affect the persistence of refugial waterholes (DERM 2010b), increasing the distance between available aquatic habitats (Pusey & Arthington 2003). Water evaporation during no-flow periods can concentrate in-stream nutrient and salinity levels. Benthic sediment build-up may also occur due to a lack of flushing, which can further decrease water quality by increasing in-stream turbidity and nutrient levels, and reduce the availability of benthic habitats (Hancock & Boulton 2005). Reduced area and volume of surface water reduces recharge and exchange with the hyporheic zone and groundwater storages (Boulton et al. 2003; Hancock & Boulton 2005). Weakened exchange between the stream and hyporheic zone can starve the interstitial environment of oxygen and promote reducing conditions, altering the nutrient dynamics of surface stream ecosystems (Hancock & Boulton 2005). High levels of water extraction from ephemeral streams can reduce lateral connectivity between river channels and associated floodplains, and decrease the frequency of longitudinal connectivity between refugial waterholes. This can severely impact nutrient availability within affected systems and limit dispersal opportunities for aquatic biota (Gregory et al. 1991; Thoms & Cullen 1998; Allibone 2000; Marsh et al. 2012). 3.2.3. Responses The increased incidence, magnitude and duration of low- and no-flow periods resulting from water extraction can severely impact aquatic ecosystem function, and a decline in aquatic biodiversity is expected (Naiman & Latterell 2005; Rolls et al. 2010). Within perennial ecosystems not well adapted to low and no flows, many aquatic species may lack the physiological and life-history characteristics needed to survive such conditions. Within temporary systems, biota are likely to be more resilient to the impacts of water extraction, and impacts are dependent on their levels of tolerance, which will vary between taxa (Poff et al. 1997; Bunn & Arthington 2002; Dewson et al. 2007). Similarly, alterations to water quality may cause in-stream conditions to deteriorate beyond the tolerance levels of some species, particularly the sensitive taxa such as Plecoptera (stonefly), Ephemeroptera (mayfly) and Trichoptera (caddisfly) (PET) (Gregory et al. 1991). In general, water extraction is likely to more heavily impact biota dependent on high-flow velocities, such as filtering macroinvertebrate species (Thoms & Cullen 1998; Bunn & Arthington 2002; Richardson & Humphries 2010). NATIONAL WATER COMMISSION — Low flows report series 12 Loss of habitat availability and connectivity resulting from water extraction also has the potential to significantly impact aquatic ecosystems. The overall reduction in habitat availability resulting from water extraction concentrates individuals within smaller areas, heightening biotic interactions. High mortality can result within contracted aquatic ecosystems, as the biotic pressures of competition, predation and disease are increased (Arthington et al. 2010). The loss of longitudinal connectivity inhibits the dispersal of aquatic biota, therefore reducing gene flow within and between affected populations (Allibone 2000; Mackay et al. 2012). Any of the above impacts can lead to a loss of taxa within affected streams, and hence cause lowered aquatic biodiversity, disrupted foodweb dynamics, and lowered ecosystem productivity and function (Gregory et al. 1991; Poff et al. 1997; Bunn & Arthington 2002; Arthington et al. 2010). Altered habitat conditions may also favour exotic species (Koehn 2004). Lowered lateral connectivity within temporary systems as a result of reduced high-flow and increased low-flow events could also impact nutrient availability within streams, interfering with primary production pathways, and hence disrupting foodweb dynamics and related properties such as system energetics and carrying capacity (Gregory et al. 1991; Thoms & Cullen 1998). The ecological impacts of surface water extraction, in conjunction with the impacts of dams and weirs in the case where they reduce downstream flows, are summarised in the conceptual model below ( Figure 4). At present no general conceptual models have been developed for the ecological impacts of groundwater extraction. Figure 4: Conceptual model of the ecological impacts of surface water extraction in conjunction with the impacts of dams and weirs in the case where they reduce downstream flows [source: DERM 2008]. NATIONAL WATER COMMISSION — Low flows report series 13 3.3. Interbasin water transfer 3.3.1. Pressure The Queensland Government has developed several large-scale initiatives for the state’s southeast collectively known as the Water Grid. The Water Grid connects water supplies, storages and treatment plants across south-east Queensland: from Noosa in the north to Coolangatta in the south, and to Toowoomba in the west. This includes the transfer of water between different basins in south-east Queensland. Water in the Wivenhoe Dam, in the Brisbane River catchment, is now supplied to Toowoomba via the Wivenhoe-Toowoomba Pipeline. Similarly, the Southern Regional Water Pipeline connects the Brisbane River catchment with the desalination plant at Tugun on the Gold Coast; and the Northern Pipeline Inter-connector links water between the Sunshine Coast and Brisbane (Queensland Water Commission 2010). Interbasin transfer schemes also exist in other parts of the state, including the Pioneer irrigation area near Mackay (Marshall et al. 2000) and the Mareeba-Dimbulah irrigation area near Cairns (Ryan et al. 2002). These ‘interbasin water transfers’ can cause changes to the flow regime, and pose a risk to aquatic ecosystems (Snaddon et al. 1998; Page et al. 2010). 3.3.2. Stressors Interbasin water transfers affect source and recipient rivers differently (Rolls et al. 2010). Flows in the recipient basin are supplemented, potentially reducing the duration of no- and low-flow periods and increasing the duration of medium to high flows. In extreme cases, interbasin water transfers could transform recipient rivers from ephemeral into perennial systems (Negus 2007). In contrast, flows in the source basin are depleted and high and medium flows may be reduced (Davies et al. 1992), with an increased frequency of low flows. The flow-related stressors and responses in source rivers are expected to be similar to those experienced as a result of water extraction (see Section 3.2). The flow-related stressors and responses in recipient rivers are expected to be similar to those experienced by constant water releases downstream of dams and weirs (see Section 3.1). 3.3.3. Responses In addition to the responses shown in the water extraction and dams and weirs PSR models, interbasin water transfer is also expected to affect source rivers by translocating biota within transferred water. This may act to create pest species problems within recipient rivers. Exotic pest species or diseases, which occur within source rivers but were not previously found within recipient rivers, may be transferred. Native aquatic species found within source rivers may also become pests within recipient rivers if they were not previously found there, and even co-existing species can be affected by introducing individuals from differing genetic source populations (Davies et al. 1992; Gibbins et al. 2000; Marshall et al. 2000; Page et al. 2010). 3.4. Disposal of coal seam gas industrial water into surface streams 3.4.1. Pressure Coal Seam Gas (CSG) is a natural gas adsorbed onto coal deposits and kept in place by pressure from surrounding groundwater. Extraction of CSG is achieved by dewatering coal seams to reduce the constraining pressure and release gas from the coal. This produces large amounts of wastewater which then requires disposal. CSG extraction is a relatively new industry in Queensland, and poses a new type of water resource issue. Whereas in the past water NATIONAL WATER COMMISSION — Low flows report series 14 resource management has typically been concerned with the allocation and storage of naturally occurring surface and groundwater supplies, the problem of excess water is relatively new. The CSG industry is rapidly expanding over large areas of Queensland, and this has prompted the need to develop appropriate guidelines for the disposal of wastewater. CSG production activities are expected to steadily increase until about the year 2018, at which time the industry is expected to produce up to 130 GL of wastewater a year (McGregor et al. 2011; Takahashi et al. 2011). DERM released its CSG Water Management Policy in June 2010, in which the preferred option for managing CSG wastewater was treatment followed by either injection into aquifers or beneficial use. It is possible, however, that instances may occur, especially in the short-term, where there is excess wastewater unable to be disposed of via the preferred methods. It has been proposed that in such cases, excess water could instead be released into surface waters. In such cases, strategies would be required to ensure water quality objectives are met and ecological values of recipient streams are protected (McGregor et al. 2011; Takahashi et al. 2011). Due to the novel nature of the CSG industry in Queensland, the impacts of wastewater release into surface waters are not yet completely understood, and empirical data is lacking. The Australian Government has allocated research funding to ensure that possible environmental impacts resulting from the CSG industry are understood and are kept below acceptable thresholds of environmental change. DERM has and continues to contribute to this research (McGregor et al. 2011; Takahashi et al. 2011) including a study to inform biological monitoring (Takahashi et al. 2011), and development of guidelines for managing flow regimes within CSGaffected systems (McGregor et al. 2011). These two components were used to inform this PSR model relating the likely impacts of CSG to flow alterations. The state’s current CSG resources are concentrated in and around the Queensland portion of the Murray-Darling Basin (QMDB) in the south-west. Freshwater systems in this area include the Border Rivers, Moonie, Balonne-Condamine, Warrego, Paroo and Bulloo drainage basins. Most of Queensland’s CSG resources underlie the Condamine-Balonne drainage basin in the QMDB and the Fitzroy drainage basin located to the QMDB’s north (McGregor et al. 2011; Takahashi et al. 2011). Natural flow regimes within the QMDB were described as being unpredictable and/or intermittent according to Kennard et al. (2010). The predominant flow class within the QMDB was ‘unpredictable summer highly intermittent’ (flow class 11). Only one naturally perennial stream was identified in the QMDB (Kennard et al. 2010; McGregor et al. 2011). 3.4.2. Stressors Takahashi et al. (2011) used a PSR approach to investigate the likely biological impacts of CSG wastewater introduction into surface streams, and to inform appropriate monitoring guidelines. They determined that changes can be expected to all of the important facets of the natural flow regime, including flow duration, timing, variability, predictability, magnitude, and rate of rise and fall. Two main flow-related stressors were identified: 1) alteration to hydrology leading to a decrease in dry spells, and 2) alteration to hydrology leading to constant flow and decreased seasonality. The introduction of CSG wastewater into surface streams therefore has the potential to alter all aspects of the low-flow regime, particularly by decreasing the magnitude and frequency of low flows, and altering the timing and variability of low flows. Within the QMDB, increased baseflow through CSG wastewater introduction has the potential to change ephemeral streams to perennial ones (McGregor et al. 2011). Flow changes will alter in-stream habitat availability, with modifications in riffle-pool sequences and loss of dry river beds (Bunn & Arthington 2002; Cockayne et al. 2010; McGregor et al. 2011). McGregor et al. (2011) modelled flow data based on water resource availability under current resource operations planning (ROP) and compared this to flow data from projected CSG water releases for the Dawson River, NATIONAL WATER COMMISSION — Low flows report series 15 Queensland. Results indicate key changes in habitat and flow characteristics likely under current CSG projections (Table 2). Table 2: Flow metrics expected for two sites in the Dawson River under current water resource development (ROP) and projected goal seam gas (CSG) flow scenarios (from McGregor et al. 2011). Location Flow metric ROP scenario CSG scenario Baroondah % time as a riffle 71 94 % time dry 24 0 % time drowned 5 6 Average duration (days) riffle-forming flows 19 85 Max. duration (days) of riffle-forming flows 390 873 Average dry spell (days) 8 0 Max. dry spell (days) 148 0 % time as a riffle 27 86 % time dry 60 0 % time drowned 13 14 Average duration (days) riffle-forming flows 8 53 Max. duration (days) of riffle-forming flows 87 426 Average dry spell (days) 45 0 Max. dry spell (days) 415 0 Taroom 3.4.3. Responses Aquatic biota are adapted to the temporary flow regime of streams in the QMDB and are therefore likely to be affected by the additional flow generated by CSG water releases. Increased streamflow and the reduced incidence of no-flow and low-flow spells are expected to negatively impact species reliant on dry or low-flow conditions for parts of their lifecycle, as well as increase the potential for invasive species to colonise affected systems (Koehn 2004; Takahashi et al. 2011). Dry river beds are unique habitats colonised by a range of terrestrial biota, and these biota will be displaced if this habitat is artificially inundated by CSG water (Steward et al. 2011). Altered flow conditions resulting in more permanent availability of aquatic habitat and altered light availability are also expected to increase macrophyte growth in these areas (Takahashi et al. 2011), which will alter production pathways within affected systems (Gregory et al. 1991). Loss of flow seasonality is expected to impact aquatic species dependent on flow-related cues for reproduction and dispersal. Loss of native species may occur, resulting in altered species richness, community composition, foodweb dynamics and overall ecosystem function (Gawne & Scholz 2006; Takahashi et al. 2011). Impacts resulting from flow alteration will be further compounded by stressors resulting from altered water quality as a result of CGS wastewater (Takahashi et al. 2011). Likely alterations to hydraulic conditions and water quality were summarised by Takahashi et al. (2011) within a conceptual model ( Figure 5). The magnitude and level of risk posed by each of these stressors and responses needs to be measured in areas where this pressure occurs to gain an accurate insight into the significance of this pressure (Takahashi et al. 2011). NATIONAL WATER COMMISSION — Low flows report series 16 Figure 5: Conceptual model illustrating alterations to hydraulic conditions and water quality resulting from coal seam gas wastewater disposal into surface waters [source: Takahashi et al. 2011]. NATIONAL WATER COMMISSION — Low flows report series 17 3.5. Climate change 3.5.1. Pressure Australian water resource management needs to incorporate the risk of climate change, which threatens to significantly alter future water availability (Khan 2008). Climate change in Queensland is predicted to cause annual mean temperatures to increase, annual precipitation to decrease, and annual runoff to decrease (Chiew & McMahon 2002; CSIRO & BOM 2007; DERM 2010a). Climate change predictions also indicate an increase in extreme events, such as floods and droughts (Hennessy et al. 2007; CSIRO & BoM 2007; DERM 2010a). 3.5.2. Stressors The increased rates of evaporation due to higher temperatures, lower runoff and the increased occurrence of droughts may result in a reduction in the magnitude, frequency and duration of streamflows. Reductions in river flow may be represented by a decrease in the duration of natural high to medium flows, and an increase in the duration of low-flow and no-flow events (Chiew & McMahon 2002). This will lead to a loss of aquatic habitats, particularly high and medium flow habitats such as riffles, and reduced persistence of aquatic habitat such as refugial waterholes during the dry season (Bunn & Arthington 2002). Many of the stressors and responses will therefore be similar to the impacts of water extraction (see the water extraction PSR model in Section 3.2). There may also be an increase in the frequency of extreme high-flow events, which could significantly disturb stream morphology and sedimentation rates, and alter in-stream habitat stability and availability. 3.5.3. Responses Changes in flow and habitat conditions are likely to cause a shift in the community composition of aquatic biota (Poff et al. 1997; Marshall et al. 2001). The availability of suitable habitats for taxa with preferences for high or medium flows may be reduced, whereas habitat for taxa with no flow preference, or a preference for low flow, may be increased. Reduced availability of refuges during droughts would affect aquatic biota with limited dispersal capabilities or low resistance to desiccation (Bunn & Arthington 2002). Extreme flood events could further disrupt community composition and ecosystem function by disrupting habitat availability, particularly taxa dependent on stable benthic habitat, similar to the impacts of dams and weirs in the case of hydroelectricity releases (see dams and weirs PSR model in Section 3.1). NATIONAL WATER COMMISSION — Low flows report series 18 4. Case studies 4.1. Walla Weir case study 4.1.1. Introduction This case study analyses biological responses resulting from construction of the Walla Weir, located on the Burnett River in south-east Queensland. Results suggest low-flow ecological linkages, which are expressed via changes to habitat characteristics, are important within this region’s riverine ecosystems. The Burnett River catchment includes the Bundaberg Irrigation Area, and is classified as highly developed, with 70 to 100 per cent of mean annual discharge diverted for use. Aquatic ecosystems within the river are therefore expected to have undergone a high level of modification as a result of these developments (Stazner & Higler 1986; Bunn & Arthington 2002). The Walla Weir is an impoundment on the Burnett River, built for the purpose of supplying irrigation water to the South Burnett region. Its construction was completed in 1998. Impounded flows and controlled releases are expected to have altered habitat conditions and impacted aquatic biota downstream and immediately upstream of the Walla Weir (Rolls et al. 2010; Bunn & Arthington 2002). The ecological impacts of the Walla Weir’s construction and operation were assessed when construction was proposed. This included an analysis of the impacts on aquatic macroinvertebrate communities upstream and downstream of the weir. The study comprised several reports that assessed the effects on macroinvertebrate communities before and during construction, as well as during operation (see Choy 1997a, b; Choy & Marshall 1999a, b, 2000). These studies investigated the implications of flow alteration on macroinvertebrate communities within the Burnett River. We intended, for the purposes of this case study, to couple the previous analysis of macroinvertebrate responses to the Walla Weir from the time of its construction with a new hydrological analysis of the nature of flow alterations caused by the weir since its construction – with a focus on low flows – as was conducted in the Wivenhoe Dam case study. Unfortunately, the gauging station located below the weir (GS 136008A) ceased operation during January 2000, and no other active gauging stations are present within the Burnett River downstream of the weir. This means that adequate data were not available to assess flow characteristics downstream of the weir in the period after its construction. Flow-ecological relationships were therefore only able to be inferred based on the responses of macroinvertebrates documented in the aforementioned studies. Based on the general PSR conceptual model for dams and weirs, a number of expectations about the impacts of the Walla Weir on macroinvertebrate communities were formed. As a result of the impoundment of water and release patterns designed to meet irrigation demand, it was expected that riffle habitats would be drowned out above the weir and dried out below the weir at times and for durations that differ from the pre-development case. Because the availability of riffle habitat is thus likely to have been highly modified, riffle taxa were expected to be highly impacted, with reduced abundance of riffle-dependent taxa, and lowered diversity within remaining riffle habitats. With respect to hydrology, it was expected that flow downstream of the weir would change from natural conditions, with a reduction in the frequency and duration of all flows, including low flows. It was expected that the frequency and duration of cease-to-flow periods would increase and the timing of events would be altered to match irrigation demand periods rather than rainfall patterns. 4.1.2. Methods The methods used in the macroinvertebrate flow requirements analysis conducted by Choy and Marshall (2000) are summarised below, as relevant to this report. A full account of the methodology is given in the relevant reports (Choy 1997a, b; Choy & Marshall 1999a, b, 2000). NATIONAL WATER COMMISSION — Low flows report series 19 Study sites Study sites were chosen within the Burnett River catchment to represent sites that would be situated upstream, downstream and impounded by the Walla Weir after its construction. A total of nine study sites were sampled in the Burnett River catchment: three downstream, two impounded, and four upstream of the Walla Weir ( Figure 6; Table 3). All sites were located on the Burnett River except the site Currajong Creek at Wallaville (GS1360076), which drains into the river. One of the downstream sites, Burnett River at Cedars Crossing (GS1360077), while on the Burnett River, was located in another impoundment: the Ben Anderson Barrage. Figure 6: Locations of sampling sites, numbered according to gauging station, in the Burnett River catchment. The Walla Weir is situated between sites GS1360069 and GS136023A. NATIONAL WATER COMMISSION — Low flows report series 20 Table 3: Site numbers and names of macroinvertebrate sampling sites within the Burnett River catchment. Position relative to the Walla Weir Gauging station number Site name Downstream 1360077 Burnett River at Cedars Crossing Downstream 1360078 Burnett River at Drinans Crossing Downstream 1360069 Burnett River at Walla Weir tailwater Impounded 1360076 Currajong Creek at Wallaville Impounded 136001B Burnett River at Walla Upstream 1360071 Burnett River at Booyal Crossing Upstream 136007A Burnett River at Figtree Creek Upstream 136012A Burnett River at Mingo Crossing Upstream 136002D Burnett River at Mt Lawless Sampling Sampling was conducted on five occasions: twice before construction, once during construction, and twice during operation of the Walla Weir. Each site was sampled between three and five times (Table 4). Aquatic macroinvertebrates were sampled from two habitat types – riffles and edges – where present at each site. Wherever both edge and riffle habitats were present, macroinvertebrates were sampled from both habitats. Riffle habitats were not sampled in impounded sites during the operation of the weir because they were flooded and therefore no longer present (Table 4). Similarly, only one downstream site (GS1360069) had riffles present during operation of the Walla Weir. Samples were live-picked on-site and macroinvertebrates stored in ethanol. Macroinvertebrates were enumerated and identified in the laboratory, mostly to family level. Table 4: Time of sampling and habitats sampled within each site in the Burnett River catchment. Sampling was conducted on five occasions: before, during and after construction of the Walla Weir. Habitats sampled were E: edge, and R: riffle. Sampling period Site Pre- Pre- construction construction Construction Operation Operation number Site position (Oct ‘96) (May ‘97) (June ‘98) (May ‘99) (Nov ‘99) 1360077 Downstream E E E E E 1360078 Downstream E, R E, R E E E 1360069 Downstream R E, R E, R E, R 1360076 Impounded R E, R E, R E E 136001B Impounded R E, R E, R E E 1360071 Upstream E, R E, R E, R 136007A Upstream R E, R E, R E, R E, R 136012A Upstream E, R E, R E, R E, R E, R 136002D Upstream E, R E, R E, R E, R Data analysis Analyses were used to assess the response of macroinvertebrate communities to assumed changes in habitat and flow conditions resulting from the construction and operation of the Walla Weir. Analysis of the flow-velocity-preference group structure of macroinvertebrate communities at each site was performed, based on the flow-preference allocation of Marshall and Marshall (in prep.). Multivariate NATIONAL WATER COMMISSION — Low flows report series 21 analyses were based on Bray-Curtis similarity matrices of the abundance of all taxa present. Cluster analyses were conducted separately for edge and riffle habitats and used to create dendrograms – to see if samples collected during the last sampling period (post-construction) separated according to site position with respect to the weir. 4.1.3. Results Flow-velocity-preference group structure In the impounded sites, only the edge habitats could be sampled during operation of the Walla Weir because riffle habitats were no longer present after it was built. No high-flow-preference taxa were found in the impounded sites during operation of the weir. These sites had high-flow-preference taxa present in both edge and riffle habitats before the weir was constructed (Figure 7, Figure 8). In edge habitats upstream of the Walla Weir, taxa from all categories of flow preference (i.e. high-, low-, and no-flow-velocity preference) were present. In contrast, sites downstream of the weir had fewer high-flow-preference taxa, with either small proportions of, or no high-flow-preference taxa present (Figure 7). This was, however, the case both before and after construction of the weir. There was little change in the proportions of flow-preference-taxa present in edge habitats before and after construction of the weir. In the riffle sites upstream of the Walla Weir, high-flow-preference taxa were present on all sampling occasions, but there were only small proportions of, or no low-flow-preference taxa present (Figure 8). In contrast, there were higher proportions of low-flow-preference taxa present at the site (GS1360069) immediately below the weir on all sampling occasions. This difference between upstream and downstream sites was especially pronounced when the weir was in operation (sampling runs 4–5), where there were no low-flow-preference taxa present in upstream sites. There was, however, little change in flow-velocity-preference group structure in riffle habitats at downstream sites before and after construction of the weir. NATIONAL WATER COMMISSION — Low flows report series 22 Figure 7: Flow-velocity-preference group structure for edge habitat samples at sites in the Burnett River in relation to the Walla Weir. Sample runs 1 and 2 refer to the pre-construction phase samples, 3 to the construction phase samples, and 4 and 5 to the operation phase samples. NATIONAL WATER COMMISSION — Low flows report series 23 Figure 8: Flow-velocity-preference group structure for riffle habitat samples at sites in the Burnett River in relation to the Walla Weir. Sample runs 1 and 2 refer to the pre-construction phase samples, 3 to the construction phase samples, and 4 and 5 to the operation phase samples. Cluster analysis Bray-Curtis measures of similarity in edge samples from the second post-construction period grouped sites into distinct groups of impounded and non-impounded sites (Figure 9). The group of impounded sites was more than 50 per cent different from other sites. This indicates that impoundment of habitats by the Walla Weir significantly altered macroinvertebrate edge assemblages. Within the impounded group, samples collected from sites above the weir were more similar to each other than the sample from the site impounded by the Ben Anderson Barrage (GS1360077). The site impounded by the barrage was approximately 36 per cent different from other sites, indicating that impoundment by the weir had a larger impact on macroinvertebrate assemblages than impoundment by the Ben Anderson Barrage. NATIONAL WATER COMMISSION — Low flows report series 24 Bray-Curtis measures of similarity in riffle samples indicated that macroinvertebrates collected from the one downstream site in which riffles were present after weir construction were significantly different from upstream sites. The difference between the downstream and upstream sites was more than 50 per cent (Figure 9). No other riffle samples were present in downstream or impounded sites. This indicates that flow alteration by the weir not only altered the distribution of riffle habitats in affected sites, but also significantly modified macroinvertebrate riffle assemblages within the remaining riffles. a) b) Figure 9: Dendrograms illustrating the separation of macroinvertebrate communities based on site position from samples taken during the second operation phase of the Walla Weir within a) edge and b) riffle habitats. Samples were grouped for sites where samples were taken from each habitat type. * indicates the impounded area at Ben Anderson Barrage whereas other impounded areas were impounded by the Walla Weir. 4.1.4. Discussion Bray-Curtis measures of similarity in macroinvertebrate community composition from edge habitats grouped the samples into two distinct groups: an impounded group and an upstream/downstream group. When this result is compared with an earlier analysis by Choy (1997b) on macroinvertebrates sampled in the same sites before the Walla Weir was constructed, there is an indication of change. The pre-construction results (Choy 1997b) showed that all of the sites (except one) in the vicinity of the proposed weir had similar taxa present; whereas post-construction results (Figure 9) showed that macroinvertebrate composition at impounded sites had become different from non-impounded sites. It is important to note, however, that the earlier analysis by Choy (1997b) included both edge and riffle samples in Bray-Curtis measures of similarity, whereas the latter analysis presented in this study used only edge samples – because after the weir was constructed there were no riffle habitats present within impounded sites. Some of the difference between the two sampling periods may therefore be attributable to this difference in data inclusion. In the pre-construction results the site that differed from the other sites near the proposed weir was the site impounded by the Ben Anderson Barrage (identified as Bingera Weir in Choy 1997b) – indicating this impoundment had altered community composition at this site before the weir was built. After construction of the weir, macroinvertebrate communities at this site were more similar to other impounded sites than non-impounded sites, despite being located downstream of the weir. This indicates that impoundment altered macroinvertebrate community composition in a consistent way. NATIONAL WATER COMMISSION — Low flows report series 25 The composition of macroinvertebrate fauna based on Bray-Curtis measures of similarity from edge habitats in sites impounded by the Walla Weir were different from those at non-impounded sites. Choy and Marshall (2000) attributed these differences to a change in reach characteristics in the impounded area from riverine to lacustrine. This is likely to reflect a change from flowing water to still water, with macroinvertebrates responding to a loss of flow velocity in impounded habitat. This is supported by analysis of the composition of macroinvertebrate fauna based on flow-velocity preferences from edge habitats, which also indicated that sites impounded by the Walla Weir became different from upstream sites after the weir was built, with a loss of high-flow-velocity-preference taxa. Analysis of riffle habitat taxa post-construction showed that the site immediately downstream of the weir (GS1360069: Walla Weir tailwater) was distinct from other sites upstream, in terms of macroinvertebrate community composition. A 50 per cent difference in taxa between this site and other riffle sites indicated the riffle at the tailwater site had been impacted by the weir (Choy & Marshall 2000). This change may relate to either of two stressors. The isolation of this riffle from all others by impoundment may have deprived it of macroinvertebrate colonists, particularly those that use stream drift as a means of dispersal, leading to a different resident fauna. The likely alteration to the frequency, timing and duration of low- and no-flow spells affecting this riffle may also have led to the loss of riffle species sensitive to this stressor. However, there is no direct evidence to support either mechanism, particularly in the absence of post-construction hydrology data. The presence of high-flow-preference taxa in the edge habitats at sites upstream of the impounded area of Walla Weir, and within impounded sites before construction of the weir, indicated that highflow-preference taxa were present in the parts of the river where flow was natural. This is because natural flow conditions provide the flow velocity required by these species within their edge habitat. The lack of high-flow-preference taxa within impounded sites after the weir was built could therefore be directly attributed to a loss of flow velocity as a consequence of the altered flow regime imposed by the weir. Season can be ruled out as a contributing factor, as high-flow-preference taxa were present at upstream sites on all occasions. The flow-preference structure of macroinvertebrate taxa in the riffle at the Walla Weir tailwater site, just downstream of the weir wall, was expected to be similar to that of riffles upstream of the impounded area due to similar habitat conditions. However, the proportion of low-flow-preference taxa was higher than proportions observed at upstream sites on all sampling occasions. This indicates that some level of difference between downstream and upstream sites exists naturally. However, the difference in macroinvertebrate flow-preference structure between upstream and downstream sites became more pronounced after construction of the weir. This was further evidence that the weir had altered the flow-velocity-preference structure of macroinvertebrate communities. The likely stressors that caused such a response could be reduced flow velocity and deteriorated quality of the water released from the weir (possibly hypolimnetic water and/or water with low dissolved oxygen and/or water being low in temperature). Once again, however, there is no direct evidence to support either mechanism. In a PSR context, the Walla Weir is the pressure in this instance. There are several hydrological stressors emanating from the pressure, and they could be acting either individually or in combination at various times. The stressors are: The conversion of a riverine system into a lacustrine waterbody, with no flow velocity in the weir pool. Alteration of the natural low-flow regime downstream of the weir. This is likely to include alteration of the timing of riffle-forming flows, and increased frequency and duration of periods when flows are too low to form riffles that are functional as macroinvertebrate habitat. Possible isolation by barrier of downstream riffles from macroinvertebrate colonisation via natural dispersal pathways. NATIONAL WATER COMMISSION — Low flows report series 26 Possible poor water quality downstream of the weir. Low flows also often create another stressor in the form of sedimentation or siltation (McKay & King 2006). Lack of flow means that sediments do not get flushed (Storey et al. 1991) and heavy sedimentation on river beds leads to anoxic decomposition of organic matter and deterioration of water quality. The biotic responses to the above hydrological stressors that have been deduced in this study are a high level of alteration in macroinvertebrate community composition within impounded sites, including the total loss of high-flow-preference macroinvertebrate taxa in the impounded zone where flow velocity is likely to be near zero, and the loss of riffle taxa in the impounded zone (as riffles no longer exist). Downstream of the weir, the biotic response was a shift toward a higher proportion of low-flowpreference taxa. The reduced number of riffles throughout the Burnett River catchment due to construction of the Walla Weir means a reduction in riffle biota in the catchment, and a reduction in connectivity between the remaining riffle populations. 4.2. Wivenhoe Dam case study 4.2.1. Introduction This case study analyses changes in flow characteristics resulting from operation of the Wivenhoe Dam, and subsequent biological responses from aquatic biota within the Brisbane River. A hydrological analysis was performed to identify the nature of low-flow alterations that have resulted from dam operation. Biological responses to these changes in flow were investigated using macroinvertebrates as biological indicators. Results indicate the nature of flow-ecological linkages present within south-east Queensland riverine ecosystems. The Brisbane River catchment is classified as highly developed, with between 70 to 100 per cent of the catchment’s mean annual discharge diverted for use. A number of significant water resource developments control water flow within the catchment, namely the Wivenhoe Dam, Somerset Dam, and Mt Crosby Weir system (DSEWPaC 2009). Aquatic ecosystems within the Brisbane River catchment are therefore expected to have undergone a high level of modification as a result of these developments (Statzner & Higler 1986; Bunn & Arthington 2002). The Wivenhoe Dam is currently the largest dam in south-east Queensland. The dam was designed by the Water Resources Commission and completed in 1985. It provides most of Brisbane’s drinking water and performs an important function in flood mitigation (Douglas et al. 2007; Seqwater 2009). The dam has significantly altered the Brisbane River’s habitat and flow characteristics upstream and downstream of the dam wall. The establishment of Lake Wivenhoe as a result of dam construction has significantly altered habitat conditions upstream of the dam, inundating an area of approximately 33.75 km2 at full storage capacity (Seqwater 2009). The dam wall is a significant barrier to the movement of biota through the system. For example, there has been no evidence of recruitment of eels or mullet upstream of the wall since construction of the dam, as both are diadromous species that require connection between the river and the sea (Arthington et al. 2000). Downstream of the dam, controlled release of water has significantly altered the natural flow regime with increased and more consistent flows (Choy et al. 2000; Seqwater 2009). All aspects of the natural flow regime are expected to have been altered as a result of water releases, including the magnitude, timing, frequency, seasonality and variability of flow, and similar low-flow characteristics (Lloyd et al. 2004; Bunn & Arthington 2002; Rolls et al. 2010). The Wivenhoe Dam is therefore expected to have impacted aquatic ecosystems within the Brisbane River downstream of the dam as a result of altered flow conditions, and corresponding alterations to in-stream habitat conditions (Statzner & Higler 1986; Choy et al. 2000; Bunn & Arthington 2002). During development of the water allocation management plan for the Brisbane River catchment, an analysis of flow requirements for aquatic macroinvertebrates was undertaken to investigate their responses to flow alterations caused by the Wivenhoe Dam (Choy et al. 2000). Although the study NATIONAL WATER COMMISSION — Low flows report series 27 was designed for the purpose of informing environmental flow requirements, the ecological consequences of altering the natural flow regime that were investigated provide information about the relationships between aquatic biota and low flows, and the ecological impacts associated with altering the low-flow regime. This case study used results from the macroinvertebrate environmental flow requirements study in conjunction with a new hydrological analysis at corresponding study sites to investigate low-flow ecological linkages. A reference site approach was used to determine the natural low-flow regime and ecological state of the Brisbane River using low-flow indices similar to those outlined by the NWC (2011) and using macroinvertebrates as biological indicators. In the hydrological analysis, the nature of flow alterations specific to low flows caused by operation of the Wivenhoe Dam was investigated. The results of the environmental flow requirements study were then used to analyse macroinvertebrate responses to these alterations. Any differences between the actual and presumed natural ecological state of the Brisbane River were analysed to determine their likely causes. Where the drivers of altered macroinvertebrate community structure appeared to be in relation to changes in the natural flow regime, it was reasonably assumed that these changes were related to the impacts of the Wivenhoe Dam. Based on the general PSR conceptual model for dams and weirs, a number of expectations about the impacts of the Wivenhoe Dam on low-flow characteristics and macroinvertebrate community composition were formed prior to analysis. As a result of consistent water releases downstream of the dam, it was expected that the duration and frequency of low-flow events would be significantly reduced compared with the natural flow regime. This was expected to affect in-stream habitat availability by drowning out riffle habitats and deepening pools. Macroinvertebrate taxa dependent on riffle habitat were expected to be the most highly impacted due to reduced habitat availability, as well as lowered dispersal opportunities resulting from increased distances between areas of suitable habitat. Lower abundance and diversity of riffle-dependent macroinvertebrate taxa were therefore expected when compared with the reference condition. 4.2.2. Methods Changes in macroinvertebrate community composition A summary of the methods conducted in the study of macroinvertebrate flow requirements within the Brisbane River by Choy et al. (2000) are presented below, as relevant to this report. A full account of the methodology is presented in Choy et al. (2000). Study sites Study sites were chosen within the Brisbane River catchment to be representative of sites that were both regulated and unregulated by the Wivenhoe Dam. A degree of water regulation was present in two of the ‘unregulated’ sites, but to a much smaller degree than at sites influenced by the Wivenhoe Dam, and usually with an opposite effect on flow – whereby water was extracted from the stream rather than supplemented, as was the case downstream of the Wivenhoe Dam. Subsequent impacts of water regulation within unregulated sites on macroinvertebrate communities were therefore expected to be minimal, and if present, in the opposite direction of impacts to sites downstream of the Wivenhoe Dam. These sites also grouped with unregulated sites during multivariate analysis, and are therefore considered to be unregulated sites for the purpose of this study. Sites unregulated by the Wivenhoe Dam are upstream of the dam, and regulated sites are downstream. This introduces the possibility that any differences between regulated and unregulated sites may be due to natural differences in relation to river position, in accordance with the river continuum concept (Vannote et al. 1980). To aid in the separation of differences between sites and identify which differences are natural and which can be attributed to the impacts of the dam, a reference approach was used. The Logan River was chosen to represent the reference condition, as it is a nearby river that drains into Moreton Bay as does the Brisbane River, and it is relatively NATIONAL WATER COMMISSION — Low flows report series 28 unregulated compared with the Brisbane. At the time of the study, the Logan River had one dam and two weirs, but these had altered in-stream hydrology by only a small amount, with flows estimated to have been reduced by 5 to 10 per cent below the pre-development state (Ruffini & Pandeya 1996). In comparison with the lower Brisbane River, flow regulation within the Logan River is of a much smaller magnitude and consists of water extraction rather than supplementation. Sampling within the Logan River was conducted (also by DERM) in an earlier study of macroinvertebrate flow requirements, as published in Choy and Marshall (1997). Sampling methodology was consistent between the Brisbane and Logan river sites. A total of eight study sites – four regulated and four unregulated – were sampled within the Brisbane River catchment (Figure 10, Table 5). A total of 10 study sites were sampled within the Logan River catchment – five upstream and five downstream – at positions approximating the position of study sites in the Brisbane River catchment (Figure 10, Table 5) (from Choy & Marshall 1997). NATIONAL WATER COMMISSION — Low flows report series 29 Figure 10: Map indicating the location of macroinvertebrate sampling sites within the Brisbane River and Logan River catchments. Site numbers are gauging station numbers. The Wivenhoe Dam is located at the base of Lake Wivenhoe. Red triangles in the Brisbane River catchment indicate regulated sites downstream of the Wivenhoe Dam, and blue triangles indicate unregulated sites. Green triangles in the Logan River catchment indicate upstream sites and yellow triangles indicate downstream sites. NATIONAL WATER COMMISSION — Low flows report series 30 Table 5: Gauging station numbers and site names of macroinvertebrate sampling sites within the Brisbane River and Logan River catchments. Unregulated sites Regulated sites Catchment Gauging station number Site name Gauging station number Site name Brisbane River 1430050 Brisbane River at Crossing 26 143001C Brisbane River at Savages Crossing 143921A Cressbrook Creek at Rosentretors Crossing 1430060 Brisbane River at Atkinsons Crossing 143009A Brisbane River at Gregors Creek 1430061 Brisbane River at Burtons Bridge 1430063 Pryde Creek below Splityard Creek Dam 1430062 Brisbane River at North Kholo Upstream sites Logan River Downstream sites Gauging station number Site name Gauging station number Site name 1450044 Logan River at Peters Place 1450050 Logan River at Paynes 145003B Logan River at Forest Home 145014A Logan River at Yarrahappini 1450048 Logan River at Wyaralong 1450049 Logan River at Bromelton 1450043 Running Creek at Drynans 145020A Logan River at Rathdowney 1450046 Running Creek at Campsite 1450045 Logan River at Maroon Dam NATIONAL WATER COMMISSION — Low flows report series 31 Sampling Sites within the Brisbane River were sampled between one and five times each, between October 1994 and May 1997 (Table 6). At each site physical and water quality data were recorded and macroinvertebrates collected, all within a 100 m reach. Physical data recorded included single readings of stream width, depth, velocity, substratum composition, riparian cover, adjacent land use, and proportion of algal and macrophyte cover. Water quality measurements included water temperature, conductivity, pH, dissolved oxygen and turbidity. Aquatic macroinvertebrate samples were collected from four different habitats, but these were not always present at all sites. Habitats included riffle/run, edge, pool and macrophyte beds. Each habitat was defined as follows: riffles were relatively shallow (<30 cm deep) and fastflowing areas of broken water, usually over stony substrates. Runs were similar to riffles but tended to be deeper and over varying substrates. Edges were defined as habitats along the bank where there is little or no current – some terrestrial vegetation or tree roots may be present along the edges, but no aquatic vegetation. Pools were zones of relatively deep and either stationary or very slow-flowing water, over varying substrates. Macrophytes were regions of dense aquatic vegetation, preferably a distance from the banks. Samples were livepicked on-site and macroinvertebrates stored in ethanol. Macroinvertebrates were enumerated and identified in the laboratory, mostly to family level. NATIONAL WATER COMMISSION — Low flows report series 32 Table 6: Time of sampling and habitats sampled within each site in the Brisbane River catchment. R: riffle/run; E: edge; P: pool; and M: macrophytes. Blank spaces indicate no samples were taken (adapted from Choy et al. 2000). Sampling period 1 2 3 4 5 6 7 Oct May Sept July Nov Feb May Site no. Site name '94 '95 '95 '96 '96 '97 '97 1430050 Brisbane River at Crossing 26 R, E, P, M R, E, P, M P, M R, E, M R, E, P, M R, E, P, M R, E, P, M R, E, P, M R, P, M 143921A Cressbrook Creek at Rosentretors Crossing R, E, M R, E, P, M R, E, P R, E, P, M R, E, M R, E 143001C Brisbane River at Savages Crossing R, E, M R, E 143009A Brisbane River at Gregors Creek 1430060 Brisbane River at Atkinsons Crossing R, E, M 1430061 Brisbane River at Burtons Bridge R, E, M 1430062 Brisbane River at North Kholo R, P, M 1430063 Pryde Creek ds Splityard Ck Dam R, P, M R, E, P, M Data analysis: comparison of regulated and unregulated sites within the Brisbane River catchment Univariate statistics (t-tests) were used to test for differences in mean values between regulated and unregulated sites within the Brisbane River for a number of community measures. These included: total taxon richness, mean taxon richness, total abundance, Shannon-Weiner diversity, abundance of uncommon taxa (calculated for each sample as the sum of the abundance categories of uncommon taxa in the sample, where uncommon taxa were defined as those which contributed less than 0.5 per cent of the total sum of abundance categories for the entire database), number of unique taxa (unique to an individual sample), and abundance of pest taxa (in this case larvae of biting Diptera taxa, including Culicidae, Simuliidae, Ceratopogonidae, and Tabanidae). These measures were used to test for overall differences between sites with values grouped for all habitats, as well as for differences within individual habitats. The significance value of all statistical tests was set at 5 per cent (α = 0.05). Multivariate statistics based on Bray-Curtis similarity measures were also used to test for differences in macroinvertebrate community composition between sites. Community composition was recorded for each sample whereby each taxon was allocated to an abundance category of 0–4, defined as follows: 0 = absent (no individuals); 1 = rare (1–2 individuals); 2 = scarce (3–4 individuals); 3 = common (5–8 individuals); and 4 = abundant (9 NATIONAL WATER COMMISSION — Low flows report series 33 individuals or more). Three datasets were constructed using this data. The first dataset consisted of abundance categories with uncommon taxa removed (those contributing less than 0.01% of the overall sum of abundance categories and less than 5 per cent of the sum of abundance categories for any one sample). The second dataset represented relative abundance, and the third dataset represented presence/absence data of each taxon within each sample. During analysis of Bray-Curtis similarities all three datasets gave similar results, and therefore only the first dataset (which maintained the most information) was used for all subsequent tests. Abundances of taxa within each sample were therefore used to represent macroinvertebrate community composition. Physical and habitat variables were range standardised (x-min/range) and matrices of Euclidean distance calculated between samples. These difference matrices and the Bray-Curtis matrix generated from the faunal abundance category data with uncommon taxa removed were then used to conduct subsequent multivariate analyses. Multivariate statistical analyses based on the macroinvertebrate abundance dataset were used to identify sources of variation. This included tests for differences in macroinvertebrate community composition between habitat types, and between sampling times, and between regulated and unregulated sites. Comparing regulated with unregulated sites in this way in addition to univariate tests was beneficial because multivariate analyses had the ability to test for differences between regulated and unregulated sites while still considering the impacts of environmental factors, and thus maintaining additional sources of variation which univariate statistics would be unable to account for. The physical and water quality variables recorded for each site were therefore included within these analyses. Differences between habitats and sampling periods were tested by two-way analyses of similarity (ANOSIMs) with habitat crossed with sampling period. These tested the hypotheses that there were differences between habitats allowing for differences between sampling periods and that there were differences between sampling periods allowing for differences between habitats. Based on the results of the multivariate analyses, which indicated that habitat type was significantly affecting macroinvertebrate community composition, subsequent analyses had to also consider habitat type as a factor. Subsequently, the analyses of difference between regulated and unregulated sites were conducted using two-way ANOSIMs which allowed for differences between habitat types, as well as separate one-way ANOSIMs for individual habitat types. The analyses within individual habitat types were, however, dependent on there being enough samples of each habitat type from both regulated and unregulated sites. Only riffle/run, edge, and macrophyte habitats had enough samples to allow this, as pool habitats did not. Edge samples were, however, indicative of pool habitat as they were only sampled from slow-flowing areas and therefore all major habitat types were represented in analyses. Macroinvertebrate community composition also appeared to vary according to sampling time. Due to the irregular sampling of sites across sampling periods, some sites were sampled later than others. By chance there were more sites that were only sampled in the last two sampling periods situated at regulated sites downstream of the Wivenhoe Dam, than there were situated at unregulated sites upstream of the Wivenhoe Dam (see Table 6). Therefore the differences in macroinvertebrate composition between sampling periods could have been due to river position or water regulation; as opposed to temporal influence per se. To separate these impacts, separate one-way ANOSIMs were conducted to test for differences between sampling periods for regulated (downstream) sites and unregulated (upstream) sites. Pairwise comparisons were made following globally significant ANOSIMs. If these differences were significant, similarity percentages (SIMPER) were calculated from the faunal data, to NATIONAL WATER COMMISSION — Low flows report series 34 indicate the taxa contributing the most to the differences. The matrices were ordinated in three dimensions using semi-strong hybrid multidimensional scaling (SSHMDS), rotated to maximise the variation in the plane of two dimensions and plotted in these two planes. Correlation vectors were calculated from the faunal and corresponding environmental datasets. The significance of correlations was calculated from 100 Monte Carlo randomisations. The correlation vectors of selected significant variables (p < 0.05) were plotted on the ordinations. Data analysis: comparison of patterns of difference between the Brisbane River and the Logan River catchments A macroinvertebrate abundance dataset was compiled for samples collected within the Logan River, calculated in the same manner as the first dataset used for the Brisbane River. Once again, physical and habitat variables were range standardised (x-min/range) and matrices of Euclidean distance calculated between samples. These difference matrices and the BrayCurtis matrix generated from the faunal abundance category data with uncommon taxa removed were then used to conduct subsequent multivariate analyses. Multivariate analyses were then used to test for differences in macroinvertebrate community composition between upstream and downstream sites within the Logan River. For comparison with results from similar tests conducted on samples from the Brisbane River, one-way ANOSIMs were also conducted for individual habitat types within the Logan River. Only edge and riffle/run habitat types had enough samples from both upstream and downstream sites within the Logan River for analysis. Patterns of difference between upstream and downstream edge and riffle/run habitats were compared between the two rivers. Any differences between the two rivers in terms of macroinvertebrate composition and environmental (physical and water quality) variables were then tested for significance, and correlation vectors were plotted using the same methodology as was conducted for the Brisbane River. Differences in the composition of macroinvertebrate communities were also analysed according to functional composition. Taxa were assigned to trophic categories (functional feeding groups) by reference to literature (Cummins 1973; Merritt & Cummins 1978; Chessman 1986; Hawking & Smith 1997) (see Table 8 for allocation of functional feeding groups to relevant taxa). Five functional feeding groups were used: collectors, filterers, grazers, shredders and predators. They were defined as follows: collectors feed on small particulate matter which they gather from the substrate or other surfaces filterers filter suspended particulate matter from the water grazers scrape periphyton (attached algae, bacteria, fungi etc.) off the substrate shredders chew large particles of plant matter such as leaves and twigs predators feed on other animals. The compositions of functional feeding groups within combined samples were compared between upstream and downstream sites in both rivers. The composition of functional feeding groups is expected to naturally fluctuate between seasons and along the river continuum in response to changing longitudinal environmental conditions (Marshall et al. 2001; Vannote et al. 1980). Comparisons were made between the two rivers, looking for differences in the patterns of change in functional feeding groups between upstream and downstream sites, and between wet and dry seasons. NATIONAL WATER COMMISSION — Low flows report series 35 If the magnitude and direction of correlation vectors between upstream and downstream sites was the same in both rivers, this would indicate that the differences were natural. If the two rivers responded differently, however, the difference in the Brisbane River may be attributable to the effects of water resource development. Changes to hydrological conditions To determine the nature of the stressor imposed by the Wivenhoe Dam, an analysis of the change in flow regime downstream before and after dam construction was conducted, with particular reference to low flows and variability. To separate the changes resulting from flow management from those resulting from natural variation in flow over the study period, analyses were also performed for the corresponding years at a reference gauge site on the Logan River with no dam present. Time series of flow (ML day-1) from gauging stations 143001C Brisbane River at Savages Crossing and 145014A Logan River at Yarrahappini (Figure 10, Table 5) were used for the analysis. Construction of the Wivenhoe Dam began in 1977 and was completed in mid-1985, so data collected during this time was removed and the time series from both sites split into pre- and post-dam periods. Part years at the start and end of datasets were removed so that the analysis was conducted on full years only and data gaps were in-filled using simple linear interpolation. The result was 19 years of pre- and 25 years of post-impact data for the Brisbane River (1959–77, 1986–2010) and eight years of equivalent pre- and 25 years of post-data for the Logan (1970–77, 1986–2010). Hydrologic metrics were calculated, compared for before- and after-impact differences, and tested for significance using the scorecard function of Indicators of Hydrologic Alteration (IHA) Software V7.1 (Richter et al. 1996). Median measures were used, rather than means, because the gauge data were non-normally distributed. Flow duration curves were produced using the Time Series Analysis module of the River Analysis Package (RAP) V 3.0.4 (Marsh et al. 2003). 4.2.3. Results Changes in macroinvertebrate community composition Differences between regulated and unregulated sites within the Brisbane River catchment When univariate measures were tested, there were no significant differences in total taxon richness between regulated and unregulated sites (p<0.05) for either grouped or individual habitats. There were, however, differences in the mean taxon richness of samples between regulated and unregulated sites. Samples from unregulated sites had a higher overall mean number of taxa when habitat types were grouped (37.1 v. 31.8 taxa), as well as a higher mean number of taxa within macrophyte habitats (24.8 v. 18.8 taxa). The number of unique taxa was also higher within macrophyte samples from unregulated sites compared with macrophyte samples from regulated sites (2.2 v. 5.5 unique taxa). There was a similar trend in grouped habitat samples and all other habitat samples except pools. The number of unique taxa in pool samples was, however, assessed from only one sample each. Therefore, higher sample numbers may have indicated more significant results. There were no other significant differences in any of the other univariate community measures tested. NATIONAL WATER COMMISSION — Low flows report series 36 When multivariate datasets were analysed, as previously mentioned, all three datasets gave similar test results. Therefore the first dataset, comprising taxa abundance (minus rare taxa), was used for all subsequent tests, and only results from this dataset are presented. There were significant differences in macroinvertebrate composition between habitat types, allowing for differences between sampling periods (Table 7). Macroinvertebrate communities within riffle/runs were significantly different from all other habitats except pools, possibly due to the low number of pool samples. Analyses of taxa contributing to the differences between habitats consistently identified Simuliidae and Hydropsychidae as contributing to the differences between riffle/runs and other habitats. Both taxa are associated with high-flow velocities according to the flow classification of Marshall and Marshall (in prep.) (Table 8). Similarly, ANOSIMs within habitats identified riffle/runs as associated with high abundances of Hydropsychidae, and low abundances of Copepoda and Coenagrionidae, the latter of which are taxa associated with low flow velocities (Table 8). Table 7: Results of pairwise comparisons between habitat types following two-way crossed ANOSIM (habitat x run, Global R = 0.191, p = 0.0043) using abundance category data. This tests the null hypothesis that there is no difference between habitats allowing for differences between runs. Shaded cells indicate pairs of habitats that were found to be significantly different (p < 0.05) Rocky pool Macrophytes Riffle Sandy pool Edge R = 0.057 p = 0.308 R = -0.138 p = 0.932 R = 0.470 p < 0.0001 R = 0.116 p = 0.229 Rocky pool - R = -0.075 p = 0.608 R = 0.666 p = 0.005 R = 0.155 p = 0.281 Macrophytes - - R = 0.194 p = 0.02 R = 0.228 p = 0.154 Riffle - - - R = 0.289 p = 0.118 There were also significant differences between sampling periods, allowing for differences between habitat types. Sampling period two was significantly different from sampling period four (R = 0.328, p = 0.029), and sampling period six was significantly different from sampling periods two and five (R = 0.432, p = 0.007 and R = 0.421, p = 0.006). This may be due to temporal effects, or river position as previously discussed. Further tests were conducted to determine the likely cause. When regulated and unregulated sites were tested separately, there was not a significant difference between sampling periods allowing for habitat differences within regulated sites. In contrast, there was a significant difference between sampling periods allowing for habitat differences within unregulated sites. Based on macroinvertebrate abundances within unregulated sites, sampling periods two, four and five grouped together in ordination space, while sampling periods one and three separated from the group (Figure 11). Analyses of associated environmental variables were conducted to determine what the likely underlying cause was for this separation. A high correlation between sampling period and flow velocity was found. When ordination was based on maximum flow velocity, sampling periods grouped in the same manner, with sampling periods one and three separate from the group (Figure 11). This indicates that the variation in faunal composition between sampling periods may be attributable to temporal differences in flow velocity. The lack of temporal differences in macroinvertebrate composition within regulated sites indicates NATIONAL WATER COMMISSION — Low flows report series 37 that much of the biologically significant variation in flow velocity may have been lost within regulated sites. Results of two-way ANOSIMs indicated significant differences in community composition between regulated and unregulated sites, allowing for differences between habitat types. Correlation vectors indicated which taxa and environmental variables were associated with the differences between regulated and unregulated sites. With respect to unregulated sites, regulated sites tended to have higher abundances of Hydropsychidae and Simuliidae (highflow-preference taxa), and lower abundances of Atyidae, Planorbidae and Copepoda (lowflow-preference taxa). Regulated sites tended to have higher values for flow velocity and width, and lower values for conductivity, alkalinity and the proportion of the substrate composed of boulders. Regulated sites also tended to have more macrophyte cover, less detritus cover, and different adjacent land use compared with unregulated sites. a) b) Figure 11: Ordinations showing the relationships between sampling periods (runs) within riffle/run habitat, based on a) macroinvertebrate abundance data allowing for differences between habitat types (stress=0.0000), and b) mean maximum flow velocity (stress = 0.0002) (adapted from Choy et al. 2000). Results of one-way ANOSIMs indicated there were also significant differences in community composition between regulated and unregulated sites within all three of the habitat types tested (riffle/run, edge and macrophyte). SIMPER identified the taxa contributing the most to the differences between regulated and unregulated samples within each habitat type. Within riffle/runs, regulated sites had higher abundances of Helicopsychidae, Pyralidae, Elmidae, Simuliidae and Hydropsychidae; and lower abundances of copepods. Regulated edges were characterised by higher abundances of Sphaeromatidae, Hydrobiidae and Hydropsychidae; and lower abundances of Acarina and Leptophlebiidae. Regulated macrophytes were characterised by higher abundances of Sphaeromatidae and Hydropsychidae; and lower abundances of Coenagrionidae, Leptophlebiidae, copepods and Atyidae. This conforms with the pattern of there being more high-flow-preference taxa in regulated sites compared with unregulated sites (see Table 8). The only exceptions are Hydrobiidae and Leptophlebiidae which are classified as having no flow preference, and Sphaeromatidae for which flow preference is not defined according to Choy et al. (2000) and Marshall and Marshall (in prep.). Patterns of difference within the Brisbane River catchment compared with the Logan River catchment One-way analyses of similarity indicated there were significant differences in macroinvertebrate community composition between upstream and downstream sites within the Logan River for both of the habitat types tested (riffle/run and edge). This was also the NATIONAL WATER COMMISSION — Low flows report series 38 case in the Brisbane River, as outlined above. This indicates that some differentiation in macroinvertebrate community composition between upstream and downstream sites is natural. Some of the difference between regulated and unregulated sites within the Brisbane River may therefore also be natural. Ordination plots were used to analyse the magnitude and direction of the differences between upstream and downstream sites in both rivers to determine how much of the differences observed in the Brisbane River were in accordance with the differences in the Logan River. Results indicated that differences in faunal composition between regulated and unregulated sites within riffle/run habitats in the Brisbane River were in accordance with the Logan River, but this was not the case within edge habitats. The magnitude and direction of difference in riffle/run macroinvertebrate composition between upstream and downstream sites in the Logan River was similar to the difference between regulated and unregulated sites within the Brisbane River, with a difference in the angle of orientation between the two rivers of 10° (refer to the similar size and orientation of vectors LR v. BR in Figure 12). In contrast, the magnitude of difference in edge macroinvertebrate composition between developed and undeveloped sites within the Brisbane River was much larger than the difference in the Logan River (refer to the size of LE v. BE vectors in Figure 12). Similarly, the orientation of the differences in edge macroinvertebrate composition between regulated and unregulated sites within the Brisbane River was visibly different in ordination space to that of the Logan River, with a difference in the angle of orientation between the two rivers of 45° (Figure 12). Also note the similar direction of the Brisbane River edge vector to both of the riffle/run vectors in Figure 12. Correlation vectors of corresponding macroinvertebrate abundance and habitat variable data gave some indication of the factors contributing to the differences between upstream and downstream sites in both rivers. The average differences between upstream and downstream riffle/run samples in both rivers and between upstream and downstream (regulated v. unregulated) edge samples in the Brisbane River were associated with upstream samples having higher average abundances of Hydropsychidae; lower average abundances of Acarina, Veliidae, Corixidae, Atyidae, Copepoda and Cladocera; and higher average values for flow velocity (Figure 12). The average differences between upstream and downstream edge samples in the Logan River were not associated with the abundances of any particular taxa but were characterised by samples from downstream sites having higher average values for total nitrogen and water turbidity, and lower average mean sediment sizes (Figure 12). These results indicate that edge habitats in the Brisbane River do not have the same longitudinal variation in environmental variables that are seen in the Logan River. The longitudinal differences in environmental variables and corresponding differences in faunal composition within the Brisbane River therefore differ from the reference condition – which indicates that they are not natural. This provides evidence that at least some of the differences between upstream and downstream (regulated v. unregulated) sites within the Brisbane River may therefore be attributable to hydrological disturbance, in this case the impacts of the Wivenhoe Dam. This depends on the assumption that the Logan River is in a comparatively natural state and that no other factors such as land use are responsible for the differences between the two rivers. NATIONAL WATER COMMISSION — Low flows report series 39 LRr HPSY BRr LRu MeanSed logVEMA BRu BEr LEu a) BEu LEr UACA logTOTN logTURB COEN VELI ATYI UCLA b) CORI COPE HAEN c) Figure 12: a) Ordination plot of average Faunal Abundance Category data with arrows showing direction and magnitude of differences between regulated (downstream) and unregulated (upstream) samples from riffle/run and edge habitats in the Brisbane and Logan rivers (stress = 0.1790). B – Brisbane River, L – Logan River, R – riffle/run, E – edge, r – regulated (downstream), u – unregulated (upstream). b) Selected significant faunal correlation vectors contributing to variation in macroinvertebrate community samples within the Brisbane River (HPSY = Hydropsychidae, UACA = Acarina, COEN = Coenagrionidae, VELI = Veliidae, ATYI = Atyidae, UCLA = Cladocera, CORI = Corixidae, COPE = Copepoda). c) Selected significant environmental correlation vectors explaining variance in macroinvertebrate communities within the Brisbane River (MeanSed = mean sediment sizes, logVEMA = log maximum velocity, logTOTN = log total nitrogen, logTURB = log water turbidity) (from Choy et al. 2000). Differences in longitudinal variation between the Brisbane and the Logan rivers were also demonstrated by results from the comparison of functional feeding compositions (Figure 13). When comparing upstream sites, the Brisbane and Logan rivers had similar patterns of change in functional feeding composition between wet and dry seasons. In both rivers, there was a decrease in the proportion of filter feeders in the dry season compared with the wet season, and an increase in predators. But when comparing wet and dry seasons at downstream sites, the two rivers showed differing patterns of change. Functional feeding composition at downstream sites within the Logan River changed considerably from wet to dry seasons, with a decrease in the proportion of filterers and increases in collectors, grazers and predators. In contrast, there was little difference in functional feeding composition between wet and dry seasons at downstream sites within the Brisbane River. Similarly, when comparing functional feeding composition between upstream and downstream sites, there was a noticeable difference in the patterns of change between the two rivers. Within the Logan River, there were considerable differences between upstream and downstream sites in both seasons. There was also a noticeable difference in functional feeding composition between upstream and downstream sites within the Brisbane River during the dry season but little difference during the wet season. Within the Brisbane River, the functional feeding composition of downstream sites in both seasons was similar to the composition of upstream sites during the wet season. These results indicate that downstream sites within the Brisbane River have been altered compared with the reference state represented by the Logan River. There has been a change in the functional feeding composition of downstream sites with macroinvertebrate communities consistently resembling those of upstream sites during the wet season. There has also been a loss of variability in functional feeding composition between seasons, and across the longitudinal gradient. NATIONAL WATER COMMISSION — Low flows report series 40 wet season Logan River Brisbane River Headwaters Headwaters dry season wet season dry season Collectors Filterers Dam Grazers Shredders Predators wet season dry season wet season dry season Figure 13: Schematic representations of the Logan and Brisbane rivers showing the proportions of macroinvertebrate functional feeding groups in samples from the wet season and dry season at upstream and downstream sites in each river (from Choy et al. 2000). NATIONAL WATER COMMISSION — Low flows report series 41 Table 8: Flow preferences and functional feeding group membership of significant macroinvertebrate taxa identified within results, according to the flow classification of Marshall and Marshall (in prep.). Flow preferences correlate with the flow requirements, larvae and adult habitat of taxa as outlined in Choy et al. (2000). Taxon Flow-preference group Functional feeding group Elmidae High flow Collector Helicopsychidae High flow Grazer Hydropsychidae High flow Filterer Planorbidae High flow Grazer Pyralidae High flow Grazer Simuliidae High flow Filterer Atyidae Low/no flow Shredder Cladocera Low/no flow Collector Coenagrionidae Low/no flow Predator Copepoda Low/no flow Collector Corixidae Low/no flow Collector Veliidae Low/no flow Predator Acarina No preference Predator Hydrobiidae No preference Grazer Leptophlebiidae No preference Collector/grazer/shredder Sphaeromatidae Not determined Shredder Changes to hydrological conditions At both the test and reference gauge sites, monthly flows were more seasonally variable in the pre-1977 period, compared with post-1985 (Figure 14). In the earlier period, high flows occurred in the first few months of the year, followed by a trough during winter and spring before a slight rise at the end of the year. In the later period, flows were more stable throughout the year at both locations. The dates of the minimum and maximum flow days have not changed significantly (Table 9) so seasonality has not been altered; flow has simply become less variable. This is likely to be largely due to climatic variation: the early period experienced regular summer flooding, particularly in the early 1970s, and the later period was dominated by drought for much of the 1990s and 2000s. However, while flows became less seasonally variable at both locations, there appeared to be differences in flow level: the Logan River experienced low-flow conditions post-1985, while higher flow volumes were maintained in the Brisbane River. Median monthly flows have decreased by more than 50 per cent in all months in the Logan River, while in the Brisbane River there has been a decrease in flows in summer and autumn and an increase in late winter and spring (Figure 15). Similar patterns are present in metrics measuring the median minimum and maximum flow magnitude. In the Logan River, there has been a decrease in both low-flow and high-flow magnitudes in the post-1985 data. In the Brisbane River, low-flow magnitudes have increased since construction of the dam, while high-flow magnitudes have decreased (Figure 16). This suggests that release strategies for the dam have led to an ‘averaging out’ of flow conditions at medium levels. NATIONAL WATER COMMISSION — Low flows report series 42 Figure 14: Comparison of median monthly flow volumes (ML day-1) in the Brisbane and Logan Rivers in the periods before construction of the Wivenhoe Dam on the Brisbane River (pre1977) and after construction (post-1985). In terms of low flows, construction of the Wivenhoe Dam has decreased the occurrence of natural low-flow events. Low-flow discharge (the 90th percentile of the flow duration curve) increased at the Brisbane River gauge from 289 ML day-1 before dam construction, to 340 ML day-1 after construction. For the same period in the Logan River, the metric decreased from 47 day-1 to 9 day-1. Table 9: The median Julian date of the one-day annual minimum flow for the Brisbane and Logan rivers for the periods before (pre-1997) and after (post-1985) construction of the Wivenhoe Dam. Brisbane River Logan River Pre-1977 Post-1985 Pre-1977 Post-1985 Julian date minimum 293 323 342 332 Julian date maximum 45 48 31 39 NATIONAL WATER COMMISSION — Low flows report series 43 Figure 15: Change in median monthly flow volumes (ML day-1) for the Brisbane and Logan rivers in the period following construction of the Wivenhoe Dam (post-1985) compared with the preceding period (pre-1977). * indicates a significant difference (P=0.05). Figure 16: Change in median low- and high-flow volume metrics for the Brisbane and Logan rivers in the period following construction of the Wivenhoe Dam (post-1985) compared with the preceding period (pre-1977). * indicates a significant difference (P=0.05). NATIONAL WATER COMMISSION — Low flows report series 44 4.2.4. Discussion Differences in macroinvertebrate communities both within the Brisbane River catchment and between it and the Logan River reference catchment indicated that in-stream biota responded to altered flow conditions. The flow characteristics to which macroinvertebrate communities responded were flow velocity and variability, and frequency and severity of low-flow events. Sources of variation between macroinvertebrate communities within the Brisbane River catchment were related to habitat type, river position and sampling time. Flow conditions appeared to be an important factor contributing to the differences between habitat types, upstream and downstream sites, and sampling times. Macroinvertebrate communities within the Brisbane River catchment are clearly responding to flow conditions, with shifts in macroinvertebrate communities observed in correspondence with shifts in flow velocity. Differences between macroinvertebrate communities within the Brisbane River catchment compared with the reference condition in the Logan River catchment also indicated that fauna were responding to changes in flow conditions. The changes in flow conditions at downstream sites caused by the Wivenhoe Dam were an ‘averaging out’ of flows with stable moderate flow levels replacing annual fluctuations between high and low flows. The macroinvertebrate response to these changes was a shift in macroinvertebrate community composition, without a loss in diversity, overall abundance or uncommon taxa. There were two main shifts in macroinvertebrate community composition associated with flow alteration caused by the dam: 1) A shift in macroinvertebrate community composition towards a higher abundance of highflow-preference taxa and a lower abundance of low-flow-preference taxa within pool habitats (represented by pool-edge habitats). This suggests one of the impacts of the Wivenhoe Dam has been a decrease in the frequency of low-flow velocities through pools and edges. Taxa with a strong preference for low flows have been reduced in abundance, and some particularly sensitive taxa may have been lost. Other studies have similarly reported that increased frequencies of high flows downstream of impoundments have adversely affected velocity sensitive taxa (e.g. Richter et al. 1997). 2) A shift in the trophic structure of macroinvertebrate communities, as indicated by functional feeding group composition. Seasonal and longitudinal patterns of change in trophic structure were altered. Stanford and Ward (1984) reported that sites downstream of dams had functional compositions which resembled headwater sites rather than sites with similar positions in unregulated rivers. This pattern was evident in the Brisbane River. Similarly, Choy and Marshall (1999a) found that water development within the Burnett catchment, Queensland, altered the physical and biological system from a temporally dynamic river with seasonal patterns to a system with a generally fixed pattern. The loss of seasonal variability in macroinvertebrate functional composition below the Wivenhoe Dam is likely to have adverse consequences for the long-term maintenance of invertebrate communities in the Brisbane River. Downstream assemblages may have lost some of their ability to adjust to flow fluctuations and other environmental changes. As a result of the more stable flow conditions, taxa downstream of the dam may be more sensitive to environmental stresses than those at unregulated sites. Communities below the dam may therefore be less able to adjust to any variations in flow and other environmental variables should they occur in future. This effect is likely to become more pronounced the longer the populations of macroinvertebrates are exposed to flows with reduced variability (Choy et al. 2000). Furthermore, loss of the natural hydro-ecological continuum, as seen by altered longitudinal patterns of community composition, is likely to disrupt ecosystem function and the movement of energy in a large section of the Brisbane River (Vannote et al. 1980). NATIONAL WATER COMMISSION — Low flows report series 45 The lack of change in riffle macroinvertebrate communities within the Brisbane River catchment compared with the reference condition is in contrast to the expectation that riffle macroinvertebrates would be the most highly impacted habitat type as a result of being drowned out by increased flows. It could be that riffle communities are more resilient to increases in flow velocity than pool communities (Reice 1991), or it could be that the sampling design was not able to capture changes in riffle communities (Marshall et al. 2001). The sampling design was intended to detect changes within individual habitats, but not to detect changes in the distribution of habitats. Changes in the flow regime may alter the distribution of riffles rather than the conditions within the remaining riffles. Impacts on riffle communities may therefore exist without detection (Marshall et al. 2001). Any riffles that were drowned out by water release from the Wivenhoe Dam would not have been sampled as riffles. Pools, in contrast, would have remained within the same habitat category for sampling despite potential deepening and increased flow velocity within individual pools. The lack of response from riffle macroinvertebrate communities therefore does not mean that riffle-dependent taxa were not affected by the Wivenhoe Dam. Instead, the sampling techniques used within the study may not have been adequate for identifying impacts on riffle taxa. To address this problem, Marshall et al. (2001) suggested that during sampling sites be surveyed for the distribution of biologically relevant habitat types and this information mapped. Changes in habitat availability could then be compared between sampling times. The response of macroinvertebrate communities to changes in low-flow conditions in the Brisbane River catchment indicates that macroinvertebrates can be used as indicators of alteration to the low-flow regime in this region. However, detected changes may or may not have broader importance in terms of the ecological values of the region (see Section 5.2). To measure macroinvertebrate community responses to low-flow conditions, a referential approach using indices with conceptual links to low-flow hydrology, such as functional feeding group composition and the proportion of flow-preference taxa, is most appropriate. This is supported by a number of other studies by DERM (Marshall et al. 2001; Choy & Marshall 2000; Marshall et al. 2000; Choy & Marshall 1999a; Marshall & Marshall in prep.) in which ecological-flow responses were demonstrated by changes in aquatic macroinvertebrate functional feeding composition and flow-preference-taxa composition or abundance. Marshall et al. (2000) used % high-flow-preference taxa, % low-flow-preference taxa, and % filterers as indices to differentiate flow-related impacts on macroinvertebrate community composition from other impacts. Marshall and Marshall (in prep.) provide a comprehensive list of the flow preferences of common Queensland taxa. In addition to changes in functional feeding group composition and the proportion of flow-preference taxa, taxa particularly sensitive to flow conditions can be identified as indicators of low-flow response. Choy and Marshall (1999b) stated that Hemiptera and Coleoptera are generally sensitive to increased flow conditions, as well as pool-edge taxa in general, as they tend to be less tolerant of high flows than pool-bed taxa. This study indicates that within south-east Queensland, the high-flow-preference taxa Elmidae, Hydropsychidae, and Simuliidae seem to be particularly sensitive to flow changes; as do the low-flow-preference taxa Acarina, Atyidae, Copepoda and Coenagrionidae. The effects of the Wivenhoe Dam on macroinvertebrate communities indicate the types of changes that occur when the low-flow regime is altered. Reduced occurrence of low flows, probably through their effect on flow velocity and depth, resulted in a loss of low-flowpreference taxa and a shift toward a higher proportion of high-flow-preference taxa. Altered trophic structure and loss of seasonality and natural hydro-ecology patterns of continuum were also observed within the Brisbane River catchment in response to a reduced occurrence of low-flow conditions. These changes are suggestive of wider changes in ecosystem function caused by reduced low flows. NATIONAL WATER COMMISSION — Low flows report series 46 5. Current monitoring relevant to low flows: issues and recommendations DERM has several freshwater ecology monitoring and research programs designed to support policy making and water management (Table 10). These programs deal with a range of management issues and questions, and thus use various methods and operate in different locations. At present DERM does not have a monitoring program explicitly for low-flow ecology. Lowflow issues may be detected and dealt with through existing monitoring and research where they are locally relevant to anthropogenic pressures and ecological values (Table 10). There are, however, a number of difficulties associated with assessing the effect of low flows on ecology based on existing hydrological and ecological data. These difficulties, and the ways in which some of them may be overcome in the context of Queensland water management, are outlined in sections 5.1 to 5.2 below. 5.1. Hydrology Availability of gauge data. – Gauges have generally been placed to collect data relating to water resource management not ecology, therefore the location of gauges may not match biological sampling sites or the location of impacts. – Sparse distribution of gauges in some catchments, particularly in western Queensland, and in low-order streams in most catchments. – Improvements could be made by including ecological considerations into future gauge-station network planning. – ‘New enabling technologies’ such as networks of dataloggers (e.g. depth, water quality) linked to central recording stations could be deployed in areas where gaugestation distribution is limited, to collect time series data for ecological assessments. The technical limitations of gauging, estimating and/or modelling low flows make it difficult to accurately measure low-flow conditions and detect the critical change from low flow to no flow. When flow stops, waterbodies providing habitat to aquatic biota often remain, which over time may contract into a series of separate waterholes before drying out. Some waterbodies are more persistent than others, and each phase of drying has implications for in-stream ecology. Gauging relies on rating curves that relate depth to discharge for a defined cross-section and this information is seldom available once flow stops. Currently depth and temperature loggers are placed in some waterholes to assess no-flow conditions. – Gauging stations, mode of gauging (monitoring) and modelling of low-flow hydrology may be issues covered in the hydrology component of this Low Flow Ecological Response and Recovery Project. There is a lack of data relating to the incidence of some low-flow pressures. For example, localised water extraction, especially during dry times, is a major threat to aquatic communities in some areas (e.g. small streams and waterholes), but information on where and when extraction occurs and the volume of water removed is generally unavailable. Understanding groundwater contribution to stream baseflow and the supplementation of refugial waterholes is essential to fully appreciate the low-flow environment. NATIONAL WATER COMMISSION — Low flows report series 47 – Hydrological modelling must incorporate groundwater contributions in a spatially explicit way so that ecological responses can be modelled accordingly. This needs to include the effects of groundwater extraction on surface water flow and aquatic habitat persistence. 5.2. Ecology Low-flow and no-flow conditions are natural occurrences in most Queensland streams and the ecology is largely evolved to cope with these events. Changes to the ecology exist where human pressures on water resources change the natural wetting and drying of streams, including the reduction of the occurrence of low and no flows through increases in flow (e.g. constant releases from dams, interbasin transfers). – It should be identified that low flows in many systems are not a stressor, but a natural and necessary part of the flow regime. The alteration of low-flow regimes by additional wetting is as equally detrimental as drying events in wetter regions. It is important not to give the impression that dry is always bad and wet is always good in terms of aquatic ecology. – Extremes and thresholds – while the ecology of Queensland streams is typically adapted to survive periods with low or no flows, increases – particularly in the duration of these spells, but also in their frequency and changes to their timing – may none-the-less represent catastrophic events leading to population failure. Our understanding of thresholds beyond which catastrophic changes occur is rudimentary, and such instances are intrinsically difficult to predict. Modelling provides a means to simulate these events, but needs to be based on realistic approximations of system parameters. Availability of ecological data. – Despite having a large dataset of existing AusRivAS-type macroinvertebrate monitoring data, when we came to assess the ecological response to the pressures identified in the low flows PSR framework, little data was available in suitable locations and time periods. This is even truer of other biotic indicators (e.g. fish, riparian vegetation) where samples often cover small areas and short time periods and sampling methods differ between projects. – Existing data was collected for a specific purpose so it may not be appropriate for answering questions for which it was not designed. For example, riffle habitat macroinvertebrates should exhibit a response to changes in low-flow conditions, but because many past monitoring projects aimed to sample a set number of riffle habitats, rather than sample in the same place regardless of the nature of the habitat at the time, trends may be difficult to detect (see Section 4.2.4). During a dry spell, sampled macroinvertebrate communities in the remaining riffles are likely to be similar, despite the fact that there may be only half the usual number of riffles in the catchment, masking the impact. As such, in some cases, it is the spatial distribution of habitats, not the communities within them, which indicate stresses. – Where low flows are identified as a potential threat to ecological communities, specific sampling, monitoring or modelling designed for this purpose should be carried out. This would include developing and testing a conceptual system understanding; selecting measures of pressure, stressor and response; and appropriate spatial and temporal sampling distribution. While this is often a high-cost undertaking, it is the best way to ensure confidence in the assessment. NATIONAL WATER COMMISSION — Low flows report series 48 The complex relationship between gauged flow and biological response – these often act at different spatial and temporal scales and are mediated by other factors and impacts – makes it difficult to draw direct conceptual links between the two (Figure 17). For example, invertebrates living in a riffle experience the local hydraulics, substrate, water quality and non-flow impacts (e.g. presence of invasive species, pollutants) and this is what determines their community structure through time. These are indirectly related to gross flow metrics but are likely to vary greatly over time and space. – Improved conceptual understanding of the flow requirements of vulnerable aquatic ecosystem components and the links between anthropogenic pressures and ecological responses is being developed within DERM through the Environmental Flows Assessment Program (EFAP) and Steam and Estuary Assessment Program (SEAP). Figure 17: Ecological responses occur in reaction to the physical and chemical conditions experienced by biota, not directly to altered flow regimes. Altered flows influence ecology by interacting with various other properties of the setting to modify the conditions the biota experience. To understand ecological responses to flow modification one must first understand the relationships between physical/chemical conditions and the ecological response, and secondly understand the relationship between flow and the provision of these conditions. Stressors other than altered flow may also influence the response, so the effectiveness of flow regime restoration or protection cannot be judged directly by the occurrence or intensity of response. Rather, it must be judged by evaluating the provision of the flow-related conditions necessary for the response to occur (Cockayne et al. 2010). Depending on their requirements, different ecosystem components may respond differently to the same flow conditions. Therefore, understanding/predicting changes in macroinvertebrate communities doesn’t necessarily translate to other taxa or processes. Flows that provide a benefit to one taxon or process may impact another, meaning choices need to be made about which to manage for. – Develop a transparent process for prioritising ecological assets and selecting the most appropriate compromise where conflicting outcomes exist. Methods are currently being developed and implemented as part of the Queensland Water Resource Plan (WRP) development and review process. NATIONAL WATER COMMISSION — Low flows report series 49 There appears to be a significant reliance on structural indicators of the ecology and how they respond to low-flow conditions. The quantification of critical ecological processes is likely to give a more direct or holistic measure of ecosystem response, particularly instream productivity which relies on the low-flow end of the hydrograph where bed mobilisation risk is low, and microbial biofilms dominate the benthos. – Low-flow-ecology research should extend to hyporheic fauna and their critical dependence on this part of the hydrograph – an area which is currently underrepresented in the scientific literature. – Instead of relying on traditional biological indicators, the system conceptualisation, along with an understanding of the nature of the threat, should be used to select the most appropriate measures of ecosystem response to impacts; for example, genetic measures of population viability. Work has begun in a number of Queensland catchments to identify the distribution of stygofauna and the effect of groundwater use. This could extend to include hyporheic fauna in the future. Changes in an indicator such as macroinvertebrate assemblage composition do not necessarily constitute an unacceptable change as a result of a stressor. Acceptable change should be implicitly linked to the ecological values society holds for the ecosystem. – Several DERM assessment programs use a risk assessment approach to present the expected impacts of environmental change on ecosystem responses (WRP assessments, SEAP). When these responses are linked to values, the outputs represent the risk of a loss of values. For example, in WRP ecological assessments, ecological assets are selected as indicators of ecological values and assessments are made of the risk posed to the values from alternative flow management options. This approach avoids the ‘so what?’ response to measured or predicted ecological change, as the answer is implicit in the metric. NATIONAL WATER COMMISSION — Low flows report series 50 Table 10: Current DERM monitoring and research programs that may relate to low flows and ecology. Relevant programs are listed, as well as the location in which they are undertaken, the ecological indicators used, the frequency of activities, and how these activities apply to our knowledge of low-flow ecology. Program Location Indicators Frequency Application to low flows Stream and Estuarine Assessment Program (SEAP) Sampling by freshwater biogeographic province (a group of catchments clustered based on the similarity of fish and macroinvertebrate communities). Reporting at both province and catchment scales. Random-stratified site selection plus some fixed long-term monitoring sites. Vary by province – selected to respond to the anthropogenic threats of importance in each region, e.g. indicators for the Wet Tropics bioregion included riparian vegetation, feral pig impact, macroinvertebrates and freshwater fish. One province sampled per year. Sampling repeated in a province approximately every eight years. Long return interval so not suitable to detect incremental responses. Uses a PSR framework to select appropriate indicators to measure threats of importance. Doesn’t focus on low flows, though prioritised threats may relate to low flows (e.g. dams). Provides the context of water management, including low flows, in relation to other threats within each biogeographic province. e.g. Change in the occurrence and stability of low flows by water infrastructure was identified as an important threat in the Wet Tropics bioregion. Pressure = dams/weirs, Stressor = change in hydrology (based on IQQM modelling), Response = proportion of invertebrate flow-preference groups. An outcome of SEAP in the Wet Tropics bioregion was targeted research to investigate the effects of changes to low flows on populations of low-flow asset taxa, e.g. Mogurnda sp. which were identified in the EFAP process. Ecosystem Health Monitoring Program (EHMP) 135 sites within SEQ catchments. All sites visited on all sample runs. Water quality (pH, conductivity, temperature, DO, nutrient levels). Macroinvertebrates (number of taxa, PET richness, SIGNAL index scores). Fish (proportion of natives, proportion of aliens, O/E). Sites sampled twice per year in autumn and spring (post-wet and pre-wet seasons). Long-term dataset includes drought and subsequent flooding – could be used to identify trajectories of decline and recovery from related impacts. Pre- and post-wet annual sampling may distinguish effects of regular seasonal dry spells. Raw data likely to be more useful for low-flow assessments than calculated indices. Program not designed for assessing the impact of low flows, which may affect confidence in results if used for that purpose. Surface Water Ambient Network Subset of DERM gauging stations throughout Queensland Flow and water quality measures (temperature, pH, EC, total nitrogen, total phosphorus, turbidity, and spot grab samples Manual water quality sampling generally occurs four times a year (more frequently in Collects ambient water quality data at a range of flow levels. This is currently used for condition and trend analyses, but could be used to inform relationships between flows and water quality for ecological assessments, particularly if used in conjunction with NATIONAL WATER COMMISSION — Low flows report series 51 Program Location Indicators Frequency Application to low flows (SWAN) (approx. 400). 190 of these are automated water quality sampling gauges and manual spot samples are collected at others for laboratory analysis. for major cations and anions). some locations) along with additional sampling to capture variability. Continuous automated sampling of some parameters at some gauges. sediment and nutrient load (during high-flow events) sampling carried out in SEQ and GBR catchments. As this project collects data on the changes in water quality parameters in disconnected waterholes over the progression of a dry spell, this could inform assessments of low-flow impacts, but it is difficult to link this information to gauged data because of the issues with gauging low flows. Environmental Flows Assessment Program (EFAP) Statewide. Specific data collected in WRP areas (catchments or groups of catchments). Flow-dependent ecological assets are selected specifically for WRP areas as indicators of the regions’ ecological values. Prioritisation of values is based on their vulnerability to the type of hydrological modification represented by the WRP/ROP. e.g. for the Fitzroy Basin WRP, assets included banana prawns, barramundi, low-flow spawning fish, refugial waterholes and riparian tree communities. Risks from flow management were calculated by comparing the water requirements of assets (expressed as facets of the flow regime) with modelled water management flow scenarios. Sampling varies by ecological asset. No monitoring as such, rather data collection where necessary to define/refine the asset’s flow dependencies. This information is then used to develop ecological response models which evaluate flow management scenarios at five- or 10-year intervals. The suite of assets selected for a WRP area will have requirements across a number of different naturally occurring flow bands, including low flows. Responses of assets to changes in flows is modelled rather than measured. Because of the influence of a range of non-flow impacts, it is impossible to attribute biotic responses directly to flow management. Local research is being undertaken within WRP areas to better understand flow-ecology relationships and responses to flow management, including low-flow conditions, e.g. effect of low flows on Tandanus tandanus movement in the Pioneer catchment, and the persistence and connectivity of refugial waterholes in dryland catchments. Gauged flow data Approximately 400 surface water gauges and 150 continuous groundwater monitoring bores statewide. Water height and discharge. Continuous monitoring. Additional groundwater bores provide manual spot measurements. Length of record varies and distribution is patchy and can be sparse in some catchments. Difficulties measuring low levels of flow, so often not possible to distinguish between low and no flow. Low-flow gauging (monitoring) and modelling issues will hopefully be captured by the hydrology component of this Low-flow Ecological Response and Recovery Project. NATIONAL WATER COMMISSION — Low flows report series 52 6. Discussion and conclusions Flow regimes play a key role in shaping the ecology of freshwater systems. The low-flow band of the hydrograph is particularly important in many catchments where it acts either as the baseline condition to which ecosystems are adapted, or a disturbance imposed naturally or anthropogenically (Bunn & Arthington 2002; Kennard et al. 2010; Rolls et al. 2010). As such, the nature and effect of low-flow impacts on ecology need to be assessed by taking into consideration the natural conditions in a catchment. While the reduction of flow volume in wet catchments through drought or extraction may impact the specialised flow-dependent biota that reside there, it is likely to have less of an effect in dryland catchments where the biota have developed strategies to survive such conditions. On the other hand, dryland ecosystems may be significantly altered by actions which reduce the number or duration of low- and noflow spells, such as water releases. Aquatic biota have developed a range of strategies to survive in the environments they inhabit. Many have life-history stages or processes linked to flow characteristics (e.g. spawning and recruitment in Ambassis agassizii is linked to stable low flows – see Cockayne et al. 2010). While for other biota, flow conditions contribute to their distribution and extent (e.g. water level, velocity and substrate determine the extent of benthic algal mats; and flood frequency and inundation area drive riparian forest communities) (Bunn & Arthington 2002; Kennard et al. 2010; Rolls et al. 2010). Changes in flow conditions – from natural – will therefore affect different organisms in different ways, and to varying extents (Arthington & Pusey 2003). The best indicators of the effects of changes to the low-flow regime will vary, depending on the region of interest and the nature of the flow change: which will determine how the changes affect habitat in the given setting (Rolls et al. 2010). In Queensland, the pressures which relate to the flow regime, and particularly to low flows, are the construction and release strategies of dams and weirs, water extraction, disposal of excess industrial water, interbasin transfers and climate change (Table 1). Within Queensland it is predicted that climate change will result in reduced surface runoff due to higher temperatures and evaporation rates, and decreases in rainfall (Chiew & McMahon 2002; CSIRO & BOM 2007; DERM 2010a). This is likely to lead to increases in the number and duration of low- and no-flow spells. Extreme droughts and storm events are also predicted to become more common (CSIRO & BOM 2007; Hennessy 2007; DERM 2010a). Interbasin transfers have different effects on the flow regime in source and recipient systems, depending on the operation of the water network. In general, recipient streams will experience increased flows, with reduced frequency of low- or no-flow periods, and lower flow variability. Source streams may experience reduced flow and water levels, though in many cases source streams are already impounded, and therefore the additional impact of reduced flows on local hydrology and biota is likely to be minimal. Disposal of excess industrial water already occurs relatively frequently in lowland and estuarine river reaches along the Queensland coast (e.g. water used to cool machinery or clean mining plants and equipment) and tends to be more of an issue of water quality than quantity. However, with the emergence of the CSG industry in dryer regions such as the Queensland Murray-Darling and Fitzroy basins, there is the potential for water disposal to significantly alter flow regimes. Modelling of potential disposal scenarios conducted to date points to a reduction in the number of zero-flow days and increased duration of flow events (McGregor et al. 2011). NATIONAL WATER COMMISSION — Low flows report series 53 Depending on their operation, dams and weirs can have a range of effects on low flows. In the impounded area, velocity is reduced and stream depth and width are vastly increased. Downstream, flow can be artificially increased by water releases, changing the variability of flow conditions and drowning out some habitats that are created by baseflow conditions; or conversely, flow can be stopped by the barrier, increasing the occurrence of low- and no-flow conditions. The case studies conducted for the Wivenhoe Dam and Walla Weir shed some light on the response of biota to the changes in flow regime imposed by these structures. After the Walla Weir was built, there was a reduction in the proportion of high-flow-preference macroinvertebrate taxa both within the impounded area and below the weir wall due to reduced flow velocities. As a result, fewer source populations of riffle taxa remain in the catchment, with increased distances between populations, which may impact their resilience at a broader scale. Operation of the Wivenhoe Dam has significantly altered the hydrology of downstream reaches by decreasing flow variability and the duration and frequency of low-flow events. This increases the velocity of flow through habitats in the river that naturally had low velocities and, as a result, the dam affected downstream aquatic macroinvertebrate communities. The macroinvertebrate community composition within pool habitats shifted towards a higher proportion of high-flow-preference taxa, in addition to changes in the functional feeding composition, with a reduction in taxa that preferred low flows. These changes give meaningful indications of the types of biological responses that occur in response to reduced low flows, as mediated by their effects on the hydraulic conditions in the river. While the Walla Weir and Wivenhoe Dam case studies go some way to explaining the response of macroinvertebrate communities to the types of modification to low flows imposed by dams and weirs, they also highlight many of the difficulties with undertaking such assessments. At the Walla Weir, for example, there are no open gauges downstream of the weir, so hydrology data that may help to confirm the nature of the stressor is unavailable. There are a number of issues, both logistical and conceptual, in attempting to determine the effect of low-flow hydrology on ecology (see Section 6). As in the Walla Weir example, existing data about flow, ecological indicators and stressors is often unavailable for the time and place of relevance; or is not suitable for the analysis because of the sampling methods or design originally employed. There are also difficulties in understanding and interpreting ecosystem responses to changes in low flow because of issues of scale, complex interactions, confounding impacts, and differing responses from ecosystem components to the same conditions. These issues, along with information gaps about the specific flow requirements of biota and processes, make selection of the most appropriate ecological indicators for a region problematic but also of critical importance. Lastly, the hydrology and ecology of ground and subsurface waters, which are intimately linked to low-flow conditions, must be included in a holistic assessment but are largely unknown. The options for managing low flows, especially naturally occurring seasonal and supraseasonal dry spells, may also be limited. In regions where water resource development is necessary to fulfill economic and social goals, a compromise must be reached between human and environmental values. In addition, there are often limitations on the capacity for management of flows, either because the necessary infrastructure doesn’t exist, or because there simply isn't enough water available during dry times to meet demands. Ideally, management of low-flow impacts will take into account the natural hydrologic conditions, the NATIONAL WATER COMMISSION — Low flows report series 54 ecological values (and the most appropriate indicators/indices of these), the water use requirements, and the intervention options within a region. We suggest that in most cases it is not appropriate to seek direct relationships between measures of flow regime change and ecological condition. Carrying out univariate or multivariate statistical analysis of biological and environmental data do not provide strong correlations, because of a combination of poor conceptual framework, inappropriate data, confounding factors and possible lag effects. This recommendation is supported by the outcomes of two large-scale studies conducted by DERM in the Condamine-Balonne and Fitzroy River catchments (Negus et al. 2004). In these studies the responses of multiple ecological condition indicators to gradients of flow alteration were investigated. Both studies demonstrated that community and process-based ecological condition indicators did not respond to flow change in a predictable way, and thus direct approaches cannot effectively inform flow management decisions. This result is corroborated by a global literature review (Poff & Zimmerman 2007) which concluded that general, quantitative patterns between flow alteration and ecological responses are not strongly evident. The absence of predictable ecological responses to flow modification is, in part, because of the influences of confounding stressors (e.g. land use gradients often correlate with gradients of flow alteration), but fundamentally it is because biota do not experience or respond directly to hydrology. Rather, flow interacts with other features to produce physical, chemical and biological conditions which are perceivable to biota and which elicit particular ecological responses (Figure 17). Fish, for example, do not perceive mean annual discharge, but they do perceive stream depth, velocity and water temperature, and react to these in predictable ways. To inform the management of flow regimes to achieve ecological outcomes, we must focus our attention on understanding what conditions trigger ecological responses, and how flow interacts with other aspects of the local setting to provide these conditions. The interactions between particular responses and the conditions that trigger them should be general and transferable. In contrast, provision of the necessary conditions by the flow regime interacting with other influences is a function of setting and thus is not transferable. Because in-stream ecological responses are influenced by stressors other than flow alteration, it is inappropriate to assume that flow management will elicit the expected responses. It follows that it is also inappropriate to evaluate the effectiveness of flow management by measuring the occurrence or the intensity of ecological response. Even if flow management is ideal, the influence of other stressors may prevent a response from occurring. For this reason we advocate evaluating the provision of the flow-related conditions required for the response as the means by which flow management should be determined and evaluated. Through the EFAP program, DERM applies an ecological risk assessment framework where the risk to specific ecological values is linked to ecological responses with critical requirements for flow-related conditions. The provision of these critical requirements is used as the measure of management success. This approach accounts for both the potential influence of confounding stressors, and the complex relationships between flow, setting and ecology. NATIONAL WATER COMMISSION — Low flows report series 55 Shortened forms ANRA Australian Natural Resources Atlas AUSRIVAS Australian River Assessment System BOM Bureau of Meteorology CSG coal seam gas CSIRO Commonwealth Scientific and Industrial Research Organisation DERM Department of Environment and Resource Management, Queensland DPSIR Driving forces, Pressures, States, Impacts, Responses DSEWPaC Department of Sustainability, Environment, Water, Population and Communities EHMP Ecosystem Health Monitoring Program EFAP Environmental Flows Assessment Program EHMP Ecosystem Health Monitoring Program FFG Functional Feeding Group GS Gauging Station GBR Great Barrier Reef IHA Indicators of Hydrologic Alteration IPCC Intergovernmental Panel on Climate Change IQQM Integrated Quantity and Quality Model ML day-1 Megalitres per day NWC National Water Commission PET richness Plecoptera (stonefly), Ephemeroptera (mayfly) and Trichoptera (caddisfly) species richness PSR Pressure-Stressor-Response QMDB Queensland Murray-Darling Basin RAP River Analysis Package ROP Resource Operation Plan SEAP Steam and Estuary Assessment Program SEQ South East Queensland NATIONAL WATER COMMISSION — Low flows report series 56 SIGNAL Stream Invertebrate Grade Number – Average Level SIMPER Similarity Percentages SWAN Surface Water Ambient Network WRP Water Resource Plan NATIONAL WATER COMMISSION — Low flows report series 57 References Allibone RM 2000, ‘Water abstraction impacts on the non-migratory galaxiids of Totara Creek’, Science for Conservation, 147:25–43. Almeida EF, Oliveira RB, Mungai R, Nessimian JL & Baptista DF 2009, ‘Effects of small dams on the benthic community of streams in an Atlantic forest area of South-eastern Brazil’, International Review of Hydrobiology, 2:179–193. Australian Natural Resources Atlas (ANRA) 2000, Australian Natural Resources Atlas: Water – Queensland – Water Resources Overview. Australian Government, Canberra. Available from: www.anra.gov.au/topics/water/pubs/state_overview/qld_ovpage.html [18/08/11]. Arthington AH, Brizga SO, Choy SC, Kennard MJ, Mackay SJ, McCosker RO, Ruffini JL & Zalucki JM 2000, Environmental flow requirements of the Brisbane River downstream from Wivenhoe Dam, South-east Queensland Water Corporation Ltd. and Centre for Catchment and In-stream Research, Griffith University, Brisbane. Arthington AH & Pusey BJ 2003, ‘Flow restoration and protection in Australian rivers’, River Research and Applications, 19:377–395. Arthington AH, Naiman RJ, McClain ME & Nilsson C 2010, ‘Preserving the biodiversity and ecological services of rivers: new challenges and research opportunities’. Freshwater Biology, 55:1–16. Boulton AJ, Humphries WF & Eberhard SM 2003, ‘Imperilled subsurface waters in Australia: biodiversity, threatening processes and conservation’, Aquatic Ecosystem Health and Management, 6:41–54. Brunke M, Hoffman A & Pusch M 2001, ‘Use of mesohabitat-specific relationships between flow velocity and river discharge to assess invertebrate minimum flow requirements’, Regulated Rivers: Research and Management, 17:667–676. Bunn SE & Arthington AH 2002, ‘Basic principles and ecological consequences of altered flow regimes for aquatic biodiversity’, Environmental Management 30: 492–507. Chessman BC 1986, ‘Dietary studies of aquatic insects from two Victorian rivers’, Australian Journal of Marine and Freshwater Research, 37:129–146. Chessman BC 1995, ‘Rapid assessment of rivers using macroinvertebrates: a procedure based on habitat-specific sampling, family-level identification and a biotic index’, Australian Journal of Ecology, 20:122–129. Chessman BC, Jones HA, Searle NK, Growns IO & Pearson MR 2010, ‘Assessing effects of flow alteration on macroinvertebrate assemblages in Australian dryland rivers’, Freshwater Biology 55: 1780–1800. Chiew FHS & McMahon TA 2002, ‘Modelling the impacts of climate change on Australian streamflow’, Hydrological Processes, 16:1235–1245. Choy SC 1997a, Walla Weir baseline study; preliminary report on the aquatic macroinvertebrates, Resource Science Centre, Department of Natural Resources, Queensland Government, Brisbane., Choy SC 1997b, Walla Weir baseline study: report on the aquatic macroinvertebrates. Freshwater biological monitoring report no. 1, Department of Natural Resources, Queensland Government, Brisbane. Choy SC & Marshall JC 1997, ‘Environmental flow requirements of aquatic invertebrates’, in Logan River trial of the building block methodology for assessing in-stream flow requirements: background papers, Eds. Arthington AH & Long GC. Centre for NATIONAL WATER COMMISSION — Low flows report series 58 Catchment and In-Stream Research Griffith University, and Department of Natural Resources, Queensland Government, Brisbane. Choy S & Marshall C 1999a, An assessment of current flow-related ecological condition of the Burnett River and nearby catchments based on aquatic macroinvertebrates. Department of Natural Resources, Queensland Government, Brisbane. —1999b, Walla Weir baseline study: construction phase report on the aquatic macroinvertebrates, Freshwater biological monitoring report no. 16. Department of Natural Resources, Queensland Government, Brisbane. —2000, Aquatic macroinvertebrate monitoring and assessment for Walla Weir: operation phase, late dry, 1999, Freshwater biological monitoring report no. 29, Department of Natural Resources, Queensland Government, Brisbane. Choy SC, Marshall JC & Conrick DL 2000, Environmental flow requirements of aquatic invertebrates in the Brisbane River downstream of Wivenhoe Dam, Freshwater biological monitoring report no. 3, Department of Natural Resources, Queensland Government, Brisbane. Cockayne B, McGregor G, Marshall J, Lobegeiger J & Menke N 2010, Fitzroy Water Resource Plan review technical report 3: ecological risk assessment, Department of Environment and Resource Management, Queensland Government, Brisbane. Cummins KW 1973, ‘Trophic relationships of aquatic insects’, Annual Review of Entomology, 18:183–206. Commonwealth Scientific and Industrial Research Organisation (CSIRO) and Bureau of Meteorology (BoM) 2007, Climate change in Australia: technical report, Australian Government, Canberra. Available from: http://www.climatechangeinaustralia.gov.au/technical_report.php [02/09/11]. Davies BR, Thoms M & Meador M 1992, ‘An assessment of the ecological impacts of interbasin water transfers, and their threats to river basin integrity and conservation’, Aquatic Conservation: Marine and Freshwater Research, 2:325–349. Deitch MJ, Kondolf GM & Merenlender AM 2009, ‘Surface water balance to evaluate the hydrological impacts of small in-stream diversions and application to the Russian River basin, California, USA’, Aquatic Conservation: Marine and Freshwater Ecosystems, 19: 274–284. Department of Environment and Resource Management (DERM) 2008, Riverine assessment in Queensland’s Central Province: Stream and Estuary Assessment Program 2008, Queensland Government, Brisbane. —2010a, Climate change in Queensland: what is the science is telling us?. Queensland Climate Change Centre of Excellence and Department of Environment and Resource Management, Queensland Government, Brisbane. —2010b, Refugial waterholes project: research highlights, Department of Environment and Resource Management, Queensland Government, Brisbane. —2011a, Understanding water resource planning, Queensland Government, Brisbane. Available from: http://www.derm.qld.gov.au/wrp/understandingwrp.html [12/08/11]. —2011b, Water monitoring data portal, Queensland Government, Brisbane. Available from: http://watermonitoring.derm.qld.gov.au/host.htm [20/09/11]. Department of Sustainability, Environment, Water, Population and Communities (DSEWPaC) 2009, Water resources – Availability – Queensland. Australian Government, Canberra. Available from: http://www.anra.gov.au/topics/water/availability/qld/index.html#SW_dev [12/08/11]. NATIONAL WATER COMMISSION — Low flows report series 59 Dewson ZS, James ABW & Death RG 2007, ‘Invertebrate community responses to experimentally reduced discharge in small streams of different water quality’, Journal of the North American Benthological Society 26: 754–766. Döll P, Fiedler K & Zhang J 2009, ‘Global-scale analysis of river flow alterations due to water withdrawals and reservoirs’, Hydrology and Earth System Sciences Discussions, 6:4773–4812. Douglas G, Palmer M, Caitcheon G & Orr P 2007, ‘Identification of sediment sources to Lake Wivenhoe south-east Queensland, Australia’, Marine and Freshwater Research, 58:793–810. Finn MA, Boulton AJ & Chessman BC 2009, ‘Ecological responses to artificial drought in two Australian rivers with differing water extraction’, Fundamental and Applied Limnology/Archiv für Hydrobiologie 175: 231–248. Gawne B & Scholz O 2006, ‘Synthesis of a new conceptual model to facilitate management of ephemeral deflation basin lakes’, Lakes and Reservoirs: Research and Management, 11:177–188. Geist J 2011, ‘Integrative freshwater ecology and biodiversity conservation’, Ecological Indicators, 11:1507–1516. Ghassemi F, Jakeman AJ & Nix HA 1995, Salinisation of land and water resources: human causes, extent, management and case studies, Centre for Resource and Environmental Studies, Australian National University, Canberra. Gibbins CN, Jeffries MJ & Soulsby C 2000, ‘Impacts of an interbasin water transfer: distribution and abundance of Micronecta poweri (Insecta: Corixidae) in the River Wear, north-east England’, Aquatic Conservation: Marine and Freshwater Ecosystems, 10:103–115. Gregory SV, Swanson FJ, McKee WA & Cummins KW 1991, ‘An ecosystem perspective of riparian zones, Focus on links between land and water’, BioScience, 41:540–551. Growns JE, Chessman BC, McEvoy PK & Wright IA 1995, ‘Rapid assessment of rivers using macroinvertebrates: case studies in the Nepean River and Blue Mountains, NSW’, Australian Journal of Ecology 20: 130–141. Hawking JH & Smith FJ 1997, Colour guide to invertebrates of Australian inland waters, Cooperative Research Centre for Freshwater Ecology, Albury. Hancock PJ & Boulton AJ 2005, ‘The effects of an environmental flow release on water quality in the hyporheic zone of the Hunter River, Australia’, Hydrobiologia, 552:75–85. Hennessy K, Fitzharris B, Bates BC, Harvey N, Howden SM, Hughes L , Salinger J & Warrick R 2007, Australia and New Zealand, Climate Change 2007: impacts, adaptation and vulnerability’, contribution of Working Group II to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change, Eds. Parry ML, Canziani OF, Palutikof JP, van der Linden PJ & Hanson CE. Cambridge University Press, Cambridge. James ABW & Suren AM 2009, ‘The response of invertebrates to a gradient of flow reduction – an in-stream channel study in a New Zealand lowland river’, Freshwater Biology, 54:2225–2242. Kennard MJ, Pusey BJ, Olden JD, Mackay SJ, Stein JL & Marsh N 2010, ‘Classification of natural flow regimes in Australia to support environmental flow management’, Freshwater Biology 55: 171–193. Khan S 2008, ‘Managing climate risks in Australia: options for water policy and irrigation management’, Australian Journal of Experimental Agriculture, 48:265–273. NATIONAL WATER COMMISSION — Low flows report series 60 Kingsford RT 2000, ‘Ecological impacts of dams, water diversions and river management on floodplain wetlands in Australia’, Austral Ecology, 25:109-127. Larned ST, Thibault D, Arscott DB & Tockner K 2010, ‘Emerging concepts on temporary river ecology’, Freshwater Biology, 55:717–738. Leigh C & Sheldon F 2008, ‘Hydrological changes and ecological impacts associated with water resources development in large floodplain rivers in the Australian tropics’, River Research and Applications, 24:1251–1270. Lloyd N, Quinn G, Thoms M, Arthington A, Gawne B, Humphries P & Walker K 2004, ‘Does flow modification cause geomorphological and ecological response in rivers? A literature review from an Australian perspective’, Technical report 1/2004, Cooperative Research Centre for Freshwater Ecology, Australia. Malmqvist B & Rundle S 2002, ‘Threats to running water ecosystem of the world’, Environmental Conservation, 29:134–153. Mackay S, Marsh N, Sheldon F & Kennard M 2012, Low-flow hydrological classification of Australia, National Water Commission, Canberra. Marsh N, Sheldon F & Rolls R 2012, Synthesis of case studies quantifying ecological responses to low flows, National Water Commission, Canberra Marsh NA, Stewardson MJ & Kennard MJ 2003, River Analysis Package, Cooperative Research Centre for Catchment Hydrology, Monash University, Melbourne. Marshall CJ & Marshall JC (in prep), Flow velocity and substrate particle size preferences of aquatic macroinvertebrates: classification of taxa into preference groups. Marshall C, Negus P & Choy S, 2000, An assessment of the current ecological condition of the Pioneer River, Sandy Creek and Bakers Creek catchments based on aquatic macroinvertebrates, Department of Environment and Resource Management, Queensland Government, Brisbane. Marshall J, Negus P, Marshall C, Choy S, Dale B & Gooda M 2001, Development of empirical relationships between flow regime and ecological condition in Queensland rivers, Freshwater biological monitoring report no. 29, Department of Natural Resources, Queensland Government, Brisbane. Marshall JC, Steward AL & Harch BD 2005, ‘Taxonomic resolution and quantification of freshwater macroinvertebrate samples from an Australian dryland river: the benefits and costs of using species abundance data’, Hydrobiologia, 572:171–194. Marshall J, McGregor G, Marshall S, Radcliffe T & Lobegeiger J 2006, Development of conceptual Pressure-Vector-Response models for Queensland’s riverine ecosystems, Department of Natural Resources, Mines and Water, Queensland Government, Brisbane. McGregor G, Marshall J & Takahashi E 2011, Stream ecosystem health response to coal seam gas water release: flow guideline, Department of Environment and Resource Management, Queensland Government, Brisbane. McKay SF & King AJ 2006, ‘Potential ecological effects of water extraction in small, unregulated streams’, River Research and Applications, 22:1023–1037. Merritt RW & Cummins KW 1978, An introduction to the aquatic insects of North America, Kendall/Hunt, Dubuque, Iowa. Metzeling L, Chessman B, Hardwick R & Wong V 2003, ‘Rapid assessment of rivers using macroinvertebrates: the role of experience, and comparisons with quantitative methods’, Hydrobiologia, 510:39–52. NATIONAL WATER COMMISSION — Low flows report series 61 Monk WA, Wood PJ, Hannah DM & Wilson DA 2008, ‘Microinvertebrate community response to inter-annual and regional river flow regime dynamics’, River Research and Applications, 24: 988–1001. Morton SR, Hoegh-Guldberg O, Lindenmayer DB, Olson M, Hughes L, McCulloch MT, McIntyre S, Nix HA, Prober SM, Saunders DA, Andersen AN, Burgman MA, Lefroy EC, Lonsdale WM, Lowe I, McMichael AJ, Parslow JS, Steffen W, Williams JE & Woinarski JCZ 2009, ‘The big ecological questions inhibiting effective environmental management in Australia’, Austral Ecology, 34:1–9. Mueller M, Pander J & Geist J 2011, ‘The effects of weirs on structural stream habitat and biological communities’, Journal of Applied Ecology, article first published online, DOI: 10.1111/j.1365-2664.2011.02035.x. Available from: http://onlinelibrary.wiley.com/doi/10.1111/j.1365-2664.2011.02035.x/abstract [25/08/11]. Naiman RJ, Bunn SE, Nilsson C, Petts GE, Pinay G & Thompson LC 2002, ‘Legitimizing fluvial ecosystems as users of water: an overview’, Environmental Management, 30:455–467. Naiman RJ & Latterell JJ 2005, ‘Principles for linking fish habitat to fisheries management and conservation’, Journal of Fish Biology, 67:166–185. Naiman RJ, Latterell JJ, Pettit NE & Olden JD 2008, ‘Flow variability and the vitality of river systems’, Geoscience, 340:629–643. Nebel S, Porter JL & Kingsford RT 2008, ‘Long-term trends of shorebird populations in eastern Australia and impacts of freshwater extraction’, Biological Conservation, 141:971–980. Negus P, Marshall C, Winning M, Cheshire K & Harch B 2004, Fitzroy Basin Resource Operation Plan ecological monitoring and assessment program: using a gradient of impact approach to identify indicators for inclusion in the design of an on-going monitoring program, Department of Natural Resources and Mines, Brisbane. Negus P 2007, Draft briefing note: ecological implications of interbasin water transfers. Department of Natural Resources and Water, Queensland Government, Brisbane. Nilsson C, Reidy CA, Dynesius M & Revenga C 2005, ‘Fragmentation and flow regulation of the world’s large river systems’, Science, 308:405–408. Olden JD & Naiman RJ 2010, ‘Incorporating thermal regimes into environmental flows assessments: modifying dam operations to restore ecosystem integrity’, Freshwater Biology, 55:86–107. Page T, Marshall J, McGregor G & Hughes J 2010, An investigation of potential implications of interbasin water transfers on the long-term genetic viability of populations of aquatic fauna, Australian Rivers Institute, Griffith University, Brisbane, and the Queensland Department of Environment and Resource Management, Queensland Government, Brisbane. Poff NL, DeCino RD & Ward JV 1991, ‘Size-dependent drift responses of mayflies to experimental hydrological variation: active predator avoidance or passive hydrodynamic displacement?’ Oecologia, 88:577–586. Poff NL, Allan JD, Bain MB, Karr JR, Prestegaard KL, Richter BD, Sparks RE & Stromberg JC 1997, ‘The natural flow regime, a paradigm for river conservation and restoration’, Bioscience, 47:769–784. Poff NL, Allan JD, Palmer MA, Hart DD, Richter BD, Arthington AH, Rogers KH, Meyer JL & Stanford JA 2003, ‘River flows and water wars: emerging science for environmental decision making’, Frontiers in Ecology and the Environment, 1:296–306. NATIONAL WATER COMMISSION — Low flows report series 62 Poff NL & Zimmerman JK 2007, Quantifying ecological responses to flow alteration: a literature review, paper presented at 10th International River Symposium, September 2007, Brisbane. Pusey BJ & Arthington AH 2003, ‘Importance of the riparian zone to the conservation and management of freshwater fish: a review’, Marine and Freshwater Research, 54:1–16. Queensland Water Commission 2010, South East Queensland water strategy, Queensland Water Commission, Queensland Government, Brisbane. Reice SR 1991, ‘Effects of experimental spates on benthic community structure in New Hope Creek, North Carolina, USA’, Internationale Vereinigung fuer Theoretische und Angewandte Limnologie Verhardlungen, 24:1691–1693. Richardson A & Humphries P 2010, ‘Reproductive traits of riverine shrimps may explain the impact of altered flow condition’, Freshwater Biology, 55:2011–2022. Richter BD, Baumgartner JV, Powell J & Braun DP 1996, ‘A method for assessing hydrologic alteration within ecosystems’, Conservation Biology, 10:1163–1174. Richter BD, Baumgartner JV, Wigington R & Braun DP 1997, ‘How much water does a river need?’ Freshwater Biology, 37:231–249. Richter BD, Baumgartner JV, Braun DP & Powell J 1998, ‘A spatial assessment of hydraulic alteration within a river network’, Regulated Rivers: Research and Management, 14:329–340. Richter BD & Thomas GA 2007, ‘Restoring environmental flows by modifying dam operations’, Ecology and Society, 12:12. Rolls R, Leigh C, Sheldon F, Kennard M & Pusey B 2010, Understanding low flows for improved water planning and management: ecological knowledge and adoption needs, Technical report, National Water Commission, Canberra. Ruffini J & Pandeya K 1996, ‘Hydrology of the Logan River’, in Trial of the South African building block methodology for defining in-stream flow requirements (IFR) of the Logan River system, south eastern Queensland, Australia. Draft starter document volume 1, introduction, workshop timetable, BBM methodology and physical aspects of the Logan River, eds. Arthington AH & Long GC, Centre for Catchment and In-Stream Research, Griffith University, Brisbane. Ryan TJ, Aland G & Cogle AL 2002, ‘Environmental condition of the Upper Mitchell River System: water quality and ecology’, Natural Heritage Trust, Queensland Government, Brisbane. Seqwater 2009, Wivenhoe Dam, Seqwater, Brisbane. Queensland Government, Brisbane. Available from: http://www.seqwater.com.au/public/source-store-treatsupply/dams/wivenhoe-dam [12/08/11]. Snaddon CD, Wishart MJ & Davies BR 1998, ‘Some implications of inter basin water transfers for river ecosystem functioning and water resources management in southern Africa’, Aquatic Ecosystem Heath and Management, 1:159–185. Stanford JA & Ward JV 1984, ‘The effects of regulation on the limnology of the Gunnison River, a North American case study’, in Regulated Rivers, eds. Lillehammer A & Saltveit SJ, Universitetsforlaget AS, Oslo. Statzner S & Higler B 1986, ‘Stream hydraulics as a major determinant of benthic invertebrate zonation patterns’, Freshwater Biology, 16:127–139. Steward AL, Marshall JC, Sheldon F, Harch B, Choy S, Bunn SE & Tockner K 2011, ‘Terrestrial invertebrates of dry river beds are not simply subsets of riparian assemblages’, Aquatic Sciences, in press. NATIONAL WATER COMMISSION — Low flows report series 63 Storey AW, Edward DH & Gazey P 1991, ‘Recovery of aquatic macroinvertebrate assemblages downstream of the Canning Dam, Western Australia’, Regulated Rivers – Research and Management, 6:213–224. Stubbington R, Wood PJ & Boulton AJ 2009, ‘Low-flow controls on benthic and hyporheic macroinvertebrate assemblages during supra-seasonal drought’, Hydrological Processes, 23:2252–2263. Takahashi E, McGregor G & Rogers S 2011, Stream ecosystem health response to coal seam gas water release: biological indicators, Department of Environment and Resource Management, Queensland Government, Brisbane. Thoms M & Cullen P 1998, ‘The impact of irrigation withdrawals on inland river systems’, Rangeland Journal, 20: 226–236. Vannote RL, Minshall GW, Cummins KW, Sedell JR & Cushing CE 1980, ‘The river continuum concept’, Canadian Journal of Fisheries and Aquatic Science, 37:130–137. Walters AW & Post DM 2011, ‘How long can you go? Impacts of low-flow disturbance on aquatic insect communities’, Ecological Applications, 21:163–174. Whittington J 2000, Technical review of elements of the WAMP process of Queensland DNR, Outcomes of a workshop held at River Glen Conference Centre on the 9th and 10th November 1999, Cooperative Research Centre for Freshwater Ecology, Canberra. Reports in the low flow series Balcombe SR & Sternberg D 2012, Fish responses to low flows in dryland rivers of western Queensland, National Water Commission, Canberra. Barma Water Resources & Sinclair Knight Merz 2012, Low-flow hydrological monitoring and modelling needs, report by for the National Water Commission, Canberra. Barmah D & Varley I 2012a, Hydrologic modelling practices for estimating low flows – stocktake, review and case studies, National Water Commission, Canberra Barmah D & Varley I 2012b, Hydrologic modelling practices for estimating low flows – guidelines, National Water Commission, Canberra Bond N 2012, Fish responses to low-flows in lowland streams: a summary of findings from the Granite Creeks system, Victoria, National Water Commission, Canberra. Bond N, Thomson J & Reich P 2012, Macroinvertebrate responses to antecedent flow, longterm flow regime characteristics and landscape context in Victorian rivers, National Water Commission, Canberra. Chessman B et al 2012, Macroinvertebrate responses to low-flow conditions in New South Wales rivers, National Water Commission, Canberra. Deane D 2012, Macroinvertebrate and fish responses to low flows in South Australian rivers, National Water Commission, Canberra. Dostine PL & Humphrey CL 2012, Macroinvertebrate responses to reduced baseflow in a stream in the monsoonal tropics of northern Australia, National Water Commission, Canberra. Hardie, SA et al 2012, Macroinvertebrate and water quality responses to low flows in Tasmanian rivers, National Water Commission, Canberra. Kitsios A et al 2012, Fish and invertebrate responses to dry season and antecedent flow in south-west Western Australian streams, National Water Commission, Canberra. NATIONAL WATER COMMISSION — Low flows report series 64 Leigh, C 2012, Macroinvertebrate responses to dry season and antecedent flow in highly seasonal streams and rivers of the wet-dry tropics, Northern Territory, National Water Commission, Canberra. Mackay S et al; 2012, Low-flow hydrological classification of Australia, National Water Commission, Canberra. Marsh N et al 2012, Synthesis of case studies quantifying ecological responses to low flows, National Water Commission, Canberra. Marsh N et al 2012, Guidance on ecological responses and hydrological modelling for lowflow water planning, National Water Commission, Canberra. Rolls R et al 2012, Review of literature quantifying ecological responses to low flows, National Water Commission, Canberra. Rolls R et al 2012, Macroinvertebrate responses to prolonged low flow in sub-tropical Australia, National Water Commission, Canberra. Sheldon F et al 2012, Early warning, compliance and diagnostic monitoring of ecological responses to low flows, National Water Commission, Canberra. Smythe-McGuiness Y et al 2012, Macroinvertebrate responses to altered low-flow hydrology in Queensland rivers, National Water Commission, Canberra. NATIONAL WATER COMMISSION — Low flows report series 65