Fish habitat in the Lower River Murray: an analysis of the nature, extent and the associated fish assemblages

(1)

Fish habitat in the Lower River Murray: an analysis of the nature, extent and the associated fish assemblages

Final Report

SARDI Publication No. F2007/000973-2 SADRI Research Report Series No. 418 Kelly Marsland, Jason Nicol and Dale McNeil

SARDI Aquatic Sciences 2 Hamra Ave, West Beach SA 5024

January 2010

(2)

Fish habitat in the Lower River Murray: an analysis of the nature, extent and the associated fish assemblages

Final Report

Kelly Marsland, Jason Nicol and Dale McNeil

January 2010

SARDI Publication No. F2007/000973-2 SADRI Research Report Series No. 418

(3)

This Publication may be cited as:

Marsland, K.B., Nicol, J.M. and McNeil, D.M. (2009). Fish habitat in the Lower River Murray: an analysis of the nature, extent and the associated fish assemblages. Final report. South Australian Research and Development Institute (Aquatic Sciences), Adelaide, 47pp. SARDI Publication No. F2007/000973-2.

South Australian Research and Development Institute SARDI Aquatic Sciences

2 Hamra Avenue West Beach, SA 5024 Telephone: (08) 8207 2400 Facsimile: (08) 8207 5481 http://www.sardi.sa.gov.au

Disclaimer.

The authors warrant that they have taken all reasonable care in producing this report. The report has been through the SARDI Aquatic Sciences internal review process, and has been formally approved for release by the Chief Scientist. Although all reasonable efforts have been made to ensure quality, SARDI Aquatic Sciences does not warrant that the information in this report is free from errors or omissions. SARDI Aquatic Sciences does not accept any liability for the contents of this report or for any consequences arising from its use or any reliance placed upon it.

This work is copyright. Apart from any use as permitted under the Copyright Act 1968, no part may be reproduced by any process without prior written permission from the author.

Printed in Adelaide: January 2010

SARDI Publication No. F2007/000973-2 SARDI Research Report Series No. 418

Authors: K.B. Marsland, J.M. Nicol and D.M. McNeil Reviewers: Sandra Leigh and Brenton Zampatti Approved by: B. Smith

Signed:

Date: 7 January 2010

Distribution: SAMDBNRM Board, DWLBC, DEH and SARDI Aquatic Sciences Library Circulation: Public Domain

(4)

Acknowledgements

The authors thank Penny Baldock and Mark Hall (PIRSA SIS) and Matt Miles (DEH) for technical support and advice for the GIS component of the project, Brenton Zampatti, Brad Hollis, Sandra Leigh and Sunil Sharma for comments on early drafts of this report and Arron Strawbridge, Michael Guederien, Neil Wellman, Phillipa Wilson, Simon Westergard and Paul Drummond for field assistance. The South Australian Murray-Darling Basin Natural Resources Management Board through their investment strategy funded this project.

(5)

List of Figures

Figure 1: Distribution of fish sampling sites across the Lower River Murray...12 Figure 2: Map of the Lower River Murray indicating the three regions used for data analysis in this report; below Lock 1, locks 1-3 and above lock 3...17 Figure 3: Graph displaying total fish diversity for each mesohabitat sampled from the three regions of the Lower River Murray (below Lock 1, locks 1-3 and above lock 3). Simpson’s Index of diversity (1 –D, where Simpson’s Index, D =Σ (n/N)² was used to calculate fish diversity.

Values range from 0 – 1, higher values indicate greater diversity...24 Figure 4: Relative abundance of fish species for each mesohabitat per region. Number of fish (CPUE) is indicated by n. Fish species are recorded in taxa code where Bid bid=silver perch, Car aur=goldfish, Cra ste=un-specked hardyhead, Cyp car=common carp, Gal mac=common galaxias, Gam hol=gambusia, Hyp spp=carp gudgeon, Mac amb=golden perch, Mac pee=Murray cod, Mel flu=Murray-Darling rainbowfish, Nem ere=bony herring, Per flu=redfin, Phi gra=flathead gudgeon, Phi mac=dwarf flathead gudgeon, Ret sem=smelt, Tan tan=freshwater catfish. ...26 Figure 5: Relative abundance of fish species for each mesohabitat per region. Number of fish (in CPUE) is indicated by n. Fish species are recorded in taxa code where Bid bid=silver perch, Car aur=goldfish, Cra ste=un-specked hardyhead, Cyp car=common carp, Gal mac=common galaxias, Gam hol=gambusia, Hyp spp=carp gudgeon, Mac amb=golden perch, Mac pee=Murray cod, Mel flu=Murray-Darling rainbowfish, Nem ere=bony herring, Per flu=redfin, Phi gra=flathead gudgeon, Phi mac=dwarf flathead gudgeon, Ret sem=smelt, Tan tan=freshwater catfish. ...27 Figure 6: Relative abundance of fish species for each mesohabitat per region. Number of fish (in CPUE) is indicated by n. Fish species are recorded in taxa code where Bid bid=silver perch, Car aur=goldfish, Cra ste=un-specked hardyhead, Cyp car=common carp, Gal mac=common galaxias, Gam hol=gambusia, Hyp spp=carp gudgeon, Mac amb=golden perch, Mac pee=Murray cod, Mel flu=Murray-Darling rainbowfish, Nem ere=bony herring, Per flu=redfin, Phi gra=flathead gudgeon, Phi mac=dwarf flathead gudgeon, Ret sem=smelt, Tan tan=freshwater catfish. ...28

(7)

List of Tables

Table 1: The 29 mesohabitat types determined for the Lower River Murray, their main habitat features and the corresponding number and percentage area of polygons. The percentage area of each habitat type was calculated by determining the total area of all polygons and the respective percent each habitat type contributed. These are to be interpreted as guides only as the width of the polygons are variable and were not drawn to scale...7 Table 2: Number of sites in each broad habitat category sampled across the South Australian River Murray. ...11 Table 3: Summary of fish caught during the 2008 survey and the conservation status of each species in South Australia (* denotes alien species)...15 Table 4: Table showing significant (P<0.05) associations between each fish species and microhabitat components, mesohabitat types and region. For each microhabitat component + indicates a positive association and – shows a negative association. Growth forms of each plant species is shown in brackets where E = emergent aquatic plant, FL = floating leaved submerged macrophyte, S = submerged macrophyte and T = tree. If a category is classified as N/A the corresponding fish species was absent (or only one individual present) in the region. ...19 Table 5: Significant microhabitat associations for each mesohabitat type for the three regions.

Growth forms of each plant species is shown in brackets where E = emergent aquatic plant, FL

= floating submerged macrophyte, S = submerged macrophyte and T = tree. If a category is classified as N/A the corresponding mesohabitat was not sampled or not present in the region.23 Table 6: Prioritisation matrix based on the diversity of the fish community and number of species listed under the EPBC Act or protected under the South Australian Fisheries Act present in a mesohabitat...32 Table 7: Prioritisation of mesohabitats based on conservation value for fish and suggested actions required (* cliffs were classified as having a low conservation value downstream of Lock 1 and medium value between Locks 1and 3 using this framework; however, they have been identified as important Murray cod habitat (Ye and Zampatti 2007) that may warrant a higher conservation value)...35

List of Appendices

Appendix 1: Simper results comparing fish assemblage between regions. ...44

(8)

Executive Summary

This report provides a description of the findings of the River Murray Fish Habitat project. This project aimed to identify priority areas of the South Australian River Murray main channel based on an assessment of fish habitat. An inventory of the different in-water and riparian (<50 cm above pool level) habitats and their distributions was undertaken in the River Murray main channel between Wellington and the New South Wales border. Mapping was undertaken from a boat and plant species (and other major habitat components such as large woody debris and man made structures) within a reach (greater than 50m in length) were recorded in situ onto a pocket computer with ArcPad software. The information was transferred onto an interactive GIS database using ArcGIS desktop software where the broad habitat types were determined based on the different habitat features and plant species present. The resultant database contains the mesohabitat data as well as snag locations, bathymetry, and aerial photographs.

At the conclusion of the mapping component 29 mesohabitat types were identified. The four most common habitat types along the main channel were classified as bare, Willows dense,

‘Typha’/’Phragmites’ and ‘Phragmites’/ Red Gum sparse. The resultant database can be used to view the habitat types, depths and number of snags within a region or queries may be preformed on the data (e.g. location or total area of a habitat type). The database may be used for various projects based in the lower River Murray, particularly studies prioritising reaches for rehabilitation or protection.

In the final phase of the study fish assemblages were sampled in a representative sub-set of the identified mesohabitats and the information used to determine relationships between fish assemblages and mesohabitats. Due to differences in fish assemblages between regions, the lower River Murray was split into three regions: below Lock 1, Locks1-3 and above Lock 3 and priority habitats identified for each region. The were no high priority mesohabitats downstream of Lock 1 (based on fish diversity and presence of protected or EPBC listed species), probably due to low water levels stranding all of the structurally diverse mesohabitats. Between Locks 1 and 3 and upstream of Lock 3 the fish community was more diverse and protected and EPBC listed species were more abundant, which may be due to the greater diversity of structural habitats. Four high and two medium conservation value mesohabitats were present between Locks 1 and 3 and four high and four medium conservation value mesohabitats were present upstream of Lock 3.

The prioritisation of mesohabitats in this report was based on a single snapshot of the fish community and did not take into account temporal changes and deil changes. In addition, there are limitations to electrofishing especially in water deeper than 2.5 m that may have resulted in

(9)

some species not being captured in certain mesohabitats (especially cliffs and willow habitats downstream of Lock 1). Nevertheless, the results from this study provide a good starting point for selecting freshwater protected areas or areas for habitat rehabilitation or revegetation projects.

(10)

1. Background and Aims

The River Murray in South Australia is highly regulated with a series of low level (~3 m) weirs and upstream abstraction reducing the natural variability of the river’s flow, inundation of the floodplain and obstructing the longitudinal movement of aquatic organisms (Gehrke et al. 1995;

Maheshwari et al. 1995). Native fish populations have declined in range and abundance since river regulation (Cadawallader 1978; Humphries et al. 2002) and in order to manage fish stocks in the River Murray it is important that the preferred habitat for these species is identified and protected or improved (Gehrke et al. 1995). Currently, there is no quantitative data on the nature and extent of fish habitats in main channel of the River Murray (except for priority sites such as the main channel adjacent to the Chowilla Anabranch (Zampatti et al. 2006).

The South Australian Integrated Natural Resources Management (INRM) Strategy, River and Floodplain Management Program 5, outlined several resource condition targets for aquatic habitats in the South Australian River Murray (SAMDINRMG 2004). Specifically, task 5.1.5 (to undertake baseline aquatic fauna and in-stream habitat data collection and assessment) in the INRM Strategy requires baseline information on in-stream habitats to be collected to gain an assessment of the current condition of these habitats, their distributions and any associated fish communities, to determine whether resource condition targets are being achieved. In accordance with this task, the South Australian Murray-Darling Basin Natural Resources Management (SAMDBNRM) Board funded this study to investigate the relationship between different aquatic and riparian habitats and fish communities in the lower River Murray.

The first stage of the study (2006-2007) involved mapping the riparian, emergent and submerged vegetation of the River Murray and a range of additional physical habitat features, specifically bathymetry, large snags, man made structures and major wetland entrances. These habitat features were used to create a GIS database incorporating bathymetry, biological and structural features. In the autumn of 2008, representative habitats along the main channel were surveyed using Sustainable Rivers Audit (SRA) electrofishing techniques. In addition to electrofishing the major mesohabitats, quantitative assessments at the microhabitat (electrofishing shot) scale were also conducted at each site. Electrofishing was used to sample fish assemblages because it is an active, non-selective sampling method that enables fish to be collected in a standard manner in the habitat that they are occupying at that point in time (Baumgartner et al. 2008). This project also offered an opportunity to investigate associations between fish assemblage and specific habitat features.

(11)

This report outlines the findings of our three year study which aimed to:

• collect baseline information of the nature, extent and types of in-stream and riparian habitats in the lower River Murray,

• identify the fish communities associated with these habitat types and

• incorporate this information into a GIS database that can be used to identify priority areas in the River Murray Main Channel based on fish habitat at a reach scale that can be used to aid managers in conservation planning and management.

The results of this survey may be used to provide baseline information on the aquatic and riparian habitat of the South Australian River Murray. Using this information in conjunction with the fish survey data priority areas may be identified to aid managers in conservation planning to ensure a comprehensive and representative reserve system of freshwater protected areas. However, only one fish survey was conducted during a period of record low inflow into South Australia; therefore, it represents a snapshot of the fish community present at that point in time. Due to the river conditions at the time of the survey and the fact that it was a “one off”

survey, the fish data and habitat associations need to be treated with some caution because it may not reflect long-term patterns or the fish community under entitlement flow (or greater) conditions.

2. Stage 1: Mesohabitat Mapping 2.1. Mesohabitat Mapping: Methods

A GIS database was constructed using ArcGIS Desktop 9.1 (ESRI, 2006). The database was populated with a combination of existing and constructed layers. The snag layer was obtained from Department for Environment and Heritage, Environmental Information and only contained the locations of large snags that are potentially hazardous to boating. The bathymetry and floodplain elevation layers were stitched together by DEH from the Department of Water, Land and Biodiversity Conservation LIDAR data (floodplain) and SA water sonar data (bathymetry) to form one layer. Topography layers, originally mapped by Department for Environment and Heritage, were obtained from the PIRSA SIS library. Features of this layer include place names and water body boundaries. The River Murray boundary was based on water body boundary mapping from the 1:50 000 Topographic GIS layer. Aerial photographs (from the 2008 fly over) provided by the SAMDBNRM Board were added to the database and used to verify the outline of the River Murray. In order for the polygons to be viewed on printed maps, the polygons had a thickness of between 50 and 70 metres. The River Murray outline was

(12)

then transferred to a hand held Trimble Recon pocket personal computer (pocket PC) equipped with GPS and Arc Pad software.

Surveys of the aquatic and riparian (from the edge of the water to 50 cm above normal pool level) habitat were undertaken along the main channel of the River Murray between the Wellington ferry and the New South Wales border in the autumn and winter of 2007, when the water level in each weir pool was at normal pool level. Mapping was undertaken from a boat and recorded on the pocket PC in situ. Habitat types were attributed to reaches greater than 50 metres in length. Changes in habitat types were marked on the pocket PC as lines. Each different habitat was marked as a point and plant species within that reach were recorded on the pocket PC in order of abundance.

Dominant emergent and overstorey plant species were labelled either dense or sparse depending on their abundance within a reach. Phragmites australis, Typha spp., Schoenoplectus validus and Bolboschoenus caldwellii were considered dense if there was greater than 50% cover within a reach and sparse if one of these species was dominant, but occupied between 15 and 50% of the reach.

Eucalyptus camaldulensis var. camaldulensis, Salix spp. and Acacia stenophylla were categorised as sparse in reaches with scattered trees present and dense in reaches where the canopy cover was greater than 50%. Other emergent and overstorey species observed during the study, which did not form dense stands or dominate a habitat, were recorded for the habitat groups in which they were found.

A reach that was categorised as bare in instances where the percentage cover of unvegetated bank was 85 % or greater (all species present have sparse abundances). Areas heavily impacted by human activity (e.g. shacks, locks, townships) along the banks of the River Murray were categorised as a modified mesohabitat. In each instance all species present in these areas were also recorded.

Habitat polygons were created in the GIS database from the lines and points recorded on the pocket PC. Each polygon was populated by a set of species in order of abundance; however, for illustrative purposes and future analyses polygons were categorised into mesohabitat types identified at the conclusion of the field component. Mesohabitat types were determined by the dominant plant species or physical features (e.g. cliffs, wetland entrances) present in the reach.

Plants were identified using keys in Cunningham et al. (1981), Sainty and Jacobs (1981), Jessop and Toelken (1986), Romanowski (1998), Sainty and Jacobs (2003) and Jessop et al. (2006). In some cases due to immature individuals or lack of floral structures, plant were identified to genus only. Nomenclature follows Barker et al. (2005).

(13)

2.2. Mesohabitat Mapping: Results

Twenty nine mesohabitat types were identified at the conclusion of the field mapping component. Each mesohabitat category is associated with a specific set of attributes is described in Table 1:

(14)

Table 1: The 29 mesohabitat types determined for the Lower River Murray, their main habitat features and the corresponding number and percentage area of polygons. The percentage area of each habitat type was calculated by determining the total area of all polygons and the respective percent each habitat type contributed. These are to be interpreted as guides only as the width of the polygons are variable and were not drawn to scale.

Mesohabitat Description of dominant habitat features No. of Polygons

% Area

‘Acacia’ Acacia stenophylla (dense) ± Eucalyptus camaldulensis var.

camaldulensis, Eucalyptus largiflorens, Salix spp.,

Phragmites australis, Typha spp., Schoenoplectus validus or submerged vegetation.

23 1.09

Bare Bare bank ± Eucalyptus camaldulensis var. camaldulensis, Eucalyptus largiflorens, Acacia stenophylla, Salix spp., Phragmites australis, Typha spp., Schoenoplectus validus, Paspalum distichum, Bolboschoenus caldwellii or submerged vegetation.

218 14.19

‘Bolboschoenus’ Bolboschoenus caldwellii (dense) with Eucalyptus camaldulensis var. camaldulensis (sparse or dense) ± Acacia stenophylla, Phragmites australis, Typha spp., Schoenoplectus validus or submerged vegetation.

4 0.09

‘Bolboschoenus’

sparse

Sparsely distributed Bolboschoenus caldwellii with Eucalyptus camaldulensis var. camaldulensis (sparse) ± Acacia stenophylla, Phragmites australis, Typha spp. or submerged vegetation.

7 0.26

Cliffs Cliffs ± Eucalyptus camaldulensis var. camaldulensis, Eucalyptus largiflorens, Salix spp., Acacia stenophylla, Phragmites australis, Typha spp. or Paspalum distichum.

120 9.20

Modified Highly modified riparian zone including shacks, locks and weirs, marinas, townships and other infrastructure along the riparian zone.

181 9.16

‘Phragmites’ Phragmites australis (dense) ± Eucalyptus camaldulensis var. camaldulensis, Eucalyptus largiflorens, Salix spp., Acacia stenophylla (all sparse), Juncus usitatus, Schoenoplectus validus, Typha spp., Bolboschoenus caldwellii, or submerged vegetation.

10 0.08

‘Phragmites’/

‘Acacia’

Phragmites australis with Acacia stenophylla ± Eucalyptus camaldulensis var. camaldulensis, Eucalyptus largiflorens, Salix spp., (all sparse), Schoenoplectus validus, Typha spp., Bolboschoenus caldwellii or submerged vegetation.

22 1.01

‘Phragmites’/

‘Bolboschoenus’

Phragmites australis with Bolboschoenus caldwellii (both dense) ± Eucalyptus camaldulensis var. camaldulensis, Acacia stenophylla, Juncus usitatus, Schoenoplectus validus, Typha spp. or submerged vegetation.

26 0.90

‘Phragmites’/

Red Gum

Phragmites australis with Eucalyptus camaldulensis var.

camaldulensis (both dense) ± Salix spp., Acacia stenophylla (all sparse), Juncus usitatus, Schoenoplectus validus, Typha spp., Bolboschoenus caldwellii or submerged vegetation.

117 5.38

(15)

% Area

‘Phragmites’/

Red Gum sparse

Phragmites australis (dense) with Eucalyptus camaldulensis var. camaldulensis (sparse) ± Salix spp., Acacia stenophylla, Eucalyptus largiflorens,, Juncus usitatus, Schoenoplectus validus, Typha spp., Bolboschoenus caldwellii, Paspalum distichum or submerged vegetation.

202 10.95

‘Phragmites’/

‘Schoenoplectus’

Phragmites australis with Schoenoplectus validus (both dense)

± Eucalyptus camaldulensis var. camaldulensis, Salix spp., Acacia stenophylla, Eucalyptus largiflorens, Juncus usitatus, Typha spp., Bolboschoenus caldwellii, Paspalum distichum or submerged vegetation.

40 1.40

‘Phragmites’ sparse Main habitat feature is Phragmites australis (sparse) with Eucalyptus camaldulensis var. camaldulensis (sparse) ± Salix spp., Acacia stenophylla, Eucalyptus largiflorens, Juncus usitatus, Schoenoplectus validus, Typha spp., Bolboschoenus caldwellii, Paspalum distichum or submerged vegetation.

121 5.39

‘Phragmites’/

Willows

Phragmites australis (dense) with Salix spp. (sparse) ± Eucalyptus camaldulensis var. camaldulensis, Acacia stenophylla, Typha spp. or submerged vegetation.

51 1.59

Red Gum dense Eucalyptus camaldulensis var. camaldulensis (dense) ± Acacia stenophylla, Phragmites australis, Typha spp., Schoenoplectus validus, Juncus usitatus, Paspalum distichum, herbs or grasses or submerged vegetation.

82 3.17

Red Gum sparse Eucalyptus camaldulensis var. camaldulensis (sparse) with herbs and grasses ± Acacia stenophylla, Phragmites australis, Typha spp., Schoenoplectus validus, Juncus usitatus or submerged vegetation.

10 0.38

‘Schoenoplectus’ Schoenoplectus validus (dense) ± Eucalyptus camaldulensis var. camaldulensis, Salix spp., Phragmites australis, Juncus usitatus, Schoenoplectus validus, Typha spp., Bolboschoenus caldwellii, Paspalum distichum or submerged vegetation.

19 0.44

Swamp Inundated area of vegetation with a poorly defined bank ± Eucalyptus camaldulensis var. camaldulensis, Salix spp., Acacia stenophylla, Phragmites australis, Juncus usitatus, Schoenoplectus validus, Typha spp., Bolboschoenus caldwellii or submerged vegetation.

67 2.49

‘Typha’/

‘Phragmites’

Phragmites australis and Typha spp. (both dense) ± Eucalyptus camaldulensis var. camaldulensis, Salix spp., Acacia stenophylla, Eucalyptus largiflorens, Juncus usitatus, Schoenoplectus validus, Bolboschoenus caldwellii, Paspalum distichum or submerged vegetation.

278 11.47

‘Typha’ Typha spp. (dense) ± Eucalyptus camaldulensis var.

camaldulensis, Salix spp., Acacia stenophylla, Eucalyptus largiflorens, Phragmites australis, Schoenoplectus validus, Bolboschoenus caldwellii or submerged vegetation.

59 1.32

(16)

% Area

‘Typha’/

‘Phragmites’ sparse

Phragmites australis, Typha spp. (both sparse) and bare soil ± Eucalyptus camaldulensis var. camaldulensis, Acacia stenophylla, Eucalyptus largiflorens, Juncus usitatus, Schoenoplectus validus, Bolboschoenus caldwellii or submerged vegetation.

43 1.75

‘Typha’ /Willows Typha spp. (dense) with Salix spp. (sparse) ± Eucalyptus camaldulensis var. camaldulensis or submerged vegetation.

9 0.22

‘Typha’ sparse Typha spp. (sparse) and bare soil ± Eucalyptus camaldulensis var. camaldulensis, Acacia stenophylla, Schoenoplectus validus or submerged vegetation.

20 0.49

‘Typha’/

‘Bolboschoenus’

Typha spp. with Bolboschoenus caldwellii (both dense) ± Eucalyptus camaldulensis var. camaldulensis, Phragmites australis or submerged vegetation.

7 0.18

‘Typha’/

‘Schoenoplectus’

Typha spp. with Schoenoplectus validus (both dense) ± Eucalyptus camaldulensis var. camaldulensis, Salix spp., Phragmites australis, Juncus usitatus or submerged vegetation.

24 0.77

‘Typha’/

Red Gum

Typha spp. with Eucalyptus camaldulensis var. camaldulensis (both dense) ± Acacia stenophylla, Salix spp., Phragmites australis, Juncus usitatus, Schoenoplectus validus or submerged vegetation.

38 0.98

Willows dense Dense Salix spp. ± Eucalyptus camaldulensis var.

camaldulensis, Typha spp., Phragmites australis, Schoenoplectus validus or submerged vegetation.

188 13.85

Willows/

Red Gum

Salix spp. and Eucalyptus camaldulensis var. camaldulensis (both sparse), ± Typha spp., Phragmites australis, Schoenoplectus validus or submerged vegetation.

8 0.23

Wetland entrance An entrance to a floodplain wetland. 175 1.56

The most widespread habitat types in the South Australian River Murray were: Bare, Willows dense, ‘Typha’/’Phragmites’ and ‘Phragmites’/Red Gum sparse, followed by cliffs and modified habitats (Table 1). The least prevalent habitat types were ‘Phragmites’ and ‘Bolboschoenus’.

Reaches classed as bare were only found upstream of Swan Reach, and were most prominent between Morgan and the NSW border, with the greatest concentration around Loxton. ‘Willows dense’ habitats were predominant between Wellington and Purnong with scattered reaches upstream. ‘Typha/Phragmites’ habitats were distributed throughout the Lower River Murray, except where ‘Willows dense’ habitats were dominant (Wellington to Purnong) and around the Loxton area where the majority of the reaches were classified as bare. The ‘Phragmites’/Red Gum sparse habitat type was also widely distributed from Purnong to the NSW border, although less common in the Loxton area and upstream of Renmark.

(17)

The database contains the habitat polygons, snag and depth layers, aerial photographs and topographical information. Users are able to view specific areas of the Lower River Murray and inspect the corresponding mesohabitats and specific species present in each polygon as well as perform queries on the habitat data. The information can also be converted into printable maps.

3. Stage 2: Fish and Microhabitat Survey 3.1. Fish and Microhabitat Survey: Methods

3.1.1. Site selection

From the 29 mesohabitats identified in the first stage of the project (Table 1), ten were selected for fish sampling (Table 2) based on the following criteria:

• Broad habitat polygons were only selected that were equal or greater than 100 metres in length to allow for sufficient area to complete six x 90 second electrofishing shots.

• Three or more polygons fitting the above criteria were required to be present in the study area.

• Polygons were ineligible to be surveyed if they were immediately downstream of a weir to avoid affecting similar sampling projects analysing fish movement in these regions and any confounding effects of the weirs themselves.

• Polygons fitting the above criteria were short listed and consideration was given to accessibility, the proximity to one or more potential sites and location to ensure an even spread of sites across the River.

• Broad habitats that were prominent and widespread across the main channel were sampled with a greater number of replicates than those that were less prevalent (Table 2).

(18)

Table 2: Number of sites in each broad habitat category sampled across the South Australian River Murray.

Broad Habitat type Number of Sites

Cliffs 10 Modified 9

‘Acacia’ 3

‘Phragmites’ Red gum sparse 3

‘Phragmites’ and Red gum 9

‘Typha’ and ‘Phragmites’ 6

Willows 9 Bare 9

Red gum 3

‘Phragmites’ sparse 6

Total sites 61

(19)

Figure 1: Distribution of fish sampling sites across the Lower River Murray.

(20)

3.2. Electrofishing

Boat electrofishing was used to sample the fish communities associated with the different mesohabitat types in the lower River Murray. It was deemed the most appropriate method because it has been proven to effectively and rapidly sample both large and small bodied fish in the littoral zone of large, turbid lowland rivers (Faragher and Rodgers 1997; Baumgartner et al. 2008) and is used extensively in the main channel of the lower River Murray (e.g. Zampatti et al. 2006). The techniques used in this survey followed those outlined in the Sustainable Rivers Audit (SRA) (Murray Darling Basin Commission 2004) with the exception that one site (composed of six shots) was confined to a single side of the river allowing consistent sampling along a specific habitat type. In addition, the data will be comparable with the Chowilla fish condition monitoring (Zampatti et al. 2008), Katarapko (Leigh et al.

2008) and Pike River projects (in progress).

Fish surveys were conducted during daylight hours from March to May 2008 using a boat mounted 7.5kW Smith Root Model GPP electrofishing system. Six 90 second (power on time) shots were conducted at each site. All fish were dip netted and placed in a recirculating well. Fish from each shot were identified and measured for length (caudal fork or total length in mm). The littoral zone of each mesohabitat (between 0 and 5 m water depth) was fished in order to stay within the effective range of the electrofishing equipment.

3.3. Microhabitat Assessments

Quantitative visual habitat assessments were carried out at each electrofishing shot following the methods used by Zampatti et al (2006). One observer estimated the percentage cover of in-stream habitat including submerged and emergent vegetation, large woody debris, physical structures and open water. Large woody debris was categorised depending on the size of the wood, such that:

• CWD 1: twigs and branches with diameters <1 cm

• CWD 2: branches with diameters 1-5 cm

• CWD 3: branches and trunks with diameters >5 cm

3.4. Data Analysis

Fish assemblage and associations with meso and microhabitat types were analysed with various non- parametric multivariate techniques using PRIMER v. 6.1.12 (Clarke and Gorley 2006) and PC-Ord v.

5.12 (McCune and Mefford 2006) software packages. The number of fish caught in each shot was transformed into catch per unit effort (CPUE) which standardises the capture of fish based on the

(21)

number of electrofishing seconds used per shot. The statistical package PRIMER was used to perform analysis of similarity (ANOSIM) and SIMPER (Clarke and Gorley 2006). ANOSIM provides information on the degree that samples are similar to each other and SIMPER compares the similarity of different groups, and indicates what factors are driving these differences. Therefore, we used this method to indicate what was driving the patterns of similarity (or dissimilarity) between regions and meso-habitats and Indicator Species Analysis (Dufrene and Legendre 1997) to determine fish habitat relationships at the micro-habitat scale (sensu Zampatti et al. 2006).

Simpson’s Index (D) was used to quantify the cumulative fish diversity of each mesohabitat in each region. Simpson’s index of diversity (1 – D) indicates the probability that two individuals randomly selected from a sample will belong to different species. The value of this index ranges between 0 and 1, with the greater the value, greater the sample diversity. There are two acceptable versions of Simpson’s Index, in this report we used the following formula:

D =Σ (n/N)

²

where n = the total CPUE of a particular species caught in a mesohabitat type within a region and N = the total CPUE of all species caught in a mesohabitat type within a region.

The relative abundance of fish species within each mesohabitat of each region was determined by pooling the total of each species caught (in CPUE) and dividing the total catch within each mesohabitat.

4. Results

4.1. Catch summary

A total of 18,176 fish were caught during the survey (Table 3). Approximately 70% of the total catch was un-specked hardyhead (Craterocephalus stercusmuscarum fulvus: n = 6450) and bony herring (Nematalosa erebi: n = 6174) with Murray-Darling rainbowfish (Melanotaenia fluviatilis: n = 1973) and smelt (Retropinna semoni: n = 1529) also abundant (Table 3). Alien species made up 6.3% of the total catch and included common carp (Cyprinus carpio: n = 922), goldfish (Caruassius auratus: n = 112), gambusia (Gambusia holbrookii: n = 65) and redfin (Perca fluviatilis: n = 55) (Table 3). Low numbers of the protected species, freshwater catfish (Tandanus tandanus: n = 6) and silver perch (Bidyanus bidyanus: n = 15), and the iconic Murray cod (Maccullochella peelii peelii: n = 11) were caught during the survey (Table 3).

(22)

Table 3: Summary of fish caught during the 2008 survey and the conservation status of each species in South Australia (* denotes alien species).

Common Name Scientific Name Total caught Conservation status

Bony herring Nematalosa erebi: 6174 -

Carp gudgeon Hypseleotris spp. 235 -

Common carp* Cyprinus carpio 922 -

Common galaxias Galaxias maculatus 39 -

Dwarf flathead gudgeon Philypnodon macrostomus 5 -

Flathead gudgeon Philypnodon grandiceps 271 -

Freshwater catfish Tandanus tandanus 6

Protected under the SA Fisheries Act

Gambusia* Gambusia holbrookii 65 -

Golden perch Macquaria ambigua ambigua 314 -

Goldfish* Caruassius auratus 112 -

Murray cod Maccullochella peelii peelii 11

Listed as vulnerable

under the Commonwealth

EPBC Act

Murray-Darling rainbowfish Melanotaenia fluviatilis 1973 -

Redfin* Perca fluviatilis 55 -

Silver perch Bidyanus bidyanus 15

Protected under the SA Fisheries Act

Australian smelt Retropinna semoni 1529 -

Un-specked hardyhead Craterocephalus stercusmuscarum

fulvus 6450 -

Total 18176

4.2. Regional comparisons of fish assemblage and habitat structure

Traditionally the South Australian section of the River Murray has been divided into four regions: the Lower Lakes, (The Barrages to Wellington, which was not part of the study area), Murraylands (Wellington to Mannum), the Gorge (Mannum to Overland Corner) and the Valley (Overland Corner to the NSW boarder) (Holt et al. 2005; Nicol et al. 2006). However, persistent drought across the basin

(23)

has resulted in low river levels below lock one. Therefore, for analytical purposes in this report, we consider the river upstream of Wellington in three distinct regions: below Lock 1, Locks 1 to 3 and above Lock 3 (Fig. 2).

ANOSIM was used to compare the extent of similarity of fish assemblage between the three regions and showed that the respective regions had significantly different fish communities (R=0.124, p=0.001). SIMPER analysis, comparing the similarity of groups and each species contribution to the similarities, indicated that the abundance of un-specked hardyheads, bony herring and Murray-Darling rainbowfish contributed the most to the differences (Appendix 1).

Indicator species analysis, used to detect associations between fish species and region, revealed that each species was significantly associated with one region except golden perch and Murray cod (Table 4). Due to significant difference in the fish community, each of the aforementioned regions will be analysed separately.

(24)

Figure 2: Map of the Lower River Murray indicating the three regions used for data analysis in this report; below Lock 1, locks 1-3 and above lock 3.

(25)

4.3. Fish assemblage and associations with mesohabitat types

ANOSIM was preformed to determine the level of similarity of fish assemblages between mesohabitats within each region. Within each region, a significantly different fish community was found between mesohabitats: below Lock 1 (R=0.085, p=0.001), locks 1-3 (R=0.187, p=0.001) and above lock 3 (R=0.091, p=0.002). Indicator species showed that un-specked hardyheads were significant indicators of modified habitats below Lock 1 and above lock 3, and Murray-Darling rainbowfish were significant indicators of willow habitats below Lock 1 and modified habitats between locks 1 and 3 (Table 4). The protected species, silver perch, was significantly associated with red gum habitats above lock 3 and the alien species common carp, goldfish and gambusia were indicators of bare (above lock 3), willows (above lock 3) and bare (locks 1-3) mesohabitats respectively (Table 4).

(26)

Table 4: Table showing significant (P<0.05) associations between each fish species and microhabitat components, mesohabitat types and region. For each microhabitat component + indicates a positive association and – shows a negative association. Growth forms of each plant species is shown in brackets where E = emergent aquatic plant, FL

= floating leaved submerged macrophyte, S = submerged macrophyte and T = tree. If a category is classified as N/A the corresponding fish species was absent (or only one individual present) in the region.

Associated Microhabitat Components Associated Mesohabitats

Common

Name Below Lock 1 Locks 1-3 Above lock 3 Below Lock 1 Locks 1-3 Above lock 3

Associated region

Silver perch N/A Not significant +CWD 3 N/A Not

significant Red gum Above lock 3

Goldfish

+Myriophyllum verrucosum (S) +Vallisneria americana (S)

+Elodea canadensis (S) +Potamogeton tricarinatus (FL)

+Vallisneria americana (S) +Tree roots

+CWD 2 -Open water

+Potamogeton crispus (S)

Not

significant Bare Not

significant Locks 1-3

Un-specked hardyhead

+Typha domingensis (E) - Rock +Vallisneria americana (S) -Elodea canadensis ( S) -Acacia stenophylla (T)

Modified Not

significant Modified Locks 1-3

Common carp

+Potamogeton crispus (S) -Salix spp. (T)

- Open water +Typha domingensis (E) +Elodea Canadensis (S) +Potamogeton tricarinatus (FL) -CWD 1

-CWD 2 -CWD 3

Not significant

Not

significant Bare Below Lock 1

Common galaxias

+Myriophyllum caput-medusae (S)

+Typha domingensis (E)

N/A N/A

Not

significant N/A N/A Below

Lock 1

Gambusia

Not significant +Typha domingensis (E) +Potamogeton crispus (S) - Open water

+Salix spp. (T)

+Schoenoplectus validus (E) Not significant

Not

significant Willows Above lock 3

(27)

Associated Microhabitat Components Associated Mesohabitats Common

Carp gudgeon

+Open water

+Potamogeton tricarinatus (FL) +Vallisneria americana (S) -Salix spp. (T)

-CWD 2

+Potamogeton tricarinatus (FL)

+Myriophyllum verrucosum (S)

-Open water -Rock

+Bolboschoenus caldwellii (E) +Typha domingensis (E) +Vallisneria americana (S) +Open water

+ Myriophyllum verrucosum (S) -Ludwigia peploides (FL) -Acacia stenophylla (T)

Not significant

‘Typha’/

‘Phragmites’ Locks 1-3

Golden perch

+CWD 1 +CWD 3

+Vallisneria americana (S)

Not significant +CWD 3

Bare Not

significant

Not significant

Not significan

t

Murray cod

-CWD 3 +Rock

Not significant +Phragmites australis (E)

Not significant

Not significan

t Murray-

Darling rainbowfish

+Roots +Salix spp. (T) -CWD 3

-Myriophyllum verrucosum (S)

-Elodea canadensis (S) -Typha domingensis (E) -Ludwigia peploides (FL)

+Open water

- Elodea canadensis (S)

-Typha domingensis (E) Willows Modified Not

significant

Above lock 3

Bony herring

-Typha domingensis (E) -Myriophyllum verrucosum (S) - Bolboschoenus caldwellii (E) -Myriophyllum verrucosum (S)

Not

significant Bare Not

significant

Below Lock 1

Redfin

+CWD 2

+Myriophyllum caput-medusae (S)

+Tree roots

+Typha domingensis (E)

+Salix spp. (T)

-Phragmites australis (E)

+Elodea canadensis (S)

Not significant

Below Lock 1

Flathead gudgeon

+CWD 2 +Salix spp.

-Potamogeton crispus (S)

+Schoenoplectus validus (E) +Vallisneria americana (S) +Tree roots

-Open water

+Vallisneria americana (S) +Paspalum distichum (E) +Open water

+Tree roots -Salix spp. (T)

Modified Not significant

Not

significant Locks 1-3

(28)

Associated Microhabitat Components Associated Mesohabitats Common

Dwarf flathead gudgeon

N/A +Vallisneria americana (S) N/A

N/A Not

significant N/A

Not significan

t

Smelt

+CWD 3 +Tree roots

- Phragmites australis (E) -Potamogeton tricarinatus (FL)

-Elodea canadensis (S) -Ludwigia peploides (FL) -Potamogeton tricarinatus (FL) -Zannichellia palustris (S) -Typha domingensis (E)

Not significant

Not

significant ‘Acacia’ Locks 1-3

Freshwater catfish

N/A N/A Not significant

N/A N/A Not

significant N/A

(29)

4.4. Fish preference for microhabitat features

Indicator Species Analysis (Dufrene and Legendre 1997) was conducted to determine the relationships between microhabitats and fish species in each region. Each species showed a preference to one or more particular microhabitat features; however, this was not always consistent across regions (Table 4).

4.5. Microhabitat features and associated mesohabitat types

As particular fish are associated with particular microhabitat components, we determined whether specific mesohabitats were associated with specific microhabitats. ANOSIM and SIMPER analysis of the microhabitat components within the three regions revealed that the abundance and type of components differed significantly (R= 0.147, p= 0.001). Indicator Species Analysis was preformed on the data to detect which microhabitat components were significantly associated with which mesohabitat in each region. Table 5 lists the microhabitat components that were significantly associated (p<0.05) with each mesohabitat in a specific region.

(30)

Table 5: Significant microhabitat associations for each mesohabitat type for the three regions. Growth forms of each plant species is shown in brackets where E = emergent aquatic plant, FL = floating submerged macrophyte, S = submerged macrophyte and T = tree. If a category is classified as N/A the corresponding mesohabitat was not sampled or not present in the region.

Associated Microhabitat Components Mesohabitat

Below Lock 1 Locks 1-3 Above lock 3

Acacia

N/A N/A Acacia

stenophylla (T) Tree roots CWD 2

Bare

CWD 3 Bolboschoenus caldwellii (E) Cyperus gymnocaulos (E) Potamogeton tricarinatus (FL) Vallisneria americana (S) CWD 2

CWD 3 Tree roots

Potamogeton tricarinatus (FL) Open water

Cliffs Rock Rock

Open water

N/A

Modified Potamogeton tricarinatus (FL) Not significant Potamogeton crispus (S)

‘Phragmites’/

Red gum

N/A Phragmites australis (E) Elodea

canadensis ( S)

‘Phragmites’/

Red gum sparse

N/A Not significant N/A

‘Phragmites’

sparse

N/A Elodea canadensis (S)

Myriophyllum verrucosum (S)

Ludwigia peploides (FL) Rock

Red gum

N/A N/A Vallisneria

americana (S) Zannichellia palustris (S) Juncus usitatus (E)

CWD 3

‘Typha’/

‘Phragmites’

N/A Typha domingensis (E)

Schoenoplectus validus (E) Potamogeton crispus (S)

Typha

domingensis (E) Phragmites australis (E) Bolboschoenus caldwellii (E) Myriophyllum verrucosum (S)

Willows

Salix spp. (T) CWD 2 Tree roots

N/A Salix spp. (T)

(31)

4.6. Fish diversity comparisons across different mesohabitat types in each region

Simpson’s Index of diversity calculates diversity by considering both evenness and number of species. Therefore those mesohabitats with the lowest cumulative diversity, such as bare below Lock 1 and willows below Lock 1 (Figure 3), have both a lower number of species and a disproportionately larger number of one or two species. Mesohabitats with a Simpson’s Index of diversity greater than 0.75 were only found upstream lock 1 (Figure 3). Those mesohabitat types with the greatest diversity were Red gum and ‘Acacia’ above lock 3 and ‘Phragmites’

sparse between locks 1-3 (Figure 3).

0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 0.55 0.6 0.65 0.7 0.75 0.8 0.85

Modified

Bare

Cliffs

Willows

Phragm ites

'/Red g um

Phr agmites' spar

se

Phragmites '/Red gum

sparse

Typha'/'Phragm ites

'

Red gum Aca

cia'

Simpsons index of diversity

below 1 locks 1-3 above 3

Figure 3: Graph displaying total fish diversity for each mesohabitat sampled from the three regions of the Lower River Murray (below Lock 1, locks 1-3 and above lock 3). Simpson’s Index of diversity (1 – D, where Simpson’s Index, D =Σ (n/N)² was used to calculate fish diversity. Values range from 0 – 1, higher values indicate greater diversity.

4.7. Relative abundance and species richness of fish in different mesohabitat types in each region

The relative abundance of fish species varied between mesohabitats (Figures 4-6). Bony herring was consistently the most abundant species below Lock 1 (Figure 4). Between locks 1-3 and above lock 3 the most abundant species varied between mesohabitats. Murray-Darling rainbow

(32)

fish was the most abundant species in modified habitats, bony herring was the most abundant species in ‘Phragmites’ sparse, ‘Phragmites’ red gum sparse, bare and cliff habitats and unspecked hardyhead was the most abundant species in ‘Typha’/’Phragmites’ and ‘Phragmites’

red gum habitats between Locks 1 and 3 (Figures 4-5). Above Lock 3 unspecked hardyhead was the most abundant species in ‘Acacia’, modified, ‘Typha’/’Phragmites’ and willow habitats and bony herring was the most abundant species in ‘Phragmites’/red gum, ‘Phragmites’ sparse, red gum, willow and bare habitats (Figures 5-6).

The most species rich mesohabitat was ‘Red gum’ (above Lock 3), with 14 species recorded (Figure 5h). Of these species, four were alien, one was listed as vulnerable nationally (Murray cod) and 2 were protected in SA (catfish and silver perch). The only other mesohabitat where both protected species, silver perch and freshwater catfish, were present was

‘Typha’/’Phragmites’ between Locks 1-3 (Figure 5d).

The number of native fish species found in any mesohabitat ranged from seven to ten. Those mesohabitats with the least number of native fish species were bare, cliffs and willow mesohabitats below Lock 1, modified and bare between Locks 1-3 and ‘Phragmites’/Red gum and willows above lock 3 (Figures 4-6). The greatest number of native fish species were present in ‘Typha’/’Phragmites’ between Locks 1-3 and Acacia and Red gum mesohabitats above Lock 3 (Figures 4-5).

Fish habitat in the Lower River Murray: an analysis of the nature, extent and the associated fish assemblages

Fish habitat in the Lower River Murray: an analysis of the nature, extent and the associated fish assemblages

Final Report

SARDI Publication No. F2007/000973-2 SADRI Research Report Series No. 418 Kelly Marsland, Jason Nicol and Dale McNeil

January 2010

Fish habitat in the Lower River Murray: an analysis of the nature, extent and the associated fish assemblages

Final Report

Kelly Marsland, Jason Nicol and Dale McNeil

January 2010

SARDI Publication No. F2007/000973-2 SADRI Research Report Series No. 418

Acknowledgements

Table of Contents

List of Figures

List of Tables

List of Appendices

Executive Summary

1. Background and Aims

2. Stage 1: Mesohabitat Mapping 2.1. Mesohabitat Mapping: Methods

2.2. Mesohabitat Mapping: Results

3. Stage 2: Fish and Microhabitat Survey 3.1. Fish and Microhabitat Survey: Methods

3.2. Electrofishing

3.3. Microhabitat Assessments

3.4. Data Analysis

D =Σ (n/N)

4. Results

4.1. Catch summary

4.2. Regional comparisons of fish assemblage and habitat structure

4.3. Fish assemblage and associations with mesohabitat types

4.4. Fish preference for microhabitat features

4.5. Microhabitat features and associated mesohabitat types

4.6. Fish diversity comparisons across different mesohabitat types in each region

4.7. Relative abundance and species richness of fish in different mesohabitat types in each region