Deterrence of Wild Waterfowl from Poultry Production Areas

Deterrence of wild waterfowl from poultry production areas: a critical review of current techniques and literature by Michael Atzeni, Darren Fielder and Bruce Thomson Deterrence of wild waterfowl from poultry production areas: a critical review of current techniques and literature

by Michael Atzeni, Darren Fielder and Bruce Thomson

January 2016

ISBN 978-1-74254-981-1 ISSN 1440-6845

Deterrence of wild waterfowl from poultry production areas: a critical review of current techniques and literature Publication No. 17/058 Project No. 009194

The information contained in this publication is intended for general use to assist public knowledge and discussion and to help improve the development of sustainable regions. You must not rely on any information contained in this publication without taking specialist advice relevant to your particular circumstances.

While reasonable care has been taken in preparing this publication to ensure that information is true and correct, the Commonwealth of Australia gives no assurance as to the accuracy of any information in this publication.

The Commonwealth of Australia, the Rural Industries Research and Development Corporation (RIRDC), the authors or contributors expressly disclaim, to the maximum extent permitted by law, all responsibility and liability to any person, arising directly or indirectly from any act or omission, or for any consequences of any such act or omission, made in reliance on the contents of this publication, whether or not caused by any negligence on the part of the Commonwealth of Australia, RIRDC, the authors or contributors.

The Commonwealth of Australia does not necessarily endorse the views in this publication.

This publication is copyright. Apart from any use as permitted under the Copyright Act 1968, all other rights are reserved. However, wide dissemination is encouraged. Requests and inquiries concerning reproduction and rights should be addressed to AgriFutures Australia Communications Team on 02 6923 6900.

Researcher Contact Details

Name: Michael Atzeni Phone: 0408155528 Email: [email protected]

In submitting this report, the researcher has agreed to AgriFutures Australia publishing this material in its edited form.

AgriFutures Australia Contact Details

Building 007, Tooma Way Charles Sturt University Locked Bag 588 Wagga Wagga NSW 2650

02 6923 6900 [email protected] www.agrifutures.com.au

Electronically published by AgriFutures Australia in January 2018 Print-on-demand by Union Offset Printing, Canberra at www.agrifutures.com.au or phone 1300 634 313

AgriFutures Australia is the new trading name for Rural Industries Research & Development Corporation (RIRDC), a statutory authority of the Federal Government established by the Primary Industries Research and Development Act 1989.

ii Foreword

Avian influenza is a constant threat to the Australian poultry industry. Aquatic birds are the main reservoir for low pathogenic avian influenza (LPAI) viruses and have been implicated in a number of highly pathogenic avian influenza (HPAI) outbreaks in commercial poultry flocks.

Various techniques have been used in other industries to deter and control waterfowl. Their application and effectiveness in the poultry industry to reduce the risk of AI transmission has not been properly investigated. This review looks critically at deterrents in the context of keeping waterfowl away from meat chicken farms. The review draws on the recent literature, various industry reports, and information from experts and manufacturers of products, and is for the benefit of the decision- makers in determining whether waterfowl control measures will be cost-effective in reducing the risk of LPAI and HPAI outbreaks, at least on some farms. The reported findings will also benefit producers looking for guidance on the best waterfowl management techniques.

This review was funded by RIRDC (now AgriFutures Australia). The report is an addition to AgriFutures Australia’s diverse range of over 2000 research publications and it forms part of our Chicken Meat Program RD&E Plan 2014-2019, which aims to increase the productivity and efficiency of chicken meat production.

Most of AgriFutures Australia’s publications are available for viewing, free downloading or purchasing online at www.agrifutures.com.au. Purchases can also be made by phoning 1300 634 313.

John Harvey Managing Director AgriFutures Australia

iii Acknowledgments

Thanks to the following people for their support, advice and information: Vivien Kite, RIRDC; Gary Sansom; QFF; John Tracey, NSW DPI; Tiggy Grillo, AWHN; Peter Scott, Scolexia; Tom Grimes; Rod Jenner; Geof Runge; Nina Kung, DAF; John Muehlebach; Colleen St Clair, University of Alberta; Bob Donely, UQ Gatton; Gary Bauer, UQ Gatton; Mark Callow; David Hedges; Brian Willmett and Des Ryan.

Abbreviations

AI Avian influenza

AIV Avian influenza virus

EAD Emergency Animal Disease

EADRA Emergency Animal Disease Response Agreement

FAO Food and Agricultural Organization of the United Nations

HPAI Highly pathogenic avian influenza

LPAI Low pathogenic avian influenza

iv Contents

Foreword ...... iii

Acknowledgments...... iv

Abbreviations ...... iv

Executive Summary ...... viii

Introduction ...... 1

Background ...... 1 Australian Surveillance ...... 1 Meat chicken production facilities ...... 2 Waterbirds on poultry farms ...... 3

Objectives ...... 4

Methodology ...... 4

Literature review ...... 4 Collation...... 4 Evaluation ...... 4

Results ...... 6

Global Avian Influenza situation ...... 6 Australian outbreaks ...... 6 High AI-risk species...... 6 Other potential vectors ...... 7 Waterfowl on meat chicken farms in the Lockyer Valley region, SEQ ...... 8 Australian waterfowl movements between countries ...... 8

Avian deterrent techniques ...... 9

Auditory deterrents ...... 9 Gas cannons ...... 9 Biosonics (bioacoustics)...... 9 Pyrotechnics ...... 10 Visual deterrents ...... 10 Effigies ...... 10 Falconry ...... 10 Drones and robots...... 10 Dogs...... 11

v Lasers ...... 11 Lights and reflectors ...... 11 Chemical repellents...... 12 Physical techniques ...... 12 Exclusion ...... 12 Decoy natural wetlands ...... 13 Habitat modification ...... 13 Lethal methods ...... 14 Waterbird management issues unique to poultry farms ...... 14 Integrated Bird Management ...... 14 Radar-activated deterrents ...... 15

Ranking deterrents ...... 16

Australian perspective ...... 17

Legislation ...... 17 Welfare and conservation ...... 18

Risk Assessment ...... 19

Factors contributing to risk ...... 19 Risk evaluation ...... 20 Cost–benefit analysis of deterrents ...... 21

Discussion ...... 23

Implications...... 24

Recommendations ...... 25

References ...... 26

Glossary ...... 31

Appendix A – Manufacturer contact details...... 32

Appendix B – Additional information ...... 35

Ball covers ...... 35 Drones and robotic birds ...... 35 Radar-activated deterrent systems ...... 35 National Avian Influenza Wild Bird (NAIWB) Surveillance Program ...... 35 Detection and ranging sensors ...... 35

Appendix C Waterfowl distributions ...... 36

vi Appendix D Waterfowl photos ...... 37

Australian Waterfowl – Selection of species featured in this review ...... 37

Tables

Table 1. Evaluation matrix applied to each identified deterrence measure...... 17

Table 2. Deterrent ranking for management of waterbirds on Australian meat chicken farms (1 is lowest; 4 is highest) ...... 17

Table 3. Approximate costs of various deterrents that could be used as part of a waterfowl management strategy on poultry farms. (Mention of specific products and brands does not imply endorsement by RIRDC or the authors)...... 22

Table C1. Australian waterfowl with extra-limital distributions outside Australia ...... 36

Figures

Figure 1 Aerial view of meat chicken farms in the Redland Bay area in Queensland, illustrating the density of farms and features of relevance to waterbirds. Existing poultry farms (A, f, g & h) along with farm dams (B), natural watercourses (C), adjacent properties and alternate land uses (D), green grassy areas where some waterfowl may potentially graze (E). (Data from Google Earth)...... 3

Figure 2. Generalised flow diagram for identifying the level of risk and appropriate deterrent systems...... 21

vii Executive Summary

What the report is about

The recent trend of increasing incidence of avian influenza (AI) outbreaks in Australia’s commercial poultry industry is of considerable concern (Fairbrother 1999; Scott 2014). Low pathogenic avian influenza (LPAI) viruses cycle naturally in wild birds, particularly in waterfowl (especially ducks and geese) and shorebirds. All commercial poultry in Australia are susceptible to AI infection as they are not vaccinated against AI and have no immunity. Exposed poultry become infected and begin spreading the virus through the flock. The LPAI virus may then mutate and manifest as highly pathogenic AI (HPAI) causing severe clinical disease with significant production losses and mortality in poultry flocks. The associated economic losses in dealing with HPAI outbreaks, including clean up and compensation costs, can be substantial.

Many Australian poultry farms have open water storages that attract waterfowl, and additional habitats such as grassed areas, pastures and crops which provide additional food sources for waterfowl and other potential AI vectors. Waterfowl can transmit LPAI viruses to poultry via direct physical contact and indirectly through faecal contamination of range areas and water supplies and runoff). There is also a risk of dust-borne and human-assisted AI transmission, for example, on clothing, footwear, vehicles and machinery. Another mechanism could be via wild terrestrial birds becoming bridge hosts between waterfowl and domestic poultry (Caron et al. 2014).

The Australian poultry industry has been fortunate to have contained the seven outbreaks of HPAI to date, all since 1976 and all due to Australian endemic H7 strains. The Asian H5N1 strain associated with human deaths overseas has not yet been detected in Australia but its introduction via migratory birds is always a possibility (East et al. 2008). A number of LPAI subtypes have been detected in wild birds in Australia. The H6 and H9 types detected in far north Queensland contain genomic sequences of Asian lineage suggesting migratory birds can, and have already, introduced overseas strains (Hoque et al. 2015).

In order to meet increasing consumer demand and to further develop a sustainable free-range sector, the Australian poultry industry needs to address the increased AI risk. In the short term, the industry as a whole need to establish what level of risk poultry farms are exposed to and whether there are economically viable, effective ways to deter waterfowl from poultry farms on an ongoing basis.

This review looks objectively at the advantages and disadvantages of various avian deterrent and control strategies for target species in target areas, day and night. The review has gathered information from the international and national literature, unpublished reports from government departments, research organisations and private industry as well as direct consultation with experts. It critically assesses avian deterrents and control methods used for waterfowl in various situations, such as tailing dams, airports, agriculture, aquaculture and oil spills, to determine those most suited to the commercial poultry industry.

Whilst evidence of AI has been detected in many endemic and migratory species, this report will focus on those Australian waterbird species commonly found on farms and therefore posing a potentially higher AI risk to poultry farms.

Who is the report targeted at?

This report on the review undertaken is aimed at those in the meat chicken industry responsible for reviewing policy and ensuring appropriate biosecurity measures are in place to minimise the chances of disease transmission from wild aquatic birds. The report may also be used to educate farm managers about waterfowl deterrence methods and strategies that are suitable for free-range farms,

viii farms that source their drinking water from surface water storages, and farms where waterfowl feed around production facilities.

Where are the relevant industries located in Australia?

The major centres of chicken meat production are near the capital cities in the eastern and southern states but becoming more regionalised with urbanisation. Often poultry farms are located in semi-rural settings with other land uses nearby that may also be primary production orientated. ABARES (2015) estimates 1.125 million tonnes of chicken meat will be produced in Australia in 2015. According to the Australian Chicken Meat Federation, the bulk of the production comes from NSW, Victoria and Queensland, although production growth is strong in SA.

All poultry farms are at risk of disease transmission via wild birds, particularly AI.

Background

The incidence of AI in poultry is on the rise worldwide (OIE 2015). Wild birds, particularly waterfowl and shorebirds, are the natural reservoir for AI viruses. The virus can be transmitted to domestic poultry through faecal contamination of water, housing and free-range areas, and on dust (Scott 2014).

In Australia, six more outbreaks of HPAI have been detected in commercial poultry farms since the first one in Victoria in 1976. The most recent outbreak was in 2013, with two layer farms infected in the Young district, NSW. Waterfowl have been implicated as the originating source of the virus in all Australian cases (Scott 2014; FAO 2015).

Current biosecurity protocols discourage meat chicken farmers from keeping domestic waterfowl and other birds, but many poultry farms have open water storages that attract waterbirds. Various waterbirds also utilise grassy areas, pastures, crops and agricultural land in the vicinity of chicken farms.

Strict requirements apply for the treatment of surface waters used for poultry drinking water to prevent this avenue of AI transmission. However, water treatment systems can fail and poultry producers make mistakes. Preventing waterbirds from utilising and potentially contaminating water supplies is problematic, without effective deterrent measures.

Aims/objectives

There are many deterrent products marketed for waterfowl and the manufacturers’ claims are often unsubstantiated. This review provides a strategic document that aims to help the Australian chicken meat industry identify appropriate waterfowl deterrents and strategies for reducing the risk of AI transmission to commercial flocks, thereby increasing the productivity and profitability of the industry for both the free-range and conventional farming sectors.

Poultry industry leaders, risk assessors, and producers seeking objective information about waterfowl deterrents and current best practice will benefit most from this review. The review also identifies key research and development areas for future funding. For example, researchers interested in waterfowl management and avian influenza will benefit from field trials that may be initiated as a result of this strategic review. The Australian community also stands to benefit through reduced AI outbreaks and associated costs. Specifically, this review aims to:

 critically review the literature on deterrents and management of birds from water bodies, range areas and other areas in the vicinity of chicken farms;  assess which species pose the most risk with consideration given to (a) ecology and behaviour (for example, habitat preferences, feeding habits), (b) distribution and (c) and likelihood of carrying AI viruses;

ix  gather anecdotal and unpublished information on the above, with an emphasis on Australian information;  provide direction for future trials of practical, cost-effective controls suited to Australian conditions; and  build on current published and grey literature knowledge including producing a scientific publication from this review.

Methods used

Information was collated from available literature, including grey literature, and from discussions with experts on pest bird management and waterbird management. The review team also consulted with people from the poultry industry, animal health, airport industry, wildlife regulation, deterrent manufacturers and deterrent users and other relevant persons.

Results/key findings

Published information about deterring waterfowl in the context of disease risk management on poultry farms is lacking, as is information about deterrence of Australian waterbirds in general. Consequently, the review draws on the extensive evidence and findings for waterfowl management for other industry applications (for example, airports and mining sites). This review also relied heavily on unpublished information and expert opinion from qualified persons in fauna management, and the authors’ knowledge and experience with Australian birds.

Netting range areas and covering water storages is generally cost prohibitive and for that reason, habitat modifications and passive and active deterrence measures would be necessary on most poultry farms to reduce the AI risk posed by wild birds.

Ensuring the property is unattractive to birds by minimising access to open water, food and roosts, should be a priority on any poultry farm, established or new (Scott, 2014; Grimes and Jackson, 2015). Proper management of production areas, particularly free-range areas, is essential (MacKenzie 2014). Eliminating non-essential water storages, including those of cooperative neighbours, reduces the number of resident waterbirds (Tom Grimes, pers. comm.). Alternatively, ponds could be covered to exclude birds.

To prevent habituation and effectiveness in the long-term, deterrence strategies must be based on sound biological and behavioural principles. Barring total exclusion from water storages and free- range areas, effective management of waterfowl will most likely require an integrated approach involving a number of deterrent techniques.

A combination of appropriate visual and acoustic deterrents which are activated only when waterfowl attempt to land in ‘no-go’ zones is suggested as the ideal strategy, as this targets the unwanted behaviour, minimises use of the deterrents, and helps to prevent habituation.

Some options to implement this approach include radar-activated systems (Nohara et al. 2011) which detect and respond to waterfowl incursions when coupled with deterrent systems that activate upon bird detection. Radar cannot distinguish similar species but can be used to distinguish birds of similar mass and behaviour. Such data can assist with understanding the timing and frequency of landing attempts, and evaluating the efficacy of the system. This method is limited somewhat by the radar technology which requires direct line of sight and works best in open flat areas.

A novel approach is to cover the water surface with weighted, floating balls to disguise it and effectively remove the attraction to waterfowl. This method has been developed commercially and may provide a long-term solution to minimising the AI risks associated with the contamination of

x drinking water and waterfowl utilising production areas. However, the balls have to be imported and are currently too expensive to consider.

Another practical and effective strategy in free-range areas and around water storages may be the use of patrol dogs such as a Maremma or other trained breeds (cattle dogs, sheep dogs) to chase waterfowl away. Dogs have been effective in the aviation industry when used as part of an integrated strategy and quickly learn to respond to the activation of other noise or visual deterrent systems. However, there are possible pathogen transmission issues with having dogs in production areas.

The use of solar- or battery-powered lights such as Fox Lights (a commercial product) to mimic moving flashlights was also identified as a possible method to deter waterbirds from landing on restricted areas at night.

The nomadic and unpredictable nature of waterbird movements in Australia, compared to the largely predictable, migratory habits of their northern hemisphere counterparts, and the size and layout of poultry farms, makes waterfowl detection and deterrence a challenging prospect on Australian poultry farms. On-farm observations and information-gathering are critical steps for assessing the AI risk and justifying the costs of waterfowl control. For instance, the species and numbers of waterfowl will depend on geographic location; size, number and nature of waterbodies in the area; food resources; time of year, and the prevailing climatic conditions. The AI risk will change accordingly.

Reducing the incidence of HPAI in Australia would save the Australian poultry industry considerable expense. HPAI is a problem ubiquitous across the commercial poultry industry and therefore needs national cooperation across the entire industry to minimise HPAI outbreaks and associated costs to industry.

Having producers put deterrent systems in place to reduce the risk of AI transmission will provide a vehicle for informed and helpful discussions on ‘best practice’ for different species and situations. In the event of a serious HPAI outbreak, proven integrated deterrent measures can be implemented quickly.

To that end, conducting trials to evaluate integrated deterrent systems for different production systems and areas is prudent.

Implications for relevant stakeholders

This review has highlighted the importance for policy makers to conduct cost–benefit analysis before deciding the level of commitment required to deter waterfowl from farms.

A cost–benefit analysis may determine that the best approach is to increase the financial commitment from government, industry and individual producers in the event of an AI outbreak. This may prove to be more economically feasible than implementing waterfowl management plans. Otherwise, the implications are:

 Enforcing stricter biosecurity measures regarding waterfowl deterrence for meat chicken farms may reduce the AI risk, in some cases considerably.

 Utilising low-maintenance, but effective, waterfowl deterrent strategies to reduce the risk of AI transmission.

 Having to assess the cost effectiveness of each strategy on a farm-by-farm basis because the AI risk is species-dependent and influenced by farm-specific factors.

xi  Developing an industry-agreed risk assessment methodology is required that factors in the measures taken to deter waterbirds under different production systems.

 Farm-based trials are necessary to determine efficacy of particular deterrent strategies before uptake can be recommended with confidence. Such field trials will need to effectively monitor ‘no go’ zones for extended periods and to identify the effect on target species.

Recommendations

The following recommendations are provided to guide future research and investment:

Detailed evaluation of key strategies

Based on this review, a more detailed comparative evaluation of the identified key deterrence strategies is warranted. The assessment will need to consider cost–benefit of each strategy.

Field trials

 Following a more detailed evaluation, field trials of the more promising strategies should be undertaken. Field trials would need to take into account temporal (daily, seasonal) and spatial variation in waterbird patterns and behaviour. Aims of such field trials would include:

 Developing a practical, cost-effective 24/7 detection and ranging system to detect likely waterbird activity on-farm, particularly, nocturnal activity in ‘no go’ zones. This would serve as a data collection system to evaluate efficacy and could become part of an automated deterrent system.

 Trialling a radar-activated on-demand deterrent system to establish the benefits of a more sophisticated, automated system to prevent habituation.

 Trialling a range of deterrent strategies across the categories of auditory, visual and physical devices, including habitat modifications.

Risk assessment

 Providing producers with a risk assessment management tool so that every meat chicken farm is assessed for AI risk by the owners.

 Surveying and characterising Australian meat chicken farms and ranking them according to their AI-risk, both individually and regionally.

 Educating producers about Australian waterfowl and encouraging the recording of their on-farm behaviour to assist risk assessment.

Economic modelling

 Conducting bio-economic modelling over a sufficient planning horizon (say 30 years) to demonstrate the merit or otherwise of investing in waterfowl deterrence strategies, taking into account industry expansion, various outbreak scenarios, and new findings about AI epidemiology.

xii Introduction

Background

Wild birds, particularly waterfowl, are the natural reservoir of low-pathogenic avian influenza (LPAI) viruses (Alexander 2000). When LPAI viruses are transmitted to susceptible poultry flocks, highly pathogenic avian influenza (HPAI) can develop, causing high morbidity and mortality in poultry. This can lead to substantial economic losses for the poultry industry and impact markets and trade.

The history, research and latest developments on HPAI have been summarised by FAO (2015). In brief, all HPAI viruses detected in poultry have belonged to the H5 and H7 haemagglutinin subtypes (FAO 2015). Until the emergence of the Asian-lineage HPAI H5N1 strains in 1996, wild birds had not been considered a primary source of HPAI viruses. The rapid spread of H5N1 strains to Europe and Africa changed this view and points to wild birds being involved in directly spreading the H5N1 HPAI over long distances. Although the evidence is still largely circumstantial, the potential spread of HPAI viruses to other countries via migratory birds is a concern to health authorities worldwide.

Australian Surveillance

The epidemiology of AI in Australia remains poorly understood despite significant contributions in this area by several researchers (Hansbro et al. 2010; Haynes et al. 2009; Klaassen et al. 2011; Tracey 2005, 2010; Tracey et al. 2004). Early detection is seen as critical. In Australia, surveillance for AI in wild birds is conducted under the National Avian Influenza Wild Bird (NAIWB) Surveillance Program which aims to better understand the role of waterbirds in transmission of AI viruses, to assess the risk to humans and production industries (Wildlife Health Australia 2015; Grillo 2015).

The NAIWB Surveillance Program activities are conducted Australia-wide. The Program has two main components: targeted surveillance and general surveillance. Targeted activities focus on sampling from waterfowl (Anseriformes) and a smaller number of shorebirds (Charadriiformes) at locations where they occur together, as well as at locations that bring waterfowl into close proximity to poultry and humans. General surveillance focuses on exclusion of AI from mass mortality and morbidity events in wild birds around Australia and the Australian Antarctic Territory. Between July 2005 and June 2014, over 81,000 wild birds were tested for influenza viruses.

To date, no HPAI viruses have been identified in Australian wild birds. However, there is evidence of infection with LPAI viruses of unique Australian lineages (Hansbro et al. 2010). While the risk of passing on LPAI to domestic flocks in Australia appears low, a dramatic recent increase in positive returns for tested waterfowl suggests that a clear potential exists. The documented increase is based on a sample of less than 4 per cent of the wild bird population (Haynes et al. 2009, Grillo 2015).

Under the Emergency Animal Disease Response Agreement (EADRA) cost-sharing arrangements have been made between the states and federal governments for dealing with avian influenza H5 and H7 viruses (both LPAI and HPAI) in poultry.

The onus is on the poultry industry to ensure AI biosecurity measures are adequate (Fairbrother 1999; Scott 2014). Under the EADRA, the expectation is that industry will actively work on reducing the AI risk. This agreement serves to drive the poultry industry in determining if there are effective, economical methods of deterring waterfowl, particularly high AI-risk species that frequent poultry farms.

There are numerous techniques and types of equipment used for bird control. Much of what is published is scattered and difficult to locate and is not specific to the poultry industry. Waterfowl

deterrence in proximity to commercial poultry farms presents a unique set of challenges for bird control experts with a knowledge gap in applied understanding of the efficacy of most methods. Ethical and conservation issues may also arise in Australia. For example, scaring water birds during drought periods may result in stress on the birds and the use of lethal methods on species that are considered threatened with extinction by State and Commonwealth Government agencies is prohibited. Legislative and policy frameworks are an important consideration and will vary with each region.

Without critically reviewing bird control products and techniques in the context of their practicality on poultry farms, inappropriate strategies may be adopted, jeopardising future efforts.

Meat chicken production facilities

Australian meat chickens are reared in batches, usually for 6-8 weeks. Some of the birds are harvested early, often around day 35. Housing flocks in tunnel ventilated sheds is the conventional way to rear meat chickens, but the growing market for free-range chicken products necessitates having a free- range area alongside the shed which the birds are free to access for part of the day, at least from an age that they are adequately feathered. Free range areas are usually about 1.5 to 2 times the size of the shed’s footprint (V. Kite, pers. comm.). Birds from different sheds are not allowed to share a common free-range area because of disease risks.

Whilst there are valid reasons for not allowing open water storages on farms in the future, it is not unusual for meat chicken farms to have one or more dams on-site. The dams may be required for production (drinking water; cooling), other livestock, irrigation (of range areas for example), for containment of run-off or for recreational purposes. Some farms have watercourses, irrigation channels or natural wetlands either on them or nearby. Most farms have open areas on or surrounding them. Some have natural bushland all around; others have pasture and cropping land. Some grow turf or border turf farms. All have silos where grain could potentially be spilt and attract wild birds.

All these features are potential habitat and food resources for waterbirds and the management of these features will be required in order to deter waterbirds.

Figure 1 shows a cluster of four poultry farms in Redland Bay, Queensland. This map illustrates the variability between farms in terms of their surface water habitats in relation to sheds and range areas. In this type of situation a co-ordinated approach may be necessary between farms to reduce the risk of AI significantly for the area.

Waterbirds on poultry farms

In the course of a year, several species of water birds could be found on poultry farms depending on the location and resources available. Any one of these species could carry LPAI virus but it is their observed or likely behaviour that determines the risk to poultry, not their mere presence.

The Anatids (ducks, geese and swans), and in particular commonly encountered dabbling and grazing species like Pacific Black Ducks (highly susceptible), Grey Teal and Australian Wood Ducks are the more obvious AI-risk species to focus on (Haynes et al. 2009; Hoque et al. 2015). In contrast to the predictable migratory patterns of northern hemisphere waterfowl, the patterns of waterfowl movement in Australia are largely nomadic, so far less predictable (Tracey et al. 2004). On any given day, there could be resident or visiting target and non-target species on-site at poultry farms, so any deterrent and control system would, ideally, have to be selective, operating all year round on an automated basis.

Objectives

The project objectives were to:

 review the literature on deterrents and management of birds, focussing on the more practical and cost-effective methods;

 gather anecdotal and unpublished information, with an emphasis on Australian information;

 provide direction for future trials into practical, cost-effective controls suited to Australian conditions; and

 build on current published and grey literature knowledge including producing a scientific publication from this review. Methodology

Literature review

Available international and national peer-reviewed and grey literature on bird management across a wide range of species and applications was collated. This includes published reviews and reports of bird deterrents and controls for the aquaculture industry (Gorenzel et al. 1994), aviation industry (Bishop et al. 2003) and horticulture industry (Tracey et al. 2007). Based on the information collected, the review focussed on deterrents and management practices relevant to controlling the high AI-risk species in and around poultry production facilities.

Species were considered high AI-risk if they were known to be highly susceptible to LPAI infection, based on AIV surveillance data (for example, Haynes et al. 2009; Hoque et al. 2015), or they commonly associated with high-risk species and therefore were potential secondary vectors to poultry farm dams, or potential bridge species to poultry production areas (Klaassen et al. 2011; Caron et al. 2015).

Collation

The review involved on-line searches using various key word combinations, accessing library collections and acquiring literature and articles from individuals with expertise in this area. Downloaded material included documents and Powerpoint presentations. Website addresses of interest were noted as were contact details for various people and organisations.

Key words included but were not limited to: scaring, hazing, controlling, managing, deterring birds, ducks, geese, swans, waterfowl, wildlife, tailing dams, airports, birdstrike, farms, poultry, horticulture, aquaculture, bird netting, bioacoustics, repellents. For ecology and ethology (behavioural studies) of the key species, searches were conducted on-line and key library references were consulted (for example, Marchant and Higgins 1990; Pizzey and Knight 2007).

Evaluation

Several substantial reports have been completed over the past 13 years reviewing avian deterrent methods (including Bishop et al. 2003, Tracey et al. 2007; Belant and Martin 2011). These reviews provided a context for waterbird and wild bird management in other situations (aquaculture, airports,

horticulture, mine tailings dams, oil spills, urban settings). From these and related publications, the review team critically evaluated how deterrence and control measures might be used in the context of meat chicken farms and reducing the AI risk.

While biosecurity is important for reducing the risk of AI, the welfare of the chickens and reputation of the meat chicken industry are paramount. Any deterrents that may unduly frighten, stress or harm the chickens, interfere with normal production, annoy neighbours or negatively impact on the industry and marketing of meat chicken products will result in negative outcomes have consequences for industry profitability.

Another important aspect to the success of any deterrent system is that they will need to be properly installed, managed and maintained for the long-term. Thus ease of use is another key consideration.

Deterrent and control methods were evaluated in terms of effectiveness, sustainability, practicality, cost, maintenance, chicken welfare and industry reputation. For economic evaluation and risk analysis, the type of production system (free-range versus fully housed) and waterfowl habitat types (dams, natural wetlands, pasture) were taken into account. Costs of deterrents and controls were standardised where possible on a per unit basis, for example, per square metre for range area, pond surface area and cultivation area.

Results

Global Avian Influenza situation

Despite outbreaks of HPAI having occurred in many countries, the epidemiology of AI is still poorly understood (OIE 2015). Food and Agriculture Organization (FAO) has provided specific guidelines for the poultry industries in dealing with AI outbreaks and minimising the risk of further infection of both domestic and wild birds (FAO 2009). This includes reducing any attractants such as food and where necessary, increasing wild bird scaring tactics on and around infected farms.

In Australia, waterfowl movements and concentrations are largely linked to the prevailing environmental conditions (Tracey et al. 2004) and so these polices are particularly relevant in times of drought when waterfowl tend to move into coastal areas and semi-rural settings. A HPAI epidemic could be devastating, not only for the poultry industry, but also the wild bird populations already succumbing to the impacts of drought. The stress on wild bird populations as a result of scaring could potentially spread disease.

Other practices, such as rice growing, have the effect of concentrating ducks, so an AI outbreak in the vicinity of rice producing areas (for example, NSW Riverina) must also be treated judiciously, with deterring practices in either industry potentially spreading the virus to other farms or sectors.

Australian outbreaks

Australia has had seven outbreaks of HPAI strains in commercial chicken flocks. Fortunately, all have been stamped out quickly, including the two related outbreaks on NSW egg farms in October 2013 (Scott 2014). Previous outbreaks were in 1976, 1985 and 1992 (all Victoria), 1994 (Queensland) and 1997 (NSW). The economic costs of outbreaks can be substantial. For example, the 1997 outbreak resulted in the destruction of a total of 310,565 chickens, 1.2M fertile chicken eggs, 261 emu chicks, and 147 emu eggs, at a total cost of $4.5M, of which $2.2M was compensation for destroyed stock (Selleck et al. 2003). The 2013 NSW outbreaks were estimated to cost $5M in total (V. Kite, pers. comm.).

High AI-risk species

The relative prevalence of LPAI in Australian wild birds, particularly in native ducks and migratory waders, has been gauged in recent years through the ongoing NAIWB Surveillance Program (Grillo 2015; Haynes et al. 2009; Tracey 2010). Dabbling ducks, particularly the genus Anas, are the major source of LPAI and, arguably, the main risk to poultry, though the evidence is largely implied (Haynes et al. 2009; Hoque et al. 2015).

All manner of transmission scenarios and bird species are possible. Theoretically, any bird, animal or human in contact with AIV could subsequently contaminate a poultry production area. In the case of HPAI outbreaks attributed to failed drinking water treatment systems, it is likely that dabbling ducks were responsible for contaminating the water supply, but it is virtually impossible to prove the source (FAO 2015).

Based on distribution, habitat preferences, species interactions and behaviour, as well as discussions with some producers and industry experts and personal experience of the authors’, the species listed below are considered the higher AI-risk species for the Australian poultry industry. It must also be considered that as development further encroaches on natural wetlands and our climate varies (either due to climate change or shorter term fluctuations), the distribution of waterfowl may change, and so

this list may require regular review to keep it relevant. A series of images of these and some other species is provided in Appendix D.

 Australian Wood Duck (Chenonetta jubata) – a widespread and familiar duck in many parks and gardens that quickly habituates to humans. The duck is a grazer likely to be found on ranges and pasture areas adjacent to chicken sheds.

 Plumed Whistling-Duck (Dendrocygna cytoni) – a gregarious species that can gather in huge flocks. They mainly camp during the day around waterholes, including farm dams, then fly out at dusk to graze on green grasses and seeds, returning at dawn. In rural areas, they regularly associate with backyard poultry, taking advantage of the free feed (personal observations). This species is likely to be found in range areas or and pasture areas adjacent to chicken sheds. Pacific Black Duck (Anas superciliosa) – a widespread, common species that can utilise all sorts of water habitats, no matter how small, including pools, drains, farm dams, creeks and channels. They are also attracted to flooded or irrigated crops and pastures. Crops can become an important source of food in dry years.

 Grey Teal (Anas gracilis) – the most widespread Australian waterfowl, responding rapidly to inland flooding, and can occupy almost any water habitat. Mainly feeds in shallow open water and marginal vegetation but readily uses agricultural habitats including farm dams, flooded or irrigated pasture, grassland, crops, roadside ditches and irrigation channels.

 Chestnut Teal (Anas castanea) – more coastal than the Grey Teal, preferring saline habitats, but also found inland and utilises freshwater wetlands including farm dams.

 Straw-necked Ibis (Threskiornis spinicollis) – a gregarious species typically found feeding on wet or dry grasslands and cultivated land. Also feeds in shallows or around the margins of wetlands, mainly freshwater. Flocks roosts together in stands of trees near water.

 Australian White Ibis (Threskiornis molucca) – a versatile, widespread species that uses a wide variety of natural and artificial wetland habitats, as well as wet grasslands or agricultural land, and open areas of sparse or low vegetation. This species is more likely to be found in the company of ducks around the margins of waterbodies and to transfer duck faeces. It is also a communal rooster.

 Cattle Egret (Ardea ibis) – a sociable small egret, which feeds mostly in paddocks and pastures in company with cattle, horses and other domestic stock. It opportunistically searches for prey disturbed by the grazing animals. It regularly congregates with other waterfowl on loafing areas and in the breeding season, forms large heronries.

Other potential vectors

Waterfowl intermingle and some species like Yellow-billed Spoonbills that favour smaller dams, may visit a number of water habitats in a single day, particularly, in drought conditions. It is possible that species other than ducks could become contaminated superficially off-farm and transfer the virus on- farm, subsequently infecting resident waterfowl. They could even be responsible for AI transmission to poultry flocks (Klaassen et al. 2011).

The following species also deserve consideration as possible vectors:

 Cormorants – three species of cormorants may occur on farm dams, particularly the Little Pied Cormorant which feeds more on crustaceans than fish; and Great and Little Black Cormorants, which prefer larger dams. Dams without fish and crustaceans are unlikely to attract these birds.

 Other egrets – three other species (Great, Intermediate, Little) feed mainly in shallow waters so the shallows of large farm dams, and ephemeral pools following heavy rain.

 Rufous Night Heron – mainly nocturnal; roosts by day and feeds by night on fish, frogs, freshwater crayfish and aquatic insects. Also recorded eating house mice, and feeding in human refuse

 Spoonbills – both species (Yellow-billed and Royal) can potentially occur in shallow areas of farm dams, the Royal preferring larger water bodies.

 Lapwings – the Masked Lapwing in particular, is a familiar bird that prefers open areas but also feeds around the edges of dams.

 Dotterels – two endemic species of dotterel (Red-kneed and Black-fronted) are found around the edges of dams, the former preferring larger water bodies.

 Rails – these species require suitable dense vegetation for cover. The larger species (swamphens, moorhens) may graze on grass and utilise temporary surface waters.

 Latham’s Snipe – this cryptic, migratory wader skulks in vegetation cover along the edge of dams, creeks and drainage channels, feeding in the shallows.

Waterfowl on meat chicken farms in the Lockyer Valley region, SEQ

The Lockyer Valley, east of Toowoomba, SEQ, is abundant in waterbird habitats and is also one of the premier birding areas in Australia (well-known nationally and internationally). Waterbird habitats include lagoons, farm dams, large water storages (lakes and dams), creeks, sewage treatment ponds and marshlands.

Three decades of observation and surveillance of birds in the Lockyer Valley has given the review authors an appreciation of the diversity, behaviour, habitat preferences and transient nature of most waterfowl in the district. Around the Gatton region, the following waterfowl species were noted on three meat chicken farms in 2013: Australian Wood Duck, Pacific Black Duck, Pink-eared Duck, Hardhead, Grey Teal, Chestnut Teal (uncommon away from coast), Eurasian Coot, and Australasian Grebe. Plumed Whistling-ducks and Dusky Moorhens were within 100 meters on a neighbouring property. These species were on farm dams and did not come near the sheds. The only species observed walking around the sheds were Brown Quail at one farm and Masked Lapwing at another.

Australian Wood Duck and Pacific Black Duck tend to have greater fidelity to a site and therefore are more likely to habituate quickly to any deterrents. What is not clear is how their behaviour influences the more transient species, but it is likely that sedentary populations attract other birds, since from a behavioural perspective, this provides advantages of greater vigilance against predators (’more eyes watching’) and also suggests to nomadic species that the area is relatively safe and productive. Thus, from a bird control perspective, it is best not to allow sedentary populations to establish, which could potentially attract other species, which are more likely to transmit disease.

Australian waterfowl movements between countries

Although genomic evidence is lacking, Australian waterfowl species with extra-limital distributions in Asia and beyond may move across their full geographic range and introduce new viruses directly or indirectly to Australia (Appendix C provides an indicative list of waterfowl species).

Avian deterrent techniques

Deterrence can be done both passively and actively, and research suggests that a combination of both measures may be needed to be effective. Passive deterrence for waterbirds in the context of poultry farms, defines various measures to ensure that the production area and surrounds are unattractive for waterfowl, and therefore do not encourage waterfowl to congregate. Active control strategies target waterfowl that are attracted to a site and aim to eliminate or reduce congregation with scaring tactics. Active control is all about outsmarting unhabituated birds so that they move on. For the purposes of this review, there are five main groups of deterrents discussed: auditory, visual, chemical, physical and lethal.

Auditory deterrents

There are various auditory deterrents available for scaring all types of birds. Habituation often becomes problematic for these methods when used indiscriminately or in isolation (Gilsdorf et al. 2002).

Gas cannons

Gas exploders using gas/propane cannons provide a temporary means of control, however, they are quickly subject to habituation by birds (Gilsdorf et al. 2002). Despite being louder than a 12-gauge shotgun, the exploders should be moved around to delay habituation for as long as possible. Various levels of success are reported, from unsuccessful (even when combined with lethal removal) at JFK Airport (Washburn et al. 2006), to highly successful in displacing Double-crested Cormorants from roosts at night (Mott et al. 1998).

Using a recording of a gas gun to deter night herons was ineffective, with habituation starting by the second night of usage (Spanier 1980). Randomly fired gas cannons used to deter Magpie Geese from grazing in young ryegrass pasture in the Lockyer Valley, southeast Queensland, were effective for less than a month (M. Bauer, personal communication). Furthermore, these geese have established a large permanent local population where they were once only seasonal.

Biosonics (bioacoustics)

Biosonics involves the playback of acoustic signals that a species uses to communicate to conspecifics. The calls are typically distress and alarm signals. They have shown promise in many short trials (for example, Jaremovic 1990) but their long-term efficacy is hard to demonstrate in practice because of confounding influences. For example, success (no habituation after many weeks of use each night) was achieved with night herons at fish-rearing ponds by using playback of natural distress calls (Spanier 1980). However, the birds had several alternative food sources available.

Mott and Timbrook (1988) measured a 71% decrease in Canada Geese (and 96% reduction when bioacoustics were coupled with pyrotechnics). Underwater recordings of boat chase noises reduced eider duck numbers in the absence of workers, but to work effectively birds had to be chased by the same boat semi-regularly (Ross and Furness 2000).

Ronconi and St. Clair (2006) used on-demand systems for waterfowl deterrence at tailings ponds and recommended that gas cannons be used in preference to effigies (visual). They suggested that deterrent techniques at their study site (oil sands) should (i) be only used when tailings ponds appear to be most attractive to migrating waterfowl, (ii) target lower flying waterfowl and shorebirds, and (iii) the technique used needed to be effective during both day and night.

In their review of sonic deterrents, Bomford and O’Brien (1990) found mixed reactions by birds, and pointed out that determining efficacy would always be problematic because experiments are difficult to control and replicate. They concluded that sonic devices provide short-term relief, at best.

Pyrotechnics

Pyrotechnics capitalise on explosive noises coupled with visual effects (bright lights and smoke) to frighten birds away. The deployment of pyrotechnics is manual and generally not cost-effective. Using rifles, shotguns and flare pistols with shell crackers bird bombs, bird whistlers and similar deterrents is possibly a better option.

Visual deterrents

Visual deterrents rely on their novelty, a startling effect or danger association. Danger associations are the most common i.e., a real predator (for example, falcon, hunter, dog), a simulated predator (effigies), predator victim or a suspicious object.

Effigies

Effigy effectiveness depends on how life-like they are, how judiciously they are used, and how attached the birds might already be to the site (e.g. resting, breeding) (Gilsdorf et al. 2002). While birds may be wary initially, they tend to habituate quickly to effigies. Night herons started to become habituated to Scarey Man® (commercial product) after four nights at a trout hatchery for instance (Andelt et al. 1997). In contrast, Boag and Lewin (1980) found a human effigy worked well on dabbling and diving ducks on small natural ponds (95% reduction; but suggest that further testing is required). Seamans and Bernhardt (2004) used dead Canadian Geese effigies to little effect with adult pairs habituating after a period of several days.

Predator models such as hawk-kites and mounted hawks and owls have short-term efficacy (Tracey et al. 2007). The more animated the better and these are best used as part of an integrated management system. A common misconception is that birds will always react to a raptor shape, when in fact birds are very in tune with which raptors pose a real threat and don’t react at all to others that do not. In Australia, only the Peregrine Falcon and the larger Australian species such as Swamp Harrier, Wedge- tailed Eagle, White-bellied Sea-Eagle and the Red Goshawk predate waterfowl and cause enough havoc for waterfowl to take flight en masse. It may be worthwhile using these species as the basis of more realistic effigies for use in Australia.

Falconry

Using falconry to disperse birds is generally impractical as it is requires trained birds, experienced handlers, and favourable daytime conditions. Other hazing techniques have proven more effective and economical. In their review of falconry as a bird-hazing technique, Erickson et al. (1990) concluded its use is limited to special situations such as at airports prone to bird aircraft strikes. Trials conducted mostly in UK using Peregrine Falcons, Gyrfalcons and Goshawks helped increase public acceptance of airport bird management programs (Dolbeer 1998). A number of European airports deploy trained raptors in their bird management strategies (Kitowski 2011). In any case, falconry is not widely available for use in Australia.

Drones and robots

Drones are remotely controlled unmanned aerial vehicles (UAVs) that are usually fitted with cameras and sensors to collect data. The increasing affordability and sophistication of drones has resulted in a

myriad of applications. They include bird applications to non-invasively survey species, and at the other extreme, harass and scare nuisance birds such as gulls, ducks and geese away from airports, beaches, parks and other public places.

Ornithologists report that wild birds are generally unperturbed by drones used for surveillance purposes, soon learning they pose no real threat (St Fleur 2015). Nevertheless, there is a critical distance at which any species will eventually take flight making drones a potentially useful tool in future, if the process of waterfowl detection and sending the drone to the detection area can be automated. For now, the Civil Aviation Safety Authority rules in Australia are too restrictive for UAVs to be launched and operated unattended, unsighted and at night. But the concept could be demonstrated with a UAV in any case, and implemented using terrestrial and amphibious robots, which may have application on poultry farms and dams.

The use of full-scale, bird of prey reproductions has proven a more practical and effective alternative to falconry at airports (Battistoni et al. 2008), particularly for dispersing large flocks. Predator-like drones combining visual and bio-acoustic scare tactics are being marketed for ‘stubborn species’ but their efficacy has not been demonstrated. Progress in developing drones with flapping wings has been made to mimic flight behaviour of raptors better (Jeffrey 2014). Indeed, the falcon and eagle Robirds developed by ClearFlight Solutions (http://clearflightsolutions.com/methods/robirds ) (for their own purposes) are very realistic in both appearance and flight behaviour and the video demonstrations on their website are impressive. However, the long-term efficacy has not yet been demonstrated and the requirement for a manual operator precludes their general use.

Dogs

Using trained working dogs has been generally effective. For example, border collies have been used to haze geese at airports (for example, Castelli and Sleggs 2000). Aviation bird management experts continue to use dogs in integrated bird management. Some Australian poultry producers use guard dogs (for example, Merremas) in free-range farms to chase wild birds and predators (G. Runge, pers. comm.). Merremas are bred to protect flocks and are a low maintenance breed. They represent a very good option for free-range farms provided they can be trained to protect the range even when the flock is indoors. Disadvantages to using these dogs are their barking, which can be annoying for nearby neighbours, and they are a potential risk for disease and pathogen transmission.

Lasers

Lasers elicit a variable response across species depending on the wavelength and transmitted light. They have gathered popularity in the airport industry as a short-term scare tactic. Field trials dispersing Double-crested Cormorants at night with lasers was successful (> 90% dispersal) but inconclusive (Glahn et al. 2000). In another application, laser lights were used successfully to deter eider ducks away from mussel farms (Ross and Furness 2000). The negatives identified in this trial were that the laser was considered expensive, labour intensive, not effective in daylight, and potentially harmful if shone in the eyes.

Lights and reflectors

In an Australian trial investigating toxic tailings dams, floating solar-powered beacons that rotated and were only activated intermittently reduced total waterfowl abundance by >90% in the short-term (Read 1999). The Masked Lapwings, Black-fronted Dotterels and Red-necked Avocets around the edge of the ponds were undeterred. Hoary-headed Grebes and Australian Grebes dived when the beacon came on but were not scared away from the pond by the technique.

In other trials reported in the literature, reflector tapes and flags were of limited use, for example, on Brent Geese in wheat (Summers and Hillmann 1990) and Herring Gulls (Belant and Ickes 1997).

Fox Lights, a commercial product said to ‘create a human-like presence in the paddock’, have been used successfully in the Australian rice industry to deter ducks (I. Whalan, pers. comm.; J. Bradford, pers. comm.) and may present a cost-effective solution on range areas, crops and pastures at night. They mimic a bright flashlight being shone in different directions and are battery-powered. However, they remain untested in terms of field trials for the poultry industry, and may be unsuitable for use in peri-urban areas, where many chicken farms are located.

Chemical repellents

The efficacy of chemical repellents on birds has been highly variable over a wide range of settings and is dependent on the chemical used, species and resource targeted. There are primary repellents, designed to elicit an immediate aversive reaction due to smell, taste, colour or irritation; and secondary repellents, intended for ingestion at sub-lethal doses to cause a distress response and an aversion to the food source (Tracey et al. 2007). Intake of secondary repellents is difficult to control and can be debilitating and sometimes fatal to the birds.

Foliar chemical sprays combined with plant growth regulator have been used to repel geese off turf grass (Blackwell et al. 1999). In a trial in Canada, methyl anthranilate (MA) was applied to grain. MA occurs naturally in some fruits and flowers and is an irritant to the trigeminal nervous system of birds. Initially, Mallards and Canada Geese strongly avoided the dosed grain but they eventually became habituated to it (Cummings et al. 1992). In the US, a new MA-based product called Avian Control® Bird Repellent is longer-lasting. The manufacturers claim it is effective on all pest birds and harmless to other animals and humans. The irritant causes temporary stress and discomfort that abates once the bird leaves the treated area.

Australian tests on seventeen chemical repellents to protect tropical fruit crops and vegetables yielded generally poor repellence (Marcsik and Clarke 1997). Bishop et al. (2003) concluded that chemical techniques are generally less effective in the field than in laboratory and cage trials, and are relatively time-consuming to implement and expensive.

In the poultry industry, chemical repellents are unlikely to be used in poultry production areas in case the poultry are adversely affected or there are chemical residue issues.

Physical techniques

Use of physical barriers and physical changes to the landscape can either prevent access to restricted areas or make them less attractive or no longer suitable for waterfowl.

Exclusion

Exclusion netting has been found to be highly effective and the best approach for protecting orchards, vineyards and fish farms where it is economically beneficial (Marcsik and Clarke 1997; Bomford and Sinclair 2002, Tracey et al. 2007). Full canopy orchard netting to exclude flying foxes costs between $23 000 and $72 000 per hectare (Rigden 2008). Therefore, the permanent exclusion of birds from large areas is generally cost-prohibitive. Exclusion netting is also logistically difficult to install on many established poultry farms and most likely, too impractical or restrictive for meat chicken farm operations. However, it may have some utility if used to cover water bodies, range areas or other areas used by waterfowl. This is also discussed below under ‘habitat modification’.

Decoy natural wetlands

The use of decoy natural wetlands for attracting and retaining native waterbirds away from the risk areas in and around poultry farms is worthy of attention in the context of reducing the risk of AI outbreaks. For this option to be viable, the natural wetland would need to be located some distance from the poultry farm and should be managed to provide available habitat for waterbirds, the idea being that making the natural water body more attractive, and a sanctuary for deterred waterbirds, will result in a decreased number and diversity of species frequenting the artificial farm storages. This strategy remains untested, and indeed may have the opposite effect, attracting birds to the general region in which farms are located. Further research and evidence-based results would be needed to confirm whether this approach has any value as a management option.

Habitat modification

Much research on passive control has focussed on habitat manipulation to reduce the attractiveness of crops to birds or to make alternative food sources more attractive, although there has been little adoption of these approaches by growers (Bomford and Sinclair 2002; Tracey et al. 2007).

Habitat manipulation has been more successful at airports where the intent is always to make it as undesirable as possible for birds, to prevent birdstrikes. The Australian Transport Safety Bureau guidelines for reducing the attraction of airports to ducks, particularly Pacific Black Duck and Australian Wood Duck, involves limiting their food supply, for example, preventing grass from seeding by mowing at appropriate times, maintaining grass at around 30 cm in non-essential areas, and reducing the attraction of water in lakes, ponds, creeks and drains on and around the airport (see www.atsb.gov.au). Advantages and disadvantages of several methods are listed on their website.

Ducks habituate quickly to inexpensive options such as flagging tape and humming wire (deterrents). Placing a 2 to 6 m wire grid over watercourses can also be ineffectual (unsuccessful in the UK and at the Gold Coast airport) and could lead to wing injuries if birds on the water take off when spooked. Netting (19-50 mm) placed over ponds is highly effective but is expensive to install, requires regular maintenance and can snare smaller birds if the mesh size is too small or if the net is not taut.

Generally, the larger and shallower the waterbird habitat, the more birds and species it attracts. Each species or group of species has its own habitat preferences. Modifying the habitat attractive to one species may well provide better conditions for another. Nonetheless, the modification of edges and large areas of suitable habitat and the removal of particularly attractive features, such as peninsulas and islands, is good practice. Long grass around waterbodies is generally a deterrent to ducks for roosting because of increased opportunities for concealment of predators.

Draining and filling in of water storages in and around poultry farms, where practical, to disperse resident waterfowl and permanently remove the water attraction has had the desired affect at a previously HPAI-infected site (Tom Grimes, pers. comm.). Negotiation with and the cooperation of neighbours with non-essential dams may also be necessary to achieve the desired outcome. Ideally, there should be no waterbird habitats in the vicinity of free-range sheds (MacKenzie 2014). It has been proposed new free-range farms should be located away from significant waterbodies such as dams, rivers and lakes.

Removal of waterbird habitats and food resources may not be feasible or prudent, but minimising temporary surface water, improving drainage systems, and managing vegetation and grass are key strategies that all poultry farms could implement to reduce their AI-risk.

Plastic flotation balls made of high-density polyethelene (HDPE) (for example, Bird Balls™, Armor Balls™; Appendix B), from here on referred to as bird balls, have been used extensively at overseas airports (Harris and Davis 1998; Belant and Martin 2011, Figure 9) to exclude waterfowl from water

storages. The balls are tipped in and form a ‘carpet’, disguising the surface. The advantages over nets are that bird balls do not restrict access to the pond because they readjust according to the water level, and they are virtually maintenance-free, whereas netting requires vigilant maintenance (Martin, Martin and Taber 1998). Bird balls would be effective on many poultry farm dams, particularly those used for poultry water. The Armor Balls™ Aqua are made of food-grade HDPE that shouldn’t taint the water.

The disadvantages of bird balls are the price and to the cost of importation(around $0.50/ball delivered). A recently patented self-filling ball (Mohammadi and Hermann 2014) is being manufactured in Hungary, however, the cost is likely to be similar. The actual life span of bird balls has not yet been clearly established though a 10-year warranty is offered and some estimate the life to be 25 years. No literature found to date mentions the logistics of replacing, storing, transporting and disposing of balls, which needs to be considered too. The balls also remain untested on natural waterbodies and their effect on the aquatic ecology and water quality remains unknown. Appendix B has more information on bird balls. These devices greatly restrict light penetration into the water and are likely to kill aquatic plants and other organisms, possibly leading to putrefaction in some cases.

Lethal methods

Lethal methods are sometimes warranted to reinforce non-lethal deterrents and to delay habituation. In a disease mitigation situation, the legalities of doing so would have to be investigated and appropriate permission sought. Shooting of waterfowl should only be conducted by experienced, licensed shooters.

In Australia, killing any native birds is illegal, except under a duck hunting permit or damage mitigation permit to reduce their impact on primary production. Each permit is issued on its individual merits and this process can be slow and administratively time consuming.

Waterbird management issues unique to poultry farms

On poultry farms there are welfare considerations for poultry, and in many cases, other livestock, as well as staff, visitors and neighbours to consider. Excessive noise is not acceptable, particularly at night. Dust exhausted from the sheds may affect deterrent equipment performance and workers may need access to areas where deterrent equipment is positioned. Equipment will need to be battery- operated or solar-powered to be able to be relocated easily.

From discussions with poultry producers and industry persons, it appears poultry habituate quickly to deterrents. That may have disadvantages since wild birds are often very aware of the behaviour of other bird species in the area and may habituate faster to a deterrent in any situation where poultry are showing no reaction.

Integrated Bird Management

History has shown that ‘quick-fix’ solutions are rarely effective in the long-term, as is the case with most animal control methods. It is suggested here that an integrated approach is needed, using a combination of habitat modification, deterrents, harassment and repellents judiciously, depending on the target species and time- and site-specific needs of the property or business concerned.

The deterrence and control of birds is usually evaluated in relation to the economic damage they would cause if no action was taken, for example, harvest losses in the agricultural and horticultural sectors. In the case of the mining industry, keeping birds off toxic tailing ponds is a conservation and welfare issue whereby failing to do so could result in legal proceedings and financial penalty (refer to Syracrude oil sands mining case (St Clair 2014)). For the aviation industry, the potential loss of

human lives drives the heavy investment in bird management, as well as the economic losses and inconveniences associated with the precautionary grounding of planes involved in birdstrikes.

In the poultry industry’s case, while it is ‘best practice’ to prevent contact with waterbirds, the logistics and costs of doing so, and the low risk of avian influenza infection, provide little incentive for most producers to take preventative measures. The cost benefits are far from clear and will be farm-specific depending on their ‘waterfowl attractiveness’, the type of farm, and their diligence with biosecurity requirements. However, things could change dramatically in the event of a serious HPAI epidemic. The need for improved poultry industry biosecurity measures for AI may increase in the future and this would require a re-evaluation of costs and benefits.

For many meat chicken farms, an automated integrated deterrent approach (Stevens et al. 2000; Muehlebach 2005) is likely to be the most effective in controlling waterfowl on poultry farms. Since some waterfowl are also active at night, this necessitates detection of flying birds at all times of the day. An integrated strategy requires determining whether they should be deterred based on distance, size and flight path, and only triggering deterrents when targeted birds approach or enter the no-go zones.

Radar-activated deterrents

Most commercially available visual and acoustic deterrents are deployed indiscriminately and not in response to specific bird activity. This eventually leads to the birds recognising them as ‘false alarms’ and becoming habituated. Effective use of deterrents requires a more discretionary and intelligent approach based on ecological and ethological principles. In the case of waterfowl on poultry farms, it is far better to dissuade them from entering a pond or range area in the first place than to act retrospectively, once the birds have had a chance to become habituated to the site. Thus the challenge is to detect flying waterfowl, recognise which are targets and only activate aversive stimuli if they attempt to land where they are not wanted.

Radar-activated on-demand systems (Stevens et al. 2000) have been developed and used for deterring waterfowl from oil sands tailing ponds (Ronconi and St. Clair 2006; Nohara et al. 2012; St Clair 2014). These have been large-scale operations requiring multiple units to monitor large expanses of water (kilometres) but arguably could be deployed on a smaller scale under the right circumstances. An Australian-made, solar-powered system being used for horticultural applications (Muehlebach 2005) could be adapted to poultry farms.

Ronconi and St. Clair (2006) compared the industry standard, randomly firing gas cannons and using stationary human effigies, to a radar-activated on-demand deterrent system which fires gas cannons and also activates large peregrine falcon effigies only when birds approach. The radar detects the birds and relays the information to a computer that automatically deploys the deterrents. They found the radar-activated system more effective at deterring birds from landing, especially shorebirds, and the cannons were more effective than the peregrine effigies. Part of the reason for this may be that the birds are less likely to habituate to the cannons because they are only fired when the birds approach.

Radar has some disadvantages, particularly size, cost and signal interference. For smaller areas, other detecting and ranging (DAR) technologies will enable more simple and flexible automated deterrent systems to be developed in future. Spinning multi-beam light DAR (LiDAR) systems, such as the Velodyne series (www.velodyne.com ), mounted on aerial or terrestrial platforms, use radially oriented lasers to generate continuous three-dimensional point clouds in a 360° horizontal plane. Analysis of these point clouds can be used to reliably discriminate and track moving objects (Börcs et al. 2013; Chan and Lichti 2015). Low-cost LiDAR solutions have been developed for a plethora of applications including collision avoidance, blind-spot sensing, traffic monitoring, 3-D image scanning, security system components, and medical imaging.

Substituting laser with LED lights is proving suitable for various applications. For example, the new Leddar™ smart sensor technology (www.leddartech.com), with its low-cost components, high range- to-power ratio, good tolerance to weather conditions and dust, and ability to resolve multiple targets simultaneously (Leddar® - Detection and Ranging Technology 2014), could be used on poultry farms instead of radar, or to complement radar, in detecting birds entering production areas, for example, in blind spots or on the ground. They could also serve as a surveillance system to help assess the efficacy of deterrents and to monitor waterfowl behaviour on-farm, particularly nocturnal routines.

The incorporation of DAR technologies into robots, drones and other unmanned vehicles is opening up a whole new world of possibilities for bird hazing applications.

Ranking deterrents

Based on the review of the international and national literature along with consultation with Australian poultry industry and experienced persons in this field, each deterrent was evaluated independently of each other using the matrix below (Table 1). The results of applying this matrix are summarised in Table 2.

Table 1. Evaluation matrix applied to each identified deterrence measure.

Success rating (from literature / other information; authors’ judgement) 1 2 3 4 Anecdotal or limited: 1 Low Low Moderate High

Literature: 2 Low Moderate High High Trial data available: 3 Moderate High High Very High

Evidence Rating

Table 2. Deterrent ranking for management of waterbirds on Australian meat chicken farms (1 is lowest; 4 is highest)

Category Type Evidence Rating Success Rating Efficacy Ranking Auditory Ultrasonic devices 1 1 Low Gas guns and cannons 2 2 Moderate Biosonic calls 3 3 High Pyrotechnics 2 2 Moderate Visual Effigy 2 1 Low Lasers and lights 3 2 High Reflector tapes and flags 2 1 Low Falconry 3 1 Moderate Drones 1 3 Moderate Dogs 2 3 High Chemical Primary 2 2 Moderate Secondary 2 1 Low Physical Exclusion 3 4 Very High Natural wetlands 1 2 Low Habitat modification 3 3 High Lethal Lethal methods 2 2 Moderate

Australian perspective

Legislation

All native species are protected by law. However, certain species of duck may be shot by licensed shooters in the open season. The bag limit and species allowed to be shot may differ depending on the relevant State or Territory legislation. Duck-hunting is banned in Queensland. Special permits to allow culling of certain pest native species are sometimes granted to help prevent crop damage. For instance, Australian rice farmers can allow professional shooters to cull ducks causing damage to their crops.

It is highly unlikely that permission will be given to kill any aquatic birds purely because they pose a potential disease risk.

Welfare and conservation

The meat chicken industry takes bird welfare and conservation issues very seriously. Tracey et al. (2007) lists various social and environmental factors affecting bird management options. In relation to waterfowl the following sub-set may apply:

 Acceptability of culling native birds

 Risk of killing non-target animals

 Consequences of the use of chemical repellents

 Animal welfare issues associated with lethal and some non-lethal techniques

 Noise pollution associated with acoustic devices and shooting

 Aesthetic acceptability of visual scaring devices and netting

 Issues associated with habitat modification and decoy feeding.

Risk Assessment

In determining how much effort, resources and money are justifiable for waterfowl deterrence on any given poultry farm a formal AI risk assessment is recommended. Risk assessment involves identifying the:

 major hazards

 level of risk the hazards pose

 most cost-effective and ethical means of reducing the risk.

Factors contributing to risk

There are many parallels between birdstrike risk assessment in the aviation industry and that needed for AI risk assessment on meat chicken farms. Therefore, a suggested airport risk assessment methodology (Allan 2000) has been adapted here (Figure 2). The following tasks are required.

The probability of being an AI source varies with species and the probability of any species infecting poultry will depend on the species’ behaviour, together with aspects of the farm’s location and structure, and adjacent environment. Some aspects to consider are as follows:

 Species and Behaviour: At the present time, due to lack of data on disease prevalence in various species, any increase in waterfowl numbers at a site can not be directly inferred to suggest increased risk of AI. It may be the case that an increase in Australian Wood Duck numbers could constitute a serious risk, whereas an increase in Pink-eared Ducks (aquatic; non-grazers) may be of little concern. Bird behaviour may also affect the risk of AI infection. For example, an increase in Purple Swamphens may presents a greater risk than Eurasian Coots because swamphens walk around the edges and graze so might be more likely to transfer contaminated faeces.

 Location and Movement: The sites where waterbirds occur and their movement patterns must be noted as part of the risk assessment. The proximity of waterfowl to range areas is likely to increase the probability of contamination. This includes consideration of flight paths.

 External Bird Movements: Bird movements away from the poultry farm also need consideration, since it is assumed that infection is transmitted from a source of infection to the farm. It follows that greatest risk might be associated with species that travel farthest, rather than comprise local and sedentary populations. In this respect, Australian waterbirds pose a significant challenge since their movements are often sporadic and unpredictable.

 Waterbird adaptation: Waterbirds adapt as required, particularly in Australia. Consequently, their behaviour is not very predictable. Sound ornithological knowledge and on-farm observations of their behaviour are critical to the risk assessment.

 Farm Location: Waterbird populations are not evenly distributed across continental Australia and, for this reason, some areas are naturally at greater risk from AI than others. Generally coastal areas and the tropics have greater species diversity, but major river systems such as the Murray- Darling may also attract a higher proportion of species. Arid areas are generally species poor, but in areas where water occurs, numbers and species may be exceptionally high.

 Farm Structure: As previously discussed, risk of AI infection may be higher in open range farms than in closed systems, and the presence of water bodies and grassy lawns or open expanses adjacent to water may increase the attractiveness of areas to waterbirds. These factors are considered to be important in determining AI infection risk.

 Adjacent Environment: Often a poultry farm area might not provide all features required by waterbirds, but if adjacent lands provide additional features, then overall attractiveness of the area increases. Thus the presence of adjacent water bodies, open pasture, cropping lands with green crops and natural drainage lines may all influence AI risk.

Risk evaluation

Factors described above are required to be assessed in a systematic manner so that the process:

 Can be repeated consistently across farms

 Takes into account ecological variables such as species composition, seasonality of occurrence, potential of species to act as disease vectors, etc.

 Is statistically valid and based upon a robust and defensible methodology (sample methodology, sample size, statistical methods).

The primary variables to be measured are those thought to contribute to the risk of AI infection. Thus:

 Identification of species and census of numbers present, preferably over a number of seasons and under different climatic conditions. Possibly the analysis should consider a ‘worst case” scenario when establishing species and numbers present. Species records should also consider Conservation Status and any relevant Government guidelines or legislation that may affect control methodologies.

 Consideration of the potential for species to act as disease vectors either through behaviour or movement patterns. More research is required to establish this factor with any degree of certainty.

 On-farm factors including the type of farm, location of poultry areas, presence of waterbird attractants (dams, herbage).

 External factors including adjacent land uses, presence of waterbird attractants and geographical location of the farm.

For each factor it is recommended that a standardised set of categories be determined and these can then be scored, based on their potential to influence risk. As more research is conducted and our understanding of AI infection increases, these scores can be reviewed and re-applied. This is of particular importance in relation to identification of high-risk AI vector species.

Figure 2. Generalised flow diagram for identifying the level of risk and appropriate deterrent systems.

Cost–benefit analysis of deterrents

A cost–benefit analysis for each deterrent and for the industry as a whole is difficult to prescribe because of the variability between poultry farms and regions within Australia and should be conducted after a risk assessment has been undertaken. The cost of deterrents can range from cheap to very expensive and often cost may not be positively correlated with effectiveness or ease of maintenance. Integrated deterrence systems involving a range of products may need to be developed specifically for individual farm circumstances and infrastructure arrangements.

Based on the information collated for this review, Table 3 provides a summary of expected costs as at the end of 2014. These prices are indicative only and may vary between brands and models and retailers. A list of retailers and products is provided in Appendix A and B.

Deterrent Units Price ($) Comments Bird Xpeller Pro V2 1 $350 Bird distress calls used. Programmable. Would need to be tailored to situation. 240V or 12V DC battery Broadband Pro Sonic and $1000 Agricultural applications Ultrasonic Propane gas cannons 1 $700- Single-, double-, multi-shot options. Activation $1000 programmable. Need to consider noise pollution laws; may be inappropriate at night Pyrotechnic charged 1 $1 Potential grass fire hazard. cartridges Australian product, Bird Frite, no longer made. Humming Line 30m $20 Short-term or complementary deterrent e.g. along 500m $100 drainage areas, over temporary water, or over favoured grassed areas in free range Scare Hawk/HawkBird 1 $40 Pools and dams. Realistic shape and ‘plumage’ but Scarer not like any Australian species. Bird Gard Kite Life like replica. Illuminated Owl 1 $100 Great horned owl likeness; non-Australian species; Australian owls all nocturnal. Unrealistic threat Holographic Ribbon 30m $20 Stretch and spiral. Circular holograms flash. Light wind required. Limited life. Perhaps Irritape 30m $70 Metallic noise and flashes Flock Reflector 1 $90 Spins and creates lights that flash; UV and weather resistant Scare Balls 3-pack $40 Balloons; bright colours; scary eyes. Terror Eyes 1 $80 Large size, bright, scary eyes and bobbing motion ScaryMan Human effigy. Inflates and deflates 7 times over a 25 day unit 1 $900 second period. Activation programmable. Siren sounds when activated. Consider noise pollution multi-use 1 $1000 laws. Multi use unit glows bright red in dark; 12V Preset timer unit 1 $1100 battery operation; recharge every couple of weeks. Variable timer 1 $200 Can reinforce with bundles of similar fabric placed elsewhere. Euro-matic 1000 ~$50-60K 100 mm diameter HDPE balls; 116 balls per sq. m; Bird Balls™ sq m delivered disguise pond surface; reduce algal growth; virtually maintenance-free; allow rainfall in; self-adjust with Armor Balls™ AQUA changing levels; reduce evaporation; have to be imported. Self-filling type cheaper to ship. Exclusion netting 1 ha $17-72K Based on flying fox netting (Rigden 2008) installed Depends on type of netting, topography, location. BirdDeter Radar System 1 ~$25K Microwave transmitter and receiver, up to 500m (including plug-in: strobe apart; radio control; deterrent sequence changed each lights, gas gun, electric activation; data logged. hawk, Scarey Man)

Discussion

Reducing the risk of AI through more proactive waterbird management on meat chicken farms is certainly possible but the means by which it can be achieved on any given farm may be quite different in strategy and cost, particularly for free-range farms wanting to deter birds from range areas.

At the present time, reliable data on the potential for many Australian waterfowl to act as AI vectors is still incomplete and further studies in this area will undoubtedly influence strategies to combat AI infection. This may well be the single most important area of uncertainty in developing an effective strategy. Ongoing observation and reassessment of techniques will be required. Professional biologists will also likely be required to conduct at least some aspects of any risk assessment so that relevant ecological information is incorporated.

The report discusses a number of deterrents trialled for a range of bird species under different circumstances and purposes, but assessing their real efficacy for the poultry industry will require long- term field experiments in a range of varying environmental settings. The overall methodology recommended here involves the application of a system that prevents habituation of target waterbirds and can remain vigilant both day and night. The actual devices deployed may depend largely on cost but also upon local conditions at the site. These conditions include both biological (species detected, waterbird resources available) and anthropogenic factors (proximity to residences, adjacent land uses).

In respect of the latter, it must be noted that the meat chicken industry has a range of sensitive issues to consider. These include bird welfare and odour nuisance. Waterbird deterrent strategies will need to be integrated with this mix of considerations so that the overall outcome is a positive one. For example, if noise pollution is an issue for a certain site, then a system such as gas cannons may not be feasible and one or more other deterrents used to compensate, or harassing waterfowl under drought conditions that may cause concern amongst the conservation fraternity.

Bioeconomic modelling could help in cost–benefit analysis and predicting the AI risk at various time intervals; monthly, seasonally or yearly, taking into account climate change and other factors that may be influencing AI epidemiology. For instance, there may be a link with HPAI and drought cycles in Australia (Klaassen et al. 2011). This could be a key to determine when to monitor and deter waterbirds more proactively, and when to wind back. This would help delay habituation.

Implications

Waterfowl management is complex at the farm level and also at the regional and national levels. Presently, the effective evaluation of an AI threat is problematic because of lack of critical biological data. This is exacerbated by the range of circumstances found on individual farms - infrastructure design and placement, farming technique and surrounding land uses, all of which influence the scope for effective waterfowl management. In addition, poultry farms need to operate within all levels of government policy and associated regulatory frameworks. These factors suggest a number of implications:

 Further research is required to fully understand the biological processes relating to the potential introduction and spread of AI.

 Assessment tools need to be developed in order to evaluate the potential risk of poultry farms to AI.

 On-ground trials of a range of deterrent systems need to be conducted to gain a better understanding of their efficacy under a range of farm and environmental conditions.

 If the risk of AI is considered to be high, then Government policy and assistance in implementing a control strategy may be required.

In the short term, peak bodies, research groups and government agencies need to adopt a strategic approach to supporting AI risk assessment and implementation or experimentation with AI preventative programs on-farm. In particular, geographic regions and/or individual farms that have experienced AI outbreaks in Australia would be a point of focus for tighter biosecurity measures. Ultimately, any farms that currently attract waterfowl are candidates for intervention as they are potentially more vulnerable to HPAI outbreaks and flow-on effects (that is, quarantining) and therefore stand to benefit more from bird management.

This review has suggested the importance for policy makers to conduct cost–benefit analysis before deciding the level of commitment required to deter waterfowl from farms. The results of such a review may determine that a greater financial commitment from government, industry and individual producers to deal with individual AI outbreaks may be a better option than spending millions of dollars across the whole industry to implement overall waterfowl management measures. In respect of short-term considerations and industry improvements:

 Meat chicken farms may reduce their AI risk, in some cases considerably, should the industry enforce stricter biosecurity measures regarding waterbird deterrence and farm management.

 While there are effective waterfowl deterrent strategies that would reduce the risk of AI transmission, arguably, the more expensive, low-maintenance options would be best for immediate trial.

Recommendations

The overarching recommendation is that the industry should support further research to confidently establish whether particular waterfowl deterrent systems will be effective and economically viable for the long-term benefit of the meat chicken industry.

More specifically, it is recommended that:

 Seasonal surveys of waterfowl and other wild birds on meat chicken farms are conducted to confirm which species are most likely to act as AI vectors in different regions at different times of the year. This will influence the selection and optimisation of deterrents potentially required.

 Desktop assessments are undertaken to estimate the likely cost of implementing and maintaining recommended deterrent strategies individually, and in combination, including the costs of:

o habitat modifications including:

. Maintaining optimal grass height and preventing it going to seed

. Improving drainage on range areas and around production sheds

. Covering or removing open water storages.

o deterrent operation and maintenance

o data gathering to assess efficacy of the system over time

 A suitable, cost-effective detection and ranging system be developed to monitor and record waterfowl activity in and around poultry production areas, and to evaluate the short- and long- term efficacy of deployed deterrent strategies.

 A proven radar-activated on-demand deterrent system is used to demonstrate how judicious use of deterrents on waterfowl will help prevent habituation, prolonging the efficacy of deterrent systems. This could be done off-farm, as an extension exercise.

 Producers are provided with a risk assessment management tool so that every meat chicken farm can be assessed for AI risk.

 Australian meat chicken farms are surveyed and ranked according to their perceived AI-risk, individually, and regionally.

 Education is provided for producers about Australian waterfowl and they are encouraged to record on-farm behaviour of waterfowl to assist risk assessment.

 Bio-economic modelling is conducted over a sufficient planning horizon (say 30 years) to demonstrate the merit or otherwise of investing in waterfowl deterrence strategies, taking into account industry expansion, various outbreak scenarios, and new findings about AI epidemiology.

References

ABARES 2015, Agricultural commodities, March Quarter 2015, Australian Bureau of Agricultural and Resource Economics, Canberra.

Alexander, DJ 2000, ‘A review of avian influenza in different bird species’, Veterinary Microbiology, vol. 74, pp. 3-13.

Allan, JR 2000, ‘A protocol for bird strike risk assessments at airports’, International Bird Strike Committee, Amsterdam, 17–21 April 2000.

Andelt, WF, Woolley, TP and Hopper, SN 1997, ‘Effectiveness of barriers, pyrotechnics, flashing lights and ScareyMan for deterring heron predation on fish’, Wildlife Society Bulletin, vol. 25, pp. 686-694.

Battistoni, V, Montemaggiori and A, Iori, P 2008, ‘Beyond falconry between tradition and modernity: a new device for bird strike hazard prevention at airports’, in Proceedings of the International Bird Strike Committee, (IBSC’08), Brasilia, Brazil, pp. 1-13.

Belant, JL and Ickes, SK 1997, ‘Mylar flags as gull deterrents’, in Proceedings of the Thirteenth Great Plains Wildlife Damage Control Workshop, Manhattan, Kansas.

Belant, JL and Martin, JA 2011, Bird harrrassment, repellent, and deterrent techniques for use on and near airports. A synthesis of airport practice, Airport Cooperative Research Program, Synthesis 23, Transportation Research Board, Washington, D.C.

Bishop, J, McKay, H, Parrott, D and Allan, J 2003, Review of international research literature regarding the effectiveness of auditory techniques and potential alternatives, Department for Environment Food and Rural Affairs.

Blackwell, BF, Seamans, TW and Dolbeer, RA 1999, ‘Plant growth regulator (Stronghold™) enhances repellency of anthraquinone formulation (Flight Control™) to Canada Geese’, Journal of Wildlife Management, vol. 63, pp. 1336-1343.

Boag, DA and Lewin, V 1980, ‘Effectiveness of three waterfowl deterrents on natural and polluted ponds’, Journal of Wildlife Management, vol. 44, pp. 145-154.

Bomford, M and O'Brien, P 1990, ‘Sonic deterrents in animal damage control: a review of device tests and effectiveness’, Wildlife Society Bulletin, vol. 18, pp. 411-422.

Bomford, M and Sinclair, R 2002, ‘Australian research on bird pests: impact, management and future directions’, Emu, vol. 102, pp. 29-45.

Börcs, A, Józsa, O and Benedek, C 2013, ‘Object extraction in urban environments from large-scale dynamic point cloud datasets’, in Proceedings of the 11th IEEE International Workshop on Content- based Multimedia Indexing (CBMI), Veszprém, Hungary, pp. 191-194.

Caron, A, Grosbois, V, Etter, E, Gaidet, N and de Garine-Wichatitsky 2014, ‘Bridge hosts for avian influenza viruses at the wildlife/domestic interface: an eco-epidemiological framework implemented in southern Africa’, Preventive Veterinary Medicine, vol. 117, pp. 590-600.

Castelli, PM and Sleggs, SE 2000, ‘Efficacy of Border Collies to control nuisance Canada geese’, Wildlife Society Bulletin, vol. 28, pp. 385-362.

Chan, TO and Lichti, DD 2015 ‘Automatic in situ calibration of a spinning beam LiDAR system in static and kinematic modes’, Remote Sensing, vol. 7, pp 10480-10500.

Cummings, JL, Otis, DL and Davis, Jr., JE 1992, ‘Diemthyl and methyl anthranilate and methiocarb deter feeding in captive Canada geese and mallards’, The Journal of Wildlife Management, vol. 56, pp. 349-355.

Dolbeer, RA 1998, ‘Evaluation of shooting and falconry to reduce bird strikes with aircraft at John F. Kennedy International Airport’, in Proceedings of International Bird Strike Committee Meeting, vol. 24, pp. 145-158.

East, IJ, Hamilton, S and Garner, G 2008, ‘Identifying areas of Australia at risk of H5N1 avian influenza infection from exposure to migratory birds: a spatial analysis’, Geospatial Health, vol. 2, pp. 203-213.

Erickson, WA, Marsh, RE and Salmon, TP 1990, ‘A review of falconry as a bird-hazing technique’, in Proceedings of the 14th Vertebrate Pest Conference, vol. 14, pp. 314-316

Fairbrother, JG 1999, ‘Emergency diseases from an Australian perspective’, in Proceedings of Australian Poultry Symposium, Sydney, Australia, pp. 23-30.

FAO 2009, Biosecurity for highly pathogenic avian influenza: issues and options. FAO, Italy.

FAO 2015, Understanding Avian Influenza, Available from: http://www.fao.org/avianflu/documents/key_ai/key_book_preface.htm [accessed 3 June 2015]

Gilsdorf, JM, Hygnstrom, SE and VerCauteren, KC 2002, ‘Use of frightening devices in wildlife damage management’, Integrated Pest Management Reviews, vol. 7, pp. 29-45.

Glahn, J, Ellis, G., Fiornelli and Dorr, B 2000, ‘Evaluation of moderate- and low-powered lasers for dispersing double-crested cormorants from their night roosts’, in Proceedings of the Ninth Wildlife Damage Management Conference, State College, Pennsylvania, USA.

Gorenzel, WP, Conte, FS and Salmon, TP 1994, Bird damage at aquaculture facilities. The handbook: prevention and control of wildlife damage, Paper 57, Institute of Agriculture and Natural Resources. http://digitalcommons.unl.edu/icwdmhandbook/57

Grillo, T 2015, Contribution to the 2012 avian influenza in wild birds surveillance program, RIRDC Publication No. 15/016.

Grimes, T and Jackson, C 2015, Code of Practice for Biosecurity in the Egg Industry, Australian Egg Corporation Limited.

Hansbro, PM, Warner, S, Tracey, JP, Arzey, KE, Selleck, P, O’Riley, K, Beckett, EL, Bunn, C, Kirkland, PD, Vijaykrishna, D, Olsen, B and Hurt, AC 2010, ‘Surveillance and analysis of avian influenza viruses, Australia’, Emerging Infectious Diseases, vol. 16, pp. 1896-1904.

Harris, RE and Davis, RA 1998, Evaluation of the efficacy of products and techniques for airport bird control, LGL Limited for Aerodrome Safety Branch, Transport Canada, Ottowa, Ontario.

Haynes, L, Arzey, E, Bell, C, Buchanan, N, Burgess, G, Cronan, V et al. 2009, ‘Australian surveillance for Avian Influenza viruses in wild birds between July 2005 and June 2007’, Australian Veterinary Journal, vol. 87, pp. 266-272.

Hoque, MA, Burgess, GW, Cheam, AL and Skerratt, LF 2015, ‘Epidemiology of avian influenza in wild aquatic birds in a biosecurity hotspot, North Queensland, Australia’, Preventive Veterinary Medicine, vol. 118, pp. 169-181.

Jaremovic, R 1990, ‘Bioacoustical scaring trials’, in Fleming, P, Temby, I and Thompson, J, (eds) National Bird Pest Workshop Proceedings, Armidale, pp. 98-110.

Jeffrey, C 2014, Robotic raptors look and fly like the real thing. Available from: [30 May 2015]

Kitowski, I 2011 ‘Civil and military birdstrikes in Europe: an ornithological perspective’, Journal of Applied Sciences, vol. 11, pp. 183-191.

Klaassen, M, Hoye, BL and Roshier, DA 2011, ‘Identifying crucial gaps in our knowledge of the life- history of avian influenza viruses: an Australian perspective’, Emu, vol. 111, pp. 103-111.

Leddar® - Detection and Ranging Technology, 2014 (video file), Available from: https://www.youtube.com/watch?v=GpADWIR8ahA [3 June 2015]

MacKenzie, M 2014, ‘Range management for disease control: Guidelines to protect your free-range flock from exotic disease’, The Drumstick Newsletter, July.

Marchant, S and Higgins, PJ 1990, Handbook of Australian, New Zealand and Antarctic Birds (HANZAB). Volume 1 Ratites to Ducks, Oxford Unity Press Australia.

Marcsik, D and Clarke, D 1997, An evaluation of control methods to reduce winged vertebrate pest damage to horticultural crops in the Top End, Technical Bulletin No. 260, Northern Territory Department of Primary Industries and Fisheries.

Martin, LR, Martin, LM and Taber, MR 1998 ‘Containment basins and bird exclusion – a historical perspective’, in Proceedings of the Eighteenth Vertebrate Pest Conference, Paper 60.

Mohammadi, P and Hermann, L 2014, Cover device and blanket for covering liquid surfaces, WIPO International Patent WO 2014/181142A1, available at < http://www.sumobrain.com/patents/wipo/Cover-device-blanket-covering- liquid/WO2014181142A1.html> [accessed 17 June 2015].

Mott, DF, Glahn, JF, Smith, PL, Reinhold, DS, Bruce, KJ and Sloan, CA 1998, ‘An evaluation of winter roost harassment for dispersing double-crested cormorants away from catfish production areas as in the Mississippi’, Wildlife Society Bulletin, vol. 26, pp. 584-591.

Mott, DF and Timbrook, SK 1988, ‘Alleviating nuisance Canada goose with acoustical stimuli’, in Crabb, AC and Marsh, RE (eds) Vertebrate Pest Conference Proceedings collection, Proceedings of the Thirteenth Vertebrate Pest Conference, Nebraska, vol. 13, pp. 301-304.

Muehlebach, J 2005, ‘Good measurement leads to better bird management’, The Australian and New Zealand Grapegrower and Winemaker, January.

Nohara, TJ, Beason, RC and Weber, P, 2011 ‘Using radar cross-section to enhance situational awareness tools for airport avian radars’, Human-Wildlife Interactions, vol. 5, pp. 210-217.

Nohara, TJ, Beason, RC and Clifford, SP 2012, ‘The role of radar-activated waterfowl deterrents on tailings ponds’, Paper presented to International Oil Sands Tailing Conference, Edmonton, Alberta, 2- 5 December.

OIE 2015, ‘Update on highly pathogenic avian influenza in animals (type H5 and H7)’, available at [accessed 20 May 2015].

Pizzey, G and Knight, F 2007, The field guide to the birds of Australia, 8th edn, Harper Collins.

Read, JL 1999, ‘A strategy for minimizing waterfowl deaths on toxic waterbodies’, Journal of Applied Ecology, vol. 36, pp. 345-350.

Rigden, P 2008, To net or not to net, 3rd edn, Queensland Department of Primary Industries and Fisheries.

Ronconi, RA. and St. Clair, CC 2006, ‘Efficacy of a radar-activated on-demand system for deterring waterfowl from oil sands tailings ponds’, Journal of Applied Ecology, vol. 3, pp. 111-119.

Ross, BP and Furness, RW 2000, Minimising the impact of eider ducks on mussel farms, University of Glasgow, Scotland.

Scott, P 2014, ‘Minimising the risk of avian influenza on free-range farms’. Paper presented at the Poultry Information Exchange 2014, PIX, Broadbeach, Queensland.

Seamans, TW and Bernhardt, GE 2004, Response of Canada geese to a dead goose effigy, USDA National Wildlife Research Center – Staff Publications, Paper 384.

Selleck, PW, Arzey, G, Kirkland, PD, Reece, RL, Gould, AR, Daniels, PW and Westbury, HA 2003, ‘An outbreak of highly pathogenic avian influenza in Australia in 1997 caused by an H7N4 virus’, Avian Disease, vol. 47, pp. 806-811.

Spanier, E 1980, ‘The use of distress calls to repel night herons (Nycticorax nycticorax) from fish ponds’, Journal of Applied Ecology, vol. 17, pp. 287-294.

St Clair, CC 2014, Final report of the Research on Avian Protection Project (2010-2014). Prepared for Alberta Justice, Edmonton, Canada.

St. Fleur, N 2015, Birds Are Mostly Cool With Drones. Available from: [30 May 2015]

Stevens, GR, Rogue, J, Weber, R and Clark, L 2000, ‘Evaluation of a radar-activated, demand- performance bird hazing system’, International Biodeterioration and Biodegradation, vol. 45, pp. 129-137.

Summers, RW and Hillman, G 1990, ‘Scaring Brent Geese (Branta bernicla) from fields of winter wheat with tape’, Crop Protection, vol. 9, pp. 459-462.

Tracey, J 2005, Targeting surveillance for avian influenza in wild birds: a pilot investigation in New South Wales, Vertebrate Pest Research Unit, NSW Department of Primary Industries.

Tracey, JP 2010, ‘Risk-based surveillance of avian influenza in Australia’s wild birds’, Wildlife Research, vol. 37, pp. 134-144.

Tracey, J, Bomford, M, Hart, Q, Saunders, G and Sinclair, R 2007, Managing Bird Damage to fruit and other horticultural crops, Bureau of Rural Sciences, Canberra.

Tracey, JP, Woods, R, Roshier, D, West, P and Saunders, GR 2004, ‘The role of wild birds in the transmission of avian influenza for Australia: an ecological perspective’, Emu, vol. 104, pp. 109-124.

Washburn, BE, Chipman, RB and Francoeur, LC 2006, ‘Evaluation of bird response to propane exploders in an airport environment’, in Proceedings of the 22nd Vertebrate Pest Conference, University of California, Davis, USA, pp. 212-215.

Wildlife Health Australia 2015, Avian Influenza Wild Bird Surveillance, Available from: [3 June 2015]

Glossary

Bio-acoustics Broadcasted alarm, distress, alert calls that have a biological relevance to target bird species. The playing of such calls indicates the presence of a potential predator and can encourage other birds to leave the area.

Biosecurity A set of measures designed to protect a population from transmissible infectious agents at national, regional and individual farm levels.

Biosonic sounds see Bio-acoustics

Distress call Call made when captured in a net or by a predator.

Emergency Animal Government and Livestock Industry Cost Sharing Deed in Respect of Disease Response Emergency Animal Disease Response, as amended from time to time. Agreement (EADRA)

Emergency Animal A response to an EAD that is, in accordance with Part A of Schedule 4, Disease Response developed by a State or Territory CVO and endorsed by the CCEAD and the Plan (EADRP) NMG and which is subject to Government and Industry Parties’ Cost Sharing.

Falcon Agile raptor with long, tapering wings that typically catch prey by diving on it from above.

Falconry Use of trained falcons to hunt prey.

Free range farm Poultry farm where poultry are allowed access outside of the shed each day after a certain age is reached into an enclosed range area.

Habituation Action of becoming accustomed to a deterrent technique due to persistent exposure, so that it no longer has desired effect.

Hawk Raptor with short, rounded wings, that typically catches prey near to the ground after a short chase, e.g. goshawk, sparrowhawk

Monofilament line Single strand line such as fishing line.

Pyrotechnics Scaring techniques that use a variety of explosive devices that emit loud, banging noises and can also produce flashes of light.

Raptor Bird of prey, including hawks, falcons, harriers, eagles, kites, owls.

Ultrasonics High sound frequencies that are inaudible to humans (i.e. >20,000 Hz).

Waterbird A bird that lives on or around water.

Waterfowl Any aquatic freshwater bird, especially any species of the family Anatidae (ducks, geese, and swans); such birds collectively.

Appendix A – Manufacturer contact details

Contact Details for Electronic, Visual and Other Bird Scare Devices:

Company Name: Vigilance Technologies Pty Ltd

Type: Acoustic and visual; radar-activated deterrent systems

Contact Person: John Muehlebach

Phone: 07 4667 0491

Other contact details: 9 Easey St Warwick QLD 4370

www.birddeter.com.au email: [email protected]

Company Name: Bird Gard Pty Ltd

Type: Acoustic

Phone: 07 5443 6344

Other contact details: PO Box 737, Cotton Tree QLD 4558

www.birdgard.com.au

Company Name: Bird-X-Peller Australia

Type: Acoustic and visual

Phone: 03 6165 0307

Other contact details: 177 Elizabeth St Hobart TAS 7000

[email protected]

Company Name: Daken Farm Equipment

Type: Gas guns- Zon bird scare gun

Phone: 02 94772599

Other contact details: 30 Salisbury Rd Hornsby NSW 2077 www.daken.com.au/zon.htm

Company Name: E.H. Cambridge & Co

Type: Gas guns

Phone: 1800 888 137 (Toll free); (08) 8391 1688 (Mt Barker)

Other contact details: 9 Mount Barker Rd, Mount Barker

Postal: P.O. Box 721 Mount Barker South Australia 5251.

Facsimile: (08) 8391 2759.

Email: [email protected];

http://www.ehcambridge.com.au

Company Name: Peaceful Pyramid Birdscarers

Type: Visual

Other contact details: http://www.peaceful-pyramid.co.uk

Company Name: Pest-Away Australia

Type: Visual, Acoustic, Roosting deterrents, Netting

Phone: 02 6493 5840

Other contact details: PO Box 115 Bermagui NSW 2546

Fax: 02 6493 5841

http://www.pestawayaust.com.au

Company Name: Phoenix Deterrent Systems

Type: Acoustic

Contact Person: Bill Johnston

Phone: 03 97390557 Mobile 0413400540

Other contact details: Email: [email protected]

Web: www.phoenixagritech.com

PO Box 167 Coldstream Vic 3770

Company Name: Bird-X

33 Type: Various

Other contact details: http://www.bird-x.com/

Company Name: Tisara (Australia) Pty Ltd

Type: Hawk bird scarer

Phone: 02 4934 8330

Other contact details: PO Box 36, Morpeth NSW 2321

[email protected]

Appendix B – Additional information

Ball covers http://www.awtti.com/armor_balls_aqua_275_cover.php http://eccllc.us/purchase/ http://eccllc.us/project/306/ http://eccllc.us/project/bird-control/ http://euro-matic.eu/en/products/self-filling-cover-ball/ http://www.sumobrain.com/patents/wipo/Cover-device-blanket-covering- liquid/WO2014181142A1.html

Drones and robotic birds http://www.dailymail.co.uk/sciencetech/article-2743272/Is-bird-Is-plane-No-s-ROBIRD-Robotic- falcons-eagles-mimic-real-predators-pests-away-airports-farms.html http://clearflightsolutions.com/methods/robirds http://www.bird-x.com/remote-control-drone-pages-64.php

Radar-activated deterrent systems http://www.birddeter.com.au/products/radar.html

National Avian Influenza Wild Bird (NAIWB) Surveillance Program https://wildlifehealthaustralia.com.au/ProgramsProjects/AvianInfluenzaWildBirdSurveillance.aspx

Detection and ranging sensors

Govers, F 2014, Review: LeddarTech's LED-based detection and tracking technology. Available from: http://www.gizmag.com/leddar-distance-sensor/30061/ [accessed 3 June 2015]

LeddarTech's Ground-Breaking Industrial Leddar® Sensor, 2014 (video file), Available from https://www.youtube.com/watch?v=DdBkhEq2xkg [accessed 3 June 2015]

PulsedLight, Inc http://0cm.us/hosting/pl3d/LIDAR-Lite-Technology-Brief.pdf [accessed 3 June 2015]

Velodyne LiDAR http://velodynelidar.com/ [accessed 30 January 2016]

Appendix C Waterfowl distributions

Table C1. Australian waterfowl with extra-limital distributions outside Australia

Australian Species Found in Papua New Found in Indonesia Found in China Guinea Magpie Goose X X Plumed Whistling-Duck X Wandering Whistling-Duck X X Black Swan X X Radjah Shelduck X X Green Pygmy-Goose X X Cotton Pygmy-Goose X X X Grey Teal X X Pacific Black Duck X X Hardhead X X Northern Shoveler X X X Australian Pelican X X Little Black Cormorant X X Little Pied Cormorant X X Great Cormorant X X X Pied Cormorant X X Australian Grebe X X Darter X X X Pacific Heron X X White-faced Heron X X Great-billed Heron X X X Pied Heron X X Rufous Night Heron X X Australian Little Bittern X X Black Bittern X X X Cattle Egret X X Little Egret X X X Intermediate Egret X X X Great Egret X X X Royal Spoonbill X X Glossy Ibis X X X Straw-necked Ibis X X Australian White Ibis X X

Appendix D Waterfowl photos

Australian Waterfowl – Selection of species featured in this review

These photographs are by Bruce Thomson, Red Leaf Environmental Pty Ltd and are available for use by RIRDC in future publications or brochures.

Australian Wood Duck (Chenonetta jubata)

Chestnut Teal (Anas gracilis)

Wandering Whistling-Duck (Dendrocygna arcuata)

Plumed Whistling-Duck (Dendrocygna eytoni)

Pacific Black Duck (Anas superciliosa)

Hardhead (Aythya australis)

Cotton Pygmy-Goose (Nettapus coromandelianus)

Magpie Goose (Anseranas semipalmata)

Australian White Ibis (Threskiornis molucca)

Glossy Ibis (Plegadis falcinellus)

Straw-necked Ibis (Threskiornis spinicollis)

Cattle Egret (juvenile) (Ardea ibis)

Rufous Night Heron (Nycticorax caledonicus)

White-faced Heron (Egretta novaehollandiae)

White-necked (Pacific) Heron (Ardea pacifica)

44 45 Deterrence of wild waterfowl from poultry production areas: a critical review of current techniques and literature by Michael Atzeni, Darren Fielder and Bruce Thomson

AgriFutures Australia Publication No. 17/058 AgriFutures Australia Project No. PRJ-009194 ISBN: 978-1-74254-981-1

AgriFutures Australia

Building 007 Tooma Way Charles Sturt University Locked Bag 588 Wagga Wagga NSW 2650

02 6923 6900 [email protected]

@AgriFuturesAU agrifutures.com.au

AgriFutures Australia is the trading name for Rural Industries Research & Development Corporation. AgriFutures is a trade mark owned by Rural Industries Research & Development Corporation.