Movement Ecology and Conservation Implications for ...

Movement Ecology and Conservation Implications for Riverine Fishes of South-eastern Australia

by Wayne M. Koster BSc

Submitted in fulfilment of the requirements for the degree of Doctor of Philosophy

Deakin University April 2015

Movement ecology and conservation implications for riverine fishes

ii


DEAKIN UNIVERSITY ACCESS TO THESIS ‐ A

I am the author of the thesis entitled Movement Ecology and Conservation Implications for Riverine Fishes of South-eastern Australia submitted for the degree of Doctor of Philosophy This thesis may be made available for consultation, loan and limited copying in accordance with the Copyright Act 1968.

'I certify that I am the student named below and that the information provided in the form is correct' Full Name: Wayne Michael Koster

Signed:

Date: 22/04/2015

iii


iv


DEAKIN UNIVERSITY CANDIDATE DECLARATION I certify the following about the thesis entitled Movement Ecology and Conservation Implications for Riverine Fishes of Southeastern Australia submitted for the degree of Doctor Philosophy a.

I am the creator of all or part of the whole work(s) (including content and layout) and that where reference is made to the work of others, due acknowledgment is given.

b.

The work(s) are not in any way a violation or infringement of any copyright, trademark, patent, or other rights whatsoever of any person.

c.

That if the work(s) have been commissioned, sponsored or supported by any organisation, I have fulfilled all of the obligations required by such contract or agreement.

I also certify that any material in the thesis which has been accepted for a degree or diploma by any university or institution is identified in the text.

'I certify that I am the student named below and that the information provided in the form is correct'

Full Name: Wayne Michael Koster Signed: Date: 22/04/2015

v


vi


Preface The data in this thesis was collected while I was employed by the Victorian Department of Environment and Primary Industries at the Arthur Rylah Institute for Environmental Research. Each of the peer-reviewed journal publication details, together with the names of my co-authors, is listed on the title page of each chapter. The details for each publication and links to the definitive versions of the paper on the journal websites are listed below. For each of the chapters, various photographs have been added. These are the only figures that have been added to the published papers. Koster, W.M. & Crook, D.A. 2008. Diurnal and nocturnal movements of river blackfish (Gadopsis marmoratus) in a south-eastern Australian upland stream. Ecology of Freshwater Fish 17: 146–154. http://onlinelibrary.wiley.com/doi/10.1111/j.1600-0633.2007.00269.x/full Koster, W.M., Dawson, D.R., Morrongiello, J.R. & Crook, D.A. 2014. Spawning season movements of Macquarie perch (Macquaria australasica) in the Yarra River, Victoria. Australian Journal of Zoology 61: 386–394. http://dx.doi.org/10.1071/ZO13054 Koster, W.M., Dawson, D.R. & Crook, D.A. 2013. Downstream spawning migration by the amphidromous Australian grayling (Prototroctes maraena) in a coastal river in south-eastern Australia. Marine and Freshwater Research 64: 31–41. http://dx.doi.org/10.1071/MF12196 Koster, W.M., Dawson, D.R., O’Mahony, D.J., Moloney, P.D. & Crook, D.A. 2014. Timing, frequency and environmental conditions associated with mainstem–tributary movement by a lowland river fish, golden perch (Macquaria ambigua). PLOS ONE 9: e96044. http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0096044#p one-0096044-g005 vii


viii


Acknowledgements To my supervisors, thanks for useful advice throughout my candidature and helpful comments on earlier versions of the manuscript: Professor Gerry Quinn, Drs Ty Matthews and Paul Jones, Deakin University, and Dr David Crook, Charles Darwin University. I am incredibly indebted to David Crook for his encouragement, guidance, enthusiasm and assistance over many years. Numerous staff at the Arthur Rylah Institute for Environmental Research assisted me with field and laboratory work. Specific acknowledgement is made at the end of each chapter. David Dawson deserves a special mention for his efforts during the many long days of fieldwork. Dr John Koehn is gratefully acknowledged for his support and advice over the last few years. Thanks to Dr Ivor Stuart for useful discussions about the movement ecology of fish. John Koehn and Ivor Stuart are also thanked for

constructive

comments on earlier versions of the manuscript. Melbourne Water, the Goulburn Broken Catchment Management Authority and the Department of Environment and Primary Industries funded this study. To my Dad, our many fishing trips to places like the Woori Yallock Creek (catching blackfish, eels and trout) helped spark my interest in fish, for which I am grateful. To my Mum, I would have loved for you to be around to see this, but I’m sure you would be proud. Jeanette Birtles (Organic Editing) is thanked for editing.

ix


x


Table of Contents Access to thesis ....................................................................................................... iii Candidate declaration .............................................................................................. v Preface.....................................................................................................................vii Acknowledgements ................................................................................................. ix Abstract ...................................................................................................................xv 1 Introduction ......................................................................................................... 1 1.1 Ecological importance of movement behaviour .................................................... 1 1.2 Movement of freshwater fishes .......................................................................... 4 1.3 Movement behaviours of Australian fish ............................................................ 14 1.4 Impacts of human activity on fish movement .................................................... 16 1.5 Structure and scope of thesis ........................................................................... 18 1.6 Thesis objectives ............................................................................................ 20 2 Diurnal and nocturnal movements of river blackfish in Armstrong Creek ........ 23 2.1 Introduction ................................................................................................... 26 2.2 Materials and methods .................................................................................... 27 2.3 Results .......................................................................................................... 31 2.4 Discussion ...................................................................................................... 39 3 Spawning season movements of Macquarie perch in the Yarra River .............. 43 3.1 Introduction ................................................................................................... 46 3.2 Materials and methods .................................................................................... 48 3.3 Results .......................................................................................................... 53 3.4 Discussion ...................................................................................................... 61 4 Downstream spawning migration by the amphidromous Australian grayling in the Bunyip River ................................................................................................ 65 4.1 Introduction ................................................................................................... 68 4.2 Materials and methods .................................................................................... 69 4.3 Results .......................................................................................................... 78 4.4 Discussion ...................................................................................................... 85 5 Timing, frequency and environmental conditions associated with mainstem– tributary movement by a lowland river fish, golden perch ............................... 93 5.1 Introduction ................................................................................................... 96 5.2 Materials and methods .................................................................................... 97 5.3 Results ........................................................................................................ 104 5.4 Discussion .................................................................................................... 113 6 General discussion and conclusions ................................................................ 119 6.1 Overview ..................................................................................................... 119 6.2 Key findings ................................................................................................. 120 6.3 Common threads and contrasts...................................................................... 128 6.4 Conservation and management implications .................................................... 135 6.5 Conclusions .................................................................................................. 141 7 References ....................................................................................................... 143

xi


List of Figures Figure 1-1. The study species: (a) river blackfish, (b) Macquarie perch, (c) Australian grayling and (d) golden perch. .................................................................................................... 19 Figure 2-1. Armstrong Creek. ......................................................................................... 28 Figure 2-2. Location of Armstrong Creek in south-eastern Australia. .................................. 29 Figure 2-3. Daily discharge (ML·day⁻ 1) (dotted lines) and distances (m) moved by six radiotagged river blackfish in Armstrong Creek downstream of the West Branch weir over the study period during the morning (○) and afternoon (•). .................................................... 32 Figure 2-4. Daily discharge (ML·day⁻ 1) (–) and distances (m) moved by five radio-tagged river blackfish in Armstrong Creek upstream of the West Branch weir over the study period during the morning (○) and afternoon (•)........................................................................ 33 Figure 2-5. Period of elevated flow in Armstrong Creek associated with use of inundated riparian areas and relatively large movements by river blackfish. ....................................... 34 Figure 2-6. Distances (m) moved and mesohabitats used by six radio-tagged river blackfish in Armstrong Creek below the West Branch weir over the 3-day diel tracking period. .............. 36 Figure 2-7. Example of an undercut bank used by river blackfish in Armstrong Creek during the study. ..................................................................................................................... 38 Figure 3-1. The Yarra River at Wonga Park...................................................................... 48 Figure 3-2. The location of the study site. ....................................................................... 49 Figure 3-3. Representative examples of the movement patterns of Macquarie perch tagged with radio-transmitters in the Yarra River at Wonga Park (a–d), Eltham (e–f) and Heidelberg (g–j). ........................................................................................................................... 56 Figure 3-4. Distances moved (median, 25th and 75th percentiles) by fish at Wonga Park (a), Eltham (b) and Heidelberg (c) between June 2011 and February 2012. ............................. 57 Figure 3-5. Predicted probability of Macquarie perch movement versus flow difference. ...... 59 Figure 3-6. Percentage of total locations of fish in each mesohabitat during each season. ... 60 Figure 4-1. The Bunyip River near Koo Wee Rup. ............................................................. 70 Figure 4-2. Map showing location of the study site. .......................................................... 71 Figure 4-3. Acoustic (left) and radio-transmitters (right) used in tagging of Australian grayling. ....................................................................................................................... 74 Figure 4-4. Driftnets set in the Bunyip River. ................................................................... 76 Figure 4-5. Movement patterns of all Australian grayling tagged with acoustic transmitters that migrated downstream in the Bunyip River in 2009 (a–g) and 2010 (h–m). .................. 80 Figure 4-6. Initiation of downstream movement of Australian grayling tagged with acoustic transmitters in the Bunyip River in 2009 (a) and 2010 (b). ................................................ 81 Figure 4-7. Scatterplot showing the relationship between the daily numbers of Australian grayling tagged with acoustic transmitters moving downstream and daily mean discharge in the Bunyip River in 2009 (a) and 2010 (b) during the downstream migration period. .......... 81 Figure 4-8. Adjusted total density of Australian grayling eggs (grey bar) and larvae (black bar) per 1000 m3 collected in driftnets in the Bunyip River. ............................................... 83 Figure 4-9. Australian grayling egg collected in the Bunyip River. ...................................... 84 Figure 5-1. The lower Goulburn (top) and Murray (bottom) rivers. .................................... 98 Figure 5-2. Map showing location of the study site. .......................................................... 99

xii


Figure 5-3. Times during which tagged fish were detected in mainstem, tributary and junction locations. ................................................................................................................... 106 Figure 5-4. Percentage of mainstem fish detected in tributary and percentage of tributary fish detected in mainstem. ................................................................................................. 108 Figure 5-5. Predicted probability of fish remaining in the Murray River versus mean weekly temperature. .............................................................................................................. 109 Figure 5-6. Predicted probability of fish remaining in the Goulburn River versus percentage change in flow. ........................................................................................................... 109 Figure 5-7. Golden perch larvae hatched from eggs collected during drift sampling. .......... 111 Figure 5-8. Period of flooding in the lower Goulburn River associated with golden perch spawning ................................................................................................................... 112 Figure 6-1.. Summary of movement behaviours of adult river blackfish, based on previous studies (top); summary of movement behaviours inferred from the present study (bottom). ................................................................................................................................. 121 Figure 6-2. Summary of movement behaviours of adult Macquarie perch in impoundments, based on previous studies (top); summary of movement behaviours in rivers inferred from the present study (bottom). ......................................................................................... 123 Figure 6-3. Summary of movement behaviours of adult Australian grayling, based on previous studies (top); summary of movement behaviours inferred from the present study (bottom). ................................................................................................................................. 125 Figure 6-4. Summary of movement behaviours of adult golden perch, based on previous studies (top); summary of movement behaviours inferred from the present study (bottom). ................................................................................................................................. 127 Figure 6-5. Example of how information from the present study has been used to inform environmental flow recommendations. .......................................................................... 136 Figure 6-6. Lower reaches of the Bunyip River, and an example of potential threat (gravel extraction) to spawning grounds. ................................................................................. 140

List of Tables Table 2-1. Total distance moved and total linear range (m) of six river blackfish during the 3day diel tracking period. ................................................................................................ 37 Table 2-2. Percentage use of mesohabitats for six river blackfish during the 3-day diel tracking period.............................................................................................................. 37 Table 3-1. Details of the Macquarie perch tagged during the study. ................................... 51 Table 3-2. Results of model-selection procedure for models comparing the effects of flow difference, mean water temperature, site, fish length and day of year on the probabilities of fish moving. ................................................................................................................. 58 Table 3-3. Model-averaged parameter estimates, standard errors and 95% confidence intervals describing the probability of fish moving, from models with some support (AICc < 4). ............................................................................................................................... 59 Table 4-1. Total number of Australian grayling eggs and larvae collected during 2008, 2009, 2010 and 2011 sampling events at the three collection sites in the Bunyip River. ................ 83 Table 5-1. Details of the tagged golden perch during the study. ...................................... 101

xiii


Table 5-2. Relative importance of predictor variables and parameter estimates for the model averages (models with ΔAICc < 4) for the transition models for movement between the Murray and the Goulburn rivers. ................................................................................... 110 Table 6-1. Summary of new information on movement and proposed revisions to existing flow recommendations. ............................................................................................... 138

xiv


Abstract Riverine fishes are among the most threatened fauna in the world, but many species remain poorly studied or are managed with little understanding of their ecological requirements. With growing concern over the impact of human activities on native fish, various strategies to conserve or restore fish populations have been developed and implemented in recent times. Understanding movement patterns of fish, and how these are influenced by or respond to environmental changes, is critical for the development and effectiveness of these strategies. For many Australian freshwater fish species, there is little or no detailed information available on their movement patterns. In this thesis, I used radio- and acoustic telemetry techniques to investigate the movement and behaviour of four species of endemic riverine fish in south-eastern Australia, all of which have declined in range and/or abundance and are of high management priority. Key objectives were: to contribute to and refine our understanding of the movement ecology of these species; to compare and contrast the movement behaviours of species with a variety of life history strategies to identify similarities (or ‘common threads’) in movement patterns; and to increase our knowledge of the movement behaviours of riverine fishes more generally in order to assist in their conservation. River blackfish Radio-telemetry techniques were used to investigate the day-to-day activity and habitat use of river blackfish (Gadopsis marmoratus) in Armstrong Creek, with emphasis on the fine spatial and temporal scales of movement and the influence of the diel period on movement between habitats. The study revealed previously undocumented aspects of the movements and behaviour of river blackfish. River blackfish were most often located within pools, but they also commonly used riffle and run habitats. They also used inundated riparian areas during a flood and made rapid, relatively long-distance movements coinciding with elevated flows. Integration of such information into management strategies (e.g. flows to maintain adequate depths through riffles to enable localised movements) has the potential for improving our capacity to provide the conditions required to conserve and restore river blackfish populations. This study also demonstrated how different conclusions regarding the xv


extent of movement of a species could be reached, depending on the temporal scale and the timing of observations. In particular, fish moved over significantly larger ranges and between mesohabitats at night, which would not have been detected using daylight tracking data only. This highlights how understanding movement dynamics at a range of spatial and temporal scales is critical for obtaining a complete picture of the movement ecology of riverine fish. Macquarie perch Radio-telemetry was used to investigate the day-to-day activity, habitat use, and spawning season movement behaviours of Macquarie perch (Macquaria australasica) in the Yarra River. The study specifically investigated whether riverine fish exhibit synchronised migrations to specific river reaches during the spawning season, as documented for lacustrine populations of the species. The tracking showed that Macquarie perch typically occupied restricted reaches of the stream. Half of the tagged fish undertook occasional upstream or downstream movements away from their usual locations, and these were particularly associated with increased flows during the spawning season, but there was no evidence of synchronised migratory behaviour or movement of multiple fish to specific locations or habitats during the spawning season. The study demonstrates that management of riverine populations of this species cannot necessarily be based on the spawning behaviour of lacustrine populations and highlights how models of movement behaviour need to consider variation in within-species life history characteristics observed in contrasting habitats. Australian grayling Acoustic telemetry and larval drift sampling were used to investigate the spawningrelated movement behaviours of Australian grayling (Prototroctes maraena) in the Bunyip River, with an emphasis on whether the species undertakes downstream spawning migrations to specific breeding grounds and whether such behaviours are associated with increased river flows. The study revealed significant new information about the movements and spawning of Australian grayling that is directly relevant to the development of conservation strategies for the species. In particular, the study demonstrated the existence of a long-distance downstream spawning migration to the lower river reaches and the importance of increased river flow as a migration cue. These findings highlight the potential impacts of in-stream barriers and altered flow xvi


regimes (resulting from water resource development and predicted climate change) on Australian grayling reproduction. My findings also provide convincing evidence of an unusual and alternate form of amphidromy, as amphidromous fishes generally do not undertake a spawning migration. Re-evaluation of terms to describe forms of diadromy would be useful for adequate characterisation of the variation in life history characteristics displayed. Golden perch Acoustic telemetry was used to investigate the movements of golden perch (Macquaria ambigua) in the Goulburn and Murray rivers. The importance of tributary and mainstem connections as corridors for movement and the influence of increased river flows and reproductive behaviour as drivers of movement were investigated. The study showed that golden perch exhibit a spatially and temporally complex pattern of movement between mainstem and tributary locations, influenced by discharge, temperature and (potentially) reproductive behaviour. These results add to growing evidence that connections between tributaries and mainstem habitats are important as corridors for fish movement and for linking populations across river networks. This demonstrates the need to consider the spatial, behavioural and demographic interdependencies of aquatic fauna across geographic management units such as rivers. Despite the distinct life history characteristics of the four study species, the findings presented in this thesis clearly revealed some common threads in the observed movement patterns. This included both restricted movement for extended periods and larger movements at specific times (including home range shifts, return movements to areas previously occupied, and movement responses to river flows). Indeed, for all four species, aspects of movement were associated with changes in river flow. This result serves to highlight the potential impact of altered flows on movement behaviours of riverine fish species under scenarios of predicted decreases in rainfall and river flow associated with climate change. The results also demonstrate how riverine fish movement behaviours are much more complex and dynamic than simple models of either restricted movement or high mobility. While my study found similarities in movement patterns of the four study species, there were also clear differences between species in the temporal and spatial patterns and functions of

xvii


movement, which demonstrates the difficulty of attempting to develop uniform rules about fish movement, and the need to determine the species-specific details of fish behaviour and life history in order to implement effective conservation and management strategies.

xviii


1 Introduction 1.1 Ecological importance of movement behaviour Since the beginning of recorded history, humans have sought to understand the movement of animals through the environment. For example, more than 2,400 years ago, Aristotle in his De Motu Animalium (On the Movement of Animals) sought explanations for trends in the movements of animals throughout the environment (cited in Nathan et al. 2008). Movement is an important factor affecting the fate and fitness of individuals, the structure and dynamics of populations, communities and ecosystems, and the evolution and diversity of life (Nathan et al. 2008). At the individual level, movement allows animals to access resources for feeding and reproduction, to avoid unfavourable environmental conditions, and to escape threats, such as predation. Movement can also influence the persistence of populations by reducing extinction risk following disturbance through the colonisation of new habitats (Bowler and Benton 2005) or by enabling gene flow among populations (Tallman and Healey 1994). Movement can shape the structure and dynamics of populations and communities, because the degree to which animals move affects the density and distribution of organisms, and thus the rates of encounter and interactions between individuals over resources (Nathan et al. 2008; Morales et al. 2010). Animals that move between habitats and ecosystems can also translocate important resources, such as nutrients; thus, the movements can have major effects on ecosystem functioning (Lundberg and Moberg 2003). Pacific salmon (Oncorhynchus spp.), for example, transport oceanderived nutrients stored in body tissue into freshwater ecosystems as they migrate from the sea to spawn in fresh water. These nutrients support the juvenile salmon after the adults die, and are also incorporated into forest ecosystems when the fish are consumed by predators, such as bears (Gende et al. 2002). For some animals, including many fishes, movement is essential for life cycle events, such as reproduction (Lucas and Baras 2001).

1


The investigation of spatial and temporal patterns of movement is critical for our understanding of animal ecology and behaviour, as well as for devising effective conservation and management strategies (Rubenstein and Hobson 2004). In the case of fish, detailed information on movement patterns is becomingly increasingly recognised as important data useful for predicting the effects of disturbances or environmental changes on riverine systems (Lonzarich et al. 2000). Such disturbances include the construction of dams, which may obstruct movement pathways (Liermann et al. 2012), or the modification of hydrological regimes, which can disrupt movement cues (Murchie et al. 2008). A detailed understanding of movement patterns, such as spatial and temporal variation in habitat preferences or requirements and the factors influencing movement, can also be important for predicting how populations might respond to management interventions, such as the provision of environmental flows or habitat restoration (e.g. re-snagging) (Koster et al. 2014). Additionally, knowledge of patterns of space utilisation is becoming increasingly recognised as important for guiding appropriate spatial design (e.g. size and positioning) of reserves or protected areas (Fogarty and Botsford 2007; Botsford et al. 2009). The investigation and understanding of the movement patterns of fishes forms a considerable body of scientific research and literature. Early research on freshwater fish movements included the seminal mark–recapture studies by Gerking (1950, 1953, 1959) and focused on concepts such as home range (the area over which an animal regularly travels during its day-to-day activities) and territory (a defended area within the home range) for stream fishes. Gerking introduced the notion that many adult fishes in streams are normally sedentary and occupy a small (20–50-m) home range, and his conclusions provided a foundation for present-day research. Recent advances in methodological approaches for tracing animal movements, such as passive integrated transponder (PIT) tags, radio- and acoustic telemetry, satellite tags, otolith chemistry and population genetics, have greatly improved our ability to examine movement patterns. We are now able, for example, to investigate daily movements and fine-scale habitat use (i.e. on a scale of metres) (David and Closs 2003; McEwan and Joy 2014), seasonal or annual trends and large-scale movements (i.e. on a scale of hundreds of kilometres) (Koehn et al. 2009; Heupel et al. 2010), connectivity of populations among landscapes (Humston et al. 2010; Benjamin et al. 2013), and the

2


role or influence of changes in environmental variables (David and Closs 2002; Taylor and Cooke 2012). However, despite a growing awareness of the importance of animal movement for survival and reproduction, our knowledge of movement patterns and behaviours is limited for most freshwater fish species (Koehn and Crook 2013). This lack of knowledge is of particular concern given that human activities have resulted in freshwater habitats and their biota, including fishes, being among the most threatened in the world (Dudgeon et al. 2006; Cooke et al. 2012).

3


1.2 Movement of freshwater fishes Movement terminology Movement behaviour in animals can be divided into two broad categories: ‘movement’ and ‘migration’ (Dingle 1996). ‘Movement’ can be defined as a change in the spatial location of an individual over time (Nathan et al. 2008) and refers to the range of movements associated with home range behaviour and accessing resources (e.g. foraging). ‘Migration’ refers to a specific type of movement that results in an alternation between two or more separate habitats, occurs with a regular periodicity, and involves a large proportion of the population (Northcote 1978). In a range of animals, including various fishes, in which populations consist of both migratory and resident individuals, migrations are referred to as ‘partial migrations’ (Chapman et al. 2012). In the case of freshwater fishes, there are two main types of migration. ‘Potamodromous’ fish species undertake migrations wholly within fresh water, whereas ‘diadromous’ fish species migrate between fresh water and marine environments. Among the diadromous fishes, several specific modes of migration are recognised, including anadromy, catadromy and amphidromy (Myers 1949; McDowall 1988). ‘Anadromous’ fishes, e.g. sea lamprey (Petromyzon marinus), enter rivers from the sea as mature adults and migrate to upstream spawning grounds, with juveniles later migrating downstream to the sea. ‘Catadromous’ fishes, e.g. anguillid eels (Anguilla spp.), enter rivers from the sea as juveniles, and adults return to the sea or estuary to spawn. ‘Amphidromous’ fishes, e.g. southern grayling and some galaxiids, mature and spawn in fresh water and the larvae drift downstream to the sea, with juveniles migrating back into fresh water.

4


Temporal, spatial and functional forms of movement Movements or migrations of fish are typically associated with three main purposes: spawning, feeding, and finding refuge (Lucas and Baras 2001). Spawning migrations of fish can involve shifts between separate habitats, such as from fresh water to marine environments (e.g. catadromous fish) and vice versa (e.g. anadromous fish), and migrations to specific spawning grounds. The best-known examples are the upstream spawning migrations of anadromous salmonids from the sea into rivers in the Northern Hemisphere (Banks 1969) and the downriver seaward migrations of anguillid eels to marine spawning grounds (van Ginneken and Maes 2005). Some fishes, e.g. Atlantic salmon (Salmo salar) and Australian bass (Macquaria novemaculeata), are ‘iteroparous’ (i.e. spawn multiple times during their life)—such species typically undertake annual spawning migrations. In contrast, ‘semelparous’ fish (e.g. anguillid eels) spawn only once and therefore only undergo a single spawning migration during their lives. Feeding movements of fish may take several forms. Many fish undertake localised foraging movements between habitat patches. Brown trout (Salmo trutta), for example, occupy pool habitats at night when resting but move to riffles during the day to feed on drifting invertebrates (Roussel et al. 1999). Longer-distance migrations for feeding also occur in some fishes. Roach (Rutilus rutilus), for instance, have been observed to migrate upstream from Lake Årungen in Norway into a tributary stream for feeding (L’Abée-Lund and Vøllestad 1987), while sharptooth catfish (Clarias gariepinus) and blunt-tooth catfish (Clarias ngamensis) migrate upstream in large shoals in the main river channels of the upper Okavango Delta, Botswana, during the drawdown of the annual flood level (to pack-hunt small fish species moving from the floodplain to the main river channel) (Merron 1993). Many fish also undergo movements to refuges. Giant kokopu (Galaxias argenteus), juvenile steelhead (Oncorhynchus mykiss) and coho salmon (Oncorhynchus kisutch), for example, have been shown to move during high flows or floods to lower-velocity habitats, such as undercut banks, off-channel habitats or tributaries (Bramblett et al. 2002; David and Closs 2002), while river catfish (Pangasius hypophthalmus) move to deeper water to avoid de-watered stream sections during low flows (Hogan et al. 2004). Movements to thermal refuges (such as tributaries) in summer to avoid high

5


water temperatures have also been reported for various fish species, such as striped bass (Morone saxatilis), rainbow trout (Oncorhynchus mykiss) and brown trout (Kaya et al. 1977; Moss 1985). Some movements are obligate steps in the life history, as in migrations by catadromous fishes (such as eels) from freshwater to marine environments for spawning (Lucas and Baras 2001). Because such movements are fundamental to the completion of the life cycle, the disruption of movement pathways (for example, by dams or weirs) can result in local population extinction. Other movements are facultative or opportunistic, although they may be important in maintaining the distribution and persistence of species (Mallen-Cooper 1999b). The dispersal of fish such as bony herring (Nematalosa erebi) and Lake Eyre golden perch (Macquaria sp. B) over large areas from isolated dry-season waterholes to newly inundated habitats, e.g. during the flooding of the Cooper Creek, Queensland, allows them to take advantage of food-rich inundated floodplains, which are important for growth and survival (Balcombe et al. 2005; Arthington and Balcombe 2011). Movements of fish occur on a broad range of temporal scales. At smaller temporal scales, many fish move on a diel (daily) basis, which can include diurnal (day), nocturnal (night) and crepuscular (twilight) components. At larger temporal scales, many fishes exhibit seasonal movement patterns. Seasonal movements to spawning areas are well known. For instance, some salmonids (such as rainbow trout and cutthroat trout (Oncorhynchus clarki)) and cyprinids (such as barbel (Barbus barbus) and nase (Chondrostoma nasus)) migrate upstream in spring to spawn in gravel beds (Bernard and Israelsen 1982; Lucas and Batley 1996; Henderson et al. 2000; Hilderbrand and Kershner 2000; Ovidio and Philippart 2008). Many fish also undergo seasonal movements unrelated to spawning. Seasonal movements to deeper refuge areas in autumn and winter have been reported, e.g. in salmonids (Cunjak 1996; Jakober et al. 1998), common carp (Cyprinus carpio) (Johnsen and Hasler 1977; Penne and Pierce 2008) and northern hog sucker (Hypentelium nigricans) (Matheney and Rabeni 1995). The spatial scale of movement also varies greatly. Some species are relatively sedentary. For instance, two-spined blackfish (Gadopsis bispinosus) are thought to

6


spend their lives within discrete stream reaches, with home ranges of adult fish estimated at ~15 m (Lintermans 1998). In contrast, other riverine fishes can move hundreds or thousands of kilometres at times. It has been estimated, for example, that piramutaba (Brachyplatystoma vaillantii) in the Amazon Basin can migrate ~3,300 km in five months from the estuary to the upper Amazon (Barthem and Goulding 1997), and small cyprinids in the Mekong River Basin are thought to migrate more than 1,000 km between Cambodia, Laos and Thailand (Baird et al. 2003). There is also great variability in the spatial pattern and directionality of movement. Movements in upstream and downstream directions within river systems occur in many fish species, such as from upland tributary streams to lowland channels and floodplain habitats and vice versa (Winemiller and Jepsen 1998). Lateral stream movements, such as between the main river channel and floodplains, are also known for many fishes. In the mid–Murray River system, for example, a range of species (such as Australian smelt (Retropinna semoni), freshwater catfish (Tandanus tandanus), carp gudgeons (Hypseleotris spp.) and common carp) move between floodplain and riverine habitats at various times (Jones and Stuart 2009; Lyon et al. 2010; Conallin et al. 2011; Koster et al. 2014). Many species also undergo vertical and horizontal movements. Juvenile sockeye salmon (Oncorhynchus nerka), for example, occupy the bottom of lakes during the day but move into the water column at night, probably to minimise predation risk and increase foraging gain (Scheuerell and Schindler 2003), and northern redbelly dace (Phoxinus eos) move from the littoral zone during the day to the pelagic zone at sunset to increase feeding opportunities on zooplankton that are more abundant in the pelagic zone (Naud and Magnan 1988).

7


Determinants of fish movement The movements and migrations of fish are influenced by a range of abiotic and biotic factors, which can act alone or in combination (Northcote 1984; Lucas and Baras 2001). Abiotic factors include hydrology, temperature, light, day length, moon phase, habitat and water quality. Biotic factors include body size, sex, ontogeny, predation pressure, density of competitors, and availability of food. Hydrology

Hydrology is a major determinant of movement for many freshwater fish species. Hydrology has been shown, for example, to influence upstream migration, downstream migration, lateral movement and fine-scale movement activity in a range of fishes (Northcote 1984; Albanese et al. 2004; Taylor and Cooke 2012). Change or variation in flow (or water level), rather than a particular flow threshold, is often cited as a trigger for the movement of fish (Ovidio et al. 1998; Baran 2006). The specific mechanisms by which hydrology affects fish movement are often unclear. Aside from eliciting direct behavioural responses, hydrological changes also influence the physical habitat and hydraulic characteristics (e.g. velocity and turbulence) of rivers, which may affect the swimming ability of fish (Mitchell 1989; Silva et al. 2011). For many species, increasing flows are associated with upstream migrations to spawning areas. Upstream migrations of salmonids (such as brown trout and cutthroat trout) into tributaries to spawn, for example, are often linked to increases in flow (Campbell 1977; Bernard and Israelsen 1982). Increased flows are also thought to trigger upstream migration of pouched lamprey (Geotria australis) and sea lamprey to spawning areas (Almeida et al. 2002; Jellyman et al. 2002). Seasonal pulses of juvenile diadromous fishes (such as common galaxias (Galaxias maculatus)) upstream into freshwater habitats have also been linked to high flows (McDowall and Eldon 1980). Downstream migrations of fish to spawn at times of increased flows are also common. Increasing flows have been shown to be a major factor, for instance, in triggering migrations of tupong (Pseudaphritis urvillii) (Crook et al. 2010) and anguillid eels (e.g. Vøllestad et al. 1986; Boubée et al. 2001; Durif and Elie 2008) downstream to the sea to spawn.

8


High flows also provide important opportunities for lateral movements of fish onto floodplains or off-channel habitats (Matheney and Rabeni 1995; Brown et al. 2001). It has been shown during periods of increasing flow, for example, that carp gudgeon and Australian smelt move from the main river channel into newly flooded off-channel habitats, then return to riverine habitats on falling levels (Lyon et al. 2010). Hydrologic events can also inhibit or decrease rates of movement. For instance, high winter flows for hydropower production or flood control caused a delay in the spawning migrations of burbot (Lota lota) in the Kootenai River, Idaho (Paragamian 2000). Periods of low flow can also be important for fish movement (Novoa 1989; Stuart and Mallen-Cooper 1999; Stuart and Berghuis 2002). In the Fitzroy River, Queensland, for example, species diversity and the number of migrating fish were found to be greatest during low flows (Stuart and Mallen-Cooper 1999). Water temperature

Water temperature has been shown to influence the movement of many freshwater fishes. Movement may be triggered when a specific water temperature level or threshold is reached, by a temperature increase or temperature decrease, or by a combination of these (Jonsson and Ruud-Hansen 1985; Swanberg 1997). Many studies have shown that the initiation of spawning migrations is strongly correlated with critical water temperatures (Northcote 1984). For instance, the onset of the spawning migration of brook lamprey (Lampetra planeri) in a Swedish stream was associated with a temperature of 7.5°C (Malmqvist 1980); the spawning migration of bull trout (Salvelinus confluentus) in the Blackfoot River, Montana, was cued by an increase in water temperature to 17.7°C (Swanberg 1997); and in the Roanoke River, North Carolina, the spawning migration of striped bass began when the water temperature reached 17–18°C (Carmichael et al. 1998). Water temperature may also interact with other factors, such as flow or photoperiod, to influence movement (Northcote 1984). Migration of brown trout in streams in the River Meuse Basin, Belgium, for instance, was initiated by a combination of water temperature variation and water level, within a thermal range of 10–12°C (Ovidio et al. 1998), and both water temperature and flow influenced the autumn migration of

9


brown trout in the River Imsa, Norway (Jonsson and Jonsson 2002). The mechanisms by which water temperature influences movement can be physiological or behavioural. Many studies have shown that water temperature directly affects the physiological performance of fishes, such as their swimming speed and activity level (Lyon et al. 2008; Breau et al. 2011). In addition to physiological responses, many fishes exhibit strong behavioural responses to water temperature, such as moving to areas to avoid undesirable or lethal water temperatures (Breau et al. 2011). In winter, cutthroat trout have been shown to move to areas of streams that are kept warmer by groundwater discharge, then to disperse as temperatures increase (Brown and Mackay 1995; Brown 1999). In summer, movement to cooler areas, such as tributaries, to avoid higher water temperatures is also common in fishes (Kaya et al. 1977; Moss 1985). Other water quality characteristics, such as turbidity and dissolved oxygen, have also been shown to influence movement (Kramer 1987; Richardson et al. 2001). As water temperature can co-vary with other factors, such as discharge or turbidity, distinguishing or separating their effects can be difficult (Taylor and Cooke 2012). Light, photoperiod and moon phase

Light–dark cycles, or diel periods, strongly influence the movement patterns of fishes. Indeed, regular shifts between distinct day and night locations are a characteristic of many fishes and are typically associated with feeding opportunities and predator avoidance (Schulz and Berg 1987; Clapp et al. 1990; Roussel and Bardonnet 1997). Bull trout, for example, have been shown to move from deep, mid-channel areas during the day to shallow, low-velocity areas along the channel margins at night (Muhlfeld et al. 2003). Photoperiod or day length can also have a strong influence on the movement of fishes. David and Closs (2003) found that the activity of giant kokopu in small creeks in New Zealand’s South Island decreased with increasing light intensity, and vice versa. Apart from behavioural responses, photoperiod can alter the physiology of fishes, such as enzyme levels and swimming speeds, which in turn can influence movement (Beeman et al. 1990; Kolok 1991; Muir et al. 1994). For example, Muir et al. (1994) found that advanced photoperiod increased gill levels of Na+/K+/ATPase (sodium-potassium adenosine, an enzyme correlated with migration rate) in spring chinook salmon (Oncorhynchus tshawytscha) in the Clear Water River, Idaho, and resulted in faster downstream movement.

10


Moon phase has also been found to be a major factor influencing fish movement. Common galaxias undertake downstream migrations on the full or new moon to spawn in the lower freshwater reaches of river estuaries just upstream of the salt wedge among tidally flooded vegetation (Benzie 1968; Mitchell 1994; Taylor 1996; Richardson and Taylor 2002). Spawning migrations in association with moon phase have also been described in other fishes, such as the characid (Prochilodus platensis) (syn. P. lineatus) in the Pilcomayo River, South America (Bayley 1973), and various tropical fishes in the Grande River, Brazil (Bizzotto et al. 2009). Moon phase has also been shown to influence other movements, such as the seaward movement of juvenile coho salmon in Lynn Creek, British Columbia, which peaks during the full moon (Mason 1975), and the seaward movement of short-finned eels (Anguilla australis) out of Darlots Creek estuary, Victoria, which increases as the moon wanes (Crook et al. 2014). The effect of moon phase on movement is thought to exert its influence on factors such as tide height (in tidally influenced waters), light intensity and predation risk (Lucas and Baras 2001). Habitat

Movement of fish can depend strongly on habitat type, quality and quantity. Movement is typically lower in higher-quality or complex habitats than in poorerquality or homogenised habitats (Gorman 1986). Bjornn (1971) found that movements of trout and salmon in field and laboratory tests were correlated with the amount of cover, with more fish emigrating from streams or troughs without rubble substrate compared with those with rubble. Similarly, red-tailed barbel (Barbus haasi) moved less in reaches with greater depth, slower current and more cover compared with when in shallower, exposed sections in a Mediterranean stream (Aparicio and De Sostoa 1999), and fallfish (Semotilus corporalis) moved less with increased habitat complexity in a Virginian stream (Albanese et al. 2004). Higher-quality or complex habitats are thought to have higher probabilities of meeting the resource needs of individuals, and thus can be associated with reduced movement behaviour (Rasmussen 2010). The spatial arrangement of habitat types (e.g. pool–riffle sequences, etc.) can also influence movement patterns (Lonzarich et al. 2000; Crook et al. 2001). For instance, riffle length was shown to affect movement patterns of fish among stream pools; fish in pools separated by long riffles moved less than fish in pools separated by short riffles (Lonzarich et al. 2000). Water body size can also 11


influence fish movement; the distance moved has been found to increase with increasing size of the water body, and is linked to factors such as individuals taking advantage of increased resource availability (Woolnough et al. 2009; Radinger and Wolter 2013). Biotic factors

Biotic factors can be important determinants of movement. Movement distances or home ranges of fishes, for example, generally increase with body size or fish length (Minns 1995; Radinger and Wolter 2013). Larger flathead catfish (Pylodictis olivaris) in the Missouri River, Missouri, moved greater distances than smaller individuals. Similarly, movements of Murray cod (Maccullochella peelii) in the Ovens River and golden perch (Macquaria ambigua) in the Murray River system were predominantly by larger fish (Reynolds 1983; Koehn et al. 2009). Size-based movement is likely related to factors such as greater resource demand and swimming ability, but can also be correlated with other factors, such as maturity or reproductive status (Woodward et al. 2005; Radinger and Wolter 2013). Predation pressure and competition can also influence movement. For instance, Gilliam and Fraser (2001) showed that in the Guanapo River, Trinidad and Tobago, the killifish (Rivulus hartii) moved more in a reach occupied by predatory fish than in a reach lacking predators, and Hansen and Closs (2005) showed that the level of diurnal activity and home range size for large dominant giant kokopu increased when the food supply was below that of normal feeding conditions. In contrast, a study of the effect of increased competition on the movement of brown trout by increasing population densities found no differences in movement patterns (Heggenes 1988). Movement behaviour can also vary with ontogeny or life history stage, and is often linked to changes in feeding and predation risk. For example, Cussac et al. (1992) found that in an Araucanian lake, Argentina, common galaxias and Patagonian pejerrey (Odontesthes microlepidotus) moved from the littoral zone (where they hatch) to the limnetic zone (where exogenous feeding begins) as free embryos, then moved to the littoral zone as larvae, and finally to the limnetic zone as juveniles. Similarly, a study by Stoffels et al. (2013) suggested that juvenile freshwater catfish may spend the first year of their life in wetland habitats before moving into rivers, then return to wetlands to reproduce.

12


Sexual differentiation in movement behaviour (both male-biased and female-biased) has also been reported in some fishes and is often linked to factors such as mating systems (Croft et al. 2003; Hanson et al. 2008). Male largemouth bass (Micropterus salmoides) move less than females and inhabit shallower areas during the breeding season, and this is thought to reflect differences in their reproductive behaviour, such as nest building (in littoral zones) and parental care by the males (Hutchings and Gerber 2002). Similarly, a recent study found a trend for greater movement during the breeding season of female freshwater catfish at night compared with males, which may also relate to nest building and guarding by the male freshwater catfish (Koster et al. 2014). Other examples of sexual differentiation in movement behaviours include female Australian bass and tupong moving further upstream into freshwater habitats, while males remain mostly in lower tidal waters, which may be linked to females seeking more productive feeding areas to increase reproductive fitness (Harris and Rowland 1996; Crook et al. 2010).

13


1.3 Movement behaviours of Australian fish Australia has a wide diversity of freshwater ecosystems, including upland perennial streams, lowland floodplain rivers, arid zone ephemeral rivers, and permanent, seasonal and intermittent lakes, billabongs and swamps. Australia is the driest inhabited continent and has the most variable rainfall and stream flows in the world, with droughts and floods a defining characteristic of the climate (Puckridge et al. 1998). Such extreme hydrologic events play a major role in determining the unique character of many of Australia’s freshwater ecosystems and the aquatic biota they support. On the Australian continent, ~206 native fish species utilise freshwater environments to complete all or part of their life cycle, which is a relatively low number by world standards (Allen et al. 2002; Lévêque et al. 2008). In contrast, ~4,475 species have been described in South America, with many awaiting formal description (Helfman et al. 2009). About 70% of the native freshwater species are endemic to Australia (Allen et al. 2002), and ~34 introduced fish species have also become established in freshwater environments (Lintermans 2004). Movement is an important component in the life history of many Australian freshwater fishes (Koehn and Crook 2013). In the inland catchments, long-distance potamodromous migrations by adult fish have been documented for several species. For example, Murray cod have been observed travelling upstream up to 130 km in spring before returning to home sites (Koehn et al. 2009), and golden perch move long distances (tens to thousands of kilometres) upstream and downstream in spring (Reynolds 1983; O’Connor et al. 2005). Migrations of juvenile fishes have also been documented, such as the upstream migrations of large numbers of juvenile golden perch, silver perch (Bidyanus bidyanus) and bony herring through fishways along the Murray River (Mallen-Cooper et al. 1995; Stuart et al. 2008; Baumgartner et al. 2014). Some species of potamodromous fish may exhibit highly flexible movement strategies. For instance, golden perch were thought to undertake long-distance upstream migrations to spawn (Reynolds 1983), but recent studies suggest that some golden perch migrate downstream to spawn (O’Connor et al. 2005). In the case of coastal rivers and streams, diadromous movements are characteristic of many fishes. For example, ~70% of the native freshwater fish species in the coastal catchments of south-eastern Australia move between freshwater and marine habitats

14


at some stage during their life cycle (Harris 1984a). Most Australian diadromous fish are catadromous (e.g. tupong and short-finned eel) or amphidromous (e.g. Australian grayling (Prototroctes maraena) and broad-finned galaxias (Galaxias brevipinnis)), but a small number of species are anadromous (e.g. pouched lamprey and shortheaded lamprey (Mordacia mordax)). Populations of some species (e.g. Australian smelt) in coastal catchments may have both diadromous and non-diadromous components (Crook et al. 2008). Recent research has resulted in significant new information on movement patterns of Australian native freshwater fish, enabling the development of more complex movement models for several species, such as golden perch and Murray cod (MallenCooper 1999b; Crook 2004a; O’Connor et al. 2005; Koehn 2009). Nevertheless, for many species there is little or no detailed information available on their movement patterns, which severely limits the development of targeted and effective conservation management strategies (Koehn and Crook 2013). There is clearly a need for greater understanding of the movements of many Australian freshwater fishes, particularly threatened species, in order to develop and refine conceptual models of their movements.

15


1.4 Impacts of human activity on fish movement Freshwater ecosystems worldwide have been greatly modified by human development (e.g. construction of dams and weirs, water regulation, habitat loss and degradation, and introduction of exotic species). These activities are linked to major changes in the ecology of freshwater ecosystems, including reduced abundance, distribution and diversity of native fish populations (Bain et al. 1988; Gehrke et al. 1995; Marchetti and Moyle 2001). Dams and weirs in particular have fragmented habitats and obstructed critical movement pathways of fish, while natural flow regimes, which are an important cue for movement in many freshwater fish, have also been greatly altered through water storage and the diversion of flows (Lucas and Baras 2001). In Australia, where water resources for human use are in great demand because of the dry climate, there has been extensive regulation and modification of freshwater ecosystems, particularly in the south-east of the continent. For example, in the Murray-Darling Basin (MDB), Australia’s largest drainage basin, more than 4,000 licensed in-stream structures have been constructed to regulate river flow (Lintermans 2007). There are also many more unlicensed structures. In some regions of the MDB, migratory fish species, such as golden perch and silver perch, have become locally extinct upstream of major barriers (Harris and Rowland 1996). Flow patterns have also been drastically altered; regulated flows at the Murray River mouth are about one-third of natural flows, and in some years there may be no flows (Maheshwari et al. 1995). Such flow alterations are considered to be a major reason for the estimated 90% decline in native fish abundance in the MDB over the past 50–100 years (MDBC 2004). Many freshwater ecosystems in the coastal drainage systems of south-eastern Australia have also been severely affected by human development. It has been estimated, for example, that impoundments have restricted the availability of stream habitats in these systems by up to about one-half (Harris 1984a, 1984b). In many systems, diadromous species (which comprise the bulk of native fish diversity and abundance) have become extinct upstream of barriers. For example, ten diadromous species, including the nationally threatened Australian grayling, disappeared from the Tallowa River upstream of Tallowa Dam after the dam was built (Gehrke et al. 2002),

16


and Australian bass disappeared from the Warragamba River upstream of the Warragamba Dam after it was built (Harris 1983). With growing concern over the impact of human activities on native freshwater fish, various strategies to conserve or restore fish populations (such as environmental flows, habitat restoration (e.g. re-snagging) and fishways) have been developed and implemented in recent times. Understanding movement patterns of fish, and how these are influenced by or respond to environmental changes, is critical for the development and effectiveness of these strategies. For example, statements on environmental flow requirements for maintaining or restoring important ecological processes (e.g. spawning migrations) have been formulated for various fish species in many Australian rivers, but there is much scientific uncertainty associated with these statements because of the lack of empirical evidence for flow–movement response relationships for the target species (Cottingham et al. 2003; Earth Tech 2006). Such information could greatly inform decision-making regarding the appropriate implementation of environmental flows for fish and also improve predictions of the ecological consequences of flow alterations. Likewise, provision for the passage of native fish past barriers to help restore populations has been undertaken in numerous Australian waterways, but for many freshwater fishes detailed knowledge of movement patterns and behaviours is needed to guide the design and construction of effective fishways (Barrett and Mallen-Cooper 2006; Stuart et al. 2008).

17


1.5 Structure and scope of thesis In this thesis, I have investigated the movement behaviours of four species of threatened freshwater fish in regulated riverine ecosystems in south-eastern Australia. The thesis focuses on key knowledge gaps associated with our understanding of the movement behaviour of the study species and consequences for conservation and management. Thesis results are interpreted and discussed in relation to the general ecology of the study species and the movement ecology of riverine fishes, together with the implications for conservation and management initiatives. All four study species are Australian natives that provide an opportunity to compare movement patterns of species with distinct life histories (Figure 1-1). River blackfish (Gadopsis marmoratus) inhabit upland rivers and are considered non-migratory (Koehn 1986; Khan et al. 2004). Macquarie perch (Macquaria australasica) inhabit upland rivers and impoundments and are considered potamodromous (Wharton 1968; Cadwallader and Rogan 1977). Australian grayling inhabit coastal rivers and are amphidromous (Berra 1987; Crook et al. 2006). Golden perch inhabit lowland rivers and lakes and are considered potamodromous (Reynolds 1983; O’Connor et al. 2005). The study species were chosen on the basis that (i) they are of high management importance, (ii) they have declined in range and/or abundance since European settlement, (iii) substantial gaps exist in our knowledge of their movement behaviours, and (iv) they allow comparisons and contrasts between species with distinct life histories. This is a thesis by publication. Chapters 2–5 present primary data that have been published in peer-reviewed scientific journals. The reference for the relevant paper and the names of the co-authors appear at the beginning of each chapter. I have chosen a thematic chapter arrangement that logically describes the movement of fish over increasingly larger temporal and spatial scales, ranging from months to years and from within individual habitats to between-river systems.

18


(a)

(b)

(c)

(d)

Figure 1-1. The study species: (a) river blackfish, (b) Macquarie perch, (c) Australian grayling and (d) golden perch.

19


1.6 Thesis objectives The overarching objective of this thesis is to examine the movement behaviours of four species of threatened freshwater fish in regulated riverine ecosystems in southeastern Australia and the consequences for conservation and management strategies. This thesis will provide data to improve our understanding of the factors influencing fish movement and fill crucial gaps in our knowledge of movement behaviours that currently limit our ability to develop targeted conservation management strategies. In addition, synthesis of findings relating to a range of species exhibiting a variety of distinct life histories may enable the identification of common factors that influence movement and that can inform and guide the management of freshwater fishes. Chapter 2 describes day-to-day activity and habitat use of river blackfish from winter to spring in Armstrong Creek, southern Victoria, using radio-telemetry, with an emphasis on fine spatial and temporal scales of movement and the role of the diel period in influencing movements between habitats. Chapter 3 describes day-to-day activity, habitat use and spawning season movement behaviours of Macquarie perch over a single spawning season in the Yarra River, southern Victoria, using radio-telemetry. It specifically investigates whether this riverine population exhibits synchronised migrations to specific river reaches during the spawning season. Chapter 4 describes the spawning-related movement behaviours of Australian grayling over four years in the Bunyip River, southern Victoria, using acoustic telemetry and egg/larval drift sampling, with an emphasis on determining whether the species undertakes downstream spawning migrations to specific breeding grounds and whether such behaviours are associated with increased river flows. Chapter 5 describes the movements of golden perch over four years in the Goulburn and Murray rivers using acoustic telemetry and egg/larval drift sampling, focusing on tributary and mainstem connections as corridors for movement and the influence of increased river flows and reproductive behaviour as drivers of movement.

20


Finally, in Chapter 6, I summarise the data chapters and the contribution of the four studies to our understanding of the movement ecology of native fish. Based on my findings, I present updated movement models for the study species developed from data in this thesis, and discuss the consequences and implications for restoration strategies, such as environmental flows. I also identify opportunities and directions for further research that may be used to inform and guide the management of fishes in riverine ecosystems in south-eastern Australia and elsewhere.

21


22


2 Diurnal and nocturnal movements of river blackfish in Armstrong Creek

This chapter has been published as: Koster, W.M. & Crook, D.A. 2008. Diurnal and nocturnal movements of river blackfish (Gadopsis marmoratus) in a south-eastern Australian upland stream. Ecology of Freshwater Fish 17: 146–154.

23


AUTHORSHIP STATEMENT Details of publication and executive author Title of Publication

Publication details

Diurnal and nocturnal movements of river blackfish (Gadopsis marmoratus) in a south‐eastern Australian upland stream Name of executive author School/Institute/Division if based at Deakin; Organisation and address if non‐Deakin

Ecology of Freshwater Fish 17: 146–154 Email or phone

Wayne Koster

03 9450 08600

Arthur Rylah Institute for Environmental Research, Department of Sustainability and Environment, 123 Brown Street, Heidelberg, 3084, Victoria, Australia

Inclusion of publication in a thesis Is it intended to include this publication in a higher degree by research (HDR) thesis?

Yes

If Yes, please complete Section 3 If No, go straight to Section 4.

HDR thesis author’s declaration Name of HDR thesis author if different from above. (If the same, write “as above”) As above

School/Institute/Division if based at Thesis title Deakin

Movement ecology and conservation implications for riverine fishes of south‐eastern Australia If there are multiple authors, give a full description of HDR thesis author’s contribution to the publication (for example, how much did you contribute to the conception of the project, the design of methodology or experimental protocol, data collection, analysis, drafting the manuscript, revising it critically for important intellectual content, etc.) I was the project leader and responsible for the experimental design (70%), fieldwork (70%), data analyses (70%), drafting the manuscript (70%) and revising it critically for important intellectual content (80%) Signature I declare that the above is an accurate description of 5/12/2014 my contribution to this paper, and the contributions of and date

other authors are as described below. Description of all author contributions Name and affiliation of author

D.A. Crook

Contribution(s) (for example, conception of the project, design of methodology or experimental protocol, data collection, analysis, drafting the manuscript, revising it critically for important intellectual content, etc.) Assisted with the experimental design, fieldwork, drafting and revising the manuscript

24


Author Declarations I agree to be named as one of the authors of this work, and confirm: i. that I have met the authorship criteria set out in the Deakin University Research Conduct Policy, ii. that there are no other authors according to these criteria, iii. that the description in Section 4 of my contribution(s) to this publication is accurate, iv. that the data on which these findings are based are stored as set out in Section 7 below. If this work is to form part of an HDR thesis as described in Sections 2 and 3, I further v. consent to the incorporation of the publication into the candidate’s HDR thesis submitted to Deakin University and, if the higher degree is awarded, the subsequent publication of the thesis by the university (subject to relevant Copyright provisions). Name of author Signature* Date D.A. Crook

5/12/14

Other contributor declarations I agree to be named as a non‐author contributor to this work. Name and affiliation of contributor

Contribution

Signature and date

Data storage The original data for this project are stored in the following locations. (The locations must be within an appropriate institutional setting. If the executive author is a Deakin staff member and data are stored outside Deakin University, permission for this must be given by the Head of Academic Unit within which the executive author is based.) Data format

Storage Location

Date lodged

Excel


5/12/14

25

Name of custodian if other than the executive author


2.1 Introduction Studies of animal movements provide valuable insights into the processes that influence a species’ presence and abundance, and are increasingly recognised as an important tool in the management and conservation of species (Baker 1982; Lucas and Baras 2000). Over the past two decades, particularly with the advancement of telemetric methods, considerable progress has been made in understanding the movements of freshwater fishes (Gowan et al. 1994; Lucas and Baras 2000). Telemetric methods have commonly been used to provide detailed information on the short-term movements of fish, including variations in movement patterns over diel periods (e.g. Matthews 1996; Harvey and Nakamoto 1999; David and Closs 2001). Miniaturisation of transmitters and the development of improved transmitter attachment and implantation techniques have also allowed longer-term studies of the movements of relatively small fish, thus providing critical information on movements at temporal scales of up to several years (e.g. Eiler 1995; Harvey and Nakamoto 1999; Irving and Modde 2000). River blackfish occur in fresh waters of south-eastern Australia. The spawning period for the species is October–December (Cadwallader and Backhouse 1983; Koehn and O’Connor 1990). Two distinct forms of river blackfish, a northern and southern form, have been identified (Sanger 1986), and a recent taxonomic work suggests that these may represent distinct species (Miller et al. 2004). The southern form (~600 mm maximum total length (TL)) grows much larger than the northern form (~300 mm maximum TL) (Cadwallader and Backhouse 1983) and is a popular angling species (Jackson et al. 1996). A decline since European settlement in the range and abundance of river blackfish has been attributed primarily to stream siltation and removal of woody debris (Lake 1971; Jackson et al. 1996). Consequently, there has been a recent emphasis on stream rehabilitation programs to conserve or restore suitable habitat conditions for river blackfish populations. For example, environmental flow recommendations specifically aimed at restoring flow patterns suitable for sustaining river blackfish populations have been established, and woody debris introductions have been undertaken to provide physical habitat for the species (Coleman 2006). Although such efforts are likely beneficial to river blackfish populations, considerable gaps in our knowledge of basic aspects of the species’ life

26


history currently hinder the development and implementation of targeted management strategies for the species. Although movements of river blackfish have been examined in several previous studies, there has been little detailed investigation of their long-term or diel movements. Koehn (1986) reported the preliminary results of a mark–recapture study conducted over a 2-year period in Armstrong Creek in southern Victoria, describing river blackfish (southern form) movements as limited, and estimating a home range of 25–30 m. Similarly, Khan et al. (2004) conducted a radio-telemetry and mark– recapture study in Birch Creek in northern Victoria in late spring to early summer (November–December) and mid-spring to late autumn (October–May), respectively, and reported that river blackfish (northern form) also displayed little movement and had a small home range (10–26 m). Distances moved between day and night were also examined but were not significantly different (Khan et al. 2004). Movements of the only other member of the genus, the two-spined blackfish, have been investigated using mark–recapture techniques in the Cotter River in the Australian Capital Territory between late spring and late autumn (November–May), and were similarly described as relatively sedentary, with a home range of ~15 m (Lintermans 1998). The current study was undertaken to investigate the movement of river blackfish using radio-telemetry at two temporal scales: (i) two to three times per week during daylight over 48 days and (ii) hourly for three consecutive days and nights. The results of the study are discussed with regard to the general ecology of river blackfish and in terms of implications for management and conservation of the species.

2.2 Materials and methods Study area The study was conducted in Armstrong Creek, a second-order tributary of the Yarra River, ~100 km east of Melbourne in south-eastern Australia (Figure 2-1, Figure 2-2). Streamflows in Armstrong Creek are partially regulated through small diversion weirs located on the eastern and western branches of the creek that supply water for use in Melbourne. Both weirs act as barriers to fish movement. The upper reaches of each creek flow through largely forested catchment. Two study sites were selected (mean stream width 4–5 m), one 250 m downstream and one 150 m upstream of the

27


diversion weir on the West Branch (Figure 2-2). The weir on the West Branch at full supply level retains ~20 ML. Daily discharge records were obtained from a gauging station on the West Branch weir, and water temperature was recorded in situ during each tracking event. The study was conducted between late winter (August) and midspring (October). Water temperatures gradually increased over this period from about 8 to 13°C. A high-discharge event (maximum flow ∼470 ML day−1) occurred in late August–early September (Figure 2-3). Apart from this event, discharge was generally much lower and more stable downstream of the weir (median flow ∼5 ML day−1) compared with upstream of the weir (median flow ∼95 ML day−1) because of flow regulation (Figure 2-3). This downstream site was previously used for studies of fish movement and habitat use (Koehn 1986; Koehn et al. 1994) and the effects of sedimentation on fish and macroinvertebrates (Doeg and Koehn 1994).

Figure 2-1. Armstrong Creek.

28


N

145.52 E W

E

t Br a nch

S

0

Wes

1

Site 2 weir weir

ch ran B st Ea

Arm stro ng

Cre ek

-37.38 S Site 1

r Rive

Ya rra

-37.40 S Australia Reefton

Figure 2-2. Location of Armstrong Creek in south-eastern Australia.

29

1 km


Fish collection and radio-tagging Seven river blackfish (southern form) were collected downstream and six upstream of the West Branch weir (mean TL 260 ± 40 mm SD, mean weight 182 ± 90 g SD) by backpack electrofishing on 18 August 2005. A radio-transmitter with an internal coil antenna and a battery life of ∼60 days (model 1040; Advanced Telemetry Systems (ATS), Isanti, MN, USA; frequency: 150 MHz; dimensions: 24 × 10 × 7 mm; weight: 2 g in air) was implanted into the body cavity of each fish through an incision of ~10 mm adjacent to the pectoral fin, extending towards the anus. The incision was not closed with sutures because trials indicated much higher survival and tag retention rates with unclosed incisions than with suturing (unpublished data). The transmitter weight to fish body weight ratio did not exceed 2%, and observations of six tagged fish held in aquaria for 6 weeks showed no apparent effects of the transmitter implantation method upon behaviour (i.e. swimming, feeding or condition). Each fish was released near its point of capture immediately after recovery from the transmitter implantation procedure. Fish monitoring Data were collected for five of the six river blackfish tagged upstream and six of the seven fish tagged downstream of the weir. Despite extensive searches, we were unable to locate the remaining two radio-tagged river blackfish and assume that the transmitters had failed before any data could be collected. Tracking began 6 days after implantation of the transmitters. Movements of each river blackfish were recorded during daylight on two or three occasions per week from late winter to mid-spring (24 August to 5 October 2005). River blackfish were located in the morning (09:00–11:00 hours) and afternoon (14:00–16:00 hours) of each day by triangulation with a handheld, three-element Yagi antenna and an ATS receiver. Because of the battery life limitations, the fish were only able to be tracked for 48 days. At the downstream site, a ford crossing downstream of the radio-tagged fish was used as a reference point for the location of fish. At the upstream site, the West Branch weir wall downstream of the radio-tagged fish was used as a reference point. The distance of each river blackfish from the relevant reference point was recorded to the nearest 1 m. At each site, markers were placed at 5-m intervals on the bank to assist with the location of fish relative to the reference points.

30


In addition, radio-tagged river blackfish downstream of the weir were tracked hourly for 71 consecutive hours from 3 to 6 October 2005 to examine diel movements. Mesohabitats for each fish location were recorded using the categories ‘riffle’, ‘run’ or ‘pool’ (after Anderson et al. 1989). Smaller-scale habitat features (undercut bank, woody debris or open channel) for each fish location were also recorded when possible. However, smaller-scale habitats for each fish location could not always be accurately determined at night. After the tracking was completed, mesohabitat composition was measured along a 200-m section of stream encompassing the most upstream and downstream diel movements.

2.3 Results Daylight movement Most of the 11 radio-tracked river blackfish remained within a short ( 0.05). The areas of stream occupied by individual river blackfish did not overlap during the three diel periods (Figure 2-6). Runs (43%) comprised the largest proportion of available mesohabitat, followed by riffles (34%) and pools (23%). Although pools comprised the least habitat, river blackfish were positioned within pools more often than within other mesohabitats during both the day (50% of observations) and night (59% of observations) (Table 2-2). During the day, river blackfish did not move between mesohabitats (Table 2-2). In contrast, at night, four of the six river blackfish moved between two or more mesohabitats (Figure 2-6, Table 2-2). At the smaller scales, four river blackfish were located exclusively in undercut banks and two fish among woody debris during the day (Figure 2-7). Although the small-scale habitats used by river blackfish could not be accurately determined at night, it was apparent that most fish were often moving between habitats, including the open channel.

35


Figure 2-6. Distances (m) moved and mesohabitats used by six radio-tagged river blackfish in Armstrong Creek below the West Branch weir over the 3-day diel tracking period. The shaded area refers to the period between sunset and sunrise. Distance (m) refers to distance upstream of the ford crossing that was used as a reference point for the location of fish. None of the locations of the six river blackfish overlapped.

36


Table 2-1. Total distance moved and total linear range (m) of six river blackfish during the 3-day diel tracking period. Total distance moved (m)

Total linear range (m)

Day

Night

Day

Night

a

7

168

1

16

b

0

36

1

12

c

9

15

2

3

d

5

223

1

25

e

0

41

0

5

f

11

68

2

9

Average ± SD

5.3 ± 4.6

91.8 ± 83.9

1.2 ± 0.7

11.7 ± 8.0

Fish

Total distance moved refers to the sum of all hourly movements. Total linear range refers to the distance between the most upstream and downstream positions.

Table 2-2. Percentage use of mesohabitats for six river blackfish during the 3-day diel tracking period. Day

Night

Fish Pool

Run

Riffle

Pool

Run

Riffle

a

0

0

100

5.1

79.5

15.4

b

100

0

0

87.2

2.5

10.3

c

100

0

0

100

0

0

d

0

100

0

66.7

28.2

5.1

e

0

100

0

0

100

0

f

100

0

0

92.3

0

7.7

37


Figure 2-7. Example of an undercut bank used by river blackfish in Armstrong Creek during the study.

38


2.4 Discussion The results of this study provide a demonstration of the importance of the temporal scale and timing of monitoring for understanding patterns of fish movement. Measurement of the movements of river blackfish during the day only, both short- and long-term, would have indicated that most individuals displayed little or no movement and were confined to distinct positions in the stream, most often an undercut bank. Incorporating intensive monitoring at night, however, showed that individuals occupy a much larger range and regularly move between mesohabitats during the night. This finding is similar to several previous radio-telemetry studies (e.g. Harvey and Nakamoto 1999; Snedden et al. 1999; Hilderbrand and Kershner 2000) that also found strong diel patterns in the movements of riverine fish. Nocturnal behaviour by river blackfish has been noted by previous authors (e.g. Koehn et al. 1994; Jackson et al. 1996). However, the increased movement at night by river blackfish contrasts with the findings of the only published study on diel activity by the species (Khan et al. 2004), which found no significant difference in the distances moved by individual fish between day and night. A possible explanation for the discrepancy between the studies is that Khan et al. (2004) studied the smaller, northern form of river blackfish, which might exhibit different behavioural characteristics to the southern form. The current study was also conducted over a longer period, with shorter tracking intervals and a greater number of fish. Khan et al. (2004) tracked three fish at 3-hourly intervals over a single diel period, compared with six fish tracked hourly over three diel periods. In the current study, fish frequently changed locations between the hourly tracking occasions at night: it is possible that tracking less fish at longer intervals over a single diel period resulted in a failure to detect variation in diel movement. Finally, there were differences in the timing of the studies and the techniques used to attach the radio-transmitters, which could potentially have affected fish behaviour. The current study was conducted between late winter and mid-spring, whereas Khan et al. (2004) conducted their radio-tracking between late spring and early summer. In the current study, radio-transmitters were implanted internally, while Khan et al. (2004) attached the radio-transmitters externally, and attached the transmitters to much smaller fish.

39


During the diel tracking, river blackfish used pools more frequently than their availability in the study reach would predict, but they were also commonly located in riffles and runs. The preference of river blackfish for pools agrees with the findings of previous studies (Jackson 1978; Koehn et al. 1994; Khan et al. 2004); however, frequent utilisation of riffles and runs has not been described in previous studies. Although river blackfish were found in riffles and runs both day and night in the current study, within these mesohabitats they were often positioned in undercut banks or among woody debris that may have afforded shelter from the surrounding fastflowing waters. Nonetheless, the results indicate that river blackfish are not confined to pools, and suggests that small-scale features, such as undercut banks or woody debris, play an important role as habitat for this species. Interestingly, the locations of the 11 radio-tagged river blackfish rarely overlapped during the study. Larger individuals of many species are known to exclude other large individuals from territorial areas, and dominance hierarchies (with subordinate smaller individuals) often form in these areas (e.g. Hughes 1992; David and Stoffels 2003). While non-tagged river blackfish were certainly present within the ranges of the radio-tagged fish during the study, it is likely that most of these were smaller than the radio-tagged fish. We undertook extensive sampling to collect fish >220 mm TL for radio-tagging. Although all fish greater than this length were radio-tagged, many smaller fish were also collected from the study reach (average density 0.07 blackfish m−2; unpublished data). Observations in aquaria have shown that individuals exhibit aggressive behaviour towards each other and that large individuals tend to dominate smaller fish (Cadwallader and Backhouse 1983; personal observation). It is possible that large river blackfish occupy non-overlapping home ranges that they share with smaller, subordinate individuals as part of a dominance hierarchy. However, further work on behavioural interactions between individual fish is required to confirm this suggestion. Over half of the fish tagged below the weir were positioned within the inundated floodplain habitats during the short-lived flood event in late August to early September. Although many fish species use flooded off-channel habitats (Ross and Baker 1983; Brown and Hartman 1988; Matheney and Rabeni 1995; Brown et al. 2001), similar observations on river blackfish have not been documented.

40


Observations of movement by river blackfish onto flooded riparian areas have potentially important implications for management of the riparian zone of streams containing river blackfish. Riparian vegetation may provide refuge habitats during flood events, as has been reported for other riverine fish species (Matheney and Rabeni 1995). A few fish downstream of the weir also undertook large movements coinciding with the increased flows, and one of these fish also moved again, coinciding with another smaller increase in flows. Previous studies have suggested that increased discharges provide important opportunities for individuals to explore and colonise other stream locations (David and Closs 2002; Crook 2004a). Movement away from established locations may also occur in response to other disturbances (e.g. handling by humans; Crook 2004a) or be associated with particular aspects of the species’ life history (e.g. spawning; Matheney and Rabeni 1995). Movements away from established locations by river blackfish during periods of high discharge were outside the October–December spawning period for the species (Cadwallader and Backhouse 1983; Koehn and O’Connor 1990) and therefore would appear to be associated with non-reproductive behaviour. The movement behaviour of river blackfish during the spawning period is an important area for future research. A few fish upstream of the weir also undertook rapid large movements downstream into the weir pool shortly after release. At the time of the transmitter implantations, flows had risen in Armstrong Creek upstream of the weir. It is possible that the elevated flows forced fish downstream into the weir pool while they were recovering from the transmitter implantations, or perhaps the movements were associated with initial post-release mobility. Although there were no physical obstructions to the upstream movement of fish from the weir pool back into the creek, these fish remained in the weir pool for the duration of the study. This finding shows that river blackfish are capable of using modified environments, such as weir pools, although their long-term viability in such environments is unclear. In conclusion, this study documents new observations on the movements and habitat use of river blackfish. Integration of such information into management strategies has the potential to improve our capacity to provide the conditions required to conserve and restore river blackfish populations. The study has also confirmed that the scale and timing of observations can lead to very different conclusions regarding the extent

41


of movement by the fish under investigation (e.g. Hilderbrand and Kershner 2000; Ovidio et al. 2000; Horton et al. 2004), and that this should be considered when appraising the movement requirements of riverine fish. Acknowledgements This study was funded by Melbourne Water. We gratefully acknowledge the assistance of Rowan Compagnoni, Cameron Padgham, John De Boer, Aaron Pie, Nigel Saville and Peter MacDonald from Melbourne Water. Jed Macdonald, John Morrongiello, Damien O’Mahony, Peter Fairbrother, Zeb Tonkin, Justin O’Mahony and Karl Pomorin from the Arthur Rylah Institute assisted with the fieldwork. Jed Macdonald and John Morrongiello deserve a special mention for their unwavering enthusiasm during the long days and nights of the diel tracking trip. Damien O’Mahony provided invaluable expertise and assistance during development of the transmitter implantation method. We also thank John Koehn, Ivor Stuart and Paul Close for constructive comments on earlier versions of the manuscript. This study was conducted under ethics permit 05/001 (Arthur Rylah Institute, Animal Ethics Committee).

42


3 Spawning season movements of Macquarie perch in the Yarra River

This chapter has been published as: Koster, W.M., Dawson, D.R., Morrongiello, J.R. & Crook, D.A. 2014. Spawning season movements of Macquarie perch (Macquaria australasica) in the Yarra River, Victoria. Australian Journal of Zoology 61: 386– 394. 43


AUTHORSHIP STATEMENT Details of publication and executive author Title of Publication

Publication details

Spawning season movements of Macquarie perch (Macquaria australasica) in the Yarra River, Victoria Name of executive author School/Institute/Division if based at Deakin; Organisation and address if non‐Deakin

Australian Journal of Zoology 61: 386–394 Email or phone

Wayne Koster

03 9450 08600


Inclusion of publication in a thesis Is it intended to include this publication in a higher degree by research (HDR) thesis?

Yes

If Yes, please complete Section 3 If No, go straight to Section 4.

HDR thesis author’s declaration Name of HDR thesis author if different from above. (If the same, write “as above”) As above

School/Institute/Division if based at Thesis title Deakin

Movement ecology and conservation implications for riverine fishes of south‐eastern Australia If there are multiple authors, give a full description of HDR thesis author’s contribution to the publication (for example, how much did you contribute to the conception of the project, the design of methodology or experimental protocol, data collection, analysis, drafting the manuscript, revising it critically for important intellectual content, etc.) I was the project leader and responsible for the experimental design (70%), fieldwork (70%), data analyses (60%), drafting the manuscript (70%) and revising it critically for important intellectual content (80%) Signature I declare that the above is an accurate description of 5/12/14 my contribution to this paper, and the contributions of and date

other authors are as described below. Description of all author contributions Name and affiliation of author

D.R. Dawson J.R. Morrongiello D.A. Crook

Contribution(s) (for example, conception of the project, design of methodology or experimental protocol, data collection, analysis, drafting the manuscript, revising it critically for important intellectual content, etc.) Assisted with the fieldwork and drafting the manuscript Provided considerable guidance and assistance with data analysis, assisted with revising the manuscript Assisted with the experimental design, fieldwork, and drafting and revising the manuscript

44


Author Declarations I agree to be named as one of the authors of this work, and confirm: vi. that I have met the authorship criteria set out in the Deakin University Research Conduct Policy, vii. that there are no other authors according to these criteria, viii. that the description in Section 4 of my contribution(s) to this publication is accurate, ix. that the data on which these findings are based are stored as set out in Section 7 below. If this work is to form part of an HDR thesis as described in Sections 2 and 3, I further x. consent to the incorporation of the publication into the candidate’s HDR thesis submitted to Deakin University and, if the higher degree is awarded, the subsequent publication of the thesis by the university (subject to relevant Copyright provisions). Name of author Signature* Date D.R. Dawson J.R. Morrongiello D.A. Crook

5/12/14 5/12/14 5/12/14

Other contributor declarations I agree to be named as a non‐author contributor to this work. Name and affiliation of contributor

Contribution

Signature and date

Data storage The original data for this project are stored in the following locations. (The locations must be within an appropriate institutional setting. If the executive author is a Deakin staff member and data are stored outside Deakin University, permission for this must be given by the Head of Academic Unit within which the executive author is based.) Data format

Storage Location

Date lodged

Excel


5/12/14

45

Name of custodian if other than the executive author


3.1 Introduction Freshwater fishes can exhibit considerable intraspecific variation in life history characteristics among populations (Mazzoni and Iglesias-Rios 2002; Blanck and Lamouroux 2007). Fishes from contrasting habitats, such as rivers and lakes, for example, have demonstrated wide variation in the timing, location and extent of reproduction and migration, with variations linked to factors such as physical habitat availability, water temperature and hydrological regimes (Mazzoni and Iglesias-Rios 2002; Chapman et al. 2006; Barthel et al. 2008; Koehn et al. 2009). From a conservation and management perspective, knowledge of intraspecific variability in life history patterns (such as spawning-related behaviours) can be important in predicting how populations might respond to management interventions or restoration efforts and can also be useful in predicting the effects of disturbances or environmental changes (Meka et al. 2003). The Macquarie perch is a nationally threatened freshwater fish species (under the Environment Protection and Biodiversity Conservation Act 1999) endemic to the MDB and coastal drainages of south-eastern Australia (Ingram et al. 2000; Lintermans 2007). The species is naturally a riverine fish but has established significant populations in several artificial impoundments (Lintermans 2013b). Macquarie perch has also been translocated into several streams outside its natural range, such as the Yarra River in southern Victoria (Lintermans 2007). Formerly abundant throughout a wide geographic range (Cadwallader and Rogan 1977; Trueman 2007), the species has undergone a major decline in range and abundance since the 1970s: most remaining populations are now small and geographically isolated (Cadwallader and Rogan 1977; Ingram et al. 2000; Lintermans 2007). Habitat degradation and loss, barriers to movement, altered flow and water temperature regimes, overfishing, alien fish species, and susceptibility to epizootic haematopoietic necrosis virus have all been suggested as likely contributors to the decline of the species (Cadwallader and Rogan 1977; Ingram et al. 2000; Lintermans 2007; Faulks et al. 2011). Understanding the behaviours and habitat requirements associated with spawning is critical for the development of strategies to conserve the remaining populations of Macquarie perch. While there is good information available for populations in 46


artificial impoundments (Wharton 1968; Cadwallader and Rogan 1977; Douglas et al. 2002; Ebner et al. 2010; Tonkin et al. 2010; Broadhurst et al. 2012a; Tonkin et al. 2012; Thiem et al. 2013), substantial gaps exist in our knowledge of the movement and spawning behaviours of riverine populations (Ebner and Lintermans 2007). A synchronised upstream spawning migration in spring to early summer has been documented for impoundment populations (Wharton 1968; Cadwallader and Rogan 1977; Douglas et al. 2002; Tonkin et al. 2010). Spawning in these populations takes place en masse in inflowing tributary streams, with fish congregating in the tail end of pools and spawning in shallow, fast-flowing habitats with gravel–cobble substrates (Cadwallader and Rogan 1977; Appleford et al. 1998; Tonkin et al. 2010). Whether riverine populations of Macquarie perch exhibit similar synchronised migratory behaviour associated with spawning or move to similar habitats during the spawning season remains unknown (Lintermans 2013b). The current study examines the movements of Macquarie perch in the Yarra River, Victoria. During the early 1900s, Macquarie perch from the Goulburn River were translocated to the Yarra River outside their natural range (Cadwallader 1981). The Yarra River population of Macquarie perch is now abundant and widespread (in ~60 km of river from Yarra Glen to Heidelberg) and supports a small recreational fishery (King et al. 2011). Spawning and recruitment (i.e. larvae and young-of-year abundances) is not evenly distributed throughout the reach of the Yarra River from Yarra Glen to Heidelberg, with most larvae and young-of-year found in the lower floodplain reach in the Eltham–Templestowe area, which raises the possibility that fish may migrate to this area to spawn (King et al. 2011). While there could also be some downstream dispersal of larvae and young-of-year to this area, once hatched, larvae rapidly develop into juveniles and occupy slow-flowing habitats i.e. pools (Broadhurst et al. 2012a). Spawning and recruitment is also variable from year to year and appears to be higher during years with moderately high within-channel flows (e.g. ~2000 ML day-1) during the pre-spawning period, followed by more stable base flows during both the spawning and rearing period (King et al. 2011). In this study, we use radio-telemetry to test the hypothesis that, like their lacustrine counterparts, riverine Macquarie perch exhibit synchronised migrations during the spawning season to specific river reaches containing shallow, fast-flowing habitats. If different movement patterns associated with spawning behaviour exist between riverine and impoundment

47


Macquarie perch populations, these need to be taken into account in the management and conservation of this threatened species.

3.2 Materials and methods Study site The study was conducted in the mid-reaches (Wonga Park to Heidelberg) of the Yarra River, a coastal stream that flows into Port Phillip Bay, Victoria, Australia (Figure 3-1, Figure 3-2). The length of the Yarra River is ~240 km, with a catchment area of ~4000 km2. Average annual discharge in the Yarra basin is ~1,100,000 ML (DWR 1989). The upper reaches of the catchment are mostly forested; the mid-reaches consist largely of cleared agricultural land, while the lower reaches are heavily urbanised. Flow in the Yarra River catchment is highly regulated by dams, particularly to supply water for use in Melbourne. The mid-reaches of the river (mean stream width 25 m) comprise riffle–run–pool sequences, with slower-flowing habitats becoming more predominant with increasing distance downstream.

Figure 3-1. The Yarra River at Wonga Park.

48


145.20 E

145.10 E -37.70 S Plenty River

Wonga Park er Riv

Eltham ra Yar

k ee Cr

N W

m llu Mu

Templestowe

-37.80 S

Warrandyte

m llu Mu

Heidelberg

E

1

0

Australia

1 km

S

Figure 3-2. The location of the study site. Black circles represent the three collection and release locations (Wonga Park, Eltham and Heidelberg).

49


Fish movement A total of 30 adult Macquarie perch (mean ± s.e. TL 294 ± 6.4 mm, range 240–362 mm, weight 366 ± 24.8 g, range 204–672 g) were collected for tagging at three sites (Wonga Park n = 14, Eltham n = 6, Heidelberg n = 10) on the Yarra River using a boat-mounted electrofishing unit between April and May 2011 (Figure 3-2, Table 3-1). The size of Macquarie perch at maturity varies considerably, with river populations tending to mature at a much smaller size (e.g. males ~100 g, females ~190 g) than lake populations (Appleford et al. 1998). The sex of the fish could not be determined at the time of tagging. Fish were transferred from the stream into an aerated, 50-L holding container and individually anaesthetised (0.03 mL AQUI-S per litre water) (AQUI-S, Lower Hutt, New Zealand). Radio-transmitters with an internal coil antenna (model F1170, ATS, Isanti, USA; frequency 150 MHz; dimensions 24 × 14 mm; weight 4 g in air; estimated battery life 220 days) were implanted into the peritoneal cavity through an incision of 15 mm, adjacent to the pectoral fin, extending towards the anus. The incision was closed with 2–3 dissolvable external sutures. Only fish >200 g were tagged to ensure that transmitter-to-fish weight ratios remained below ~2% (Winter 1996). Each fish was placed into a recovery net positioned in the stream channel. Once the fish were observed to maintain their balance and freely swim throughout the holding net they were released near their point of capture (i.e. within 15 m).

50


Table 3-1. Details of the Macquarie perch tagged during the study. Site

Tag ID

Length (mm)

Weight (g)

50% linear range (m)



Wonga Park

W1

246

218

59

938

1273

W2

253

240

52

139

161

W3

260

300

35

122

5247

W4

272

300

62

825

1309

W5

288

314

56

217

248

W6

292

334

250

394

421

W7

300

374

58

275

313

W8

319

478

29

947

1276

W9*

320

436

–

–

–

W10

321

442

4

20

55

W11

324

492

39

3523

3571

W12

330

514

209

727

821

W13

336

562

20

82

119

W14

354

578

68

2098

2122

E1

258

216

29

62

82

E2*

260

208

–

–

–

E3

261

218

49

237

308

E4*

261

224

–

–

–

E5

274

240

21

225

559

E6#

305

340

49

572

572

H1

240

204

36

692

797

H2

261

270

35

1139

1749

H3

264

262

19

81

280

H4

278

292

9

96

466

H5

297

292

75

199

206

H6

306

398

28

471

610

H7#

320

512

14

51

53

H8

330

494

9

24

448

H9#

350

578

14

44

44

H10

362

672

15

83

93

Eltham

Heidelberg

* = Tag expelled or fish deceased; # = Tag could not be located 150–200 days post release.

51


Fish were located once during daylight hours (09:00–17:00 hours) every 2–4 weeks between late May 2011 and early February 2012, and one to two times per week between September and early December 2011 to coincide with the known spawning season of the species in the Yarra River (King et al. 2011). Locations of fish were determined from a small aluminium boat by triangulation using a handheld, threeelement Yagi antenna and an ATS receiver (R4100). This technique has been shown to have high location accuracy (i.e. within 2 m) (David and Closs 2002). Locations were recorded using a handheld Garmin 60CSx GPS (Garmin International, Olathe, Kansas, USA) unit, plotted in Arc View 3.3 (Environmental Systems Research Institute, Redlands, California, USA), and the longitudinal distance between tracking locations along the mid-channel was calculated in metres using the Arc View measure tool to determine the distances moved. To examine habitat use, mesohabitat composition (riffle, run or pool) for each site was measured at the completion of tracking by taking GPS coordinates wherever a change in mesohabitat type occurred along a length of stream encompassing the most upstream and downstream movements of the radio-tagged fish. These coordinates were then plotted in Arc View 3.3 to create a map of mesohabitat composition. The mesohabitats for each fish location were then determined by plotting the fish locations on the mesohabitat map in Arc View. The criteria used to classify mesohabitat types were riffle: fast flowing with broken water surface; pool: slow flowing with smooth water surface; and run: in between pool and riffle with wavy water surface (after Anderson et al. 1989; Jowett 1993; Broadhurst et al. 2011). Total linear ranges for each fish were estimated by determining the distance along the river channel between the most upstream and downstream locations. To examine the influence of occasional long-distance movement on home range estimates, we also calculated 90% linear ranges and 50% linear ranges (i.e. core areas) for each fish by calculating the minimum distance containing at least 90% and 50% of the locations, respectively (Crook 2004a). Generalised linear mixed models fitted with a binomial distribution were used to examine the probabilities of fish moving (defined as moves >200 m, i.e. away from core areas; see Results) during the spawning period (Bestley et al. 2009; Zuur et al. 2009; Dudgeon et al. 2013) using the lme4 package in R 2.14.0 (R Development Core

52


Team 2013). The explanatory variables examined were: (i) flow difference (maximum – minimum daily flow in megalitres (ML) in the intervening period between tracking events; (ii) water temperature (mean daily water temperature in the intervening period between tracking events); (iii) site; (iv) fish length (mm); and (v) day of year (tests whether movements are synchronised among individuals). We examined 32 models of increasing additive fixed-effect complexity; the most complex considered all parameters. All covariates were mean-centred to aid model convergence and interpretation, and flow difference was log10 transformed to satisfy the model assumption of linearity between a predictor and the link function. A random intercept for each individual fish was included in the model because the data consisted of repeated measurements through time for each of multiple individuals. Preliminary analysis indicated the presence of significant autocorrelation in the data whereby the movement state at time t1 was related to that at time t0. A lagged movement term (t−1) was added to the models, which successfully accounted for the autocorrelation. Relative support for each of the models was assessed by calculating Akaike’s Information Criterion (AIC) corrected for small sample (AICc) (Burnham and Anderson 2002). AICc values were rescaled as the difference between each model and the model with the lowest AICc (∆AICc), because their value is relative only to other models constructed using the same dataset. The larger a model’s ∆AICc, the less plausible that it is the best model given the data. A chi-square test of homogeneity was also used to test that there was no association between habitat use and season.

3.3 Results Radio-transmitters from three (10%) of the tagged fish remained in the same location from the commencement of tracking to the end of the study (Table 3-1). We concluded that these transmitters had been expelled or the fish had suffered mortality. Two of the fish could not be located beyond 150 days post release (i.e. late September), and another fish could not be located after 200 days (i.e. late November), despite extensive searches in the Yarra River and the Plenty River (a major tributary that enters the Yarra River within the study region) (Table 3-1). The apparent loss of these fish from the study area could be due to a number of factors, including removal (e.g. avian predation or angling) or transmitter failure. 53


Most of the fish tracked during the study remained close to their point of capture. However, two individuals at Wonga Park (W4: 272 mm TL; W12: 330 mm TL) undertook large-scale (5–6-km) movements from their capture locations into pools further upstream shortly after their release (i.e. within 4 weeks) (see Figure 3-3a). These individuals remained 4–6 km upstream of their capture locations throughout the rest of the study. Because fish were tracked less frequently (i.e. monthly) early in the study (and outside the spawning season), the environmental conditions associated with these movements were not recorded, although there was a large rise in flow from 307 to 4569 ML day-1 in early May, which coincided with the release of the fish. Water temperature around this time ranged from 11 to 14C. Three individuals also undertook longer-distance (1–5-km) movements away from their usual locations later in the study in November (Figure 3-3b, c, h). One individual at Wonga Park (W3: 260 mm TL) shifted 5 km upstream from a pool to a riffle in late November, before returning to the area it had occupied previously within 1 week (Figure 3-3b). One individual at Wonga Park (W11: 324 mm TL) shifted 2 km downstream between pools (Figure 3-3c), and one fish at Heidelberg (H2: 261 mm TL) shifted 1 km upstream from a pool to a run (Figure 3-3h) in mid-November; they remained at these locations until the completion of the tracking. These shifts followed increases in flow (e.g. from 961 to 4,509 ML day-1 in mid-November, and from 1,223 to 5,245 ML day-1 in late November). Water temperature around this time ranged from 17 to 19C. Most fish occupied small lengths of stream; the median total linear range was 448 m, and the 50% linear ranges of nearly all fish (25 of 27) were