The state of Queensland, Department of Primary Industries and Fisheries and the Queensland Murray-Darling Committee Inc

Transcription

1 Queensland the Smart State Characteristics of fish fauna of the Macintyre and Dumaresq Rivers and Macintyre Brook A report to the Queensland Murray-Darling Committee Prepared by Adam Butcher Department of Primary Industries & Fisheries May 2007

2 May 2007 While every care has been taken in preparing this publication, the State of Queensland accepts no responsibility for decisions or actions taken as a result of any data, information, statement or advice, expressed or implied, contained in this report. The Department of Primary Industries and Fisheries has taken all reasonable steps to ensure that the information contained in this publication is accurate at the time of publication. Readers should ensure that they make appropriate enquiries to determine whether new information is available on the particular subject matter. The state of Queensland, Department of Primary Industries and Fisheries and the Queensland Murray-Darling Committee Inc Copyright protects this publication. Except as permitted by the Copyright Act 1968(Cth), reproduction by any means (photocopying, electronic, mechanical, recording of otherwise), making available online, electronic transmission or other publication of this material is prohibited without prior written permission from the Department of Primary Industries and Fisheries or the Queensland Murray-Darling Committee Inc. Inquiries should be addressed to: Director General Chief Executive Officer Department of Primary Industries and Fisheries Queensland Murray-Darling Committee Inc. Southern Fisheries Centre PO Box 6043 GPO Box 46 Toowoomba West Qld 4350 Brisbane Qld 4001

3 Contents Executive summary.. ii 1 Project objectives Project Methodology Background Geographical scope Methods 5.1 Data sources Data gaps Species biology/ecology Native Species Ambassis agassizii.. 11 Bidyanus bidyanus.. 12 Craterocephalus amniculus Craterocephalus stercusmuscarum fulvus.. 14 Gadopsis marmoratus Galaxias olidus Hypseleotris spp.. 17 Leiopotherapon unicolour Maccullochella peelii peelii Macquaria ambigua Melanotaenia fluviatilis Mogurnda adspersa Nemotalosa erebi Philypnodon grandiceps Retropinna semoni.. 27 Tandanus tandanus Alien Species Carassius auratus 30 Cyprinus carpio Gambusia holbrooki Perca fluviatilis Information gaps in the biology/ecology The river continuum concept The demonstration reach concept Review of sites on the proposed demonstration reach Recommendations on the suitability of sites Recommendations References. 64 i

4 Executive Summary The Queensland Murray-Darling Committee (QMDC) and the Border Rivers-Gwydir Catchment Management Authority (BRGCMA) are in the process of establishing a demonstration reach on the Macintyre River system. A key component of the investigative stage is to determine the appropriate length of the demonstration reach. This document informs the debate on demonstration reach length by characterising the species present within the geographical scope. The geographical coverage of this review is the Border Rivers region of southern Queensland and northern New South Wales. It includes waters from the Dumaresq and Macintyre Rivers, and Macintyre Brook and their tributaries downstream of Glenlyon Dam, Pindari Reservoir, and Coolmunda Dam, respectively. The downstream limit is near Toobeah on the Macintyre River. Ten data sets were identified from research projects in this area. These projects have collectively identified 16 native and four alien species in this section of the Border Rivers catchment. Although data collection has occurred between 1901 and 2007, very few of the data sets have spatial, temporal or technical continuity, limiting their usefulness for direct comparisons. At least five native fish and one alien species have a restricted distribution within the Border Rivers region. Species such as flathead gudgeons (Philypnodon grandiceps), river blackfish (Gadopsis marmoratus), Agassiz perchlet (Ambassis agassizii) and silver perch (Bidyanus bidyanus) are rare species in both river systems. On the other hand, gudgeons (Hypseletoris sp.), Murray River rainbow fish (Melanotaenia fluviatilis) and Darling hardyhead (Craterocephalus amniculus) are all relatively abundant in the Border Rivers region. Other species have variable distributions and abundance, being common in one river system, but not the other. The Border Rivers region would be suitable for the proposed demonstration reach. Three large-bodied species are known to undertake extensive migrations during their life-cycle, but populations could be accommodated within any demonstration reach site. Many of the lesser native species would not require extensive longitudinal distances to complete their life-cycle. Rather, it is the presence or absence of critical habitat over a short spatial scale that would be a key influence to their population success. However, there is still a concern about adequate offstream access. Many smaller species require access to either instream or offstream backwaters for some of their life-cycle. Both the habitat and access have diminished with the advent of flow regulation and strategies addressing this issue must be considered if we are to adequately accommodate the full suite of native freshwater fish present in the Border Rivers region of the Murray-Darling Basin. ii

5 Project Title: QMDC native fish strategy Demonstration Reach Macintyre River Phase 1: Characteristics of fish fauna of the Macintyre and Dumaresq Rivers and Macintyre Brook Prepared by Reviewed by Adam Butcher, DPI&F, Queensland Adjunct Assoc. Professor Mark Lintermans Research Program Coordinator Native Fish Strategy Murray-Darling Basin Commission & Dr. Peter Jackson, Consultant 1. Project objectives 1. To use existing data to determine the characteristics of the fish faunas of the Macintyre and Dumaresq Rivers, and Macintyre Brook. 2. To use this information to make recommendations on the appropriate length of a demonstration reach in the Macintyre River that would impact on the fish fauna. 3. To use this information to comment on the suitability of using sites on the Macintyre Brook or Dumaresq Rivers as control and treatment sites for the demonstration reach. 4. To provide information on knowledge gaps in relation to the fish faunas of these rivers. 2. Project Methodology The project will involve: 1. Identification of, negotiation for access to, and collation of existing databases on fish faunas in the three rivers with regard to species present, distributions and relative abundances. 2. Identification of data gaps. 3. Documenting the life cycles of the species present, particularly in relation to normal home range, migrations, spawning and juvenile rearing areas from existing information sources. 4. Identify information gaps for individual species. 1

6 5. Review the river continuum concept to identify the influence of riverscapes on fish dynamics, and define key concepts that underpin demonstration reach principles. 6. Use this information to make recommendations on suitable demonstration reach lengths that encompass and enhance previously defined fish faunas. 7. Review the proposed control and demonstration sites for suitability according to the criteria specified in Section Use this information to make recommendations on the suitability of sites on the Macintyre Brook or Dumaresq River as control sites, or to recommend other options for reach and control sites. 9. Provide recommendations on any additional work related to the fish fauna that needs to be done before the locality and length of the demonstration reach, and the locality of control sites can be determined. 3. Background Current estimates put native fish populations in the Murray-Darling Basin at 10% of the pre-european settlement levels (Anon. 2004). They are in urgent need of rehabilitation. This need for regeneration of fish populations is driven by the belief that they are an allegory of riverine health. As the largest river system in Australia, the Murray-Darling Basin supports the largest social and agricultural community outside the capitals of each state. A recent survey of the Murray-Darling Basin has classified 95% of the river length as degraded, with 30% as modified substantially from its original condition (Norris et al. 2001). While there are a range of reasons for the decline in fish populations in the Murray-Darling Basin (Anon. 2004), it is apparent that their decline is commensurate with the decline in river health. The Murray-Darling Basin Commission recently released a 10 year plan for fish rehabilitation (Anon 2004) to ensure that the significant declines in native fish populations within the Murray-Darling Basin are addressed and rehabilitated. These are to be achieved through 13 objectives, condensed into the following six driving actions: rehabilitating fish habitat; protecting fish habitat; managing riverine structures; controlling alien fish species; protecting threatened native fish species; and managing fish translocation and stocking. A recent Murray-Darling Basin Commission workshop on native fish habitat rehabilitation and management recommended five key actions to drive future habitat rehabilitation and management actions (Lintermans et al. 2004). These were: protecting fish habitat; managing rivers holistically; involving stakeholders in riverine protection and restorations; 2

7 building knowledge and capacity for integrated management of aquatic habitats; and developing demonstration reaches. Demonstration reaches are reaches of river where a number of interventions are implemented (e.g.: restoring fish passage, rehabilitating riparian vegetation, reestablishing woody debris, controlling alien fish, etc.) to demonstrate how a coordinated program will beneficially impact on populations of native fishes. Demonstration reaches are favoured by many because they demonstrate an integrated approach, using several rehabilitation actions that can achieve a quantifiable improvement in fish populations. They engage community awareness and support, focus the attentions of funding agencies, and provide scientific knowledge of rivers and fish (Barrett 2004a). There are several demonstration reaches in the southern states, but none yet in Queensland. The southern examples appear to be allied with pre-existing management actions. The Queensland Murray-Darling Committee (QMDC) and the Border Rivers-Gwydir Catchment Management Authority (BRGCMA) are in the process of establishing a demonstration reach on the Macintyre River system. The Dumaresq, Severn and Mole Rivers and the Macintyre Brook have been included in the brief as there may be control and/or reference sites established on these rivers. The project is currently in the investigative stage with a focus on determining the current condition of the river and identifying what interventions are needed and where they should happen. A key component of the investigative stage is to determine the appropriate length of the demonstration reach. Recent discussions at a demonstration reach workshop in Canberra have indicated that the length of a reach should be governed by the fish fauna in question and the life cycles of the fish species involved. For example, will the reach be long enough to impact on the fish species present or will some species move outside the reach during their life cycle and be influenced by factors other than those occurring in the reach. This document informs the debate on demonstration reach length by characterising the species present within the geographical scope. 4. Geographical scope The geographical coverage of this review is the Border Rivers region of southern Queensland and northern New South Wales, from the Great Dividing Range (~ E), west to Toobeah (~ E), north to the southern slopes of Mt. Bodumba (~28 20 S), and south to the northern extremities of the New England Tablelands of northern New South Wales (~29.25 S) (Fig. 1). It includes waters from the Dumaresq and Macintyre Rivers, and Macintyre Brook, and their tributaries downstream of three major impoundments: Coolmunda Dam on Macintyre Brook; Glenlyon Dam on Pike Creek, a tributary of the Dumaresq River; and Pindari Reservoir on the Severn River (NSW), a tributary of the Macintyre River. 3

8 Figure 1: Geographical area where the proposed demonstration reach and control and reference sites will occur. Data supplied by QMDC. The region comprises two distinct hydrological zones. These were defined by Moffat and Voller (2002) as upper and lower foothills based on altitude, and can be reconciled to the Sustainable Rivers Audit (SRA) terminology developed by Whittington et al. (2001). The SRA defines geomorphology on two scales: Functional Process Zones and Valley Process Zones. Functional Process Zones are segments of river that have similar discharge and sediment regimes as defined by their gradient, stream power, valley dimensions and boundary material. Valley Process Zones are an aggregation of Functional Process Zones and are geomorphically similar regions within a river valley. They are described as regions of sediment source, transport and deposition. The zones contained within the Border Rivers region of the Murray- Darling Basin comprise: 1. the upper foothills/sediment source zone metres above sea level (a.s.l.) occurs within between Stanthorpe/Inverell and Inglewood/Texas/ Wallangara. This includes parts of the Macintyre Brook between Lake Coolmunda and Inglewood, the Dumaresq River between Pike Creek and Texas, the Macintyre River above Wallangara, and the Severn River (NSW) between Ashford and the Pindari Reservoir. These sections are predominantly pool and riffle habitats with semi-permanent flowing clear water. Submerged aquatic plants are common. 4

9 2. the lower foothills/sediment transport zone metres a.s.l. between the upper foothills zone and Toobeah on the Macintyre River. This region is characterised by a network of pools, intermittently connected during floods. The sides of pools are often clay or mud, while the bottoms are covered with sand or silt. Areas of bedrock and rubble are common. Waters are quite turbid, visibility is usually less than 30 centimetres, and submerged aquatic vegetation is patchy and uncommon. 5. Methods 5.1 Data sources Relevant data has been identified from ten different sources (Table 1) and these were contacted and asked to provide information on the sites sampled during each project, and the fish species found. These projects cover a range of sites from the lower foothills region of the Macintyre River below Goondiwindi, to the upper foothills region of the Macintyre, Dumaresq and Mole Rivers on the north-western slopes of the New England Tablelands. Table 1: Data sources contacted during compilation of this report. Abbreviations are listed below the table. Dataset # 1 Organisation Project Sampling location DPI&F 1 2 DPI&F/ Fisheries (NSW) 2 3 NRM&W 3 4 UNE 4 5 DNR (NSW) 5 / UNE 4 6 DPI&F DPI&F 1 CU 6 Meso-scale movements SRA tri-annual survey Critical flows for fish life-history strategies and flows Pindari Fish Monitoring Project Regional impoundment stocking surveys Macintyre River fishway surveys Anabranch geomorphology and productivity Dryland Refugia project Macintyre River between Goondiwindi and Mungindi Severn, Dumaresq, Macintyre and Mole Rivers, and adjacent tributaries Macintyre River between Goondiwindi and Mungindi Dumaresq and Mole Rivers Severn, Macintyre and Mole Rivers Coolmunda and Glen Lyon dams Boggabilla Weir Fishway, Macintyre River, (only studied pit-tagable sized fish) Anabranches to the Macintyre River below Goondiwindi 9 ARI (GU) 7 Macintyre River at Goondiwindi, Talwood-Boomi, Mungindi, and Collarenebri. 10 Qld Museum Historical records Opportunistic sampling in the Border Rivers region Abbreviations are: 1. DPI&F Department of Primary Industries and Fisheries, (QLD) 2. Fisheries (NSW) New South Wales Fisheries 3. NRM&W Natural Resources, Mines and Water (QLD) 4. UNE University of New England, Armidale 5. DNR (NSW) Department of Natural Resources, (NSW) 6. CU Canberra University, Canberra 7. ARI (GU) Australian Rivers Institute, Griffith University 5

10 These projects have collectively identified 16 native and four alien species in this section of the Border Rivers catchment (Table 2). This compares to the more than 35 native and 11 alien fish species in the Murray-Darling Basin, some of which are estuarine, and others are restricted in their distribution to highland habitats (more than 600 metres a.s.l.). Although data collection has occurred between 1901 and 2007, very few of the data sets have spatial, temporal or technical continuity (see Table 3). This limits their usefulness for making meaningful comparisons between data sets, or defining temporal changes in abundance for individual species. However, it does allow us to establish distributions for individual species. The SRA data set contains the most temporally contiguous data, although the number of sites and gear used varies in different years. However, there was a consistent level of sampling in the years of 2001, 2002 and A relative abundance index (Table 4) can be inferred by sub-sampling the data from these years across all uniform sample sites. This is not a true relative abundance index, though, because of the sampling gear diversity between years and sites, and the differences in levels of sampling between the two rivers. However, the inferred relative abundance index can be used to inform the species characterisation debate (see section 5.3) at both the Border Rivers regional scale (combined) and the two individual river valleys. Table 2: List of data sets with number of sites sampled, gear type used and river system sampled. Gear type is abbreviated in the table and abbreviations are expanded below the table. 6 Data set Year # sites Gear used River system M, F, BtE Macintyre River, below Goondiwindi M, F, BtE Macintyre River, below Goondiwindi M, F, BtE Macintyre River, below Goondiwindi BtE, CMT Beardy & Macintyre River, Tenterfield Creek BtE, CMT Severn River (NSW) CMT? Macintyre River BtE Macintyre, Dumaresq, Severn (NSW) Rivers BPE, BtE, Tenterfield Creek CMT, F BtE, CMT, F Beardy, Deepwater, Macintyre & Mole Rivers, Tenterfield Creek BtE, CMT, F Beardy, Deepwater, Macintyre, Mole & Severn (NSW) Rivers, Tenterfield Creek BtE, CMT, F Beardy, Dumaresq, Macintyre, Mole & Severn (NSW) Rivers, Tenterfield Creek LDN Dumaresq & Macintyre Rivers BtE, CMT, F Beardy, Deepwater, Dumaresq, Macintyre, Mole & Severn (NSW) Rivers, Tenterfield Creek MGn, CMT, F Severn (QLD), Dumaresq & Macintyre Rivers MGn, CMT, F Severn (QLD), Dumaresq & Macintyre Rivers MGn, CMT, Severn (QLD), Dumaresq & Macintyre F LDN, S, CMT Rivers Macintyre, Severn (NSW), Mole & Dumaresq Rivers

11 Data set Year # sites Gear used River system LDN, S, CMT Macintyre, Severn (NSW), Mole & Dumaresq Rivers LDN, S, CMT Macintyre, Severn (NSW), Mole & Dumaresq Rivers BtE Macintyre, Severn (NSW) & Mole Rivers BtE Macintyre, Severn (NSW) & Mole Rivers BtE Macintyre, Severn (NSW) & Mole Rivers BtE Macintyre, Severn (NSW) & Mole Rivers BtE Coolmunda and Glen Lyon dams current Bte Below and above Boggabilla weir, Macintyre River 8 Macintyre River between Goondiwindi and Talwood F, S, LDN Macintyre River, below Goondiwindi various unknown Border Rivers and tributaries Gear abbreviations are: BPE Backpack Electrofishing; BtE Boat Electrofishing; CMT Collapsible Minnow Traps; F Fyke nets; LDN Larval Drift Nets; M Mini fykes; MGn Multi-mesh Gill nets; S Seine net. Table 3: Combined list of all species of fish captured during the projects listed in Table 1. Presence in data set # refers to the number of each data set in Table 1. Species Presence in data set # Relative distribution Ambassis agassizii 1,2,3,4, 7,8,9 7 Bidyanus bidyanus 1,2,3, 5,6,7 6 Craterocephalus amniculus 2, 1 Craterocephalus stercusmuscarum fulvus 1,2,3,4,5, 7 6 Gadopsis marmoratus 3, 1 Galaxias olidus 2,3, 2 Hypseleotris spp 1,2,3,4,5, 7, 9 7 Leiopotherapon unicolor 1,2,3,4,5,6,7,8,9 9 Maccullochella peelii peelii 1,2,3,4,5,6,7, 9 8 Macquaria ambigua 1,2,3,4,5,6,7, 9 8 Melanotaenia fluviatilis 1,2,3,4, 7,8,9 7 Mogurnda adspersa 2,3,4, 3 Nemotalosa erebi 1,2,3,4, 7, 9 6 Philypnodon grandiceps 1, 1 Retropinna semoni 1,2,3,4,5, 9 6 Tandanus tandanus 1,2,3, 5,6,7 6 Carassius auratus* 1, 3,4, 7 4 Cyprinus carpio* 1, 3,4,5,6,7,8,9 8 Gambusia holbrooki* 1, 3,4,5, 9 5 Perca fluviatilis* 2, 1 *denotes alien species. 7

12 Table 4: Inferred Relative Abundance Index of native fish caught in the Macintyre and Dumaresq River valleys. Note that neither the flathead gudgeon (Philypnodon grandiceps) nor river blackfish (Gadopsis marmoratus) were recorded in data set 2 and thus are not included in this table. 8 Species Relative Abundance Macintyre River Relative Abundance Dumaresq River Relative Abundance combined Ambassis agassizii Bidyanus bidyanus *Carassius auratus Craterocephalus amniculus Craterocephalus stercusmuscarum *Cyprinus carpio Galaxias olidus *Gambusia holbrooki Hypseleotris spp Leiopotherapon unicolor Maccullochella peelii peelii Macquaria ambigua Melanotaenia fluviatilis Mogurnda adspersa Nematalosa erebi *Oncorhynchus mykiss *Perca fluviatilis Retropinna semoni Tandanus tandanus *denotes alien species. The presence/absence of any single species in each data set is governed by several factors. On a mechanical scale, this includes factors such as the selectivity of sampling methods used and sites sampled of any single species. On a whole of catchment spatial scale, the geographical distribution of each species will be an influencing factor. On a single river scale, access to suitable habitat via longitudinal and lateral connectivity will influence their presence/absence. On a reach scale, availability of suitable habitat will be a primary determining factor. Thus, the relative abundance of each species outlined in table three cannot be determined by an integrated assessment across all data sets. However, the number of data sets in which a species is found can be used to infer a relative distribution within the Macintyre- Dumaresq Rivers system, with any score below four indicating a restricted distribution (providing the reader accepts the mechanical and spatial factors influencing these results). A quick examination of table three highlights the very restricted distribution of five species: the Darling hardyhead (Craterocephalus amniculus), the river blackfish (Gadopsis marmoratus), the mountain galaxias (Galaxias olidus), the purple-spotted gudgeon (Mogurnda adspersa), and the flathead gudgeon (Phylipnodon grandiceps). The Darling hardyhead and the mountain galaxias are usually restricted to upper foothill waters (greater than 600 metres a.s.l.) in the Macintyre-Dumaresq River system. Similarly, the purple-spotted gudgeon, while declining in the southern Murray-Darling Basin, is found in the Border Rivers above 240 metres (a.s.l.) and would not be expected to be present in many of the data sets. The river blackfish is more common in waters above 150 metres (a.s.l.) in the southern Murray-Darling Basin, and it is listed as present in the Macintyre River (data set 4), but restricted to upper foothill waters. The flathead gudgeon has only been found in the lower foothill

13 waters of the Macintyre River (data set 1), and it appears to be quite rare in the Border Rivers region of the Murray-Darling Basin (see section 5.3, below). At least five native fish and one alien species have a restricted distribution within the Border Rivers region. While this might imply rarity of these species, their absence is probably explained by the diversity of sample location between data sets (i.e.: not sampling in locations where these fish might occur). The Relative Abundance Indices (RBI) in table four give a more accurate indication of species abundance, although being an arbitrary relative scale, it is not possible to compare between species (i.e.: a species with an RBI of four does not mean it is twice as abundant as a species with an RBI of two). From table four, it is apparent that species such as Philypnodon grandiceps and Gadopsis marmoratus are rare as they are not present in any of the SRA data for the Border Rivers, and only occur in one database. Ambassis agassizii and Bidyanus bidyanus are also rare species in both river systems. On the other hand, Hypseletoris sp., Melanotaenia fluviatilis and Craterocephalus amniculus are all relatively abundant in the Border Rivers region. We can also infer that Nematalosa erebi is relatively common in the Macintyre River, but not as common in the Dumaresq River (only three have been recorded in the SRA database since 2000 for the Dumaresq River and tributaries). Other species with divergent abundance, depending on the river, include Leiopotherapon unicolor and Macquaria ambigua, which are more common in the Macintyre River, and Galaxis olidus, Mogurnda adspersa and Tandanus tandanus, which are more common in the Dumaresq River. 5.2 Data gaps Examining table one, it is apparent that there are data gaps, particularly from Macintyre Brook. The SRA database contains the only record of sampling in this water body, at Whetstone Weir which was sampled in This water body has not been surveyed adequately to determine its species suite. Lake Coolmunda, upstream on Macintyre Brook, has been sampled for stocked fish occasionally by DPI&F officers using electrofishing, but there are no records for any other instream sampling. If this is to be used as a suitable demonstration reach control or reference site (as has been suggested by QMDC), then there will have to be some survey work to determine species assemblages and habitat condition. Other areas in need of further information are the impacts of declining lateral connectivity (caused by drought and floodplain alteration) on fish abundances, and the impacts declining longitudinal connectivity (caused by barriers) to fish distribution and abundance. There are numerous weirs and dams on each river, and the degree of impeded connectivity of each needs to be ascertained. This is discussed further in section Species Biology/Ecology The idealised fish life-cycle involves changes in body-shape, behaviour, habitat, diet and sexual development, which comprise the life-history, (Fig. 2). At each stage in their development, there may be a need to move to a more suitable habitat. Access to this habitat may provide suitable food and shelter resources, or shelter from predation, thereby enhancing the chance of survival and recruitment. These habitats can be in backwaters or main channel habitats within the mainstream, or offstream in anabranches or billabongs. 9

14 Collectively these ontogentic shifts can be divided into three main functions: reproduction; nursery phase; and recruitment. Some or all of these phases will be applicable to the life-histories of most native freshwater fish. Recruitment Sexually mature adult population May migrate to spawning grounds Reproduction eggs May be a juvenile migration to adult habitat Nursery May be affixed to substrate or drift in current to a suitable nursery area juveniles May be a movement to postlarval or adult habitat larvae Figure 2: The idealised model of a fish life-cycle, highlighting potential temporal patterns for migration. Much of the biology/ecology information for each species below comes from publications on studies within the Murray-Darling Basin, supplemented by excellent information available from studies within their distribution elsewhere (e.g.: Pusey et al. 2004). The combined information is presented species by species below, following the order presented in table 2. Native Species Ambassis agassizii Common Name/s: Agassiz s glassfish, Olive perchlet Family: Chandidae Discovered by: Steindachner, 1866 Most of the information about this species comes from studies in the eastern Queensland coastal drainages. However, it is a fish with a wide distribution along eastern Australian coastal rivers between northern Queensland and the central-north coast of New South Wales (Pusey et al. 2004). Historically, it enjoyed a wide distribution within the Murray-Darling Basin, but is now considered to be threatened in the southern states (Vic, S.A.), and uncommon in southern New South Wales (Harris & Gehrke 1996). It has been recorded in the Dumaresq and Macintyre Rivers (Morris et al., 2001), but is quite rare in both distribution (Fig. 3) and abundance (Table 4). Ambassis agassizii is commonly found in pools of water with a moderate depth (around 0.5 metres depth) and slow velocity (less than 0.05 m.sec -1 ). In south-eastern Queensland, it has been collected in waters of centimetres depth and velocities up to 0.44 m.sec -1 (Pusey et al. 2004). It is a pelagic species, 10

15 commonly found over fine sediment substrates (mud, sand, clay). While not demonstrating a strong affinity to banks, it is commonly found close to cover such as aquatic macrophytes and filamentous algae, occasionally aggregating into loose schools during upstream movements (Morris et al. 2001). A. agassizii has been shown to have a wide temperature ( C), and oxygen level tolerance (0.3 to 19.5 mg.l -1 ) (Pusey et al. 2004). Figure 3: Recorded distribution of Ambassis agassizii in the Border Rivers region of the Murray-Darling Basin, based on the ten datasets in table one. This species has a lifespan of approximately four years, but are sexually mature after 12 months (Moffatt & Voller, 2002). A. agassizii is a serial spawner with an extended reproductive season from spring through to autumn (Milton & Arthington 1983), but with a peak in late spring to early summer. Spawning cues are unknown, but aquariabased research has shown that it can be triggered by increasing photoperiod and water temperature. Thus, it may not be driven by rising water levels or flooding. Spawning has been shown to occur amongst fine-leafed vegetation (Leggett & Merrick 1986). On the east coast, larvae have been observed in large schools in surface waters of tributaries, moving into the mainstream aquatic vegetation after reaching 8-12 millimetres in length (Pusey et al. 2004). They have been reported to disperse and forage during darkness, and aggregate around cover during daylight (Allen & Burgess 1990). A. agassizii has been observed to undertake upstream movements in east coast waters, after flood events, but movement is restricted to periods during lower velocity after flood events. It has been observed to actively move into offstream billabongs in the lower foothill waters of the Macintyre River during in-bank flow events (Hutchison, pers. comm.). It is known to feed extensively on aquatic insects and micro-crustaceans (Hansen 1995). Pusey et al. (2004) have reported that threats to its survival in the Murray-Darling Basin include alien fish species, habitat degradation and flow regulation. Alien species are thought to actively feed on larvae and juveniles. Disturbance of riparian habitat leading to increased turbidity can reduce macrophyte vegetation. Loss of flow or artificial flows during breeding can cause loss of eggs from macrophyte vegetation, reducing the chance of successful recruitment. Bidyanus bidyanus Common name/s: Silver perch, black bream 11

16 Family: Terapontidae Discovered by: Mitchell 1838 This species once had a wide distribution throughout the Murray-Darling Basin, but has declined markedly throughout the lower Murray-Darling Basin. Self-sustaining populations are now thought to be restricted to the Queensland Murray-Darling Basin (Moffatt & Voller 2002), and the upper Murray River. In the Border Rivers region, it has a relatively wide distribution (Fig. 4) assisted, in part, by local stocking programs, but a rare abundance (Table 4). A shoaling species, they are commonly found in slow or standing pools, often in association with large woody debris, or reeds, as well as fast flowing turbid waters (Koehn & O Connor 1990) of lowland rivers. They are often found in open water, or in association with submerged or emergent aquatic vegetation. Bidyanus bidyanus is reportedly able to tolerate a wide range of temperature ( C) (Lake 1966a). Figure 4: Recorded distribution of Bidyanus bidyanus in the Border Rivers region of the Murray-Darling Basin, based on the ten datasets in table one. B. bidyanus have been reported to live up to 26 years (Mallen-Cooper & Stuart 2003), and can reach up to eight kilograms, although less than 16 years and kilograms is more common in the southern Murray-Darling Basin (Morris et al. 2001). Males mature at approximately three years, while females take up to five years to mature. These fish are known to school and migrate upstream to spawn (Allen et al. 2002), usually in spring or summer when water temperatures exceed 23 0 C (Lake 1966a), on the back of a receding flood. Eggs are pelagic and readily displaced by current. Larvae are free swimming by five days old and strongly phototactic. This is thought to assist their wide distribution across the flood plain. Post-larval juveniles are found 18 days after the eggs are fertilized. Juveniles feed primarily on filamentous algae and phytoplankton, although some zooplankton is also in their diet (Lake 1966b). Adults feed on aquatic micro-invertebrates and algae (Moffatt and Voller 2002). B. bidyanus is listed as critically endangered in Victoria, and vulnerable in New South Wales (Morris et al. 2001). Factors leading to a decline in B. bidyanus populations throughout the Murray-Darling Basin include instream habitat degradation, alterations to river flow and water temperature regimes, and instream barriers (such as weirs and dams) to spawning migrations (Morris et al. 2001). These threats are thought to impact on both spawning and recruitment success, leading to localised population extinctions. 12

17 Craterocephalus amniculus Common name: Darling River hardyhead Family: Atherinidae Discovered by: Crowley & Ivanstoff, 1988 Craterocephalus amniculus is endemic in the upper tributaries of the Darling River, including the Condamine, Peel, Namoi, Macintyre and Cockburn Rivers, and Warialda, Tenterfield and Boiling Down Creeks (Ivanstoff & Cowley 1966). In these upland regions, it is relatively common (Table 4), but restricted in its distribution (Fig. 5). Adults are found in small schools (10-15) in clear slow-flowing waters or in weed at the edges of such waters (Ivanstoff & Cowley 1966). Little is known about the biology or ecology of this species other than it is commonly found in waters less than 700 metres a.s.l., but can be found as low as 240 metres a.s.l. (Dean Gilligan pers. comm.). It is probably very similar in biology to the fly-specked hardyhead (see below), but exploits a different ecological niche. Figure 5: Recorded distribution of Craterocephalus amniculus in the Border Rivers region of the Murray-Darling Basin, based on the ten datasets in table one. Note that additional records of C. amniculus can be found in the upland waters to the east of map. Craterocephalus stercusmuscarum fulvus Common name: Fly-specked hardyhead Family: Atherinidae Discovered by: Ivanstoff, Crowley and Allen, 1986 The Craterocephalus stercusmuscarum fulvus is widely distributed across northern Australia from Timor Sea drainages in Northern Territory, to the east coast rivers of Queensland, down to the Tweed River, and historically inland into much of the Murray-Darling Basin (Pusey et al. 2004). In 1986, it was divided into two subspecies: a northern variant (Craterocephalus stercusmuscarum stercusmuscarum) and a southern variant (Craterocephalus stercusmuscarum fulvus) by Ivanstoff et al. (1986). The southern variant, C.s. fulvus, is moderately common in coastal rivers of southern Queensland from the Elliot to the Tweed Rivers. Historically, it occurred throughout much of the Murray-Darling Basin, although it was patchy in its distribution. Currently, it is found in patches of the Queensland and northern New South Wales Murray-Darling Basin including the Border Rivers, Moonie and Condamine-Balonne catchments (Moffatt & Voller 2002). It has a wide distribution 13

18 within the Border Rivers region (Fig. 6), and has a relatively low, but not rare, level of abundance (Table 4). It is also found in patches in Victoria and South Australia, but has disappeared from much of the southern Murray-Darling Basin. On the east coast, it is a pelagic schooling species, found in pools or runs with moderately low flow (less than 0.4 to 0.86 m.sec -1 ) in depths of 10 to 60 centimetres, over medium to fine substrates (sand, fine to course gravel) (Pusey et al. 2004). It has been collected in areas closely associated with macrophyte vegetation, filamentous algae or submerged marginal vegetation. It has been reported to congregate in areas where streams flow into still water (Allen et al. 2002). C. s. fulvus is known to tolerate temperatures ranging from 12.4 to C, and oxygen levels of 2.9 to 19.5 mg.l -1 (Pusey et al. 2004). Figure 6: Recorded distribution of Craterocephalus stercusmuscarum fulvus in the Border Rivers region of the Murray-Darling Basin, based on the ten datasets in table one. C. s. fulvus live up to three years, but are sexually mature within 12 months. They have an extended spawning season from late winter through to summer, but a peak occurs in late winter to early spring (Pusey et al. 2004). The spawning stimulus is unknown, but corresponds to increasing photoperiod and water temperature. Spawning takes place in aquatic macrophytes and submerged vegetation (Milton & Arthington 1983). Evidence from fishway sampling on the east Queensland coast indicates that for the northern variant (C. s. stercusmuscarum), low numbers move all year round, but a peak in upstream migration occurs during summer as a consequence of flow events (Stuart & Berghuis 1999). These movements are considered to be dispersal/recolonisation events, especially after dry periods. Movements of the southern variant are less well studied. C. s. fulvus is a microphagic carnivore consuming aquatic insects and micro-crustaceans, as well as aquatic algae and macrophytes. It is listed as endangered in New South Wales and Victoria (Treadwell and Hardwick 2003). Its spawning habitats are threatened by erosion and increased siltation (Pusey et al. 2004). Being a facultative migrator, it is susceptible to instream barriers that prevent dispersal/recolonisation. It is vulnerable to competition and predation interactions with alien species. Unseasonable flows during peak reproductive periods are likely to diminish reproductive success (Morris et al. 2001). 14

19 Gadopsis marmoratus Common name: River blackfish Family: Discovered by: Richardson, 1848 Gadopsis marmoratus enjoy a wide distribution throughout the lower Murray-Darling Basin and remnant populations can still be found in the uplands zones of the Condamine and Border Rivers of Queensland and New South Wales (Moffatt & Voller 2002). It is also found in many coastal streams of south-east Australia and Tasmania (Morris et al. 2001). While common in patches, it has suffered a general decline in total numbers throughout its range, most probably due to siltation of habitat and over fishing (Morris et al. 2001). It is quite rare within the Macintyre-Dumaresq River systems (Table 4), occurring only in the upland waters of tributaries of the Dumaresq River (Fig. 7). They exhibit strong site fidelity, with most fish being associated with bank undercuts during daytime, and moving to open and boulder habitats at night, presumably to feed (Khan et al. 2004). It is a hardy species commonly found in slow flowing stream pools at depths from 20 to 60 centimetres (Koehn et al. 1994). It has a preference for slow moving water (less than 0.20 m.sec - 1 ), but has been recorded in waters of 0.04 to 0.34 m.sec -1, and in temperatures ranging from five to 20 C. Adult G. marmoratus are tolerant of slightly brackish water (up to 10 ppt) (Allen et al.2002). Figure 7: Recorded distribution of Gadopsis marmoratus in the Border Rivers region of the Murray-Darling Basin, based on the ten datasets in table one. In the northern Murray-Darling Basin, G. marmoratus has a reported lifespan of over six years, but is reproductively mature at two years (Moffatt & Voller 2002). The spawning season runs from November to January, when adults form pairs and spawn in hollow logs (Jackson 1968). The male parent actively guards the eggs until hatching. Larval and juvenile habitats are unknown, but given the highly territorial behaviour of adults, it is highly likely that they undertake dispersal migrations (Morris et al. 2001). This is a predatory species, feeding on aquatic macroinvertebrates, terrestrial insects and small fish, such as gudgeons (Moffatt & Voller 2002). G. marmoratus has undergone a marked decline in its range, and probably abundance throughout the Murray-Darling Basin. This has been attributed to a high susceptibility to over fishing associated with strong site fidelity and low fecundity. A decline of habitat quality associated with riparian destruction and agricultural run-off, causing 15

20 siltation of habitat may also be a contributing factor (Morris et al. 2001). It has a preference for lower water temperatures and thus is restricted to the uplands region of Condamine and Severn (QLD) Rivers (Michael Hutchison, pers. comm.). Galaxias olidus Common name: Mountain galaxias Family: Discovered by: Günther, 1866 Galaxias olidus are found in upland streams draining both sides of the Great Dividing Range from southern Queensland to South Australia and Kangaroo Island (Allen et al. 2002). It has become scarce in areas accessible to alien fish species, such as brown trout (Salmo trutta) and rainbow trout (Oncorhynchus mykiss). It is quite rare in the Macintyre River system, being restricted to waters that are higher than 400 metres (a.s.l.) (Table 4). It is much more common in the upper regions of the Dumaresq River system (Fig. 8). Adults are found in clear pools of mountain streams, shoaling around rocks or logs, or in open water (Allen et al. 2002). They are tolerant of very cold water, having been caught in ice-melt fed streams (Koehn & O Connor 1990). Figure 8: Recorded distribution of Galaxias olidus in the Border Rivers region of the Murray-Darling Basin, based on the ten datasets in table one. G. olidus live for four years, and are reproductively mature by one (Moffatt & Voller 2002). Spawning occurs from August to late October when water temperatures range between eight and 10 C (O Connor & Koehn 1991). Adults move upstream to shallow riffle areas to lay demersal, adhesive eggs over gravel substrates (Allen et al. 2002). Eggs hatch about 46 days later and feeding juveniles are found another eight days later (O Connor & Koehn 1991). Juveniles and adults can be found shoaling in the same pools (McDowell & Fulton 1996). Their diet consists primarily of aquatic invertebrates and terrestrial invertebrates, with larger fish consuming a greater proportion of terrestrial vertebrates, especially in areas with substantial overhanging riparian vegetation (Cadwallader et al. 1980). The biggest threat to the survival of this species is predation from alien fish such as brown trout and rainbow trout (Koehn & O Connor 1990). This, combined with barriers to spawning migrations, has probably led to localised population depletions. The loss of riparian vegetation, leading to declining dietary sources of terrestrial invertebrates, and increased sediment loads and subsequent loss of spawning habitat may have also depressed successful recruitment for some populations. 16

21 Hypseleotris spp. Common name: Carp gudgeons Family: Eleotridae Discovered by: Hypseleotris spp. of the Murray-Darling Basin are a complex group of small fish. While initially divided into western carp gudgeon (Hypseleotris klunzingeri Ogilby, 1898), Midgley s carp gudgeon (Hypseleotris species one), lake s carp gudgeon (Hypseleotris species two), and Murray-Darling carp gudgeon (Hypseleotris species three), recent genetic research indicates that they may in fact be a group of five species, but capable of inter-breeding (Bertozzi et al. 2000). Murray-Darling Basin Commission protocols require these species be classified as a group (Hypseleotris spp.) until further work can elucidate the specific differences. Collectively, Hypseleotris spp. have a Murray-Darling Basin-wide distribution, as well as occurring in many coastal streams from central New South Wales to central Queensland, the Bulloo River and Cooper Creek (Lake Eyre Basin). They are one of the most widely distributed (Fig. 9) and abundant species (Table 4) within the Macintyre-Dumaresq River system. They have also been widely translocated (Pusey et al. 2004). They occur in a variety of lentic and lotic habitats. Although generally considered to be a benthic species, they also occur in the water column over a range of depths from 10 to 50 centimetres. They are found on sediments ranging from mud to coarse gravel, usually associated with aquatic vegetation, leaf litter, undercut banks or root balls (i.e.: structures that may afford protection). Figure 9: Recorded distribution of Hypseleotris spp. in the Border Rivers region of the Murray-Darling Basin, based on the ten datasets in table one. Adults are tolerant of a wide range of water temperatures (eight to 32 C) being caught in southern mountain streams and arid desert water holes. They are found in anoxic to hyper-oxic water (0.6 to 13 mg.l -1 ), and in a range of turbidities (0.5 to more than 680 NTU) (Pusey et al. 2004). Hypseleotris spp. are a short lived group (two to three years) of small fish (seven centimetres, maximum). Adults are sexually mature by 12 months and spawning occurs from spring to early summer in the Murray-Darling Basin when water temperatures rise above 22.3 C (Anderson et al. 1961). Males establish a territory and pair with a female. The adhesive eggs are attached to the underside of large gravel or 17

22 aquatic plants and are actively defended by the male during incubation (Cadwallader & Backhouse 1983). Hypseleotris spp. are known to spawn during floodplain inundation and this, coupled with an extended spawning season, may be regarded as an adaptation to the unpredictable onset of the wet season flooding. Juveniles have been caught in similar locations to the adults. Juveniles have been found in the lower foothills zone of the Macintyre River in November 2005, in the absence of any natural flow (Hutchison, pers. comm.). Hypseleotris spp. are a macrophagic carnivore, feeding primarily on aquatic insects, microcrustaceans, and other macroinvertebrates (Balcombe & Humphries 2006). Small scale mass migrations have been recorded in coastal streams (Berghuis et al. 2000) where large aggregations have been observed below instream barriers. However, these were postulated to be recolonisation movements after downstream displacements by flood events. Being a small fish, Hypseleotris spp. form a significant link in freshwater trophodymanic webs. Although not considered to be threatened, their larvae are susceptible to predation from alien fish species such as eastern gambusia (Gambusia holbrooki). Also, being a periodically migratory fish (Hutchison pers. comm.), they would be susceptible to instream barriers to passage. Leiopotherapon unicolor Common name/s: Spangled perch, bobby Family: Terapontidae Discovered by: Günther 1859 Leiopotherapon unicolor are one of the most widely distributed freshwater fish in Australia (Pusey et al. 2004), being found in coastal rivers from Geralton (Western Australia), north and east to Newcastle (New South Wales), and inland in the Murray- Darling, Lake Eyre, Bulloo-Bancannia and Western Plateau drainage Basins. It is found in a wide variety of habitats including desert springs, bores, billabongs, impoundments, channels, rivers and streams. It is widely distributed within the Macintyre-Dumaresq River system (Fig. 10), but is more abundant in the Macintyre than the Dumaresq River (Table 4). Its distribution south is thought to be limited by the 4.4 C isotherm (Lewellyn 1963). It is commonly found over muddy, sand and fine gravel substrates in still to low velocity water bodies and can be found in isolated water holes that may not have been connected to the main channels for many years. It is usually associated with a benthic habitat, or in association with aquatic macrophytes, large woody debris or root masses. It is a hardy fish capable of tolerating a wide range of salinity (0.2 to / 00 ), oxygen (0.4 mg.l -1 ) and turbidity levels (1.5 to more than 680 NTU) (Pusey et al. 2004). 18

23 Figure 10: Recorded distribution of Leiopotherapon unicolor in the Border Rivers region of the Murray-Darling Basin, based on the ten datasets in table one. L. unicolor live up to five years and are sexually mature within the first 12 months. They breed during the summer wet season cued by water temperatures rising above 20 C (Lewellyn 1963). Adults broadcast spawn over an inundated floodplain and the eggs sink to the substrate. The larvae hatch within three days and are feeding within four days of spawning. Juveniles metamorphose by 28 to 35 days of hatching (Pusey et al. 2004). They are a highly predatory fish consuming a variety of aquatic organisms and around 10 % vegetable matter (algae, aquatic macrophytes and terrestrial vegetation). L. unicolor are one of the most active migratory freshwater fish in Australia. They have been observed to travel up very small streams, leaping up to one metre to surmount low instream barriers such as culverts. They are thought to undertake both spawning and dispersal migrations (Pusey et al. 2004). This fish is not considered to be threatened in the Murray-Darling Basin, but is known to be of less abundance in regulated rivers when compared to unregulated rivers (Gehrke 1996). Maccullochella peelii peelii Common name: Murray cod Family: Percichthyidae Discovered by: Mitchell 1838 This species was once widely distributed within the Murray-Darling Basin from the upper Condamine River down to the lower reaches of the Murray River in South Australia. However, over fishing and river regulation have led to a decline in populations throughout its native range (Rowland 1989), although increasing numbers have been recorded recently from the Queensland section of the Murray-Darling Basin. It is widely distributed (Fig. 11) and reasonably common (Table 4) within the Border Rivers region, but, as it is the target of a large stocking program, it is difficult to tell if this is a result of natural populations, or the stocking programs. It is a large fish growing to over one metre in length with historical records of it being more than 100 kilograms (Moffatt & Voller 2002). This species is found in a range of habitats from small clear rocky streams in the uplands of northern New South Wales to large turbid waterholes of the lowland 19

24 creeks and rivers. It prefers deeper waterholes with large woody debris, rocks or an undercut bank. There is little information available about its environmental tolerances. Figure 11: Recorded distribution of Maccullochella peelii peelii in the Border Rivers region of the Murray-Darling Basin, based on the ten datasets in table one. Being the largest freshwater fish in Australia, it is not surprising that Maccullochella peelii peelii can live for more than 40 years, and that they take up to four years to attain sexual maturity. In the southern Murray-Darling Basin, they spawn between October and December, cued by water temperatures rising above 15 C (Koehn & Harrington 2006). They are an annual spawner, but spawning success is closely linked to high flow events during the breeding season (Rowland 1998). The mature adults will migrate upstream some 80 to 100 kilometres into a small anabranch, forming pairs that spawn adhesive eggs in hollow logs, on snags, under rocks or firm clay (Moffatt & Voller 2002). The male will remain and guard the nest until hatching. Larvae disperse downstream in the current, becoming active feeders 10 to 19 days after hatching (Humphries 2005). Juveniles feed on zooplankton, aquatic insects and small native fish such as gudgeons. Adults feed on other fishes, crustaceans and molluscs (Harris & Rowland 1996). M.p. peelii is threatened by several factors including over fishing, loss of spawning habitat, inability to migrate past barriers, river regulation and cold water pollution (Morris et al. 2001). In the mid 1800s to 1930s it was subject to over fishing by a large commercial fishery, and post 1950s has been over fished by a recreational sector that consider it to be the premier freshwater angling fish in eastern Australia (Rowland 1989). Floodplain management since the early 1900s, with the construction of weirs and dams, has severely inhibited the annual spawning migration, changed the flow regime to water on demand into the irrigation industry, often leading to cold water releases from dams that are detrimental to the survival of larvae (Todd et al. 2005), and reduced the frequency and magnitude of flooding, which has had a detrimental effect on larval survival. Recent agricultural developments within the floodplain environment have removed access to many anabranch sites for spawning and removed many snags from rivers and creeks. 20

25 Macquaria ambigua Common names: Golden perch, yellowbelly Family: Percichthyidae Discovered by: Richardson, 1845 Macquaria ambigua is widely spread throughout the Murray-Darling Basin, except at higher altitudes and above large impoundments (Allen et al. 2002), and the Fitzroy River in central Queensland. A separate species, Macquaria species B, occurs in the Lake Eyre and Bulloo River catchments (Musyl & Keenan 1992). M. ambigua has been widely translocated to many rivers and impoundments in south east Queensland and northern New South Wales. They are widely distributed within the Macintyre- Dumaresq River system (Fig. 12) but appear to be much more common in the Macintyre than the Dumaresq River (Table 4). Like cod, they have been extensively stocked and this dichotomy of abundance may reflect stocking activity in Queensland waters. They inhabit generally turbid, slow-flowing rivers, creeks, billabongs and backwaters (Morris et al. 2001). They prefer deep pools containing cover such as dead trees or fallen timber, undercut banks or rocky ledges (Moffat & Voller 2002). They are reported to withstand temperatures ranging from four to 36 C and salinities up to 33 0 / 00 (Merrick & Schmida 1984), oxygen levels ranging from three to 15 mg.l - 1, and turbidity above 400 NTU (Pusey et al. 2004). Figure 12: Recorded distribution of Macquaria ambigua in the Border Rivers region of the Murray-Darling Basin, based on the ten datasets in table one. Research from the middle Murray River region indicates that M. ambigua live for more than 26 years, with males and females attaining sexual maturity in two and four years respectively (Mallen-Coooper & Stuart 2003). They are active spawners from October to March, cued by the addition of water to their environment, after water temperatures have exceeded 23.6 C (Lake 1966a). This does not have to be a major flood (King 2003), but inundation of dry ground leads to important plankton blooms, which are used as food by the larvae (Lake 1966a). Larvae shoal up after four days then disperse at five days and commence feeding on microcrustaceans, zooplankton and aquatic insects (Cadwallader & Backhouse 1983). Juveniles and adults are opportunistic carnivores, feeding on fish, shrimp and yabbies (Morris et al. 2001). M. ambigua are macro-scale migrators, with adults travelling long-distances upstream during high flow events to colonise habitats that are only occasionally accessible. Juveniles and sub-adults make strong upstream dispersal migrations during spring and 21

26 summer, stimulated by rising water levels. Peak migratory behaviour is during evening and early morning (Harris & Rowland 1996). M. ambigua populations are threatened by loss of access to floodplains, unnatural flows and barriers to migration. Weir and dams obstruct natural migratory paths to isolated upstream water holes. Unseasonable water flows, released for irrigation, can make the floodplain environment uninhabitable for larvae and juvenile M. ambigua, which diminishes recruitment success (Morris et al. 2001). Melanotaenia fluviatilis Common name/s: Murray River rainbowfish, Crimson-spotted rainbowfish Family: Melanotaeniidae Discovered by: Castelnau, 1868 This small species (up to 90 millimetres, maximum) occurs throughout the Murray- Darling Basin in all catchments, but generally in the lowland areas (Allen 1996, Moffat & Voller 2002). There is unpublished genetic work by Unmack (cited by Morris et al. 2001) that suggests the Melanotaenia fluviatilis from the Paroo, Warrego and upper Darling Rivers is actually the desert rainbowfish, M. splendida tatei (Zeitz 1896). However, Allen et al. (2002) consider M. splendida tatei to be confined to the central desert areas and Lake Eyre Basin (Coopers Creek, Diamantina and Georgina Rivers), and the central-eastern Barkley Tablelands of the Northern Territory. M. fluviatilis is considered to be relatively abundant in much of the northern Murray- Darling Basin, but is declining in the Victorian and New South Wales regions (Allen et al. 2002). It is both abundant (Table 4) and widely distributed in the Macintyre and Dumaresq Rivers (Fig. 12). It is found in a wide range of habitats including rivers, streams, billabongs, swamps and drains, preferring slow-flowing or still waters with dense aquatic vegetation (Cadwalader & Backhouse 1983). This is the most southerly ranging species of Melanotaenia and the only species adapted to cold winter temperatures (eight to 28 C) (Allen et al. 2002). They have been caught in quite turbid waters (0.5 to 528 NTU). Figure 13: Recorded distribution of Melanotaenia fluviatilis in the Border Rivers region of the Murray-Darling Basin, based on the ten datasets in table one. This species lives for three years, attaining sexual maturity after 12 months. They breed between October and January cued by rising water temperatures. Females can 22

27 spawn up to three to four times per day over several days. Adults pair for a single spawning of up to 10 demersal eggs. Eggs adhere to aquatic plants and larvae hatch in six to seven days. Larvae are carnivorous, aggregating near the water surface to feed on zooplankton (Merrick & Schmida 1984). Adults feed on aquatic invertebrates and terrestrial insects (Cadwalader & Backhouse 1983). Although not considered to be a migrating species, large numbers have been observed to aggregate below weirs during summer (M. Hutchison, pers. comm.). The biggest threat to this species is the loss of aquatic vegetation, from riparian disturbance, leading to erosion/siltation. Other factors include larval predation by alien species such as the eastern gambusia (G. holbrooki). It is also possible that instream barriers may prove to be a hindrance to re-colonisation of upstream habitats. Mogurnda adspersa Common name/s: Purple-spotted gudgeon, southern purple-spotted gudgeon Family: Eleotridae Discovered by: Castelnau, 1868 Mogurnda adspersa was once distributed in eastern coastal streams from central Cape York, south to the Clarence River, northern New South Wales, and west into the Murray-Darling Basin and some coastal drainages of South Australia (Pusey et al. 2004). Its range in the Murray-Darling Basin is now severely restricted to the upper Condamine River, the Border Rivers region of Queensland-New South Wales and patchy populations in the southern Murray-Darling Basin. It is listed as endangered in New South Wales and critically endangered in Victoria (Morris et al. 2001). It is more common to the upper foothills regions of the Dumaresq River (Fig. 14) than the Macintyre River (Table 4). It is reported to be a shallow (10 to 60 centimetre) pooldwelling species commonly found in slow-flowing (less than 0.1 m.sec -1 ) weedy areas (Briggs 1998). However, Moffat and Voller (2002) state M. adspersa are more commonly associated with rock or cobble habitat in the Border Rivers uplands region (higher than 200 metres a.s.l.). Figure 14: Recorded distribution of Mogurnda adspersa in the Border Rivers region of the Murray-Darling Basin, based on the ten datasets in table one. It is tolerant of a wide range of temperatures (10.5 to 20 C) in the Murray-Darling Basin (Briggs 1998), but able to cope with temperatures up to 32 C in south-eastern Queensland (Pusey et al. 2004), and tolerant of low oxygen levels (0.6 to 12.8 mg.l - 1 ). Although it can tolerate high turbidity (0.2 to 200 NTU) (Pusey et al. 2004) it 23

28 appears to prefer clear water (averaging 6 NTU). M. adspersa are tolerant of a wide range of salinities and have been captured from uplands freshwater streams (840 metres a.s.l.), down to estuarine waters (Hansen 1988). This is a small fish (up to 12 centimetres maximum) that lives up to four years of age. It is mature at 12 months (Moffatt & Voller 2002) and spawns during the wet season from November to April (Allen et al. 2002). It is a serial spawner, and, although spawning cues are unknown, they are speculated to be rising temperature (more than 20 C) and photoperiod (Hansen 1988). Flooding is not essential to spawning (Larson & Hoese 1996). Males are territorial, and establish a nesting site. Adults pair off and the female lays between 100 to 1300 small adhesive eggs onto the hard substrate (rocks, woody debris, broad-leafed aquatic vegetation). The male will fan and defend the nest until hatching three to nine days later (temperature dependant). The diet of juveniles (mainly zooplankton) and adults are reported to be similar (Allen et al. 2002), although prey size is obviously related to gape size. Prey consists primarily of aquatic insects, terrestrial invertebrates, molluscs and other microinvertebrates (Pusey et al. 2004). There is very little literature on their movements, although Pusey et al. (2004) speculate that on the east coast, they may have a facultative mass dispersal phase, although they are rarely reported from fishway studies. The decline of M. adspersa in the Murray-Darling Basin has been correlated with the invasion of eastern gambusia (G. holbrooki) by Wager and Jackson (1993). It is highly likely that M. adspersa are also susceptible to instream barriers to dispersal migration/recolonisation, and rapid fluctuations in water levels brought on by river regulation during reproduction and recruitment. Riparian disturbance, causing sedimentation, and loss of habitat may also influence recruitment success. Nemotalosa erebi Common name/s: Bony bream, bony herring Family: Clupeidae Discovered by:günther, 1868 Nemotalosa erebi is the most widely spread freshwater fish in Australia (Pusey et al. 2004). It is found in coastal drainages from the Pilbara and Kimberly regions of Western Australia, across the Northern Territory, including its arid interior, the arid interior of South Australia, the Murray-Darling Basin and coastal Queensland rivers as far south as the Albert River. In these areas it is an abundant species, one of the few to thrive since European settlement (Moffatt & Voller 2002). It is found in a range of habitats from saline lakes to lowland rivers and streams, billabongs and floodplains. It thrives in impoundments and is only restricted from higher, cooler, fast-flowing waters possibly because of an intolerance to lower water temperatures (Puckridge & Walker 1990). It is a schooling species and, in the Murray-Darling Basin, is often found in slow or still water commonly swimming in open water or near large woody debris. It is widely distributed in the lowland and lower foothills zones of both the Macintyre and Dumaresq Rivers (Fig. 15), but is much more abundant in the Macintyre River (Table 4). It is tolerant of a wide range of salinities and temperature (eight to 29 C), but is intolerant of low oxygen levels, being the first fish to perish when ephemeral water bodies begin to dry out (Allen et al 2002). It has been found in highly turbid waters (more than 600 NTU) in the Macintyre River (Hutchison, pers. comm.). 24

29 Figure 15: Recorded distribution of Nemotalosa erebi in the Border Rivers region of the Murray-Darling Basin, based on the ten datasets in table one. Although usually a smaller fish (around 10 to 15 centimetres maximum) it is known to reach more than 40 centimetres and eight years of age (Moffatt & Voller 2002). It is sexually mature at two years of age and spawns in summer, irrespective of the water level, with the onset of water temperatures rising above 18 to 20 C (Puckeridge & Walker 1990). Spawning takes place in shallow still water such as sandy embayments/backwaters. Eggs are scattered randomly and develop rapidly, suggesting that larval survival is closely reliant on availability of phytoplankton blooms. Adults are primarily detritiphores, but also consume algae, microcrustaceans and aquatic insects. N. erebi are a highly migratory species, able to rapidly recolonise habitat when conditions are conducive. They are one of the most common species found traversing fishways (Mallen-Cooper et al. 1995, Pusey et al. 2004) but are restricted to these migrations during daylight hours only. Their migrations are divided into reproductive movements by adult fish and recolonisation movements by juvenile and sub-adult fish. N. erebi is not threatened in the Murray-Darling Basin. It is an important component of the trophic web in any freshwater habitat within its distribution. Being highly migratory, it is susceptible to barriers to migration, and conditions that lead to anoxic water chemistry. Philypnodon grandiceps Common name: Flathead gudgeon Family: Eleotridae Discovered by: Krefft, 1864 Philypnodon grandiceps are a widely spread species, occurring in coastal drainages from the Burdekin River in central northern Queensland to the Gawler River, South Australia, Kangaroo Island and northern Tasmania. It is also patchy throughout much of the Murray-Darling Basin (Pusey et al. 2004, Allen et al. 2002). It has not been recorded in the Border Rivers region before the Murray-Darling Basin Commission funded meso-scale movement project (Hutchison, pers. comm., Fig. 16), and is considered to be extremely rare in this location (one record only, 83 millimetres TL). P. grandiceps has been caught in a variety of lentic and lotic habitats including rivers, streams, floodplain billabongs and wetlands up to 520 metres a.s.l. in the Murray- Darling Basin (Harris & Gehrke 1996). It is common to low gradient, moderate depth 25

30 pools (10 to 60 centimetres) and runs (less than 0.1 m.sec -1 ) in south-east Queensland (Pusey et al. 2004) in areas with medium to course substrate and aquatic vegetation, leaf litter, undercut banks or root masses. In the Murray-Darling Basin it is known to frequent areas with fine substrate, lying motionless on the substrate, but capable of rapid bursts of movement over short distances to avoid capture, or pursue prey (Cadwallader & Backhouse 1983). It is tolerant of a wide range of water quality (11 to 31 C, 2.6 to 12 mg.l -1 D.O., 0.6 to 460 NTU) (Pusey et al. 2004). Figure 16: Recorded distribution of Philypnodon grandiceps in the Border Rivers region of the Murray-Darling Basin, based on the ten datasets in table one. The maximum age and age at maturity of this small fish (less than 12 centimetres) are unknown, but it breeds in freshwater mainly from mid-spring to summer, although larvae are found from spring to autumn (Humphries et al. 2002). Spawning may be cued by rising temperatures (higher than 18 C). Adults engage in elaborate courtship. The adhesive eggs are deposited on a hard substrate such as rock or wood debris and guarded by the male until hatching four to six days after fertilisation. Larval habitat is unknown, but juveniles are common to runs and pools of regulated section of the Campaspe River, north-west Victoria (Humphries et al. 2002), feeding on microcrustaceans. Adult P. grandiceps are considered to be microphagic carnivores, feeding primarily on aquatic insects, molluscs, and other macro-invertebrates. Obligative dispersal migration has been recorded for this species by several authors working in east coast streams (summarised by Pusey et al. 2004) and its use of larval drift as a dispersal mechanism has led Humphries et al. (2002) to classify it as a facultative potamodrous migratory species. Although not listed as threatened, P. grandiceps is extremely rare in the Border Rivers region of the Murray-Darling Basin, and has undergone reductions in distribution in most other areas within the Murray-Darling Basin (Pusey et al. 2004). Its survival is threatened by barriers to movement on key stages of its lifecycle, loss of habitat to agriculture in lowland areas, and predation by alien/translocated species (Pusey et al. 2004). 26

31 Retropinna semoni Common name: Australian Smelt Family: Retropinnidae Discovered by: Weber, 1895 Retropinna semoni are a relatively widespread freshwater species in eastern and southern Australia. It is found in coastal flowing freshwaters from central Queensland to south-eastern South Australia and throughout the Murray-Darling Basin and Lake Eyre Basin (Pusey et al. 2004). It is found in a wide range of macro habitats in both lowland and upland zones and into the headwaters of streams in southern Victoria, New South Wales and Queensland. It is widely distributed (Fig. 17) and relatively common in the Macintyre and Dumaresq Rivers (Table 4). It is a species with a preference for moving water microhabitats (0.2 to 1 m.sec -1 ) such as riffles/runs, often observed in the shallow (less than 60 centimetres) slack-water eddies adjacent to high energy discharge points (Harris & Gehrke 1996), but can also be found in abundance in lakes. However, it is also found in inland lakes and impoundments such as dams and weir pools. It is commonly found over intermediate to course-sized substrates such as sand and gravel, often in association with submerged aquatic macrophytes and filamentous algae. Being a pelagic schooling species, it is commonly found in the upper portion of the water column (Hansen 1989), in open water during high discharge and in close association with coarse substrates, aquatic vegetation and leaf litter during low discharge or no flow. R. semoni is tolerant of poor water quality, but intolerant to capture/handling (Moffatt & Voller 2002). Figure 17: Recorded distribution of Retropina semoni in the Border Rivers region of the Murray-Darling Basin, based on the ten datasets in table one. It has been captured in waters with a wide range of temperatures (8 to 32 0 C), low levels of oxygen (0.6 to 16 mg.l -1 ), and high levels of turbidity (up to 680 NTU) (Harris & Gehrke 1996). R. semoni has been known to live for over four years, but is largely an annual species, attaining sexual maturity in less than one year. Spawning can occur for over nine months of the year in the Murray-Darling Basin, but peaks in winter and early spring (Milton & Arthington 1985) during low and stable discharge periods when larvae and juveniles have a greater chance of encountering high densities of small prey. Spawning can extend into the summer months during periods of higher and variable discharge (Humphries et al. 1999). Adhesive eggs are broadcast by the spawning shoals of fish and adhere to aquatic vegetation and gravel substrate. R. semoni is 27

32 known to undertake short upstream migrations (Moffatt & Voller 2002), usually prior to reproduction. Individuals have been recorded both descending and ascending barrages on coastal Queensland streams (Stuart & Berghuis 1999). Generally there is thought to be some facultative potamodry as a dispersal mechanism for juveniles and sub-adults (McDowall 1996). Tandanus tandanus Common name/s: Eel-tailed catfish, Jewfish, freshwater catfish Family: Plotosidae Discovered by: Mitchell, 1838 This species enjoys a wide distribution along the east coast of Queensland from the wet tropics of Cape Tribulation south to the Manning River of central-northern New South Wales (Pusey et al. 2004). While once common throughout the Murray-Darling Basin, it has undergone a decline in many southern parts of the Basin (Moffatt & Voller 2002). Although this species is found in a wide range of habitats on the east coast, in the Murray-Darling Basin it is thought to prefer turbid waterholes of the lowland zones. It is widely distributed throughout the Macintyre and Dumaresq Rivers (Fig. 18) and is common in the Macintyre and abundant in the Dumaresq (Table 4). It is commonly caught near large woody debris during the daytime in pools with abundant riparian vegetation, moving about the waterhole at night to feed. Along the east coast, it is commonly found in waterholes with still or a slow moving current (less than 0.2 m.sec -1 ), with sand or gravel substrates (necessary for its breeding cycle) (Pusey et al. 2004). Adults are usually solitary, although juveniles are known to aggregate into schools and frequent shallower waters than the adults. Although some authors have reported a close association of this species with aquatic vegetation in the Murray-Darling Basin, recent reports conclude that the significance of this association is unclear and warrants further investigation (Pusey et al. 2004). T. tandanus is tolerant of a wide range of temperatures (8.3 to C) and laboratory experiments have shown that juveniles will survive short exposures to temperatures down to 4 0 C. This is a hardy species capable of surviving in low oxygen levels (0.3 to 16.1 mg.l -1 ) (Clunie & Koehn 2001), and highly turbid waters of the Murray-Darling Basin (0.2 to 910 NTU). Figure 18: Recorded distribution of Tandanus tandanus in the Border Rivers region of the Murray-Darling Basin, based on the ten datasets in table one. T. tandanus will live up to 15 years, attaining sexual maturity within three to five years (Davis 1996a). The breeding season extends from spring to late summer with a 28

33 peak between January and March. Adults pair and build a nest in coarse sediments (Davis 1996b). It is thought that rising temperatures are a primary trigger for spawning behaviour, although rising water levels may also be a factor (Davis 1996c). The males will guard the nest until the eggs hatch. It has been suggested that the time of nest building and spawning coincides with stable conditions of low flows, but that post hatching juveniles take advantage of floodplain inundation associated with late summer flooding to access additional food resources (Davis 1996d). Adults are thought to be generally sedentary, although there are reports of T. tandanus ascending fishways in the winter months, and it has been suggested that these fish were migrating back upstream after flood downstream displacement (Stuart & Berghuis 1999). Juveniles are probably more mobile than adults, being engaged in dispersal/recolonisation movements (Reynolds 1983). However, the overall degree of site fidelity shown by this species indicates their susceptibility to local populations to anthropogenic disturbances. T. tandanus is primarily an opportunistic carnivore, consuming aquatic invertebrates, fish, microcrustaceans and terrestrial invertebrates in varying concentrations, depending on availability (Davis 1996c). T. tandanus is listed as vulnerable in Victoria in the lower Murray-Darling Basin, but not in the Queensland portion of the Murray-Darling Basin (Morris et al. 2001). Survival of the species is threatened by a range of anthropogenic interactions including introduction of alien fish that disrupt nests and prey on the eggs and larvae, impoundments that flood suitable nesting habitat, draw-downs of impounded waters that strand nests, irregular artificial flow regimes that disturb the breeding cycle, riparian habitat destruction that increases siltation and decreases aquatic vegetation, and de-snagging to increase stream flow (and decrease flood duration). Pusey et al. (2004) have linked their basin wide decline to the development of water infrastructure, while alien species introductions, agricultural runoff and riparian destruction are linked to localised population declines. Clunie and Koehn (2001a - not sighted) have prepared a recovery plan for this species. Alien species Carassius auratus Common name: Goldfish Family: Cyprinidae Carassius auratus were first introduced into Australia in the 1860s as an ornamental fish (Brumley 1996). They are now widespread in the Murray-Darling Basin. Populations of C. auratus often establish in impoundments, building to substantial numbers before declining after the stocking of predatory species such as Murray Cod (Maccullochella peelii peelii), Golden Perch (Macquaria ambigua), and trout (Salmo trutta and Onchorynchus mykiss), which consume large numbers of goldfish. They are widely distributed in both the Macintyre and Dumaresq Rivers (Fig. 19) and common in both river systems (Table 4). While usually found in warm, slow-flowing lowland rivers or lakes, they are also found in association with freshwater vegetation and slower-flowing areas of upland rivers and streams. They are tolerant of high summer temperatures and low oxygen concentrations (Allen et al. 2002). They spawn during summer with fish maturing at 100 to 150 millimetres in length. Eggs are laid amongst freshwater plants and hatch in about one week. Their diet includes small crustaceans, freshwater insect larvae, plant material and detritus (Brumley 1996). 29

34 Figure 19: Recorded distribution of Crassius auratus in the Border Rivers region of the Murray-Darling Basin, based on the ten datasets in table one. They have been linked to the introduction of 'Goldfish ulcer' disease in Australia, but are otherwise considered a 'benign' introduction, with few or no adverse impacts documented. Cyprinus carpio Common name/s: Carp, European Carp, Common Carp, Koi Carp Family: Cyprinidae Cyprinus carpiowere first introduced into Australia sometime between 1850 and 1870, but remained in two relatively confined locations (Sydney and the Murrumbidgee Irrigation Area) until the early 1960s (Brumley 1996). These two populations were different strains of the one species and showed no sign of spreading. In the early 1960s a third strain, the Boolarra strain, was illegally introduced by a fish farmer in Victoria. A large, but ultimately unsuccessful eradication program was mounted and the Boolarra strain rapidly colonised watercourses throughout Australia. A recent genetic study of C. carpio in Australia has identified a fourth strain (Koi) in the ACT and Tasmania (Davis et al. 1998). C. carpio are present in the majority of Murray-Darling Basin slopes and lowland rivers and creeks, and some upland streams as well. They are widely distributed in the Macintyre and Dumaresq River systems (Fig. 20) but are more common in the Macintyre than the Dumaresq River (Table 4). They are usually associated with warm, slow-flowing lowland rivers or lakes and are tolerant of a wide range of environmental conditions such as extremely low levels of dissolved oxygen (Koehn et al. 2000). 30

35 Figure 20: Recorded distribution of Cyprinus carpio in the Border Rivers region of the Murray-Darling Basin, based on the ten datasets in table one. Males are sexually mature at two to three years (300 millimetres) and females at three to four years (350 millimetres), and spawning usually occurs in spring and summer when water temperatures are between 17 and 25 C. Spawning fish congregate in shallow water to lay adhesive eggs in clumps on freshwater vegetation, logs and submerged grass. The small eggs (0.5 millimetre diameter) hatch in two to six days, depending on water temperature (Brumley 1996). C. carpio feed by 'mumbling' in the sediment on the bottom or banks of water bodies. This involves sucking in sediment, sorting the edible items from the inedible sediment, and expelling the sediment through the gill openings. Dietary items include zooplankton, freshwater insect larvae, crustaceans, molluscs and, to a lesser extent, plant material (Brumley 1996). C. carpio are vectors of the parasitic copepod Lernaea sp., which infests a range of native and alien fish species. The impacts of C. carpio are not clear but their feeding behaviour has led to considerable concern that C. carpio may be increasing turbidity levels in waterways, and undermining river banks. Their high abundance in many streams and lakes indicates they are probably competing with native fish for food and space. Gambusia holbrooki Common name/s: Eastern gambusia, Gambusia, Mosquitofish, Top Minnow, Plague Minnow Family: Poeciliidae Native to rivers draining to the Gulf of Mexico, G. holbrooki were introduced into Australia in the 1920s for mosquito control. Further introductions were made by health authorities in the 1930s and the species was distributed to many military camps during World War Two. The species is now widely distributed throughout Australia, in farm dams, slow-flowing waters and shallow wetlands (Allen et al. 2002). They are widely distributed throughout the Macintyre and Dumaresq Rivers (Fig. 21), but are far more abundant in the Dumaresq River than the Macintyre (Table 4). G. holbrooki have an affinity for peripheral margins of still or slow flowing waters, and in amongst freshwater plants (McDowell 1996a). Their tolerance of a wide range of water temperatures, oxygen levels, salinities and turbidity, coupled with their ability to breed rapidly, has enabled them to densely populate many habitats, in the absence of predators (Lloyd et al. 1986). 31

36 Figure 21: Recorded distribution of Gambusia holbrooki in the Border Rivers region of the Murray-Darling Basin, based on the ten datasets in table one. G. holbrooki mature at about 25 millimetres long, with the fertilised eggs developing inside the female and live-born young being a few millimetres long at birth. Maturity can be reached after only two months and individuals can breed several times a year (Pen & Potter 1991). They are primarily carnivorous with the diet containing a range of small freshwater invertebrates and wind-blown terrestrial insects (Pen & Potter 1991). While often referred to as Mosquitofish, mosquito larvae are only a minor component of their diet. They are an aggressive species and will chase and fin-nibble fish much larger than them. They also prey on the eggs of native fish and amphibians. G. holbrooki are implicated in the decline of at least nine species of native fish in Australia. Perca fluviatilis Common name/s: Redfin Perch, Redfin, English Perch, European Perch Family: Percidae Native to the cool-temperate waters of the Northern Hemisphere, Perca fluviatilis were first introduced to Tasmania in 1862 and to Victoria in This species is widely distributed throughout the temperate portion of the Murray-Darling Basin, but absent from the Queensland portion of the Basin. They have been found in head waters of the Dumaresq (Beardy River) and the Macintyre Rivers (NSW Severn River), usually higher than 700 metres a.s.l., but have also been recorded from upper foothill waters as low as 240 metres a.s.l (Fig. 22) during colder months of the year. They are quite rare in these headwaters (Table 4) and only two records (data set two only) exist from waters inside the proposed demonstration reach. Their distribution is largely explained by their intolerance of temperatures higher than 31ºC (McDowell 1996). Adult P. fluviatilis are usually found in slow-flowing or still water habitats, especially where freshwater plants are abundant. P. fluviatilis generally mature after two to three years, but males may mature at the end of the first year. Spawning occurs in spring when water temperature reaches 12 C, with thousands of two to three millimetre diameter eggs laid as gelatinous ribbons amongst freshwater plants. Eggs hatch in one to two weeks and juveniles will 32

37 often form large schools (McDowell 1996b). Their diet includes crustaceans, zooplankton and small native fish like eastern gambusia. P. fluviatilis are a primary host for the Epizootic Haematopoietic Necrosis Virus (Langdon 1989). This virus, unique to Australia, also affects Macquarie Perch, Silver Perch and Mountain Galaxias (Galaxias olidus). P. fluviatilis are also a voracious predator, consuming small native species such as carp gudgeons, juvenile Murray Cod and juvenile Golden Perch. Figure 22: Recorded distribution of Perca fluviatilis in the Border Rivers region of the Murray-Darling Basin, based on the ten datasets in table one. 33

38 5.4 Information gaps in the biology/ecology of each species There has been a large investment in research into the biology and life history of many native fish, documenting the habitat requirements of several key species, such as Murray cod and golden perch (Anon. 2004). This has led to further investment to restore appropriate habitat, such as large woody debris (Crook & Robertson 1999, Crook et al. 2001, Nicol et al. 2004, Bond & Lake 2005), reduce impediments to passage through the construction of fishways on key barriers (Close & Aland 2001, Gehrke et al. 2002, Stuart & Berghuis 2002, Baumgartner 2003, Koehn et al. 2003, Koster 2003, Mallen-Cooper 2004, Baumgartner 2006), re-establish connectivity between river and floodplain environments (Koehn & Nicol 1998, King et al. 2003) and restore elements of the natural flow regime that trigger key life history stages, such as migration and spawning (Koehn et al. 2003). However, there are still key knowledge gaps in the specific biology and habitat requirements of many small native fish species (Table 5). Some of the areas where knowledge is present for individual species rely on inferred information gained for research into the same species, but in locations outside the Basin. The utility of such information needs to be carefully assessed as it may not be appropriate to infer that behaviour of populations occurring in coastal tropical streams will be the same as populations occurring in semi-arid Murray-Darling Basin waters. The knowledge base for alien species is also highly patchy, depending on the species in question. It is important to understand the ecology of alien species so that we can gain an insight of their overall impact on native fish species in the Border Rivers region. From table five, it is apparent that there are at least 11 of the 15 species that require further information. They are the olive perchlet (Ambassis agassizii), the Darling River hardyhead (Craterocephalus amniculus), the fly-specked hardyhead (Craterocephalus stercusmuscarum fulvus), the River Blackfish (Gadpsis marmoratus), the mountain galaxias (Galaxias olidus), the carp gudgeon complex (Hypseleotirs spp.), the Murray-Darling rainbowfish (Melanotaenia fluviatilis), the purple-spotted gudgeon ( Mogurnda adspersa), the bony herring (Nemotalosa erebi), the flathead gudgeon (Phylipnodon grandiceps), and the Australian smelt (retropinna semoni). Key information gaps fall into two main categories: the need for migration; and knowledge of their spawning, or nursery habitat. For many species, the habitat requirements are unknown (Table 6). Of those that we do have knowledge about, some require aquatic vegetation, or cobbles/gravel, or large woody debris for spawning habitat. Several species are known to require access to backwaters, both instream or on the floodplain, or amongst macrophyte beds for their larval habitat. Some are known to require slow-flowing waters either amongst aquatic vegetation, or off-stream in floodplain channels for juvenile habitat. A second important information gap is the longitudinal linkages between the egg/larval/juvenile/adult habitats (Table 7), and how individuals move between them to complete their life-cycle. Some of this information is being addressed by a Murray- Darling Basin Commission funded project investigating meso-scale movement. However, as this project is based in the lower foothills zone of the Macintyre River, it is anticipated that some additional information will still be required from other hydrographical zones, and from high flow (flood) years 34

39 Table 5: Summary of knowledge and gaps for native freshwater fish species present in the Border Rivers region of the Murray-Darling Basin. Species Common Name Good Distribution Adult Habitat Physiological Tolerances Migration Reproductive cues Spawning habitat Ambassis agassizii Agassiz perchlet ( ) ( ) ( ) ( ) ( ) Bidyanus bidyanus Silver perch Craterocephalus Darling River hardyhead (r) amniculus Craterocephalus Un-specked hardyhead (p) ( ) ( ) ( ) stercusmuscarum fulvus Gadopsis marmoratus River blackfish (no) Galaxias olidus Mountain glaxias (no) Hypseleotris spp Carp gudgeons ( ) Leiopotherapon Spangled perch ( ) unicolor Maccullochella peelii Cod Macquaria ambigua Golden perch Murray-Darling Melanotaenia fluviatilis rainbowfish Mogurnda adspersa Purple-spotted gudgeon ( ) Nemotalosa erebi Bony bream Philypnodon Flathead gudgeon ( ) ( ) ( ) grandiceps Retropinna semoni Australian smelt - Tandanus tandanus Eel-tailed catfish - Table symbols are: indicates knowledge is recorded indicates that knowledge is not recorded ( ) indicates that knowledge is known from studies outside the Murray-Darling Basin (no) indicates that knowledge is recorded, but the event does not happen (p) indicates patchy distribution (r) indicates restricted distribution (Treadwell and Hardwick 2003) Larval habitat Juvenile habitat Diet QMDC Macintyre River Fish fauna characterisation QDPI&F 35

40 Table 6: Habitat requirements for native freshwater fish species in the Border Rivers region of the Murray-Darling basin, recorded from the 10 data sets outlined in table one. Species Spawning Habitat Larval Habitat Juvenile Habitat Adult Habitat Ambassis agassizii Bidyanus bidyanus Craterocephalus amniculus Craterocephalus stercusmuscarum fulvus Gadopsis marmoratus Galaxias olidus Hypseleotris spp Leiopotherapon unicolor Maccullochella peelii Adhesive eggs laid amongst fine leaved aquatic vegetation Semi-buoyant, pelagic eggs released into water column near surface in flooded backwaters of low gradient streams Surface waters of tributaries Benthic habitats in backwaters & floodplain areas rich in zooplankton Amongst aquatic vegetation in main channel Floodplain & channel waters Not yet recorded Not yet recorded Not yet recorded Adhesive eggs laid over rocks & crevices/weedy areas Adhesive eggs laid in hollow snags or spaces between boulders Adhesive eggs laid amongst cobbles in riffle areas Adhesive eggs laid in shallow flooded backwaters amongst macrophytes & woody debris Demersal eggs dispersed over shallow substrates & vegetation in channel & across floodplain Adhesive eggs laid on hard substrates & inside hollow logs Not yet recorded Not yet recorded Silt/detritus substrate Not yet recorded Data deficient - habitat likely to be similar to adults, from loose shoals in pools Not yet recorded Not yet recorded Pelagic, drifting in benthic habitats in main channel; not recorded from floodplain habitats Not yet recorded - habitat likely to be similar to adults, from loose shoals in pools Shallow ponded areas & inchannel habitats during low flows Juveniles found in shallow inchannel waters, similar to adult habitat Juveniles found in shallow inchannel waters associated with cover Shoals in slow/still water with fine sediment substrate, close macrophytes & woody debris Slow/standing pools with nearby cover from littoral macrophytes & woody debris Slow-flowing, clear waters of small creeks & streams, in shallows or at surface, frequently among aquatic vegetation Shoals in moderate depth pools/runs with fine sediment, usually in shallow, vegetated areas Slow flowing pools with abundant cover from woody debris & undercut banks and coble open water at night High mountain streams with fringing and overhanging vegetation, gravel & boulders Forms shoals in turbid, slow flowing shallows over substrates including bedrock, sand, gravel & mud with cover from boulders & aquatic vegetation Found in rivers, billabongs, lakes, isolated dams, bores & wells, also recorded from ephemeral pools & across inundated floodplains over substrates including bedrock, sand, gravel & mud with cover from boulders & aquatic vegetation In deep holes with habitats that provide high levels of cover from woody debris, rocks, undercut banks or overhanging vegetation 36 QMDC Macintyre River Fish fauna characterisation QDPI&F

41 Macquaria ambigua Melanotaenia fluviatilis Mogurnda adspersa Nemotalosa erebi Philypnodon grandiceps Retropinna semoni Tandanus tandanus Pelagic eggs released in water column near surface in-channel or on floodplain Adhesive eggs laid amongst macrophytes Adhesive eggs laid over hard substrates, rocks, woody debris, broad-leafed aquatic vegetation Semi-buoyant eggs dispersed in shallow sandy substrate backwaters during floods Adhesive eggs laid over hard substrates rocks, woody debris, broad-leafed aquatic vegetation Adhesive eggs scattered over macrophytes or gravel substrate Eggs laid in circular nest in shallow sand & gravel substrates Larvae are rarely observed Shoal in shaded backwaters & still littoral zones, near surface Juveniles associated with slow flowing backwaters, deep pools, anabranches & inundated floodplain habitats, amongst woody debris Not yet recorded, but probably similar to adults Not yet recorded Not yet recorded Shoal in backwaters of floodplain habitats, creeks & weir pools, associated with phytoplankton blooms Slow flowing backwaters, pools & littoral zones Backwaters & still littoral habitats in main channels & billabongs Forms loose shoals over gravel & mud substrate with cover from macrophytes & detritus Not yet recorded, but probably similar probably to adults Runs & run/pool interface Not yet recorded, but probably similar to adults Not yet recorded Warm, turbid slow flowing rivers, floodplain lakes and anabranches often associated with deep pools, woody debris & overhead cover Shoals near surface of still or slow flowing waters in backwaters of large rivers, streams & billabongs, with submerged vegetation, woody debris & overhead cover Amongst fringing vegetation & hard benthic substrates in still or very slow flowing areas in clear streams & off-channel habitats such as large floodplain lakes, billabongs & small terminal lakes Forms shoals over muddy & vegetated littoral zones in standing & flowing waters in main channel & floodplain areas near large woody debris Slow flowing lowland reaches in shallow littoral habitats with abundant submerged cover from overhanging banks, woody debris or vegetation Shoals in large numbers in eddies of highdischarge points amongst coarse substrates close to macrophytes Solitary species in benthic habitats in lakes and sluggish turbid streams in deeper holes near leaf litter and woody debris with coarse substrate and fringing vegetation QMDC Macintyre River Fish fauna characterisation QDPI&F 37

42 Table 7: Migration requirements for native freshwater fish species in the Border Rivers region of the Murray-Darling basin, recorded from the 10 data sets outlined in table one. Species Spawning Migration Larval Migration Juvenile Migration Ambassis agassizii Data deficient Data deficient Some evidence of off stream movement during increased flows Bidyanus bidyanus Long upstream migration Larvae swept downstream juveniles move upstream & downstream during increased flows Craterocephalus amniculus Data deficient Data deficient Data deficient Craterocephalus stercusmuscarum fulvus Gadopsis marmoratus No Galaxias olidus Hypseleotris spp Data deficient - known to aggregate below barriers during flows Possibly undertake a compensatory short upstream migration Data deficient - known to aggregate below barriers during flows Data deficient Data deficient Large scale movement unlikely due to parental care Large scale movement unlikely due to parental care downstream displacement of larvae Downstream dispersal of juveniles Facultative dispersal movement in drift Data deficient Leiopotherapon unicolor Upstream migration Data deficient Data deficient Maccullochella peelii peelii Macquaria ambigua Melanotaenia fluviatilis Facultative upstream migration of some individuals Facultative long distance upstream migration Data deficient - known to aggregate below barriers during flows Obligate larval drift downstream during spring & early summer Eggs & larvae swept downstream juveniles may move upstream during increased flows Data deficient Data deficient Mogurnda adspersa No Data deficient Data deficient Nemotalosa erebi Data deficient Larvae move around in flood waters Philypnodon grandiceps Data deficient Retropinna semoni Data deficient Data deficient Tandanus tandanus No Data deficient Facultative dispersal movement in drift Non-spawning, within pool movement recorded Data deficient Possible upstream dispersal movement in spring - summer during increased flows Possible upstream dispersal movement in spring - summer during increased flows 38 QMDC Macintyre River Fish fauna characterisation QDPI&F

43 5.5 The river continuum concept: a holistic approach to understanding fish in rivers To understand a river, we need to visualise the physical variables within a river system as a continuous gradient of physical conditions between the headwaters and the mouth (Vannote et al. 1980). Within any river, the biological element moves towards a balance between a tendency for efficient use of energy inputs through resource partitioning (food, substrate, etc.) and an opposing tendency for a uniform rate of energy processing throughout the year, and downstream communities are fashioned to capitalise on upstream processing inefficiencies. Separate to, but contiguous with the river, are the floodplains. These are distinct because they do not depend on upstream processing inefficiencies of organic matter, although their nutrient pool is influenced by periodic lateral exchange of water and sediments with the main channel (Junk et al. 1989). Thus, the position of a floodplain within the river network is not a primary determinant of the processes that occur within the stream, but the potential of the floodplain to add to the energy system of the river is enormous. Any fish species that can access the floodplain energy sources may derive significant advantage over instream restricted fish species. This gives rise to the flood-pulse concept (Junk et al. 1989) where fish productivity is strongly related to the extent of accessible floodplains (Arthington et al. 2005), whereas the main river is used as a migration route by most of the fishes. In Queensland, flood events are usually associated with summer and autumn rain, at a time when temperatures are highest and drivers for productivity are maximised. From the biological perspective, the river can also be viewed over a variety of scales ranging from catchment (several rivers) to the individual river, to a reach within the river, and down to the micro-habitat occupied by a particular fish species during a particular phase of its life-cycle. Fausch et al. (2002) present a persuasive argument for a holistic view of the river to understand how processes interacting across scales set the context for stream fishes and their habitats. Ecological studies that focus on fish in one environment only, may fail to identify the causal links to a declining population. The recruitment bottleneck is a good example when habitat for adult fish may be perfectly adequate, but access to juvenile habitat or spawning habitat may have been lost or reduced in another part of the riverscape, leading to reduced recruitment. Thus, a study focusing only on environments where the adults occur could completely miss this crucial causal linkage. This idea is highlighted in Schlosser s (1991) dynamic landscape model for stream fish ecology. This model comprises a holistic spatial arrangement of spawning, feeding, rearing and refugia habitats and the necessity for movement among them for fish to complete their life history. Thus, an awareness of fish biology and ecology, the need to move between habitats and reaches, and knowledge of why these movements occur and their relative importance is a fundamental step in understanding how to sustain or enhance fish populations in a region. Within the Murray-Darling Basin, some native species are able to complete their entire life history without moving beyond a small home-range, e.g. river blackfish (Gadopsis marmoratus), (Khan et al., 2004), two-spined blackfish (Gadopsis bispinosus), (Lintermans, 1998) and mountain galaxias (Galaxias olidus) (Berra, 1963). However, most species need to move during their life history, (whether QMDC Macintyre River Fish fauna characterisation 39

44 between different instream habitats, the river and floodplain habitats, or long distance movements), and impediments to movement and habitat loss can have a major impact. Thus, it is important to know the habitat requirements for each stage of a fish s life history, the temporal and spatial requirements governing their need for movement, and the key threats to their life-cycle. Such information can lead to informed management decisions for developing reach restoration strategies. These have been summarised by Treadwell and Hardwick (2003) and are presented below in Table 8. Table 8: Key threats to each native fish species recorded from the Border Rivers region of the Murray-Darling Basin (Treadwell & Hardwick 2003). Species Key threats Loss of habitat (macrophyte beds), altered flow regimes and Ambassis agassizii possibly increased salinity and water temperatures. Flow regulation, barriers to movement, cold water pollution, and Bidyanus bidyanus unknown impacts of stocking on the gene pool of remnant populations Craterocephalus amniculus Craterocephalus stercusmuscarum fulvus Gadopsis marmoratus Galaxias olidus Hypseleotris spp Leiopotherapon unicolor Maccullochella peelii Macquaria ambigua Melanotaenia fluviatilis Mogurnda adspersa Nemotalosa erebi Philypnodon grandiceps Retropinna semoni Tandanus tandanus Reduced distribution due to habitat degradation Loss of habitat and floodplain connectivity, barriers, thermal pollution, increasing salinity of deflation basin lakes and predation and competition for habitat from introduced fish species such as carp, redfin perch and eastern gambusia Habitat degradation, particularly de-snagging and sedimentation that has reduced available cover and filled the interstitial spaces in the cobble substrate, and introduced fish species such as trout Predation by trout and sedimentation of spawning sites Loss of aquatic vegetation as habitat and sites for egg attachment and parasite infestation Barriers to migration, reduced flow in rivers, loss of macrophyte beds and cold water pollution River regulation, barriers to migration and loss of critical habitat are the major threats to Murray Cod. De-snagging has significantly reduced the optimum habitat for cod in the Basin River regulation, thermal pollution, flow alteration, barriers to upstream spawning areas River regulation, altered flows and competition from mosquito fish Competition and predation by introduced species, increasing salinity and loss of habitat No major threats, although barriers to movement, impacts to floodplain habitats, and cold water pollution may pose risks Loss of habitat and limited information on the biology Barriers to movement and predation from exotic species may pose risks Competition with Carp, loss of suitable breeding habitat, barriers to movement and thermal pollution The range of threats to the survival of native freshwater fish in Australia has been documented by several authors (Kearney et al. 1999, Morris et al. 2001, Pusey et al. 2004, Barrett 2004b), and, from table eight, can be summarised into seven categories: flow regulation causing loss of connectivity and access to critical habitat (both instream and floodplain/anabranch habitat), and reversal/loss of seasonality of flows; infrastructure dams, weirs and crossings causing barriers to movement; loss of habitat de-snagging, infilling from siltation, loss of shading and root mass; 40 QMDC Macintyre River Fish fauna characterisation

45 thermal pollution cold water from dam releases and hot water from riparian vegetation clearing; increased salinity caused by excessive land clearing, or raised water tables from ground water sourced irrigation; introduced fish either alien or translocated species can cause excessive competition or predation; lack of knowledge leading to ill-informed management decisions. These can be addressed by rehabilitation/mitigation activities and are best aggregated in a demonstration reach. 5.6 The demonstration reach concept and the ideal demonstration reach length The key concepts that define a demonstration reach are that it needs to promote aquatic habitat rehabilitation, promote community interest and participation, have well defined objectives at it s start, and be viewed as a long-term commitment by all stakeholders (Barrett 2004b). A successful demonstration reach has three key factors that underlie its success. Firstly, it needs to comprise a suite of management actions that collectively enhances the fish community. Secondly, when management actions are strategically located between good riverine habitat, they can lead to improved riverine health on a much larger scale. Thus, it is very important to ensure that any demonstration reach is of sufficient spatial and temporal scale to demonstrate a quantifiable result. Thirdly, demonstration reach projects must be aimed at community engagement and stewardship by integrating intervention actions with community and stakeholder groups, and focusing funding agencies towards coinvestment for a better return on financial inputs. To be successful, any demonstration reach program will have to be extensive in duration (maybe 10 years or more), although some interventions such as riparian rehabilitation may show a positive response after only one or two years. This is because most benefits to fish populations may not be demonstrable in the short term, but rather over a generational time frame of five to 50 years. This is not unexpected, given the time frame of European occupation that has accompanied the decline in native fish populations in the Murray-Darling Basin. To maintain this continuity, demonstration reach programs need to be well organised and coordinated to ensure that effective on-ground work is achieved and communication is targeted to multiple interest groups. This will require substantial support by the local stakeholder groups, as well as relevant regional NRM, State, and Federal agencies, both in principle and financially over the extended time frame of the demonstration reach project. The time frame for any demonstration reach must be adequate to engender the public attitudinal shift necessary to foster responsible stewardship of adjacent lands, to make the necessary investment to achieve large spatial scale improvement of instream habitat, and to allow the river system adequate time to respond to the intervention. Ideally, a favourable response by a river system would be reflected in an ecologically balanced increase in fish populations. Demonstration reach rehabilitation actions cover a range of activities and a range of social acceptance. These may include (in no particular order): riparian fencing to keep stock out, with or without limited crash grazing; off stream watering for stock; QMDC Macintyre River Fish fauna characterisation 41

46 woody weed control; bank stabilisation to control erosion hotspots; instream habitat complexity rehabilitation including rock/cobble and large woody debris introduction; alien fish removal/control; fish passage enhancement such as fishway construction; fish migration barrier removal; restriction or removal of non-seasonal stream flow; removal or restriction of seasonal flow harvesting; negotiated allocation of impoundment water for environmental flow; limited stocking of threatened species to re-establish viable native populations; scientific research into establishing key restrictions to viable native fish populations. It is important to understand that no single intervention provides a panacea to declining fish stocks. Rather, a range of actions will be necessary to achieve the desired outcome. It is equally important to understand that each action has a cost, socially, economically and environmentally. The willingness of user groups to accept these costs varies according to the impact on particular groups. The idealised demonstration reach length is very much dependant on the suite of species present, both current and historically. Understandably, for smaller fish it will be shorter, but for larger fish it may be much longer. To understand the individual species requirements, we need to examine what habitat and movements are necessary within the life-cycle. Fish move for a range of reasons such as breeding, dispersal/ recolonisation, moving to nursery habitats, foraging and parasite removal. Within the Border Rivers region of the Murray-Darling Basin, there are at least 16 species of native fish (Moffat & Voller 2002), some of which have some aspects of their movement requirements well documented (e.g.: Golden perch, Murray cod), but there are still large knowledge gaps for the majority of native species, particularly small species (Treadwell & Hardwick 2003). Fishway monitoring has provided some information on upstream movements of several, mainly larger, native species (Berghuis & Boradfoot, 2005 unpublished report) but there is little information available on downstream (O Connor et al. 2003, 2004, 2005) or lateral movement requirements, and limited information on small species or juvenile life stages (Treadwell & Hardwick 2003). The Queensland Murray-Darling Committee has proposed that the Macintyre River demonstration reach should be in the Border Rivers region somewhere between the lower foothills of the Macintyre River, near Toobeah, 69 kilometres west of Goondiwindi, upstream into the Severn River (NSW) below Pindari reservoir, and into the Dumaresq River, terminating at the junction of the Dumaresq River and Pike Creek. This is a region in excess of 250 kilometres in river length. The selection of a demonstration reach within this region is dependant on various social, economic and environmental factors, but from a fish ecology and social point of view this region has several beneficial aspects. It is contained within a single geographical region (western slopes of the New England Tableland) and would provide 42 QMDC Macintyre River Fish fauna characterisation

47 very good public access to key hotspots along the demonstration reach length. Some of these hotspots are reviewed in section 5.6. There are three species of fish, present in the proposed demonstration reach that undertake large-scale migration during their life history. These are the silver perch (Bidyanus bidyanus), the Murray cod ( Maccullochella peelii, peelii), and the golden perch (Macquaria ambigua). Silver perch are known to undertake a long pre-spawning migration in summer, to areas behind the peak of a flood (Reynolds 1983), provided that there are no barriers to that migration. Mature Murray cod will migrate upstream some 80 to 100 kilometres into a small anabranch, forming pairs that spawn adhesive eggs in hollow logs, on snags, under rocks or firm clay (Moffatt & Voller 2002). The male will remain and guard the nest until hatching. Adults will often return downstream to their territory. Golden perch have been recorded to migrate over large distances in excess of 1000 kilometres (Reynolds 1983). However, a recent study in the Murray River indicated that only a small proportion of fish undertake such extensive migrations, usually associated with spring spawning behaviour, and many Golden perch return to a home range (O Connor et al. 2005). There is insufficient data for most of the other species present in the Border Rivers region. As most of these are smaller fish than the three highly mobile species described above, it is highly unlikely that any would require extensive longitudinal distances to complete their life-cycle. Rather, it is the presence or absence of critical habitat over a short spatial scale that would be a key influence to their population success. It is quite conceivable that most of the species present in the Border Rivers region could be accommodated within a shorter demonstration reach project of some 30 to 50 kilometres in length, provided that adequate habitat exists or can be rehabilitated. However, there is still a concern about adequate offstream access. Many smaller species require access to either instream or offstream backwaters for some of their life-cycle. Both the habitat and access have diminished with the advent of flow regulation and strategies addressing this issue must be considered if we are to adequately accommodate the full suite of native freshwater fish present in the Border Rivers region of the Murray-Darling Basin. 5.7 Review of sites on the proposed demonstration reach During late March, the author of this report, the Native Fish Strategy Coordinator for Queensland, and a representative from QMDC travelled through the Macintyre Brook, Dumaresq and Macintyre River valleys between Bonshaw, northern New South Wales, and Toobeah on the New South Wales-Queensland border. The following observations were made during this ground-truthing site visit, supplemented by comments from a DPI&F report on Fish Barriers in the Queensland section of the Murray-Darling Basin (Piltz unpublished). Dumaresq River The Dumaresq River is a tributary of the Macintyre River. Beginning in the uplands near Sundown National Park, the river flows southwest where it defines the states border, and joins the Macintyre River near the town of Yelarbon. Major tributaries of the Dumaresq River include the Mole and Severn (QLD) Rivers and Pike Creek. This system is considered by DPI&F fishway officers to be of importance to fish migration because of the low number of significant barriers throughout a larger portion of the system. QMDC Macintyre River Fish fauna characterisation 43

48 However, a major installation, Glenlyon dam (254,310ML), on Pike Creek was built to deliver water on demand down the Severn and Macintyre Rivers for the irrigation industry. Water from this dam is delivered via an off-take located at the bottom near the dam wall and can be delivered at up to 1000ML per day for irrigation purposes. There is some evidence that releases up to 350ML per day may cause cold water plumes for some 30 to 40 kilometres downstream (Glenn Wilson pers. comm.). Water delivered at 1000ML will impart a cold water plume much further downstream and may have significant impact on native fish, especially during critical periods such as spring (rising water temperature) when such cold water plumes may negatively influence spawning success. This water could also interfere with natural water levels in critical habitat such as egg laying sites and backwater nursery areas during critical periods in the life history of native fish. The Bonshaw weir is located 126 kilometres upstream of the Dumaresq-Macintyre Rivers junction, just outside Bonshaw (NSW). It is a stepped construction built upon a rocky base (Fig. 23). Rocky areas protrude from the riverbed at the base of the barrier, dividing the immediate downstream pools. Approximately 10 to 40 metres below the weir is a smaller timber barrier, which is in disrepair, which diverts water in several directions. A culvert road crossing has been installed across the river 50 metres downstream from the base of the barrier. Riffles and deeper pools are continually linked further downstream (Fig. 24) with good instream habitat including, rocky/gravel beds, fallen timber and overhanging trees. Some silting was evident directly above the weir, but this was not excessive. The riparian vegetation above the weir is only a thin strip, but is in a satisfactory state. The habitat includes fallen trees, overhanging vegetation and steep banks throughout the reach (Fig. 25). This region had good grass cover and no significant erosion was noted, indicating good bank stabilisation. Figure 23: The stepped Bonshaw Weir (left), with rocky base and timber barrier (right), in disrepair, diverting the water in several directions. 44 QMDC Macintyre River Fish fauna characterisation

49 Figure 24: Riffle and deeper pools downstream of Bonshaw Weir (photo Piltz) Figure 25: Thin strip of riparian vegetation lines the upstream reaches, with instream woody debris. Carp were observed in the downstream pool of the Bonshaw weir, and the site would benefit from programs to remove alien fish. The riparian vegetation on both sides of the weir pool is in good condition, but very thin and may not be an adequate source for future instream large woody debris (Fig. 25). It would benefit from some riparian rehabilitation to improve the density of this buffer. The south bank is setup as a camping reserve and would provide good public access. Immediately downstream of the weir, the banks are poorly vegetated and would benefit from some rehabilitation. Depending upon further investigation into the river s hydrology, a rock ramp fishway appears to be the most appropriate infrastructure to facilitate fish passage over this weir. This could be designed to circumvent the first downstream pool, avoiding backwater habitat more suited to invasive pest species such as carp. Myall Creek joins the Dumaresq River three kilometres downstream of Bonshaw. This area is a stock reserve with creek access for watering. Myall Creek is mostly dry QMDC Macintyre River Fish fauna characterisation 45

50 with several semi-permanent watering holes. It has sections of reasonable riparian vegetation, but does exhibit areas of severe bank erosion from cattle access to watering points. These have led to instream sediment slug development. It would benefit from establishing off-stream watering points for transient stock and riparian rehabilitation. Between Bonshaw (NSW) and Yelarbon (QLD) is a typical rural property adjacent to the Dumaresq River. Like many properties that abut the river, the occupants harvest water for stock and irrigation. The steep southern bank is mostly well vegetated, but has been cleared almost to the change in slope (Fig. 26). The northern bank is much flatter and vegetation has been thinned to facilitate grass development for stock feed. Where the riparian buffer zone has been cleared too close to the bank, signs of erosion exist, and there is evidence that the landholder has identified this as a potential problem and attempted to address this by dumping chopped vegetation into the void (Fig. 27). This site would be of particular concern to the landholder as it is adjacent to his pump shed. Carp were observed feeding on the substrate in the long waterhole downstream of the pump. This site, like many others along the river banks, would benefit from an intensive riparian rehabilitation program to control erosion and stabilise the bank, and some carp control. However, unlike several other sites, this site has good public access, and provided access can be negotiated, would prove a good demonstration reach action site. Figure 26: Typical river banks of the Dumaresq, steep on one side and gentle on the other. Note openness of northern bank, and gravel substrate. 46 QMDC Macintyre River Fish fauna characterisation

51 Figure 27: Grazier instigated erosion control (chopped branches, tin and wire) on steep southern bank of Dumaresq River. Downstream of the Bonshaw weir is the Cunningham Weir, approximately 68 kilometres upstream of the junction between the Macintyre and Dumaresq Rivers. It is a 4.6 metre barrier, constructed from a variety of materials ranging from timber, rocks and metal sheeting (Fig. 28). As a weir, this structure is of questionable integrity as there is water passing through the structure in many locations. However, it is still a significant barrier to fish migration. The weir pool is long (more than two kilometres) and utilised by landholders for stock watering and irrigation. The banks are well vegetated and appear to be stable except in several points where stock access water. At these locations the banks are eroding and sediment slugs have filled in the stream bed. Below the barrier is a large pool, about two to three metres deep, which shallows a further 60 metres downstream. This pool contained large schools of adult bony bream, rainbow fish, carp gudgeons and harydheads, as well as carp and eastern gambusia. Below the pool is a road crossing constructed from culverts (Fig. 29); the water deepens beyond this into adjoining pools. The riparian vegetation adjacent to the culvert is in moderate to poor condition, but the instream cobble appears to have prevented significant erosion and stream bed infill. QMDC Macintyre River Fish fauna characterisation 47

52 Figure 28: Cunningham Weir leaking through the wall in several locations, and the downstream pool. Figure 29: Downstream road crossing and culvert at Cunningham Weir (photo Piltz) Located at a major road reserve between Yetman (NSW) and Yelarbon (QLD), the Cunningham Weir has good public access. The weir is a substantial structure (4.6 metres high) and of questionable integrity. If the owners of the structure were to upgrade this facility, it would present a good opportunity to build a vertical slot fishway to facilitate fish passage. Both the weir and downstream pools would benefit from introducing habitat complexity via large woody debris. The downstream banks are highly degraded, but appear stable. They would benefit from riparian rehabilitation as this would supply future demand for stream shading and habitat complexity. A further 11 kilometres downstream of Cunningham Weir is the Glenarbon Weir (Fig not sighted). It has a non-functional fishway installed (Fig not sighted), which is a pool pass structure (Piltz pers. comm.). At the base of the barrier on the 48 QMDC Macintyre River Fish fauna characterisation

53 right hand side, outlet works are running over a rocky area as shown. The pool downstream does not appear to be disjointed and contains good quality habitat such as rock cobble areas throughout the riverbed. The riparian vegetation both up-stream and downstream, which includes canopy cover and good under-story preventing erosion, was in fair condition. Above the weir the instream habitat, including fallen trees, aquatic and overhanging vegetation (Fig not sighted), was plentiful. No silting was obvious directly above or below the barrier. Figure 30: Glenarbon Weir with a fishway (photo Piltz). Outlet Figure 31: Down stream pool at Glenarbon Weir showing the fishway outlet (marked) and the quality of habitat (photo Piltz). Further fish survey work and hydrology detail would be required to ascertain if fish passage is necessary at the Glenarbon Weir, but DPI&F fishway officers felt that a rock-ramp fishway would be best suited for this weir, in place of the existing but nonfunctional vertical-slot fishway, as structure height is not great. Additional works QMDC Macintyre River Fish fauna characterisation 49

54 would involve riparian rehabilitation and probably alien fish removal (carp buster program). Figure 32: Instream habitat (fallen trees) upstream of Glenarbon Weir (photo Piltz). Dumaresq River Summary The Dumaresq River locations surveyed all possessed essential riparian and instream habitat, with variation in structure types. All these sites were typical of an upper foothills zone river, with nearly continuous water flow, good water clarity, and a stream bed largely of cobble and gravel. It is one of the few areas in the Queensland portion of the Murray-Darling Basin where dense aggregations of aquatic macrophytes occur. However, this is a regulated river with the 253,000ML Glenlyon Dam located on a major tributary, Pike Creek. Water releases from this dam are thought to contribute cold water pollution for many tens of kilometres downstream and interfere with flow heights during critical stages of native fish life history. Several other tributaries such as the Mole and Severn (QLD) Rivers, and Tenterfield Creek are unregulated and are a source of natural flow. If the three minor barriers were to have fish passage infrastructure installed, it would open up 240 kilometres of water that fish could navigate, taking the open water up to Nundubbemere Falls. Carp have been observed near all three weirs and any project to install fish passage on either of these weirs would also require a concurrent alien fish removal programme. Various forms of disturbance were noted at all sites along the Dumaresq River, including cattle trampling of banks, removal of riparian vegetation, and erosion gullies leading to some instream sediment slugs. However, the over-all river health appears to be in good condition with modification of important habitats remaining at minimal levels. This is also an attractive river for public access with dedicated camping reserves established at many points with good all weather access. Macintyre Brook The Macintyre Brook is a tributary of the Dumaresq River, branching approximately five kilometres south of Yelarbon, and running in a north-easterly direction east of Inglewood. It contains four major barriers along the 80 kilometre stretch of river below the 75,000ML Coolmunda Dam. This dam is recognised as an eastern border to 50 QMDC Macintyre River Fish fauna characterisation

55 the region where a demonstration reach could be established. There are two main reasons for this. Firstly, at 16 metres in height, it would be difficult to facilitate fish passage past Coolmunda Dam wall. Secondly, because there are no reports of alien fish above this barrier, it may be best if no fish passage above the dam wall was built. Inglewood Weir (not sighted refer to Piltz report) is approximately 25 kilometres downstream of the Coolmunda Dam wall. This is a concrete structure (Fig. 33) developed to supply water to the township of Inglewood. On one side, the weir pool banks are stable, but open to provide an amenity to an adjacent golf course. On the other side bank disturbance is medium due to the presence of agricultural practices on adjacent land, although riparian vegetation is in fair condition with little to no erosion being present on the steep banks. There is little silting in the weir pool, except closer to the bank, which is covered with reeds and other macrophytes. Instream habitat is notable and varying, including aquatic plants, fallen timber and overhanging vegetation. Figure 33: Inglewood Weir (photo Piltz). Downstream of the weir is a 10 metre long pool that has been filled in by silting, and then a culvert road crossing. Immediately below this road crossing, a riffle zone has been created due to a bottleneck resulting from the culvert. This area has good riparian vegetation providing shading, and complex instream habitat. Riffle zones, pools at various depths and aquatic plants are all present, as well as other diverse habitats, although some silting is evident on the southern side. QMDC Macintyre River Fish fauna characterisation 51

56 Figure 34: Riparian and instream habitat downstream of Inglewood weir (photo Piltz). The Inglewood Weir has good public access and would benefit from a riparian rehabilitation program. This would ensure the future supply of large woody debris for instream habitat complexity. Fish passage could be achieved by either a rock-ramp or a vertical slot fishway. This would depend on hydrology information, the close proximity of the road-crossing downstream, and the silting at the base of the structure. Whetstone Weir (not sighted), a further 16 kilometres downstream of Inglewood Weir, is a five metre structure of timber pylons and rails, and has been packed with gravel and rock (Fig.35). The sides have been secured by concrete pads. The weir pool has good riparian vegetation and instream habitat (Fig. 36), although there is some bank vegetation disturbance. Bank stability is in good condition with no erosion or silting being noted upstream of the weir. At the base of the weir structure is a large 40 metre pool of variable depth up to 2.5 metre, and below this pool there is a concrete road crossing (Fig. 37), with five square channels under the road. This enables suitable downstream flow with the possibility of allowing adequate passage for fish, although light penetration appeared poor. 52 QMDC Macintyre River Fish fauna characterisation

57 Figure 35: Whetstone Weir on Macintyre Brook (photo Piltz). Figure 36: Whetstone Weir pool on Macintyre Brook (photo Piltz). Figure 37: Road crossing below Whetstone Weir (photo Piltz). QMDC Macintyre River Fish fauna characterisation 53

58 Numerous riffle zones are present downstream of the road crossing (Fig. 38), along with a quantity of deeper pools at various depths. The instream habitat and riparian vegetation is in good condition, including large woody debris, rocks, gravel beds and aquatic vegetation. Figure 38: Downstream of Whetstone Weir road crossing with good stream habitat (photo Piltz). This region of Macintyre Brook represents some very good habitat. Water turbidity is not quite as low as similar regions on the Dumaresq River, but riparian zones and instream habitat complexity is good. Fish passage is a problem and considerable structure refurbishment may be required as much of the Whetstone Weir main structure is in fair to poor condition. A vertical slot fishway would be the most appropriate structure for this weir as the barrier is five metres high (Piltz pers. comm.). A third weir, Ben Dor, occurs some 19 kilometres upstream of the Macintyre Brook- Dumaresq River junction. At 5.6 metres in height, it is a significant barrier to fish movement. This site has not been surveyed. Some 10 kilometres upstream of the Macintyre Brook-Dumaresq River junction is the 1.2 metre high Sunny Girl Weir (not sighted). This is a privately owned structure built of timber posts and rails (Fig. 39). Water is constantly being released via gaps throughout the length of the weir. Tall grasses, trees and debris are across the length of the barrier. Upstream riparian vegetation is in very good condition with no visible sign of silting or erosion (Fig. 40). The riparian vegetation is notable with ample overhanging trees and bank cover, while instream aquatic vegetation and timber habitat are diverse and complex. This reach of river varies between 30 to 40 metres wide with significant depth throughout. 54 QMDC Macintyre River Fish fauna characterisation

59 Figure 39: Sunny Girl Weir downstream of woody debris crossing the water body (photo Piltz). Figure 40: Sunny Girl Weir pool (photo Piltz). Downstream of the barrier is a long pool (about 20 metres by five metres) approximately 1.5 metre deep in places and terminating at a road crossing (Fig. 41). This crossing has good flow and light penetration under the bridge. The riparian vegetation is in excellent condition with no silting or erosion. Instream structures include rocky areas, over-hanging vegetation and fallen timber. QMDC Macintyre River Fish fauna characterisation 55

60 Figure 41: View of Sunny Girl Weir from road crossing downstream (photo Piltz). This area provides a good example of integrating adjacent agricultural practices with good stream stewardship. The barrier itself is in disrepair and significant refurbishment would be required if fish passage infrastructure, such as a rock ramp fishway, was to be built on the barrier. Macintyre Brook Summary Macintyre Brook provides a good potential reference zone for the demonstration reach project. Although smaller in size than either the Dumaresq or Macintyre Rivers, it is generally in less disturbed condition. It is within a similar hydrological zone to the Dumaresq and Macintyre Rivers (upper foothills) and has only four barriers, all within 44 kilometres of each other. Fish habitat throughout this stretch of river shows little evidence of human influence (other than weir installations and some recreational fishing activities) compared to the other catchments. The habitat along this system is both diverse and complex and the brook provides significant potential for natural fish populations. Although Macintyre Brook is not of great length, the appropriate fish habitat is substantial, enhancing native fish species chances of recruitment for the whole catchment. One drawback is the downstream extent of cold water pollution from Coolmunda Dam. Macintyre River The Macintyre River runs from Mt Rumbee, south-west of Glenn Innes, west through Inverell, and north-west through Wallangra before being joined by the Severn River (NSW) near Wallangra, and then the Dumaresq River some 25 kilometres east of Goondiwindi. Pindari Dam is located on the Severn River (NSW) between Ashford and Emmaville. Although it is a deep dam of 322,000ML, and capable of delivering up to 3000ML per day, it has multiple off-takes at various depths and thus cold water pollution is not such a significant issue with this dam as in other rivers, except during blue-green algal blooms, when bottom water is mixed with surface water to dilute the bloom. However, flow regime changes can occur when out-of-season irrigation flows 56 QMDC Macintyre River Fish fauna characterisation

61 are released that do not coincide with natural flows. These can impact on native fish access to critical habitats. Much of the Macintyre east of Wallangra is in the uplands zone (more than 600 metres a.s.l.). Below Wallangra it conforms to an upper foothills category (Moffat and Voller 2002), and then to lower foothills west of Yetman to the western border of the proposed demonstration reach at Toobeah (65 kilometres west of Goondiwindi). In the uplands zone, it is typical of many rivers in the region, having nearly continuous cool, clear water flow, with abundant over-hanging vegetation and diverse and complex instream habitats, such as large woody debris, rocks, cobbles, deep pools, riffles and runs, and deep pools. Aquatic vegetation is not common. There are numerous areas between Wallangra and Yetman where the Macintyre River conforms to the typical upper foothills habitat, having pools and riffles similar to the uplands, but a greater predominance of cobble and gravel substrate. Aquatic macrophytes are common and riparian vegetation is both dense and diverse. However, there are also large tracts of river that pass through agricultural land. In some of these areas the riparian vegetation has been cleared for grazing and agricultural purposes. In these areas, the banks exhibit severe signs of erosion from stock access to watering points. Instream habitat is simplified due to infill from siltation and water turbidity is high. The area suffers from infestations of large woody weeds, such as willow, privet, blackberry and osage orange. Further downstream into the lower foothills zone of the river between Goondiwindi and Yetman there are two major barriers and a common theme in land use. This zone has shallow and deep pools intermittently connected during flow events. The banks are predominantly clay and mud, while the bottom is usually covered in silt or sand with the occasional rock bar. The water is highly turbid (Fig. 42). Riparian buffers do exist, and can be more than 50 metres thick (Fig 43). Adjacent land use is almost entirely agricultural and banks show periodic signs of erosion and impact from stock access (Fig. 44). Instream habitat is often compromised by siltation, although in many areas there is an abundance of large woody debris. The two large barriers, Goondiwindi and Boggabilla Weirs, form significant barriers to fish migration. While fishways have been built on both, recent surveys question the efficacy of each to fish passage (Berghuis pers. comm.). Recent surveys below the Goondiwindi Weir have also found aggregations of rainbow fish, bony bream, hardyheads and gudgeons, as well as carp, golden and spangled perch. It is thought that smaller native fish have great difficulty negotiating this barrier. QMDC Macintyre River Fish fauna characterisation 57

62 Figure 42: Highly turbid waters of Macintyre River. Figure 43: Riparian vegetation on the Macintyre River ranges from over 50 metres thick (left) to narrow, or absent (right). 58 QMDC Macintyre River Fish fauna characterisation

63 Figure 44: Signs of erosion and stock access on the bank of the Macintyre River. Macintyre River Summary This river has several attributes that lend it to a demonstration reach. The river traverses a range of hydrological zones from the uplands to the lower foothill regions. Within these zones there are areas with good riparian habitat and adjacent highly disturbed areas. The river has a well documented suite of native fish fauna, and is very accessible by the public in key areas. There are some fish migration barriers in need of remediation and these are located near major population centres. There is evidence of local public interest in rehabilitation in the area, such as the Yetman Fishing Club s submission to round five of the NSW Recreational Fishing Trust for a fish rehabilitation program. This application is for work in the Macintyre River catchment near Yetman. The Yetman Fishing Club, in collaboration with the Border Rivers-Gwydir Catchment Management Authority plan to revegetate two kilometres of riparian habitat, manage woody weed infestations, protect 200 metres of bank from soil erosion, develop various recreational fishing infrastructure and signage, hold a carp buster program and build a boat ramp. There are several factors that detract from the Macintyre River as a demonstration reach site, such as the two major barriers to fish migration and non-seasonal irrigation flows. This river would benefit from programs to remove alien fish. QMDC Macintyre River Fish fauna characterisation 59