Novel methods for managing freshwater refuges against climate change in southern Australia

Supporting document 4: Evaluating small barrier removal to improve refuge connectivity: a global review of barrier decommissioning and a process for southern Australia in a drying climate

Stephen Beatty, Mark Allen, Alan Lymbery, Tim Storer, Gillian White, David Morgan and Tom Ryan

NOVEL METHODS FOR MANAGING FRESHWATER REFUGES AGAINST CLIMATE CHANGE IN SOUTHERN AUSTRALIA

SUPPORTING DOCUMENT 4: Evaluating small barrier removal to improve refuge connectivity - a global review of barrier decommissioning and a process for southern Australia in a drying climate

AUTHORS

Stephen Beatty – Murdoch University Mark Allen – Murdoch University Alan Lymbery – Murdoch University Tim Storer – Department of Water, Western Australia Gillian White – Department of Water, Western Australia David Morgan – Murdoch University Tom Ryan – Murdoch University

Published by the National Climate Change Adaptation Research Facility

ISBN: 978-1-925039-54-2 NCCARF Publication 83/13

© 2013 Murdoch University

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 from the copyright holder.

Please cite this report as: Beatty, S, Allen, M, Lymbery, A, Storer, T, White, G, Morgan, D & Ryan, T 2013, Novel methods for managing freshwater refuges against climate change in southern Australia: Evaluating small barrier removal to improve refuge connectivity - a global review of barrier decommissioning and a process for southern Australia in a drying climate, National Climate Change Adaptation Research Facility, Gold Coast, 73 pp.

Acknowledgements This work was carried out with financial support from the Australian Government (through the Department of Climate Change and Energy Efficiency and the National Water Commission) and the National Climate Change Adaptation Research Facility (NCCARF).

The role of NCCARF is to lead the research community in a national interdisciplinary effort to generate the information needed by decision-makers in government, business and in vulnerable sectors and communities to manage the risk of climate change impacts.

Many thanks go to Belinda Robson for the development and management of the project. The barriers sub-project would also like to thank the following people whose input was greatly appreciated: Kath Lynch (Department of Water, WA/Geocatch), Lynn Heppell (Wilson Inlet Catchment Committee, WA), Mark Waud (Oyster Harbour Catchment Group, WA), Chris Sharpe (DPI NSW), John Koehn (Arthur Rylah Institute, Victoria), Michelle Miller (Australian National Committee on Large Dams), Rob Freeman (Inland Fisheries Service, Tasmania), Tim Marsden (Department of Agriculture, Fisheries and Forestry, Queensland), Richard Stewart (Department of Agriculture, Fisheries and Forestry, Queensland), Brian Stockwell (Department of State Development Infrastructure and Planning, Queensland), Paul Duncanson (North Queensland Dry Tropics NRM), Nick Bond (Griffith University), Brent Nottage (Helispecs), and Harry Dunkley (Helispecs).

Disclaimer The views expressed herein are not necessarily the views of the Commonwealth or NCCARF, and neither the Commonwealth nor NCCARF accept responsibility for information or advice contained herein.

Cover image: © Stephen Beatty

Table of contents

ABSTRACT ...... 1 EXECUTIVE SUMMARY ...... 2 1. OBJECTIVES OF THE RESEARCH ...... 8 2. INSTREAM BARRIER REVIEW ...... 9 2.1 Global trends in the construction of instream barriers ...... 9 2.2 Impacts of instream barriers ...... 11 2.2.1 Biological impacts on fishes...... 11 2.2.2 Instream barriers and refuges in southern Australia: impacts and the influence of climate change ...... 13 2.3 Restoring fish passage: barrier modification versus removal ...... 14 2.4 Impacts of barrier removal ...... 16 2.4.1 Hydrological restoration and improved river connectivity ...... 16 2.4.2 Elimination of barriers as aquatic ‘predators’ ...... 17 2.4.3 Ecological community shifts...... 17 2.4.4 Sedimentation ...... 18 2.4.5 Spread of exotic species ...... 19 2.4.6 Impacts of barrier removal on habitat availability and influence of climate change in southern Australia ...... 19 2.4.7 Cost ...... 20 2.5 Opposition to barrier removal ...... 20 2.6 Barrier removal advocacy ...... 22 2.7 Barrier removals globally ...... 23 2.7.1 North America ...... 23 2.7.2 Europe ...... 24 2.7.3 Australia ...... 24 2.8 Barrier prioritisation ...... 26 2.8.1 Score and ranking and optimisation modelling ...... 26 2.8.2 Prioritisation methods in practice ...... 28 2.9 Knowledge gaps ...... 31 3. PROCESS FOR SOUTHERN AUSTRALIA ...... 33 3.1 Project planning and development...... 35 3.2 Barrier identification ...... 37 3.3 Detailed barrier assessment and prioritisation ...... 37 3.4 Barrier mitigation on-ground works and monitoring ...... 41 3.5 Project extension ...... 42 4. PILOT STUDIES ...... 43 4.1 General methods ...... 43 4.2 Pilot study 1 – Ludlow River ...... 44 4.3 Pilot study 2 – Mitchell River...... 48 4.4 Pilot study 3 – Goodga River ...... 50 4.5 General discussion of pilot study results ...... 52 5. CONCLUSION ...... 54 REFERENCES ...... 55 APPENDIX 1 ...... 69 APPENDIX 2 ...... 72

SD 4: Evaluating small barrier removal to improve refuge connectivity iii

List of figures

Figure 1: Large dam creation in Australia (data from the Australian National Committee of Large Dams)...... 10 Figure 2: A conceptual example of optimisation modelling for prioritising barrier removal...... 27 Figure 3: Flow chart of the process for identification and mitigation of instream barriers in southern Australia...... 34 Figure 4: Seasonal migratory movements (top) of fishes of the Goodga River, south-western Australia: green = adults, blue = larvae, red = juveniles...... 36 Figure 5: Map summarising key results of the Ludlow River barrier prioritisation process...... 47 Figure 6: Map summarising the key results of the Mitchell River barrier prioritisation process...... 49 Figure 7: Map summarising the key results of the Goodga River barrier prioritisation process...... 51

List of tables

Table 1: Score and ranking system developed to prioritise instream barriers for mitigation or removal in southern Australia...... 39 Table 2: Summary of key results of the barrier identification phase of the process conducted within three pilot study catchments ...... 46 Table 3: Summary of key efficiency results of the barrier identification phase of the process conducted within the three pilot study catchments...... 46

iv SD 4: Evaluating small barrier removal to improve refuge connectivity

ABSTRACT

Severe surface flow reduction is exacerbating the impacts of the existing stressors on aquatic ecosystems in southern Australia. River regulation via the creation of instream barriers such as dams and weirs is known to have wide ranging impacts on aquatic ecosystems and mitigating these barriers can greatly benefit aquatic fauna, particularly freshwater fishes. Given the continued reduction in surface flows and aging of barrier infrastructure, there will be an increasing need to assess, prioritise and decommission such structures particularly in southern Australia. Whilst considerable attention has been given to barrier mitigation and prioritisation in eastern Australia, there is less information on how barriers impact the ecosystems in the often intermittent rivers across southern Australia; particularly with regard to the relationship between barriers and refuge pools. There is also limited information on the impacts of complete barrier removal as opposed to barrier retrofitting and there is no consistent process for identifying and prioritising barriers for mitigation or removal across southern Australia. The current project aimed to review the global literature on the impacts of barriers and the processes that exist for prioritising their removal. It then aimed to develop and trial a process for instream barrier prioritisation tailored specifically for systems in southern Australian.

Mitigation of instream barriers through the construction of fishways has been increasingly undertaken in Australia and internationally to reconnect fish communities however, whilst often partially effective, they are limited in terms of fully reconnecting fish communities and are also costly. Complete removal of barriers has therefore increasingly been undertaken particularly in the United States and Europe to completely reconnect river reaches. Both strong advocacy and opposition can exist to instream barrier removal projects and it is vital to have broad stakeholder involvement; particularly at a local level.

Artificial instream barriers can also actually create important refuge habitats that may increase in significance particularly as surface flow reductions continue in southern Australia. However, spatial information on refuge pools across southern Australia is limited as are their relative ecological significance. Barrier prioritisation processes in Australia have usually utilised score and ranking systems but little regard has been given to optimising multiple barrier mitigations. The barrier identification and prioritisation process we developed is a stepwise protocol that is underpinned by broad stakeholder involvement and can be applied on multiple scales from single rivers to multiple catchments. It identifies existing information on barriers and aquatic fauna and also confirms barrier and refuge information using a cost-effective surveying protocol. It includes a score and ranking system that weights both the positive and potential negative impacts of removing instream barriers and incorporates information on species diversity, habitat availability, and spatial information on barriers and refuges.

The process was trialled in three catchments in south-western Australia. Information on potential barrier locations and fish distributions was obtained by accessing GIS and distributional databases and undertaking local landholder surveys. The rapid aerial survey technique was found to be highly effective at confirming GIS information and identifying new barriers. The score and ranking system revealed that the least modified catchment had the highest scoring barriers. The information contained in this review will be of considerable interest to managers of fluvial ecosystems in temperate Australia and the prioritisation process will be a valuable and easily implemented tool in identifying and mitigating the impacts of in-stream barriers in southern Australia in a drying climate.

SD 4: Evaluating small barrier removal to improve refuge connectivity 1

EXECUTIVE SUMMARY

Many aquatic systems in southern Australia have distinct seasonal flow periods resulting from Mediterranean and cool temperate climates. However, parts of southern Australia, particularly south-western Australia, have undergone major rainfall and surface flow reductions in recent decades. For instance, total annual discharge in rivers in south-western Australia has reduced by ca 50% since 1970. This reduction is predicted to continue as is the number of no-flow days in rivers of the region. The aquatic fauna of southern Australia is adapted to the prevailing climatic regimes. Most freshwater fishes, for example, breed during higher flow periods in winter and spring, with many undergoing upstream spawning migrations in rivers and subsequent contraction of populations to permanent refuge pools during dry summer and autumn periods. Reduced rainfall and flow has major implications for these lifecycles in terms of reducing (spatially and temporally) both access to spawning habitats and the amount and longevity of refuge habitat available during baseflow.

A reduction in rainfall and surface flow is also predicted to reduce the viability and value of water supply dams as reflected recently by shifts in strategies to securing water supply for major capital cities in Australia. This, coupled with the finite lifespan of dams, will result in increasing numbers of dams becoming redundant in southern Australia. Instream barriers are known to cause numerous negative impacts to the lifecycles of both obligate freshwater and diadromous fishes, as well as affecting other components of the aquatic biota and ecological processes of lotic systems. Therefore, retrofitting or removing barriers can provide enormous benefit in terms of increasing river connectivity and facilitating fish migration, which in turn can increase habitat availability and population viability. However, it should also be noted that artificial pools are often created immediately upstream of barriers and can provide refuge habitat and these may become more important as naturally occurring pools in intermittent or ephemeral systems of southern Australia are lost in drying climatic conditions.

Many instream barrier inventory and remediation projects have been undertaken in recent years in Australia, particularly in the Murray Darling Basin and on the east coast drainages; however no consistent, standardised guidelines or methodologies for identifying and prioritising barriers for mitigation have yet been developed across jurisdictions. The consideration of the relative importance of natural and artificial refuge pools and how barrier mitigation or removal can impact them has not previously been determined in Australia. Nor has there been a review that focuses on issues surrounding the removal of instream barriers as an option for barrier mitigation.

To this end, the initial aim of the current study was to review globally the approaches taken to identify and remove instream barriers, and to identify and discuss the positive and negative impacts of instream barrier removal. Using this information, the study then aimed to develop a process for identifying and prioritising redundant instream barriers for removal that may be relevant throughout southern Australia in light of past and ongoing flow reductions. Whilst the process results in the prioritisation of all barriers for mitigation including those that is non-redundant, we deliberately focus on issues surrounding barrier mitigation through actual physical removal (rather than fishway design and construction) as this is the least understood yet potentially the most beneficial and cost-effective means of improving refuge connectivity if undertaken in a rigorous manner. The study was undertaken so that managers and decision makers have a thorough understanding of global trends in dam construction and removal, and the state of knowledge on impacts of removing barriers, including issues that may need to be addressed when undertaking barrier removal projects. Social, economic and ecological impacts of barrier removal are reviewed including levels of public support

2 SD 4: Evaluating small barrier removal to improve refuge connectivity and advocacy for removals, the causes of and how best to address opposition to removals, and cost-effectiveness of dam removal versus retrofitting with fish passage infrastructure.

Most of the literature on barrier removal projects relates to large dams (i.e. typically with a crest height of >10-15m) despite small barriers (i.e. <10m crest height) being more commonly removed. However, many (not all) of the lessons learnt from large dam removal are broadly applicable to smaller structures that can equally represent major obstructions to river connectivity and fish migrations. Dam creation increased markedly following the Second World War and peaked in most parts of the world including Australia between 1960 and 1980; although a high rate of dam construction has continued since then in many countries including China and India. However, in the United States and Australia, the construction of dams has slowed since the 1980s. Dam decommissioning now exceeds construction in the United States, and as of 2010 a total of 888 mostly small (i.e. <10m high) dams had been removed there. Whilst most dams that have been removed are small redundant structures, increasing numbers of larger dams have recently been removed in the United States. In Europe in 2000, more than 10,000 dams (>3m height) were approaching their concession renewal date and in Spain alone more than 300 mostly small dams had been removed. This trend highlights the increasing recognition of the negative impacts of dams and the cost-effectiveness of their removal cf upkeep and maintenance costs. This has particular relevance for southern Australia due to ongoing flow reductions reducing the economic value of these structures.

Dams are removed for one or a combination of four major reasons: 1) Unacceptable level of impacts. 2) Excessive expense to maintain. 3) No longer serving a useful purpose. 4) Legislation or policy for removal based on the preceding three factors.

Determining the number of instream barriers in a given geographical area can be difficult, particularly at larger spatial scales, yet is an important component of any removal process; including the one proposed in the current study (see below). The actual removal of small barriers is often driven or managed by local communities or government agencies, but has, in the past, been hampered by a lack of consistency in data collection and reporting, a lack of coordination or integration of studies across disciplines within projects, and a lack of centralised data management. Much of the information on dam decommissioning and removal projects and processes is available only in the grey literature. The Clearing House for Dam Removal goes some way towards addressing data collation on removal projects and should be accessed for existing information and also contributed to as part of future barrier removal projects in southern Australia.

Given the known ecological impacts of barriers, their removal also has a number of positive impacts that may include: • Increased fish passage for migratory species. • Reduced favourable habitat for exotic species. • Halting of thermal pollution. • Reinstating of natural sediment and nutrient transport processes.

However, removals are an ecological disturbance in their own right and therefore exert some negative impacts that vary in scale depending on a number of factors. The major negative ecological impacts can be sedimentation of downstream habitats causing impacts on fish and macroinvertebrates, pollution by toxins trapped in the sediment on

SD 4: Evaluating small barrier removal to improve refuge connectivity 3 the downstream environment, loss of populations of native species that exist within the lentic waterbody, and spread of exotic species. Careful consideration should be given to these impacts, and others summarised in this review, to ensure appropriate decisions to remove barriers are made in light of stream rehabilitation objectives. Many of these impacts can be mitigated by careful planning and assessment of each project, particularly with regard to managing the issue of sedimentation. Prudent strategies include undertaking the decommissioning at an appropriate time and rate, and ensuring that exposed reservoir banks are stabilised and revegetated. Robust environmental monitoring prior to and following barrier removal is strongly advised in order to quantify impacts and facilitate adaptive management throughout each project.

Whilst there are strong dam removal advocacy groups particularly overseas, there is often strong opposition to dam removal that reflects the social significance of these projects. Social concerns include:

• Support for existing services that the dam provides (e.g. stock watering, irrigation, fishing, swimming, boating). • Perceived high cost of removal. • Historical value of the dam or impoundment • Concerns of a lack of recovery of the streamline to an acceptable state. • Loss of tourism opportunities. • Loss of property value, and • Uncertainty or conflict over ownership.

To help mitigate opposition, it has been shown that stakeholder involvement, education and effective consultation are important throughout the project in order to understand and address opposition to barrier removal projects. As the process developed in the current project focuses largely on the removal of small redundant structures, ensuring that all stakeholders are engaged in determining the level of redundancy is advisable. This engagement may be driven by the full range of project partners and organisations.

The growing worldwide interest in riverine restoration through instream barrier mitigation/removal coupled with a limited amount of funding and resources to tackle the issue has given rise to a number of different methods of prioritising the order in which barriers are mitigated or removed. The most commonly used methods are simple score and ranking systems using a range of ecological, logistical and socio-economic assessment criteria. These methods can be applied rapidly and require little in the way of technical expertise, but have been criticised by some researchers as being overly simplistic. Recently, a number of more robust optimisation based methods have been developed which factor in cumulative impacts of multiple barriers within catchments. These methods aim to mitigate barriers in order to maximise the increase in longitudinal habitat connectivity given a particular financial budget. However, it is unclear if they will become popular given the specialist skills and knowledge needed to apply them as well as the criticism that some researchers have recently levelled at these methods.

In Australia over the past 15 years, a number of processes have been implemented in various States for identifying, assessing, and prioritising instream barriers for mitigation works. The first comprehensive project was undertaken in NSW and the process methodology developed has largely been adapted for similar programs in other States with varying degrees of modification. Reflecting the ongoing value of barriers particularly for irrigation purposes, there has been a tendency in Australia for barrier mitigation through the construction of fishways rather than physical removal. However,

4 SD 4: Evaluating small barrier removal to improve refuge connectivity fishways are usually not completely effective at passing all resident species or all life phases of particular species; therefore it is likely that physical removal as a barrier mitigation option will come under increasing consideration in the future particularly if changes in water allocation occur that result in less need for open irrigation systems.

The process proposed in this report consists of five phases, each containing a number of steps. These phases are: 1) Project planning and development. 2) Barrier identification. 3) Detailed barrier assessment and prioritisation. 4) Barrier mitigation works and monitoring. 5) Project extension.

The process is underpinned by high levels of local engagement driven principally, but not solely, by the relevant Catchment Management Agency (CMA). These groups usually have a strong relationship with private landholders and environment groups, and are also custodians of useful data on the aquatic ecosystems within their region. Early engagement of the relevant CMA (including providing appropriate financial support for their participation) through Steering Committee membership is recommended. The Steering Committee should also consist of relevant State Government agencies, traditional owner groups, land/rivercare groups, aquatic biologists/ecologists, irrigation/landholder groups, and an environmental engineer with experience in instream barriers. The CMA should be responsible for organising community workshops where project partners (i.e. Steering Committee members) can gather local stakeholder support, outline the proposed project (purpose and scale), and seek input on its development. Subsequently, community forums should be run at regular intervals throughout the process to provide updates, seek ongoing input on its progress, and address any concerns.

Information on small instream barriers is known to be sketchy and unreliable. The second phase of the process reviews all available information on the barriers and refuge habitat in the catchment(s) of interest. This includes accessing GIS databases, where available, and also surveying landholders with riparian frontage in order to determine if they are aware of potential barriers or baseflow refuge pools on their property or elsewhere. These data are then used to create a shortlist of potential instream barriers and refuge pools of interest. Locations and other relevant information for sites of interest gathered up to this point are then verified firstly using aerial or high resolution satellite imagery, followed by aerial survey by helicopter, and finally ground truthing. In conducting three pilot studies of the barrier identification phase of the process in south-western Australia, we determined that aerial surveying of instream barriers is a cost-effective step in the validation process.

All available information on distribution of fishes and other important or endangered aquatic macrofauna in the project area is reviewed. Expert opinion from a broad range of relevant stakeholders should be sought as to whether the information on the distribution and life-cycles of fish communities is adequate for decisions to be made on both the level of impacts of the existing barriers, and also the potential positive (e.g. increasing habitat range) and negative impacts (e.g. loss of a critical dry season refuge, spread of exotic species) of removing barriers. Acceptable levels of uncertainty and weighting of scores (see below) based on those levels should be made. Knowledge gaps in the understanding of the ecosystem should be prioritised and resources allocated to address them. Of major consideration are the position of the barrier in relation to dry season refuges, and the relative conservation value of dry

SD 4: Evaluating small barrier removal to improve refuge connectivity 5 season refuge habitats, if any, offered by the water body created upstream of the barrier.

We develop a score and ranking system drawing in part on those previously used elsewhere in Australia. The ranking system takes into account biodiversity values, spatial factors, hydrological factors, habitat factors and, importantly, potential negative impacts of removing or mitigating barriers such as loss of refuge habitat, facilitating the spread of exotic species and stakeholder opposition to proposed barrier removal project.

Depending on the scale of the project, interaction between barrier removals should be considered and if large numbers of barriers are identified, optimisation modelling should be considered. Using the prioritised list of barriers, levels of redundancy for each should be assessed. Those barriers assessed as non-redundant but high priority should then be considered for retrofitting (e.g. fishway construction), and those high priority redundant structures slated for removal.

A robust pre- and post- removal ecological monitoring program should be undertaken (including sediment and water quality, fish and macroinvertebrate communities, and riparian vegetation) to monitor and quantify the positive and negative impacts. Actual barrier removal should involve careful planning and be conducted under the supervision of a suitably qualified environmental or structural engineer. Of major consideration is the timing and rate of the removal. Post-removal rehabilitation of the drained reservoir is usually required to help stabilise and recreate the riparian zone and may consist of both physical and ecological phases. Finally, in addition to the above, stakeholder workshops should be held periodically throughout the project. Implementation of an effective project extension program is also necessary to create broader community understanding and support for barrier removal projects.

Within the scope of the current project, we trialled a number of the process in three south-western Australian catchments. The pilot studies highlighted both the efficiency and value of several aspects of the prioritisation process. The results indicated that whilst the GIS database of potential barriers was extremely valuable in formulating shortlists of barriers worthy of further investigation within the catchments, aerial truthing revealed a consistent level of false positives in the database in terms of barriers to fish migration and also detected numerous additional barriers. The study identified the spatial distributions of both refuge habitats and aquatic species within catchments as major knowledge gaps. These knowledge gaps would have readily been addressed by completing the process, however, this was outside the scope of the current project. The barriers within the least disturbed catchment scored much higher (i.e. approximately double) than those in the two more highly regulated catchments.

Given the fact that barrier prioritisation processes and barrier removal projects, both internationally and in Australia, have mostly occurred during the last ~20 years, there remains a number of knowledge gaps relating to the identification, prioritisation, and mitigation of in-stream barriers that are relevant to southern Australia. These include:

• Robust information on the location and impacts of small barriers. • Fine scale (i.e. to river reach and even refuge pool) distribution and ecology of freshwater fishes and other important aquatic fauna. • Understanding the short and long-term impacts of sedimentation on downstream ecosystems following barrier removal.

6 SD 4: Evaluating small barrier removal to improve refuge connectivity

• Quantifying overall ecosystem change following removal of barriers such as fish diversity and population increases in specific river reaches (cf recolonisation). • Understanding and thereby developing ways to mitigate stakeholder opposition to barrier removal. • Understanding the relative cost-effectiveness of maintaining instream barriers, compared with their retrofitting or removal.

Finally, understanding the location and ecological value of freshwater refuge habitats across southern Australia will become increasingly important when considering barrier mitigation and removal, particularly given the projected changes that are likely to occur with continued drying climatic conditions.

SD 4: Evaluating small barrier removal to improve refuge connectivity 7

1. OBJECTIVES OF THE RESEARCH

There is genuine concern that many freshwater fishes across southern Australia may not be able to withstand the combined effects of severe projected climate change and existing anthropogenic stressors such as habitat degradation, introduced species, and river regulation and therefore extinctions of the rarest and vulnerable species may be difficult to avoid without direct human intervention (Morrongiello et al. 2011). Other aquatic biota are also susceptible to shifts in community composition and population viability due to these effects (Stanley et al. 1997; Davies 2010; Robson et al. 2011). Instream barriers such as dams, weirs, and road crossings are known to have numerous detrimental effects on aquatic ecosystems, particularly on freshwater fish communities.

Whilst barrier inventory, prioritisation and remediation projects have been undertaken over the past ~two decades in Australia, particularly in the Murray Darling Basin and on the east coast drainages; no consistent, standardised guidelines or methodologies for identifying and prioritising barriers for mitigation have yet been developed across jurisdictions. Furthermore, the consideration of the relative importance of natural and artificial refuge habitats and how barrier mitigation or removal can impact them has not previously been determined in Australia. Nor has there been a review that focuses on issues surrounding the complete removal of instream barriers as an option for barrier remediation.

Therefore, the initial aim of the current study was to undertake a global review of the literature on the approaches and processes that have been taken to identify, prioritise and remove instream barriers. The review also aimed to comprehensively review the positive and negative impacts that the decommissioning of instream barriers can have so that managers and researchers in southern Australia can make informed decisions on their management. Using the information gathered in the review, the study then aimed to develop a process for identifying and prioritising instream barriers for remediation (including complete removal) that may be relevant throughout southern Australia taking into account the increasing significance of refuges to its aquatic fauna in light of projected reductions in surface water availability.

The effects of instream barriers on refuge connectivity in aquatic systems are highly relevant for fishes, and there have been many studies around the world into the impacts of instream barriers on fish movements. Instream barriers are also known to impact on other biota and ecological processes more generally. While the current study does review those impacts to some extent, it deliberately focuses on the issues surrounding the impact of instream barriers on fishes. The process subsequently developed for barrier prioritisation herein, while focused on fishes, aims to be readily adaptable for including information on any component of the aquatic fauna in a catchment or region of interest.

8 SD 4: Evaluating small barrier removal to improve refuge connectivity

2. INSTREAM BARRIER REVIEW

2.1 Global trends in the construction of instream barriers Humans are exercising an increasing degree of control over the most vital natural resource upon which civilisation depends – fresh water. The construction of instream barriers such as dams and weirs across river channels is the most prevalent means by which this has occurred, with the number of dams worldwide estimated to be in the millions (Katopodis & Aadland 2006). Dams supply water for domestic needs, irrigation, waste containment, recreation, and the generation of hydro-electric power, while other human-made instream barriers (e.g. weirs, flood gates, road crossings, and locks) are constructed for flood and erosion control, flow regulation, land access and waterway transportation purposes (Katopodis & Aadland 2006; O’Hanley 2011).

Information and literature on small dams (i.e. crest height <10m) is less available than for large dams (i.e. crest height >10-15 m), in fact, as far as we are aware there is no global estimate on the total number of small dams (as also found by the review of Lejon et al. 2009). However, lessons, procedures, and even impacts of large dams can often be applied to smaller structures which are the focus of the current sub-project. The rate of construction of large (>4 storeys or 15 m high) dams increased rapidly in the period following the Second World War, prior to which there were ~5,000 globally (WCD 2000). By the end of last century this number had increased to 45,000 (worth a total estimated value of $2 trillion) with the greatest number found in China (~22,000; an increase from just 22 prior to 1949) (WCD 2000). Dam construction peaked in the early 1970s (~5,000 being constructed) and the average age of dams as at the year 2000 was 35 years old (WCD 2000). However, China and India continue to construct dams at a phenomenal rate with an estimated 40% of the ~1,700 under construction in the year 2000 being in India (WCD 2000). In the United States, the total number of dams (large and small) was estimated at ~80,000 structures >3 m in height and, astonishingly, ~2,000,000 total river blockages (Service 2011).

In Australia, there are 231 large storage dams (SoE 2011), however, The Australian National Committee on Large Dams Incorporated (ANCOLD) lists ~560 large (i.e. >10 m height) dam projects (which includes projects that have altered or retrofitted existing dams). The construction rate of dams in Australia increased rapidly from 1950, peaking at ~90-100 dams/yr between 1960-1990, and has steadily declined since that time (Figure 1).

SD 4: Evaluating small barrier removal to improve refuge connectivity 9

ACT Large dam projects in Australia NSW 120 NT QLD SA 100 TAS VIC WA 80

60

40 Number of projects Numberof

20

0

1870 1880 1890 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 2010 2020

Decade

Figure 1: Large dam creation in Australia (data from the Australian National Committee of Large Dams).

Pressures on water resource development in Australia are highlighted by predictions that consumption across capital cities will increase (relative to 2009) by 39-49% in 2026, rising to 64-107% in 2056 (SoE 2011). Water security through diversity of supply has been developed as the new paradigm in Australia over the past decade, and this is reflected by the increase in ‘manufactured’ water sources; particularly water recycling and desalinisation (SoE 2011). For example, desalinised water supply in Australia increased from 231 ML in 2004-2005 to 33,270 ML in 2008-2009 (SoE 2011). Given the past and projected rainfall and surface flow reductions across southern Australia (particularly in the south-west, see Silberstein et al. 2012), current planning is focusing on water conservation, desalinisation, and recycling to meet demands, and this explains the reductions in major dam construction over the past 20 years (Figure 1). For example, in Perth, a 10-15% reduction in rainfall since the 1970s has resulted in a 50% reduction in surface water flow into large storage dams, and by 2060 it is expected that dams will no longer provide a reliable contribution to the water supply (Water Corporation 2009; SoE 2011). In the Victorian jurisdiction of the Murray-Darling Division, surface water diversions during dry years are expected to fall by 35->50% under median 2030 climate scenarios (SoE 2011). Therefore, along with the finite lifespan that instream structures such as dams have, their reduced value due to flow reductions in southern Australia will increase the rate at which they become redundant from an economic perspective.

Despite the pressure to secure water supplies, many recent water extraction projects (surface and groundwater) have been rejected due to overwhelming public opposition

10 SD 4: Evaluating small barrier removal to improve refuge connectivity and associated impacts on environmental grounds. For example, a dam project at Traveston Crossing on the Mary River in Queensland was rejected in 2009 by the Federal Government on environmental grounds; particularly due to its likely impact on the Australian Lungfish, Mary River Cod and Mary River Turtle and associated environmental concerns and fishway inadequacies in the region including Paradise Dam on the Burnett River (Walker 2008). Likewise, the south-west Yarragadee groundwater proposal in Western Australia was rejected in 2007 due to unacceptable environmental impact along with agricultural water supply issues, for example, the latter project would have impacted listed threatened fish species including Balston’s Pygmy Perch and Western Mud Minnow (Beatty et al. 2009). These projects highlight how listed threatened species can play a considerable role in determining the fate of water resource development projects. It is less clear, however, how the trend in construction of small dams and other instream barriers has been evolving in Australia.

2.2 Impacts of instream barriers There have been many detailed reviews and studies into the impacts of instream barriers on fluvial environments (see for example Ward & Stanford 1979; eds Craig & Kemper 1987; Poff et al. 1997; Bunn & Arthington 2002; Arthington 2012). A brief summary of the major hydrological and ecological impacts is provided in this section, with a focus on fishes and their migratory movements. Whilst there is a comprehensive body of literature on this subject, extensive research still remains to be done in this field, particularly addressing the shortfall in knowledge of basic life history characteristics of the vast majority of aquatic species, in southern Australia and around the world.

Dams cause inundation of river valleys, in the process destroying riparian vegetation and associated faunal communities, and create lotic habitat where lentic habitat previously existed (Bednarek 2001; Bunn & Arthington 2002; Gregory et al. 2002; Katopodis & Aadland 2006). Shifts in the physico-chemical water profile within and downstream of dams are also common (Stanford et al. 1996; Bednarek 2001; Gehrke et al. 2001). The alteration of natural hydrological regimes and reduction of peak flows caused by dams and other barriers is often profound (Poff et al. 1997; Bunn & Arthington 2002). There are innumerable examples around the world of rivers that rarely or never discharge to the ocean as a result of damming and flow regulation, including some of the largest systems on the planet (Arthington 2012). Curtailment of natural processes of flooding in catchments reduces connectivity between river channels and floodplain habitats, and disrupts natural processes of erosion, sediment transport and deposition, channel scouring, and nutrient cycling (Junk et al. 1989; Kondolf 1997; Katopodis & Aadland 2006). The trapping of sediment and organic debris can interrupt ecological processes in rivers, altering habitats and the composition and abundance of biotic communities downstream of barriers, particularly in heavily impounded systems (Katapodis & Aadland 2006). 2.2.1 Biological impacts on fishes Biological impacts of instream barriers typically manifest as a loss of aquatic faunal diversity and abundance, and the contraction of the distributional range of native species (Li et al. 1987; Lucas & Baras 2001; Gehrke et al. 2002; Catalano et al. 2007; Slawski et al. 2008; O’Hanley 2011). The instinct of migratory fishes to move upstream results in unnaturally large congregations below impassable barriers and increasing exposure to predation (Buchanan et al. 1981; Gregory et al. 2002; Morgan & Beatty 2003). More significantly however, barriers can reduce access or prevent these species from reaching habitats which may be necessary for spawning, feeding, nursery areas or as refuge during drought or flood (Northcote 1998; Lucas & Baras 2001). This effect is especially pronounced for diadromous species that must migrate between

SD 4: Evaluating small barrier removal to improve refuge connectivity 11 marine and freshwater habitats to complete their life-cycle. For example, the proliferation of dams in North America has cut off access to an estimated 40-80% of anadromous salmonid spawning habitat, contributing to significant declines of a number of species (WWF 2001; Gregory et al. 2002; Sheer & Steel 2006). In south-eastern Australia, the proliferation of instream barriers has been implicated in the dramatic decline of native fish species such as the Australian Bass, Macquarie Perch and Golden Perch (Harris 1984; Lake & Marchant 1990). Potamodromous species (i.e. those that undertake migrations between different freshwater sections of rivers), including many native southern Australian freshwater fishes, are also affected by instream barriers (Thorncraft & Harris 2000; Allen et al. 2002; Morgan et al. 2011).

Barriers can also impede downstream movements of catadromous (e.g. Barramundi) and potamodromous species (e.g. Murray Cod, Golden Perch and Silver Perch) and limit the dispersal of propagules and larvae of species that spawn in the headwaters of streams (NSW DPI 2006a). Furthermore, habitat fragmentation in catchments with reduced longitudinal connectivity due to barriers can result in isolation and a loss of genetic diversity, compromising the sustainability of affected populations (Wofford et al. 2005; Katopodis & Aadland 2006).`

Species that migrate in order to reproduce are mainly impacted due to the physical restriction on movements imposed by barriers, however, alteration of the natural flow regime can also impact negatively as this provides a requisite cue for many species to commence spawning (Nesler et al. 1988; Poff et al. 1997; Bunn & Arthington 2002; Naiman et al. 2002). In seasonally flowing watersheds that are highly impounded, variation from a natural flow regime is especially apparent during the shoulder period between base and peak flow conditions when the majority of surface runoff is captured behind instream barriers with greatly reduced spillover (SKM 2007). For aquatic species that rely on flow triggers to mature and spawn, the potential duration of breeding periods can be shortened substantially.

Artificial barriers can also have a negative impact on fishes through the alteration of instream habitat. Flow regulation has been demonstrated to reduce habitat quality and complexity (Beechie et al. 1994), both of which have been shown to be crucial determinants of the structure of aquatic species assemblages (Meffe & Sheldon 1988; Pusey et al. 1993, 2000). Heavily modified catchments with altered flow regimes are also more favourable to invasive exotic species which tend to be habitat generalists (Moyle 1986; Arthington et al. 1990; Stanford et al. 1996; Morgan et al. 2002; Gehrke & Harris 2001; Clavero et al. 2004; NSW DPI 2006a). Once established, exotic species can exert a multitude of negative impacts upon native species including predation, competition, degradation of aquatic habitats and water quality, and spreading of diseases and parasites (Fletcher et al. 1985; Morgan et al. 2004; Lymbery et al. 2010; Olden & Cucherousset 2011).

The significance of these impacts on fluvial systems and resident fauna is reflected in the recognition of instream barriers as a “threatening process” under various Acts of legislation in two Australian states (i.e. New South Wales and Victoria) (NSW DPI 2006a, DSE 2003), in the River Murray as part of the River Murray Act 2003 and the the Murray Darling Basin Native Fish Strategy 2003-013 (MDBC 2003), nationally as a direct impact on threatened species under the EPBC Act 1999 and elsewhere globally. In North America for example, over half of the listed endangered species are recognised as such due to the impacts of river infrastructure (Losos et al. 1995) and in recent years federal courts in the United States have ordered dam managers to address the impact that dams are having on water quality in accordance with various state and federal policies including the Clean Water Act (Gregory et al. 2002). Legislative acknowledgment of the impact of instream barriers has also occurred in

12 SD 4: Evaluating small barrier removal to improve refuge connectivity

Europe in the past decade with the ratification of the EU Water Framework Directive (Kemp & O’Hanley 2010). 2.2.2 Instream barriers and refuges in southern Australia: impacts and the influence of climate change The impacts of instream barriers outlined above are globally pertinent but are exacerbated in Australia where there is a paucity of permanent freshwater habitats over vast tracts of the southern half of the continent. Many rivers here are highly seasonal, due to prevailing climatic conditions, ceasing to flow for many months each year during which time large portions of catchments consist of isolated pools separated by sections devoid of all surface water. The native aquatic biota have adapted life history strategies to persist in these environmental conditions, with perhaps the most remarkable example being the Salamanderfish Lepidogalaxias salamandroides. This species occupies the ephemeral peat-flat swamps of the south-western corner of Australia and is capable of burrowing into the substrate when all surface water evaporates and aestivating during the dry summer/autumn period until winter rains return to refill its habitat (Allen & Berra 1989; Berra & Allen 1989). A number of other fish species, including many representatives of the family Galaxiidae, utilise lower order tributary streams as spawning and nursery areas during periods of flow before retreating to permanent mainstream refuge pools when dry conditions set in (see for example Pen & Potter 1991a). Instream barriers can impede these seasonal movements and also cause mortality if fish become stranded in non-permanent stream sections when flows subside. Stranding is likely to have a greater impact on juvenile fish, as has been demonstrated for some salmonids in North America (Bradford et al. 1995; Bradford 1997), because younger fish have presumably not yet acquired appropriate behavioural responses to fluctuating stream flow.

Climate change is predicted to impart increasing stress on the resilience of freshwater fauna in southern Australia (Davies, 2010; Morrongiello et al. 2011), especially in catchments where natural hydrological regimes have been heavily modified by instream barriers (Hughes 2003; Robson et al. 2011). In the south-west of Western Australia for example, winter rainfall has decreased substantially since the middle part of last century, a trend that has been particularly evident since the 1970s (Suppiah et al. 2007). Rainfall has declined by 10-15% since that time with a concomitant reduction in surface discharge into potable dams of ~50% (DoW 2010) and temperatures have increased (IOCI 2008). There has also been a trend of lower autumn and early winter rainfall and increased late winter and summer rainfall (IOCI 2008). Further reductions in rainfall (~8%) and surface water discharge (~25%) are predicted over the next 20 years (Suppiah et al. 2007; CSIRO 2009a,b; Silberstein et al. 2012). The number of no-flow days in some rivers in south-western Australia is also predicted to increase by up to 4 months by 2030 (Barron et al. 2012). Furthermore, the length of drought periods and days of extreme heat are also predicted to increase leading to higher evaporation rates (CSIRO 2009a,b). Consequently, the amount of available permanent refuge habitat, connectivity between refuge pools, and the overall aquatic faunal carrying capacity of catchments throughout the region are likely to decline.

Future low flow scenarios will exacerbate the impacts of instream barriers on native aquatic fauna through increased temporal disruption of river connectivity and restriction of access to permanent refuge habitat. Paradoxically, pools formed behind instream barriers are likely to comprise a greater proportion of the available permanent refuge habitat in some catchments in southern Australia (Gasith & Resh 1999). For example, in surveying the distribution of fishes in south-western Australia, Morgan et al. (1998) found many artificially created refuges supporting native fish populations during low

SD 4: Evaluating small barrier removal to improve refuge connectivity 13 flow periods. Therefore, when developing and implementing barrier removal or mitigation projects, it is pertinent to consider each barrier in terms of its potential positive and negative impacts on resident aquatic fauna.

2.3 Restoring fish passage: barrier modification versus removal Enhancement of fish passage over instream barriers can be achieved by removal (breaching), repair/modification, or installation of a fish passage structure (Kemp & O’Hanley 2011). Barrier removal is generally accepted as the most effective method for achieving this goal but is often not undertaken due to issues related to the current use and ownership of the structure. River infrastructure provides goods and services that may be deemed too expensive or impractical to source elsewhere, and there is often overwhelming opposition to proposed decommissioning projects (Gregory et al. 2002). Retrofitting of barriers with fish passage structures (i.e. fishways or fish ladders) is therefore generally more common.

Despite the phenomenal rate of dam construction over the latter part of last century, increased instances of water shortages in many parts of the world have fuelled a push to determine how effective dams really are for water supply, and simultaneously to compare their benefits with their impacts (Stanley & Doyle 2003). The World Commission on Dams (WCD), founded in 1997 following an IUCN World Bank sponsored workshop, issued a comprehensive report in 2000 entitled Dams and Development: A new framework for decision-making (WCD 2000). The report aimed to develop a balanced and sustainable framework, including criteria and guidelines, for decisions relating to water and energy resources to help overcome the wide variety of issues and conflicts that dams have historically caused. The report included a global review on dams, their performance (economic, environmental, social), and also examples and broad principles for addressing existing dams, including decommissioning (contained within Strategic Priority 3). As well as highlighting the flaws in Environmental Impact Assessments (EIAs) of dam construction when they had occurred, the review also stated, “one of the most disturbing findings….was the lack of monitoring of the impacts of dams and the complete failure to conduct proper ex-post evaluations of performance and impacts.” It also highlighted the problem of the lack of fishways on larger dams (noting the situation at that time in Australia; which has improved over the last decade, see section below on Australia). For example, <10% of the 1,825 hydropower dams in the US, and <5% of the 450 largest South African dams had fish passage facilities (WCD 2000).

Fishways are low-gradient, stair-like structures featuring a series of steps interspersed with resting pools, designed to dissipate flow velocity and turbulence thereby allowing fish to swim up and over instream barriers (see Katopodis 1992; Mallen-Cooper 2000 for comprehensive reviews of fishways). Globally, fishways have proven most effective at facilitating the passage of large-bodied salmonids but often block or impede smaller- bodied, slower swimming species (Katopodis & Aadland 2006). Many fishways around the world have proven to be largely ineffective at passing fish. For example, ~73% of 34 fishways constructed in the Glomma and Laagen Basin in Norway were found to be working ineffectively or not at all (WCD 2000). In most EU countries, the effective provision of fish passage over instream barriers is considered to be a legal obligation of waterway managers, with a further requirement for the improvement of fishway efficiency (eds Kroes et. al. 2006). In Australia, fisheries and conservation organisations in New South Wales, Queensland and Victoria are seeking to legislate similar responsibilities for waterway managers. As such, an approved operational

14 SD 4: Evaluating small barrier removal to improve refuge connectivity protocol will need to be established and regular monitoring of fishway performance and efficiency will be required (O’Brien et. al. 2010).

In Australia, native fish species tend not to possess the equivalent swimming and jumping ability of northern hemisphere salmonids (Mallen-Cooper 1992) and an inappropriately designed fishway can be as impassable to many species as the barrier that it was designed to mitigate (Mallen-Cooper 1994). For example, some fishway structures in south-western Australia have proven to be highly effective at passing fast- swimming Galaxias species but mostly ineffective for passing other species probably due to excessive water velocity or turbulence in the fishway and the poorer swimming ability of those sympatric species (Morgan & Beatty 2004a,b, 2006; Beatty & Allen 2008; Beatty et al. 2007, 2012). The Western Pygmy Perch, a poorer swimmer in comparison to sympatric Western Minnow, was shown to be capable of ascending a rock-ramp fishway in appreciable numbers, but only in the latter part of its breeding season at sub-peak flow velocities, thus limiting the magnitude of seasonal upstream dispersal and reducing the potential for population expansion and genetic exchange (Beatty et al. 2012).

The most effective means of increasing the longitudinal connectivity of rivers is the physical removal of instream barriers. Barrier removal is generally undertaken for four main reasons: 1) level of impact is unacceptably high (ecological or social), 2) maintenance is prohibitively expensive (financial), 3) the barrier no longer serves a useful purpose (redundancy) and, 4) law or policy driven (resulting from one or more of the preceding reasons) (Stanley & Doyle 2003; Lejon et al. 2009). In many cases the decision to remove barriers has been motivated as much by financial reasons as environmental ones, as the costs to repair aging dams and install fish passages to bring them into line with required safety and environmental standards often greatly outweigh removal costs (Graber 2003; Graf 2003; Katopodis & Aadland 2006; Gordos et al. 2007). In other instances, barriers have been removed as they have become disused and redundant or hazardous to human safety (e.g. Chaplin et al. 2005; Burroughs et al. 2010; I&I NSW 2009). Instream barrier decommissioning activity has increased markedly in the last two decades around the globe with the majority being relatively small (e.g. Bednarek 2001; Stanley & Doyle 2003; Bernhardt et al. 2005; Pess et al. 2008; I&I NSW 2009; American Rivers 2010; Service 2011). Whilst barrier decommissioning has delivered unquestioned benefits through restoration of ecological processes and reviving of fisheries, it is vital that careful assessment, planning, monitoring, and public consultation take place to address opposition to dam removals, and also reduce the potential for negative ecological and social impacts that removal may cause; particularly regarding sediment flushing (see WCD 2000; Stanley & Doyle 2003; Service 2011).

As highlighted in the current review, there are currently no consistent guidelines for the decommissioning of redundant dams in Australia although this has been identified as a future project (M. Miller, ANCOLD pers. comm.). Much of the information on dam decommissioning is centred around the United States where most projects have occurred (Graf 2003; Pohl 2003; Stanley & Doyle 2003), and also Europe (e.g. Lejon et al. 2009). The bulk of information (with many notable exceptions mentioned in the current review) is found only in the grey literature on a project by project, jurisdictional (often State) basis, or collated by certain environmental advocacy groups (see section below Dam decommissioning advocacy). In 2002, a workshop was held on dam removal research in the United States in which contributors covered social, environmental, and economic aspects, as well as providing a census of dam removals that had occurred up until that time (ed. Graf 2003). Pohl (2003) states that the existing data on removal were limited by: inconsistencies in reporting styles, lack of coordination or integration of studies on specific dam removals, and lack of centralised

SD 4: Evaluating small barrier removal to improve refuge connectivity 15 data management. As we discuss below, the latter issue has been at least partially addressed with the creation, in 2006, of a data repository for dam removal related projects, their associated studies and also planning decisions which will greatly aid managers and decision makers in tackling future decommissioning projects.

2.4 Impacts of barrier removal Given the numerous negative impacts associated with barrier construction, the removal of barriers has many ecological, social and economic benefits (e.g. Graf 2003). However, barrier removals can, on the other hand, cause profound ecosystem changes (e.g. loss of the lentic ecosystem) and also social change (e.g. loss of recreational use of a waterbody). Therefore instream barrier removal decisions should be viewed as trade-offs between positive and negative impacts (e.g. Bednarek 2001; Stanley & Doyle 2003; eds Wilson et al. 2007). Decisions to remove or repair dams have, in the past, often been made using incomplete or inaccurate information (Born et al. 1998). Given the potential positive and negative impacts, it has been proposed that dam removal should be considered an ecological disturbance in its own right (Stanley & Doyle 2003) and therefore consideration must be given to qualitative and quantitative ecological changes, and the temporal and spatial scales over which they operate. Hart et al. (2003) proposed a stressor-response conceptual model to determine the effects of existing dams, with the difference between the stressor-response curve and reference condition providing a measure of predicted change in ecological integrity that would occur through the removal of a dam. However, this study was limited by a lack of existing data on small dams (i.e. the type being considered for decommissioning) required to convert the conceptual model into a predictive form. Below we outline the major impacts of barrier removal and provide some discussion of positive and negative aspects. 2.4.1 Hydrological restoration and improved river connectivity Removing instream barriers causes artificial lentic habitat to be transformed back to its original lotic state and also restores more natural temperature and sediment transport regimes (Bednarak 2001; Gregory et al. 2002; Katapodis & Aadland 2006; Feld et al. 2011). These ecological benefits are most prevalent near former barrier sites and immediately downstream but the zone of influence can extend much further throughout catchments, even as far as near coastal marine waters in the case of sediment transport (DOI 1995; Bednarak 2001; Gregory et al. 2002).

In previous case studies on dam removals in the United States, researchers postulated that the re-establishment of a more natural regime of fluctuating river flows contributed (along with improved river connectivity) to observed increases in fish diversity, abundance, and biomass at formerly impacted sites, as well as improved water quality (Hill et al. 1993; Estes et al. 1993; Kanehl et al. 1997). This was, in part, due to substrate stabilisation and consolidation that occurred under the restored flow regime following dam removal which led to an increase in macrophyte cover (e.g. Hill et al. 1993; Estes et al. 1993). The flushing of accumulated sediment at formerly impounded sites can also lead to the restoration of aquatic habitats such as gravel and cobble bottoms which are crucial spawning habitats for salmonids (Dadswell 1996; Kanehl et al. 1997).

Altered flow regimes in impounded river systems disrupt migratory movements of resident aquatic fauna and improved longitudinal connectivity resulting from the removal of instream barriers is repeatedly identified as a major positive impact in many studies (e.g. Nelson & Pajak 1990; Simons & Simons 1991; Shuman 1995). Increased access to habitat upstream of former barriers has been documented to lead to significant increases in recruitment and productivity within relatively short timeframes

16 SD 4: Evaluating small barrier removal to improve refuge connectivity following dam removal, particularly for anadromous salmonids (Scully et al. 1990; Beamer et al. 1998; Pess et al. 2008). Barrier removal has resulted in some species recolonising areas from which they were previously extirpated due to barrier construction (Iversen et al. 1993; DOI 1995). The improvement of lateral connectivity between rivers and riparian habitats including floodplains and backwaters is another potential positive impact of dam removal that has in some cases been credited with contributing to increases in not only resident aquatic fauna, but dependent terrestrial fauna as well (Kaufman 1992; Hill et al. 1993; Shuman 1995). In places like southern Australia where rainfall and surface runoff are declining, the ailing health of many intermittent freshwater ecosystems is becoming a pressing issue. For years now, many such systems have been kept on life support through the release of so-called ‘environmental flows’ from impoundments. With predicted further reductions in rainfall, it stands to reason that the complete removal of instream barriers may be increasingly considered as a restorative measure as it would enhance longitudinal and lateral river connectivity, thereby increasing the resilience of freshwater systems to climate change. 2.4.2 Elimination of barriers as aquatic ‘predators’ Gregory et al. (2002) raised the notion that instream barriers, particularly large hydropower dams, could be considered a ‘predator’ in their own right due to their capacity to inflict injury and death to aquatic fauna. By removing these structures, injuries sustained by aquatic organisms while swimming through hydropower turbines or over large vertical drops and spillways are effectively eliminated (Travnicheck et al. 1993; Dadswell 1996; Coutant & Whitney 2000). Researchers have estimated that such injuries result in the mortality of between 5-20% of immature salmonids undertaking downstream migrations in the Columbia Basin in the western United States (Raymond 1979; Skalski 1998). Furthermore, elevated mortality rates due to natural predators consuming migratory aquatic fauna that congregate below barriers are also reduced when barriers are removed. 2.4.3 Ecological community shifts Instream barrier removals have also been documented to cause shifts in community composition (ASCE 1997; Born et al. 1998). These changes may include decreases in the abundance and diversity of exotic species which tend to be habitat generalists that favour artificial lentic environments created by dams, and increases in the diversity and abundance of native species that are adapted to the natural lotic habitats of restored rivers (Staggs et al. 1995; Kanehl et al. 1997).

The ecological impacts of removing instream barriers are not always predictable (Gregory et al. 2002), and it is generally accepted that timeframes in the order of decades, or perhaps even longer, are required before the full benefits of barrier removals are realised (Katapodis & Aadland 2006; Feld et al. 2011). Unfortunately, long-term environmental and biological monitoring activities that are necessary to determine the exact nature and extent of ecological impacts are often not carried out following river restoration projects (Feld et al. 2011). In the United States for example, less than 10% of river restoration programs include monitoring (Bernhardt et al. 2005; Palmer et al. 2005, 2007). To further illustrate this point, at least one government agency in Australia has stated that ongoing monitoring of fish populations is not required following barrier removal as it is assumed that fish passage is restored entirely (State Water Corporation 2011). Some researchers, ourselves included, do not share this viewpoint and maintain that ongoing monitoring of aquatic communities is essential for auditing the outcomes of river restoration efforts and determining how future projects can best be approached (Stewart & Marsden 2006).

SD 4: Evaluating small barrier removal to improve refuge connectivity 17

2.4.4 Sedimentation Impact of sediment flushing is often the major concern of managers when assessing dam decommissioning proposals (Randle 2003; Rathburn & Wohl 2003; Stanley & Doyle 2003). Importantly, as many dams were created for industrial purposes or within industrial areas, accumulation of toxins, heavy metals, and nutrients can cause serious environmental damage and costly cleanups (e.g. Shuman 1995; Stanley & Doyle 2003). Sediment movement is a natural process in river systems, creating vital benthic habitat diversity and transporting nutrients (Kondolf 1997; Wood & Armitage 1997). Dams disrupt the natural sediment transport process by causing fine sediment to settle and accumulate and also block the downstream movement of larger debris (Kondolf 1997; Bednarek 2001). When dams are removed accumulated sediment is flushed and can have serious impacts on downstream habitats via erosion, first vertically then laterally (Bednarek 2001; Gregory et al. 2002; Stanley & Doyle 2003).

Although there have only been a limited number of studies examining sediment transport associated with dam removal, and very few on intermittent or ephemeral systems (e.g. Hart et al. 2003; Randle 2003; Neave et al. 2009), this process is known to cause considerable downstream deposition of sediment over time, which varies depending on sediment amount and type in the reservoir, flow regimes, and deconstruction techniques, amongst other factors (Bednarek 2001; Rathburn & Wohl 2003; Doyle et al. 2003). In all dam removals, there is a delay in downstream sedimentation with the lag time depending on rate of sediment erosion in the reservoir, thickness of the accumulated layer of finer sediments in the reservoir, and the length of the reservoir (Doyle et al. 2003).

For small (i.e. the focus of the current sub-project) or even modest sized dams, sediment flushing and associated ecological impacts are generally small in scale and short lived (Pollard & Reed 2004; Thomson et al. 2005; Orr et al. 2006; Feld et al. 2011; Hart et al. 2002; Stanley et al. 2002; Doyle et al. 2003). For example, in examining the effects of removal of a small dam on downstream macroinvertebrate and algal assemblages, Thomson et al. (2005) concluded that sedimentation significantly reduced species densities in the short term (12 months), however, overall assemblage composition did not change significantly and the study concluded that for small dams impacts may be minor and temporary. Pollard and Reed (2004) demonstrated that no significant change in benthic macroinvertebrates occurred at sites downstream of a decommissioned dam in Wisconsin, rather, a shift in functional feeding groups occurred at the upstream reference site, a response that was attributed to reduced silt coverage. However, whilst de-silting (not complete removal of the dam) of a small Victorian reservoir did not affect the long term diversity and density of macroinvertebrates, densities of River Blackfish were reduced by between 59-93% between 250 m – 2.7 km downstream in the months following (Doeg & Koehn 1994). Some recovery of the species was detected in the subsequent two years of monitoring, although not to pre-disturbance densities (Doeg & Koehn 1994).

Although downstream ecological changes due to sedimentation may be relatively minor and temporary, an understanding of sediment composition (e.g. concentrations of toxins and nutrients) and instream sediment transport processes should comprise part of instream barrier removal assessments (Randle 2003; Rathburn & Wohl 2003). Minimising the impact of sediment transport during removal projects can be achieved by appropriate timing and pacing of removal, along with other mechanical removal or stabilisation approaches such as grading or armouring river banks with geotextiles or riprap (Randle 2003; Rathburn & Wohl 2003). Furthermore, sound baseline ecological data should be obtained for the downstream system to assess the potential impacts of sedimentation, develop mitigation strategies, and monitor the impacts and recovery

18 SD 4: Evaluating small barrier removal to improve refuge connectivity during and after barrier removal. Monitoring programs should also span sufficient timeframes to detect longer term change and recovery of affected ecosystems (i.e. Coe et al. 2008). 2.4.5 Spread of exotic species Artificial water bodies such as reservoirs, particularly those commonly utilised for human recreation, are often hot spots for exotic species (Bunn & Arthington 2002). Concerns have been raised among experts in the United States over the trade-off of increasing benefit to native species through barrier removal against the potential spread of exotic species (Fausch et al. 2006; McLaughlin et al. 2007). However, dispersal of exotics following barrier removal can be difficult to predict. Stanley et al. (2007) conducted fish surveys before and after the removal of two low-head dams on Boulder Creek (Wisconsin, USA) to determine if exotic Brown Trout moved into habitat upstream of the former barrier sites, but found that two years after dam removal that they had not. A feasibility study into removal of the Nicholson Reservoir dam in the Australian State of Victoria highlighted that exotic Carp were restricted to reaches below the dam and although native diadromous species would benefit from dam removal, the potential for Carp to spread further upstream (Browning et al. 2007) contributed to the recommendation to shelve this project. It is critical to understand the distributions and life cycles of both native and exotic fishes within catchments when assessing whether dam removal (of any size) should be undertaken.

The retention of instream barriers is often used as a conservation management tool. For example the threatened Barred Galaxias in Victoria of the upper Goulburn catchment is being actively managed with instream barriers to prevent the upstream migration of Brown Trout, a known predator of many Galaxias species (Raadik et. al 2010). Furthermore, in the Great Lakes catchments in North America, new barriers have actually been installed to attempt to halt the upstream spread of invasive fishes and protect resident native aquatic fauna upstream (Velez-Espino et al. 2011). 2.4.6 Impacts of barrier removal on habitat availability and influence of climate change in southern Australia Rapid de-watering of reservoirs results in near 100% mortality of aquatic fauna populations; aside from mobile fish (Stanley & Doyle 2003). However, colonisation of re-created lotic habitats with species that prefer such habitats can occur rapidly. For example, Kanehl et al. (1997) found that fish communities changed rapidly following dam removal with a reduction in exotic Carp and re-colonisation and natural recruitment by native species.

Up until 2002, there had been no dam decommissioning projects in the United States that were mandated through the Endangered Species Act (Pohl 2003). Although the impacts of dams on endangered species are relatively well known, the level of impact reversal through dam removal is less clear, aside from enabling improved passage and re-colonisation of fish. However, threatened species legislation enacted in Australia (i.e. the Environment Protection and Biodiversity Conservation Act 1999) now ensures that careful assessment of the positive and negative impacts of barrier removal on threatened species occurs; just as it does in the assessment of new water resource developments (e.g. SoE 2011).

As mentioned previously, many aquatic systems in southern Australia are naturally intermittent with distinct seasonal flow periods. Many freshwater fishes in this region have potamodromous life cycles that involve breeding during high flow periods in winter and spring before retreating to main channels of rivers or refuge pools during baseflow periods (Beatty et al. 2009). Artificial lentic systems that have been created through

SD 4: Evaluating small barrier removal to improve refuge connectivity 19 the construction of not only large potable water supplies but also small structures such as gauging station weirs or even artificially excavated pools used for fire management, are thought to act as important dry season refuges for many native freshwater fishes, including threatened species (Morgan et al. 1998; Beatty et al. 2011; Lintermans 2012). In order to determine whether or not a barrier removal will result in a net ecological benefit, it is crucial to understand the ecology of the reservoir that would be impacted by the removal of the barrier along with both the upstream and downstream ecosystems.

Continued rainfall and flow reductions projected for much of southern Australia, particularly the south-west, has direct implications for the conservation value of artificial refuges created by instream barriers (e.g. weir pools, reservoirs). The loss of naturally occurring refuge pools due to climate change for example, may increase the importance, from a conservation perspective, of these artificial refuge habitats across southern Australia in the future.

Conversely, instream barriers can prevent migratory species from accessing important spawning/nursery/feeding grounds or refuge habitats (Northcote 1998; Lucas & Baras 2001), and removal of such barriers would result in an increase in species resilience to climate change induced flow reductions. Therefore careful consideration must be given to how climate change is likely to influence the relative impacts of instream barriers prior to their removal. 2.4.7 Cost Economics is the usual driver for barrier removal as it is recognised as a cost-effective restorative tool (Sarakinos & Johnson 2003) particularly when weighed up against the often considerable cost of maintaining ailing instream barriers (Service 2011; Lejon et al. 2009). Repair cost can be on average three to five times greater than removal cost. In the United States it was estimated that more than 2,200 of 10,000 flood-control dams required maintenance at an estimated cost of $543 million (NRCS 2000). In fact, by 2020, Bednarek (2001) noted that the vast majority of major dams in the United States would reach or have passed their original life-expectancy. Between 1960-2002, 80 small dams (<8 m) had been removed in the State of Wisconsin at an average cost of ~$115k, much less than the average cost of repair of ~$700k each (Sarakinos & Johnson 2003). Many European dams were also approaching their life expectancy in the early part of this century. Although, at least superficially, the cost of dam removal is usually much less than the cost of repair, it is also important to consider the cost of appropriate impact assessments and ongoing monitoring in determining overall project costs. As with the other potential negative factors of dam removal, cost of the removal project should also be weighed against the cost of ongoing maintenance along with existing economic value and future forecasts of economic benefit of maintaining the structure. Undertaking these analyses in a rigorous and transparent manner will not only allow proper cost-benefit decisions to be made, but also help mitigate opposition to removals.

2.5 Opposition to barrier removal Social impacts of the decommissioning of barriers, particularly dams, can give rise to strident opposition. These impacts or concerns include: cost of the removal, historical value of the dam or impoundment, support for existing services that the dam provides such as: irrigation (a major issue and source of conflict in the Murray Darling Basin), aesthetic value, maintenance of water level, limiting tidal influence, loss of tourism opportunities, recreational or commercial fishing opportunities, or concerns over the impact on the lentic ecosystem including threatened species (Sarakinos & Johnson 2003; Lejon et al. 2009).

20 SD 4: Evaluating small barrier removal to improve refuge connectivity

The most common societal concerns surround the loss of a high-value (either actual or perceived) amenity (i.e. the impoundment). Such concerns can be exacerbated if it is perceived that decisions are being made outside the community (e.g. by State government departments or national environmental groups), which can often add to a feeling of community powerlessness and stimulate opposition. Community opposition to even rigorously assessed dam decommissioning projects can also be fuelled by a general lack of understanding of the function and values of rivers and dams (Sarakinos & Johnson 2003).

Such opposition can result in the downgrading of a full barrier removal project to a do- nothing situation or a compromised restorative approach such as fish passage construction. For example, Lejon et al. (2009) reviewed the outcomes of a number of proposed dam decommissioning projects in Sweden. Of the 17 proposed dam removal projects reviewed only six were ultimately removed, nine were not removed (for a variety of reasons outlined in the current review), and two were removed but replaced with new dams. Sarakinos and Johnson (2003) reviewed the social aspects of small dam (<8 m) removal in Wisconsin. The 80 dams that were removed between 1960- 2002 were identified as mostly in need of significant repair and/or no longer provided significant economic benefit. They identified nine major community concerns relating to dam removal, including: fears of reductions in flow / disappearance of the river, increased flooding, that the drained impoundment would not be rehabilitated (naturally or otherwise) and therefore become a mud-basin, loss of land due to government acquisition, decline in property values, increased disease, loss of historical monument, reduced fishing quality, and loss of recreational amenity.

Ownership of dams is often a driver of conflict around decision making. In 2003, Wisconsin had ~38,000 registered dams of which 75% were owned privately or by municipalities (Sarikinos & Johnson 2003). The ownership of formerly submerged land that becomes exposed following dam removal can be complicated when there are numerous landholders around the waterbody. This can lead to not only debate on public versus private ownership, but also on the extent of public access to restored streamlines.

Concerns over lost recreational amenities, particularly angling opportunities, are a major source of opposition to dam removal and can raise economic concerns within local communities if impoundments attract tourists and visitors. For example, in south- western Australia reservoirs are important components of the Smooth Marron (the third largest species of freshwater crayfish in the world) fishery and much debate has occurred on securing and maintaining public access to potable and irrigation water supply dams (de Graaf et al. 2010). Aesthetic concerns surrounding dam removal projects are often considerable. Many of these relate to personal preference and perceptions on what constitutes an aesthetically desirable view (i.e. still water versus rehabilitated streamline). Furthermore, public scepticism of the likely outcome of environmental restoration of former impoundment sites can also drive opposition. Helping the local community to envision how the rehabilitated site will look (through artists impressions, computer generated images, or real world examples of other projects) is extremely important in alleviating these concerns (Sarakinos & Johnson 2003).

It is imperative that appropriate tools in the field of social science are applied to barrier removal projects (such as surveys and social media). Social marketing in relation to dam removal projects was explored by Johnson and Graber (2002) whose aim was to shift human behaviour rather than just raise awareness thereby allowing acceptance of new concepts (such as removing a well-known dam). This can be achieved by gaining

SD 4: Evaluating small barrier removal to improve refuge connectivity 21 an understanding of the beliefs or concerns that the community has towards dam removal by means of inexpensive surveys and then addressing those concerns (Sarakinos & Johnson 2003). Notably, surveys conducted during a Wisconsin dam removal project, revealed a community preference for actions that delivered increased access to the restored river (e.g. walkways and boat access) rather than other restoration actions such as fish stocking (Sarakinos & Johnson 2003). This was taken into account by managers who subsequently constructed a walkway along the river as well as an access point to a popular fishing hole; a good example of how seeking community input can guide effective management decisions in decommissioning dams.

The importance of stakeholder engagement and education in all aspects and at all stages of dam decommissioning to ensure the public are engaged and empowered during the process cannot be overstated. As concluded by Sarakinos and Johnson (2003), this engagement should be driven not only by scientists involved in these projects, but also resource agencies, university extension services, NGOs and dam owners themselves. In doing so this will ensure that a broad range of stakeholders are engaged throughout the decision making process (Sarakinos & Johnson 2003).

2.6 Barrier removal advocacy As discussed, the impacts of dams can be broad ranging and often directly affect livelihoods; particularly in developing countries where large numbers of people can be displaced or have fisheries upon which they rely decimated (WCD 2000). This has led to a wide range of local and international barrier removal activism and advocacy groups being formed. There are numerous examples of advocacy groups that have been formed to campaign for removal of specific structures, as well as larger groups that advocate on a wider scale. Notable organisations include International Rivers (http://www.internationalrivers.org) whose focus is the endorsement of dam removal projects and, more generally, the sustainable use of riverine resources. They have representatives in South America, Africa, and Asia. Founded in 1985, they publish the quarterly World Rivers Review in which dam removal and remediation are strongly promoted. American Rivers (http://www.americanrivers.org) is a conservation organisation founded in 1973 that advocates for river health. They have been involved in numerous dam decommissioning projects throughout the United States and publish up to date information on dams slated for removal. However, Pohl (2003) noted that some users of its database had concerns over the advocacy nature of the organisation and information on decommissioned dams did not always distinguish between those that had been completely removed versus those that were just breached. European Rivers Network (http://www.rivernet.org/ern.htm) is another advocacy group focusing on sustainable management and protection of rivers (including dam removals) in Europe.

In Australia, there are numerous State based and local environmental activism groups that advocate for river restoration. Further, larger environmental advocacy groups (e.g. Australian Conservation Foundation) have campaigned strongly in the past to prevent dams being constructed (e.g. the proposed Franklin Dam project in Tasmania during the 1970s). The Lake Pedder Restoration Committee has been fighting for the restoration of Lake Pedder in Tasmania which was flooded for hydroelectricity in 1972. The Lake Pedder movement has existed in various forms including the precursor to the Australian political Greens Party, the United Tasmania Group (UTG), formed in 1972. They continue to advocate strongly for the restoration of the lake. There is, however, no coordinated national group advocating for dam removals more broadly across Australia.

22 SD 4: Evaluating small barrier removal to improve refuge connectivity

The highly controversial Pak Mun Dam on the Mun River (the largest tributary of the Mekong) in Thailand is a notable case study of a campaign for the decommissioning of a dam. The dam was constructed ~5 km from the confluence with the Mekong in 1994 for the purpose of generating hydroelectricity (Roberts 2001). The project was constructed by the Electricity Generating Authority of Thailand (EGAT) with support from the World Bank, however, there were serious issues with the EIA process; assessments were conducted ~10 years prior to project approval, based on a different dam design at a different location and the EIA was not subsequently revised (Amornsakchai et al. 2000). The fishway constructed over the dam proved ineffective and it is estimated that ~20,000 people have been affected by consequent reductions in fish populations with 6,202 households compensated for loss of fisheries during its construction (Roberts 2001). Since 1999 there has been an ongoing local community and international campaign, led by International Rivers, to have the dam decommissioned (Amornsakchai et al. 2000).

There are also strong advocates for building and maintaining dam infrastructure given the benefits that can be provided (e.g. hydroelectric power generation, potable water, recreational opportunities, visual amenity, etc). An example of one such body is the International Commission on Large Dams (ICOLD) which was created in 1928 with the Australian committee (ANCOLD) subsequently formed in 1937. ICOLD have a committee on dam decommissioning on which ANCOLD have a representative. Although ANCOLD publishes widely on dam construction and operation, no specific guidelines exist on the decommissioning of large dams (http://www.ancold.org.au/publications.asp). However, this has been identified as an area of future work by ANCOLD (M. Miller, ANCOLD, pers. comm.).

2.7 Barrier removals globally 2.7.1 North America The rivers of the United States are the most heavily regulated in the world relative to their extent (Heinz Center 2002). More than 500 dams (mostly older, small structures) were decommissioned in the United States in the latter part of the 20th century and the decommissioning rate exceeded the construction rate in the final two years of that century (WCD 2000; Pohl 2003). In 2002, a census of dam removals that occurred during the 20th century was provided by Pohl (2003). Importantly, this census included small dams (i.e. crest heights as low as 1.8 m) and was based on criteria stipulated for inclusion in the National Inventory of Dams. The number of dams removed was found to have increased during the latter part of the century, but the average height of dams removed in all decades was <10 m, and the ecological consequences of removals were largely unknown.

According to American Rivers the total number of dam removals in the United States stood at 888 in 2010 with 450 of these having occurred between the years 1999 and 2010 (American Rivers 2010). Service (2011) (also quoting American Rivers) revealed a positive trend in dam removals between 1990-2010, mostly of small, obsolete structures in the eastern states (led by Pennsylvania), but also highlighted a recent increase in large dam removals in western states in an attempt to save salmon runs. In some states, criteria for assessing dam removals have been set. For example, Warner and Pejchar (2001) produced criteria for dam removal and applied them in four separate case studies in California. In Canada, Donnelly et al. (2004) observed that the successful removal and river rehabilitation of the Finlayson Dam in 2001-2002 was the first documented case to have occurred in that country that had followed a planned process.

SD 4: Evaluating small barrier removal to improve refuge connectivity 23

The most comprehensive resource for information on dam decommissioning is the Clearinghouse for Dam Removal Information (CDRI), which is a collaborative archival resource (University of California, Riverside and California State University, San Bernardino Libraries) and a special project within the broader Water Resources Collections and Archives (WRCA). The latter collection manages >200 archival collections, has 200,000 technical reports, 1,500 newsletters, >5,000 maps, and >100,000 historic photos and videos. CDRI is an online inter-disciplinary repository that aims to collect documents and publications on dam removal projects so that decision makers can easily access information such as EPA decisions, techniques, previous impact assessments, engineering approaches, ecological information, etc., when planning or assessing dam removal projects. Its major appeal is that it includes published peer-reviewed information (i.e. scientific journals) as well as grey literature sources from the United States and elsewhere (including Australian reports and publications) on the decommissioning of dams.

The CDRI began in 2006 in response to a recommendation made in the Aspen Institute report, Dam Removal: A New Option for a New Century (2002). The search engine is straightforward to use and is rather like a journal database with search terms including: Keywords, Document Title, Author, Dam Name, River Name, State (with an international search option) (see http://library.ucr.edu/wrca/collections/cdri/search.html). Authors or agencies can easily upload new relevant information. As an example, searching for the term ‘fish’ produced 491 unique publications. A search for ‘macroinvertebrates’ produced just 12 resources. This reflects a general bias in impact assessments towards fish; presumably as they have a greater economic and/or social value. Along with usual database search information (i.e. title, links to electronic version of publication, authors, date of publication, abstract, etc.), each item includes information on dam and river name (and often latitude and longitude), year of dam construction and removal. If links to electronic versions of reports are not available, requests for copies can be made directly to CDRI. The CDRI is a leading resource on dam removal projects but focuses largely on the United States. It contains invaluable resources on this subject and we recommend a similar archive be set up for Australian projects, or that effort be made to encourage the uploading of more resources relevant to Australia to the CDRI. 2.7.2 Europe As in North America, dam decommissioning in Europe has also increased over the past two decades (European Rivers Network; Lejon et al. 2009). In 2000, it was estimated that >10,000 dams (>3 m height) would be required to renew their concessions in the ensuing two decades. European Rivers Network stated that >300 mostly small dams in Spain had been removed (see WWF 2009). Lejon et al. (2009) noted that of 17 dams in Sweden considered for dam removal, only six were actually removed (see above, Opposition to barrier removal). 2.7.3 Australia Authorities in NSW have been the most proactive in undertaking barrier mitigation projects including removals of redundant structures (I&I NSW 2009). This process commenced in the 1990s with investigations into the impact of instream barriers on native fish passage by Williams et al. (1996) and followed by Pethebridge et al. (1998), and Thorncraft and Harris (2000). This work laid the foundations for the Initial Weir Review (NSW Fisheries 2002), in which over 1,000 weir structures were assessed and a shortlist of 109 priority structures were identified using a ranking method developed by Pethebridge et al. (1998). This was followed by a more thorough investigation of the shortlisted structures in the NSW Detailed Weir Review (NSW DPI 2006a) and in the

24 SD 4: Evaluating small barrier removal to improve refuge connectivity same year a report was published in which the impact on aquatic habitat of over 1,700 road crossings was also assessed (NSW DPI 2006b).

Initial investigations of the shortlisted instream barriers were aimed at identifying structures having the most severe impact on fish passage and aquatic habitat quality in each Catchment Management Area (NSW DPI 2006a). These barriers were then assessed using a broad range of social, ecological, cultural and logistical criteria to refine the prioritisation of structures requiring mitigation and a number of options for improving fish passage were presented along with estimated costings (NSW DPI 2006a). Detailed planning and undertaking of recommended works followed a process of engagement and consultation with stakeholders during which awareness was raised and support garnered for each individual barrier mitigation project (I&I NSW 2009). In some instances, lack of stakeholder support resulted in the shelving of recommended plans (I&I NSW 2009).

Instream barrier reviews in NSW have culminated in mitigation works being undertaken at 94 different sites including the removal of 14 redundant structures (i.e. weirs and road crossings), with over 1,200 km of stream habitat opened up (I&I NSW 2009). More recently the NSW State Water Corporation have undertaken a project entitled ‘Fish Superhighways’ in which all river infrastructure managed and operated by the department (over 300 dams, weirs, and flow regulators) has been assessed and prioritised for work to improve fish passage in accordance with NSW legislation (State Water Corporation 2011). As of 2011 this project resulted in the opening up of over 1,300 km of stream to migratory fish species through mitigation works on 18 instream barriers, including six weir removals (State Water 2012). A further 1,400 km of stream will be opened up following the planned installation of 21 additional fishway structures and the removal of four more redundant weirs across NSW (State Water 2012).

In Victoria strategies for improving fish passage have largely focused on fishway installations with over 60 structures being constructed in recent decades under the ‘State Fishway Program’ launched in 1999 (DSE 2003; O’Brien et al. 2010). A review of redundant weirs was later released (DSE 2003) following the example of the NSW Initial Weir Review. A total of 232 instream structures were identified as “potentially redundant” throughout Victoria and, of these, 27 of the highest priority barriers (i.e. the top three from each of the 9 CMAs assessed) were recommended for more detailed investigation (DSE 2003). As at 2010, only a quarter of the shortlisted structures had been decommissioned, primarily due to a lack of funding (O’Brien et al. 2010). A number of ‘unauthorised instream structures’ were reported to have been removed in north-eastern Victoria as well, but no official records have been published (O’Brien et al. 2010).

One instream structure that was not part of the original shortlist is the Nicholson Reservoir in East Gippsland. This dam has a crest height of 15m and a capacity of 640 ML and has been unused in recent years due to an adjustment of water licensing allocations in the region, resulting in its identification as a potentially redundant asset (Browning et al. 2007). It has been proposed for a possible decommissioning and, to this end, was the subject of a detailed feasibility study (Browning et al. 2007). Whilst the predicted significant environmental benefits and low social costs lend weight to support decommissioning, insufficient availability of funds and bureaucratic complexities of approving on-ground works were cited as reasons contributing to the decision to hold the project back (Browning et al. 2007). If this project goes ahead, it will be the largest dam removal project ever undertaken in Australia.

The Inland Fisheries service in Tasmania conducted a survey of barriers to fish passage in coastal Tasmanian rivers during 2001-2002 (Nelson 2003a). The survey

SD 4: Evaluating small barrier removal to improve refuge connectivity 25 recommendations resulted in the removal of eight weirs and the installation of fishways or other mitigation measures on five additional barriers (Nelson 2003b; IFS 2004).

In Queensland, a range of fishways have been installed across the state in recent decades (see for examples Marsden et al. 2006; Stockwell et al. 2008) but physical barrier removal, while being the stated preferred barrier mitigation option of governmental agencies, has rarely been undertaken due to difficulties in obtaining community support (DAFF 2012).

The Murray-Darling river system, Australia’s largest, has seen a number of new fishways installed at weirs and barrages along its length as a result of the “Sea to Hume Dam” initiative (MDBC 2008). This collaborative project between governmental and non-governmental agencies in South Australia, Victoria and NSW commenced in 2001 with the objective of constructing 14 new fishway structures and restoring fish passage to over 2,000 km of river for migratory fishes (MDBC 2008). A number of fishways have also been installed on various tidal barrages in South Australia (see McNeil et al. 2010; Zampatti et al. 2012). Elsewhere in Australia instream barrier removals and the installation of fishways has been somewhat limited. For example there have only been seven fishways constructed in Western Australia with the first being installed as recently as 2000 (Morgan & Beatty 2004a, b, 2006; Beatty et al. 2007, 2012; King & Torre 2007; Beatty & Allen 2008) and there are no documented reports of instream barrier removal projects having taken place in South Australia, Western Australia or the Territories.

2.8 Barrier prioritisation 2.8.1 Score and ranking and optimisation modelling The proliferation of instream barriers and their deleterious ecological effects has already been summarised in this report. Interest in mitigating these impacts through fishway construction and barrier removal has gained considerable momentum across the globe at all levels from local communities to inter-governmental collaborations (in Europe for example). Like many environmental issues, the scale of this problem vastly outweighs the resources currently being made available to confront it, therefore prioritisation of which barriers to mitigate or remove is an area of growing importance.

The vast majority of barrier prioritisation methods have used score and rank techniques (Kemp & O’Hanley 2010; and see for example Pethebridge et al. 1998; Taylor & Love 2003; WDFW 2009) where barriers within a given spatial range are scored based on ecological, physical and financial characteristics, the sum of which is used to determine the priority order for mitigation under given budgetary constraints (Kemp and O’Hanley 2010). The main advantage of these methods lies in their simplicity and speed of application (Kemp & O’Hanley 2010). However, speed and simplicity come at the cost of efficiency according to some authors who cite that insufficient consideration is given to multiple barriers within catchments, which can result in minimal habitat gains for migratory species, and this is seen by some as a major shortcoming of these methods (O’Hanley & Tomberlin 2005).

Kemp and O’Hanley (2010) comprehensively reviewed barrier mitigation planning and prioritisation processes and argued strongly for the use of more robust optimisation based models (see Figure 2). The holistic approach of considering cumulative effects of multiple barrier networks on habitat connectivity and fish passability within catchments (rather than considering each barrier independently) and cost-effectiveness were cited as the main advantages of these methods. At a minimum, the computation of optimisation models requires data for length and/or habitat quality of river reaches

26 SD 4: Evaluating small barrier removal to improve refuge connectivity between barriers, location and fish passability of barriers, and removal costs (Kemp & O’Hanley 2010; O’Hanley 2011). However, optimisation models are highly adaptable and able to incorporate any number of additional criteria (e.g. biological, socio- economical, hydrological) and weighting parameters depending on catchment-specific scenarios and restoration objectives (O’Hanley 2011; also see for example Kuby et al. 2005; Zheng et al. 2009).

2.0 km 1.5 km N5 4

N2’ 0.5 km N4 Budget = $10k Original condition 1 km 3 Remove barrier #2 N2 = 2 km 0.5 km N2’ = 3.5 km 0.5 km 2.0 km 1 1.5 km N5 N3 0.5 km N1 4 2 0.5 km N4 2.0 km 3 1 km Budget = $20k N3 1.5 km N2 0.5 km Remove barrier #3, #4 2 0.5 km 0.5 km N2’’ = 4.5 km 1 1 km N2’’ 0.5 km N1 0.5 km

0.5 km 1 0.5 km N1

Figure 2: A conceptual example of optimisation modelling for prioritising barrier removal. The aim is to maximise catchment connectivity distances (i.e. maximise single largest continuous section of suitable unimpeded river). In this example, the dendritic system has five nodes (N) separated by four impassable barriers. Each barrier costs $10k to remove. The most appropriate barrier to remove with a budget of $10k would be to remove barrier #2 which would create 3.5km of unimpeded river. However, if the budget was increased to $20k, barriers #3 and 4 would be best removed that would create 4.5 km of unimpeded river. The process can weight habitat suitability and length, and it is assumed removal completely overcomes the impact. Adapted from O’Hanley and Tomberlin (2005).

SD 4: Evaluating small barrier removal to improve refuge connectivity 27

A major drawback is the complexity of the theory and programming of optimisation models which has undoubtedly restricted the extent of their use in real-world applications. This is likely to remain the case unless a ‘turn-key’ application becomes available that is accessible and easily interpretable to non-specialists, or there is a worldwide upsurge in the training of scientists capable of applying optimisation models. Kemp and O’Hanley (2010) contend that the lack of coordination between authorities from different global regions has resulted in wasted funding through inefficient and sometimes erroneous planning and execution of barrier mitigation prioritisation, as well as duplication of effort. They advocate adopting coordinated and rigorous optimisation based methods (Kemp & O’Hanley 2010).

Webb & Padgham (2009) contest that the use of optimisation modelling is not generally warranted, highlighting that solutions apply only to the specific study in question and are generally not widely applicable. They developed an analytic framework incorporating stream reach quality and connectivity in both an upstream and downstream direction to determine transitional probabilities between reaches (or nodes). They concluded that as the overall effects of multiple remediation projects will usually equal their sum effects in terms of fish populations, there is almost never a need to examine interactive effects on multiple modifications (either to stream reach quality or connectivity) and that accurate scoring of individual modifications should be used to prioritise instream barrier mitigation works (Webb & Padgham 2009). 2.8.2 Prioritisation methods in practice In the western United States a number of similar protocols have been developed by various Fish and Wildlife agencies for the prioritisation of barrier mitigation projects, the focus of which has aimed to maximise the benefits for migratory salmonids which are important both commercially and recreationally, and in some catchments critically endangered (Kemp & O’Hanley 2010). The Washington State Department of Fish and Wildlife (WDFW), for example, have developed a comprehensive operational framework involving field assessment of fish passability of barriers and stream habitat quality in close proximity to the structures (WDFW 2009). All data are maintained in a centralised database and a scoring system is used to prioritise barriers for mitigation based on the potential increase in fish productivity taking into account interspecific competition, habitat gain, conservation status, and financial costs of barrier mitigation (WDFW 2009).

This protocol was applied across the State of Washington in 2007 identifying ~6,500 culvert barriers of which 225 had been removed or mitigated by 2009 (WDFW 2009). Similar protocols have also been developed in the states of Oregon (ODFW 2010) and California, with the latter also maintaining a GIS database of barriers and stream hydrology for use in prioritisation processes (CSCC 2004; PAD 2009). The situtation is not as well organised in the eastern United States where barrier inventories have been compiled in an uncoordinated and unreliable fashion with substantial overlap of effort (Kemp & O’Hanley 2010). The federal US Fish and Wildlife Service has also developed a nationwide barrier inventory along similar lines to the Californian PAD project (GeoFin 2012), however it has been labelled “simplistic” by Kemp and O’Hanley (2010) as many important ecological, socio-economic, and financial factors are not taken into account.

In Europe there has been a tendency for stand-alone assessments, generally of large dams, but the focus has shifted to broader scale evaluations of all forms of barriers since the enactment of the EU Water Framework Directive (Kemp & O’Hanley 2010). Instream barrier inventories have been compiled and mapped in Belgium (see Monden et al. 2000) and the Netherlands (see eds Kroes et al. 2006), with the information being used to develop a priority list of barriers requiring mitigation. German authorities have

28 SD 4: Evaluating small barrier removal to improve refuge connectivity developed a standardised assessment methodology applicable at both the site and catchment scale, and a database for archiving barrier information (Dumont 2005). Any barriers that do not meet EU Water Framework Directive standards of allowing fish to pass in both upstream and downstream directions are prioritised for mitigation works (Dumont 2005). Score and rank procedures have also been used in Austria to compile a national barrier database (Zitek et al. 2008). In England and Wales a standardised barrier assessment methodology is in its infancy but the protocol has been criticised by Kemp & O’Hanley (2010) for not considering multiple barrier networks and relying too heavily on subjective expert opinion rather than objective data. The Scottish Environmental Protection Agency have also recently begun compiling a national barrier database and the development of a broad-scale barrier assessment protocol (Kemp et al. 2008).

In Australia the first reported instance of a barrier mitigation prioritisation process was in NSW (Pethebridge et al. 1998). This method involved an initial desktop assessment of topographic maps to identify potential instream barriers followed by correspondence with relevant State and local authorities to gather more information. Additional data were then collected in the field, photographs taken, and all information archived in a database. The data collected included type of barrier, structural attributes (e.g. height, width, head loss, crest-tailwater differential), weir pool characteristics, stream/catchment characteristics (e.g. distance to proximal barriers, flow regimes, drownout frequency, habitat condition upstream), information on resident fish assemblages (if known), and a range of other information (e.g. age of structure, owner, status of use, presence of existing fishway, availability of data from previous environmental studies in the area, etc.). A total score was calculated for each barrier based on 11 different criteria and used to determine the priority ranking for remediation. This method was later modified to incorporate a more comprehensive suite of data on barriers during the NSW Detailed Weir Review Project (NSW DPI 2006a). The prioritisation was also divided into two phases, with rankings initially determined using stream habitat value (i.e. habitat class, location in catchment, number of downstream barriers, potential habitat to be opened up) and structural criteria (i.e. headloss, drown out frequency, whether barrier was run-of-the-river or undershot by pipework). The second phase involved further refinement of the initial rankings by scoring environmental criteria (i.e. instream habitat condition, riparian condition, presence of threatened species) and modification criteria (i.e. structure redundancy, ease and cost of remediation) (NSW DPI 2006a).

While officially there has been no centralised, coordinated approach to barrier prioritisation driven by the Commonwealth government, agencies in most Australian states have developed their own assessment and prioritisation methods, most of which use a similar score and rank procedure to that of Pethebridge et al. (1998). A review of redundant weirs in Victoria, for example, took place in 2003 as a direct consequence of the “success” of the Initial Weir Review in neighbouring NSW, however the methodology for assessment consisted of a desktop review only and the prioritisation process considered some alternative criteria (e.g. potential threat to human safety, potential adverse impacts of removal such as spread of exotic species) and was more qualitative in nature (DSE 2003).

A Tasmanian weir review conducted by the Inland Fisheries Service took place in 2001-2002 in which there was a stated objective to remove a total of 10 redundant weirs (Nelson 2003a). This outcome was achieved through an initial desktop review of information on weirs and through communication with field officers and local land holders (Nelson 2003a). The review resulted in the identification of 48 catchments of interest, limited to more developed parts of the state, and was followed by collection of additional data in the field, including sampling of local fish assemblages (Nelson

SD 4: Evaluating small barrier removal to improve refuge connectivity 29

2003a). Unfortunately, no detail was reported on the methodology used to determine which sites were prioritised for removal/mitigation (Nelson 2003a). Despite recommendations for further prioritisation of barriers for removal as a means of rehabilitating Tasmanian native fishes (Nelson 2003a), nothing further has taken place and no staff are currently assigned to this area (R. Freeman, Inland Fisheries Service, pers. comm.).

In Queensland there have been numerous barrier mitigation prioritisation projects undertaken in the past decade in various parts of the State (Stewart & Marsden 2006; Carter et al. 2007; Stockwell et al. 2008; Lawrence et al. 2010; Lawson et al. 2010). There has been no standardisation of the methods used, although all have used similar score and rank methodology. Barrier mitigation efforts in Queensland have resulted almost exclusively in the installation of fishways rather than barrier removals. The first barrier review process in the State was undertaken in the Reef Coast region between Cairns and Brisbane (Stewart & Marsden 2006). Barriers were prioritised using a two- step score and rank process, very similar to that used in the NSW Detailed Weir Review (see above). A range of mainly biological/ecological criteria were first assessed to rapidly compile a list of priority barriers for further investigation, whereupon a range of secondary socio-economic criteria were assessed (Stewart & Marsden 2006). Two novel criteria used in this secondary phase of assessment included a barrier’s value as a demonstration site (for raising awareness of the issue), and ease/cost of conducting ongoing fish monitoring at the site. Monitoring (both before and after mitigation works) was identified as a vital step in the overall process as a means of validating the assumptions underpinning the use of barrier mitigation as a fish rehabilitation tool (Stewart & Marsden 2006).

Stockwell et al. (2008) conducted a barrier review in another part of Queensland utilising a three-step process where the broader area of interest was first divided into sub-regions that were prioritised according to an ecological risk assessment model where biodiversity values (i.e. presence of threatened species, habitat quality and quantity) were given the highest weighting. Barriers within the priority sub-regions were then analysed using a decision tree and scored mainly according to ecological criteria to generate a shortlist of high priority barriers for further investigation. The final phase assessed mainly socio-economic criteria including local stakeholder capacity for assistance and support, and a quantified cost benefit analysis (Stockwell et al. 2008). The robustness of this methodology was enhanced considerably by the incorporation of uncertainty in the scoring of some criteria (i.e. those assessed using unvalidated desktop review of existing data) by use of confidence intervals, as well as the undertaking of sensitivity testing whereby the weighting of criteria was varied and reanalysed to determine whether ranking order remained similar under different scoring scenarios (Stockwell et al. 2008).

Carter et al. (2007) applied yet another spin on prioritisation methods in an assessment of the Burdekin Dry Tropics NRM region, using similar criteria but a different scoring system, weighted heavily towards the most desirable scenario (from a prioritisation perspective) for each criteria. A simplified version of this process (using less criteria) was also used in another barrier review project undertaken in the Wet Tropics NRM region (see Lawson et al. 2010).

Western Australia is currently in the process of compiling a statewide GIS database of information pertaining to instream barriers for the purpose of prioritising instream barriers for fish passage enhancement. The methodology for compiling the inventory and prioritising barriers uses similar assessment criteria to the NSW model with more emphasis on fish data (e.g. historical vs current distribution, life history information, potential for spread of exotic species through barrier removal) (Storer & Norton 2011).

30 SD 4: Evaluating small barrier removal to improve refuge connectivity

The process consists of two-stages: a preliminary desktop assessment using the GIS database and existing literature to determine potential barriers in an area or catchment of interest; followed by more detailed assessment, usually involving field validation, a cost benefit analysis, and finally the prioritisation of candidate barriers (Norton & Storer 2011). Despite the inflexible ranking method, the authors advise that it be used only as a guide when making decisions regarding prioritisation of barriers as not all factors are taken into account or the relative importance of some criteria can vary depending on geographic location for instance (Norton & Storer 2011). At this stage the process has not been applied in any region of Western Australia.

2.9 Knowledge gaps As mentioned, there is a general lack of scientific knowledge regarding the effects of dam removal or mitigation (Pohl 2003; Stanley & Doyle 2003). From the current review, many knowledge gaps associated with instream barrier decommissioning and removal that have been previously identified elsewhere (e.g. WRD 2000; Graf 2003; Stanley & Doyle 2003; Lejon et al. 2009; Service 2011) remain largely unaddressed (particularly in southern Australia).

These include: 1) Lack of accurate information on barrier locations and characteristics. Whilst information often exists for large dams, information on the location and impact of small barriers is much less available. 2) Refuge pool locations. There is currently no readily available remote sensing technology to accurately map refuge pools on a fine scale, particularly in forested catchments. 3) There are considerable gaps in the understanding of the migration patterns, habitat requirements and fine scale distribution of many freshwater fish communities across southern Australia. 4) Information associated with social research on impacts of instream barrier removal. Given many small dams are situated on private property, or privately owned on publically owned waterbodies, social impacts may affect specific land holders and/or the broader community. More research is required into the socioeconomic impacts of barrier removal and how best to address them. 5) Accurate economic modelling of instream barrier decommissioning projects. In the past, removal costs have often been overestimated and the costs of retaining barriers underestimated. Restoration and pre-post monitoring costs are often not considered when deciding whether barriers should be removed or retained. 6) A general lack of catchment-scale impact assessments of instream barrier decommissioning projects. As discussed, although relatively complex compared to individual score and ranking systems, broader scale impact models should be considered when planning barrier decommissioning (or remediation) projects where appropriate. Optimisation modelling approaches can be utilised to this end. 7) Predicting the magnitude and scale of ecological impacts can be extremely difficult and careful planning of monitoring programs is required to quantify these changes. The use (both access and contribution towards) of a data repository on dam decommissioning, specifically the CDRI, should be encouraged to ensure researchers and managers are aware of previous impacts so that impacts of future projects are better able to be predicted and effectively monitored. Importantly, little information exists on long-term ecological impacts of instream barrier decommissioning such as overall increases in fish population sizes cf simply determining changes in distributions.

SD 4: Evaluating small barrier removal to improve refuge connectivity 31

The current review has clearly highlighted that careful assessment of the effects of instream barriers (i.e. their economic, social and ecological value and impact) need to be weighed against the impacts (positive and negative) that might occur from their decommissioning. Although no consistent guidelines for decommissioning small or large instream barriers across Australia currently exist, careful and multi-disciplinary planning, monitoring and effective public communication strategies should be implemented to ensure appropriate decisions are made to mitigate opposition to dam removals, and also ensure that potentially negative ecological and social impacts that may be caused by sudden dam removal are minimised. Should a compromise to a full barrier decommissioning be decided upon, it is crucial to involve a broad range of stakeholders in the development of this compromise in order to ensure it is successful and that further opposition and conflict are avoided (Lejon et al. 2009).

32 SD 4: Evaluating small barrier removal to improve refuge connectivity

3. PROCESS FOR SOUTHERN AUSTRALIA

We have developed a prioritisation process for the identification and mitigation of instream barriers in southern Australia (Figure 3). The process consists of five phases within which are a number of steps. The process can be applied to projects of different spatial scales from a single river to multiple catchments. The process is underpinned by broad stakeholder engagement, particularly at a local level.

SD 4: Evaluating small barrier removal to improve refuge connectivity 33

Instream Barrier Identification and Prioritisation Process for southern Australia Contact CMAs and form Steering Permit applications for Committee on-ground works

Develop project scope, scale, budget, responsibilities, extension Stakeholder workshop to Pre-mitigation monitoring present and refine final prioritised list, ascertain Desktop/GIS review (barriers, fish, level of redundancy, and Barrier mitigation works refuges) and ID knowledge gaps determine mitigation and site rehabilitation

Project development Project approach for each barrier Minor/ Major Removal Fishway/ none culvert

Address Produce final prioritised Post- Post- knowledge gaps list of barriers removal mitigation Landholder survey for (fish sediment ecological barriers and refuges distribution, monitoring monitoring Prioritisation movements, Consider optimisation (seasonally barrier and based modelling of for at least Pre-select barriers of refuge multiple barrier 2 years) interest for further information) mitigations

investigation Barrier & monitoringworks mitigation

Assess barriers according Aerial survey of barriers for initial to score and rank validation and assessment methods Ongoing project extension and promotion including use Compile data (complementary of barrier mitigation works as a demonstration site (if

ground truthing where necessary) for Project

potential fish barriers extension appropriate) Barrier ID, validation & assessment & validation ID, Barrier

Figure 3: Flow chart of the process for identification and mitigation of instream barriers in southern Australia. N.B. steps in bold typeface were completed as part of the three pilot studies as detailed in the current report.

34 SD 4: Evaluating small barrier removal to improve refuge connectivity

3.1 Project planning and development The importance of engaging local stakeholders has been demonstrated to be crucial in delivering successful outcomes in barrier removal/mitigation projects. Therefore, we recommend contacting local Catchment Management Authorities (CMAs) and Land/Rivercare groups early in the process to garner local stakeholder support. We highly recommend that such groups also have membership on a Steering Committee that should be formed for each project. Other Committee members could inlcude appropriate State and Local Government Department representatives, traditional owner groups, aquatic biologists/ecologists, irrigation/landholder groups, and an environmental engineer with relevant river infrastructure experience. The Committee should be resourced adequately to allow for regular meetings to take place throughout the life of the project; including the post-removal monitoring phase.

The initial phase of the process involves the arbitrary designation of the project scale. The choice of which region(s), or if coverage is to be more limited, which catchment(s) to run the instream barrier assessment and prioritision process will largely depend on the amount of resources available as well as the conservation issues and objectives of the area in question. Catchments that house threatened or endangered taxa, particularly any which undertake migrationary movements will be of greater interest and will likely be assessed earlier than catchments that do not. However, the question of which catchment(s) will be the focus of the project may depend on numerous other political, social and ecological factors.

Once the scale has been decided, a thorough desktop review of available literature and databases should follow in which all relevant data and information on the biological, hydrological, and ecological characteristics of the catchment(s) of interest are compiled. In addition to any available GIS databases that may exist on instream barriers, an analysis of available aerial imagery should be undertaken if practical (e.g. in predominantly cleared catchments where barriers are readily visible in aerial imagery) to help determine the locations and extent of barriers in the catchment(s) of interest. Spatial and ecological information on the aquatic fauna within the study area should be compiled and any migratory patterns (if known) matched against the hydrology of the system in order to guide several aspects of the score and ranking system. An example of how this could be presented is provided in Figure 4. This example provides the expected migratory patterns of native fish species and an historic flow hydrograph of the Goodga River. It demonstrates that adults of all four native fish species in the Goodga are dependent on autumn to spring flow peaks for upstream migration. The management or removal of instream barriers that impede important peak flows should therefore be given higher priority. However, it also demonstrates that the Common Jollytail (Galaxias maculatus) migrates upstream throughout the year given suitable stream flows. As such, fish passage would be required under all flow conditions for this species.

SD 4: Evaluating small barrier removal to improve refuge connectivity 35

Figure 4: Seasonal migratory movements (top) of fishes of the Goodga River, south-western Australia: green = adults, blue = larvae, red = juveniles. Arrows indicate upstream and downstream movements. Also presented (bottom) is the mean monthly historical discharge within the system (discharge data from Department of Water, Government of Western Australia).

We advise that contact be made with land holders who have riparian frontage in the catchment(s) of interest, and this should be undertaken with the assistance of the relevant CMA in order to determine the most appropriate approach. We recommend this be done initially by either mailing out information about the barrier identification/prioritisation process and including a convenient means for providing data on instream barriers and refuge pools via a simple questionnaire (see Appendix 1), or advertising and conducting a landholder information workshop. Any additional unpublished information or data on instream barriers and refuge pools provided by land holders or other stakeholders is then collated with data from the desktop review.

An assessment should then be made of knowledge gaps that may limit or prevent the project moving forward. These are likely to be related to: 1) lack of information on barrier locations (e.g. lack of GIS barrier database for that region), and 2) lack of information on the aquatic fauna communities within the study area. The Steering Committee should seek relevant external advice in assessing the amount and quality of information that exists to ensure it is adequate to successfully conduct the project. Any significant knowledge gaps should be addressed before continuing in order to prevent an incomplete or inaccurate process being pursued. This may involve conducting seasonal aquatic fauna surveys, and/or developing a GIS database of instream barriers.

36 SD 4: Evaluating small barrier removal to improve refuge connectivity

3.2 Barrier identification After the Committee is satisfied that adequate information exists to proceed with the process and following the desktop review of barriers and ecological information, aerial and/or ground-truthing of barriers and refuge pools should be undertaken to validate the data gathered thus far. Any previously undocumented barriers (artificial and natural) are also able to be identified and mapped during this phase. Either of the two truthing methods (i.e. aerial or ground) may be more economically viable or suitable depending on a range of factors including catchment size, length, tributary complexity, extent of road access, flight time to site from base of operations, etc. A combination of both methods is likely to prove most effective, i.e. rapid aerial assessment followed by more detailed targeted ground-truthing. Whilst seemingly expensive, the aerial survey method via helicopter should be considered as an initial phase of the truthing of instream barriers. In our experience running the pilot studies (see 4. Pilot studies below), we found this to be an extremely effective and efficient method for validating GIS derived information on barriers and refuges, particularly in remote or inaccessible (to vehicles) river reaches. During the aerial surveys an initial screening assessment of all shortlisted barriers can be made based on whether the structure is: a) on the actual streamline, and; b) has a head loss of >0.2m (i.e. the approximate limit of leaping ability for small bodied native fishes of southern Australia). If the barrier does not meet both of these criteria, it may be excluded from the assessment process.

It is important that aerial and/or ground truthing surveys be undertaken during periods of high flow to assess instream barriers as well as during base flow to assess refuge habitats in intermittent southern Australian rivers as these are both important factors that drive freshwater fish distributions and explain community structures. Once all data are collated, owners or managers of instream barriers should be contacted to attend a workshop where levels of redundancy for instream barriers can be ascertained and a final short list of potentially redundant barriers produced for further detailed priority assessment.

3.3 Detailed barrier assessment and prioritisation The above process will result in a list of potential barriers to fish movements within a catchment that is then subjected to further detailed assessment and prioritisation of removal or mitigation works. Steps are included in the process for thorough consultation with barrier owners/managers to workshop prioritisation strategies. However, all barriers, regardless of their potential redundancy, will be subjected to an initial assessment using a score and rank methodology adapted from those used by the Department of Water, Government of Western Australia (see Norton & Storer 2011) and Ryan et al. (2010) and therefore the current process can also be used to determine barriers that should be mitigated through retrofitting of fish passage infrastructure.

The proposed score and rank method uses a number of criteria including: biodiversity values at the barrier site; location within the catchment and relative to other barriers both upstream and downstream; drown out frequency; habitat quality and availability upstream of the barrier; restriction of access to refuge habitat, and; some potential negative impacts of removing or mitigating barriers such as facilitating the spread of exotic species and any local stakeholder opposition to proposed barrier removal projects (see Table 1). Criteria are weighted differently according to their varying levels of importance in the prioritisation process. For instance, the criteria for an increase in upstream refuge access has the highest weighting reflecting the focus of the current project. Other highly weighted criteria include species diversity and presence of threatened species downstream of the barrier site, and potential for increase in upstream access if the barrier were to be mitigated or removed (Table 1).

SD 4: Evaluating small barrier removal to improve refuge connectivity 37

A number of supplementary scoring criteria are also proposed for barriers that provide a permanent refuge habitat. In these instances, the supplementary score applies a number of negatively weighted criteria aimed at assessing the importance of the refuge habitat created by the barrier structure relative to other refuge habitats in the catchment (Table 1). It is included to provide an evaluation of a scenario where a barrier providing refuge habitat is removed rather than mitigated by means of a fishway installation.

The results of the score and ranking phase produces a prioritised list of barriers for mitigation or removal which is then presented to stakeholders at a workshop for feedback and further refinement. Note that synergistic impacts of removing multiple barriers within catchments can be considered, but the optimisation based methods used to assess such impacts will usually result in the same or similar outcomes in terms of barrier prioritisation ranking as if all barriers were assesed individually (refer to section 2.8.1 Score and ranking and optimisation modelling).

38 SD 4: Evaluating small barrier removal to improve refuge connectivity

Table 1: Score and ranking system developed to prioritise instream barriers for mitigation or removal in southern Australia. POSITIVE CRITERIA Weighting Rating Score 3 2 1 0 -1 (Weighting x Rating) 1: Fish present (or expected) within stream section Threatened fish 5 migratory migration likely local not present n/a 5 x rating movements only Native fish diversity (% catchment/regional 5 >80% 50-79% 25-49% <25% n/a 5 x rating species represented) sub-total (max. 30) 2: Barrier location and relative size Increase in upstream access if barrier mitigated (% 5 >15% 9-15% 2-8% <2% n/a 5 x rating of total stream length) Percentage of total stream length upstream of 3 >80% 50-79% 25-49% 5-24% <5% 3 x rating barrier Percentage of total fish barriers in catchment 3 <10% 25-49% 50-79% >80% n/a 3 x rating located downstream Fish access through drownout (% flow period) 3 <25% 25-49% 50-75% >75% n/a 3 x rating

sub-total (max. 42) 3: Habitat quality and availability upstream of barrier Quality of riparian buffer upstream of barrier 2 Excellent Good (mostly Moderate Poor (mostly n/a 2 x rating (near pristine) undisturbed)) (patchy/semi- degraded) degraded) Water quality upstream of barrier 2 Excellent Good Moderate Poor n/a 2 x rating Increase in upstream refuge access if barrier 8 >30% 15-29% 5-14% <5% n/a 8 x rating mitigated (% of total refuge habitat in catchment) Percentage of year flowing at barrier 3 100% 40-99% 5-39% <5% n/a 3 x rating

sub-total (max. 45) TOTAL (+) (max. 117)

SD 4: Evaluating small barrier removal to improve refuge connectivity 39

Negative criteria Weighting Rating 4: Negative impacts of barrier mitigation 3 2 1 0 Barrier preventing spread of Alien fish -5 confirmed likely possible unlikely - 5 x rating Lack of stakeholder support -1 confirmed likely possible unlikely - 1 x rating

TOTAL (-) (max. -18)

GRAND TOTAL = TOTAL (+) + TOTAL (-)

SUPPLEMENTARY SCORE: Additional negative impacts subtracted from Grand Total under a barrier removal scenario (*apply these scores only when a barrier pool provides a refuge habitat in baseflow)

Biodiversity value of barrier pool as a baseflow refuge -5 (threatened (50-79% of all (25-49% of all (<25% of all - 5 x rating habitat species present fish) fish) fish) and/or >80% of all fish) Species uniqueness in barrier pool (compared to other -10 2 or more 2 or more 1 species (not zero -10 x rating known refuges) species species threatened) uniqueness including at (none least 1 threatened) threatened Relative proportion of total baseflow refuge habitat created -5 >30% 15-29% 5-14% <5% - 5 x rating by barrier Proximity to nearest refuge -2 >5 km 2-5 km 1-2 km <1 km - 2 x rating Major downstream sedimentation risk -4 confirmed likely possible unlikely - 4 x rating SUPP. TOTAL (-) (max. -81) SUPP. GRAND TOTAL = GRAND TOTAL - SUPP. TOTAL (-)

40 SD 4: Evaluating small barrier removal to improve refuge connectivity

3.4 Barrier mitigation on-ground works and monitoring Once all parties are in agreement with the proposed barrier removal/mitigation plans the on-ground works phase of the process begins. Pre-removal ecological monitoring should take place while awaiting the approval of permits for undertaking the removals in order to ensure the impact of the barrier removal is quantified. It is crucial that the monitoring program is appropriately resourced and therefore is factored into the initial budget development phase of the project. The scope and scale of the ecological monitoring program will be dependent on the barrier, the aquatic ecosystem, and the level of existing information on that ecosystem. Sediment quality and load in the reservoir (should one exist) upstream of the barrier should be assessed to determine the potential for downstream sedimentation of habitat or release of toxic material. A riparian vegetation survey at the site would be required should disturbance projected during the site works phase be predicted.

It is important that hydrological and ecological monitoring be undertaken over an appropriate timeframe so that any impacts of the removal can be objectively documented. Broadly speaking, pre- and post-removal assessment of the downstream and upstream aquatic fauna community, water quality, and sediment loads should take place seasonally for a period of at least a year prior and two years post-removal using established survey techniques such as those used in the development and monitoring of biotic indices of ecosystem health. This will allow quantification of the outcomes of the project in terms of enhancing the distribution, population connectivity, recruitment success and abundance of resident species. Ideally, post-removal monitoring should be undertaken over a much longer timespan than two years, but it is likely that typical project budgetary constraints will prevent this (see discussion of this topic above in section 2.4.3 Ecological community shifts). Given the short-term impacts of sedimentation following instream barrier removals, it is recommended that sediment and ecological monitoring also be undertaken intensively at downstream sites following barrier removal. The monitoring program should be developed with input from suitably experienced scientists and be conducted by appropriately qualified or supervised personnel.

Barrier removal/mitigation should be undertaken by suitably qualified contractors under the supervision of structural and/or environmental engineers to ensure that works are conducted safely and efficiently. Appropriate consideration should also be given to minimize environmental disturbance in general, and particularly, river bank and instream habitat disturbance.

Depending on the scale of the project and the location of each barrier that is removed/mitigated, some degree of site rehabilitation will usually be required particularly on any newly exposed banks where a reservoir previously existed. The nature and scale of the rehabilitation should be decided by the Steering Committee with particular input from the CMA representatives and the environmental engineer. This may include both physical and ecological rehabilitation of the streamline. Minimising the impact of sediment transport during removal projects can be achieved by appropriate timing and pacing of removal, along with other mechanical removal or stabilisation approaches such as grading or armouring river banks with geotextiles or riprap.

SD 4: Evaluating small barrier removal to improve refuge connectivity 41

3.5 Project extension As with all natural resource management projects, every effort should be made to ensure that the project is promoted widely to maximise the public education benefit as increasing awareness of the factors impacting aquatic ecosystems is crucial in their long-term conservation. The Steering Committee should formulate a project extension plan that includes not only the stakeholder workshops and information sessions discussed above, but extension to a wider audience through media releases and updates on appropriate websites.

Demonstration sites for restoration projects are widely used to raise awareness and support for on-ground activities and we recommend that barrier removal sites be considered for use as demonstration sites if deemed appropriate by the Steering Committee. This may involve the erecting of information signs or plaques illustrating conditions at the site before and after works as well as inviting local stakeholders, media, and interested members of the local community to attend and observe the removal works should it be feasible.

42 SD 4: Evaluating small barrier removal to improve refuge connectivity

4. PILOT STUDIES

Three pilot studies were conducted to trial the process of identifying and prioritising instream barriers as outlined above. Three catchments in south-western Australia for which considerable pre-existing spatial data were available on fish were selected representing different scales of human disturbance: i.e. extensively cleared in the lower and middle reaches with relatively undisturbed headwaters (Ludlow River); extensively cleared in the middle reaches and headwaters with relatively pristine lower reaches (Goodga River), and; relatively pristine throughout (Mitchell River).

4.1 General methods A review of the geodatabase of potential instream fish barriers maintained by the Department of Water, Government of Western Australia was initially conducted for each catchment. Additionally, flyers and questionnaires were prepared and mailed out to riparian land holders (whose contact details were provided by local CMAs) in each of the catchments to raise awareness of the issue of instream barriers and gather basic data on barriers and refuge pools (see Appendix 1). Information provided by land holders can be an effective and economical (cf helicopter survey) means of helping to address what are, more often than not, crucial knowledge gaps in southern Australian rivers, i.e. the location and extent of fish refuge habitat and instream barriers. This method cannot be wholly relied upon however, as there are a number of shortcomings of this form of data acquisition including the lack of a guaranteed response from land holders to questionnaires resulting in spatially incomplete data sets, and the fact that it is only applicable in rurally developed sections of catchments. Substantial portions of many rivers in southern Australia flow through land that is managed by governmental agencies or private companies including Crown Land, conservation reserves, mining leases, timber plantations, and other uninhabited areas.

Following the review of information from the database and questionnaire responses, a number of instream barriers of interest were pre-selected for further investigation. These included those lying on or close to the main river channel in the target catchments and the majority of structures identified in the DoW barrier database as having a high priority impact level. These impact levels are estimated via an assessment of “…barrier type, river size, river flow regime and any validated structural data available” operating under the assumption that impacts will be highest in larger, permanent river systems (Norton & Storer 2011). Additionally, at least one barrier was pre-selected on each major tributary, to expand the spatial coverage within the study catchments during the aerial surveys. The number of barriers pre-selected for each catchment was limited to 50 or less in order to reduce the amount of time and expenditure required for the aerial surveys. Aerial barrier surveys took place during October 2012 (i.e. mid-spring), which allowed the assessment of head loss at barrier sites during a ‘shoulder’ period of stream flow when local native fishes are known to be actively moving in catchments. It is advisable to time aerial barrier surveys to coincide with a ‘shoulder’ period rather than the ‘peak’ flow period to avoid difficulties associated with assessing barriers that are drowned out.

A range of data was collected for instream barriers during aerial surveys including GPS co-ordinates, barrier type and function, construction material, head loss, and the nature of any flow at the barrier site. Stream width and accessiblity of the barrier site, as well as surrounding land use(s), and whether the barrier appeared to be redundant were also recorded. An observation confidence score (between 1-10, with 10 being most confident) was assigned for each barrier as visibility from the air varied within catchments, and was especially reduced at sites with dense riparian vegetation cover.

SD 4: Evaluating small barrier removal to improve refuge connectivity 43

These confidence scores had no practical use in the case of these pilot studies, but in a genuine implementation of the prioritisation process would be useful for identifying barriers requiring supplementary ground-truthing. Photographs were also taken to compile a visual record of all instream barriers assessed. Data sheets were prepared prior to aerial surveys and used to enter data manually during the first survey flight, but were deemed too cumbersome for practical use. Subsequently, a digital voice recorder was used and found to be far more efficient as the observer was able to maintain visual contact with the barrier below while recording all necessary data.

Following the survey, data were collated into a spreadsheet and a prioritisation score was calculated for each barrier. Some of the data required to calculate an accurate prioritisation score were incomplete or unavailable (e.g. fish distribution, water quality, location and extent of refuge habitat, stream flow, barrier drown-out frequency, stakeholder support, sedimentation) and were therefore estimated for the purpose of applying the score and rank methods for these pilot studies.

It should be noted that we omitted an important step in the planning and execution of the aerial surveys, in failing to notify riparian land holders of the date, time and proposed route of the helicopter flights. Following our surveys, we received a complaint from a land holder that the low flying helicopter (which flies at around 50 – 60 m above ground level) startled livestock and could have resulted in damage to property, fences, and/or the animals themselves. We strongly advise that permission to fly at low altitude be sought well in advance from land holders prior to conducting any aerial survey work over rural land, firstly as a courtesy, and secondly to avoid stimulating ill will and opposition towards the barrier prioritisation process among the local community.

4.2 Pilot study 1 – Ludlow River Seven of the 23 land holders contacted provided a response to the questionnaire on instream barriers and permanent pools. One instream barrier (a potentially redundant ford/crossing) and four permanent river pools were identified.

The Ludlow River catchment was estimated to comprise around 107 km of stream length (including major tributaries). A desktop review of the instream barrier database (DoW) revealed a total of 288 potential fish barriers throughout the catchment of which 44 were pre-selected for investigation by helicopter (Table 2, Figure 5). Of the pre- selected barriers, 59.1% were either unable to located from the helicopter (n = 5) or deemed to not represent a fish barrier (n = 21) (i.e. located offstream and/or with a head loss of less than 0.2 m). During the flight, 20 additional potential fish barriers (including one natural barrier) that were not logged on the DoW database were recorded and assessed in the Ludlow catchment. Of all the barriers surveyed (i.e. pre- selected and newly found) only around 8% matched the criteria for consideration of removal, i.e. they were identified both as a genuine fish barrier (i.e. onstream and with a head loss greater than 0.2 m) and potentially redundant (Table 2). However, to ascertain whether these barriers are genuine candidates for removal, the data gathered during the aerial survey would need to be ground truthed and the owner/manager of the barriers would also need to be consulted.

The total survey flight time was 4:00 (hours:minutes) which included 2:35 of survey time and a 1:25 return commute from the helicopter’s base of operations (Table 3). The total stream distance surveyed from the air was 72.7 km (approximately 68% of total stream length in the catchment) giving a cost per surveyed stream kilometre of AUD$49.54 and a cost per barrier of AUD$58.06. These costs were based on a rate of

44 SD 4: Evaluating small barrier removal to improve refuge connectivity

AUD$900 per hour of helicopter running time (Table 3). It should be noted that a similar inspection via road would have required substantially more time (at an estimated ~10 barriers per day) and would not have been possible in some of the more remote locations within the catchment.

Prioritisation scores in the Ludlow averaged 23.2 (± 1.67), which was 19.8% of the maximum possible obtainable score (see Appendix 2). All but one of the five highest priority barriers were located on the main channel. The highest ranking was given to a drop structure, built for the purpose of controlling erosion in a straight section of river channel, and currently also used as a stream gauging station (K. Seewraj, pers. comm. see Figure 5). This structure was the furthest downstream of all barriers in the catchment; located approximately 4 km upriver from the point where the Ludlow drains into Wonnerup Estuary. Furthermore, it was the only barrier in the catchment that we believe provides a permanent refuge pool habitat. Applying the supplementary scoring criteria to this barrier in a hypothetical scenario where the barrier is removed rather than mitigated with a fishway, reduced the score by 10, changing its ranking from first to fifth on the priority list. The score was downgraded due to the barrier pool having biodiversity value, in that it is located in a section of the river that houses the full complement of resident fish species in the Ludlow catchment.

SD 4: Evaluating small barrier removal to improve refuge connectivity 45

Table 2: Summary of key results of the barrier identification phase of the process conducted within three pilot study catchments. N.B. DoW GIS = GIS database (Department of Water, Government of Western Australia) of potential fish barriers in Western Australia. River Total % pre- % of pre- % of pre-selected Number of additional % of ALL surveyed number selected for selected DoW DoW GIS features potential fish barriers barriers (pre- of aerial GIS features found not to be identified during selected + new) potential assessment unable to be potential fish helicopter survey identified as fish fish located from barriers barriers AND barriers helicopter Artificial Natural potentially (DoW redundant GIS) Ludlow 288 15.3% (44) 11.4% (5 of 44) 59.1% (26 of 44) 19 1 7.8% (5 of 64)

Mitchell 102 29.4% (30) 16.7% (5 of 30) 56.7% (17 of 30) 2 8 0% (0 of 40)

Goodga 152 21.7% (33) 6.1% (2 of 33) 69.7% (23 of 33) 21 0 0% (0 of 54)

Table 3: Summary of key efficiency results of the barrier identification phase of the process conducted within the three pilot study catchments. River Survey Commute Total flight Survey cost Stream distance Cost per unit Cost per Flight time time (@ $900/hr) surveyed (km) stream distance barrier ($) time (hr:min) (hr:min) ($/km) (hr:min)

Ludlow 2:35 1:25 4:00 $3,600 72.7 $49.54/km $58.06 Mitchell 1:47 1:50 3:37 $3,255 41.3 $78.82/km $81.38 Goodga 1:27 1:51 3:18 $2,970 23.7 $125.22/km $55.00 Totals $9,825 137.7

46 SD 4: Evaluating small barrier removal to improve refuge connectivity

Figure 5: Map summarising key results of the Ludlow River barrier prioritisation process. Sites pre-selected from the GIS database are numbered (1-44), red circles indicate sites deemed to be potential fish barriers and blue circles sites deemed not to be potential fish barriers (based on the aerial survey, see text for details). Fish diversity data at six sites are presented as pie graphs, conceptual representation of refuge pools are shown as blue rectangles, and aerial photos of three example fish barriers are provided as are their respective priority scores.

SD 4: Evaluating small barrier removal to improve refuge connectivity 47

4.3 Pilot study 2 – Mitchell River The Mitchell River catchment is mostly undisturbed, flowing predominantly through National Park and State Forest. As such, contact details for only five riparian land holders were obtained from the local CMA, two of which were private companies with timber plantation holdings. Only one response was received to our questionnaire with no instream barriers or permanent pools reported.

The catchment, including major tributaries, was estimated to comprise 58 km of stream length in total. A review of the DoW database revealed a total of 102 potential fish barriers in the catchment of which 30 were pre-selected for aerial survey (Table 2, Figure 6). Of the pre-selected barriers, 60% were either unable to be located from the helicopter (n = 5) or assessed as not representing a genuine fish barrier (n = 13) (i.e. located offstream and/or with a head loss of less than 0.2 m). Two additional human- made potential fish barriers (not logged on the DoW database) were identified and assessed in the Mitchell catchment. A number of natural barriers were also identified during the survey, eight of which were identified as being significant fish barriers. In some instances, sections of stream (up to 1 km in length) containing multiple, closely- spaced natural barriers (e.g. a series of small waterfalls and cascades) were recorded as a single barrier to save time during the flight. None of the artificial barriers surveyed in the Mitchell matched the criteria for removal.

The total survey flight time was 3:37 (hours:minutes), including 1:47 of survey time and 1:50 transit. The total stream distance surveyed was 41.3 km (approximately 71% of total stream length in the catchment) giving a cost per surveyed stream kilometre of AUD$78.82 and a cost per barrier of AUD$81.38 (Table 3). It should be noted that most of this catchment is inaccessible by vehicle, therefore a comprehensive land- based barrier survey would not have been feasible.

The mean prioritisation score for barriers in the Mitchell was 44.0 (± 7.55), which was 37.6% of the maximum possible score (Appendix 2). The top five ranked barriers were all situated on the main channel of the Mitchell, with the highest ranking (score of 101) given to a vehicle track crossing (i.e. ford) located just upstream of the confluence with the Hay River (see Figure 6). The second highest ranked barrier was a v-notch weir located in the middle reaches of the catchment, used by DoW as a stream gauging station (Figure 6, also illustrated on the frontispiece of the report). This structure provides a permanent refuge pool and when supplementary scoring criteria were applied (in the hypothetical scenario where the barrier is removed rather than mitigated with a fishway), its score dropped from 83 to 20.

This substantial reduction reflects the significant biodiversity value of this weir-pool refuge habitat estimated to constitute around 10% of the total refuge habitat in the entire Mitchell catchment and housing two threatened fish species. One other barrier in the Mitchell catchment provided a permanent refuge pool: a pipe culvert located immediately downstream of an artificially constructed water point used for fire management by the Department of Environment and Conservation, Government of Western Australia. Supplementary scoring reduced the score for this barrier from 52 (ranked fourth highest) to 23 (ranked third lowest), also reflecting its importance as a refuge habitat for one threatened species and providing an estimated 10% of the total refuge habitat in the catchment.

48 SD 4: Evaluating small barrier removal to improve refuge connectivity

Figure 6: Map summarising the key results of the Mitchell River barrier prioritisation process. Sites pre-selected from the GIS database are numbered (1-30) with red circles indicating sites deemed to be potential fish barriers and blue circles sites deemed not to be potential fish barriers (based on the aerial survey, see text for details). Fish diversity data at 11 sites are presented as pie graphs, conceptual representation of refuge pools are shown as blue rectangles, and aerial photos of two example fish barriers are provided as are their respective priority scores.

SD 4: Evaluating small barrier removal to improve refuge connectivity 49

4.4 Pilot study 3 – Goodga River Fourteen riparian land holders were contacted in the Goodga River catchment of which six provided a questionnaire response. Information was provided for only one instream barrier: a pipe culvert on a property boundary track, identified by the land holder as not redundant. No permanent pools were identified by land holders; however some basic information on periodicity of flow was volunteered by two separate land holders.

This was the smallest of the three catchments surveyed comprising a total stream length of 33.4 km and was found to have 152 barriers in total on the DoW fish barrier database, making it the most highly modified of the three catchments assessed with 4.55 barriers per stream km cf 1.77 for the Mitchell and 2.69 for the Ludlow. Thirty- three barriers were pre-selected for aerial survey of which 67.7% were either unable to located from the helicopter (n = 2) or deemed to not represent a genuine fish barrier (n = 20) (Table 2). In addition, 21 artificial fish barriers were identified and assessed during the Goodga aerial survey, although two of these were included in the DoW database but had not been arbitrarily pre-selected for aerial assessment. None of the barriers surveyed matched the criteria for consideration of removal (i.e. onstream, with a head loss in excess of 0.2 m and potentially redundant. No natural barriers were identified during the survey.

The total survey flight time was 3:18 (hours:minutes), including 1:27 of survey time and 1:51 of transit (Table 3). The total stream distance surveyed was 23.7 km (approximately 71% of total stream length in the catchment) giving a cost per surveyed stream kilometre of AUD$125.22 and a cost per barrier of AUD$55.00 (Table 3).

The mean prioritisation score for barriers in the Goodga was 24.1 (± 3.65), or 20.6% of the theoretical maximum score of 117 (Appendix 2). The top ranked barrier with a score of 100 was the v-notch weir gauging station located about two kilometres upstream of the point where the Goodga drains into Moates Lake. This weir has an operational vertical-slot fishway (Figure 7) but is still a barrier for a number of fish species unable to utilise the fishway infrastructure. The weir forms a refuge pool habitat and when supplementary scoring criteria were applied, despite a drop in score from 100 to 75, it remained the top ranked barrier in the catchment.

50 SD 4: Evaluating small barrier removal to improve refuge connectivity

Figure 7: Map summarising the key results of the Goodga River barrier prioritisation process. Sites pre-selected from the GIS database are numbered (1-33) with red circles indicating sites deemed to be potential fish barriers and blue circles sites deemed not to be potential fish barriers (based on the aerial survey, see text for details). Fish diversity data at nine sites are presented as pie graphs, conceptual representation of refuge habitat shown as blue rectangles, and an aerial photo of an example fish barrier is provided as is its priority score.

SD 4: Evaluating small barrier removal to improve refuge connectivity 51

4.5 General discussion of pilot study results One of the primary objectives of undertaking the pilot study component of this project was to test the concept that aerial survey via helicopter is an effective method for identifying and assessing instream barriers in southern Australia. Although aerial survey costs may appear high (i.e. ~AUD$900 per hour), they compare very favourably to the alternative of vehicle-based ground truthing due to the time efficiency involved. To illustrate this point, during the current pilot studies a total flight time of approximately 11 hours (undertaken over two days) yielded data for a total of 158 different instream barriers. We estimate that to survey the same number of barriers by vehicle would have required at least 16 field days. This is assuming that 10 sites could be surveyed per day but this may be an underestimate of the time involved depending on factors including distances between barriers, terrain, and river access issues on private land, to name a few. Moreover, approximately 30% of the barriers would either not have been able to be surveyed or would not have even been detected if ground-based methods were used, as large sections of the catchments were inaccessible to vehicles. This would have included artificial barriers that were not in the GIS database, as well as all natural barriers.

Some barriers proved difficult to assess from the helicopter mostly because of obscured line of sight visibility due to dense riparian vegetation cover. Data for characteristics such as head loss, barrier type, and construction material could not be recorded with absolute certainty in such cases. Therefore an observation confidence score was recorded for each barrier assessed. In a real world application of this methodology, observation confidence scores below a certain threshold would necessitate ground-truthing of data for the barriers in question assuming they were accessible by vehicle. Potential barrier redundancy was another characteristic that could only be estimated by aerial survey as this was based on tenuous evidence at best; including the visual state of repair of the barrier and/or evidence of recent use (e.g. well-worn vehicle tracks on a ford would indicate that the barrier was not redundant). Clearly, consultation with barrier owners/managers is required in order to ascertain the true redundancy status of barriers.

Basic information about the project and a questionnaire were mailed out to 42 land holders across the three catchments of which 13 responded (i.e. 30.95% response rate). Most reported that no instream barriers or permanent refuge pools existed on their properties, but data were volunteered for two instream barriers (including one potentially redundant ford/river crossing) and four permanent pools, which aided the calculation of the prioritisation scores. The land holder response rate was higher than we anticipated, but the fact that over two-thirds did not respond provides solid evidence that such methods cannot be solely relied upon in the pursuit of a comprehensive data set on barriers and refuges. Nonetheless, we believe this is a worthwhile activity as it is an effective means by which local land holders can be informed of and engaged in the barrier prioritisation process.

The prioritisation scores calculated for barriers across the three catchments yielded some results worthy of discussion. Mean barrier scores in the Ludlow (23.2 ± 1.67) and Goodga (24.1 ± 3.65) catchments were very similar. These scores were around 20% of the maximum possible score (i.e. 117) reflecting the relatively high degree of modification and disturbance in both catchments. The Goodga is predominantly developed for agriculture and timber plantations whereas the Ludlow is developed for agriculture and sand mining. The mean barrier score in the Mitchell (44.0 ± 7.55) was almost double those of the other two catchments, reflecting its largely undisturbed and relatively pristine condition, as well as the presence of three listed threatened fish species (i.e. Western Mud Minnow, Balston’s Pygmy Perch, Little Pygmy Perch).

52 SD 4: Evaluating small barrier removal to improve refuge connectivity

The low mean score in the Goodga catchment was surprising at first given the presence of three threatened fish species in the catchment (i.e. Western Mud Minnow, Balston’s Pygmy Perch, and Trout Minnow). However, a closer inspection of the distribution of refuge pools and instream barriers explained this result; most of the threatened species occur only in the lower reaches of the catchment where there are few barriers and a high proportion of the total refuge habitat in the catchment (including Moates Lake).

The highest ranking barriers in all three rivers tended to be located on the main river channel (as opposed to tributaries) and in the lower reaches of their respective catchments. These sites typically house the highest proportions of native fish species and restrict access to sizeable sections of habitat further upstream. Four artificial barriers were identified as providing a refuge pool habitat in the pilot studies and were accordingly evaluated under the supplementary scoring criteria to ascertain their conservation value relative to other refuge habitat in their respective catchments. In all cases this resulted in the lowering of scores, and in some instances a lowering of the priority ranking for consideration of removal. One of these barriers was the v-notch weir on the Goodga River which is used by DoW for stream gauging. This weir has an operational vertical-slot fishway that was installed in 2004 (Figure 7) primarily to facilitate upstream migratory movement of the Trout Minnow, a Critically Endangered species. However, despite the presence of the fishway, the weir remains a barrier for a number of other fish species that are unable to negotiate the fishway and are restricted to downstream reaches. Supplementary scoring criteria for the hypothetical removal of the barrier resulted in a reduction of the prioritisation score of 100 to 75. The fact that this was still the highest score for the catchment indicated that complete removal of the barrier warrants consideration.

In summary, the pilot studies have demonstrated that using a combination of desktop reviews of existing barrier information, landholder questionnaires, and aerial surveys is an extremely effective approach to the identification phase of the barrier prioritisation process. The calculation of prioritisation scores, while using an incomplete data set, resulted in all of the visually obvious higher priority barriers being ranked as such. Therefore, it proved to be a useful trial of this phase of the process.

SD 4: Evaluating small barrier removal to improve refuge connectivity 53

5. CONCLUSION

Lotic systems of southern Australia are predicted to continue to be impacted by surface flow reductions. These reductions will not only impact on the value of current instream barriers, but will also increase the need to enhance river connectivity. The resilience of many freshwater fishes will be dependent on maintaining migratory pathways to enable them to access refuge habitats and spawning grounds. Complete barrier removal and decommissioning as opposed to maintaining or retrofitting can have many environmental and economic benefits. However, in any barrier mitigation project, there is a need for a thorough understanding of the ecosystem, particularly the resident fish communities, in order to avoid unintended negative impacts. Barriers themselves, particularly in intermittent or ephemeral systems such as many of those in southern Australia, can create valuable permanent refuge habitats (e.g. weir pools, reservoirs) and this should be carefully considered during barrier mitigation assessments across this region. There can also be strong opposition to removing instream barriers and this is best approached by ensuring stakeholder involvement throughout the process particularly at the local level.

The stepwise barrier prioritisation process developed in the current project is underpinned by broad stakeholder engagement; with a particular focus on local level involvement. An understanding of the location and characteristics of instream barriers and refuge habitat, and the spatial distribution of the resident aquatic fauna community (with emphasis on fishes) in the target system is required. The approach we developed of combining landholder surveys, GIS database analysis, and conducting rapid aerial assessment (followed by eventual selective ground truthing) are key components of the identification phase of the process. The score and ranking system developed specifically incorporates the potential positive and negative impacts of barrier mitigation particularly with regard to the issue of refuge provision and accessibility. The three case studies in south-western Australia highlight both the efficiency and value of several aspects of the prioritisation process. The results indicated that whilst the GIS barrier database used in our study was extremely valuable in formulating shortlists of barriers worthy of further investigation, aerial truthing revealed a consistent level of false positives in the database in terms of fish barriers and also detected numerous additional barriers. The barriers within the least disturbed catchment scored much higher (i.e. approximately double) than those in the two other, more regulated catchments.

The study has identified a number of knowledge gaps pertaining to the impacts of instream barriers and their mitigation and removal that need to be addressed; particularly with regard to the location and characteristics of refuge habitat and the movement and ecology of freshwater fishes. However, the process for mitigating instream barriers for southern Australia developed here will be a valuable addition to management of aquatic ecosystems, and will help to increase resilience of fishes by reconnecting populations, expanding habitat availability and enhancing access to freshwater refuges.

54 SD 4: Evaluating small barrier removal to improve refuge connectivity

REFERENCES

Allen, GR & Berra, TM 1989, ‘Life history aspects of the West Australian salamanderfish, Lepidogalaxias salamandroides Mees’, Records of the Western Australian Museum, vol. 14, pp. 253-267.

Allen, GR, Midgley, SH & Allen, M 2002, Field guide to the freshwater fishes of Australia, Western Australian Museum, Perth.

American Rivers 2010, 60 dams removed to restore rivers in 2010, available from

Armornsakchai, S, Annez, P, Vongvisessomjai, S, Choowaew, S, Kunurat, P, Nippanon, J, Schouten, R, Sripapatrprasite, P, Vaddhanaphuti, C, Vidthayanon, C, Wirojanagud, W, & Watana, E 2000, Pak Mun Dam Mekong River Basin Thailand. Report to the World Commission on Dams (WCD).

Arthington, AH 2012, Environmental flows : saving rivers in the third millennium, University of California Press, Berkeley.

Arthington, AH, Hamlet, S & Blühdorn, DR 1990, ‘The role of habitat disturbance in the establishment of introduced warm-water fishes in Australia’ in Introduced and translocated fishes and their ecological effects, ed DA Pollard, Bureau of Rural Resources Proceedings No. 8, Australian Government Printing Service, Canberra, pp. 61-66.

ASCE 1997, Guidelines for retirement of dams and hydroelectric facilities, American Society of Civil Engineers, New York.

Barron, O, Silberstein, R, Ali, R, Donohue, R, McFarlane, DJ, Davies, P, Hodgson, G, Smart, N & Donn, M 2012, ‘Climate change effects on water-dependent ecosystems in south-western Australia’, Journal of Hydrology, vols. 434-435, pp. 95-109.

Beamer, E, Beechie, T & Klochak, J 1998, A strategy for implementation, effectiveness, and validation monitoring of habitat restoration projects, with two examples from the Skagit River basin, Washington, Completion report (Cost Share Agreement CCS-94- 04-05-01-050) to U.S. Forest Service, Sedro Woolley, Washington.

Beatty, S, Allen, M, & Keleher, J 2012, Assessment of the Rushy Creek fishway system, south .‐ western Report toAustralia the Department of Water, Western Australia.

Beatty, S, Allen, M, Morgan, D, Strehlow, K & Janicke, G 2011, Ecological significance of the fish, macroinvertebrates and riparian vegetation in the Mitchell River catchment, Centre for Fish & Fisheries Research, Murdoch University report to the Water Corporation of Western Australia.

Beatty, SJ, McAleer, FJ & Morgan, DL 2009, ‘Migration patterns of fishes of the Blackwood River and relationships to groundwater intrusion’, Centre for Fish & Fisheries Research, Murdoch University report to the Department of Water, Government of Western Australia, viewed 15 September 2012,

SD 4: Evaluating small barrier removal to improve refuge connectivity 55

Beatty, SJ, Morgan, DL & Torre, A 2007, ‘Restoring ecological connectivity in the Margaret River: Western Australia’s first rock-ramp fishways’, Ecological Management & Restoration, vol. 8, no. 3, pp. 224-229.

Bednarek, AT 2001. ‘Undamming rivers: a review of the ecological impacts of dam removal’, Environmental Management, vol. 27, pp. 803–814.

Beechie, T, Beamer, E & Wasserman, L 1994, ‘Estimating coho salmon rearing habitat and smolt production losses in a large river basin, and implications for habitat restoration’, North American Journal of Fisheries Management, vol. 14, pp. 797-811.

Bernhardt, ES, Palmer, MA, Allan, JD, Alexander, G, Barnas, K, Brooks, S, Carr, J, Clayton, S, Dahm, C, Follstad-Shah, J, Galat, D, Gloss, S, Goodwin, P, Hart, DD, Hassett, B, Jenkinson, R, Katz, S, Kondolf, GM, Lake, PS, Lave, R, Meyer, JL, O’Donnell, TK, Pagano, L, Powell, B & Sudduth, E 2005, ‘Synthesizing US river restoration efforts’, Science, vol. 308, pp. 636–637.

Berra, TM & Allen, GR 1989, ‘Burrowing, emergence, behavior, and functional morphology of the Australian salamanderfish, Lepidogalaxias salamandroides’, Fisheries, vol. 14, pp. 2-10.

Born, SM, Genskow, KD, Filbert, TL, Hernandez-Mora, N, Keefer, ML & White, KA 1998, ‘Socioeconomic and institutional dimensions of dam removals - the Wisconsin experience’, Environmental Management, vol. 22, no. 3, pp. 359-370.

Bradford, MJ 1997, ‘An experimental study of stranding of juvenile salmonids on gravel bars and in side channels during rapid flow fluctuations’, Regulated Rivers: Research and Management, vol. 13, pp. 395-401.

Bradford, MJ, Taylor, GC, Allan, JA & Higgins, PS 1995, ‘An experimental study of stranding of juvenile coho salmon and rainbow trout during rapid flow decreases in winter conditions’, North American Journal of Fisheries Management, vol. 15, pp. 473- 479.

Browning, K, Ryan, H & O’Donnell, E 2007, Feasibility study into the decommissioning of East Gippsland Water’s Nicholson Reservoir, Sinclair Knight Merz, Malvern, Victoria.

Buchanan, DV, Hooton, RM & Moring, JR 1981, ‘Northern squawfish (Ptychocheilus oregonensis) predation on juvenile salmonids in sections of the Willamette River basin, Oregon’, Canadian Journal of Fisheries and Aquatic Sciences, vol. 38, pp. 360-364.

Bunn, SE & Arthington, AH 2002, ‘Basic principles and ecological consequences of altered flow regimes for aquatic biodiversity’, Environmental Management, vol. 30, no. 4, pp. 492–507.

Burroughs, BA, Hayes, DB, Klomp, KD, Hansen, JF & Kiskak, J 2010, ‘The effects of the Stronach Dam removal on fish in the Pine River, Manistee County Michigan’, Transactions of the American Fisheries Society, vol. 139, pp. 1595-1613.

Carter, J, Tait, J, Kapitzke, R & Corfield, J 2007, Burdekin Dry Tropics NRM Region fish passage study, Final report, Alluvium.

Catalano, MJ, Bozek, MA & Pellett, TD 2007, ‘Effects of dam removal on fish assemblage structure and spatial distributions in the Baraboo River, Wisconsin’, North American Journal of Fisheries Management, vol. 27, pp. 519-530.

56 SD 4: Evaluating small barrier removal to improve refuge connectivity

Chaplin, JJ, Brightbill, RA & Bilger, MD 2005, Effects of removing Good Hope Mill Dam on selected physical, chemical, and biological characteristics of Conodoguinet Creek, Cumberland County, Pennsylvania, U.S. Geological Survey Scientific Investigations Report 2005-5226.

Clavero, M, Blanco-Garrido, F & Prenda, J 2004, ‘Fish fauna in Iberian Mediterranean river basins: biodiversity, introduced species and damming impacts’, Aquatic Conservation: Marine and Freshwater Ecosystems, vol. 15, pp. 415-426.

Coe, HJ, Duda, JJ, Kloehn, KK, McHenry, ML & Morley, SA 2008, ‘Benthic invertebrates and periphyton in the Elwha River Basin: current conditions and predicted response to dam removal’, Northwest Science, vol. 82, pp. 179-196.

Coutant, CC & Whitney, RR 2000, Fish behavior in relation to passage through hydropower turbines: a review’, Transactions of the American Fisheries Society, vol. 129, no. 2, pp. 351-380.

Craig, JF & Kemper, JB (eds) 1987, Regulated streams: advances in ecology, Plenum Press, New York.

CSCC 2004, Inventory of barriers to fish passage in California’s coastal watersheds, California State Coastal Conservancy (CSSC), Oakland, California.

CSIRO 2009a, Surface water yields in south-west Western Australia, a report to the Australian government from the CSIRO south-west Western Australia sustainable yields project, CSIRO water for a healthy country flagship, Australia.

CSIRO 2009b, Groundwater yields in south-west Western Australia, a report to the Australian government from the CSIRO south-west Western Australia sustainable yields project, CSIRO water for a healthy country flagship, Australia.

Dadswell, MJ 1996, The removal of Edwards Dam, Kennebec River, Maine: Its effects on the restoration of anadromous fishes, Draft environmental impact statement, Kennebec River, Maine.

Davies, PM 2010, ‘Climate change implications for river restoration in global biodiversity hotspots’, Restoration Ecology, vol. 18, pp. 261-268. de Graaf, M, Beatty, S & Molony, BM 2010, Evaluation of the recreational marron fishery against environmental change and human interaction, Final report to the Fisheries Research and Development Corporation on Project No. 2003/027, eds D Baxter & R Larsen, Fisheries research report no. 211, Department of Fisheries, Western Australia.

Department of Agriculture, Fisheries & Forestry (DAFF) 2012, Human impacts on wetlands. Available from: [30 July 2012].

Doeg, TJ & Koehn, JD 1994, ‘Effects of draining and desilting a small weir on downstream fish and macroinvertebrates’, Regulated Rivers: Research & Management, vol. 9, no. 4, pp. 263-277.

DOI 1995, Final environmental impact statement: Elwha River ecosystem restoration, Olympic National Park, Washington, Department of the Interior.

SD 4: Evaluating small barrier removal to improve refuge connectivity 57

Donnelly, CR, King, L & Phillips, M 2004, ‘An evaluation of the impacts of the removal of the Finlayson Dam’, Paper presented at the 24th Annual USSD Conference, St. Louis, Missouri.

DoW 2010, South West regional water plan: supporting detail, Department of Water, Perth, Western Australia.

Doyle, MW, Stanley, EH & Harbor, JM 2003, ‘Channel adjustments following two dam removals in Wisconsin’, Water Resources Research, vol. 39, no. 1, 1011, 15 pp.

DSE 2003, State fishway program redundant weir review, Final report prepared for Catchment and Water Division, Department of Sustainability and Environment, Victoria.

Dumont, U 2005, Handbuch Querbauwerke, Herausgeber: Ministerium fur Umwelt und Naturschutz, Landwirtschaft und Verbraucherschutz des Landes NRW, Dusseldorf.

Estes, JR, Myers, RA & Mantini, L 1993, Fisheries investigations of Newnans Lake, State of Florida, Game and Fresh Water Fish Commission, A Wallop-Breaux Project F- 55-6, Study IV.

Fausch KD, Rieman, BE, Dunham, JB & Young, MK 2006, Strategies for conserving native salmonid populations at risk from nonnative fish invasion: tradeoffs in using barriers to upstream movement, USDA Forest Service, Rocky Mountain Research Station, GTR-RMRS-174, Fort Collins, Colorado.

Feld, CK, Birk, S, Bradley, DC, Hering, D, Kail, J, Marzin, A, Melcher, A, Nemitz, D, Pedersen, ML, Pletterbauer, F, Pont, D, Verdonschot, PFM & Friberg, N 2011, ‘From natural to degraded rivers and back again: a test of restoration ecology and practice’, Advances in Ecological Research, vol. 44, pp. 119-209.

Fletcher, AR, Morison, AK & Hume, DJ 1985, ‘Effects of carp (Cyprinus carpio L.) on aquatic vegetation and turbidity of waterbodies in the lower Goulburn River Basin’, Australian Journal of Marine & Freshwater Research, vol. 36, pp. 311-327.

Gasith, A & Resh, VH 1999, ‘Streams in mediterranenan climate regions: abiotic influences and biotic responses to predictable seasonal events’, Annual Review of Ecology and Systematics, vol. 30, pp. 51–81.

Gehrke, PC, Gilligan, DM & Barwick, M 2001, Fish communities and migration into the Shoalhaven River: before construction of a fishway, NSW Fisheries final report series, no. 26.

Gehrke, PC, Gilligan, DM & Barwick, M 2002, ‘Changes in fish communities of the Shoalhaven River 20 years after construction of Tallowa Dam, Australia’, River Research and Applications, vol. 18, pp. 265-286.

Gehrke, PC & Harris, JH 2001, ‘Regional-scale effects of flow regulation on lowland riverine fish communities in New South Wales, Australia’, Regulated Rivers: Resource Management, vol. 17, pp. 369-391.

GeoFin (2012) Geospatial Fisheries Information Network website, US Fish and Wildlife Service, available from: [30 August 2012].

58 SD 4: Evaluating small barrier removal to improve refuge connectivity

Gordos, M, Nichols, S, Lay, C, Townsend, A, Grove, C, Walsh, S & Copeland, C 2007, ‘Audit and remediation of fish passage barriers in coastal NSW’, Proceedings of the 5th Australian stream management conference. Australian rivers: making a difference. Charles Sturt University, Thurgoona, New South Wales, pp. 109-114.

Graber, B 2003, ‘Potential economic benefits of small dam removal’, in Dam removal research - status and prospects, ed WL Graf, The H. John Heinz III Center for Science, Economics and the Environment, Washington, DC, pp. 56-66.

Graf, WL 2003, ‘Summary and perspective’ in Dam removal research - status and prospects, ed WL Graf , editor. The H. John Heinz III Center for Science, Economics and the Environment, Washington, DC, pp. 1-22.

Gregory, S, Hiram, L & Li, J 2002, ‘The conceptual basis for ecological responses to dam removal’, BioScience, vol. 52, pp. 713-723.

Harris, JH 1984, ‘Impoundment of coastal south-eastern Australia, and a review of its relevance to fish migrations’, Australian Zoologist, vol. 21, pp. 235-250.

Hart, DD, Johnson, TE, Bushaw-Newton, KL, Horwitz, RJ, Bednarek, AT, Charles, DF, Kreeger, DA & Velinsky, DJ 2002, ‘Dam removal: challenges and opportunities for ecological research and river restoration’, BioScience, vol. 52, no. 8, pp.669-681.

Hart, DD, Johnson, TE, Bushaw-Newton, KL, Horwitz, RJ & Pizzuto, JE 2003, ‘Ecological effects of dam removal: an integrative case study and risk assessment framework for prediction’ in Dam removal research: status and prospects, ed WL Graf, The H. John Heinz III Center for Science, Economics and the Environment, Washington, DC, pp. 67-80.

Heinz Center 2002, Dam removal: science and decision making, H.J. Heinz Center for Science, Economics and the Environment, Washington DC.

Hill, MJ, Long, EA & Hardin, S 1993, ‘Effects of dam removal on Dead Lake, Chipola River, Florida’, Apalachicola River Watershed Investigations, Florida Game and Fresh Water Fish Commission, A Wallop-Breaux Project F-39-R.

Hughes, L 2003, ‘Climate change and Australia: trends, projections and impacts’, Austral Ecology, vol. 28, pp. 423–443.

IFS 2004, A survey of fish barriers in coastal Tasmanian rivers, preliminary investigation, Inland Fisheries Service, Tasmania.

Indian Ocean Climate Initiative (IOCI) 2008, Project 1.2: South-west Western Australia’s regional surface climate and weather systems, IOCI stage 3, milestone report no. 1, (research to the end of 2008).

Industry and Investment NSW (I&I NSW) 2009, Bringing back the fish – improving fish passage and aquatic habitat in coastal NSW, Final report to the Southern Rivers Catchment Management Authority, Industry and Investment NSW, Cronulla.

Iversen, TM, Kronvang, B, Madsen, BL, Markmann, P & Nielsen, MB 1993, ‘Re- establishment of Danish streams: restoration and maintenance measures’, Aquatic Conservation: Marine and Freshwater Ecosystems, vol. 3, pp. 73-92.

SD 4: Evaluating small barrier removal to improve refuge connectivity 59

Johnson, SE & Graber, BE 2002, ‘Enlisting the social sciences in decisions about dam removal’, BioScience, vol. 52, no. 8, pp. 731-738.

Junk, WJ, Bayley, PB & Sparks, RE 1989, ‘The flood pulse concept in riverfloodplain systems’ in Proceedings of the international large river symposium, ed DP Dodge, Ottawa, Ontario, Department of fisheries and oceans, Canadian special publication of fisheries and aquatic sciences, vol. 106, pp. 110–127.

Kanehl, PD, Lyons, J & Nelson, JE 1997, ‘Changes in the habitat and fish community of the Milwaukee River, Wisconsin, following removal of the Woolen Mills Dam’, North American Journal of Fisheries Management, vol. 17, pp. 387-400.

Katopodis, C 1992, Introduction to Fishway Design, Freshwater Institute, Department of Fisheries and Oceans, Winnipeg.

Katopodis, C & Aadland, LP 2006, ‘Effective dam removal and river channel restoration approaches’, IAHR International Journal of River Basin Management, Special Issue on Ecohydraulics, vol. 4, no. 3, pp. 153-168.

Kaufman, JH 1992, ‘Effects on wildlife of restoring the riverine forest in the Rodman pool area’ in The case for restoring the free-flowing Ocklawaha River, Florida Defenders of the Environment, Gainesville, Florida, pp. 38-41.

Kemp, PS & O’Hanley, JR 2010, ‘Evaluation of barriers to fish migration and prioritisation of removal and mitigation projects’, Fisheries Management and Ecology, vol. 17, pp. 297-322.

Kemp, PS, Russon, IJ, Waterson, B, O’Hanley, J & Pess GR 2008, Recommendations for a ‘‘coarse-resolution rapid-assessment’’ methodology to assess barriers to fish migration and associated prioritization tools – Final report, Stirling: Scottish Environment Protection Agency.

King, C & Torre, A 2007, ‘Reconnecting off-channel habitats to waterways - using engineering techniques to restore fish passage, Proceedings of WA Wetland Management Conference, 2 February 2007.

Kondolf, GM 1997, ‘Hungry water: effects of dams and gravel mining on river channels’, Environmental Management, vol. 21, no. 4, pp. 533-551.

Kroes MJ, Gough, P, Schollema, PP & Wanningen, H (eds) 2006, From sea to source: practical guidance for restoration of fish migration in European rivers, Groningen: Interreg IIIC project ‘‘Community Rivers’’, available from [15 May 2012].

Kuby, MJ, Fagan, WF, ReVelle, CS & Graf, WL 2005, ‘A multiobjective optimization model for dam removal: an example of trading off salmon passage with hydropower and water storage in the Willamette basin’, Advances in Water Resources, vol. 28, pp. 845-855.

Lake, PS & Marchant, R 1990, ‘Australian upland streams: ecological degradation and possible restoration’, Proceedings of the Ecological Society of Australia, vol. 16, pp. 79- 91.

Lawrence, M, Sully, D, Beumer, J & Couchman, D 2010, Fisheries guidelines for conducting an inventory of instream structures in coastal Queensland (revised version),

60 SD 4: Evaluating small barrier removal to improve refuge connectivity

Department of Employment, Economic Development and Innovation, Fish habitat guideline FHG 007 update.

Lawson, T, Kroon, F, Russell, J & Thuesen, P 2010, Audit and prioritisation of physical barriers to fish passage in the Wet Tropics region, Milestone report, MTSRF project 2.6.2, CSIRO Sustainable Ecosystems, Atherton, Queensland.

Lejon, AGC, Malm Renöfält, B & Nilsson, C 2009, ‘Conflicts associated with dam removal in Sweden’, Ecology and Society, vol. 14, no. 2, p 4.

Li, HW, Schreck, CB, Bond, CE & Rexstad, E 1987, ‘Factors influencing changes in fish assemblages of Pacific Northwest streams’ in Community and evolutionary ecology of North American stream fishes, eds WJ Matthews & DC Heins, University of Oklahoma Press, Norman, pp. 193–202.

Lintermans, M 2012, ‘Managing potential impacts of reservoir enlargement on threatened Macquaria australisica and Gadopsis bispinosus in southeastern Australia’, Endangered Species Research, vol. 16, pp. 1-16.

Losos, E, Hayes, J, Phillips, A, Wilcove, D & Alkire, C 1995, ‘Taxpayer-subsidized resource extraction harms species’, BioScience, vol. 45, pp. 446-455.

Lucas, MC & Baras, E 2001, Migration of freshwater fishes. Blackwell Science, Oxford.

Lymbery, AJ, Hassan, M, Morgan, DL, Beatty, SJ & Doupe, RG 2010, ‘Parasites of native and exotic freshwater fishes in south-western Australia’, Journal of Fish Biology, vol. 76, pp. 1770-1785.

Mallen-Cooper, M 1992, ‘The swimming ability of juvenile Australian bass, Macquaria novemaculeata (Steindachner), and juvenile barramundi, Lates calcarifer (Bloch), in an experimental vertical-slot fishway’, Australian Journal of Marine and Freshwater Research, vol. 43, pp. 823-834.

Mallen-Cooper, M 1994, ‘Swimming ability of adult golden perch, Macquaria ambigua (Percicthyidae), and adult silver perch, Bidyanus bidyanus (Teraponidae), in an experimental vertical slot fishway’, Australian Journal of Marine and Freshwater Research, vol. 45, pp. 191-198.

Mallen-Cooper, M 2000, Review of fish passage in NSW, NSW Fisheries, Cronulla.

McLaughlin, RL, Hallet, A, Pratt, TC, O’Connor, LM & McDonald, DG 2007, ‘Research to guide use of barriers, traps and fishways to control sea lamprey’, Journal of Great Lakes Research, vol. 33, special issue no. 2, pp. 7-19.

McNeil, DG, Wilson, PJ, Fredberg, JF & Westergaard, S 2010, Fish passage at the Breakout Creek fishway, River Torrens, South Australia, Report to the Adelaide and Mount Lofty Ranges Natural Resources Management Board, South Australian Research and Development Institute (Aquatic Sciences), Adelaide, SARDI publication no. F2009/000409-1, SARDI Research Report Series no. 440.

Meffe, GK & Sheldon, AL 1988, ‘The influence of habitat structure on fish assemblage composition in southeastern blackwater streams’, The American Midland Naturalist, vol. 120, pp. 225-240.

SD 4: Evaluating small barrier removal to improve refuge connectivity 61

Monden, S, De Charleroy, D & Van Liefferinge, C 2000, ‘Inventory of fish migration barriers on ecological and strategic important rivers in the Flemish region (Belgium)’, Proceedings of the International Symposium on Freshwater Fish Conservation: Options for the future, 30 October - 4 November 2000, Albufeira.

Morgan, D & Beatty, S 2004a, Margaret River Fishway, Report to Margaret River Regional Environment Centre, Freshwater Fish Research, Murdoch University, Western Australia.

Morgan, D & Beatty, S 2004b, Monitoring the Lion’s Weir Fishway – Hotham River, Western Australia, Report to the Department of Environment, Centre for Fish and Fisheries Research, Murdoch University, Western Australia.

Morgan, DL & Beatty, SJ 2006, ‘Use of a vertical-slot fishway by galaxiids in Western Australia’, Ecology of Freshwater Fish, vol. 15, pp. 500-509.

Morgan DL, Beatty SJ, Klunzinger, MW, Allen, MG & Burnham, QE 2011, A Field Guide to Freshwater Fishes, Crayfishes & Mussels of South-Western Australia, SERCUL & Freshwater Fish Group & Fish Health Unit, Murdoch University, Murdoch, Western Australia.

Morgan, DL, Gill, HS, Maddern, MG & Beatty, SJ 2004, ‘Distribution and impacts of introduced freshwater fishes in Western Australia’, New Zealand Journal of Marine and Freshwater Research, vol. 38, pp. 511-523.

Morgan, DL, Gill, HS & Potter, IC 1998, ‘Distribution, identification and biology of freshwater fishes in southwestern Australia’, Records of the Western Australian Museum Supplement, no. 56, pp. 1-97.

Morrongiello, JR, Beatty, SJ, Bennett, JC, Crook, DA, Ikedife, DNEN, Kennard, MJ, Kerezy, A, Lintermans, M, McNeil, DJ, Pusey, BJ & Rayner, T 2011, ‘Climate change and its implications for Australia’s freshwater fish’, Marine and Freshwater Research, vol. 62, pp. 1082-1098.

Moyle, PB 1986, ‘Fish introductions into North America: patterns and ecological impact’ in Ecology of biological invasions of North America and Hawaii, eds HA Mooney & JA Drake, Springer-Verlag, New York, pp. 27-43.

Murray-Darling Basin Commission (MDBC) 2003, Native fish strategy for the Murray- Darling Basin 2003-2013, Murray-Darling Basin Commission, Canberra.

Murray-Darling Basin Commission (MDBC) 2008, The sea to Hume Dam: restoring fish passage in the Murray River, ed J Barrett, Murray-Darling Basin Commission, Canberra, ACT.

Naiman, RJ, Bunn, SE, Nilsson, C, Petts, GE, Pinay, G & Thompson, LC 2002, ‘Legitimizing fluvial ecosystems as users of water: an overview’, Environmental Management, vol. 30, pp. 455-467.

Neave, M, Rayburg, S & Swan, A 2009, ‘River channel change following dam removal in an ephemeral stream’, Australian Geographer, vol. 40, no. 2, pp. 235-246.

Nelson, JE & Pajak, P 1990, ‘Fish habitat restoration following dam removal on a warmwater river’ in The restoration of Midwestern stream habitat, American Fisheries

62 SD 4: Evaluating small barrier removal to improve refuge connectivity

Society, North Central Division, Rivers and Streams Technical Committee Symposium Proceedings at the 52nd Midwest Fish and Wildlife Conference, December 1990, Minneapolis, Minnesota.

Nelson, M 2003a, A survey of fish barriers in coastal Tasmanian rivers, preliminary investigation, Inland Fisheries Service internal report, June 2003.

Nelson, M 2003b, Weir removal project summary, Inland Fisheries Service internal report, March 2003.

Nesler, TP, Muth, RT & Wasowicz, AF 1988, ‘Evidence for baseline flow spikes as spawning cues for Colorado Squawfish in the Yampa River, Colorado’, American Fisheries Society Symposium, vol. 5, pp.68-79.

Northcote, TG 1998, ‘Migratory behavior of fish and its significance to movement through riverine fish passage facilities’ in Fish Migration and Fish Bypasses, eds M Jungwirth, S Schmutz & S Weiss, Fishing News Books, Blackwell Science, Oxford, pp. 3-18.

Norton, S & Storer T 2011, Prioritising fish barriers in Western Australia, Water science technical series report, Department of Water, Perth, Western Australia. NRCS 2000, A report to Congress on aging watershed infrastructure, US Department of Agriculture, Natural Resources Conservation Service, Washington DC.

NSW Department of Primary Industries (NSW DPI) 2006a, Reducing the impact of weirs on aquatic habitat - New South Wales detailed weir review. Murray CMA region, Report to the New South Wales Environmental Trust, NSW Department of Primary Industries, Flemington, NSW.

NSW Department of Primary Industries (NSW DPI) 2006b, Reducing the impact of road crossings on aquatic habitat in coastal waterways, Report to the New South Wales Environmental Trust, NSW Department of Primary Industries, Flemington, NSW.

NSW Fisheries 2002, Initial weir and floodgate review, Report for the State Weir Review Committee, NSW Fisheries, Ballina.

O’Brien, T, Ryan, T, Stuart, I & Saddlier, S 2010, Review of fishways in Victoria 1996– 2009, Arthur Rylah Institute for Environmental Research, Technical Report Series No. 216, Department of Sustainability and Environment, Heidelberg, Victoria.

ODFW 2010, Oregon Fish Passage Barrier Data Standard, Version 1.1, Oregon Department of Fish and Wildlife (ODFW), Salem, Oregon, available at [21 June 2012].

O’Hanley, JR 2011, ‘Open rivers: Barrier removal planning and restoration of free- flowing rivers’, Journal of Environmental Management, vol. 92, pp. 3112-3120.

O’Hanley, JR & Tomberlin, D 2005, ‘Optimizing the removal of small fish passage barriers’, Environmental Modeling and Assessment, vol. 10, pp. 85-98.

Olden, JD & Cucherousset, J 2011, ‘Ecological impacts of nonnative freshwater fishes, Fisheries, vol. 36, pp. 215-230.

SD 4: Evaluating small barrier removal to improve refuge connectivity 63

Orr, CH, Rogers, KL & Stanley, EH 2006, ‘Channel morphology and P uptake following removal of a small dam’, Journal of the North American Benthological Society, vol. 25, pp. 556-568.

PAD 2009, California fish Passage Assessment Database (PAD) project: methodology and documentation, Pacific States Marine Fisheries Commission, Sacramento, California.

Palmer, MA, Allen, JD, Meyer, J & Bernhardt, ES 2007, ‘River restoration in the twenty- first century: data and experimental knowledge to inform future efforts’, Restoration Ecology, vol. 15, no. 3, pp. 472-481.

Palmer, MA, Bernhardt, ES, Allen, JD, Lake, PS, Alexander, G, Brooks, S, Carr, J, Clayton, S, Dahm, CN, Follstad Shah, J, Galat, DL, Loss, SG, Goodwin, P, Hart, DD, Hassett, B, Jenkinson, R, Kondolf, GM, Lave, R, Meyer, JL, O’Donnell, TK, Pagano, L & Sudduth, E 2005, ‘Standards for ecologically successful river restoration’, Journal of Applied Ecology, vol. 42, pp.208-217.

Pen, LJ & Potter, IC 1990, ‘Biology of the nightfish, Bostockia porosa Castelnau, in a south western Australian river’, Australian Journal of Marine and Freshwater Research, vol. 41, pp. 627-645.

Pen LJ & Potter IC 1991a, ‘Biology of the western minnow, Galaxias occidentalis Ogilby (Teleostii: Galaxiidae), in a south-western Australian river’, Hydrobiologia, vol. 211, pp. 89-100.

Pen LJ & Potter IC 1991b, ‘The biology of the western pygmy perch, Edelia vittata, and comparisons with two other teleost species endemic to south-western Australia’, Environmental Biology of Fishes, vol. 31, pp. 365-380.

Pess, GR, McHenry, ML, Beechie, TJ & Davies, J 2008, ‘Biological impacts of the Elwha River dams and potential salmonid responses to dam removal’, Northwest Science vol. 82 (special issue), pp. 72-90.

Pethebridge, R, Lugg, A & Harris, J 1998, Obstructions to fish passage in New South Wales south coast streams, Final report series 4, Cooperative Research Centre for Freshwater Ecology and NSW Fisheries, Cronulla.

Poff, NL, Allan, JD, Bain, MB, Karr, JR, Prestegaard, KL, Richter, BD, Sparks, RE & Stromberg, JC 1997, ‘The natural flow regime’, BioScience, vol. 47, pp. 769-784.

Pohl, MM 2003, ‘American dam removal census: available data and data needs’ in Dam removal research - status and prospects, ed WL Graf, The H. John Heinz III Center for Science, Economics and the Environment, Washington, DC, pp. 29-39. Pollard, AI & Reed, T 2004, ‘Benthic invertebrate assemblage change following dam removal in a Wisconsin stream’, Hydrobiologia, vol.513, pp. 51-58.

Pusey, BJ, Arthington, AH & Read, MG 1993, ‘Spatial and temporal variation in fish assemblage structure in the Mary River, south-east Queensland: the influence of habitat structure’, Environmental Biology of Fishes, vol. 37, pp. 355-380.

Pusey, BJ, Kennard, MJ & Arthington, AH 2000, ‘Discharge variability and the development of predictive models relating stream fish assemblage structure to habitat in north-eastern Australia’, Ecology of Freshwater Fishes, vol. 9, pp. 30-50.

64 SD 4: Evaluating small barrier removal to improve refuge connectivity

Raadik, TA, Fairbrother, PS & Smith, SJ 2010, National Recovery Plan for barred galaxias Galaxias fucus, Department of Sustainability and Environment, Melbourne.

Randle, TJ 2003, ‘Dam removal and sediment management’ in Dam removal research - status and prospects, ed WL Graf, The H. John Heinz III Center for Science, Economics and the Environment, Washington, DC, pp. 81-104.

Randle, TJ 2003, ‘Dam removal and sediment management’ in Dam removal research - status and prospects, ed WL Graf, The H. John Heinz III Center for Science, Economics and the Environment, Washington, DC, pp. 81-104.

Rathburn, SL & Wohl, EE 2003, ‘Sedimentation hazards downstream from reservoirs’ in Dam removal research - status and prospects, ed WL Graf, The H. John Heinz III Center for Science, Economics and the Environment, Washington, DC, pp. 105-118.

Raymond, HL 1979, ‘Effects of dams and impoundments on migrations of juvenile chinook salmon and steelhead from the Snake River, 1966 to 1975’, Transactions of the American Fisheries Society, vol. 108, pp. 505-529.

Roberts, TR 2001, ‘On the river of no returns: Thailand’s Pak Mun Dam and its fish ladder’, Natural History Bulletin - Siam Society, vol. 49, pp. 189-230.

Robson, BJ, Chester, ET & Austin, CM 2011, ‘Why life history information matters: predicting the role of drought refuges for sustaining macroinvertebrate biodiversity in non-perennial streams subject to increased periods of drying’, Marine and Freshwater Research, vol. 62, pp. 801–810.

Ryan, T, Stuart, I, O’Brien, T & Saddlier, S 2010, ‘Instream barrier assessment and prioritisation of the Lower Barwon River and Otways streams’, Report for the Corangamite Catchment Management Authority.

Sarakinos, H & Johnson, SE 2003, ‘Social perspectives on dam removal’ in Dam removal research - status and prospects, ed WL Graf, The H. John Heinz III Center for Science, Economics and the Environment, Washington, DC, pp. 40-55.

Scully, RJ, Leitzinger, EJ & Petrosky, CE 1990, Idaho Habitat Evaluation for Off-Site Mitigation Record, Bonneville Power Administration, Annual Report 1988, Contract No. DE-A179-84BP133881, Portland, Oregon.

Service, RF 2011, ‘Will busting dams boost salmon?’, Science, vol. 334, pp. 888-892, available at [23 July 2012].

Sheer, MB & Steel, EA 2006, ‘Lost watersheds: barriers, aquatic habitat connectivity, and salmon persistence in the Willamette and Lower Columbia River basins’, Transactions of the American Fisheries Society, vol. 135, no.6, pp. 1654-1669.

Simons, RK & Simons, DB 1991, ‘Sediment problems associated with dam removal- Muskegon River, Michigan’ in Hydraulic engineering, Proceedings of the 1991 national conference of the American Society of Civil Engineers, Nashville, Tennessee, pp. 680- 685.

Skalski, JR 1998, ‘Estimating season-wide survival rates of outmigrating salmon smolt in the Snake River, Washington’, Canadian Journal of Fisheries and Aquatic Sciences, vol. 55, pp. 761-769.

SD 4: Evaluating small barrier removal to improve refuge connectivity 65

SKM 2007, Impacts of farm dams in seven catchments in Western Australia, Report prepared for the Department of Water, Western Australia.

Shuman, JR 1995, ‘Environmental considerations for assessing dam removal alternatives for river restoration’, Regulated Rivers: Research & Management, vol.11, pp. 249–261.

Silberstein, RP, Aryal, SK, Durrant, J, Pearcey, M, Braccia, M, Charles, SP, Boniecka, L, Hodgson, GA, Bari, MA, Viney, NR & McFarlane, DJ 2012, ‘Climate change and runoff in south-western Australia’, Journal of Hydrology, vol. 475, pp. 441-455.

Slawski, TM, Veraldi, FM, Pescitelli, SM & Pauers, MJ 2008, ‘Effects of tributary spatial position, urbanization, and multiple low-head dams on warmwater fish community structure in a Midwestern stream’, North American Journal of Fisheries Management, vol. 28, pp. 1020-1035.

SoE 2011, Australian State of the Environment Report, State of the Environment Committee, Report to the Department of Sustainability, Environment, Water, Population and Communities, Australian Commonwealth Government.

Staggs, M, Lyons, J & Visser, K 1995, ‘Habitat restoration following dam removal on the Milwaukee River at West Bend’ in Wisconsin’s biodiversity as a management issue: a report to Department of Natural Resources managers, Wisconsin Department of Natural Resources, pp. 202-203.

Stanford, JA, Ward, JV, Liss, WJ, Frissell, CA, Williams, RN, Lichatowich, JA & Coutant, CC 1996, ‘A general protocol for restoration of regulated rivers’, Regulated Rivers: Research & Management, vol. 12, pp. 391-413.

Stanley, EH, Catalano, MJ, Mercado-Silva, N & Orr, CH 2007, ‘Effects of dam removal on brook trout in a Wisconsin stream’, River Research & Applications, vol. 23, pp. 792- 798.

Stanley, EH & Doyle, MW 2003, ‘Trading off: the ecological effects of dam removal’, Frontiers in Ecology and the Environment, vol. 1, no. 1, pp. 15-22.

Stanley, EH, Fisher, SG & Grimm, NB 1997, ‘Ecosystem expansion and contraction in streams’, Bioscience, vol. 47, pp. 427-435.

Stanley, EH, Luebke, MA, Doyle, MW & Marshall, DW 2002, ‘Short-term changes in channel form and macroinvertebrate communities following low-head dam removal’, Journal of the North American Benthological Society, vol. 21, pp. 172-87.

State Water 2012, Completed fish passage, available from: [25 July 2012].

State Water Corporation 2011, Fish Superhighways, Internal report by the NSW State Water Corporation.

Stewart, R & Marsden, T 2006, Reef coast freshwater fish habitat rehabilitation strategy, Report for the Department of Primary Industries and Fisheries, Queensland.

66 SD 4: Evaluating small barrier removal to improve refuge connectivity

Stockwell, B, Fennesy, R, Berghuis, A, Johnston, B & Hutchison, M 2008, Burnett Mary regional biopass strategy – reconnecting the dreamtime’s rainbow serpent, Report to the Department of Primary Industries and Fisheries, Brisbane, Queensland.

Storer, T & Norton, S 2011, An introduction to fish barriers and fishways for Western Australia, Water science technical series report, Department of Water, Perth, Western Australia.

Suppiah, R, Hennessy, KJ, Whetton, PH, McInnes, K, Macadam, I, Bathols, J, Ricketts, J & Page, CM 2007, ‘Australian climate change projections derived from simulations performed for the IPCC 4th assessment report’, Australian Meteorological Magazine, vol. 131, pp. 131-152.

Taylor, RN & Love, M 2003, California salmonid stream habitat restoration manual - Part IX fish passage evaluation at stream crossings, California Department of Fish and Game, Sacramento, California.

Thomson, JR, Hart, DD, Charles, DF, Nightingale, TL & Winter, DM 2005, ‘Effects of removal of a small dam on downstream macroinvertebrate and algal assemblages in a Pennsylvania stream’, Journal of the North American Benthological Society, vol. 24, pp. 192-207.

Thorncraft, G & Harris, JH 2000, Fish passage and fishways in NSW: a status report, Cooperative Research Centre for Freshwater Ecology Technical Report 1/2000, Canberra, ACT.

Travnicheck, VH, Zale, AV & Fisher, WL 1993, Entrainment of ichthyoplankton by a warmwater hydroelectric facility’, Transactions of the American Fisheries Society, vol. 122, pp. 709-716.

Velez-Espino, LA, McLaughlin, RL, Jones, ML & Pratt, TC 2011, ‘Demographic analysis of trade-offs with deliberate fragmentation of streams: control of invasive species versus protection of native species’, Biological Conservation, vol. 144, no. 3, pp. 1068-1080.

Walker, KF 2008, Environmental impact statement for Traveston Crossing Dam (Mary River, Queensland): a review with regard for species of concern under the EPBC Act 1999, Report to the Department of Environment, Water, Heritage and Arts, Canberra.

Ward, JV & Stanford, JA 1979, The ecology of regulated streams, Plenum Press, New York.

Warner, K & Pejchar, L 2001, ‘A river might run through it again: criteria for consideration of dam removal and interim lessons from California’, Environmental Management, vol. 28, no. 5, pp. 561-575.

Washington Department of Fish and Wildlife (WDFW) 2009, Fish passage and surface water diversion screening assessment and prioritization manual, Washington Department of Fish and Wildlife, Olympia, Washington.

Water Corporation 2009, Water forever – towards climate resilience, Water Corporation, Perth, Western Australia, available from [6 July 2012]

SD 4: Evaluating small barrier removal to improve refuge connectivity 67

Webb, JA & Padgham, M 2009, ‘Patterns of dispersal through stream networks respond simply to multiple structural modifications’, Proceedings of the 18th World IMACS / MODSIM Congress, Cairns, Australia 13-17 July 2009, available from

Williams, RJ, Watford, FA & Taylor, MA 1996, A summary of aspects of FRDC project 94/041 “Restoration of estuarine fisheries habitat” relevant to tidal obstructions in New South Wales estuaries, NSW Fisheries Research Institute, Cronulla, NSW.

Wofford, JEB, Gresswell, RE & Banks, MA 2005, ‘Influence of barriers to movement on within-watershed genetic variation of coastal cutthroat trout’, Ecological Applications, vol. 15, pp. 628-637.

Wood, PJ & Armintage, PD 1997, ‘Biological effects of fine sediment in the lotic environment’, Environmental Management, vol. 21, no. 2, pp. 203-217.

World Commission on Dams (WCD) 2000, Dams and development: a new framewok for decision-making, Earthscan Publications, London.

Wilson, AL, Dehaan, RL, Watts, RJ, Page, KJ, Bowmer, KH & Curtis, A (eds) 2007, Proceedings of the 5th Australian Stream Management Conference, Australian rivers: making a difference, Charles Sturt University, Thurgoona, New South Wales.

WWF 2001, The status of wild Atlantic Salmon: a river by river sssessment, World Wildlife Fund (WWF).

WWF 2009, Liberando ríos – propuestas de WWF para el desmantelamiento de presas en España, WWF Spain, Madrid.

Zampatti, BP, Bice, CM & Jennings, PR 2012, Fish assemblage response and fishway effectiveness at Goolwa, Tauwitchere and Hunters Creek Barrages in 2010/11, South Australian Research and Development Institute (Aquatic Sciences), Adelaide, SARDI Publication no. F2011/000186-2, SARDI Research Report Series No. 605.

Zheng, PQ, Hobbs, BF & Koonce, JF 2009, ‘Optimizing multiple dam removals under multiple objectives: linking tributary habitat and the Lake Erie ecosystem’, Water Resources Research, vol. 45, W12417.

Zitek, A, Haidvogl, G, Jungwirth, M, Pavlas, P & Schmutz, S 2008, ‘An ecologically- based strategic guideline for restoring the longitudinal connectivity for fish in running waters of Austria’, 4th ECCR International Conference on River Restoration, 16–21 June 2008, Venice, European Centre for River Restoration (ECRR).

68 SD 4: Evaluating small barrier removal to improve refuge connectivity

APPENDIX 1

Example of flyer and Questionnaire mailed to riparian landholders.

SD 4: Evaluating small barrier removal to improve refuge connectivity 69

Appendix 1 (continued).

70 SD 4: Evaluating small barrier removal to improve refuge connectivity

Appendix 1 (continued).

SD 4: Evaluating small barrier removal to improve refuge connectivity 71

APPENDIX 2

Scores and rankings for instream barrier prioritisation in the three pilot study catchments in south-western Australia.

Goodga prioritisation rankings Mitchell prioritisation rankings Ludlow prioritisation rankings Rank Score Supp Barrier # Rank Score Supp Barrier # Rank Score Supp Barrier # 1 100 75 4 1 101 3 1 42 32 3 2 60 45 5/6 2 83 20 16 2 38 80 3 53 28 3 54 8 3 37 32 4 46 102 4 52 23 21 4 36 31 5 37 18 5 45 22 5 29 30 6 33 79 6 44 15 6 29 75 7 32 84 7 39 13 7 28 36 8 30 23 8 30 19 8 27 39 9 30 30 68 9 24 29 9 24 35 10 29 27 10 22 6 10 23 61 11 23 74 11 21 30 11 23 65 12 23 93 12 13 14 12 23 68 13 20 103 13 22 34 14 19 19 108 Mean 44 (± 7.55 s.e.) 14 21 37 15 18 7 15 21 43 16 18 101 16 20 26 17 18 105 17 17 25 18 17 -2 30 18 17 33 19 15 11 19 17 41 20 14 94 20 15 62 21 14 99 21 15 63 22 13 98 22 15 64

72 SD 4: Evaluating small barrier removal to improve refuge connectivity

Goodga prioritisation rankings Mitchell prioritisation rankings Ludlow prioritisation rankings 23 12 12 110 23 15 66 24 10 106 24 15 67 25 9 4 109 25 10 22 26 8 33 27 8 96 Mean 23.16 (± 1.67 s.e.)

Appendix 2 (continued).

Goodga prioritisation rankings Mitchell prioritisation rankings Ludlow prioritisation rankings

28 7 100 29 5 112 30 2 107

Mean 24.1 (± 3.65 s.e.)

SD 4: Evaluating small barrier removal to improve refuge connectivity 73