Risk assessment for Devilbend and Bittern Reservoirs

Lance Lloyd and Peter Newall

3^rd September 2009
Final Version 2

Download the PDF version of this document: Translocation risk assessment for Devilbend and Bittern Reservoirs for stocking select recreational (fish) species (PDF - 782.0 KB)

Table of Contents

1 Introduction
1.1 Aim And Scope Of The Risk Assessment
1.2 Structure Of The Risk Assessment

2 Background Information
2.1 Devilbend And Bittern Reservoirs
2.2 Devilbend And Balcombe Creek System (Potential Receiving Water)
2.3 Fish Species Proposed To Be Translocated
2.3.1 Australian Bass
2.3.2 Brown And Rainbow Trout

3 Project Methods
3.1 Task 1: Review And Scoping
3.2 Task 2: Information Review
3.3 Task 3: Field Inspection
3.4 Task 4: Risk Assessment
3.4.1 Identifying And Incorporating The Values
3.4.2 Risk Characterisation
3.4.3 Management Action Recommendations
3.5 Task 5: Reporting

4 Values & Threats
4.1 Values
4.1.1 Current fish community in Devilbend & Bittern Reservoirs (threats to the potential angling value and natural ecosystem value). This value could be impacted through the following pathways:
4.1.2 Current fish community in Devilbend Creek and Balcombe Creek system (must at least maintain current ecological condition). This value could be impacted through the following pathways:
4.1.3 Significant fish species (Dwarf Galaxiid) in Devilbend Creek and Balcombe Creek system. This value could be impacted through the following pathways:
4.1.4 White-bellied Sea Eagle population. This value could be impacted through the following pathways:
4.1.5 Blue-billed Duck population. This value could be impacted through the following pathways:
4.2 Threats
4.2.1 Predation in Reservoirs (fish)
4.2.2 Predation in Reservoirs (macroinvertebrates)
4.2.3 Competition in Reservoirs (fish impacts)
4.2.4 Introduction of Disease to Reservoirs
4.2.5 Predation in Creeks
4.2.6 Predation in Creeks (macroinvertebrates)
4.2.7 Competition in Creeks (juvenile stages of escapees)
4.2.8 Introduction of Disease to Creeks
4.2.9 Reduced food for Sea Eagles
4.2.10 Reduced food for Blue-billed Duck
4.3 Ecosystem Level Impacts

5 Statement Of Investigation

6 Conceptual Model

7 Risk Analysis
7.1 Risk Analysis Results And Rationale For The Value 'Fish Community In Reservoirs'
7.2 Risk Analysis Results And Rationale For The Value 'Fish Community In Devilbend Creek And Balcombe Creek System'
7.3 Risk Analysis Results And Rationale For The Value 'Dwarf Galaxiid Population In Devilbend Creek And Balcombe Creek System'
7.4 Risk Analysis Results And Rationale For The Value 'White-Bellied Sea Eagle Residents'
7.5 Risk Analysis Results And Rationale For The Value 'Blue-Billed Duck Population'
7.6 Knowledge Gaps And Assumptions

8 Conclusions
8.1 Risk Mitigation

9 References

10 Appendix 1: Tep Risk Assessment Proforma

1 Introduction

Parks Victoria is in the final stages of drafting a Management Plan for the Devilbend Natural Features Reserve. The planning will focus on developing an environmentally sustainable park and ensuring that any recreational use is consistent with conservation objectives. A range of activities are being considered for approval, with advice provided by an Advisory Group. The activities being considered include recreational fishing. This is consistent with recent Victorian Government 2006 election commitments and current Government policy to look at options for developing new fishing opportunities and fisheries. One of the identified new opportunities is the Devilbend Reservoir on the Mornington Peninsula.

One of the steps in developing the Devilbend and Bittern Reservoirs as stocked fisheries is to conduct a risk assessment for submission to the Translocation Evaluation Panel (TEP). The risk assessment must demonstrate that the translocation meets the requirements of the Guidelines for Assessing Translocations of Live Aquatic Organisms in Victoria. The established risk evaluation process includes consideration of the values of the site and the potential impacts of the translocated species.

Some of the environmental values of the site are high and there is community support for an ecologically viable, conservation-focussed, park. The site supports more than 150 native plant species and 80 significant fauna species. The nationally vulnerable Dwarf Galaxias fish has historically been found in the catch drain and nearby creeks (though not in recent surveys; McGuckin 2007), but has never been found in the reservoirs themselves.

The Growling Grass Frog was last recorded at the site in 1978, the reasons for its absence are unknown but it is known this species are subject to impacts from exotic fish species (Heard et al. 2004) such as Gambusia and Redfin but also require extensive grasslands next to permanent water. The most recent survey by Monash (Thompson and Walker 2009) did not detect this species. Regardless the addition of further predatory species is unlikely to increase the risks that the species, if found would face at Devilbend and Bittern Reservoirs, and is not considered further.

The reserve is home to the threatened Blue-billed Duck and the White-bellied Sea-Eagle, the latter species being dependent upon aquatic fauna such as fish and macro-crustaceans (Stamation and Lyon 2009). As such these two species of bird represent other significant bird species which may be present or are newly recognised onsite (such as the recent sighting of nesting Caspian Terns, Jo Richards, Parks Victoria pers. comm.) as one species that represents the piscivorous species (Sea Eagle) and the other represents the invertebrate dependent species (Blue-Bill Duck).

There are also several exotic aquatic species such as Redfin Perch, Eastern Gambusia, Goldfish and Marron (McGuckin 2001 & 2007; Thompson and Walker In Prep). There is also community support for a range of recreational activities in the Reserve, in particular recreational angling and development of a stocked fishery in the reservoirs.

The impacts or threats posed by the stocked fish include predation upon, or competition with, significant species; ecosystem effects; genetic impacts; and disease introduction. The risk assessment needs to consider all these values and impacts (taking account of the likelihood and consequences of these impacts occurring), however there may be mitigation options which exist, and assessing these will allow documenting any residual risk that managers can take into account in proceeding or not with any proposed translocation.

1.1 Aim and Scope of the Risk Assessment

The objective of this project is to undertake an independent, scientific risk assessment of the stocking of select recreational fish species in Devilbend and Bittern Reservoirs. The risk assessment is to be based on the approved proforma for submission to the Translocation Evaluation Panel (TEP). The proforma assesses environmental risks, but also takes into account social and economic factors of the proposed translocation to allow an informed decision in regard to the proposal.

1.2 Structure of the Risk Assessment

There are two documents making up this risk assessment. This document which follows the general approach to ecological risk assessment adopted by EPA Victoria (EPA Victoria 2004) and used in several recent projects by Lloyd Environmental (e.g. Lloyd et al. 2008, Newall and Lloyd 2008) and a second document, the TEP Risk Assessment Proforma, which uses the outputs from this primary report. The TEP Risk Assessment Proforma is provided as Appendix 1 to this document and must be read in conjunction with this supporting risk assessment.

2 Background Information

2.1 Devilbend and Bittern Reservoirs

Bittern and Devilbend reservoirs are decommissioned Melbourne Water storages which were previously filled by transferring water directly into the storage piped from nearby reservoirs. The local catchment inflow is excluded using a catch-drain (Figure 2.1 & Figure 2.2). Bittern is much smaller than Devilbend in capacity and drains into Devilbend Reservoir. Devilbend Reservoir overflows into Devilbend Creek and then into Balcombe Creek (Figure 2.3).

Figure 2.1 Devilbend Reservoir and its upstream catchments (GHD 2008)
(note the catch-drain runs along the southern edge of Bittern Reservoir and all the way along Devilbend Reservoir into Devilbend Creek to completely by-pass these storages)

Figure 2.2 Hydrodynamics diagram of the Devilbend and Bittern Reservoirs and catchments (GHD 2008)

Figure 2.3 Devilbend Reservoir and its downstream catchment

Devilbend and Bittern Reservoirs are constructed habitats but are well established with complex aquatic habitats such as woody debris, rocky sections, gravel substrates and extensive aquatic vegetation and also excellent water quality. These sites would provide excellent habitat for many fish species. There are several exotic species already established, such as of Redfin Perch, Eastern Gambusia and Marron, however, these are unlikely to affect survival of any stocked species due to excellent cover (extensive aquatic vegetation beds) available.

Currently, only the Devilbend Reservoir spills in around 20% of years under historic conditions and around 7% under median climate change scenario conditions. Bittern does not currently spill and is less likely to under climate change scenarios. Two options for managing reservoir water levels have been examined (based on the median climate change scenario) – removal of catch drains (Option 1) and winter fill option (where 50% of flows from May to Oct would be diverted from local catchments into the reservoir – Option 2). Removal of the catch drains means that Bittern Reservoir would spill 80% of years (up from 0% with catch drains) and that Devilbend would spill 85% of years. Unregulated the winter fill option means that Bittern and Devilbend would spill about 55% of years (Table 2.1; Figure 2.4 & Figure 2.5).

Table 2.1 Reservoir Spill Frequency under various Management Scenarios (GHD 2008)

Under the Parks Victoria preferred operating regime (Winterfill - Option 2), these reservoirs are essentially open systems. At the current spillway levels (and proposed inflows from the catch drains), water would leave both reservoirs about one in every two years. Bittern would spill into Devilbend Reservoir only. Devilbend would spill into Devilbend Creek via its spillway. The spillway is a broad concrete crest which spills directly into the creek, via a steep fall in a concrete channel. There are currently no screens or trash racks to prevent fish movement over the spillway.

This risk assessment is based on the Winterfill - Option 2. It should be noted that relevant approvals to undertake this option are being progressed by Parks Victoria with Melbourne Water. In the event that these approvals are delayed or not received or if this option is not adopted then the associated risk of spill would be greatly reduced. As highlighted under the climate change scenario Bittern would never spill and Devilbend would spill only 1 in 12 years.

Conversely, if Option 1 is considered in management of environmental flows to Devilbend Creek then the frequency of spills and mitigation measures proposed would need to be reviewed to ensure the risks are managed.

Figure 2.4 Devilbend and Bittern Reservoir water levels under historical and climate change streamflow scenarios (GHD 2008). Note: red lines indicate reservoir spill levels.

Figure 2.5 Devilbend and Bittern Reservoir water levels under Base and 2 Management streamflow scenarios (GHD 2008). Note: red lines indicate reservoir spill levels.

2.2 Devilbend and Balcombe Creek System (potential receiving water)

The Devilbend and Balcombe Creek System are the potential receiving waters for overflows of the Devilbend and Bittern Reservoirs. It is a small creek which arises near Mt Eliza and flows into Port Phillip Bay near Mt Martha. The majority of Devilbend Creek flows through largely cleared farmland, there is dense but narrow riparian habitat and good quality fish habitat. Devilbend Creek and its tributary streams Balcombe and Tuerong Creeks have suitable freshwater habitat provided by the tree-lined creeks, deep pools and constructed lakes in the system. Melbourne Water (2007) rate these streams as having overall moderate environmental value but aquatic life being rated excellent and good water quality, vegetation, habitat and stream stability. The sediment is predominantly muddy or silty substrates in the freshwater reaches and sandy substrate in the estuary (ANGA 2005).

The water quality of Balcombe Creek has been monitored by Melbourne Water monthly since 1993 which indicates that the oxygen and temperature environment is within the tolerance range of both trout species with only 1 recording below the critical minima as indicated for trout of 2.5mg/L (Molony 2001). This measure occurred in March 2009 and very few low DO events (below 4mg/L) in the entire record. Of the 16 years of record, it indicates that a record of below 2.5mg/L is less than an 1 in 100 year event and the 10^th percentile is 4.21 mg/L (meaning it is better than 4.21 mg/L for 90% of the time). This indicates that Balcombe Creek is a well oxygenated system. The temperature of the stream is always within suitable temperatures for both trout species (see section 2.3.2).

It should be noted that this data is from Balcombe Creek and that conditions may be different in Devilbend Creek. However, it has been assumed water quality is similar as factors which would negatively impacts upon water quality (lack of riparian vegetation, shallow pools) are countered by positive factors (more reliable, historically, catchment streamflows along Devilbend Creek, Melbourne Water 2007).

Table 2.2 Temperature and Dissolved Oxygen Statistics (data supplied by Melbourne Water) for Balcombe Creek

Balcombe Creek has a wider riparian zone but a smaller and less reliable (in terms of freshwater flows) catchment. Devilbend Creek has historically had a reliable flow, with most flows diverted around the upstream Reservoirs. This may change with some of the higher flows being diverted into the Reservoirs in future management scenarios but base flows should remain strong, although diminished due to climate change (GHD 2008). The system contains Short-finned Eel, Common Galaxias, Spotted Galaxias, Flat-headed Gudgeon, Dwarf Galaxias and Southern Pygmy Perch (ANGFA 2005).

The Devilbend, Tuerong and Balcombe Creek have been known as one of the more reliable systems for Dwarf Galaxias on the Mornington Peninsula (McGuckin 2007) until the drought and no fish have been found in surveys of the catch drains or the reservoirs (McGuckin 2007; Thompson and Walker, In Prep.). Figure 2.6 illustrates the distribution of Dwarf Galaxias within the catchment over time. Dwarf Galaxias have never been captured within either Devilbend or Bittern Reservoirs despite suitable habitat being present within the reservoirs for this species. Nonetheless, surveys in the recent years have located some populations in Balcombe Creek and Tuerong Creek, downstream of Devilbend Reservoir (ANGFA 2005, SEITA 2008). The Dwarf Galaxias which were found in Tuerong Creek in 2007 were part of the EES studies for the Frankston ByPass (SEITA 2008).

The estuarine section of Balcombe Creek also contains Yellow-eye Mullet, Trevally and Black Bream. Estuary Perch have been recorded in nearby estuaries and are possibly within the Balcombe Creek estuary as well. The Balcombe Estuary is one of the highest value estuaries in the Port Phillip and Westernport Bays region (BERG 2008; Melbourne Water 2007).

Figure 2.6 Dwarf Galaxias distribution (adapted from McGuckin 2007) updated with additional identification locations of surveyed fish in 2005 and 2007

2.3 Fish Species Proposed to be Translocated

Three species are proposed for translocation:

• Australian Bass, Macquaria novemaculeata
• Rainbow Trout, Oncorhynchus mykiss
• Brown Trout, Salmo trutta

2.3.1 Australian Bass

Australian Bass is a native species but its range extends along the east coast of Australia around to Wilson's Promontory (i.e. its natural range does not extend as far west as Port Phillip Bay). Australian Bass have high salinity tolerances (up to sea water) but spend most of their time in freshwaters but eggs or sperm do not survive in freshwaters. Australian Bass breed at temperatures between 14-19^oC, temperature tolerances of adults unknown but they inhabit areas of fast flowing waters and are likely to be quite cold tolerant.

The diet of Australian bass varies significantly between habitat and season, however insects, fish and large crustaceans are the most important prey types (Harris 1985). Harris (1985) found that almost every available prey type was included in the diet of Australian Bass such as fish (the most important food recorded in Australian Bass); insects; crustaceans; and terrestrial vertebrates (such as skinks, frogs and birds) and plant material.

It is unknown at present where the source of Australian Bass will be derived. It has been assumed for the purposes of the risk assessment that all fish will be produced in a facility which has an ongoing stock health surveillance program that meets the department requirements and where there have been no unexplained disease outbreaks at the source facility in the past 24 months. In addition, because of the uncertainty of the source of the Australian Bass stock, it has been assumed that the Australian Bass to be translocated would be assessed (using a sub-sample of the consignment being sacrificed and tested for the presence of notifiable diseases) and provided with a certificate of stock health.

2.3.2 Brown and Rainbow Trout

The salmonid species (Brown and Rainbow trout) are not native to the region but are stocked extensively in Victoria for recreational or commercial purposes. For optimum growth and survival, Brown & Rainbow trout require cool, well oxygenated waters of low salinity with Rainbow Trout having slightly higher temperature tolerances compared to Browns. However, both species are adaptable and broadly tolerant of water quality and temperature conditions enabling them to inhabit a relatively wide range of systems within SE Australia (Arthington & McKenzie 1997).

Molony (2001) summarised the environmental requirements of both Trout species and these show that the critical thermal maxima (CTM) for Rainbow Trout is approximately 24 - 26°C, 10-22°C as an optimal range. Brown trout are reported as having slightly lower maximum CTMs of up to 23.5°C - 26.7°C, with an optimal range of 8 - 17°C. Therefore, both species of trout can tolerate a range of high temperatures with Rainbow Trout tolerating slightly higher temperature than Brown Trout, especially when reproductive success is examined (Molony 2001). Both species can withstand salinities up to 30 (just below sea water). Dissolved oxygen is one of the most critical requirements for trout survival with neither species found in waters below 2.5 mg/L though generally above 4 or 5 mg/L is required for successful growth and survival (Molony 2001).

Brown and Rainbow Trout both feed on a variety of animals including aquatic macro-invertebrates, (crustaceans, molluscs, insects), small fishes and any terrestrial insects which fall into the water.

Brown Trout live for up 22 years whereas Rainbow have maximum ages of up to 11 years (Fishbase 2009). In practice, most Brown Trout live up to 4-5 years (or longer) with Rainbow Trout perhaps only 3-4 years but occasional individuals do survive up to 14 years (for Brown Trout) and somewhat less for Rainbow Trout (Douglas and Kylie 2004; Kylie & Douglas 2008). Brown Trout juveniles (> 0+) are known to be highly dispersive, typically moving downstream from upper sub-catchments with age (Davies 2005).

Stock would be sourced from the Snobs Creek aquaculture facility and stocked as yearlings 150-300gm. The source of the stock will be aquaculture bred fish (in the case of Salmonoids either the facility at Snobs Creek or suppliers to Fisheries Victoria). It has been assumed for the purposes of the risk assessment that all fish will be produced in a facility which has an ongoing stock health surveillance program that meets the department requirements and where there have been no unexplained disease outbreaks at the source facility in the past 24 months.

3 Project Methods

The following project plan consisted of 5 major tasks:

• Task 1: Review and Scoping
• Task 2: Information Review
• Task 3: Field Inspection
• Task 4: Risk Assessment
• Task 5: Reporting

The following project details detail the tasks and the outputs.

3.1 Task 1: Review and Scoping

The project began with an inception meeting with Fisheries Victoria to ensure the objectives of the project were understood, and all the relevant background information was collected and collated. The meeting agreed upon the major tasks and timeframes to enable the project plan to be refined.

3.2 Task 2: Information Review

The second task developed an "information log" (or Info-Log), to document the reports and information resources available to the project. The Info-Log was used to summarise which documents were obtained for the project. The Info-Log was updated throughout the project. This information was reviewed and a list of potential issues identified. This list of issues was assessed in the field and later in an internal consultant's workshop.

3.3 Task 3: Field Inspection

The field inspection allowed an onsite meeting with Fisheries Victoria and Parks Victoria, was used to confirm the list of issues and to inspect the site. The site inspection examined the system in terms of suitable habitat for the fish, determined the connectivity with up-stream and down-stream waterways, identified potentially affected waters, and assessed any identified issues. These issues were then fed into the preliminary risk assessment.

3.4 Task 4: Risk Assessment

The risk assessment followed the Victoria EPA-developed Ecological Risk Assessment process, which has six steps:

1. Develop an explicit statement of investigation (problem formulation) for the risk assessment;
2. Identification of the values associated with the waters to be stocked and any connected waterways;
3. Identification of the threats posed to the values associated with the proposed translocations;
4. Undertake a preliminary risk assessment with existing information and local knowledge;
5. Documentation of gaps identified and assumptions made during the process; and,
6. Recommendations for appropriate risk mitigation actions to protect values and reduce threats to these values.

The Translocation Risk Assessment proforma which is used for assess submissions to the Translocation Evaluation Panel was filled out as part of step 4 and shown in Appendix 1.

In this study, the risk analysis was semi-quantitative, that is, risks were ranked, based on known data or literature and extrapolated to the site. The analysis is therefore described as "preliminary", to distinguish it from a quantitative analysis. The preliminary risk analysis required the determination of a consequence level and likelihood (or probability) rating of each threat causing any impacts each value. The qualitative nature of the analyses required documentation of the logic used in allocating ratings. This is provided in the sections below.

3.4.1 Identifying and incorporating the values

Key values of the site and its surrounding catchment were identified through a review of relevant literature, discussions with site managers, and a site inspection.

The risk analysis process is designed to qualify the risks associated with the translocation of the fish species. The process uses a qualitative analysis to describe the magnitude of the potential consequences and the likelihood that the consequences may occur. This form of analysis is broadly accepted for environmental purposes (Australian Standards, 2004) because of the limited knowledge of ecosystem thresholds and ecosystem responses associated with natural resource management actions.

In this risk assessment, a consequence table was designed based on the selected endpoints (refer Section 5), displaying a range of severity levels of consequence to each endpoint. The consequence ratings for each threat were allocated on the basis of the potential for impact upon each value – in essence a "worst-case scenario" rating. A likelihood scale was also derived, categorising the probabilities of each threat being realised. The consequence and likelihood tables are presented in a later section (Section 7 – Risk Assessment). The tables were used in a workshop of the authors to determine the risks of the stocking proposals, and then rank the species in order of risk rating (risk prioritisation). It is important to recognise that this approach provides a general guide only, as there can be problems associated with allocating a numerical value to a categorical ranking. For this reason, the risk rankings are considered as indicative only, with lines of logic being presented to facilitate examination of the decision-making.

3.4.2 Risk Characterisation

The outcomes of the preliminary analysis were subsequently described in terms of their relationship to the site and its surrounding catchment (risk characterisation). This process defined what the risk analysis means for the creeks. The process is completed with discussion of uncertainties and consideration of limitations resulting from knowledge gaps and assumptions.

3.4.3 Management Action Recommendations

The key threats, their risk ratings and associated knowledge gaps/assumptions were subsequently tabulated with recommended management actions for risk reduction. Management actions can include ameliorative actions to lessen risks, monitoring (with adaptive management planning) and/or further assessments.

3.5 Task 5: Reporting

The draft risk assessment was produced to allow written and verbal comments to be made on the report. In order to facilitate this, a formal presentation was made in which the report's findings were presented to key Fisheries Victoria staff after the draft document was delivered. The risk assessment was finalised by responding to comments received.

4 Values & Threats

4.1 Values

A list of values for the Devilbend and Bittern Reservoirs, and the Devilbend and Balcombe Creeks was restricted to those values that could potentially be impacted by the introduction of brown trout, rainbow trout and Australian Bass. These included:

1. The existing fish community in Devilbend & Bittern Reservoirs (the primary value being potential angling opportunities and natural ecosystem value)
2. The current fish community in Devilbend Creek and Balcombe Creek system (the primary value being the natural ecosystem apart from the threatened Dwarf Galaxiid)
3. A significant fish species (Dwarf Galaxiid) in Devilbend Creek and Balcombe Creek system (the primary value being the continued support for species listed as nationally vulnerable)
4. The White-bellied Sea Eagle (a threatened species within Victoria) nest within the Devilbend Reserve (this also represents other significant bird species which feed upon fish)
5. The Blue-billed Duck (a threatened species within Victoria) population that uses Devilbend Reservoir (this also represents other significant bird species which feed upon macroinvertebrates)

The potential pathways of the key threats posed to each value, through stocking with brown trout and rainbow trout and Australian Bass are listed in Sections 4.1.1 to 4.1.5.

4.1.1 Current fish community in Devilbend & Bittern Reservoirs (threats to the potential angling value and natural ecosystem value). This value could be impacted through the following pathways:

• Mature Brown Trout, Rainbow Trout and Australian Bass may prey upon existing fish species in Devilbend & Bittern Reservoirs, changing the structure of current fish assemblages in the reservoirs and impacting on the angling quality and natural ecosystem value of the water bodies;
• Brown Trout, Rainbow Trout and Australian Bass may compete with existing fish species in Devilbend & Bittern Reservoirs, leading to the loss or reduction of some species; and
• Stocking of Brown Trout, Rainbow Trout and Australian Bass may introduce disease and/or parasites to the waterbodies, threatening existing fish species in the reservoirs.

4.1.2 Current fish community in Devilbend Creek and Balcombe Creek system (must at least maintain current ecological condition). This value could be impacted through the following pathways:

• Mature Brown Trout, Rainbow Trout and Australian Bass may prey upon existing fish species in Devilbend Creek and Balcombe Creek system
• Stocked juvenile Brown Trout and Rainbow Trout and stocked and breeding Australian Bass may compete with existing fish species in Devilbend Creek and Balcombe Creek system
• Stocking of Brown Trout, Rainbow Trout and Australian Bass may introduce disease and/or parasites to the waterbodies, threatening existing fish species in Devilbend Creek and Balcombe Creek system

4.1.3 Significant fish species (Dwarf Galaxiid) in Devilbend Creek and Balcombe Creek system. This value could be impacted through the following pathways:

• Mature Brown Trout, Rainbow Trout and Australian Bass may prey upon Dwarf Galaxiids in the Devilbend Creek and Balcombe Creek system
• Juvenile Brown Trout, Rainbow Trout and Australian Bass may compete with Dwarf Galaxiids in Devilbend Creek and Balcombe Creek system
• Stocking of Brown Trout, Rainbow Trout and Australian Bass may introduce disease and/or parasites to the waterbodies, threatening Dwarf Galaxiids in Devilbend Creek and Balcombe Creek system

4.1.4 White-bellied Sea Eagle population. This value could be impacted through the following pathways:

• Introduction of Brown Trout, Rainbow Trout and Australian Bass may impact on current fish community in Devilbend Reservoir, reducing food availability for resident White-bellied Sea Eagles
• Introduction of Brown Trout, Rainbow Trout and Australian Bass may impact on current fish community in Devilbend Creek and Balcombe Creek system, reducing food availability for resident White-bellied Sea Eagles
• This would also represent other significant bird species which feed upon fish.

4.1.5 Blue-billed Duck population. This value could be impacted through the following pathways:

• Introduction of Brown Trout, Rainbow Trout and Australian Bass may impact on current macroinvertebrate biomass in Devilbend Reservoir, reducing food availability for Blue-billed Duck
• This would also represent other significant bird species which feed upon macroinvertebrates

4.2 Threats

The threats presented below were used in the qualitative risk assessment (see Section 7). Although largely a rewording of some potential pathways presented in Section 4.1 (above), each threat that was used in the qualitative risk assessment needs to be described as a discrete component.

4.2.1 Predation in Reservoirs (fish)

The introduction of the three piscivorous species to the reservoirs creates a predation threat to the populations of existing species. All three species are aggressive feeders and could feed on juveniles and adults of smaller fish species within the reservoirs.

4.2.2 Predation in Reservoirs (macroinvertebrates)

All three species proposed for introduction are known to be predators of macroinvertebrates. Diets for the introduced species will vary with age of the individuals and would range from small zooplankton being consumed by juveniles, through larval stages of larger insects, to larger crustaceans such as yabbies.

4.2.3 Competition in Reservoirs (fish impacts)

The proposed species for introduction have the potential to compete with existing species in the reservoirs for food and habitat. In particular, the existing population of introduced Redfin Perch may have overlapping resource requirements with the three proposed introductions. To some extent, juveniles of the proposed introductions may compete with the smaller native species in the reservoirs.

4.2.4 Introduction of Disease to Reservoirs

There is a potential threat that an individual of any of the species proposed for introduction may be carrying pathogenic bacteria, viruses, fungi or protozoa. If any pathogens establish, they may impact on the biota of the reservoirs.

Escape from Reservoirs

Escape from the reservoir is the key process for the threats 4.2.5 to 4.2.8 to occur and this is therefore described. During large rainfall events, the reservoirs may overflow and discharge into the Devilbend Creek. During these overflows there is a potential threat of biota from the reservoirs to be washed into the Devilbend and Balcombe Creek system. The probability of fish escaping will depend upon frequency of spills, the type of species (Bass are more likely to be triggered to swim downstream during a rain event [Harris 1985 and Harris et al. 1986]; a percentage of young brown trout are also triggered to swim downstream in such conditions [Davies 2005]) and the magnitude of the spill.

4.2.5 Predation in Creeks

Similar to predation in the reservoirs, there exists the potential threat of predation by the proposed introductions upon the juveniles and adults of smaller fish species within the creek systems.

4.2.6 Predation in Creeks (macroinvertebrates)

Predation by the proposed introductions upon the creek macroinvertebrate fauna is also a risk to the values identified in the study. In particular, any juveniles that escape from the reservoirs or offspring of adults (in the case of Australian Bass) that escape pose a threat to the macroinvertebrate fauna of the system of small receiving creeks and the estuary.

4.2.7 Competition in Creeks (juvenile stages of escapees)

The adults of the species proposed for introduction are unlikely to compete with the fish species present in the creek systems. However, juveniles of the proposed introductions may compete with the current fish fauna within the creeks.

4.2.8 Introduction of Disease to Creeks

Similar to the reservoirs, there is a potential threat that an individual of any of the species proposed for introduction may be carrying pathogenic bacteria, viruses, fungi or protozoa. Any introduced pathogens could escape to the creek systems and establish, subsequently impacting upon the creeks' biota.

4.2.9 Reduced food for Sea Eagles

Any impact on the number of individuals or species within the reservoirs or creek systems has the potential to threaten the food resources of the resident White-bellied Sea Eagles of the site.

4.2.10 Reduced food for Blue-billed Duck

Any impact on the number of individuals or species of macroinvertebrates within the reservoirs or creek systems has the potential to threaten the food resources of the resident Blue-billed Duck population of the site.

4.3 Ecosystem level impacts

Stocking alone (at the levels contemplated with the large Devilbend and Bittern Reservoirs) is unlikely to result in the complex ecosystem level impacts such as restructuring food webs and changing nutrient cycling (Lloyd et al 1998 & 2000; Berg et al 1997; Benndorf 1990). This is further supported by the already large population of the piscivore, Redfin, present within the reservoirs (McGuckin 2001) and the likelihood that further predator introductions would not lead to an increased predatory impact on native fish or other ecosystem components.

In fact, Benndorf (1990) showed that high piscivore diversity increases stability and reliability of food web manipulations. Drenner (2002) found most studies did not find piscivore effects on phytoplankton biomass and therefore did not support the trophic cascading interactions hypothesis. In Lake Hume, high densities of planktivorous stages of alien European perch (Perca fluviatilis) did not translate into declining trends for plankton in the reservoir (Matveev and Matveeva 2005). The authors concluded that the community was structured by nutrient levels and water level fluctuations (Matveev and Matveeva 2005).

Clearly, these type of impacts are complex to understand and little information is available to provide valid input to this semi-quantitative risk assessment.

5 Statement Of Investigation

Within a risk assessment, a statement of investigation clearly states the question being asked of the assessment and defines the measurement endpoints that can be used to assess impacts to the key values of the ecosystem.

The investigative question for this risk assessment is:

What are the risks posed to the current assemblages and populations of Devilbend and Bittern Reservoirs and the Devilbend/Balcombe Creek systems by translocating (introducing) Brown Trout, Rainbow Trout and Australian Bass?

Using the values and threats identified in Section 4, the specific values and their measurement endpoints are presented in Table 5.1.

Table 5.1: Measurement endpoints for each value in the risk assessment

6 Conceptual Model

The conceptual model for the translocation of Brown Trout, Rainbow Trout and Australian Bass to Devilbend and Bittern Reservoirs (Figure 6.1 & 6.2) displays pathways for interaction within the reservoirs as well as pathways for interactions in the event of these species escaping to the Devilbend and Balcombe Creek system. Each set of interactions includes the possibility of predation, competition and disease introduction. The pathways ultimately link to the five values listed in Table 5.1.

The stocking of the reservoirs with piscivorous predator species creates a potential threat to the structure and functioning of the current fish assemblages within the reservoirs, through predation upon the smaller species that occur in the reservoirs. Similarly, there is a potential threat of the stocked species competing with existing species for food (e.g. macroinvertebrates) and habitat, thereby impacting on the current fish assemblages. Clearly, the introduction of diseases to the water bodies would be a potential impact on the fish assemblages within the reservoirs and therefore subsequently fish eating birds (such as the Sea Eagle).

The potential for the stocked fish to reduce the biomass of aquatic macroinvertebrates within the reservoirs (represented by the brown box in Figure 6.1 & 6.2) also has the potential to impact the population of Blue-billed Duck and other like feeders that use the reservoirs, as the food for these birds includes the larvae of midge, caddisflies and dragonflies (Stamation and Lyon 2009).

The potential interactions between threats and values increase in the event of escapes from the reservoirs to the receiving creek systems. The complexity of the interactions also increases in the event of an escape. Two reasons for this increase in number and complexity of interactions are: (i) the fish assemblages in the creek systems are valued more as functioning aquatic ecosystems rather than for angling; and (ii) there is a significant fish species, in limited numbers, within these creek systems.

Figure 6.1: Conceptual Model of threat pathways to values in Devilbend Reservoir study area

Figure 6.2: Conceptual Model of threat pathways to values in Bittern Reservoir study area

7 Risk Analysis

The qualitative risk analysis required the determination of a consequence level and likelihood (or probability) of each threat causing any impacts on the values, for each of the three species proposed for introduction. The qualitative nature of the analyses required documentation of the logic used in allocating ratings. This is provided in the sections below. The introductions were evaluated by allocating a risk score to each species under each threat to relevant values. The risk scores were calculated as the product of consequence (Table 7.1) and likelihood (Table 7.2) under typical conditions within the reservoirs.

Table 7.1: Consequence levels of impacts on the values of the study area

Table 7.2: Likelihood ratings for threats to the values of the study area

Risk is simply calculated by the multiplication of both likelihood and consequence scores and the resulting score is qualitatively assigned into 3 categories of low, medium and high risk (Table 7.3).

Table 7.3: Final Risk Scores for the Assessment

In this study, the risk posed by stocking the three proposed fish species was firstly calculated individually for each species.

Table 7.4 provides the consequence and likelihood scores for the risk assessment for Devilbend Reservoir for Brown Trout (BT), Rainbow Trout (RT) and Australian Bass (AB).
Table 7.5 provides the risk scores for the risk analysis for Devilbend Reservoir for the three species.

The rationale for the allocation of the consequence and likelihood scores for each threat is presented for each value in Sections 7.1 to 7.5, below. The results for Bittern Reservoir were identical to those of Devilbend Reservoir, with the exception of the Australian Bass results, as there is no proposal to stock Bittern Reservoir with Australian Bass.

7.1 Risk analysis results and rationale for the value 'Fish Community in Reservoirs'

As the geographic scope of this value was the reservoirs, the five threats relating to impacts on the creek systems were not assessed for this value (shaded grey in Tables 7.4 and 7.5). The first threat evaluated was 'Predation in Reservoirs (fish impacts)'. As noted in Table 5.1, the measurement endpoint for this value is the angling worth and natural ecosystem value. Given that the reservoirs currently contain a large, predatory exotic fish population (Redfin Perch; McGuckin 2001), the introduction of one or more species of piscivorous fish was considered to have a low or moderate consequence (rating 1 for Rainbow Trout because of shorter lifespan and rating 2 for the other two species) on the angling worth and the natural ecosystem value of the reservoirs.. In order for stocked species to become established, it is likely that some the Redfin Perch will be displaced by the fish introductions and predator pressure would not increase. Therefore it would be of no consequence to the angling worth and natural ecosystem value of the reservoirs. Redfin are considered as voracious, or more so, than all three proposed species. Further, the introductions would be considered better angling species, resulting in an improvement to the angling worth. The likelihood of any negative impacts was rated 'unlikely' (rating 2). This led to a Low risk score for all three proposed species (2 for Rainbow Trout; 4 for Brown Trout and Australian Bass).

Similar to the threat of fish predation, the threat of macroinvertebrate predation impacting on the current fish community is also low due to the existing population of Redfin Perch in the reservoirs. Redfin Perch are predators of aquatic macroinvertebrates and the consequences of increased competition with the proposed introduced species were considered low or moderate consequence in terms of angling worth (rating 1 for Rainbow Trout because of shorter lifespan and rating 2 for the other two species). The likelihood of any negative impacts was rated 'unlikely' (rating 2), leading to a Low risk score for all three proposed species (2 for Rainbow Trout; 4 for Brown Trout and Australian Bass).

The threat to angling and natural ecosystem value posed by overall competition (for habitat as well as food sources) between the proposed introductions and the current assemblages has a slightly higher consequence (rating 2) for the Brown Trout and Australian Bass, due to their longevity. In contrast, the lower life expectancy of Rainbow Trout kept the consequence rating to 1 (low). For all three species, any impact was considered 'unlikely' due to the existing population of Redfin Perch in the reservoirs. This resulted in a risk score of 4 (Low) for the Brown Trout and the Australian Bass, and a score of 2 (also Low) for the Rainbow Trout.

A key assumption of this assessment is that rigorous procedures will be followed in any translocation, including adherence to the procedures set out in the Protocols for the Translocation of Fish in Victorian Inland Public Waters (Department of Environment and Primary Industries 2005) for eliminating introductions of pathogens. Assuming these procedures are followed correctly, the possible consequences and likelihood of that threat being realised were rated as 'low' and 'highly unlikely' respectively (rating 1 and 1), leading to Low risk score of 1.

As discussed above, the introduction of additional fish species is considered a low risk to the existing angling or natural ecosystem value of the reservoirs. The proposed introductions will add new species to the existing fish community but is not expected to impact the availability of fish for angling or natural ecosystem values of the reservoirs. This is also the case for the availability of fish as prey for the site's resident White-bellied Sea Eagles. As the site is likely to be one of several food sources for the Sea Eagles and because there is a low risk to changes in the fishing worth of the site, the introductions will have a low consequence on the Sea Eagle residents. Similarly, the likelihood of any impacts is 'very unlikely' resulting in a Low risk score of (1 X 1 =) 1.

7.2 Risk analysis results and rationale for the value 'Fish Community in Devilbend Creek and Balcombe Creek system'

As the geographic scope of this value was the creek systems, the four threats relating to impacts on the reservoirs were not assessed for this value (shaded grey in Tables 7.4 and 7.5). This value is markedly different to the previous, as the focus is on the ecological condition of the fish community, downstream of the reservoirs . The creek systems are small, with dense riparian cover and containing small native species (e.g. Hardyhead, Gudgeons, Pygmy Perch, and Dwarf Galaxias) and shortfinned eels.

Review of the hydrological modelling for the site (GHD 2008) suggests that the reservoirs are expected to overflow in 55% of years under the preferred management scenario (see section 2). The risks of predation and competition from translocated fish are underpinned by escape of fish.

Without spill or escape mitigation this creates a high likelihood of the proposed introduced species escaping and surviving to the receiving creek systems. If the escaped individuals are juvenile, they are likely to compete with the smaller native species in the creek systems. Consequences of an escape are likely to be medium- or short-term for the two Trout species, as they will not breed if they escape. For the Australian Bass, however, there is a possibility that they might breed and/or hybridise with Estuary Perch in the Balcombe Creek estuary (if present) or those in nearby estuaries. Even a few individuals of these predatory species, if they were to become established to a site within a creek, they are likely to exert an ongoing predatory pressure on small native fish for 2-4 years (in the case of trout) and even longer for Australian Bass, even if none of the breed. However, there is some chance that Australian Bass will breed and establish self-sustaining populations.

In the event of escape the introduction of new predators to the creek system are likely to have ongoing consequences for the existing fish community (given the low numbers of exotic predators already present within the creek system). The longevity of Brown Trout and the potential for Australian Bass to breed within the system means that the consequences for these impacts could be long-term (rating 3), whereas the shorter life-span for Rainbow Trout suggests that consequences would more likely be restricted to the short- to medium-term (rating 2). For all three species, the likelihood of an impact was assessed as highly likely (likelihood rating 4) given the predatory nature of the species and their high likelihood of escape. Therefore the risk rating for impacts of predation on the creek fish communities was calculated as 12 (High) for the Brown Trout and Australian Bass, and 8 (Moderate) for the Rainbow Trout.

Similar to the threat of fish predation, the threat of macroinvertebrate predation impacting on the current fish community of the creek systems is also highly likely (given the low numbers of exotic predators already present within the creek system) for all three species (likelihood rating 4). All three species are considered voracious feeders and the longevity of Brown Trout and the potential for Australian Bass to breed within the system means that the consequences for these impacts could be long-term (rating 3), giving both these species a High risk rating of 12. The shorter life-span for Rainbow Trout suggests that consequences would more likely be restricted to the short- to medium-term (rating 2), resulting in a risk rating of 8 (Moderate) for the Rainbow Trout.

The threat of juvenile stages of the escaped species competing with native fish in the creek systems was considered 'likely' (likelihood rating = 3) to have some impact. Brown Trout and Australian Bass have a tendency to travel downstream on flow events, so this makes it more likely that these species will escape from the reservoir if it spills (Harris 1985, Harris et al. 1986; Davies 2005). The two trout species would be new competitors. Although they will not breed within the creek systems and any escaped fish will not remain for more than the short-to medium term, they could be replaced by reservoir overflows, having a potential long-term impact and resulting in a consequence rating of 3. The Australian Bass also received a consequence rating of 3 as although it is a native species similar to other possible native competitors such as Estuary Perch, the fact that it may be able to breed within the system means that it may have ongoing and higher numbers of juveniles in the systems. Therefore all three species received a risk rating of 9 for the threat of competition with the creek fish communities. This risk is reinforced by the stocking strategy likely to see fish stocked each year for 5 years, in the first instance.

Similar to the threat of diseases impacting the fish community at the reservoir, the possible consequences and likelihood of that threat being realised in the fish communities of the creek systems were rated as 'low' and 'highly unlikely' respectively (rating 1 and 1), leading to a Low risk score of 1. The fact that the fish communities of the creek system are unlikely to provide substantial food for the White-bellied Sea Eagles resulted in there being a 'highly unlikely' likelihood rating and a low consequence of impacts upon the fish community with respect to this value.

7.3 Risk analysis results and rationale for the value 'Dwarf Galaxiid population in Devilbend Creek and Balcombe Creek system'

Again, with the geographic scope of this value being restricted to the creek systems, the four threats relating to impacts on the reservoirs were not assessed for this value. For all three proposed species, the likelihood of escaping and individuals establishing was regarded as highly likely and as all the translocated fish are long- lived predators (in comparison to the short-lived Dwarf Galaxiids), the consequences to the Dwarf Galaxiid population could be long-term.

The introduction and survival of new predators to the creek system will have consequences for the existing Dwarf Galaxiid population. The longevity of Brown Trout and the potential for Australian Bass to breed within the system means that the consequences for these impacts could be long-term (rating 3), whereas the shorter life-span for Rainbow Trout suggests that consequences would more likely be restricted to the short- to medium-term (rating 2). For all three species, the likelihood of an impact was assessed as highly likely (likelihood rating 4) given the predatory nature of the species and their high likelihood of escape. Therefore the risk rating for impacts of predation on the Dwarf Galaxiid population was calculated as 12 (High) for the Brown Trout and Australian Bass, and 8 (Moderate) for the Rainbow Trout.

The impact, rationale and assessment outcomes for the impact of the introduced species upon macroinvertebrates for Dwarf Galaxiids are the same as for the general fish community of the creeks. The threat of reduced macroinvertebrates due to increased predation impacting on the Dwarf Galaxiids is considered highly likely for all three species (likelihood rating 4). Similarly, the longevity of Brown Trout and the potential for Australian Bass to breed within the system indicates potential long-term consequences (rating 3), giving both these species a High risk rating of 12. The shorter life-span for Rainbow Trout indicates that consequences would more likely be restricted to the short- to medium-term (rating 2), resulting in a risk rating of 8 (Moderate) for the Rainbow Trout.

The threat of juvenile stages of the escaped species competing with Dwarf Galaxiids in the creek systems was considered 'likely' (likelihood rating = 3) to have some impact. The two trout species would be new competitors. Although they will not breed within the creek systems and any escaped trout will not remain for more than the short- to medium term, they could be regularly replaced by reservoir overflows, having a potential long-term impact and resulting in a consequence rating of 3. The Australian Bass also received a consequence rating of 3 as although it is a native species similar to other possible native competitors such as Estuary Perch, the fact that it may be able to breed within the system means that it may have ongoing and higher numbers of juveniles in the systems (the rating is lessened somewhat due to the fact that this is a predator similar to locally native fish (Estuary Perch). Therefore all three species received a risk rating of 9 for the threat of competition with the Dwarf Galaxiid population.

Similar to the threat of diseases impacting the fish community at the reservoir, the possible consequences and likelihood of that threat being realised in the Dwarf Galaxiid population in the creek systems were rated as 'low' and 'highly unlikely' respectively (rating 1 and 1), leading to a Low risk score of 1. The fact that the Dwarf Galaxiid population of the creek system are unlikely to provide substantial food for the White-bellied Sea Eagles resulted in there being a 'highly unlikely' likelihood rating and a low consequence of impacts upon the fish community with respect to this value.

7.4 Risk analysis results and rationale for the value 'White-bellied Sea Eagle residents'

The introduction of additional fish species to the reservoirs and creek systems is highly unlikely (rating 1) to lead to any impact on bird populations (consequence rating 1) as establishment of new predators within the system will be a replacement of existing predators. This is more likely to result in a change of the community assemblage than a change in structure or biomass. For this reason, all risks to the White-bellied Sea Eagle residents received a risk rating of 1 (Low risk). This would also apply to other piscivorous birds in the reservoir.

7.5 Risk analysis results and rationale for the value 'Blue-billed Duck Population'

The only threat from the proposed introductions that was considered a possible risk to Blue-billed Ducks was the increased predation of macroinvertebrates in the reservoirs. This is because Blue-billed ducks are common on the reservoirs and macroinvertebrates form part of their diet. However, given that the reservoirs currently contain a large predatory fish population (Redfin Perch), the introduction of one or more species of piscivorous fish was considered to have a low consequence (rating 1). If the introductions resulted in the Redfin Perch being replaced as the top predator, it is unlikely (likelihood rating 2) to have any consequence upon the macroinvertebrates and hence the Blue-billed Ducks. This led to a Low risk score of 2 for all three proposed species. This would also apply to other macro-invertebrate eating birds in the reservoir.

Table 7.4: Risk Assessment for Devilbend Reservoir (Consequence X Likelihood). BT = Brown Trout, RT = Rainbow Trout, AB = Australian Bass

Table 7.5: Risk Scores for Devilbend Reservoir. BT = Brown Trout, RT = Rainbow Trout, AB = Australian Bass

7.6 Knowledge Gaps and Assumptions

A lack of complete knowledge and understanding of environmental systems requires the use of assumptions when undertaking risk assessments. In this study, the major knowledge gaps centre on the likelihood of individuals escaping, surviving escape from the reservoirs and the successful establishment of escapees in the downstream environments. This is particularly relevant in the context of the existing piscivore (Redfin Perch) in Devilbend Reservoir. McGuckin (2001) showed large numbers of Redfin Perch present within the reservoirs using a range of gear and this is used as the basis of the risk assessment. Thompson and Walker (2009) also concluded that Redfin were likely to be the dominant large bodied fish in the reservoirs despite low numbers caught in their survey (as their techniques were unlikely to reflect these population levels).

Redfin Perch have a maximum life expectancy of 22 years (Fishbase 2009 – see www.fishbase.org) with many individuals living about half this time (Dr Michael Shirley. SKM New Zealand, pers comm.). As the current probability of overflow from the reservoir is 20% (i.e. one year in five), it is reasonable to expect that this species has had the opportunity to escape and establish to the downstream environment. Surveys of the creeks systems (ANGFA 2005, McGuckin 2001, 2007, SEITA 2008) have not recorded Redfin Perch (McGuckin 2001 lists Redfin present within the creek in a table but data in his appendices shows the species was not recorded), indicating that either the surveys have been undertaken at times or places or in ways that were not conducive to catching this species, or that Redfin Perch do not escape and establish in the creek systems. A more likely scenario is that any escaped fish had died out when surveys were done. The last spill was in 1996 and the next survey in the system downstream of the reservoirs was 2001 (McGuckin 2001). Natural mortality would mean fewer fish through time, making these harder to detect.

In this risk assessment, we have assumed that there is a possibility that the proposed stocking species can escape, with or without the Redfin Perch. Brown Trout and Australian Bass have a tendency to travel downstream on flow events (Harris 1985, Harris et al. 1986, Davies 2005), so this makes it more likely that these species will escape from the reservoir if it spills. We have also assumed that if the Redfin Perch is already in the creek systems (though in low numbers) then the addition of new predators will amplify, rather than replace, predation and competition pressures on the native species in the creeks. Any additional predator species is likely to add a further pressure on the survival of Dwarf Galaxias populations, which is not currently present.

The situation is different in the reservoirs as these have large populations of Redfin, so it is assumed that introduced predators (stocked species) would need to displace some of these fish to become established, so impacts on macroinvertebrates and fish in the Reservoirs would be low compared to the current situation.

8 Conclusions

The risks associated with stocking Brown Trout, Rainbow Trout and Australian Bass into Devilbend and Bittern Reservoirs can be divided into two groups, those risks restricted to impacts within the reservoirs and those risks that may impact on the ecosystem in the downstream receiving waters. All risks assessed within the reservoir have low risk ratings for all three species. However, risks associated with escape from Devilbend Reservoir have resulted in moderate risks from Rainbow Trout and high risks from Brown Trout and Australian Bass.

In the absence of effective controls (to prevent fish escape) the stockings are likely to trigger a referral to DEWHA to assess if this is a controlled action on Dwarf Galaxias (under the EPBC Act), given the current spill frequencies. These risks would be higher if the proposed management strategy of breaching the catch drains is implemented. There is a documented process for consideration of any referral under the EPBC Act. However, these risks could be reduced to low and not significant, if risk mitigation measures (such as those based on the strategy outlined below) are implemented, and these substantially reduce the likelihood of stocked species escaping from the reservoirs.

8.1 Risk Mitigation

A risk management strategy would be required to reduce the identified moderate and high risks to no significant impact. It is clear that a multi-management response would be most effective. This could include:

1. Reduce frequency of spills from Devilbend Reservoir to a very low probability by controlled releases through a meshed outlet structure (which would need to be designed or tested for effectiveness).
2. Investigate the feasibility of controlling escape from Devilbend Reservoir via the spillway through use of screens, implementing a multilayered approach (trash rack, followed by appropriately sized screen(s)). This control strategy (steps 1 above and this step) would need detailed engineering, hydrology and fish biologist advice, to ensure it was successful in preventing fish escape.
3. Stock one and two year old Rainbow and Brown Trout into Devilbend and Bittern Reservoirs. In the lowered risk of escape (.ie. after implementation of point 1 & 2 above) the consequence of escape would be reduced due to their average longevity surviving only 3-5 years and not being able to reproduce. The size of this species at stocking would also assist in development of effective screening.
4. Stock Estuary Perch instead of Australian Bass as the native species in Devilbend Reservoir. This species is endemic to the catchment and could be propagated from a local genetic population.

The disease risks are low from the proposed translocations if standard biosecurity procedures are followed.

Finally as stated earlier, this risk assessment is based on the Winterfill - Option 2. It should be noted that relevant approvals to undertake this option are being progressed by Parks Victoria. In the event that these approvals are delayed or not received or if this option is not adopted then the associated risk of spill would be reduced. As highlighted under the climate change scenario Bittern would never spill and Devilbend would spill only 1 in 12 years. The risks associated with the stockings would then need to be re-evaluated as some would reduce further. Further, other strategies to manage reservoir water levels may also increase spill frequency, so if these are employed then this will also affect the risks identified in this report. If the basis of future hydrological management is changed, the risks assessed would need to be reviewed.

9 References

Arthington, A.H. & McKenzie, F. 1997. Review of impacts of displaced/introduced fauna associated with inland waters, Australia. State of the Environment Technical Paper Series (Inland Waters), Department of the Environment, Canberra.

Benndorf, J. 1990. Conditions for effective biomanipulation; conclusions derived from whole-lake experiments in Europe. Hydrobiologia 200/201: 187-203.

Berg, S., Jeppesen, E. & Søndergaard, M. 1997. Pike (Esox lucius L.) stocking as a biomanipulation tool: 1. Effects on the fish population in Lake Lyng, Denmark. Hydrobiologia 342/343: 311–318.

BERG. 2008. The Balcombe Estuary Reserves. The Balcombe Estuary Rehabilitation Group Website (http://www.berg.org.au/; accessed Dec 2008).

Davies, P.E. 2005. A review of the flow and flow-related habitat requirements of Tasmanian native and introduced freshwater fish. APPENDIX E Report Series WA 07/03 of DPIW (2007). The Tasmanian Environmental Flows Framework. Technical Report. Water Assessment Branch, Department of Primary Industries and Water, Hobart, Tasmania. Technical Report No. WA 07/02.

Department of Environment and Primary Industries. 2005. Protocols for the Translocation of Fish in Victorian Inland Public Waters. Fisheries Victoria Management Report Series No. 24.

Douglas, J. & Hall, K. 2004 Lake Wendouree Fisheries Assessment. Department of Environment and Primary Industries.

Drenner, R.W., Hambright, R.K. David. 2002. Piscivores, Trophic Cascades, and Lake Management. The Scientific World JOURNAL. Vol 2: 284-307.

EPA Victoria. 2004. Risk-Based Assessment of Ecosystem Protection in Ambient Waters. EPA Victoria Publication No. 961, October 2004. ISBN 0 7306 7644 7. Southbank, Victoria.

GHD. 2008. Report on Hydrological Analyses. Devilbend Reservoir Modelling Report. GHD Report to Parks Victoria.

GROWNS, I. and JAMES, M. 2005. Relationships between river flows and recreational catches of Australian bass. Journal of Fish Biology (2005) 66, 404–416.

Hall, K. and Douglas J. 2008 Lake Modewarre Creel Survey. Fisheries Victoria Research Report. Series No. 33.

Harris, J. H. 1986. Reproduction of the Australian bass, Macquaria novemaculeata, (Perciformes: Percichthyidae) in the Sydney Basin. Australian Journal of Marine and Freshwater Research 37, 209–235.

Heard, G., Robertson, P. & Scroggie, M. (2004). The ecology and conservation status of the Growling Grass Frog (Litoria raniformis) within Merri Creek corridor. Second Report: Additional field survey and site monitoring. Wildlife Profiles Report to DSE and Yarra Valley Water.

Lloyd, L., Sandercock, C., Shirley, M. and Cottingham, P. (1998). Conceptual Models Describing Ecological Processes within Lake Mokoan. Lake Mokoan Restoration Program Technical Report No. 2/98. Report prepared for Goulburn-Murray Water by WATER ECOscience, WES Report No. 220/98.

Lloyd, L., Sandercock, C., Shirley, M. and Cottingham, P. (2000). Updated Report: Conceptual Models Describing Ecological Processes Within Lake Mokoan. Lake Mokoan Restoration Program Technical Report No. 21/00. Report prepared for Goulburn-Murray Water by AWT Victoria, Report No. 485/00.

Lloyd, L.N., Newall, P.R., Loffler, T. and Knight, C.D. 2008. Tullaroop Creek Flows Ecological Risk Assessment. Lloyd Environmental report to Central Highlands Water, Mt Waverley, Victoria.

Matveev V.F. and Matveeva. 2005. Seasonal succession and long-term stability of pelagic community in a productive reservoir. Marine and Freshwater Research 56(8) 1137–1149.

Melbourne Water. 2007. Port Phillip and Westernport Regional Health Strategy. Melbourne Water, and Port Phillip and Westernport Catchment, Management Authority, Melbourne.

Molony, B. 2001. Environmental requirements and tolerances of Rainbow trout (Oncorhynchus mykiss) and Brown trout (Salmo trutta) with special reference to Western Australia: A review. FISHERIES RESEARCH REPORT NO. 130, 2001. Fisheries Research Division, WA Marine Research Laboratories, Dept of Fisheries.

McGuckin J. (2001). Fish survey of Devilbend and Bittern Reservoirs, Devilbend and Tuerong Creeks. A report prepared by Streamline Research for Melbourne Water.

McGuckin, J. 2007. A survey for dwarf galaxias in the vicinity of Bittern and Devilbend Reservoirs (following a prolonged drought). A Streamline Research Report for Parks Victoria, August 2007.

Newall, P.R. and Lloyd, L.N. 2008. Ecological Risk Assessment of Modified Releases to the Merri Creek: Final Report. Lloyd Environmental Pty Ltd Report to Yarra Valley Water. Mt Waverley, Victoria.

SEITA. 2008. The Environmental Effects Statement (EES) for the Frankston Bypass. Southern and Eastern Integrated Transport Authority (SEITA), Melbourne Victoria.

Stamation, K. and Loyn, R (2008) Impact of recreational access and stocking on waterbirds and shorebirds – Devilbend Reserve. Arthur Rylah Institute for Environmental Research Technical Report Series No. xxx. Department of Environment and Primary Industries, Heidelberg, Victoria.

Thompson, R. and Walker, M. 2009. DEVILBEND AQUATIC HABITATS ECOSYSTEM STUDY. Parks Victoria Contracted Report # #####. School of Biological Sciences and Australian Centre for Biodiversity Monash University.

10 APPENDIX 1: TEP RISK ASSESSMENT PROFORMA

Risk Assessment Proforma – Devilbend & Bittern Reservoirs.

Risk Assessment Proforma and Instructions for the Translocation of Live Aquatic Organisms

1. Proponent and Risk Assessor Details.

1.1 Proponent

1.2 Risk Assessor

2. Escape or Release - Likelihood

2.1 Will the transport medium and equipment be treated before transport?

2.2 Will the transport medium be treated using appropriate methods after the translocation?

Important note for questions 2.3 - 2.8

Questions 2.3 -2.8 applies to closed and semi-closed systems only. If your translocation event is the stocking of organisms into a system that is defined as an open system please disregard questions 2.3 -2.8 and proceed to question 2.9. A definition of each system is provided in the Glossary of terms.

2.3 How close and accessible are nearby watercourses?

2.4 Is the facility fully enclosed and secure from unauthorised access?

2.5 Based on knowledge of the facility's waste water treatment and disposal, and the capability to contain all life stages of the organism, are any life stages likely to be released from the facility during normal operations?

2.6 Based on knowledge of the facility's waste water filtration, sterilisation and disposal, are any diseases present in the facility that are likely to escape?

2.7 Does the facility have adequate contingency plans in the event of a technical failure?

2.8 Have local environmental issues (e.g. flooding) been considered in containment planning?

2.9 What is the nature of any disease surveillance programs in the source area and/or facility?

2.10 Are there any disease, parasite or unexplained mortality issues in the source area?

2.11 Will the consignment be reliably certified free of known diseases, and if so by whom?

2.12 What is the OIE disease zoning status of the source and destination areas?

2.13 What quarantine processes and/or treatments will the consignment be subject to?

2.14 Are undesirable species (e.g. parasites, blue green algae, fish) likely to be translocated with the consignment that are not currently found in the target location?

2.15 Will the consignment be reliably certified free of undesirable accompanying species? If so, by whom?

2.16 Based on the answers to Questions 2.1 to 2.15, what is the likelihood of escape?

3. Escape / Release - Consequences

3.1 What species (including diseases and parasites) are likely to escape?

3.2 In the event of an escape, what life stages (e.g. gametes, fertilised eggs, juveniles, adults, etc.) are likely to escape?

3.3 In the event of an escape how many individuals are likely to escape?

3.4 Based on the answers to Questions 3.1 to 3.3, what are the consequences of escape?

4. Survival - Likelihood

4.1 Is the natural and/or current range of the species/genetic stock known?

4.2 Are the temperature and water quality requirements for survival known and are they available in the potential receiving waters?

4.3 Are the habitat requirements for survival known and are they available in the potential receiving waters?

4.4 Are the food requirements of the species known and are they available in the potential receiving waters?

4.5 How `natural` is the target area (some species colonise disturbed areas more effectively)?

4.6 For diseases and parasites, are suitable hosts likely to be available in the target area?

4.7 Based on the answers to Questions 4.1 to 4.6, what is the likelihood of survival?

5. Survival Consequences

5.1 Is the species endemic to the target area?

5.2 Is the species currently found in the target area?

5.3 Is the species likely to be a significant competitor/predator in the target area?

5.4 Is the species likely to alter the physical environment?

5.5 Is the species likely to destabilise local plant communities?

5.6 What effects are any released diseases or parasites likely to have in the potential receiving waters without completing their full life cycle?

5.7 Based on the answers to Questions 5.1 to 5.6, what are the consequences of survival?

6. Establishment - Likelihood

6.1 Are the environmental requirements for the completion of all stages of the life cycle known and are they available in the potential receiving waters?

6.2 For diseases and parasites, are carriers and hosts required for the completion of all stages of the life cycle known; and are they available in the potential receiving waters?

6.3 Is the ability of the species to hybridise with local species known?

6.4 Based on the answers to Questions 6.1 to 6.3, what is the likelihood of establishment?

7. Establishment - Consequences

7.1 How `natural` are the potential receiving waters (in unique or pristine areas, the consequences of establishment are likely to be considered to be more important)?

7.2 Are there any endangered or rare species in the potential receiving waters?

7.3 Is the species subject to an eradication or minimisation program in the target area?

7.4 Is the organism genetically modified?

7.5 Should the species establish in water bodies, is it likely that it can be eradicated?

7.6 Based on knowledge of the species' growth, reproductive characteristics and behaviour, is the species likely to displace local species in similar ecological niches?

7.7 Based on knowledge of the species' behaviour and physical characteristics, is it likely to be a significant predator in the potential receiving waters?

7.8 Based on knowledge of the species' behaviour and physical characteristics, is it likely to alter the physical environment in the potential receiving waters?

7.9 Based on knowledge of the species' behaviour and physical characteristics, is it likely to destabilise plant communities in the potential receiving waters?

7.10 Is the consignment of the same genetic stock as local populations?

7.11 What effects are any released diseases or parasites likely to have in the potential receiving waters?

7.12 Based on the answers to Questions 7.1 to 7.11, what are the consequences of establishment?

8. Socio-Economic

8.1 What are the social or community impacts associated with your proposal to translocate?

8.2 What are the estimated economic impacts associated with your proposal to translocate?

9. References

BERG. 2008. The Balcombe Estuary Reserves. The Balcombe Estuary Rehabilitation Group Website (http://www.berg.org.au/; accessed Dec 2008). Davies, P. 2005. A review of the flow and flow-related habitat requirements of Tasmanian native and introduced freshwater fish. Report to DPIWE Water Assessment and Planning Branch, Freshwater Systems Pty Ltd, Hobart Tasmania.

Department of Environment and Primary Industries. 2005. Protocols for the Translocation of Fish in Victorian Inland Public Waters. Fisheries Victoria Management Report Series No. 24.

Douglas, J. & Hall, K. 2004 Lake Wendouree Fisheries Assessment. (former) Department of Environment and Primary Industries. EPA Victoria. 2004. Risk-Based Assessment of Ecosystem Protection in Ambient Waters. EPA Victoria Publication No. 961, October 2004. ISBN 0 7306 7644 7. Southbank, Victoria.

GHD. 2008. Report on Hydrological Analyses. Devilbend Reservoir Modelling Report. GHD Report to Parks Victoria.

GROWNS, I. and JAMES, M. 2005. Relationships between river flows and recreational catches of Australian bass. Journal of Fish Biology (2005) 66, 404–416. Hall, K. and Douglas J. 2008 Lake Modewarre Creel Survey. Fisheries Victoria Research Report. Series No. 33. Harris, J. H. 1986. Reproduction of the Australian bass, Macquaria novemaculeata, (Perciformes: Percichthyidae) in the Sydney Basin. Australian Journal of Marine and Freshwater Research 37, 209–235. Henry, G.W. and Lyle, J.M. 2001. The National Recreational and Indigenous Fishing Survey. Tasmanian Aquaculture & Fisheries Institute Report. FRDC Project No. 99/158.

Jerry, D. R., Raadik, T. A., Cairns, S. C. & Baverstock, P. R. (1999). Evidence for natural interspecific hybridization between the Australian bass (Macquaria novemaculeata) and estuary perch (M. colonorum). Marine and Freshwater Research 50, 661–666.

Melbourne Water. 2007. Port Phillip and Westernport Regional Health Strategy. Melbourne Water, and Port Phillip and

Westernport Catchment, Management Authority, Melbourne. Molony, B. 2001. Environmental requirements and tolerances of Rainbow trout (Oncorhynchus mykiss) and Brown trout (Salmo trutta) with special reference to Western Australia: A review. FISHERIES RESEARCH REPORT NO. 130, 2001. Fisheries Research Division, WA Marine Research Laboratories, Dept of Fisheries.

McGuckin J. (2001). Fish survey of Devilbend and Bittern Reservoirs, Devilbend and Tuerong Creeks. A report prepared by Streamline Research for Melbourne Water. McGuckin, J. 2007. A survey for dwarf galaxias in the vicinity of Bittern and Devilbend Reservoirs (following a prolonged drought). A Streamline Research Report for Parks Victoria, August 2007.

SEITA. 2008. The Environmental Effects Statement (EES) for the Frankston Bypass. Southern and Eastern Integrated

Transport Authority (SEITA), Melbourne Victoria. Stamation, K. and Loyn, R (2008) Impact of recreational access and stocking on waterbirds and shorebirds – Devilbend Reserve. Arthur Rylah Institute for Environmental Research Technical Report Series No. xxx. Department of Sustainability and Environment, Heidelberg, Victoria.

Thompson, R. and Walker, M. 2009. DEVILBEND AQUATIC HABITATS ECOSYSTEM STUDY. Parks Victoria Contracted Report # #####. School of Biological Sciences and Australian Centre for Biodiversity Monash University.

Appendix 1

Disinfection of Fish Farms
Disinfection and Method of use as per OIE, International Aquatic Animal Health Code 2003.
ISBN 92-9044-478-9

a. Dangerous. Personnel must be protected. Protect skin, eyes and respiratory tract with a suitable barrier.
* The concentrations indicated are those for the active substance. NB Chemicals must be approved for the prescribed use and used according to the manufacturer's specifications