Victorian Environmental Flows Monitoring and Assessment Program: Monitoring and assessing environmental flow releases in the Campaspe River

Full text

Victorian environmental flows
monitoring and assessment program
Monitoring and evaluation of environmental flow
releases in the Campaspe River
August 2006
(Published Online April 2009)
Yung En Chee, Angus Webb, Peter Cottingham and Mike Stewardson
eWater Cooperative Research Centre
eWater Limited ABN 47 115 422 903
Building 15, University of Canberra,
ACT 2601, Australia
Phone +61 2 6201 5168
Fax +61 2 6201 5038
1
www.ewatercrc.com.au
eWater Cooperative Research Centre
eWater Limited ABN 47 115 422 903
Building 15, University of Canberra,
ACT 2601, Australia
Phone +61 2 6201 5168
Fax +61 2 6201 5038
www.ewatercrc.com.au

Monitoring and evaluation of environmental flow releases in the Campaspe River
Victorian environmental flows monitoring and
assessment program:
Monitoring and evaluation of environmental
flow releases in the Campaspe River
Yung En Chee, Angus Webb, Mike Stewardson & Peter Cottingham
The University of Melbourne and eWater CRC
September 2006
(Published online April 2009)
A report prepared for the Department of Sustainability & Environment, Victoria
i

Monitoring and evaluation of environmental flow releases in the Campaspe River
Acknowledgements
We thank the Department of Sustainability & Environment (Victoria) as the originators and owners of
the statewide monitoring framework, and who also fund the research and implementation of VEFMAP.
In particular we thank Sabine Schreiber, Paul Bennet, Paul Lay and Jane Doolan for their support and
guidance. The members of the scientific advisory panel (Alison King, Angela Arthington, Mark
Kennard, Gerry Quinn, Barbara Downes, Wayne Tennant) and other scientific advisers (Sam Lake,
Jane Roberts, Terry Hillman, Leon Metzeling) provided careful review of the conceptual bases for the
monitoring programs, and were a ready source of information on monitoring methodologies. We also
thank the CMA waterway managers and Environmental Water Reserve officers for their input to the
VEFMAP process, and for embracing the statewide approach to environmental flows monitoring.
Lastly, we thank Payam Ghadiran (map production) and Victoria Allen (editorial assistance) for their
help in preparing this report.
Contact for more information:
Dr Angus Webb
Department of Resource Management and Geography
The University of Melbourne
Victoria 3010, Australia.
jawebb@unimelb.edu.au
This report is part of a set of eight, peer-reviewed reports commissioned and funded by the Victorian Department
of Sustainability & Environment, and it is published with that Department’s permission.
Please cite this report as:
Chee, Y., Webb, A., Cottingham, P. and Stewardson, M (2006) Victorian Environmental Flows Monitoring and
Assessment Program: Monitoring and assessing environmental flow releases in the Campaspe River. Report
prepared for the North Central Catchment Management Authority and the Department of Sustainability and
Environment. e-Water Cooperative Research Centre, Melbourne
.http://ewatercrc.com.au/reports/VEFMAP_Campaspe-River.pdf
© eWater Cooperative Research Centre 2008
This report is copyright. It may be reproduced without permission for purposes of research, scientific
advancement, academic discussion, record-keeping, free distribution, educational use or other public benefit,
provided that any such reproduction acknowledges eWater CRC and the title and authors of the report. All
commercial rights are reserved.
Published online December 2008
ISBN 978-1-921543-15-9
eWater CRC
Innovation Centre, University of Canberra
ACT 2601, Australia
Phone (02) 6201 5168
Fax (02) 6201 5038
Email info@ewatercrc.com.au
Web www.ewatercrc.com.au
eWater CRC is a cooperative joint venture whose work supports the ecologically and economically sustainable
use of Australia’s water and river systems. eWater CRC was established in 2005 as a successor to the CRCs for
Freshwater Ecology and Catchment Hydrology, under the Australian Government’s Cooperative Research
Centres Program.
ii

Monitoring and evaluation of environmental flow releases in the Campaspe River
Contents
1 Introduction ................................................................................................................................................... 5
1.1 Victorian Environmental Flows Monitoring & Assessment Program ....................................................... 5
1.2 VEFMAP Process and Report Structure ................................................................................................ 6
1.3 Environmental flow objectives for the Campaspe River ......................................................................... 8
2 Conceptual models, hypotheses & response variables relevant to the Campaspe River .................... 12
2.1 Intended uses for conceptual models ................................................................................................... 12
2.2 Model development .............................................................................................................................. 12
2.3 Geomorphic Processes ........................................................................................................................ 14
2.3.1 Description of Conceptual Model ..................................................................................................... 14
2.3.2 Selected Sub-Hypotheses & Response Variables to be Monitored ................................................. 16
2.3.2.1 Complementary Research Issues ........................................................................................... 18
2.4 Habitat Processes & Macroinvertebrates ............................................................................................. 19
2.4.1 Description of Conceptual Model ..................................................................................................... 19
2.4.2 Selected Sub-Hypotheses & Response Variables to be Monitored ................................................. 23
2.4.2.1 Complementary Research Issues ........................................................................................... 26
2.5 Aquatic and Riparian Vegetation .......................................................................................................... 27
2.5.1 Description of Conceptual Model ..................................................................................................... 27
2.5.2 Selected Sub-Hypotheses & Response Variables to be Monitored ................................................. 32
2.5.2.1 Complementary Research Issues ......................................................................................... 333
2.6 Native Fish – Spawning & Recruitment ................................................................................................ 33
2.6.1 Description of Conceptual Model ................................................................................................... 333
2.6.2 Selected Sub-Hypotheses & Response Variables to be Monitored ............................................... 377
2.6.2.1 Complementary Research Issues ........................................................................................... 38
2.7 Water Quality ....................................................................................................................................... 39
2.7.1.1 Complementary Research Issues ........................................................................................... 40
2.8 Summary of Variable Definitions & Sampling Timing and Protocols .................................................... 40
2.9 Refining the Conceptual Models and Variables for the Campaspe River ............................................. 41
2.10 Potentially Important Adverse Effects .................................................................................................. 54
3 Monitoring program design and data analysis ......................................................................................... 55
3.1 Background .......................................................................................................................................... 55
3.1.1 Practical constraints for monitoring the effects of environmental flows ............................................ 55
3.1.2 Proposed approach to the VEFMAP monitoring program design .................................................... 56
3.2 Bayesian hierarchical approach to data analysis ................................................................................. 58
3.3 Reach-by-Reach Monitoring Design .................................................................................................... 59
iii

Monitoring and evaluation of environmental flow releases in the Campaspe River
iv
3.3.1 Reach 1: Coliban River, Malmsbury Reservoir to Lake Eppalock .................................................... 61
3.3.1.1 Reach Description ................................................................................................................... 61
3.3.1.2 Intended Program of Environmental Flows Delivery ............................................................... 62
3.3.1.3 Monitoring Recommendations ................................................................................................ 64
3.3.2 Reach 2: Lake Eppalock to Campaspe Weir ................................................................................... 65
3.3.2.1 Reach Description ................................................................................................................... 65
3.3.2.2 Intended Program of Environmental Flows Delivery ............................................................... 66
3.3.2.3 Monitoring Recommendations ................................................................................................ 69
3.3.3 Reach 3: Campaspe Weir to Campaspe Siphon ............................................................................. 70
3.3.3.1 Reach Description ................................................................................................................... 70
3.3.3.2 Intended Program of Environmental Flows Delivery ............................................................... 71
3.3.3.3 Monitoring Recommendations ................................................................................................ 73
3.3.4 Reach 4: Campaspe Siphon to River Murray .................................................................................. 75
3.3.4.1 Reach Description ................................................................................................................... 74
3.3.4.2 Intended Program of Environmental Flows Delivery ............................................................. 765
3.3.4.3 Monitoring Recommendations ................................................................................................ 77
3.4 Other conceptual models and hypotheses of local importance in each reach ...................................... 79
4 Implementing the Monitoring Design ........................................................................................................ 82
4.1 VEFMAP implementation group ........................................................................................................... 82
4.2 Quality assurance/quality control ......................................................................................................... 82
5 References ................................................................................................................................................... 87
Appendix 1: Environmental Flow Recommendations for Reaches of the Campaspe River (from SKM
2006b) ........................................................................................................................................................... 94
Appendix 2: Selection of Conceptual Models ................................................................................................... 97
Appendix 3: Bayesian Hierarchical Modelling & Illustrative Case-study...................................................... 125

Monitoring and evaluation of environmental flow releases in the Campaspe River
1. Introduction
1.1 Victorian Environmental Flows Monitoring & Assessment Program
The provision of water to meet environmental objectives is a key feature of the Victorian River Health
Strategy (DNRE 2002), which recognises the importance of the flow regime to river function and
health. To this end, the Victorian Government is establishing Environmental Water Reserves (EWRs)
that define a legally recognised share of water to be set aside to maintain or improve the
environmental values of Victoria’s river systems (DSE 2004). Water will be delivered as environmental
flows to achieve specific ecosystem outcomes in a number of Victoria’s large regulated rivers.
It is important to demonstrate whether or not the EWRs are achieving the desired ecosystem
outcomes. The delivery of environmental flows represents a considerable investment in river
protection and rehabilitation, especially given the competing demands for consumptive uses of water.
Future decisions about the provision of environmental flows will rely on evidence that demonstrates
the benefits or otherwise of these water allocations. Additionally, the large-scale delivery of
environmental flows is a relatively new form of river rehabilitation. Evaluating ecosystem responses to
changes in the flow regime will provide valuable information to support future decision-making within
an adaptive management cycle.
The intention of the Victorian Government (Cottingham et al. 2005b) is to:
Evaluate ecosystem responses to environmental flows in the eight high-priority regulated
rivers that are to receive enhancements (of various degrees) to their flow regime.
The Victorian Environmental Flow Monitoring and Assessment Program (VEFMAP) has been
established to coordinate the monitoring of ecosystem responses to environmental flows. To establish
a robust and scientifically defensible monitoring program, the Department of Sustainability and
Environment (DSE) requires:
• A consistent, scientifically defensible, framework for monitoring environmental flows across
Victoria;
• Detailed, hypothesis based, monitoring plans for each of the eight high-priority rivers where the
delivery of environmental flows is expected or underway;
• Sufficient flexibility in the monitoring framework and plans so that they can be adapted in light of
changing conditions and information generated from annual analysis of monitoring data;
• Ongoing scientific support to review the data and critically analyse the monitoring programs as
implemented by the Catchment Management Authorities (CMAs). A full-scale data analysis and a
review of progress against the program objectives for each monitoring program are anticipated
every three years.
5

Monitoring and evaluation of environmental flow releases in the Campaspe River
The high-priority rivers (and associated environmental flow studies) to be initially included in the
Statewide program are the:
• Broken River (Cottingham et al. 2001);
• Goulburn River (Cottingham et al. 2003a, b);
• Campaspe River (Marchant et al. 1997, SKM 2005, 2006a, b);
• Loddon River (LREFSP 2002a, b, 2005);
• Thomson River (Earth Tech 2003);
• Macalister River (SKM 2003a);
• Wimmera River (SKM 2001, 2002, 2003c);
• Glenelg River (SKM 2003b).
Separate reports discuss environmental flow monitoring in each of these rivers. This report deals with
monitoring and assessment of environmental flows in the Campaspe River.
1.2 VEFMAP Process and Report Structure
The Statewide program is being developed and delivered in three main stages:
i. Development of an overarching Victorian (Statewide) framework for monitoring ecosystem
response to environmental flow releases (Cottingham et al. 2005a, b);
ii. Development of targeted monitoring and assessment plans for individual river systems (this report
and those for the other rivers);
iii. Data analysis and interpretation, and program review after three years, including testing the value
of taking a statewide approach to monitoring environmental flows.
For Stage 2, the framework and associated recommendations from Stage 1 (Cottingham et al. 2005b)
were applied to develop the present monitoring and evaluation plan.
In developing this monitoring and evaluation plan, the general approach has been to:
• Define the conceptual understanding of flow-ecology relationships and the hypotheses to be
tested using the original environmental flow reports and other literature;
• For the conceptual models developed, seek feedback from Scientific Panel members (involved in
the original environmental flow studies for the individual river systems), the project Advisory Panel,
DSE and CMA staff;
• Confirm key conceptual models and hypotheses that will form the basis of the monitoring and
evaluation program.
• Consider the EWR releases expected for the next 2-3 years in individual rivers. Confirm the
relevant hypotheses to be tested and from these, what variables are to be monitored in each river
reach;
6

Monitoring and evaluation of environmental flow releases in the Campaspe River
• Examine current monitoring arrangements in each river system (if any) and discuss how this
aligns with recommendations in the VEFMAP;
• Consider Bayesian and other analyses that are appropriate, their assumptions and data
requirements, and implications for the study design and interpretation of results.
The logical process used to arrive at recommendations for monitoring, along with the structure of the
report is summarized in Figure 1. In each of the reaches previously identified for environmental flow
enhancement in the Campaspe River (Marchant et al. 1997, SKM 2006), environmental flows and
other major influences (e.g. land use) will drive the ecosystem responses. We have synthesised the
conceptual models previously developed to obtain integrated conceptual models that illustrate our
belief of how certain ecosystem components will respond to environmental flows. The models suggest
measurement endpoints (e.g. bank erosion, fish abundance) that can be obtained from various field
programs (e.g. channel surveys and electro-fishing for the two endpoints above). We expect these
endpoints will respond to environmental flows, and these responses will be tested using Bayesian or
other analytical approaches. The models also contain areas of uncertainty that reduce our ability to
predict the effects of environmental flows on certain ecosystem outcomes. These knowledge gaps are
noted, and recommended as questions for specific research to be carried out concurrently with the
monitoring program. Failure to carry out such research will not necessarily prevent predictions from
being made, but they are likely to be more uncertain.
Figure 1. Process followed in developing the individual monitoring plans. The report is also
structured according to this logic.
7

Monitoring and evaluation of environmental flow releases in the Campaspe River
1.3 Environmental flow objectives for the Campaspe River
Environmental flow objectives for the Campaspe River (SKM 2006, Table 1) were based on
biodiversity and hydrological considerations. Flow recommendations to meet these objectives
(Appendix 1) were developed for the following reaches of the Campaspe River (see Figure 2):
• Reach 1 – Coliban River, Malmsbury Reservoir to Lake Eppalock;
• Reach 2 – Lake Eppalock to Campaspe Weir;
• Reach 3 - Campaspe Weir to Campaspe Siphon;
• Reach 4 - Campaspe Siphon to Murray River;
The flow recommendations were designed to meet biodiversity and flow objectives. Biodiversity
objectives related to a desired future state of key listed or threatened fauna within each reach, flora
and fauna of value, or those with a strong relationship with flow. Biodiversity objectives also included
objectives for the physical nature of the channel or ecological processes (e.g. connectivity) that have
an indirect influence on biodiversity through some physical response. Flow objectives related to the
important flow components required to achieve the Biodiversity Objectives.
8

Monitoring and evaluation of environmental flow releases in the Campaspe River
Table 1. Summary of Campaspe River environmental flow objectives (from SKM 2006)
Reach 1 – Coliban River between Malmsbury Reservoir and Lake Eppalock.
Reach 2 – Campaspe River between Lake Eppalock and Campaspe Weir.
9

Monitoring and evaluation of environmental flow releases in the Campaspe River
10
Reach 3 – Campaspe River between Campaspe Weir and Campaspe Siphon.
Reach 4 – Campaspe River between Campaspe Siphon and the River Murray.

Monitoring and evaluation of environmental flow releases in the Campaspe River
Figure 2. Map of Campaspe
River catchment showing
mean annual discharge of
the environmental flow
reaches as different shades
of blue, where darker blue
indicates greater relative
discharge, as well as flood
extent associated with
reaches of the Campaspe
River. The boundaries of
the reaches are marked by
pink bars.
11

Monitoring and evaluation of environmental flow releases in the Campaspe River
2 Conceptual models, hypotheses & response variables
relevant to the Campaspe River
2.1 Intended uses for conceptual models
The proposed conceptual models are described in the following sections to illustrate how models
and/or relationships that will underpin monitoring and testing of Statewide hypotheses may be derived.
Knowledge gaps, key questions and potential monitoring endpoints have been highlighted. The
models presented are generic, and would need to be modified to be more river-specific.
At this stage however, it is not proposed that these models will form the basis of quantitative models to
be used to make predictions about the expected level of response to a change in flow regime in any
particular system. Hence, the river-specific modifications have not been made. Rather, they seek to
synthesise our current understanding, to depict qualitatively our expectation of how an environmental
component will react to a given flow intervention. At this early stage of knowledge concerning
responses of the biota to flows, the models state our expectation of:
1) Which aspect of the flow regime (and when and where) to measure in order to confirm that the
hypothesised flow requirements for the endpoint are being met;
2) What ecological endpoint (or endpoints) to measure to determine if a response has occurred;
3) Where there are known major gaps in our knowledge as to the effects of an aspect of the flow
regime on the ecological endpoint/s. This can then be used as a guide for the conducting of
specific research projects designed to fill such gaps;
4) The presence of other overriding effects within the causal network that may prevent the beneficial
effects of environmental flows from being realised (e.g. cold water pollution).
We recognise that no conceptual representation will be complete or adequate for all circumstances.
Thus there is no expectation that these models provide a detailed representation of any particular
system. However, as these models summarise the best scientific understanding we currently have on
the relationship between specific flow components and their desired ecological outcomes, they are
crucial to formulating and testing the key questions of this project. Refinements and additions to the
models should only be undertaken if there is a belief that this will lead to a change in how we monitor
whether or not an ecological endpoint is being affected, and by which aspects of the flow regime or
another overriding influence.
2.2 Model development
We performed an exhaustive review of the scientific and management literature that collated and
summarised all the conceptual models and predicted ecosystem responses associated with the
environmental flow recommendations for the eight high-priority river systems. The review also
documented the relevant evidence cited in support of the conceptual models and the environmental
flow recommendations, and incorporated more recent evidence that adds to our conceptual
understanding. The draft was circulated amongst members of the Scientific Panels involved in the
environmental flow studies and to other scientists of high international standing in stream ecology and
management for comment.
12

Monitoring and evaluation of environmental flow releases in the Campaspe River
This process resulted in a very large collection of conceptual models and potential hypotheses to be
tested (Appendix 2). We applied a number of criteria to select the groups of conceptual models and
hypotheses that would be addressed both for the Campaspe River and at the Statewide level.
Conceptual models and hypotheses were selected according to the criteria that they should:
1. Be scientifically ‘sound’ – well-founded and supported by appropriate theoretical or empirical data
from scientific studies and/or expert opinion;
2. Involve responses to recommended environmental flows that will be detectable - expected
responses must be of sufficient magnitude to be detectable within a useful management
timeframe (nominally 10 years);
3. Address questions that are relevant to the Victorian River Health Strategy (VHRS);
4. Where possible, have general applicability to multiple reaches within Victorian rivers receiving an
Environmental Water Reserve;
5. Be realistic, given the quantity of water available for implementing the recommended
environmental flow releases;
6. Be targeted to components of the flow regime that can most feasibly be returned to a more natural
pattern using the environmental flow recommendations;
7. Acknowledge potential constraints on ecosystem response because of river-specific
characteristics and/or regulation activities (e.g. cold water releases from large dams);
8. Make use of available relevant historical data;
9. Address potentially adverse outcomes associated with implementing the recommended
environmental flow releases (e.g. blackwater events).
The individual conceptual models assembled by the review were stratified into groups (e.g. fish
responses, geomorphological responses) from which we chose a number of groups to develop into
synthesized conceptual models. A subjective rating system was applied to each group for criteria 1-3
above. The details of this process and the final groups of models chosen for development are outlined
in Appendix 2.
The various groups of conceptual models were synthesized into diagrammatic representations
supported by descriptions using the information from the review, and on occasions other information
from literature sources (detailed in each section below). More inclusive models were developed for
geomorphic responses, biochemical responses, habitat, macroinvertebrates, fish, and aquatic and
riparian vegetation. These conceptual models were collectively reviewed in a workshop (30th March
2006) attended by scientific experts and senior DSE and CMA staff (see Acknowledgements).
Adjustments to the conceptual models were made following feedback from this workshop. A decision
was taken at this time to drop biochemical responses from further consideration due to a lack of
current knowledge about the ecological implications of specific values for measures of production and
respiration, which could make it difficult to set targets. In addition, the macroinvertebrate model was
subsumed into the habitat model because all the important environmental requirements for
macroinvertebrates should be provided in an environment where sufficient habitat exists (see § 2.4). In
general, the conceptual modelling approach to the development of the monitoring and evaluation
program was strongly endorsed by workshop attendees.
In the sections below we describe the final conceptual models, presenting each diagrammatically, and
with a comprehensive verbal description supported by references. The models lead to the
13

Monitoring and evaluation of environmental flow releases in the Campaspe River
identification of possible monitoring variables, which are identified within each of the conceptual model
sections, and then summarised in a table in § 2.8 along with recommendations for monitoring
methods. The full range of possible variables has been presented, and will thus need to be refined for
each river system. In identifying response variables for monitoring, we have been guided by the
following considerations:
1) Relevance – variable must be demonstrably linked to components in the conceptual model;
2) Responsiveness - variable should respond to the planned intervention at the spatial and time
scales of interest;
3) Reliability – variable can be measured in a reliable and reproducible way;
4) Interpretability – what does it mean with respect to the issue of concern? Can one obtain
meaningful interpretations that are useful for drawing inferences, making decisions and/or
reporting? For example, some multivariate measures are difficult to interpret and their practical
value may be limited;
5) Cost effectiveness.
2.3 Geomorphic Processes
2.3.1 Description of Conceptual Model
It is likely that many channels have not fully adjusted to the altered flow and sediment regimes
introduced through regulation over the last century. For this reason we must consider the geomorphic
effects of additional environmental flows in the context of on-going channel adjustment, not a change
imposed on channels at equilibrium. Flow modifications in Victorian regulated rivers generally occur at
two main locations along the channel. An initial stage of modification occurs at the reservoir, normally
located in the upland section of the river. Flow may be diverted directly from the reservoir but is more
often released in the dry season and diverted once the river emerges onto areas of floodplain farming.
The effect of flow regulation is attenuated downstream of the reservoir by unregulated tributary
inflows.
The historic effect of irrigation and urban water supply reservoirs on geomorphic processes and the
downstream floodplain and channel geometry will depend on release policy, catchment physiography,
distance downstream of the reservoir and time since construction of the reservoir (Knighton 1998).
Limited understanding of these effects makes generalization of these responses difficult. However,
there is likely to be a general response of sediment starvation immediately downstream of the
reservoir as a consequence of sediment trapping in the reservoir (Zone 1 in Figure 3). In Zone 1 we
would expect erosion, channel enlargement, bed degradation, and armouring. The extent of sediment
starvation will depend on the size of the reservoir relative to inflow volumes (the sediment trapping
efficiency) and the impact of flow release policy on the flood regime. Consequent downstream erosion
will be inhibited by resistant boundary material and in particular bedrock. As erosion progresses, the
bed will become armoured thus slowing the process of erosion. Bed degradation downstream of dams
has also been reported to reduce channel gradients and hence erosive forces. In most cases, high
flow pulses released in Zone 1 will either have no effect or enhance these erosional processes.
Further downstream, unregulated tributaries with a higher-gradient than the main channel may deliver
sediment loads that exceed the transport capacity of the main channel. In Zone 2 we would expect
development of in-channel bars and slow development of benches leading to bed aggradation and
channel narrowing. Additional high flows released as part of an environmental flow may impede or
14

Monitoring and evaluation of environmental flow releases in the Campaspe River
reverse these trends by scouring sediments from the channel bed and promoting bank erosion and
channel widening.
Further downstream in Zone 3, large lowland sediment stores will tend to buffer effects of altered
upstream sediment regime and responses will be slow and difficult to detect. Flow diversions from the
midland or lowland river could result in reduced flows in the lowland channel leading to sedimentation
of pools and narrowing through development of benches. High flow freshes in this section will flush
fine sediments deposited in pools and promote bank erosion which can off-set the narrowing effects of
bench development.
Figure 3. Conceptualisation of the change in sediment and flow regimes downstream of dams.
The X-axis represents distance downstream from the reservoir, with zone 1 being immediately
downstream. Solid line: regulatied flow conditions; dashed line: with environmental flow
allocation.
Petts & Gurnell (2005) present a thorough review of the effects of dam on channel morphology and
point out that there can be great variation in the rate of response as a consequence of variable
sediment loads (relative to sediment transport capacity), variable erosion resistance of the channel
bed and banks and variable rates of growth and colonization by riparian vegetation (which has the
potential to stabilize sediment deposits). Figure 4 conceptualizes the geomorphic responses to flow
regulation adapted from the above discussion and models presented in Petts & Gurnell (2005). Habitat
responses are related to changes in channel bed level, width and stability.
The main drivers of morphological change are flow and sediment and these are represented in the
third level of Figure 4 by (i) flood magnitude, (ii) duration and frequency of high in-channel flows and
(iii) total sediment load. In a regulated river, these will be a function of unregulated tributary flows and
the influence of upstream impoundments and diversions. These are represented in the first row of the
diagram. As one moves downstream of a dam, the flow and sediment regime is modified through the
effects of catchment physiography including the location of confluences. These effects are
represented by the box labelled “valley geometry” although this is likely to be better represented as a
multi-dimensional effect of valley relief and the proportion of catchment upstream of impoundments.
15

Monitoring and evaluation of environmental flow releases in the Campaspe River
The “response times” box refers to the rate of responses to flow regulation. Fast response rates are
hypothesized to increase with sediment supply (relative to transport capacity) which is a function of
flow regime, sediment load and channel slope (closely related to valley slope although slightly
modified by sinuosity). Rapid colonisation of sediment deposits by riparian vegetation and reduced
bed and bank resistance will also promote rapid adjustment in depositional and erosional phases
respectively.
The types of adjustment and subsequent effects on physical habitats are described at the bottom of
the diagram. In Zone 1, one would expect channel widening, bed degradation and a reduction in the
frequency of geomorphically significant events. In Zone 2, changes would typically include bed
aggradation, bench development and eventually channel narrowing. With reduced frequency of high
flow events, one might expect a reduced frequency of geomorphically significant events although
oversupply of sediments at confluences may offset this to some extent. With increased geomorphic
stability, one would expect reduced complexity of the bed and bank morphology. In such situations,
the geomorphic objective of environmental flows is commonly to promote geomorphic events (which
involve both erosion and deposition of bed and bank material) to increase channel complexity or
specific habitats associated with complex channels. Such habitats might be deeper pools, bed
sediments free of fine particulates, benches of different heights and billabongs formed through
meander cut-offs. These habitats are all products of natural erosion and deposition processes in
dynamic channels and can be adversely affected by regulation.
2.3.2 Selected Sub-Hypotheses & Response Variables to be Monitored
Based on this conceptual model, we propose that key endpoints to be monitored in the VEFMAP are:
(i) changes in channel geometry (i.e. channel width, depth and complexity); (ii) changes in channel
alignment (i.e. rates of bank erosion and deposition on benches); and (iii) changes in the frequency of
geomorphologically significant events (i.e. frequency of events during which bed sediments are
redistributed or there is development of meander bends and benches) (Table 2). The first two aspects
can be monitored using periodic re-survey of the channel. The third requires targeted and ongoing
monitoring. These monitoring tasks are covered by the channel features and channel dynamics
surveys outlined in Table 6, which includes information on variable definitions, sample timing and
sampling protocols.
We recommend monitoring in carefully selected reaches associated within Zone 2 in Figure 3. There is
little hope of returning natural geomorphic processes to Zone 1 using environmental flows because the
fundamental problem is sediment starvation. Responses of river morphology to environmental flow
releases in Zone 3 will be slow (decades to centuries) and probably not detectable within a practical
management timeframe. Zone 2 represents the zone in which the intended geomorphic effects of
environmental flows may be detectable. Zone 2 begins at the first major input of sediment to the
regulated river downstream of the dam. This zone extends downstream to either (a) the base of an
alluvial fan forming where the river flows onto an unconfined floodplain or (b) the estuary.
16

Monitoring and evaluation of environmental flow releases in the Campaspe River
Figure 4. Conceptual model of geomorphic responses to flow regulation.
17

Monitoring and evaluation of environmental flow releases in the Campaspe River
It is probably best to target monitoring in river reaches at the upstream end of Zone 2 to have the
greatest chance of detecting responses to environmental flows in the short term. Further downstream,
the effects will be attenuated and may proceed at a slower rate. Where there is a second major point
of flow regulation (e.g. a diversion weir), it may be prudent to locate further geomorphic monitoring
below this point if it is in Zone 2. Site location will need to consider the proximity of upstream
tributaries supplying sediment. There is likely to be a dynamic zone downstream of tributary
confluences, with increased channel complexity associated with depositional features. Ideally,
monitoring would be undertaken downstream of the immediate zone of influence of these tributaries.
Table 2. Geomorphic conceptual model - summary of subhypotheses and corresponding
variables associated with the various flow components.
Flow
Component
Sub-hypotheses Variables
Winter-
Spring
Freshes
In Zone 2 reaches, does increased
frequency of winter-spring fresh events:
a) increase the frequency of
geomorphologically significant events
(e.g. redistribution of bed and bank
sediments)?
b) increase channel complexity (e.g.
areas of the stream bed which are flushed
free of fine deposits, deeper pools and
variability in bench elevations)?
c) increase channel width and depth?
d) increase rates of meander
development (i.e. bank erosion on the
outside bank, point bar development,
increased sinuosity and eventually bend
cut-off and billabong formation)?
Frequency of channel disturbances
Frequency of bed disturbances
Rate of bench deposition
Bed complexity
Bench development and variability
Mean channel top width, cross-section
area and thalweg depth
Bank erosion on outside of meander
bends
Point bar development
Bankfull
Flows
As for winter-spring freshes As for winter-spring freshes
Overbank
Flows
As for winter-spring freshes As for winter-spring freshes
2.3.2.1 Complementary Research Issues
A non-exhaustive list of complementary research issues that have been identified as being relevant to
the conceptual model are as follows:
a) Surveying extent of bed armouring downstream of dams and monitoring changes through time;
b) Effect of riparian vegetation on channel narrowing downstream of dams;
c) Effect of tributary sediment load on channel form and habitat complexity downstream of tributaries
in regulated rivers.
18

Monitoring and evaluation of environmental flow releases in the Campaspe River
2.4 Habitat Processes & Macroinvertebrates
2.4.1 Description of Conceptual Model
Physical factors of ecological significance include stream-flow, current velocity, channel shape, water
depth, substrate, and temperature and water quality indicators such as dissolved oxygen (DO)
concentration, salinity and pH. Habitats are defined by a complex interaction of physical factors and
the ecological requirements of aquatic flora and fauna, such as light, shelter, food and flow-mediated
chemical exchanges.
The conceptual model is for a reach (ranging from 10s to >100 km in length) in a regulated mid-to-
lowland river. It focuses on hypotheses relating to the maintenance of hydraulic habitat for vegetation,
invertebrates and fish during the summer-autumn low flow period and habitat creation associated with
the reinstatement of more natural levels of winter-spring base flows and patterns of freshes (Figure 5).
The effects of other flow components (i.e. summer high flows, bankfull and overbank flows) are less
understood from the point of view of habitat creation and maintenance, but have other known positive
ecosystem-level effects, which are discussed for the other conceptual models. The individual
conceptual models for vegetation and fish spawning and recruitment (presented later) provide greater
detail on flow relationships and ecological requirements at various stages in the life histories of these
groups.
Habitat patches in rivers are formed by interactions of hydrology, geomorphological features (e.g.
pools, runs, bars, benches, overhanging banks and anabranches) and structural elements (e.g.
boulders, tree roots, coarse woody debris and macrophytes). These habitat patches are dynamic and
respond to various characteristics of the flow regime. For instance, freshes can create new habitat
patches through inundation where none existed previously and can alter the nature of a habitat patch
from a pool to a run. The persistence of habitat patches depends on the temporal characteristics of the
flow regime (e.g. timing, duration, frequency and variation of various flow features).
Provision of adequate levels of summer-autumn low flows and freshes helps to maintain adequate
depth and water quality in permanent pools, riffle/run sections and shallow water areas as well as
provide longitudinal connectivity. The combination of high temperatures, high evaporation rates and
lower flows over summer can lead to a decrease in surface area and volume of surface water and
increases in extremes of physicochemical water quality parameters such as dissolved oxygen (DO)
concentration. Loss of continuous surface water may lead to the drying of riffle/run sections and
shallow areas to the detriment of biota dependent on these habitats (e.g. macrophytes, invertebrates
and larval and juvenile fish). Furthermore, loss of lateral and longitudinal connectivity may affect
processes such as drift which may be a necessary step in the life history of some macroinvertebrate
species and fish. Adequate levels of low flows should maintain shallow water areas, trickling flows
over riffle/run sections and adequate depth for maintaining connectivity and passage of biota to
alternative habitats.
During periods of drying, permanent pools are critical for faunal persistence by providing a refuge in
which individuals can survive to recolonize when streams reconnect (Boulton 2003, Magoulick &
Kobza 2003). The provision of summer-autumn low flows and freshes of adequate magnitude,
frequency and duration helps to replenish and maintain water quality in permanent pools, through
such mechanisms as dilution of salt and nutrients and the increase of DO by mixing/aeration.
Improved water quality may increase the number and diversity of pollution-sensitive taxa. Such
freshes may also re-establish temporary connectivity along the stream channel allowing for the
dispersal of mobile organisms to alternative habitats.
19

Monitoring and evaluation of environmental flow releases in the Campaspe River
In systems with seasonal flow inversion, constant high water levels in summer can alter the distribution
and reduce the area and persistence of shallow and slow water patches favoured by some in-channel
macrophytes and larval, juvenile and small-bodied fish.
‘Slackwater’ areas, typically small shallow areas of still water formed by sand bars, woody debris and
bank morphology, have also been found to contain many more larval and small-bodied fish than
flowing water patches in lowland rivers. In their study involving hydraulic manipulations to create
slackwater and flowing water patches, an order of magnitude more fish and shrimp were collected
from slackwaters, both created and natural (Humphries et al. 2006). Slackwaters have been
hypothesized as providing refuge from current (and therefore energetic advantages) for the young
stages of fish and shrimp and/or predation and as sites where food is abundant (Humphries et al.
1999, 2006, King 2004a, b, Richardson et al. 2004). These hypotheses are largely untested. However,
some workers have suggested that refuge from current may be a more important factor than food
availability in explaining why slackwaters are favoured. Several studies have found evidence that it is
energetically advantageous for fish larvae to inhabit still or low-velocity patches (Flore & Keckeis 1998,
Matthews 1998, Flore et al. 2001). King (2004a) has also shown that the density of benthic meiofauna,
seemingly an important food source for larval fish, was not different between still and flowing habitats
in the Broken River. Similarly, Humphries et al. (2006) found that the density of benthic meiofauna and
zooplankton did not differ consistently between flowing and slackwater patches. They concluded that
the greater abundance of fish and shrimp in created slackwater patches and lower abundance in
created flow patches could not be explained by the density of potential prey. They did however, find a
significant difference in the community composition of benthic meiofauna between slackwater and
flowing patches and also found that slackwater patches had a greater amount of benthic organic
matter – a potential food resource for shrimp (Burns & Walker 2000).
Reinstatement of more natural levels of winter-spring baseflows will produce a sustained increase in
channel depth over the winter-spring period. The provision of winter-spring freshes will produce
temporary additional increases in channel depth that will vary depending on the magnitude and
duration of the fresh. Increase in channel stage height will increase the volume of pool habitat. Where
the increase in stage height results in inundation of physical structures and features (e.g. in-channel
macrophytes, channel-edge macrophytes, tree roots, woody debris, branch piles, in-channel bars,
benches and overhanging/undercut banks), it makes these substrates available for colonisation and
attachment by invertebrates. When inundated these physical features provide important habitat for
feeding, shelter, current refuge and spawning, and represent an increase in the quantity, diversity and
complexity of physical habitat for both invertebrates and fish (Crowder & Cooper 1982, O’Connor
1991, Crook & Robertson 1999).
Freshes during winter and spring may lead to the flushing of fine sediments and organic mater from
areas of coarse streambed substrate (e.g. riffles). This flushing reduces armouring of the stream bed
and leads to greater availability of interstitial spaces in the coarse substrate. These spaces are
available as habitat for invertebrates. It is widely assumed that flushing flows as described here will
have beneficial effects for macroinvertebrate assemblages in riffle environments, but there has been
little investigation as to whether this is the case (P.S. Lake, pers. comm.). Thus at present, the link
between sediment flushing and invertebrate response remains as a hypothesis to be tested through
monitoring.
Reinstatement of more natural levels of winter-spring baseflows increases the overall area of shallow
and slow water within the channel. Increases in flow velocity can increase the area of riffle/run habitat,
and together with the increase in shallow and slow water habitat and pool habitat, contributes to an
overall increase in the quantity and diversity of different hydraulic habitat types.
20

Monitoring and evaluation of environmental flow releases in the Campaspe River
21
Shallow and slow water areas provide suitable conditions for the growth of aquatic macrophytes. Run
areas also provide a site for production of some species that thrive in elevated velocities (eg, some
species of Vallisneria and Myriophyllum, M.J. Kennard, pers. comm.). Hence, the increased availability
of shallow and slow waters, as well as runs is expected to lead to increased abundance of
macrophytes over the main growing season in spring. Macrophytes are widely recognized as
important habitat for some invertebrates and fish (Crowder & Cooper 1982; Minshall 1984; Humphries
1996). Macrophytes are a direct and indirect source of food (e.g. from epiphytic periphyton growing on
them and organic detritus found at their base), provide protection from predators and current, and
increase the amount of available habitat per unit area of substrate (Crowder & Cooper 1982; Minshall
1984; Newman 1991; Weatherhead & James 1991). Seasonal growth and increased abundance of in-
channel macrophytes with a range of growth-forms creates additional physical substrate and
contributes to the increase in the quantity and diversity of in-stream physical habitat.
Slow water habitats produce higher densities of planktonic microinvertebrates due to increased
residence time (Ferrari et al. 1989, Pace et al. 1992, Basu & Pick 1996, King 2004a). Hence,
increased availability of slow water habitat can be expected to increase the abundance of
microinvertebrates. Diversity in flow velocities per se may also be important for macroinvertebrate
community diversity because moderate-fast current velocity riffle/run habitats may support species
specialised for those conditions. While this seems reasonable and plausible, there is uncertainty about
the link between flow velocity habitat and its influence on invertebrate community diversity/condition
(L. Metzeling, pers. comm.), and existing studies suggest complex interactions with other biological
processes (e.g. Hart & Finelli 1999).
An increase in the quantity and diversity of habitats should increase assemblage diversity by allowing
aquatic organisms with different habitat requirements to coexist. For example, the complexity of
macroinvertebrate communities and the abundance of individual species have been correlated with
habitat complexity created by woody debris, macrophytes and organic debris, coarse substrates and
substrate stability (Schlosser 1982; O’Connor 1991; Cobb et al. 1992, Pusey & Kennard 1996; Bond &
Lake 2003; Pusey & Arthington 2003; Pusey et al. 1993, 1995, 1998, 2000, 2004).
Direct knowledge of the effects of environmental flows on macroinvertebrate assemblages is poor.
However, given the links between habitat diversity and species diversity, potential macroinvertebrate
responses to flow manipulation are captured by the conceptual model for habitat processes. The
expected responses of macroinvertebrates to environmental flows have been presented only in terms
of the expected effects of flow on macroinvertebrate habitat, and the potential subsequent effects on
assemblages.

Monitoring and evaluation of environmental flow releases in the Campaspe River
Flow
Reinstatement of more natural
levels of Winter-Spring Baseflow
Increase in
Inchannel
Shallow & Slow
Water Area
Increase in Abundance
of Submerged &
Amphbious
Macrophytes
Increase in Area
of Riffle and/or
Run Habitat
Winter-Spring
Freshes
Increase in Inchannel
Flow Velocity
Fish Immigration to Reach
Sustained Inundation of
Inchannel Macrophytes,
Channel Edge
Macrophytes,Tree Roots,
Branch Piles, Woody
Debris, Inchannel Bars,
Overhanging/Undercut
Banks
Increase in Channel
Stage Height
Temporary
Inundation of Higher-
level Channel Edge
Macrophytes, Tree
Roots, Branch Piles,
Woody Debris, Bars,
Benches,
Overhanging/
Undercut Banks
Invertebrate
Community
Structure
Increase in Fish
Recruitment
Fish Assemblages &
Population Structure
Increase in Quantity &
Diversity of Flow
Velocity Habitats
Increase in Quantity
& Diversity of
Physical Habitat
Increase in
Invertebrate
Abundance
Increase in
Micro-
invertebrate
Abundance
Summer-Autumn Low
Flows & Freshes
Maintain Inchannel
Shallow & Slow
Water Area & WQ
Maintain Adequate
Area, Depth & WQ
in Riffle/Run Habitat
Replenish & Maintain
Adequate Volume,
Depth & WQ in
Permanent Pools
Maintain
Connectivity
Maintain Habitat
for Submerged &
Amphibious
Macrophytes
Maintain Shallow &
Slow Water Habitat for
Invertebrates & Fish
Maintain ‘Refuge’
Habitat to Ensure
Persistence of
Invertebrates & Fish
Allow Passage to
Alternative Habitats
Increase in
Volume of
Pool Habitat
Flush Accumulations of Fine
Sediments & Organic Matter from
Coarse Streambed Substrate
Increase Availability of
Interstitial Spaces in Coarse
Streambed Substrate
Figure 5. Conceptual model of habitat processes. Abbreviation used: WQ = water quality.
22

Monitoring and evaluation of environmental flow releases in the Campaspe River
We have deliberately chosen this type of conceptual model for two reasons. First, by focussing on
habitat, the causal network for macroinvertebrates is kept relatively simple. Simple conceptual models
stand a far better chance of being developed into predictive numerical models in the future. Second,
we believe that the flow characteristics necessary to fulfil the habitat requirements for
macroinvertebrates will, as a general rule, lead to the provision of other resources necessary for the
maintenance of assemblages. For instance:
• Water quality may affect macroinvertebrates. The provision of adequate flows, particularly freshes,
during periods of low flow should maintain water quality at adequate levels.
• Freshes that lead to bench inundation may stimulate the hatching of some macroinvertebrate
species that produce resting stages in these previously dry environments.
Higher flows will lead to the introduction of organic matter into the stream system, stimulating primary
productivity and providing food resources for macroinvertebrates.
There are certainly other linkages that could be made between flows and effects on
macroinvertebrates. We believe however, that there are no important linkages that will occur
independently of those flow characteristics that will provide adequate macroinvertebrate habitat. This
approach to modelling macroinvertebrate response to flows does not address the question of where
colonists are to come from. It may be that macroinvertebrates in heavily flow impacted systems take
some time to respond to a change in flow regime because of the need for passive dispersal of
colonists from less impacted tributaries, which may be well upstream of the target reaches.
With respect to fish assemblages and fish population structure, an increase in the quantity and
diversity of physical habitat and flow velocities habitat, as well as an increase in food supply due to
increased invertebrate abundance is predicted to increase survival of young fish and consequently
recruitment of fish within the reach. Increased habitat and food resources are also likely to increase
the health and survival of adult fish, and may also lead to an increase in fish immigration to the reach.
Both factors will impact on the fish community structure within the reach. It is noted that many fish
species are very territorial, so the availability of habitat resources may be a more influential limiting
factor on fish community dynamics than the availability of food resources (P.S. Lake, pers. comm.).
2.4.2 Selected Sub-Hypotheses & Response Variables to be Monitored
First, the monitoring program should determine whether the implemented flows deliver the expected
habitat features. Then, monitoring should attempt to determine whether implemented flows result in
the expected ecological responses (Table 3). The achievement of the expected abiotic features not
accompanied by the expected ecological response may indicate gaps or errors in our conceptual
understanding of the system, or the presence of other limiting factors. The relevant monitoring field
program, variable definitions and sampling timing and sampling protocols are presented in Table 6.
23

Monitoring and evaluation of environmental flow releases in the Campaspe River
Table 3. Habitat processes and macroinvertebrates conceptual model - summary of
subhypotheses and corresponding variables associated with the various flow components.
Flow
Component
Subhypotheses Variables
Summer-
Autumn Low
Flow &
Freshes
Do implemented environmental flows
maintain in-channel shallow and slow
water area?
Shallow and slow water area
Do implemented environmental flows
maintain adequate area and depth of
at least 0.1 m in shallow, slow water
and riffle/run habitats?
Riffle/Run depth and area
Do implemented environmental flows
maintain adequate volume and depth
in permanent pools?
Permanent pool depth and volume
Do implemented environmental flows
maintain connectivity?
Connectivity
Do implemented environmental flows
maintain macroinvertebrate
community structure?
Number of invertebrate families index
AUSRIVAS score
SIGNAL biotic index
EPT biotic index
Presence/Absence and number of ‘flow-
sensitive’ taxa
Do implemented environmental flows
increase fish recruitment?
See conceptual model for Fish Spawning &
Recruitment
Do implemented environmental flows
maintain fish assemblages and/or
population structure?
Fish species composition
Relative abundance of adult/sub-adult native
and exotic fish species
Population structure and size class
distribution of native and exotic fish species
Winter-
Spring
Baseflows
Do implemented environmental flows
increase in-channel shallow and slow
water area?
Shallow and slow water area
Do implemented environmental flows
increase area of riffle and/or run
habitat?
Riffle and/or Run area
Do implemented environmental flows
increase volume of permanent pool
habitats
Permanent pool depth and volume
Do implemented environmental flows
result in sustained inundation of in-
channel macrophytes, channel edge
macrophytes, tree roots, woody
debris, branch piles, in-channel bars,
overhanging or undercut banks?
Inundation of representative physical habitat
features
Do implemented environmental flows
increase abundance of macrophytes?
See conceptual model for Aquatic and
Riparian Vegetation
Cover of submerged and amphibious species
in Zone A
Do implemented environmental flows
improve macroinvertebrate
community structure?
Number of invertebrate families index
AUSRIVAS score
SIGNAL biotic index
EPT biotic index
Presence/Absence and number of ‘flow-
sensitive’ taxa
Do implemented environmental flows
improve fish assemblages and/or
population structure?
Fish species composition
Relative abundance of adult/sub-adult native
and exotic fish species
Population structure and size class
distribution of native and exotic fish species
24

Monitoring and evaluation of environmental flow releases in the Campaspe River
Winter-
Spring
Freshes
Do implemented environmental flows
increase area of riffle and/or run
habitat?
Riffle and/or Run area
Do implemented environmental flows
increase volume of pool habitats
Permanent pool depth and volume
Do implemented environmental flows
result in temporary inundation of
higher-level channel edge
macrophytes, tree roots, woody
debris, bars, benches,
overhanging/undercut banks?
Inundation of higher elevation representative
physical habitat features
Do implemented environmental flows
improve macroinvertebrate
community structure?
Number of invertebrate families index
AUSRIVAS score
SIGNAL biotic index
EPT biotic index
Presence/Absence and number of ‘flow-
sensitive’ taxa
Do implemented environmental flows
improve fish assemblages and/or
population structure?
Fish species composition
Relative abundance of adult/sub-adult native
and exotic fish species
Population structure and size class
distribution of native and exotic fish species
The physical dimensions and water quality parameters at shallow and slow water, riffle/run, and pool
habitats will show short term responses to implemented environmental flows. At this stage we are not
proposing specific monitoring of water quality parameters at permanent pools, shallow and slow water
and run habitats. The question of whether water quality in these microhabitats during the summer-
autumn low flow period is sufficiently different to that measured at VWQMN stations (and therefore
warrants dedicated monitoring) is flagged as a complementary research issue. With regard to this, we
propose that priority be given to investigating water quality in permanent pools because they have a
greater potential for developing water quality problems (particularly at depth) due to temperature (and
in some rivers, salinity) stratification.
Detecting responses of macroinvertebrates to flow augmentation at the scale of a river reach is made
difficult by the high levels of small-scale spatial and temporal variability that characterize
macroinvertebrate assemblages (Rosenberg & Resh 1993, Downes et al. 2000, 2006). In addition,
species-level identifications require advanced taxonomic skills, and quantitative sampling is time-
consuming with most samples being dominated by a few numerically abundant species. In response
to the difficulties of sampling, various standardised rapid-sampling protocols have been developed
(e.g. Barbour et al. 1999, Growns et al. 1995, 1997). These protocols are often characterised by
presence/absence sampling (rather than quantitative), and by the identification of macroinvertebrates
to coarser taxonomic levels, usually Family. Presence/absence sampling precludes the possibility of
analysing changes in the abundance of individual species in response to flow augmentation. It also
reduces the amount of information that can be incorporated in multivariate analyses. Similarly,
collecting taxonomic data at the family level prevents examination of species-specific effects of
environmental flows. Some authors have argued that the effects of large-scale interventions may be
more easily detected by analyses conducted at coarser taxonomic levels than for data collected at
species level (e.g. Warwick 1988a, b). They hypothesise that species are likely to respond to fine
scale natural environmental differences, but changes at coarser taxonomic levels are more likely to be
seen in response to larger-scale ‘treatments’. Data for marine benthic infauna support this hypothesis
(Warwick 1988a, b). In freshwater studies, data collected at family level give a similar picture of
community composition to those collected at species level (Marchant et al. 1995, Hewlett 2000).
25

Monitoring and evaluation of environmental flow releases in the Campaspe River
There are marked cost advantages to employing a rapid sampling protocol. Many more samples can
be processed for same investment of time and money, and the level of training required of the
operators is far less. We support the use of the EPA rapid bioassessment protocol (EPA Victoria
2003). The only exception to this would be extra sampling effort designed to pick up ‘flow-sensitive’
taxa. One response to flow augmentation may be the immigration of flow-sensitive species (L.
Metzeling, pers. comm.). Such taxa could be used as a simple bioindicator of the success or otherwise
of flow augmentation. The first requirement is for a list of such taxa to be developed through expert
consultation. With regards to subsequent monitoring, it will be necessary to invest some extra training
and effort in the laboratory identification of flow-sensitive taxa, as these taxa are likely to require
identification to species level, rather than family-level as specified in the EPA rapid bioassessment
protocol (EPA Victoria 2003).
Data collected using rapid sampling protocols are most amenable to analysis using either multivariate
statistics, or some form of index that collapses the multivariate data to some univariate measure.
Because of the nature of community level data (many species, many low or zero counts), non-
parametric multivariate analyses (Clarke 1993) are usually appropriate. Results may be visualised as
ordinations of the multivariate data. Such results are problematic from the point of view of target-
setting and inference of ‘improvement’. While it is possible to track an assemblage through time in an
ordination, we cannot from the ordination alone infer whether or not the assemblage condition has
improved, merely that is has (or has not) changed. The only solution is to define a target assemblage
(i.e. species list) and to include this in the multivariate analysis as another ‘sample’. It can then be
determined whether the actual assemblage is moving towards this state through time. Defining a
target assemblage for lowland rivers, however, would be very difficult (G.P. Quinn, pers. comm.).
Multivariate analyses are, however, powerful visual tools for communication of findings, and are very
sensitive in picking up changes in assemblages. There are also a number of indices that present
community-level data as a univariate response (e.g. AusRivAS; Simpson et al. 1997, SIGNAL;
Chessman 1985, RivPACS, Wright et al. 1985). Built into these indices is some assumption about
what constitutes a ‘better’ assemblage, and a higher index score implies an improvement. Thus it is
possible to infer improvement in an assemblage from the index scores alone, and target setting is also
possible. These indices are also often amenable to analysis by commonly-used statistical methods.
2.4.2.1 Complementary Research Issues
A non-exhaustive list of complementary research issues that have been identified as being relevant to
the conceptual model are as follows:
a) Is water quality in permanent pools, shallow and slow water and run habitats sufficiently different
to that measured at VWQMN stations to justify dedicated monitoring, and how does relative water
quality respond to environmental flows?
b) Role of permanent pools as refuge habitat ensuring persistence and allowing recolonization
c) Connectivity characteristics for upstream and downstream movement of macroinvertebrates and
fish for movement to alternative habitats. When is this connectivity most important?
d) From the original environmental flows documents, it was the general consensus of opinion that the
direct effects of flow augmentation on macroinvertebrate assemblages are likely to be minor
compared to other potential restoration actions, such as the introduction of large woody debris to
the system. This assertion could be investigated
26

Monitoring and evaluation of environmental flow releases in the Campaspe River
2.5 Aquatic and Riparian Vegetation
2.5.1 Description of Conceptual Model
Streams and their riparian zones are nonequilibrium ecosystems that provide habitat for a wide range
of plants with a variety of adaptations (Nilsson & Svedmark 2002). The flow regime interacts with river
geomorphic features to produce different seasonal patterns of availability and persistence of free
water in the different portions of a channel. These temporal and spatial patterns of wetting and drying
are a major determinant of plant community structure, floristics and dynamics.
The conceptual model is for a reach (ranging from 10s to >100 km in length) in a regulated mid- to
lowland, unconfined river. The conceptual model concentrates on the processes of habitat
maintenance, seasonal growth as well as plant establishment from seed germination in response to
flow manipulation. It focuses on the main growing seasons of spring and summer, when day length is
longer, which maximises the energy available for photosynthesis, and temperature is higher, which
maximises rates of physiological processes (Roberts et al. 2000). It depicts predicted responses of key
attributes such as spatial distribution of measures of community structure and floristic composition, as
well as system viability in the form of regeneration. We do not distinguish between native and exotic
species, as the major effects predictable with current knowledge apply to functional groups (see
below) rather than individual species. We expect that native and exotic species within the same
functional group will respond similarly to a given flow regime. Responses are predicted to (a) stable
flow conditions (spring and summer baseflows) as well as (b) fluctuating flow conditions (spring-
summer freshes and bankfull flows) (Figure 6).
There is a consistent and distinct transverse zonation about streams as vegetation responds to
transverse environmental gradients such as soil moisture conditions. Following Christie & Clark
(1999), three channel zones may be defined. These zones are:
• Zone A: from mid-channel to stream margin (or the area covered by water during times of
baseflow)
• Zone B: from stream margin to a point mid-way up the flank of the bank (or the area that is
infrequently inundated)
• Zone C: from mid-way up the flank of the bank to just beyond the top of the bank
Following Brock & Casanova (1997), three main groups of plant species may be distinguished
according to the amount of free water in which species grow. These groups are described as
‘terrestrial’, ‘amphibious’ and ‘submerged’. The terrestrial group may be further split into species that
germinate, grow and reproduce in either ‘dry’ or ‘damp’ places (Casanova & Brock 2000). The
amphibious group includes species found throughout the wet-dry ecotone and may be further divided
into ‘fluctuation-responders’ and ‘fluctuation-tolerators’. Fluctuation-responders germinate in flooded
conditions, grow in both flooded and damp conditions using their ability to alter their growth pattern or
morphology in response to the presence or absence of water and reproduce above the water surface
(Casanova & Brock 2000). This group includes some floating-leaved species (e.g. Nymphoidea spp.)
and species with some degree of morphological plasticity (e.g. Myriophyllum spp.) (Brock & Casanova
1997). Fluctuation-tolerators germinate in damp or flooded conditions; tolerate variation in water levels
without major changes in growth or morphology, and some species may reproduce above the water
surface (Casanova & Brock 2000). This group includes low-growing species (e.g. Hydrocotyle spp.)
27

Monitoring and evaluation of environmental flow releases in the Campaspe River
and emergent species (e.g. Eleocharis spp., Persicaria spp. and Typha spp.). The submerged group is
not split and includes species that germinate, grow and reproduce under water (e.g. Vallisneria spp.).
Alternating wet and dry periods within a flow regime can affect seed bank germination and
establishment. For instance, wet periods modify oxygen availability in the soil; mediate decomposition
of organic matter and subsequent concentrations of nutrients and toxic substances; stimulate or inhibit
germination; suppress terrestrial or flood-intolerant plants; and alter the light regime depending on
turbidity and depth of inundation (Casanova & Brock 2000). Dry periods following flooding may
desiccate and kill submersed species, but may also stimulate or inhibit germination depending on a
species’ ability to respond to or tolerate fluctuations in flooding and drying (Casanova & Brock 2000).
The following sections draw heavily on the work of Casanova & Brock (2000) who used experiments
to investigate the relative importance of depth, duration and frequency of inundation on plant
community germination from wetland seed banks. They found that duration, frequency of flooding and
depth all affected plant community development in some way, but concluded that duration of
individual inundation events was the major determinant of plant community composition. Frequency
and depth exerted a secondary influence which further differentiated plant community composition.
This study was relatively short-term compared to the time frame of response to environmental flows of
some species (e.g. River Red Gum), but provides the best indication available of the types of
responses that might be seen in early life history stages, and also of responses of terrestrial species to
inundation. Casanova & Brock’s (2000) main experimental results are summarized below.
Under a continuously flooded treatment (16 weeks of inundation), the resultant plant community was
found to be dominated by submerged species and some amphibious fluctuation-responder species.
Overall species richness was low.
Long inundation events lasting more than two weeks at a time produced a plant community dominated
by amphibious fluctuation-responder species, although some terrestrial ‘damp’ species also occurred.
Overall species richness and biomass was also low.
Shorter inundation events lasting less than two weeks produced a species-rich plant community
dominated by amphibious fluctuation-responder, fluctuation-tolerator and terrestrial ‘damp’ species. A
similar suite of functional groups resulted when short inundation events occurred at a higher frequency
(i.e. more than twice within 16 weeks), but the biomass of terrestrial ‘damp’ species was lower under
the regime of increased inundation frequency. In addition, Casanova & Brock (2000) found that water
depth during short and frequent inundation events determined dominance within the amphibious
group, with fluctuation-tolerator species favoured at shallower depths and fluctuation-responder
species favoured at greater depths.
When the seed bank was wetted and maintained in a damp condition without flooding and drying, the
resultant plant community was dominated by terrestrial ‘damp’ and ‘dry’ species, but amphibious
fluctuation-tolerator species were also present. Of the various treatments, this resulted in the highest
species-richness, but most species had low biomass.
In summary, duration of inundation events determined what combination of submerged, amphibious or
terrestrial species germinated and established. Protracted inundation tends to favour germination,
establishment and growth of submerged species although some amphibious fluctuation-responder
species are also expected. Shorter durations of inundation allow amphibious species to germinate and
establish and the brief periods of anoxia are tolerable by terrestrial ‘damp’ species. Frequency governs
the length of dry phases between inundation events and influences the relative survival and growth of
28

Monitoring and evaluation of environmental flow releases in the Campaspe River
species in the terrestrial and amphibious groups. Long, dry intervals enable terrestrial ‘damp’ species
to establish, flourish and attain high biomass. Whereas more frequent inundation, implying shorter dry
intervals, might favour amphibious fluctuation-responder and fluctuation-tolerator species over
terrestrial ‘damp’ species. Depth affects light levels and the ability of emergent species to reach the
surface and is an important determinant of plant community composition when water levels are stable.
However, in the context of a fluctuating water regime, depth alone is less important than duration and
frequency of inundation in influencing plant community composition (Casanova & Brock 2000).
This conceptual model does not consider the contribution of vegetative reproduction and dispersal
processes to plant community establishment and development. In this restricted context, the
composition of resulting plant communities is dependant on the composition and viability of the seed
bank present in the soil. Different flow regimes can select for different plant communities, but only if
the seed bank contains the potential for different plant communities to develop (Casanova & Brock
2000). The seed bank of a species-poor site might only allow the development of a community of
limited species richness and diversity. The relative proportion of exotic plant species to total plant
species will also depend on the composition and viability of the seeds present in the seed bank, as
well as differences in the capacity of exotics to exploit (or alternatively, to withstand) the prevailing
wetting-drying regime in the relevant channel zone.
During spring baseflows, increasing temperatures and shallow, slow water and run areas in Zone A
provide favourable conditions for germination and seasonal growth of submerged and some
amphibious fluctuation-responder macrophyte species. Run areas also provide a site for some species
that thrive in elevated velocities (eg. some species of Vallisneria and Myriophyllum, M.J. Kennard,
pers. comm.). Where regulation activities have reduced the magnitude of spring baseflows,
reinstatement of more natural (i.e. higher) flows should increase the amount of shallow and slow water
and run habitat within the river channel for submerged and some amphibious fluctuation-responder
species. If the river channel is geomorphically complex, the provision of spring-summer baseflows and
freshes that sustain inundation of geomorphic features in Zone A (e.g. channel bed, low-lying bars and
benches, channel edges, runners and anabranches), will increase the quantity of damp or flooded
substrate and ensure reliable water supplies for the germination and growth of submerged and
amphibious fluctuation-responder species. This may be reflected in an increased number of
submerged and amphibious fluctuation-responder species and better growth performance such as
greater height, more projected foliage cover and greater stem densities of plants in Zone A.
In addition to the above effects, increased baseflows will affect terrestrial ‘dry’ species (including
agricultural weeds) in the channel. Terrestrial ‘dry’ species that may have encroached into Zone A
during dry periods lack the physiological adaptations needed to survive sustained inundation and are
expected to drown and dieback. Sustained inundation will also inhibit germination of these species.
These conditions should result in lower species richness and biomass of terrestrial ‘dry’ species in
Zone A.
Spring freshes may inundate higher-elevation geomorphic features in Zone A such as in-channel bars
and islands as well as benches, runners and anabranches in Zone B. Long spring freshes that last for
more than two weeks at a time may be expected to result in a plant community dominated by
amphibious fluctuation-responder species and some terrestrial ‘damp’ species within the areas
inundated.
Short, infrequent (e.g. one or two) spring freshes that last for less than two weeks at a time and are
separated by long dry intervals may result in a species-rich plant community dominated by amphibious
fluctuation-responder, fluctuation-tolerator and terrestrial ‘damp’ species within the areas inundated. If
short freshes occur with greater frequency, a similar combination of amphibious fluctuation-responder,
29

Monitoring and evaluation of environmental flow releases in the Campaspe River
fluctuation-tolerator and terrestrial ‘damp’ species would be expected, but the biomass of terrestrial
‘damp’ species would likely be lower.
Spring freshes and bankfull flows may wet geomorphic features in Zones B and C (e.g. high-level
benches, upper banks, runners and anabranches) dampening the soil without flooding. This may
result in:
a) germination and establishment of mainly terrestrial ‘damp’ and ‘dry’ species but also some
amphibious fluctuation-tolerator species
b) growth pulse in terrestrial species including riparian seedlings and saplings
c) increased vigour in the canopy of terrestrial trees and shrubs in the riparian zone.
Provision of summer low (base) flows helps to maintain the area and water quality of shallow and slow
water and run habitats for submerged and amphibious species through the remainder of the growing
season. Provision of summer freshes results in wetting of Zone A geomorphic features such as in-
channel bars, low-lying benches, channel edges, runners and anabranches. This may alleviate
desiccation of in situ macrophytes that have become exposed as the channel stage height dropped
over summer. It may also result in improvement in the condition of the canopy of riparian trees and
shrubs that are adjacent to the wetted channel zones.
In systems with seasonal flow inversion, extended periods of artificially high late-spring and summer
flows can alter the distribution and reduce the area and persistence of shallow and slow water habitats
for macrophytes. Increased summer flow magnitude also often means deeper and colder water, which
may translate to poorer growing conditions for submerged macrophytes (J. Roberts, pers. comm.).
Shear stress associated with high velocity flows also increase the risk of mechanical damage to
plants, of parts breaking off, and of emerging or floating leaves being dragged underwater, reducing
rates of photosynthesis and consequently growth (Madsen et al. 2001). Elevated summer flows with
little variation in water level also means a shift in the inundation pattern of geomorphic features in
Zone A (and possibly B), from transitory inundation to prolonged inundation. This may favour
germination and establishment of submerged and amphibious fluctuation-responder species over
terrestrial ‘damp’ species. The likely consequence is an altered plant community composition with
lower species richness.
30

Monitoring and evaluation of environmental flow releases in the Campaspe River
Figure 6. Conceptual model of aquatic and riparian vegetation responses to Spring and
Summer flows. Zone A: from mid-channel to stream margin (or the area covered by water
during times of baseflow); Zone B: from stream margin to a point mid-way up the flank of the
bank (or the area that is infrequently inundated): Zone C: from mid-way up the flank of the bank
to just beyond the top of the bank.
31

Monitoring and evaluation of environmental flow releases in the Campaspe River
To investigate the effects of flow manipulations on aquatic and riparian vegetation, it is proposed that
variables relating to community structure, floristic composition and regeneration be monitored in each
channel zone.
2.5.2 Selected Sub-Hypotheses & Response Variables to be Monitored
First, the monitoring program should determine whether the implemented flows deliver the expected
habitat features. Then, monitoring should attempt to determine whether implemented flows result in
the expected ecological responses (Table 4). The achievement of the expected abiotic features not
accompanied by the expected ecological response may indicate gaps or errors in our conceptual
understanding of the system, or the presence of other limiting factors. The relevant monitoring field
program, variable definitions and sampling timing and sampling protocols are presented in Table 6.
Table 4. Aquatic and riparian vegetation conceptual model - summary of subhypotheses and
corresponding variables associated with the various flow components.
Flow
Component
Subhypotheses Variables
Spring
Baseflow
Do implemented environmental flows
increase in-channel shallow and slow water
area?
Shallow and slow water area
Do implemented environmental flows
increase run area?
Run depth and area
Do implemented environmental flows result
in sustained inundation of channel bed,
channel edges, in-channel bars, low-lying
benches, runners and anabranches in Zone
A?
Inundation of geomorphic features in Zone A
Do implemented environmental flows
a) increase germination and seasonal
growth of submerged and amphibious
fluctuation-responder species in Zone A?
b) reduce species richness of terrestrial ‘dry’
species in Zone A?
Cover of submerged and amphibious species in
Zone A
Species composition, number of submerged,
amphibious and terrestrial species in Zone A
Proportion of exotic plant species
Spring
Freshes
What is the pattern of inundation and drying
in Zones A & B imposed by the implemented
environmental flows?
What is the composition of the resultant
plant community?
Cover of amphibious and terrestrial species in
Zones A & B
Species composition, number of amphibious and
terrestrial species in Zones A & B
Proportion of exotic plant species
Spring
Freshes &
Bankfull
Flows
Do implemented environmental flows wet
high-level benches, upper banks, runners
and anabranches in Zones B & C?
Wetting of geomorphic features in Zones B & C
Do implemented environmental flows
increase germination and establishment of
terrestrial ‘damp’, terrestrial ‘dry’ and
amphibious fluctuation-tolerator species?
Species composition, number of amphibious and
terrestrial species in Zones B & C
Proportion of exotic plant species
Germination of seedlings of overstorey and
midstorey species
Do implemented environmental flows
improve canopy condition of in situ riparian
trees and shrubs?
Canopy condition
Summer
Baseflows
Do implemented environmental flows
maintain area of in-channel shallow and
slow water and run habitats?
See conceptual model for Habitat Processes
Shallow and slow water area
Run depth and area
Summer
Freshes
Do implemented environmental flows wet in-
channel bars, low-lying benches, channel
edges, runners and anabranches in Zone A?
Wetting of geomorphic features in Zone A
Do implemented environmental flows
improve canopy condition of adjacent
riparian trees and shrubs?
Canopy condition
32

Monitoring and evaluation of environmental flow releases in the Campaspe River
2.5.2.1 Complementary Research Issues
A non-exhaustive list of complementary research issues that have been identified as being relevant to
the conceptual model are as follows:
a) Minimum conditions necessary to drown and affect dieback of invasive terrestrial agricultural weed
species in channel zones A and B.
b) Variation in growth pulse response in riparian seedlings and saplings to spring-summer fresh
events of differing frequency, magnitude and duration.
c) The conceptual model focussed on seed germination responses to the spring freshes. However, it
may also be important to understand the relative importance of vegetative reproduction
(regeneration from pre-existing plant parts such as rhizomes, stolons and turions) on the
maintenance and recruitment of aquatic and riparian vegetation. It is possible that regeneration
from rhizomes and other plant parts may be similar to seed germination responses, but this
comparison has apparently not yet been done (J. Roberts, pers. comm.). This question represents
a knowledge gap which warrants further investigation.
d) Another potential area of investigation is to understand the role of winter-spring floods and the
consequences of their absence on aquatic and riparian vegetation (J. Roberts, pers. comm.).
However, this conceptual model was not articulated in any of the environmental flow study reports
and there is little ecological research that addresses this issue (J. Roberts, pers. comm.). Hence
this is flagged here as another knowledge gap that needs to be addressed.
2.6 Native Fish – Spawning & Recruitment
2.6.1 Description of Conceptual Model
The conceptual model is for a hypothetical regulated lowland river that contains both diadromous and
non-diadromous species. This distinction should be borne in mind when using the model. It focuses on
the pre-spawning habitat and spawning and recruitment requirements of diadromous and non-
diadromous fish.
River fishes are typically seasonal in their breeding habits, with temperature and flow being the two
major factors that dictate when fish spawn (Humphries et al. 1999). Fish tend to spawn during the
warmest months of the year, partly because rates of egg, embryo and larva development are
positively correlated with temperature, and partly because food for larvae and juveniles is most
abundant at this time of the year in temperate systems (Jobling 1995). Rates of growth and
development are critical to larval survival because the longer an individual spends as a highly
vulnerable larva, the greater the cumulative risk of predation by larger fish (Jobling 1995). Also the
larger a fish is, the greater its swimming ability and potentially, the greater its ability to forage and to
avoid predators (Margulies 1990, Bone et al. 1995). In temperate systems, high temperature and
flooding may only be weakly correlated. In cases where high temperatures and flooding do not
coincide, temperature will often be the dominant variable that influences the timing of spawning
33

Monitoring and evaluation of environmental flow releases in the Campaspe River
(Humphries et al. 1999). There is a great variety in the spawning styles of Australian freshwater fish
species.
Humphries & Lake (2000) note that river regulation can affect fish via reproduction and recruitment.
The effects of river regulation on reproduction include the removal of appropriate conditions for gonad
maturation, migrations, pre-spawning interactions and spawning (Humphries & Lake 2000). River
regulation may also impact upon recruitment when it decouples the production of larvae and the
environmental conditions needed to sustain them until they become juveniles. This may include
desiccation or dispersal of eggs and/or larvae and the loss of hatching and/or rearing habitat
(Humphries & Lake 2000). Reproduction effects are manifested as a failure to spawn or an absence of
viable eggs and larvae following spawning. Recruitment effects imply that larvae are produced but do
not survive to become juveniles (Humphries & Lake 2000).
Environmental flow recommendations have been developed to address some aspects of the
reproductive and recruitment effects of river regulation, and these have been incorporated into the
different sections of the conceptual model (Figure 7). They are briefly explained in the following notes.
High flows in autumn and early winter (represented here as freshes and bankfull flows) are required to
trigger spawning in diadromous fish species such as galaxiids, eels and Australian Grayling (Koehn &
O’Connor 1990). Following spawning, high flows (represented here as freshes) within the same
period, are required to transport the larvae to the sea/estuary (Koehn & O’Connor 1990). This requires
that sufficient flows for larval transport be maintained along the entire river length between the
spawning zones and river mouth.
Reinstatement of more natural levels of baseflows as well as freshes in winter and early spring will
increase the overall quantity and diversity of in-stream habitat and food resources (see conceptual
model on Habitat Processes) and provide favourable feeding and growth conditions for adult fish.
These conditions may be important for physiological preparation and conditioning of adults prior to
spawning. We note that while this is widely believed to be true and considered to be appropriate, this
assumption has not been tested (A.J. King, pers. comm.).
The occurrence of spring/early summer bankfull or overbank flows may be a direct spawning trigger
for some non-diadromous fish species such as Golden Perch and Silver Perch, which appear to be
‘flood-specialist’ species (Lake 1967). If the adult fish have had the benefit of pre-spawning
conditioning in favourable habitats over the winter-spring period, they may produce a greater number
of larvae. Bankfull flows, while not leaving the main river channel, may result in the inundation of low-
lying runners and anabranches. Overbank flows are expected to inundate low-lying runners and
anabranches as well as floodplain areas including billabongs and wetlands. These areas become
slackwater habitat (velocity < 0.01 m/s) after recession of the bankfull or overbank flows and are highly
productive environments in spring/early summer. They provide suitable hatching, rearing, feeding and
refuge environments for non-diadromous larvae. The bankfull and overbank flows also lead to
increased habitat within the channel and the introduction of nutrients to the system, with resulting
effects on primary productivity, and in turn, food resources for larval fish. In general terms, the
spring/summer bankfull and overbank flows can be expected to result in greater numbers of fish larvae
of all breeding strategies recruiting into the population of juvenile fish (King et al. 2003a).
The reinstatement of more natural levels of spring and early-summer baseflows will help to maintain
in-stream habitats for adult fish and slackwater habitat for larval fish. Low flow conditions in spring,
summer and autumn may also provide favourable spawning conditions for ‘low flow specialists’ such
as Carp Gudgeons and Gambusia (King et al. 2003a) as well as generalist species such as Australian
34

Monitoring and evaluation of environmental flow releases in the Campaspe River
Smelt and Flathead Gudgeon. Humphries et al. (1999) noted that spawning and recruitment into
juvenile stocks of Flathead Gudgeon, Australian Smelt, Crimson-spotted Rainbowfish and three
species of Carp Gudgeon can occur in mid-summer when the prospect of flooding is remote but the
predictability of high temperatures and low flows are high.
Slackwater areas have also been found to contain many more fish than flowing water patches in
lowland rivers. Humphries et al. (2006) collected an order of magnitude more fish and shrimp from
slackwaters, both created and natural. King (2004b) found widespread use of natural slackwaters by
the larvae and juveniles of most species of fish, irrespective of whether they were limnophilic or
rheophilic as adults. Humphries et al. (2006) also reported that recent work (Price, unpublished data)
has indicated that the abundance of fish larvae increases with slackwater area.
Slackwaters have been hypothesized as providing refuge from current for the young stages of fish and
shrimp and/or predation and as sites where food is abundant (Humphries et al. 1999, 2006; King
2004a, b, Richardson et al. 2004). Some workers believe that mortality due to starvation is highest
during the critical stages of first feeding, when larvae are poorly developed and have limited mobility
(Bone et al. 1995). Low flow conditions may concentrate appropriately sized prey in sufficient densities
and slackwater habitats tend to have higher densities of microinvertebrates due to increased water
residence time (Ferrari et al. 1989, Pace et al. 1992, Basu and Pick 1996, King 2004b). These
conditions may increase the likelihood of larvae surviving through the critical stages of first feeding.
There is also evidence that in energetic terms, it is advantageous for fish larvae to be associated with
still or low-velocity patches (Flore & Keckeis 1998, Matthews 1998, Flore et al. 2001) , presumably
because the expend less energy to retain their position in the water column.
Although there remains debate over whether refuge from current or food availability is more important
in explaining the preference of larvae and young fish for slackwaters, the importance of slackwaters as
rearing, feeding and refuge habitats for larvae and young fish appears to be well-supported. The
maintenance of adequate baseflows throughout spring and summer to maintain essential slackwater
habitats is therefore expected to increase the number of fish completing the larval stage.
In reaches directly downstream of major regulatory structures, excessive rates of rise during the
spawning season may lead to the loss of fish eggs and larvae. In addition, in systems with seasonal
flow inversion, constant high water levels in summer can alter the distribution and reduce the area and
persistence of slackwater habitats, thereby potentially affecting the recruitment success of fish
(Humphries et al. 2006).
35

Monitoring and evaluation of environmental flow releases in the Campaspe River
Figure 7. Conceptual model for fish spawning and recruitment into the juvenile population.
36

Monitoring and evaluation of environmental flow releases in the Campaspe River
2.6.2 Selected Sub-Hypotheses & Response Variables to be Monitored
First, the monitoring program should determine whether the implemented flows deliver the expected
habitat features. Then, monitoring should attempt to determine whether implemented flows result in
the expected ecological responses (Table 5). The achievement of the expected abiotic features not
accompanied by the expected ecological response may indicate gaps or errors in our conceptual
understanding of the system, or the presence of other limiting factors. The relevant monitoring field
program, variable definitions and sampling timing and sampling protocols are presented in Table 6.
Table 5. Native fish spawning and recruitment conceptual model - summary of subhypotheses
and corresponding variables associated with the various flow components. Note that
hypotheses concerning adult fish are generally dealt with under the Habtat Processes
concentual model.
Flow
Component
Subhypotheses Variables
Autumn-early
Winter
Freshes/Bankfull
Flows
Do implemented environmental flows
trigger spawning of diadromous fish?
(Only relevant in river reaches inhabited
by diadromous fish species such as
galaxiids, eels and Australian Grayling)
Presence/Absence of diadromous fish
larvae
Winter-Spring
Baseflows and
Winter-Spring
Freshes
Do implemented environmental flows
increase overall quantity and diversity of
instream habitat?
See conceptual model for Habitat
Processes
• Shallow and slow water area
• Run area
• Permanent pool depth and
volume
• Inundation of in-channel
physical habitat features
• Inundation of higher elevation
physical habitat features
• In-channel and littoral cover of
macrophytes
Spring-early
Summer
Bankfull Flows
Do implemented environmental flows
inundate low-lying runners and
anabranches to create increased
slackwater habitat?
Area of slackwater habitat in runners
and anabranches
Do implemented environmental flows
increase the number of fish completing
larval stages?
Density of post-larval fish
Spring-early
Summer
Overbank Flows
Do implemented environmental flows
inundate low-lying runners and
anabranches to create increased
slackwater habitat?
Area of slackwater habitat in runners
and anabranches
Do implemented environmental flows
inundate floodplain areas to create
increased slackwater habitat?
Area of slackwater habitat in floodplain
Do implemented environmental flows
provide appropriate conditions for
spawning and larval production of ‘flood
specialist’ non-diadromous fish species?
Presence/Absence of ‘flood specialist’
non-diadromous fish larvae
Do implemented environmental flows
increase the number of fish completing
larval stages?
Density of post-larval fish
Spring-early
Summer
Do implemented environmental flows
provide appropriate conditions for
spawning and larval production of ‘low
Presence/Absence of ‘low flow
specialist’ and generalist fish larvae
37

Monitoring and evaluation of environmental flow releases in the Campaspe River
Baseflows
Spring-early
Summer
Baseflows (cont)
flow specialist’ and generalist fish
species?
Do implemented environmental flows
maintain adequate instream habitat for
adult and larval fish?
See conceptual model for Habitat
Processes
• Shallow and slow water area
• Run area
• Permanent pool depth and
volume
• Connectivity
Do implemented environmental flows
increase the number of fish completing
larval stages?
Density of post-larval fish
Summer-
Autumn Low
Flows
Do implemented environmental flows
maintain adequate instream habitat for
adult and larval fish?
See conceptual model for Habitat
Processes
• Shallow and slow water area
• Run area
• Permanent pool depth and
volume
• Connectivity
Do implemented environmental flows
increase the number of fish completing
larval stages?
Density of post-larval fish
2.6.2.1 Complementary Research Issues
A non-exhaustive list of complementary research issues that have been identified as being relevant to
the conceptual model are as follows:
a) Larval transport for diadromous fish. What levels of flow are required to successfully transport
diadromous fish larvae to estuarine/marine habitats? This has important implications for the
allowable level of extraction in the downstream reaches of major rivers.
b) Role of winter-spring baseflows and freshes in creating conditions conducive for pre-spawning
conditioning and pre-spawning interactions for adult fish. As noted above, it is widely assumed,
but not known that pre-spawning conditions are an important determinant for the success of
spawning, and therefore recruitment. Pre-spawning conditions will vary between sites and years,
and such data on conditions coupled with spawning in recruitment data may reveal the importance
(or otherwise) of pre-spawning conditions.
c) ‘Condition factor’ may be useful as an indicator of ‘fish health’ based on length/weight ratios.
However, basic research is required to test the utility of this index, and to develop standards for
individual species.
d) What is the impact of exotic fish species on native fish assemblages, and how is this affected by
environmental flows?
38

Monitoring and evaluation of environmental flow releases in the Campaspe River
2.7 Water Quality
In designing the monitoring program, we have sought to link flow-related effects back to ecosystem
endpoints through conceptual models. Water quality per se has not been identified as an endpoint to
be improved through the provision of environmental flows, and is more likely to be seen as a link
between flow and a given ecosystem endpoint.
However, improved water quality is frequently cited as one of the goals in the various environmental
flow reports – notwithstanding the fact that the improved water quality is often linked to some
ecological effect. Moreover, there was support at the inaugural meeting of the VEFMAP
Implementation Committee to include water quality monitoring as a program in its own right.
The reasons for including water quality can be summarised as follows:
• Water quality is already well-accepted by regulators and the community as an indicator of
environmental health through such programs as Waterwatch, the VWQMN, and through the
inclusion of water quality variables in the State Environment Protection Policy (SEPP; EPA 2001a,
b).
• We expect that water quality variables will react to environmental flow events more quickly than
ecological variables. This possibility alone makes water quality a strong candidate for the
VEFMAP program.
• Water quality variables do not show the same level of small-scale unexplained variation that is
typical of many ecological variables. Thus it should be easier to demonstrate a beneficial effect of
flows on water quality.
• Water quality could affect most of the ecosystem endpoints outlined above, and knowledge of the
effects of environmental flows on water quality may help to improve models of ecosystem
response to flow events.
Inclusion of water quality in the monitoring program falls outside of the protocol laid down to select
other endpoints for monitoring. Hence, no conceptual model has been developed for the effects of
different flow components on water quality, and no sub-hypotheses from environmental flow reports
have been identified. The inclusion of water quality has been justified on other grounds, and it would
be circular logic to go back and justify its inclusion using the same process as for the other endpoints.
A standard water quality monitoring program broadly of the type employed at VWQMN stations would
be sufficient to address whether overall water quality changes in response to environmental flows. The
program already instituted in the Glenelg River to the monitor the effects of environmental flows on
water quality (SKM 2006c) should be used as a guide, although this program was designed with null-
hypothesis testing of water quality responses to flow in mind, which contrasts our own
recommendations to quantify flow-response relationships within a Bayesian framework (see below).
The number of parameters to be monitored would be less than that of VWQMN sites, and these are
detailed in Table 6. If a VWQMN site is located close to a chosen environmental flow monitoring site,
the data from that site may be used to reduce costs. It is likely, however, that additional water quality
monitoring data will need to be collected at some environmental flow monitoring sites in order to give
sufficient coverage of the reaches.
Because water quality monitoring has not been justified based on a conceptual model, there are no
criteria for including it or excluding it based on reach specific characteristics. We recommend
39

Monitoring and evaluation of environmental flow releases in the Campaspe River
monitoring water quality in all reaches, and do not discuss it individually within the reach by reach
recommendations. Conceptual refinements of the VEFMAP program in the future may lead to
revisions of the recommendation.
2.7.1.1 Complementary Research Issues
Small-scale spatial patchiness in water quality is poorly understood. As stated above, we do not know
whether water quality in slackwater areas and pools will be of different quality to that in the main
channel. Results from the Glenelg River (SKM 2006c) suggest that water quality in pools differs from
the main channel when pools were > 3 m deep. Whether this result will hold for other rivers and years
is unknown. It would be a relatively simply research project to determine whether a single monitoring
point per site can effectively represent the water quality for the whole site.
Also not addressed by the current recommendations for water quality monitoring is the question
monitoring specifically tailored to particular flow assess the effects of particular flow components. Two
potential negative effects of environmental flows on water quality are blackwater events and saline
fronts (see § 2.10). In order to assess whether environmental flows are causing such events, fine
temporal-scale monitoring around specific flow components will be required. This is a different
research project to the standard water quality monitoring being recommended here.
2.8 Summary of Variable Definitions & Sampling Timing and Protocols
We have identified six field programs (in addition to monitoring flow), with several of these consisting
of several subcomponents. The programs are delineated by background colour in Table 6. These
programs cover the conceptual models, and associated hypotheses and variables developed for the
statewide program. In addition, we have provided definitions of variables, recommendations for the
timing and frequency of sampling, and sampling protocols for each variable (Table 6). River-specific
monitoring programs will be based on a selection of these hypotheses and associated variables.
We recognise several types of variables in Table 6. Flow-related effects are seen as the primary
drivers of response in the rivers, and are labelled as such. Other variables will also influence
responses in the rivers, and these may be built into models as covariates that explain why some sites
react differently to similar flows. Some of these covariates will be collected as part of the field
programs. Other covariates will relate to data collected in the field programs but will need to be
collected independently. This latter group of covariates appear at the end of Table 6, colour coded
according to the field program they are most relevant to. We have considered two main types of
responses: intermediate endpoints and endpoints. The intermediate endpoint classification recognises
the central role that habitat is expected to play in mediating ecological response to flow releases.
Although habitat may not be the only requirement for the ecological response to be realised, there is a
strong expectation that no response will be seen unless there is an improvement in habitat attributable
to the environmental flow. It is necessary to monitor habitat to see if this requirement is being met.
Habitat data may also then be built into ecological response models as covariate data as described
above. Some ecological endpoints (such as in channel macrophytes) may also be intermediate
endpoints (and covariates) for other ecological responses (e.g. macrophytes act as macroinvertebrate
habitat). Thus the classification of variables is not intended to be mutually exclusive, and there is likely
to be discussion about the labels assigned. The labels below reflect our current opinion about the
most likely use for the data.
40

Monitoring and evaluation of environmental flow releases in the Campaspe River
41
2.9 Refining the Conceptual Models and Variables for the Campaspe River
The conceptual models presented above and the potential monitoring variables presented in Table 6
are generic in that they have not been specifically developed for the Campaspe River. In designing the
detailed monitoring program for the Campaspe, contractors will need to take account of details in the
reach-by-reach monitoring recommendations (§ 3.3).

Monitoring and evaluation of environmental flow releases in the Campaspe River
Table 6. Summary of field programs, variable definitions and sampling timing, frequency and protocols. Background colour refers to the field
program with which the individual variables are associated. Cells on multiple rows have been merged where the same description applies to more
than one variable. The words (cont. below) indicates that the cell breaks onto the next page.
Field
Program
Type Variable Definition & Measure Timing & Frequency Sampling Protocol
Driver Flow Timing, magnitude, frequency and
duration of flow components (base
flows, freshes and bankfull flows)
Continuous Obtainable from gauging records
Channel Dynamics
Endpoint Frequency of
channel
disturbances
Geomorphically significant events
are defined as events where bed
or bank sediments are mobilised.
Geomorphic events will be
identified by increased turbidity in
the main channel associated with
increased flows in the main
channel. This monitoring
technique is novel and may
require some development
following the first year of
monitoring.
Continuous sampling Continuous monitoring of turbidity in the
main channel and upstream tributary
using automated turbidity sensors. The
magnitude of the event will be measured
relative to baseflow turbidity rather than
absolute values eliminating concerns of
“drift” in turbidity observations.
Observations of turbidity in the upstream
tributary are required to remove effects of
tributary inputs from observed fluctuations
in turbidity within the main channel.
Endpoint Frequency of bed
disturbances
A bed disturbance is defined as a
flow event during which there is
movement of coarser bed
sediments. This and the following
endpoint (rate of sediment
deposition on benches) should be
related to the frequency of
geomorphically significant events.
After the passage of
Winter-Spring freshes or
bankfull or overbank flows
Monitoring of painted lines on exposed
point bars (recorded in photographs)
Endpoint Rate of bench
deposition
In an alluvial channel we would
expect there to be erosion of bank
material on the outside bank of
meander bends and deposition of
finer fraction of these bank
sediments in slackwater areas
such as benches. Thus deposition
on benches is a surrogate
measure of bank erosion rates
and channel dynamics.
After the passage of
Winter-Spring freshes or
bankfull or overbank flows
Sample mass of sediment deposited on
sediment mats (e.g. artificial turf mats)
placed on bench surfaces. The number of
replicate mats to be used will depend on
bench area and size of sediment mats.
Sediment may either be removed by
water using a high pressure washer or
through careful brushing after drying
(Steiger et al. 2003, Vietz et al. 2006).
42

Monitoring and evaluation of environmental flow releases in the Campaspe River
Channel Features Survey
Endpoint Bed complexity Bed complexity can be
characterised by analysis of the
longitudinal bed profile
surveyed along the channel
thalweg (Bartley & Rutherfurd
2005).
Once every 5 years
Channel survey: cross-sectional and
longitudinal bed profiles. Survey
should use at least 15 permanently
marked cross-sections surveyed to a
fixed datum. The monitoring site
should include at least one full
meander wavelength.
Endpoint Bench
development and
variability
Bench development can be
measured by the increased cross-
sectional area occupied by
benches. Variability in benches
can be measured by variability in
their elevation relative to bankfull
and/or water surface at the time of
survey.
Endpoint Mean channel top
width, cross-
section area and
thalweg depth
Channel size is characterised by
the top width, depth at the thalweg
and the total cross-section area.
Appropriate and consistent means
should be used to identify the
bankfull level.
Bank erosion on
outside of
meander bends
Bank erosion on the outside of
meander bends are a surrogate
measure for the rate of meander
development. This measured by
re-survey of cross-section profiles
at meander bends.
Endpoint Point bar
development
Point bar development on the
inside of meander bends is a
surrogate measure for the rate of
meander development.
Field
Program
Type Variable Definition & Measure Timing & Frequency Sampling Protocol
43

Monitoring and evaluation of environmental flow releases in the Campaspe River
Habitat Field Survey (Repeated)
Intermediat
e endpoint Shallow and
slow water area Shallow and slow water areas
are defined as areas where the
depth is between 0.1-0.3 m and
the velocity is < 0.1 m/s.
e.g. m2 of shallow and slow
water area per m2 of stream
channel area
During Summer-Autumn
low flows and Winter-
Spring baseflows
Field survey at 3-4 stages over the
range of baseflows typically
encountered during Summer-Autumn
and Winter-Spring baseflows. The
relevant variable (i.e. shallow and slow
water area, riffle/run area and
permanent pool depth and volume) is
interpolated from these surveys.
Survey from channel bank towards the
channel centre until depth exceeds 0.3
m or velocity exceeds 0.1 m/s for
shallow and slow water area.
Riffle/run areas can be identified
visually
Intermediate
endpoint Riffle/Run depth
and area
Riffle areas are defined as regions
with coarse bed material and
shallow, fast-flowing water.
Run areas are defined as regions
with low to moderate laminar flow
and smooth, unbroken water
surface.
e.g. m2 of riffle/run area per m2 of
stream channel area
Intermediate
endpoint Connectivity Maximum channel depth in
shallow cross-sections
During Summer-Autumn
low flows and freshes
Covariate Size class
distribution of
streambed
substrate
e.g. % cover of substrate size
categories where substrate size
categories include:
• Bedrock
• Boulder >256 mm
• Cobble 64-256mm
• Pebble 16-64mm
• Gravel 2-16mm
• Sand 0.06-2mm
• Silt/Clay <0.06mm
Annually, during periods of
low flow
Visual rating using EPA Rapid
Bioassessment protocol (EPA Victoria
2003) substrate size categories along with
% cover categories
Covariate Organic matter Refers to fine and coarse organic
material (e.g. leaf-packs, twigs,
branch piles, root masses etc)
e.g. % cover of organic material
per m2 of stream channel area
Visual rating using % cover categories.
See for example, Anderson & QNRM
(2003) and EPA Victoria (2003).
Field
Program
Type Variable Definition & Measure Timing & Frequency Sampling Protocol
44

Monitoring and evaluation of environmental flow releases in the Campaspe River
Field
Program
Type Variable Definition & Measure Timing & Frequency Sampling Protocol
Habitat Field Survey (Repeated) (cont.)
Covariate Woody debris
loading
e.g. Woody debris (≥ 0.1 m in
diameter) loading per unit stream
area (e.g. volume loading such as
m3 of woody debris per m2 of
stream channel area or surface
area loading such as m2 of woody
debris per m2 of stream channel
area)
Once every 5 years, during
periods of low flow
We recommend quantifying woody debris
loading using a combination of the
census method (Gippel et al. 1996) and
the line-intersect method (van Wagner
1968, Wallace & Benke 1984). The
census method entails measurement of
the diameter and length of every piece of
WD above a predetermined threshold size
(e.g. min. 0.1 m diameter, min. 1 m
length) within a given stream area. The
line-intersect method involves recording
the diameter of every piece of WD (above
the predetermined threshold size)
intersected by a transect of given length.
The census method gives accurate,
repeatable estimates of woody debris
loading, but is intensive and time-
consuming. The line-intersect method is
much quicker and provides consistent,
repeatable estimates, but may grossly
overestimate the actual loading (Marsh et
al. 1999). Both methods should be used
in the initial survey to calibrate the line-
intersect method (see Marsh et al. 1999
for full details). Woody debris loading can
then be estimated in subsequent (repeat)
surveys using the line-intersect method.
Habitat Field
Survey
(Post-Event)
Intermediate
endpoint Area of
slackwater habitat
in runners and
anabranches
Slackwater areas are defined as
areas where the velocity is <0.01
m/s. In practice, these are areas
where the flow is imperceptible.
e.g. m2 of slackwater area per m2
of stream channel/floodplain area
Following Spring-early
Summer bankfull or
overbank flows
Same protocols as for quantifying
‘Shallow and slow water area’ and
‘Riffle/Run area’
Intermediate
endpoint Area of
slackwater habitat
in floodplain
45

Monitoring and evaluation of environmental flow releases in the Campaspe River
Habitat Survey in Conjunction with One-dimensional Hydraulic Modelling
Intermediat
e endpoint Permanent pool
depth and
volume
Depth and volume of water in
selected permanent pools
within the monitoring site
During Summer-Autumn
baseflows and following
Summer-Autumn freshes
A stage recorder should be installed at
the downstream end of the monitoring
site. A rating curve should be
established for the downstream cross-
section based on measurements of
discharge or observations of
discharge at a nearby gauge. Cross-
sections should be surveyed using at
least 15 permanently marked cross-
sections surveyed to a fixed datum.
HEC-RAS or similar hydraulic model
should be used to model water surface
profiles along the monitoring site over
the range of in-channel flow.
Observations of stage at a number of
points along the channel should be
used to verify the model at a range of
flow magnitudes from low flows up to
bankfull. The model can then be used
to estimate the depth and volume of
water in selected permanent pools
within the monitoring site during
Summer-Autumn baseflows and
following Summer-Autumn freshes.
Intermediate
endpoint Inundation of
representative
physical habitat
features
Physical habitat features include
in-channel macrophytes, channel
edge macrophytes, tree roots,
branch piles, woody debris, in-
channel bars and overhanging or
undercut banks
During Winter-Spring
baseflows
A feature survey should be used to locate
representative physical habitat features
and geomorphic features (in channel
zones A, B and C) and establish their
elevation at the monitoring site.
The hydraulic model developed according
to the protocol described in the cell
directly above can then be used to
estimate the discharge at which physical
habitat and geomorphic features (in the
various channel zones) are inundated or
wetted.
Intermediate
endpoint Inundation of
higher elevation
representative
physical habitat
features
Higher elevation physical habitat
features include channel edge
macrophytes, tree roots, woody
debris, branch piles, bars,
benches and overhanging or
undercut banks
During Winter-Spring
freshes
Field
Program
Type Variable Definition & Measure Timing & Frequency Sampling Protocol
46

Monitoring and evaluation of environmental flow releases in the Campaspe River
Field
Program
Type Variable Definition & Measure Timing & Frequency Sampling Protocol
Habitat Survey
with 1D
modelling (cont.)
Intermediate
endpoint Inundation/
Wetting of
geomorphic
features in Zone A
Geomorphic features in Zone A
include channel bed, channel
edges, low-lying bars and
benches, runners and
anabranches
During Spring baseflows
and Summer freshes
Repeat feature surveys are required once
every three years as the distribution of
physical habitat features may be altered
by events such as large flow events, bank
slumping, tree fall and macrophyte
growth/dieback.
Intermediate
endpoint Wetting of
geomorphic
features in Zones
B & C
Geomorphic features in Zones B
& C include higher elevation bars
and benches, upper banks,
runners and anabranches
Macroinvertebrate Survey
Endpoint Number of
invertebrate
families
Number of invertebrate families Autumn & Spring in same
year OR
Spring & Autumn in
consecutive year
EPA Rapid Bioassessment protocol (EPA
Victoria 2003) with separate assessments
for riffle & edge habitats.
Must be based on data from sampling in
both Autumn & Spring.
Family-level sampling would provide cost-
savings over species-level sampling.
However, the environmental flows to be
implemented are likely to represent subtle
rather than substantial changes to the
current flow regime in many rivers. There
is uncertainty over whether
macroinvertebrate responses to minor
flow augmentation can be detected at the
family-level. Consequently, we
recommend sampling to species level for
an initial 3 years. The data should then be
reviewed to determine if it is necessary to
continue sampling to the species-level.
Endpoint AUSRIVAS score AUSRIVAS predicts the
invertebrates that should be
present in specific stream habitats
under reference conditions. By
comparing the number of
expected families with the number
of families actually found, a ratio
can be calculated for each test
site. This ratio is expressed as the
observed number of
families/expected number of
families (the O/E score).
Endpoint SIGNAL biotic
index
Index of water quality based on
the tolerance of aquatic biota to
pollution. Calculated by summing
together the sensitivity grades for
each of the families found at a site
that have been assigned a
sensitivity grade, and then dividing
by the number of graded families
present.
47

Monitoring and evaluation of environmental flow releases in the Campaspe River
Field
Program
Type Variable Definition & Measure Timing & Frequency Sampling Protocol
Macroinvertebrate Survey (cont.)
Endpoint EPT biotic index Total number of families in the
generally pollution-sensitive insect
orders of Ephemeroptera
(mayflies), Plecoptera (stoneflies)
and Trichoptera (caddisflies).
Calculated by summing together
the number of families in these
three orders present at a site.
May not be that useful in
lowland streams and rivers
where such taxa are rare.
See above See above
Endpoint Presence /
Absence and
number of ‘flow-
sensitive’ taxa
e.g. some leptophlebiid mayfly
species
EPA Rapid Bioassessment protocol +
identification of ‘flow-sensitive’ taxa.
Expert input required to generate list of
‘flow-sensitive’ taxa, possibly including
keys for identification of different life
history stages of these taxa.
Species-level data from the first three
years could be used to inform the creation
of the ‘flow-sensitive’ species list
Vegetation Survey
Intermediate
endpoint &
Endpoint
In-channel and
littoral cover of
macrophytes
Includes submerged and
amphibious (e.g. free-floating,
floating-leafed and emergent)
macrophytes
e.g. % cover of in-channel
macrophytes per m2 of stream
channel area
If assessing as Habitat
intermediate endpoint,
sample during late-Spring
baseflows
If assessing as Vegetation
endpoint, sample in late-
Spring for quick
responders, and late-
Summer for integrative
effects over whole growing
season
If assessing as Habitat intermediate
endpoint, visual rating of each functional
group using ordinal scales such as the
Braun-Blanquet % cover scale or % cover
categories will suffice (see Werren &
Arthington 2002).
If assessing as Vegetation endpoint,
quadrat sampling recommended.
Cover by species is desirable but requires
species-level knowledge. Cover by
growth-form (e.g. submerged, free-
floating, floating-leafed and emergent)
may be sufficient?
(cont. below)
Intermediate
endpoint &
Endpoint
Cover of
submerged and
amphibious
species in Zone A
Cover refers to the real proportion
of a horizontal plane at a given
height occupied by vegetation
biomass. A coarse indicator of the
amount of vegetation occupying
different canopy layers, but also a
measure of the quantity of canopy
available to absorb sunlight.
(cont. below)
48

Monitoring and evaluation of environmental flow releases in the Campaspe River
Field
Program
Type Variable Definition & Measure Timing & Frequency Sampling Protocol
Vegetation Survey (cont.)
Endpoint Cover of
amphibious and
terrestrial species
in Zones A & B
e.g. % cover of submerged
macrophytes per m2 of Zone A
area
Late Spring-early Summer
(preferably a fixed time
such as 7 weeks after the
passage of Spring freshes).
Standard protocols do not exist. Will need
to do some experimentation on minimum
sample sizes required.
Endpoint Species
composition,
number of
submerged,
amphibious and
terrestrial species
in Zone A
Assessment of species present at
a site provides information on the
vegetation’s structural and floristic
diversity and weediness.
Vegetation diversity is usually
interpreted as an indicator of the
community’s stability and capacity
to respond to disturbance
(Baldwin et al. 2005).
In late-Spring for quick
responders, and late-
Summer for integrative
effects over whole growing
season
Species composition data may be
obtained by quadrat or point sampling in
the relevant channel zone.
Plant collection and identification are
likely to be required to determine species
composition. Plant identification to
species level can be time-consuming and
may require taxonomic expertise. May be
difficult to assess submerged species.
Endpoint Species
composition,
number of
amphibious and
terrestrial species
in Zones B & C
Late Spring-early Summer
(preferably a fixed time
such as 7 weeks after the
passage of Spring freshes).
Endpoint Proportion of
exotic plant
species
Proportion of total number of plant
species that are exotic species in
each relevant zone
Obtainable from species composition data
Endpoint Germination of
seedlings of
overstorey and
midstorey species
Number of native and exotic
seedlings of overstorey and
midstorey species per m2 of Zone
C area
Summer (preferably a fixed
time such as 7 weeks after
the passage of Spring
freshes) OR
Fixed time after the
passage of bankfull/
overbank flows
Quadrat sampling in the channel zone C
Identification of exotic weed seedlings
should be fairly easy, but identification of
native seedlings may be more difficult (J.
Catford, pers. comm.).
49

Monitoring and evaluation of environmental flow releases in the Campaspe River
Field
Program
Type Variable Definition & Measure Timing & Frequency Sampling Protocol
Vegetation Survey (cont.)
Endpoint Canopy condition The physiological condition of
vegetation affects its survivorship,
growth, reproduction, habitat
quality and ability to perform
ecosystem functions. Assessment
of the condition of a plant’s
canopy provides information about
the physiological condition of that
individual (Baldwin et al. 2005).
e.g.
a) visual rating of canopy
condition for trees of
different age/size classes
Dyer & Roberts (2006)
recommend that canopy
condition be assessed
annually in early summer
(and preferably a fixed time
(e.g. 7 weeks) after the
passage of spring
fresh(es).
And possibly at a fixed time
after the passage of
summer freshes
Canopy condition can be assessed using
a visual rating system aided by % ‘cover’
diagrams. Baldwin et al. (2005)
recommended that the assessment of
canopy condition be based on visual
survey of four components: i) % of
branches which are dead; ii) % of canopy
represented as epicormic growth; iii) % of
canopy which is discoloured; and iv)
canopy density (% cover) for that
individual. They suggest that on-ground
measurement might involve applying the
visual ratings system to 10 randomly
selected individuals for each species
under study. Dyer & Roberts (2006)
suggest a procedure involving the visual
rating of canopy condition for a random
sample of 30-50 trees in a pre-designated
area at each site. They also recommend
stratification if the trees are clearly of
different age/size classes.
Other methods include spherical
densiometers and the use of fish-eye lens
photography and computerised analyses
of images using appropriate software
(Werren & Arthington 2002).
Fish abundance
and composition
survey
Endpoint Fish species
composition and
distribution
e.g. Ratio of species actually
collected at a site compared to
total suite of species believed to
be present in the reach (from past
records, anecdotal information,
Sustainable Rivers Audit models
etc.)
Report on pilot studies for
the Sustainable Rivers
Audit (SRA) considered
that fish were best sampled
during baseflow conditions
and that Autumn was the
best time to sample fish
(MDBC 2004).
Adult fish may be sampled using
backpack, bank- and boat-mounted
electro-fishing and fyke nets, although
fyke netting has generally been found to
be a cost-ineffective survey method (A.J.
King pers. comm.).
(cont. below)
50

Monitoring and evaluation of environmental flow releases in the Campaspe River
Field
Program
Type Variable Definition & Measure Timing & Frequency Sampling Protocol
Fish abundance and
composition survey
(cont.)
Endpoint Relative
abundance of
adult/sub-adult
native and exotic
fish species
eg Abundance of adult/sub-adult
native and exotic fish per unit
effort (expressed as catch per unit
effort).
See above Individual fish measured to the nearest
millimeter with Total Length (TL) for
round-tailed fish and Caudal Fork Length
(LCF) for fork-tailed fish.
In the SRA, all individual fish larger than
15mm were counted and identified to
species. Individuals smaller than 15mm
were not counted in the sample as there
were concerns that the gear types used
(electro-fishing, fyke and light traps) might
be relatively ineffective for fish of that size
and smaller (MDBC 2004).
Endpoint Population
structure & Size
class distributions
of native and
exotic fish species
e.g. Size class distribution by
length or weight of native and
exotic fish species
Larval Fish Survey
Endpoint Presence/
Absence of
diadromous fish
larvae
Diadromous fish species include
galaxiids, eels and Australian
Grayling
Throughout Autumn and
Winter.
Minimum sampling
frequency of once a month.
According to Humphries &
Lake (2000) fish larvae
sampling must be carried
out at least monthly
because of the variation in
the behaviour of larvae of
different species of fish.
A variety of methods may need to be used
for sampling fish larvae, such as the use
of drift nets and light traps in run areas
and trawl nets and light traps in pool areas
(Humphries & Lake 2000).
Humphries et al. (2002) reported that drift
nets (500 μm mesh, 1.5 m long with a 0.5
m diameter mouth and tapered to a 90mm
diameter cod end to which a reducing
bottle was fitted) and light traps were used
in runs and plankton tow nets and light
traps were used in pools. They also
reported that seine netting proved
relatively ineffective in collecting fish
larvae of any species.
(cont. below)
Endpoint Presence/
Absence of ‘flood
specialist’ non-
diadromous fish
larvae
‘Flood specialist’ species include
Golden Perch and Silver Perch
Throughout Spring and
Summer.
Minimum sampling
frequency of once a month.
51

Monitoring and evaluation of environmental flow releases in the Campaspe River
Larval Fish Survey
(cont.)
Endpoint Presence/
Absence of ‘low
flow specialist’ &
generalist fish
larvae
‘Low flow specialist’ species
include Crimson-spotted
Rainbow fish and Carp
Gudgeons. Generalist fish
species include Australian
Smelt and Flathead Gudgeon.
See above Data from drift and tow nets may be
expressed as number of larvae per unit
volume of water (m3).
Data from light traps may be expressed
as number of fish per trap.
Endpoint Density of post-
larval fish
e.g.
a) number of larvae per unit
volume of water (m3)
b) number of fish per trap
Water
Quality
Survey
Endpoint Standard Water
Quality
parameters
Parameters: pH, total phosphorus,
total nitrogen, turbidity, suspended
solids, dissolved oxygen,
temperature, salinity.
Monthly as a minimum
frequency. Fortnightly
preferred
Follow standard protocols for sampling
methodology used by the VWQMN.
Covariate Bed/Bank
Erosivity
Resistance to erosion
characterized categorically (low
medium, high)
Initial survey, followed by
infrequent repeats as
channel evolves (5 yrs +)
Method described in Annandale (1996)
Covariate Rate of riparian
grass growth on
benches and point
bars
Grass cover Surveys at the start and
end of the Spring-Summer
period.
Survey of grass cover on benches and
bars
Covariate Active habitat
management
activities
e.g.
a) Re-snagging
b) Other activities
Anecdotal data
Covariate Revegetation
activities
e.g.
a)Passive regeneration by fencing
off/restricting stock access
riparian zone
b)Direct seeding and/or replanting
Covariate Stock access e.g.
a) No stock access any time
b) Partial stock access (e.g.
watering points)
Field
Program
Type Variable Definition & Measure Timing & Frequency Sampling Protocol
52

Monitoring and evaluation of environmental flow releases in the Campaspe River
53
Field
Program
Type Variable Definition & Measure Timing & Frequency Sampling Protocol
Covariate Instream
barrier(s) to fish
movement
e.g.
Presence/absence of effective
instream barriers to fish
movement into the reach
Covariate Fish stocking
practices
e.g.
a) Fish species stocked.
b) Number, size class, location &
frequency of stocking

Monitoring and evaluation of environmental flow releases in the Campaspe River
2.10 Potentially Important Adverse Effects
The aim of this program is to determine whether the desired ecological outcomes from the
implementation of environmental flow components have occurred. It should be noted that the
implementation of environmental flows may also be accompanied by non-desirable outcomes. The
FLOWS method is based on the natural flow paradigm (Poff et al. 1997), which was developed as a
conceptual model to explain the relationship between flow and ecological outcomes. In the transfer of
this scientific, explanatory model to the management context it has become clear that practical
restrictions often prevent the application of a flow regime in its entirety (e.g. infrastructure risks
associated with overbank flows generally make the delivery of a flow component unrealistic, or there
simply is not enough water to deliver all components). Commonly only a limited number of flow
components can be delivered, and at this stage it is unclear what this means for the conceptual links
in the natural flow paradigm. Future work should include a more comprehensive analysis of the
potential risks associated with such partial deliveries of flow components.
Aside from the high-level concern noted above, other more easily definable adverse effects may occur
as a result of the instigation of an environmental flows program. Such adverse effects include, but are
not restricted to:
a) Provision of environmental flows from below the thermocline of a storage may cause cold water
pollution.
b) Increased shallow and slow water areas following reinstatement of more natural levels of
baseflows year-round may increase the availability of habitat for adult and sub-adult carp.
c) Increased slackwater areas may increase the availability of suitable habitat for carp larvae and
juveniles and consequently improve survival, recruitment and population size of carp in the reach.
d) Spring and summer freshes/bankfull/overbank flows may flush eggs, larval and juvenile fish from
slackwater habitats.
e) Winter-Spring freshes/bankfull/overbank flows may deliver large quantities of organic matter into
the river and cause blackwater events. If these events lead to low oxygen levels in constrained
(i.e. between weirs) sections of the river, it may lead to fish kills and other detrimental biotic
effects.
f) Winter-spring freshes/bankfull/overbank flows may flush highly saline and/or deoxygenated water
from the bottom of stratified pools into the water column to the detriment of aquatic biota.
g) Environmental flow events may lead to the dispersion of invasive plant and animal species. At
present, very little is known about dispersal and proliferation in relation to environmental flows and
there is little scientific data, but it is likely that the benefits of environmental flows to native flora
and fauna will also be experienced by invasive species.
54

Monitoring and evaluation of environmental flow releases in the Campaspe River
3 Monitoring program design and data analysis
A monitoring design consists of directions that stipulate what, where, when and how many
observations or sampling units should be taken to assess whether or not a change has occurred in a
system. The design of a monitoring program should link the delivery of environmental flows
components to the relevant environmental and ecological responses outlined in the conceptual
models.
3.1 Background
3.1.1 Practical constraints for monitoring the effects of environmental flows
Fully replicated monitoring studies that compare changes in response variables at Control, Impact and
Reference locations both Before and After the prescribed environmental flows (BACI designs and
derivatives; Underwood 1991, 1992, 1994) would enable the strongest conclusions (sensu Downes et
al. 2002) about whether the interventions are having the predicted effect (Cottingham et al. 2005a, b).
However, a number of factors prevent the application of BACI-type designs to many of the river
systems in the statewide program. Some of these are as follows:
1. Lack of ‘Before’ data. Defining ‘Before’ conditions is not straightforward as a clear definition of
‘Before’ conditions is not possible. Environmental flows implementation will not be a ‘step change’
whereby all recommendations are delivered in full after a certain date. Rather, it will be an
incremental and variable process, contingent on natural variability in climatic and hydrologic
conditions, as well as operational and logistical constraints. Furthermore, some environmental flow
components have already been delivered in some of the eight nominated rivers, and it is unlikely
that data that can be used to test Before-After hypotheses have already been collected. Extant
data collected prior to environmental flow provision as part of existing monitoring programs might
be useful, but it is unlikely that such data will be suitable for testing the hypotheses of interest at
the appropriate spatial and temporal scale (Chessman & Jones 2001).
2. Lack of Control locations. It can be very difficult to identify a suitable set of Control locations in
other river systems, because there are likely to be important environmental differences in
physiographic, geomorphic, hydrologic and ecological characteristics. In addition, there are likely
to be differences in water resource development and historical and current land use. One might
expect that the upstream reaches or tributaries of rivers might be able to act as controls. But in
reality there may be natural systematic, longitudinal changes in geomorphological,
physicochemical and biological characteristics in the river or between the main river channel and
its tributaries (Downes et al. 2002). For example, we expect to see natural changes in
composition of the river biota proceeding from upstream to downstream environments (Vannote et
al. 1980, Schlosser 1982).
3. Lack of Reference locations. Given that nearly all of our catchments are developed to some
degree, Reference locations can also be difficult or impossible to identify. If upstream reaches and
tributaries are chosen as Reference locations, one would also need to be aware of the possibility
of natural systematic, longitudinal differences in geomorphological, physicochemical and biological
characteristics of the river or differences between the main river channel and its tributaries.
55

Monitoring and evaluation of environmental flow releases in the Campaspe River
4. No Factorial Treatment. BACI-type designs rely on being able to identify a site as being either
Control or Impact. Implicitly, this assumes that all Impact sites are subjected to a similar level of
the treatment. Thus, for environmental flows, we either consider a site as ‘flow-enhanced’ or not.
In reality, different sites will receive different proportions of the recommended flows, depending on
water availability and competing uses. The amount delivered from year to year at any one site will
also vary for the same reasons. Thus the environmental flows ‘Impact’ will be continuous rather
than factorial in nature. It would possible to draw an arbitrary line through this continuum and
describe all flow enhancement less than a certain amount as Control, and all above it as Impact,
but this would be very poor practice and would be less likely to tell us anything useful about the
effects of environmental flows than would an analysis that treats the flow ‘treatment’ as a
continuous variable.
3.1.2 Proposed approach to the VEFMAP monitoring program design
The difficulty in applying BACI-type designs means there is a need for targeted monitoring and
innovative analytical approaches that will enable us to accommodate and, if possible profit from the
anticipated variation in environmental flow implementation, starting conditions and lack of
Control/Reference locations.
Based on the expectation that different sites within river reaches are likely to be subject to different
environmental flow regimes, we recommend that response variables be measured at multiple sites
within each river, representing different levels of a continuous ‘treatment’ along a continuum of
environmental flow interventions. Thus, information will be collected on flow and response along a
gradient, allowing us to build up a picture of the relationship between flow and response. This is in
contrast to the BACI approach, which collects information at the two extreme ends of the gradient and
applies a test to see whether or not the two ends are significantly different to one another.
For responses that are discrete in time (i.e. they can be conceptually tied to the flow regime at the site
over the last year or perhaps slightly longer), variation in the amount of environmental flows delivered
over time will also result in response data that can be treated as a function of the continuous flow
‘treatment’. For such responses, there will be periods of time at most sites when environmental flows
are not delivered at all due to a lack of available water. Data collected during these years would lie at
one extreme end of the ‘treatment’ value (i.e. no treatment applied). These data can be thought of as
‘control’ in the sense of the BACI analysis, and will be very useful for inferring causal linkages between
flow and response. Conversely, for some rivers, it may be possible to find reference sites on nearby
rivers that are sufficiently similar to the ‘treatment’ river to be considered in the same analysis. Data
from such sites would lie at the other end of distribution of treatment values (i.e. natural flow regime),
and would again be very useful for inferring cause and effect. On the whole, however, the approach
steps away from the BACI-type designs, and recognizes that we are unlikely to be able to define
specific Control or Reference sites, that the concept of Before/After is problematic for environmental
flows interventions, and that the environmental flow ‘treatment’ can not be treated as a factorial
Control/Impact variable. The approach relies on collecting data at multiple sites and times, and
treating these in the broadest sense, as replicate measurements. That means that where the data
permit it, results from multiple sites, reaches and potentially rivers will be combined in analyses to infer
the effects of flow releases. The challenge will be to account for other variation that will inevitably
occur between sites and times so that the effect of different flow regimes, and not of other extraneous
variables, is being tested in statistical analyses. As stated above, the aim of this type of program will
be to quantify and describe the relationship between flow augmentation and various ecosystem
responses, and this is in contrast to BACI-type designs that primarily aim to infer whether or not an
effect has occurred through a null hypothesis test, without necessarily quantifying the size of that
effect (although this is possible). For example, we might seek to quantify the relationship between the
56

Monitoring and evaluation of environmental flow releases in the Campaspe River
frequency of freshes and germination of a certain species of riparian vegetation. We can use this
relationship to infer whether or not there has been a beneficial effect of enhanced flows at individual
sites, and also to make predictions about the potential benefit of different flow regimes. This example
is developed in detail in Appendix 3.
Spatial variability is characteristic of river systems, and a sample taken at one place will differ from
another sample taken at a second location. Even where flow delivery is approximately equal (most
likely at the reach scale), replicate sites are necessary to provide some measure of the site-to-site
variability of each response variable, and to provide a better estimate of average response at the
reach scale. This is important for statistical analysis of any contrast between reaches. For a given level
of natural variability in a response variable, more precise estimates can be obtained with a larger
number of samples. However, the larger the number of sites, the more intensive and expensive the
monitoring effort becomes. We recommend a similar compromise to that proposed by Sharpe & Quinn
(2004), which is to use a minimum of two sites per reach so that one can at least obtain an estimate of
variability, but to use more sites per reach if the resources allow for it. If there are previous data that
allow a calculation of the between site variability, preliminary power analyses may be useful for
determining the number of sites necessary for certain variables. Power analysis techniques are well-
developed for frequentist null-hypothesis testing analysis frameworks. Little work has been done for
power analysis of Bayesian techniques, although the principles will be the same (Cottingham et al.
2005). However, even if Bayesian techniques are to be used for final data analysis, a power analysis
done with frequentist techniques will give an indication of the necessary degree of site replication, and
is likely to be conservative.
Generic criteria for selecting monitoring sites within a river reach are outlined as follows:
1. Representative – sites should be physically representative or typical of the reach in terms of
characteristics such as hydrology, channel morphology, abundance of instream vegetation and
woody debris and riparian vegetation. Sites should not be located immediately downstream of
major tributary confluences, or at road crossings, bridges, gauging stations, weir pools and any
other built structures that may have created artificial flow and habitat characteristics.
2. Proximity to gauging station – sites should be located at reasonable proximity to an operational
gauge that can provide reliable flow data (including contributions from tributaries) for tracking flow
characteristics such as the timing, magnitude, duration and frequency of the various flow
components. Sites that were assessed using HECRAS modelling during the original
environmental flows studies should be used if they are suitable by the other criteria supplied. For
these sites we have a parameterised model of expected changes to habitat components under
different flow regimes.
3. Accessibility – should be reasonably accessible. As Sharpe & Quinn (2004) point out
accessibility has been an issue of over-riding concern and while accessible sites may not be
representative, some compromise may be inevitable.
4. ‘Independence’ – response variables may be spatially auto-correlated between upstream and
downstream sites. It is difficult to know a priori how far apart we have to keep measurements to
ensure that they are independent estimates. We follow Sharpe & Quinn (2004) and recommend
that sites should be at least 1 km apart. If necessary, the effects of spatial autocorrelation can be
accounted for during the analysis of data.
5. Availability of relevant historical data. Environmental flows may already have commenced in
some rivers, either in full or in part, so the opportunity to collect ‘before’ data might have passed.
The availability of relevant historical data may provide us with some information for making useful
inferences once the monitoring data are in hand. Such data may exist as part of routine monitoring
programs or specific research programs that have previously been carried out.
57

Monitoring and evaluation of environmental flow releases in the Campaspe River
Given the requirements detailed above, particularly ones such as proximity to a gauge and
accessibility, sites cannot be chosen at random from all possible sites within a reach. Nevertheless, it
is important to try and ensure as far as possible that the chosen sites are physically representative of
the reach in which they are located. This is not a trivial exercise as some of the reaches span a great
distance (e.g. 60->100 km long) and the notion of ‘typical’ becomes more tenuous in those situations.
We recommend the following practical approach for site selection:
1. Plot the (1) annual mean discharge, (2) valley width and (3) meander wavelength against river
distance for the entire length of river to be monitored (These data are available from the eWater
project team if required).
2. Draw reach boundaries (as delineated in the original environmental flow study) on this graph and
confirm that there are no major step changes in these three characteristics within the reaches. If
there are, consider adding an additional reach (with a division at the largest step change in the
relevant attribute.
3. Visually classify each reach into two or three sub-reach types based on meander wavelength and
valley width (local expertise will help with this). The river may switch back and forth between sub-
reach types (e.g. as the valley contracts and widens then contracts again)
4. On the graph produced in step 2, locate (1) major tributaries (2) active streamflow gauging
stations (3) channel cross-sectional survey sites used to develop hydraulic models for the
environmental flow study and (4) sections of the river where access is feasible for monitoring
(allowing for the possibility of arranging access with riparian landholders where this is known).
5. Choose to sample sites within each of the two or three sub-reach types where (i) access is
possible; (ii) are not located at or within 1 km of a major tributary confluence; (iii) are not unique in
terms of wavelength, geomorphology, valley width or other known characteristic of the river (e.g.
sites with recent engineering works should not be used) and (iv) mean flow is as close as possible
to that of an active streamflow gauge. As discussed above, f the channel survey sites used to
develop hydraulic models for the environmental flow study satisfy these criteria, they should be
selected.
3.2 Bayesian hierarchical approach to data analysis
The data can be analysed by any statistical method that can accommodate the following features:
• The main driver of ecological effect, stream flow, is continuous, rather than categorical
• There will be other drivers of ecological effects that will vary between replicate measurements.
• Given the linear nature of rivers, data may be spatially auto-correlated. Data collected at the same
site over time may be temporally auto-correlated.
• The method can utilise data from multiple sites and/or times to infer the effects of environmental
flows.
We advocate the use of a regression-based approach within a Bayesian hierarchical modelling (BHM)
framework. This approach may have certain advantages with regards to the inclusion of data from
multiple, partially different, sites within a single analysis. It will also allow the inclusion of prior
knowledge of the effects of flow on the biota to be formally incorporated in models.
58

Monitoring and evaluation of environmental flow releases in the Campaspe River
A regression-based approach describes a mathematical relationship between variables based on the
data available. This is in contrast to BACI type designs that explicitly test for differences between two
or more data sets.
The basis of Bayesian statistical modelling is briefly covered in Cottingham et al. (2005b). We believe
that the central advantage of Bayesian hierarchical modelling, in particular, to the analysis of data in
the VEFMAP program is a property known as ‘borrowing strength’ (Gelman et al. 1995). Practically,
this means that the data from one site will lead to stronger conclusions being drawn from the data at a
second site when the two sites are considered in the same model. The site-level conclusions are
stronger for each site than would have been possible if the data were considered in separate
analyses. Given the relatively small number of sites likely to be sampled in each reach / river, and
given that data may only be collected once or twice a year, any means to strengthen conclusions from
these sparse data sets needs to be applied.
The increase in inferential strength relies on the two sites behaving in similar (but not necessarily)
identical fashion to the flow augmentation, and that the two sites also be similar in terms of other
environmental variables. This second condition is likely to be problematic for many sites. However, it is
also possible to build site-specific differences into Bayesian hierarchical models such that the results
can still be considered together (Gelman et al. 1995). If sites show very different responses to flow
augmentations, the flow regimes are completely different, or site specific differences are too great, the
data should not be analysed within the same model. Such differences would show up in pre-analysis
checks of the data that are mandatory no matter what statistical approach is being used.
It is important to note that we are not necessarily committed to the use of Bayesian statistics by the
design of monitoring program being advocated. Nor is the design of the program being driven by a
particular desire to use BHM. Although our recommended program would not be amenable to analysis
by a BACI model, this is a result of the spatially and temporally variable nature of the environmental
flows interventions, along with restrictions in the types of sites available, and the overall budget for
monitoring. BHM is one practical approach to the type of data that will be produced when studying the
effects of environmental flows. Other approaches that can accommodate the points above could also
be used. These would include multiple regressions of various types and meta-analysis. However, we
believe that these other approaches would provide inferior results compared to Bayesian modelling.
An introduction to Bayesian statistics was provided in the VEFMAP Stage 1 report (Cottingham et al.
2005b; Appendix 1). In addition, a technical discussion of Bayesian hierarchical modelling is provided
in Appendix 3 of this report, along with a hypothetical case-study that demonstrates the effects of
considering data from multiple sites, reaches and rivers within the same model.
3.3 Reach-by-Reach Monitoring Design
The information presented in the sections below should be used by consultants to design the
individual monitoring programs. In particular, Table 11 provides a summary of the field programs
(detailed in Table 6) that apply to each reach in the Campaspe River. Thus, in order to use this report
in the development of a specific monitoring program, the user should first refer to Table 11 to identify
which broad sections of Table 6 are relevant to a particular reach in the Campaspe River. Using this
information, the user can then refer to the relevant section in Table 6 to get specific details about
variables to be considered, and details on how these variables should be sampled. In addition, the
59

Monitoring and evaluation of environmental flow releases in the Campaspe River
user can refer back to the tables summarising conceptual models in § 2.3-2.7 to identify the
hypotheses that are being tested by sampling particular variables. The user can also relate these
hypotheses to the conceptual understanding of the relationship between environmental flows and
objectives by referring to the details of the conceptual models presented in these sections.
As mentioned above (§ 2.9), the details of the individual reaches will determine which aspects of the
generic conceptual models are applicable, and hence which variables should be monitored.
Differences in the proposed delivery of water compared to the original recommendations may also
affect what variables to monitor. A detailed monitoring program for a particular year need to be based
on information about the likely volumes of water committed for that water year (i.e. based on the
annual watering plan) and the flow components that this water will deliver. The program then monitors
the expected ecological outcomes for those flow components. For each reach, the sections below
make general recommendations for monitoring. When designing detailed monitoring programs,
consultants should consider the following:
a) Reach description (hydrology including current flow regime; seasonal flow inversions, etc., special
features of geomorphology, barriers, condition of vegetation, fish, and macroinvertebrates).
b) Intended program of delivery of environmental flows with reference to recommended
environmental flows (including details on the priority of different flow components and delivery
rules/conditions, if any exist). This will change from year to year, and the program will need to be
updated accordingly.
c) Comparison of recommended environmental flows with i) current delivery of environmental flows
and ii) intended program of delivery of environmental flows.
d) The differences to the current flow regime that will be caused by the intended delivery of
environmental flows.
e) Statewide hypotheses that apply in the reach based on information provided on intended program
of delivery of environmental flows.
f) Availability of relevant historical and current data pertaining to response variables to be monitored
for hypothesis-testing.
g) Implications of intended program of delivery of environmental flows and availability of
historical/current data for the testing of statewide hypotheses.
h) Any reach-specific characteristics that might have implications for i) site selection; ii) timing of
sampling; iii) number of samples; iv) frequency of sampling in the monitoring of response
variables.
In this report, we have concentrated on the environmental flow reaches, and have not attempted to
identify reference systems. Consultants will need to identify whether potential reference systems exist
(there may already be information on proposed reference systems), and whether parallel monitoring to
that in the environmental flows reaches can be implemented in the reference system within budget.
Similarly, we have not tried to identify existing data sets that may be useful as ‘Before’ data. The
existence of such data should be investigated, and the program designed with these data in mind. A
prime example of such a data set is that collected in the Campaspe and Broken Rivers during the
1990s (e.g. Humphries & Lake 2000, Humphries et al. 2002, Humphries & Cook 2004).
60

Monitoring and evaluation of environmental flow releases in the Campaspe River
Our treatment in the sections below of the specific links between ecological response hypotheses and
various flow components is also superficial. This was unavoidable, due to the requirement to provide
broad monitoring recommendations across eight rivers. As part of the fine-tuning of field programs
process, consultants should return to the original environmental flows recommendations and to the
information contained in the conceptual models in § 2 and Appendix 2 to ensure that monitoring
carried out relates to the specific hypotheses developed for the river.
3.3.1 Reach 1: Coliban River, Malmsbury Reservoir to Lake Eppalock
3.3.1.1 Reach Description
This section is based on information in Marchant et al. (1997) and SKM (2005, 2006a, b).
Flow seasonality has not been greatly affected under regulation, but flow magnitudes have been
reduced in summer-autumn and early winter. Tributaries contributing to this reach include Sandy
Creek (entering near Taradale) and Myrtle Creek (entering about 10km u/s Lake Eppalock). Inflows
from these major tributaries and general runoff substantially influence flow in the Coliban River.
Stratification and deoxygenation occurs in the lower sections of the Coliban River. Although
Malmsbury Reservoir has been identified as a source of potential cold water pollution (Ryan et al.
2001), no data are available to assess the effect of the reservoir on water temperatures in Coliban
River (SKM 2006a).
The Coliban River downstream of Malmsbury Reservoir primarily flows through Basalt Plains and
Granite Hills landforms with occasional Sedimentary Hills deposits. At Phillips Road, the channel was
relatively uniform and narrow (<10m wide at bankfull). Instream boulders constrict flow in some places
and small bars and benches were evident. This part of the reach did not appear to be subject to major
sand impacts. Downstream of Phillips Road, there were a number of small waterfalls and rocky
cascades which are distinct features and represent natural barriers to fish. Sections upstream of
Sandy Creek and near Lyell Road contain pool and riffle sequences. However, in other sections, many
channel features have been smothered or infilled by sand inputs from Sandy Creek and other local
sources. At Lyell Road, the channel was about 15-20m wide at bankfull and was characterised by a
series of pools and gravel riffles. One section of the right hand bank was a steep bedrock wall but the
rest of the Lyell Road site has gently sloping grassed banks. There is no developed floodplain in this
reach with only occasional alluvial flats observed (SKM 2005, 2006a, b).
In the mid- to late-1990s, EPA macroinvertebrate sampling was carried out at a single site within this
reach (Lyell Road). Macroinvertebrates from both riffle and edge habitats were assessed as Band A in
AUSRIVAS (1.0 and 1.05 respectively, equivalent to ‘reference condition’). Taxon richness was high,
with 27 families recorded from riffle habitats and 34 families recorded from edge habitats. SIGNAL
scores were 5.7 for riffle habitats and 5.4 for edge habitats. These scores were indicative of mild
pollution levels. In addition, extensive macroinvertebrate sampling was done during the Campaspe
Flow Manipulation Project (Humphries and Cook 2004).
SKM (2006a) noted that few species of instream plants were observed during field site inspections.
The most commonly observed species was the native, perennial, Water Ribbon (Triglochin procerum).
Small patches of Ribbon Weeds (Vallisneria spp.) and Pondweeds (Potamogeton spp.) were also
sometimes observed. At Phillips Road, the riparian zone had a River Red Gum overstorey, shrub mid-
61

Monitoring and evaluation of environmental flow releases in the Campaspe River
storey with Wattles (Acacia spp.), Bottlebrush and some exotic species (e.g. Willow, Gorse and
Blackberry) and a grassy ground layer. At Lyell Road, the overstorey contained River Red Gums, but
the understorey was sparse. In other parts of this reach, some remnant patches of streambank
shrubland persist in the riparian zone but many sections are heavily impacted by willows, gorse and
blackberries.
According to SKM (2005), the Phillips Road site was selected for field assessment and channel cross-
section survey (for developing the hydraulic model used in the FLOWS methodology), because it was
one of the few locations in the upper part of Reach 1 that was not severely impacted by Willows. The
underlying motivation was to allow assessment of channel features that may have been lost from other
sections due to Willow impacts. However, this may mean that the Phillips Road site is not
representative of the upper part of Reach 1, and this should be considered during site selection.
Native fish species that have been recorded in this reach include Trout Cod and Macquarie Perch
(which have been stocked as part of a recovery program for threatened species), River Blackfish,
Mountain Galaxid, Spotted Galaxid, Flathead Gudgeon and Australian Smelt (SKM 2006a). Other
native fish species that have not been recorded but which are likely to have occurred historically in the
reach include Murray Cod, Golden Perch, Silver Perch and other Gudgeons (Marchant et al. 1997).
Extensive fish data were also collected during the Campaspe Flow Manipulation Project (Humphries
and Cook 2004). Exotic fish species that have been recorded in this reach include Brown Trout,
European Perch, Tench, Goldfish and Carp. The reach has a number of small waterfalls and rocky
cascades which represent may natural barriers to fish.
3.3.1.2 Intended Program of Environmental Flows Delivery
Apart from ‘passing’ flows and inter-valley transfer (IVT) water, all water for environmental flows is
to be sourced from sales water. Implementation of environmental flow recommendations is therefore
contingent and constrained by how much sales water is available and when it can be used. This
complicates the issue of whether flow components will be delivered in full, in part or not at all in any
given year. Inter-Valley Transfer (IVT) refers to water from the Waranga Channel entering the
Campaspe River at Campaspe Siphon (and flowing through Campaspe Reach 4) to provide flows for
the Murray River.
The intended program for delivery of environmental flows is summarized in Table 7 along with details
on the modelled ‘natural’ and recorded (current) flow regime and recommended environmental flows.
62

Monitoring and evaluation of environmental flow releases in the Campaspe River
Table 7. Comparison table showing a) recommended environmental flows, b) modelled ‘natural’
vs recorded flow regime and c) intended program of delivery of environmental flows for Reach
1 of the Campaspe River.
Coliban 1 Malmsbury Reservoir to Lake Eppalock
Compliance Point: Lyell Road
Gauge: 406215
Season Recommendation Modeled ‘Natural’ vs Recorded* Intended
Dec-May Low Flow 5 ML/d
throughout Dec-May
Modeled ‘natural’: flows fell below 5 ML/d
about 2.2 times per Dec-May period, 20
days
Recorded: pattern of summer-autumn low
flows is similar to modeled ‘natural’
‘Passing’ flow of
8 ML/d or natural
(whichever is
lower)
Dec-May Fresh 100 ML/day, 1
per year, 3 days
Fresh 200 ML/day, 1
per year, 3 days
Max. rate of rise: Q2
< 2.8Q1; Max. rate of
fall: Q2 > 0.65 Q1.
Modeled ‘natural’: 100 ML/d fresh would
have occurred about 3 times per Dec-May
period, 3 days
Recorded: 100 ML/d fresh occurs about
once per Dec-May period, 2 days
Undetermined
June-Nov Low Flow 35 ML/d
throughout June-Nov
Modeled ‘natural’: flows fell below 35
ML/d about 2.3 times per June-Nov
period, 10 days
Recorded: flows fall below 35 ML/d about
2.8 times per June-Nov period, 15 days
Recommended
flow of 35 ML/d
throughout June-
Nov, may be
supplied in part
June-Nov Fresh 700 ML/d, 4
per year, 3 days.
Max. rate of rise: Q2
< 2.8Q1; Max. rate of
fall: Q2 > 0.65 Q1.
Modeled ‘natural’: 700 ML/d fresh would
have occurred about 4.5 times per June-
Nov period, 3 days. Most commonly in
July and August.
Recorded: 700 ML/d fresh occurs about
3.3 times per June-Nov period, 3 days.
Most commonly in September and
August.
Undetermined
Aug-Sept Bankfull Flow 12,000
ML/d, 1 in 3 years, 1
day. Max. rate of rise:
Q2 < 2.8Q1; Max. rate
of fall: Q2 > 0.65 Q1.
Modeled ‘natural’: 12,000 ML/d bankfull
would have occurred about 1 in 3 years, 1
day. Most commonly in September and
August.
Recorded: pattern of bankfull flow is
similar to modeled ‘natural’
Undetermined
*The recorded flow regime refers to actual current use, including the effect of impoundments and diversions.
Modeled ‘natural’ flow is the flow regime that would exist under current land use conditions if no diversion or
storage of water. The modeled ‘natural’ and recorded (current) frequency is based on the average number of
(specified) events that will occur in any 100 year period and the duration is the median value over the period of
record. Note: The period (and number of years) of flow data used for the comparison of modeled ‘natural’ and
recorded flows was not stated in SKM (2006a, b).
Key Features:
• Summer-autumn baseflow magnitude is expected to be compliant with recommendations. If
delivered as intended, it would represent an improvement from the current situation (i.e.
recorded/current summer-autumn baseflow regime).
• Summer-autumn fresh provision undetermined.
• Winter-spring baseflow magnitude may be supplied in part.
• Winter-spring fresh provision undetermined.
• Bankfull provision undetermined.
63

Monitoring and evaluation of environmental flow releases in the Campaspe River
3.3.1.3 Monitoring Recommendations
The field monitoring programs applicable to this reach are shown in Table 11.
The reach lies between two reservoirs, and tributary inputs are limited. Geomorphologically, most of
the reach will be classified as zone 1, and we would not expect any beneficial effects of environmental
flows on channel form or dynamics. Downstream of the confluence with Myrtle Ck, we might see
geomorphic effects of flows (i.e. this stretch is zone 2), but there is probably insufficient river length (~
10 km) before Lake Eppalock to establish replicate geomorphology monitoring sites within reach 1. A
single site could be set up, but this would only provide weak inference of any effects at the reach
scale. We see geomorphic monitoring in this reach as a low priority, but it can be undertaken below
the confluence with Myrtle Ck if there is sufficient support. Other channel surveys may still be
necessary to inform one-dimensional hydraulic modelling of habitat features.
The recommended summer/autumn low flows should be supplied. The recommended winter/spring
baseflows may be supplied in part. It will be important to determine whether these deliveries result in
adequate areas of shallow and slow water habitat for macrophytes, macroinvertebrates and larval and
juvenile fish. This should be investigated via the habitat field survey program.
Overbank flows are not part of the intended delivery of environmental flows. These will occur naturally
on occasions, but their frequency will not change relative to current conditions. Accordingly, there is no
need monitor off-stream slackwater habitats as post-event surveys.
The habitat survey in conjunction with one-dimensional hydraulic modelling will track whether the
winter-spring flows delivered are effective in maintaining permanent pool depth and volume, and
providing sustained inundation of representative physical habitat features.
The single EPA macroinvertebrate sampling occasion indicates a relatively high-quality
macroinvertebrate assemblage. The macroinvertebrate survey program will enhance baseline
information on macroinvertebrate community structure for the reach, and will complement data
collected during the Campaspe Flow Manipulation Project.
Few instream macrophytes have been observed in this reach, and riparian vegetation is largely
dominated by Willows (SKM 2006a). However, quantitative baseline information is not available.
Because of the requirement to source environmental water largely from sales water, spring freshes,
bankfull and overbank flows that might improve canopy condition of riparian trees and shrubs, and
increase germination and establishment of amphibious and terrestrial species are at present not part
of the intended program of delivery of environmental flows. These flow components will occur in the
future, although possibly at lower rates than recommended. Hydraulic-modelling assisted habitat
surveys, examining the inundation of geomorphic features in channel zones A, B and C, along with
vegetation surveys of variables such as cover, species composition, canopy condition and germination
of seedlings will provide valuable baseline data on community structure, floristic composition and
regeneration of overstorey and midstorey plant species.
There are a number of native fish species that are expected to occur but are not yet recorded in this
reach. A fish abundance and composition fish survey program will enhance existing baseline data on
the fish community. Information gained from this survey can be used to fine tune larval sampling
(include or exclude certain parts of the year), but the previously recorded presence of generalist non-
diadromous species and the self-sustaining landlocked population of the diadromous Spotted Galaxid
is likely to mean that year-round larval sampling is required. Larval sampling will also help to
64

Monitoring and evaluation of environmental flow releases in the Campaspe River
determine whether the stocked populations of Trout Cod and Macquarie Perch are breeding, and
whether environmental flows may assist these populations to become self sustaining.
3.3.2 Reach 2: Lake Eppalock to Campaspe Weir
3.3.2.1 Reach Description
This section is based on information in Marchant et al. (1997) and SKM (2005, 2006a, b).
Flow seasonality and magnitudes in this reach are affected by the operation of Lake Eppalock. In
particular, regulation has resulted in severe reduction of the magnitude of Winter-Spring flows from
May-Aug. During what is usually a high flow period, flows may be reduced to <10 ML/d and may even
cease to flow occasionally (SKM 2006a, b). Summer irrigation flows in this reach result in seasonal
flow inversion and artificially elevated flows for about six months over the Summer-Autumn period
(Humphries et al. 2002).
Unregulated tributaries contributing to this reach include Axe Creek, Forest Creek and Mt Pleasant
Creek. The main hydrologic inputs are from Axe Creek as the latter two are ephemeral (SKM 2006a).
Salinity levels in the section immediately downstream from Axe Creek inflow was noted as a matter of
concern. In addition, high organic matter loads and short duration spates from these tributaries have
been linked with blackwater events (McGuckin & Doeg 2001 cited in SKM 2006b).
Lake Eppalock is one of the few large Victorian dams with a multi-level off-take that allows water to be
harvested from different heights in the water column. However, varying storage levels and specific
release operating procedures means that Lake Eppalock may nevertheless be a source of cold water
pollution if cold water is released from below the thermocline. Data presented in Ryan et al. (2001)
showed a marked depression of summer maximum temperatures immediately downstream of the
dam. Similarly, according to Humphries et al. (2002), temperature in the section of the Campaspe
River immediately downstream of the dam was affected by hypolimnetic releases. A recent study
found that summer average daily temperatures immediately downstream of Lake Eppalock were 2-5
°C lower than temperatures in the Campaspe River upstream of Lake Eppalock (SKM 2006a).
Decreased temperatures persisted as far downstream as Axedale. At Barnadown further downstream,
summer water temperatures were comparable to the Campaspe River upstream of Lake Eppalock. In
winter, thermal impacts associated with dam releases were often reversed and there was some
evidence of slightly elevated winter temperatures in the Campaspe River downstream of Lake
Eppalock (SKM 2006a).
The Campaspe River flows through Basalt Plains until just below Axe Creek and then flows through
Alluvial Plain deposits (SKM 2005). The floodplain is confined between high Shepparton Formation
(clay) terraces. In this reach, streambed material changes from gravel at the upstream end to coarse
sand (with no gravel) at the lower end (Marchant et al. 1997). The channel was generally between 12-
30 m wide and of moderate sinuosity throughout the reach. The stream was characterised by very
long (up to 1 km), deep and still/slow-flowing pools interspersed with shallow runs. Within the channel
there were some sand drapes, point bars, mid-channel bars and a series of low (1-2 m) and higher
level benches, including a major 2-2.5 m bench below Lake Eppalock (Marchant et al. 1997). With
summer irrigation flows, low-level benches are inundated during summer and autumn, when they
would historically have been dry. Instream islands created by anabranch channels were another
significant feature. At Doakes Reserve, there was a large pool approximately 20-25 m wide and
downstream of that, a large island with a narrow riffle/run anabranch channel along the left hand side
65

Monitoring and evaluation of environmental flow releases in the Campaspe River
(SKM 2005). Downstream of English's Bridge, the channel has steep, eroded banks and the stream
consisted of relatively shallow pools with distinct backwater and edge habitats filled with woody debris
and some Triglochin spp. (SKM 2005).
Marchant et al. (1997) found that the macroinvertebrate community in this reach was generally typical
of that expected in a lowland river. However, the reach also supported a number of filter-feeding
species that are normally associated with cool, faster-flowing upland streams. It was hypothesized that
the abundance of filter-feeding species such as caddisfly larvae (Hydropsychidae) and blackfly larvae
(Simuliidae) may be due to favourable feeding conditions caused by large quantities of phytoplankton
from Lake Eppalock (Marchant et al. 1997). In addition, such taxa might be favoured by summer
irrigation releases that elevate flow and reduce water temperatures downstream of Lake Eppalock at a
time when lowland streams would normally contract and warm up (SKM 2006a). In addition, extensive
macroinvertebrate sampling was done during the Campaspe Flow Manipulation Project (Humphries
and Cook 2004).
In the mid- to late-1990s, EPA macroinvertebrate sampling was carried out at a single site
downstream of Lake Eppalock. Macroinvertebrates from riffle habitats were assessed as Band B in
AUSRIVAS (0.56, equivalent to ‘significantly impaired) and edge habitats were assessed as Band A in
AUSRIVAS (0.89, equivalent to ‘reference condition’). Taxon richness was lower than the Reach 1 site
at Lyell Road, with 17 families recorded from riffle habitats and 23 families recorded from edge
habitats. SIGNAL scores were 4.8 for riffle habitats and 5.0 for edge habitats. These scores were
indicative of moderate pollution levels.
SKM (2006a) noted that few species of instream plants were observed during field site inspections.
The most commonly observed species was the native, perennial, Water Ribbon (Triglochin procerum).
Small patches of Ribbon Weeds (Vallisneria spp.) and Pondweeds (Potamogeton spp.) were also
sometimes observed. In this reach, dense stands of Phragmites australis and/or Typha spp. were
present in many places. Most of the reach had a continuous riparian zone comprising a River Red
Gum overstorey and a grassy ground layer. There was a patchy midstorey in the riparian zone which
supported native shrubs such as Bottlebrush, Wattles and River Red Gum seedlings.
Native fish species that have been recorded in this reach include Murray Cod, Golden Perch, River
Blackfish, Mountain Galaxid, Spotted Galaxid, Flathead Gudgeon and Australian Smelt (SKM 2006a).
Of these, two native species – Flathead Gudgeon and Australian Smelt were collected as larvae in
sampling carried out between October 1995 and April 1999 (Humphries et al. 2002). Other native fish
species that have not been recorded but which are likely to have occurred historically in the reach
include Trout Cod, Macquarie Perch, Silver Perch, Freshwater Catfish, Bony Bream and other
Gudgeons (Marchant et al. 1997). Exotic fish species recorded in this reach include Brown Trout,
European Perch, Goldfish, Carp and Gambusia. Of these, European Perch, Carp and Gambusia were
collected as larvae in sampling carried out between October 1995 and April 1999 (Humphries et al.
2002). Other fish data were also collected during the Campaspe Flow Manipulation Project
(Humphries and Cook 2004).Campaspe Weir is a significant barrier to adult fish movement both up-
and downstream.
3.3.2.2 Intended Program of Environmental Flows Delivery
Apart from ‘passing’ flows and inter-valley transfer (IVT) water, all water for environmental flows is
to be sourced from sales water. Implementation of environmental flow recommendations is therefore
contingent and constrained by how much sales water is available and when it can be used. This
complicates the issue of whether flow components will be delivered in full, in part or not at all in any
66

Monitoring and evaluation of environmental flow releases in the Campaspe River
given year. Inter-Valley Transfer (IVT) refers to water from the Waranga Channel entering the
Campaspe River at Campaspe Siphon (and flowing through Campaspe Reach 4) to provide flows for
the Murray River.
Information on ‘passing’ flows was provided by K. Stanislawski, NCCMA. The intended program for
delivery of environmental flows is summarized in Table 8 along with details on the modelled ‘natural’
and/or recorded (current) flow regime and recommended environmental flows.
Table 8. Comparison table showing a) recommended environmental flows, b) modelled ‘natural’
vs recorded flow regime and c) intended program of delivery of environmental flows for Reach
2 of the Campaspe River.
Campaspe 2 Lake Eppalock to Campaspe Weir
Compliance Point: Doakes Reserve
Gauge: 406207
Season Recommendation Modeled ‘Natural’ vs
Recorded* Intended
Dec-May Cease to Flow 0
ML/d, once per
Dec-May period,
14 days during
January/February
Modeled ‘natural’: cease to
flow events about 1.3 times
per Dec-May period, 13
days. Most commonly in Jan
& Feb.
Recorded: CTF events 1 in 2
Dec-May periods, 5 days.
Most commonly in May at
the end of the irrigation
season when discharge from
Lake Eppalock is cut off for
maintenance.
Unknown
Dec-May Low Flow 10 ML/d
(or natural)
throughout Dec-
May
Modeled ‘natural’: flows less
than 10 ML/d about 2.5
times per Dec-May period,
17 days Recorded: flows
less than 10 ML/d about 0.8
times per Dec-May period, 5
days
If Eppalock storage volume:
a) ≤150 GL, ‘passing’ flow of 10
ML/d or natural for all months
between Dec-May
b) >150 GL but ≤ 200 GL,
‘passing’ flow of 50 ML/d or
natural for all months between
Dec-May
c) >200 GL but ≤ 250 GL,
‘passing’ flow of 80 ML/d or
natural for all months between
Dec-May
d) >250 GL, ‘passing’ flow of
• 90 ML/d for Dec, Jan, Mar,
May & Jun
• 80 ML/d for Feb & Apr
or natural.
Dec-May Fresh 100 ML/d, 3
per year (or
natural), 5 days.
Preferably
between Feb-
May1. Max. rate of
rise: Q2 < 2.3Q1;
Max. rate of fall:
Q2 > 0.65 Q1.
Modeled ‘natural’: 100 ML/d
fresh about 4 times per Dec-
May period, 5 days. Most
commonly in Dec, May &
Apr.
Recorded: 100 ML/d flows
occur about 1.6 times per
Dec-May period, 110 days.
Commencing most
commonly in Dec.
Undetermined
June-Nov Low Flow 100
ML/d (or natural)
throughout June-
Modeled ‘natural’: flows fall
below 100ML/d about 2.5
times per June-Nov period, 8
If Eppalock storage volume:
a) ≤150 GL, ‘passing’ flow of 10
ML/d or natural for all months
67

Monitoring and evaluation of environmental flow releases in the Campaspe River
Nov days. Most commonly in Jun
& Nov.
Recorded: flows fall below
100 ML/d about 1.5 times
per Jun-Nov period, 70 days.
Beginning most commonly in
Jun.
between Jun-Nov
b) >150 GL but ≤ 200 GL,
‘passing’ flow of 50 ML/d or
natural for all months between
Jun-Nov
c) >200 GL but ≤ 250 GL,
‘passing’ flow of 80 ML/d or
natural for all months between
Jun-Nov
d) >250 GL, ‘passing’ flow of
• 90 ML/d for Jun
• 150 ML/d for Jul & Nov
• 200 ML/d for Aug-Oct
or natural.
June-Nov 1,000 ML/d, 4 per
year (or natural), 4
days. Max. rate of
rise: Q2 < 2.3Q1;
Max. rate of fall:
Q2 > 0.65 Q1.
Modeled ‘natural’: 1,000
ML/d flows would have
occurred about 4.3 times per
June-Nov period, 4 days.
Most commonly in Jul, Jun &
Oct.
Recorded: 1,000 ML/d flows
occur about 1 in 2 Jun-Nov
periods, 10 days. Beginning
most commonly in Oct, Sep
& Nov.
Undetermined
Aug-Sept Bankfull Flow
10,000 ML/d, 1 per
year (or natural), 2
days. Max. rate of
rise: Q2 < 2.3Q1;
Max. rate of fall:
Q2 > 0.65 Q1.
Modeled ‘natural’: 10,000
ML/d flows about 1.3 times
per year, 2 days. Most
commonly in Aug & Sep.
Recorded: 10,000 ML/d
flows occur about 1 in 2
years, 2 days. Most
commonly in Oct & Sep.
Undetermined
Aug-Sept Overbank Flow
12,000 ML/d, 1 per
year, 1 days. Max.
rate of rise: Q2 <
2.3Q1; Max. rate of
fall: Q2 > 0.65 Q1.
Modeled ‘natural’: 12,000
ML/d flows about 1.1 times
per year, 1 day. Most
commonly in Aug & Sep.
Recorded: 12,000 ML/d
flows occur about 1 in 2
years, 2 days. Most
commonly in Oct & Sep.
Undetermined
1So as not to flush backwater habitats in early summer when larval and juvenile fish are abundant.
*The recorded flow regime refers to actual current use, including the effect of impoundments and diversions.
Modeled ‘natural’ flow is the flow regime that would exist under current land use conditions if no diversion or
storage of water. The modeled ‘natural’ and recorded (current) frequency is based on the average number of
(specified) events that will occur in any 100 year period and the duration is the median value over the period of
record. Note: The period (and number of years) of flow data used for the comparison of modeled ‘natural’ and
recorded flows was not stated in SKM (2006a, b).
Key Features:
• Summer-autumn cease to flow period recommended.
• Summer-autumn baseflow magnitude is expected to be compliant with or higher than
recommendations. Depending on volume of water in Lake Eppalock, these flows may be
substantially greater than recommended.
• Summer-autumn fresh provision undetermined.
68

Monitoring and evaluation of environmental flow releases in the Campaspe River
• Winter-spring baseflow magnitude only to be supplied if Lake Eppalock contains > 250 Gl.
• Winter-spring fresh provision undetermined.
• Bankfull provision undetermined.
• Overbank provision undetermined.
3.3.2.3 Monitoring Recommendations
The field monitoring programs applicable to this reach are shown in Table 11.
Part of this reach will be suitable for geomorphic monitoring. The area immediately downstream of
Lake Eppalock will be geomorphic zone 1, and we do not recommend channel surveys in this stretch
beyond those necessary to inform one-dimensional hydraulic modelling of habitat feature.
Downstream of the confluence with Forrest creek will be geomorphic zone 2 and beneficial effects of
environmental flows on the channel should be investigated with channel dynamics and channel
features surveys.
Summer/autumn low flows will exceed recommendations when Lake Eppalock contains more that 150
Gl. Presumably this extra water is to be used for irrigation purposes. From the flow rule provided for
low flows (10 Ml/d), we also assume there is no special provision for a summer cease to flow period
beyond those biennial events that occur due to maintenance of the Lake Eppalock outlet (Table 8). As
this reach will often be affected by artificially elevated summer flows, it will be important to determine
whether adequate areas of shallow and slow water are maintained for macrophytes,
macroinvertebrates and larval and juvenile fish over the irrigation season when flows are artificially
elevated. This should be investigated via the habitat field survey program.
In contrast, winter/spring baseflows will usually fall short of recommendations, and often by quite a
large margin. The recommendation will only be met if Lake Eppalock contains more than 250 Gl,
which is approximately 80% of its capacity. The habitat survey in conjunction with one-dimensional
hydraulic modelling will track whether the winter-spring flows delivered are effective in maintaining
permanent pool depth and volume, maintaining channel connectivity in the face of reduced flows, and
providing sustained inundation of representative physical habitat features.
Overbank flows are not part of the intended delivery of environmental flows. These will occur naturally
on occasions, but their frequency will not change relative to current conditions. Accordingly, there is no
need monitor off-stream slackwater habitats as post-event surveys.
The macroinvertebrate survey program will help to furnish baseline information on macroinvertebrate
community structure for the reach, and will complement data collected during the Campaspe Flow
Manipulation Project. In particular, it will be important to note whether the re-instatement of more
natural levels of summer flow (if they occur) causes the loss from the assemblage of filter feeding
species more typically associated with faster upland streams as found by Marchant et al. (1997).
Few instream macrophytes have been observed in this reach, but the riparian zone is reasonably
continuous with an overstorey in good condition (SKM 2006a). Quantitative baseline information on
vegetation condition is not available. Because of the requirement to source environmental water
largely from sales water, spring freshes, bankfull and overbank flows that might improve canopy
condition of riparian trees and shrubs, and increase germination and establishment of amphibious and
69

Monitoring and evaluation of environmental flow releases in the Campaspe River
terrestrial species are at present not part of the intended program of delivery of environmental flows.
These flow components will occur in the future, although probably at lower rates than recommended.
This reach will also continue to be affected by a seasonal flow inversion, which may have detrimental
effects for vegetation, especially instream species. Hydraulic-modelling assisted habitat surveys,
examining the inundation of geomorphic features in channel zones A, B and C, along with vegetation
surveys of variables such as cover, species composition, canopy condition and germination of
seedlings will provide valuable baseline data on community structure, floristic composition and
regeneration of overstorey and midstorey plant species.
There are a number of native fish species that are expected to occur but are not yet recorded in this
reach. A fish abundance and composition fish survey program will enhance existing baseline data on
the fish community. Information gained from this survey can be used to fine tune larval sampling
(include or exclude certain parts of the year), but the previously recorded presence of generalist non-
diadromous species and the self-sustaining landlocked population of the diadromous Spotted Galaxid
is likely to mean that year-round larval sampling is required.
3.3.3 Reach 3: Campaspe Weir to Campaspe Siphon
3.3.3.1 Reach Description
This section is based on information in Marchant et al. (1997) and SKM (2005, 2006a, b).
Flow magnitudes in this reach are affected by the operation of Lake Eppalock. In particular, regulation
has resulted in reduction of the magnitude of Winter-Spring flows. During what is usually a high flow
period, flows may be reduced to <10 ML/d and may even cease to flow occasionally (SKM 2006a, b).
Seasonal flow inversion has resulted in artificially elevated flows for about two months over the
Summer-Autumn period (Humphries et al. 2002).
In this reach, the Campaspe River flows through Alluvial Plain sediments (SKM 2005). The channel
was of similar width to that in Reach 2 (i.e. 12-30 m wide), deeply incised and had very high and steep
banks. Within the channel there were a series of benches at range of elevations including a major 2-
2.5 m bench similar to that in Reach 2. There were no very large pools or defined riffles (SKM 2006a,
b). The stream consisted mostly of run areas characterized by alternating shallow (<1 m) and deep (1-
3 m) regions (Humphries et al. 2002). Streambed material was predominantly medium sand and SKM
(2005) noted that the reach had a very high volume of large woody debris. The floodplain continued to
be confined between high terraces but the reach tended to open out and flood runners and other
secondary channels were observed (SKM 2006a, b).
At Bryant's Lane, the channel was about 30-40 m wide at bankfull and the banks were high and steep-
sided. Sections of the bank around large trees were eroded. At Spencer Road (about 2 km upstream
of Rochester), bankfull channel width was approximately 35-40 m. The channel had steep banks and
a distinct bench at half bankfull height. The left hand side of the channel had a large billabong that
would most likely fill during most moderate to high flow events.
The macroinvertebrate community in this reach was generally typical of that expected in a lowland
river. In the mid- to late-1990s, EPA macroinvertebrate sampling was carried out at a single site within
this reach (Rochester). Macroinvertebrates from edge habitats were assessed as Band A in
AUSRIVAS (0.99, equivalent to ‘reference condition’). Taxon richness was high, with 29 families
70

Monitoring and evaluation of environmental flow releases in the Campaspe River
recorded from edge habitats. The SIGNAL score was 5.2 for edge habitats which was indicative of
mild pollution levels. In addition, extensive macroinvertebrate sampling was done during the
Campaspe Flow Manipulation Project (Humphries and Cook 2004).
SKM (2006a) noted that few species of instream plants were observed during field site inspections.
The most commonly observed species was the native, perennial, Water Ribbon (Triglochin procerum).
Small patches of Ribbon Weeds (Vallisneria spp.) and Pondweeds (Potamogeton spp.) were also
sometimes observed. In this reach, dense stands of Phragmites australis were common. The riparian
zone in most of the reach had a River Red Gum overstorey and a patchy midstorey of native and
exotic shrubs.
Native fish species that have been recorded in this reach include Murray Cod, Golden Perch, River
Blackfish, Flathead Gudgeon, Carp Gudgeons and Australian Smelt (SKM 2006a). Of these, Flathead
Gudgeon, Australian Smelt and Carp Gudgeons were collected as larvae in sampling carried out
between October 1995 and April 1999 (Humphries et al. 2002). Other native fish species that have not
been recorded but which are likely to have occurred historically in the reach include Trout Cod,
Macquarie Perch, Silver Perch, Mountain Galaxid, Flathead Galaxid, Freshwater Catfish, Bony Bream
and other Gudgeons (Marchant et al. 1997). Exotic fish species which have been recorded in this
reach include European Perch, Goldfish, Carp and Gambusia. Of these, European Perch, Carp and
Gambusia were collected as larvae in sampling carried out between October 1995 and April 1999
(Humphries et al. 2002). Other fish data were also collected during the Campaspe Flow Manipulation
Project (Humphries and Cook 2004).
Campaspe Siphon might present a barrier to fish movement (particularly for larval and juvenile fish)
but is drowned out at moderate to high flows and therefore would be passable by fish at these times.
3.3.3.2 Intended Program of Environmental Flows Delivery
Apart from ‘passing’ flows and inter-valley transfer (IVT) water, all water for environmental flows is
to be sourced from sales water. Implementation of environmental flow recommendations is therefore
contingent and constrained by how much sales water is available and when it can be used.
Consequently, this complicates the issue of whether flow components will be delivered in full, in part or
not at all in any given year. Inter-Valley Transfer (IVT) refers to water from the Waranga Channel
entering the Campaspe River at Campaspe Siphon (and flowing through Campaspe Reach 4) to
provide flows for the Murray River.
71

Monitoring and evaluation of environmental flow releases in the Campaspe River
Table 9. Comparison table showing a) recommended environmental flows, b) modelled ‘natural’
vs recorded flow regime and c) intended program of delivery of environmental flows for Reach
3 of the Campaspe River.
Campaspe 3 Campaspe Weir to Campaspe Siphon
Compliance Point: Spencer Road
Gauge: 406202
Season Recommendation Modeled ‘Natural’ vs Recorded* Intended
Dec-May Low Flow 10 ML/d
(and not more than
20 ML/d1) throughout
Dec-May
Modeled ‘natural’: flows fell below
10 ML/d about 2 times per Dec-
May period, 16 days. Most
commonly in Jan & Feb.
Recorded: flows fall below 10
ML/d about 2.5 times per Dec-
May period, 11 days. Most
commonly in Apr & Jan.
‘Passing’ flow of 35 ML/d
or natural (whichever is
lower). ‘Passing’ flows for
Reach 4 are released
from Campaspe Weir.
Dec-May Fresh 100 ML/d, 3
per year (or natural),
6 days. Preferably
between Feb-May2.
Max. rate of rise: Q2
< 2.3Q1; Max. rate of
fall: Q2 > 0.65 Q1.
Modeled ‘natural’: 100 ML/d flows
occurred about 4 times per Dec-
May period, 6 days. Most
commonly in May & Apr.
Recorded: 100 ML/d flows occur
about 1.8 times per Dec-May
period, 8 days. Most commonly in
Feb & May.
Undetermined
June-Nov Low Flow 200 ML/d
throughout June-Nov
Modeled ‘natural’: flows fell below
200 ML/d about 2.7 times per
June-Nov period, 12 days. Most
commonly in Jun, Nov & Oct.
Recorded: flows fall below 200
ML/d about 2.5 times per June-
Nov period, 28 days. Most
commonly in Jun, Oct & Nov.
Dependant on flows
provided in Reach 2
June-Nov 1,500 ML/d, 4 per
year (or natural), 4
days. Spread
throughout the
winter-spring period.
Max. rate of rise: Q2
< 2.3Q1; Max. rate of
fall: Q2 > 0.65 Q1.
Modeled ‘natural’: 1,500 ML/d
flows occurred about 4 times per
June-Nov period, 4 days. Most
commonly in Jul, Jun & Oct.
Recorded: 1,500 ML/d flows
occur about 1.8 times per June-
Nov period, 3 days. Most
commonly in Aug & Sep.
Undetermined
Aug-Sept Bankfull Flow 8,000
ML/d, 2 per year (or
natural), 2 days. Max.
rate of rise: Q2 <
2.3Q1; Max. rate of
fall: Q2 > 0.65 Q1.
Modeled ‘natural’: 8,000 ML/d
flows occurred about 2.3 times
per year, 2 days. Most commonly
in Aug & Jul.
Recorded: 8,000 ML/d flows
occur about 0.8 times per year, 2
days. Most commonly in Oct &
Sep.
Undetermined
Aug-Sept Overbank Flow
12,000 ML/d, 1 per
year, 1 day. Max. rate
of rise: Q2 < 2.3Q1;
Max. rate of fall: Q2 >
0.65 Q1.
Modeled ‘natural’: 12,000 ML/d
flows occurred about 1.6 times
per year, 1 day. Most commonly
in Aug & Sep.
Recorded: 12,000 ML/d flows
occur about 0.6 times per year, 2
days. Most commonly in Oct &
Sep.
Undetermined
1 This value is subject to review after planned work assessing the behaviour of saline pools and slackwaters in
different flow conditions.
2 So as not to flush backwater habitats in early summer when larval and juvenile fish are abundant.
72

Monitoring and evaluation of environmental flow releases in the Campaspe River
*The recorded flow regime refers to actual current use, including the effect of impoundments and diversions.
Modeled ‘natural’ flow is the flow regime that would exist under current land use conditions if no diversion or
storage of water. The modeled ‘natural’ and recorded (current) frequency is based on the average number of
(specified) events that will occur in any 100 year period and the duration is the median value over the period of
record. Note: The period (and number of years) of flow data used for the comparison of modeled ‘natural’ and
recorded flows was not stated in SKM (2006a, b).
Information on ‘passing’ flows was provided by K. Stanislawski, NCCMA. The intended program for
delivery of environmental flows is summarized in
Table 9 along with details on the modelled ‘natural’ and/or recorded (current) flow regime and
recommended environmental flows.
Key Features:
• Summer-autumn baseflow magnitude is expected to normally be higher than the recommended
upper limit. These flows will also be affected by the provision of flows for reach 4 from Campaspe
Weir.
• Summer-autumn fresh provision undetermined.
• Winter-spring baseflow magnitude dependent on flow provided to reach 2, which will often be
below recommendations.
• Winter-spring fresh provision undetermined.
• Bankfull provision undetermined.
• Overbank provision undetermined.
3.3.3.3 Monitoring Recommendations
The field monitoring programs applicable to this reach are shown in Table 11.
This reach marks a transitional zone between geomorphic zones 2 and 3. Thus we would expect any
geomorphic response to environmental flows to be quite slow, and it is unlikely that a detectable
change will occur within a reasonable time frame (~ 10 years). Thus, no channel surveys are
recommended beyond those necessary to inform one-dimensional hydraulic modelling of habitat
features (see below).
73

Monitoring and evaluation of environmental flow releases in the Campaspe River
Summer/autumn low flows will exceed recommendations most of the time. Importantly, the intended
passing flow is greater than the recommended maximum summer low flow. It will be important to
determine whether adequate areas of shallow and slow water are maintained for macrophytes,
macroinvertebrates and larval and juvenile fish over the summer/autumn period. This should be
investigated via the habitat field survey program. Specific water quality surveys may determine
whether the increased summer low flows disrupt saline pools as alluded to in
Table 9.
Given the scenarios for reach 2, and looking at the recoded history for reach 3, we expect that
winter/spring baseflows will often fall short of recommendations, and perhaps by quite a large amount.
The habitat survey in conjunction with one-dimensional hydraulic modelling will track whether the
winter-spring flows delivered are effective in maintaining permanent pool depth and volume,
maintaining channel connectivity in the face of reduced flows, and providing sustained inundation of
representative physical habitat features.
Overbank flows are not part of the intended delivery of environmental flows. These will occur naturally
on occasions, but their frequency will not change relative to current conditions. Accordingly, there is no
need monitor off-stream slackwater habitats as post-event surveys.
There are few macroinvertebrate data for this reach, but those that exist indicate a relatively high-
quality assemblage. The macroinvertebrate survey program will help to furnish baseline information on
macroinvertebrate community structure for the reach, and will complement data collected during the
Campaspe Flow Manipulation Project.
Few instream macrophytes have been observed in this reach, and much of the riparian zone is
affected by exotic species (SKM 2006a). Quantitative baseline information on vegetation condition is
not available. Because of the requirement to source environmental water largely from sales water,
spring freshes, bankfull and overbank flows that might improve canopy condition of riparian trees and
shrubs, and increase germination and establishment of amphibious and terrestrial species are at
present not part of the intended program of delivery of environmental flows. These flow components
will occur in the future, although probably at lower rates than recommended. This reach will also
continue to be affected by a seasonal flow inversion, which may have detrimental effects for
vegetation, especially instream species. Hydraulic-modelling assisted habitat surveys, examining the
inundation of geomorphic features in channel zones A, B and C, along with vegetation surveys of
variables such as cover, species composition, canopy condition and germination of seedlings will
provide valuable baseline data on community structure, floristic composition and regeneration of
overstorey and midstorey plant species.
74

Monitoring and evaluation of environmental flow releases in the Campaspe River
A fish abundance and composition fish survey program will enhance existing baseline data on the fish
community. Information gained from this survey can be used to fine tune larval sampling (include or
exclude certain parts of the year).
3.3.4 Reach 4: Campaspe Siphon to River Murray
3.3.4.1 Reach Description
This section is based on information in Marchant et al. (1997) and SKM (2005, 2006a, b).
Flow magnitudes in this reach are affected by the operation of Lake Eppalock. In particular, regulation
has resulted in reduction of the magnitude of Winter-Spring flows. During what is usually a high flow
period, flows may be reduced to <10 ML/d and may even cease to flow occasionally (SKM 2006a, b).
This reach is only slightly affected by summer irrigation flows, but the artificial elevation of flows has
meant the loss of summer cease to flow events. Saline pools near Echuca have been noted as a
water quality problem in this reach.
In this reach, the Campaspe River flows through Alluvial Plain landforms (SKM 2005). The channel
was incised with steep banks. The channel exhibited increasingly sinuous toward the River Murray
junction. Within the channel there were a series of benches at range of elevations. There were no very
large pools (SKM 2006a, b). The stream consisted mostly of run areas characterized by alternating
shallow (<1 m) and deep (1-3 m) regions (Humphries et al. 2002). Streambed substrate was fine to
medium sand and SKM (2005) noted that the reach had a high load of large woody debris. The
floodplain continued to be open out in this reach.
At Strathallan, the channel was relatively uniform and incised, with steep sides. The wetted channel
meandered through the bottom of the main channel and low benches were present in the bottom of
the channel. At the end of Campbells Road, the channel was wide and less steeply incised than at the
Strahallan site and there was a high load of fine sediment.
No specific information on macroinvertebrates was provided for this reach. However, extensive
macroinvertebrate sampling was done during the Campaspe Flow Manipulation Project (Humphries
and Cook 2004).
SKM (2006a) noted that few species of instream plants were observed during field site inspections.
The most commonly observed species was the native, perennial, Water Ribbon (Triglochin procerum).
Small patches of Ribbon Weeds (Vallisneria spp.) and Pondweeds (Potamogeton spp.) were also
sometimes observed. At Strathallan, small stands of Typha spp. were present. The riparian zone had
a River Red Gum overstorey and a patchy midstorey of native shrubs such as Blackwood (Acacia
melanoxylon) and Bottlebrush (Callistemon seeberi). At Campbells Road, the riparian zone had a mix
of established and young River Red Gums with most of the younger trees growing in the main
channel.
Native fish species that have been recorded in this reach include Murray Cod, Golden Perch, Silver
Perch, Macquarie Perch, River Blackfish, Flathead Galaxid, Flathead Gudgeon, Carp Gudgeons and
Australian Smelt (SKM 2006a). Of these, Flathead Gudgeon, Australian Smelt and Carp-Gudgeons
were collected as larvae in sampling carried out between October 1995 and April 1999 (Humphries et
al. 2002). Other native fish species that have not been recorded but which are likely to have occurred
75

Monitoring and evaluation of environmental flow releases in the Campaspe River
historically in the reach include Trout Cod, Mountain Galaxid, Freshwater Catfish, Bony Bream and
other Gudgeons (Marchant et al. 1997). Exotic fish species that have been recorded in this reach
include European Perch, Goldfish, Carp, Gambusia and Oriental Weatherloach. Of these, European
Perch, Carp and Gambusia were collected as larvae in sampling carried out between October 1995
and April 1999 (Humphries et al. 2002). Other fish data were also collected during the Campaspe Flow
Manipulation Project (Humphries and Cook 2004).
3.3.4.2 Intended Program of Environmental Flows Delivery
Apart from ‘passing’ flows and inter-valley transfer (IVT) water, all water for environmental flows is
to be sourced from sales water. Implementation of environmental flow recommendations is therefore
contingent and constrained by how much sales water is available and when it can be used.
Consequently, this complicates the issue of whether flow components will be delivered in full, in part or
not at all in any given year. Inter-Valley Transfer (IVT) refers to water from the Waranga Channel
entering the Campaspe River at Campaspe Siphon (and flowing through Campaspe Reach 4) to
provide flows for the Murray River.
Information on ‘passing’ flows was provided by K. Stanislawski, NCCMA. The intended program for
delivery of environmental flows is summarized in Table 10 along with details on the modelled ‘natural’
and/or recorded (current) flow regime and recommended environmental flows.
Key Features:
• Summer-autumn baseflow magnitude is expected to normally be higher than the recommended
upper limit, and will be based on the amount of water in Lake Eppalock.
• Summer-autumn fresh provision undetermined.
• Intended magnitude for winter/spring baseflows will always be substantially less than the
recommended baseflow.
• Winter-spring fresh provision undetermined.
• Bankfull provision undetermined.
• Overbank provision undetermined.
Table 10. Comparison table showing a) recommended environmental flows, b) modelled
‘natural’ vs recorded flow regime and c) intended program of delivery of environmental flows
for Reach 4 of the Campaspe River.
Campaspe 4 Campaspe Siphon to Murray River
Compliance Point: Campbells Road
Gauge: 406265
Season Recommendation Modeled ‘Natural’ vs Recorded* Intended
Dec-May Low Flow 10 ML/d
(and not more than
20 ML/d1) throughout
Dec-May
Modeled ‘natural’: flows fell below
10 ML/d about once per Dec-May
period, 25 days. Most commonly in
Feb & Jan.
Recorded: flows fall below 10 ML/d
about 1.5 times per Dec-May
period, 17 days. Most commonly in
Feb, Dec & Mar.
If Eppalock storage
≤200 GL, ‘passing’ flow
of 35 ML/d throughout
Dec-May, provided by
35 ML/d release from
Campaspe Weir.
Dec-May Fresh 100 ML/d, 3
per year2, 6 days.
Modeled ‘natural’: 100 ML/d flows
occurred about 4 times per Dec-
Undetermined
76

Monitoring and evaluation of environmental flow releases in the Campaspe River
Preferably between
Feb-May3. Max. rate
of rise: Q2 < 2.3Q1;
Max. rate of fall: Q2 >
0.65 Q1.
May period, 7 days. Most
commonly in Dec, May & Apr.
Recorded: 100 ML/d flows occur
about 1.8 times per Dec-May
period, 5 days. Most commonly in
Jan, May & Dec.
June-Nov Low Flow 200 ML/d
(or natural)
throughout June-Nov
Modeled ‘natural’: flows fell below
200 ML/d about 2.7 times per June-
Nov period, 12 days. Most
commonly in Jun, Nov & Oct.
Recorded: flows fall below 200
ML/d about 2.3 times per June-Nov
period, 32 days. Most commonly in
Jun & Oct.
If Eppalock storage
volume:
a) ≤200 GL, ‘passing’
flow
• 35 ML/d for Jun
• 20 ML/d for Jul-
Nov
b) >200 GL, ‘passing’
flow of 70 ML/d for all
months betw Jun-Nov
June-Nov 1,500 ML/d, 2 per
year (or natural), 4
days. Spread
throughout the
winter-spring period.
Max. rate of rise: Q2
< 2.3Q1; Max. rate of
fall: Q2 > 0.65 Q1.
Modeled ‘natural’: 1,500 ML/d flows
occurred about 4 times per June-
Nov period, 4 days. Most
commonly in Jul, Oct & Jun.
Recorded: 1,500 ML/d flows occur
about 1.8 times per June-Nov
period, 3 days. Most commonly in
Aug & Sep.
Undetermined
Aug-Sept Bankfull Flow 9,000
ML/d, 2 per year (or
natural), 2 days. Max.
rate of rise: Q2 <
2.3Q1; Max. rate of
fall: Q2 > 0.65 Q1.
Modeled ‘natural’: 9,000 ML/d flows
occurred about 2 times per year, 2
days. Most commonly in Aug, Jul &
Sep.
Recorded: 9,000 ML/d flows occur
about 0.7 times per year, 2 days.
Most commonly in Oct & Sep.
Undetermined
1 This value is subject to review after planned work assessing the behaviour of saline pools and slackwaters in
different flow conditions.
2 Additional freshes may be released from Dec-Feb to manage water quality if necessary.
3 So as not to flush backwater habitats in early summer when larval and juvenile fish are abundant.
*The recorded flow regime refers to actual current use, including the effect of impoundments and diversions.
Modeled ‘natural’ flow is the flow regime that would exist under current land use conditions if no diversion or
storage of water. The modeled ‘natural’ and recorded (current) frequency is based on the average number of
(specified) events that will occur in any 100 year period and the duration is the median value over the period of
record. Note: The period (and number of years) of flow data used for the comparison of modeled ‘natural’ and
recorded flows was not stated in SKM (2006a, b).
3.3.4.3 Monitoring Recommendations
The field monitoring programs applicable to this reach are shown in Table 11.
This reach lies in geomorphic zone 3. Thus we would expect any geomorphic response to
environmental flows to be quite slow, and it is unlikely that a detectable change will occur within a
reasonable time frame (~ 10 years). Thus, no channel surveys are recommended beyond those
necessary to inform one-dimensional hydraulic modelling of habitat features (see below).
77

Monitoring and evaluation of environmental flow releases in the Campaspe River
Summer/autumn low flows will exceed recommendations most of the time. Importantly, the intended
passing flow is greater than the recommended maximum summer low flow. It will be important to
determine whether adequate areas of shallow and slow water are maintained for macrophytes,
macroinvertebrates and larval and juvenile fish over the summer/autumn period. This should be
investigated via the habitat field survey program. Specific water quality surveys may determine
whether the increased summer low flows disrupt saline pools as alluded to in Table 10.
The release rules for winter spring baseflows are based on the volume of water in Lake Eppalock, and
the winter spring baseflow will range from 10-35% of the recommended magnitude. The habitat survey
in conjunction with one-dimensional hydraulic modelling will track whether the winter-spring flows
delivered are enough to maintain permanent pool depth and volume, maintain channel connectivity in
the face of reduced flows, and to provide sustained inundation of representative physical habitat
features.
Overbank flows are not part of the intended delivery of environmental flows. These will occur naturally
on occasions, but their frequency will not change relative to current conditions. Accordingly, there is no
need monitor off-stream slackwater habitats as post-event surveys.
There were no macroinvertebrate data acknowledged in the environmental flows report for this reach.
However, data may have been collected during the Campaspe Flow Manipulation Project. The
macroinvertebrate survey program will help to furnish baseline information on macroinvertebrate
community structure for the reach.
Few instream macrophytes have been observed in this reach. Varied riparian quality has been noted,
with some riparian ‘regeneration’ occurring within the channel (SKM 2006a). Quantitative baseline
information on vegetation condition is not available. Because of the requirement to source
environmental water largely from sales water, spring freshes, bankfull and overbank flows that might
improve canopy condition of riparian trees and shrubs, and increase germination and establishment of
amphibious and terrestrial species are at present not part of the intended program of delivery of
environmental flows. These flow components will occur in the future, although probably at lower rates
than recommended. Although summer effects are less than for the upstream reaches, this reach will
continue to be affected by a seasonal flow inversion, which may have detrimental effects for
vegetation, especially instream species. Hydraulic-modelling assisted habitat surveys, examining the
inundation of geomorphic features in channel zones A, B and C, along with vegetation surveys of
variables such as cover, species composition, canopy condition and germination of seedlings will
provide valuable baseline data on community structure, floristic composition and regeneration of
overstorey and midstorey plant species.
There are a number of native fish species that are expected to occur but are not yet recorded in this
reach. A fish abundance and composition fish survey program will enhance existing baseline data on
the fish community. Information gained from this survey can be used to fine tune larval sampling
(include or exclude certain parts of the year). At present, diadromous species have not been recorded
from the reach, but they are expected to occur, which would necessitate winter/spring larval sampling.
78

Monitoring and evaluation of environmental flow releases in the Campaspe River
79
Table 11. Summary of recommended field survey programs for each reach. Subcomponents of
the indivudal programs that are applicable within each reach must be determined through
examining the detailed monitoring recommendations and the reach descriptions (above). The
field programs are described in detail in Table 6.
Field Program
Reach 1 Coliban River,
Malmsbury Reservoir to
Lake Eppalock
Reach 2 Lake Eppalock
to Campaspe Weir
Reach 3 Campaspe Weir
to Campaspe Siphon
Reach 4 Campaspe
Siphon to River Murray
Flow 9 9 9 9
Channel Dynamics 9
1
9
Channel Features Survey 9
1
9
Habitat Field Survey (Repeated) 9 9 9 9
Habitat Field Survey (Post-Event)
Habitat Survey in Conjunction with One-dimensional
Hydraulic Modelling 9 9 9 9
Macroinvertebrate Survey 9 9 9 9
Vegetation Survey 9 9 9 9
Fish Abundance and Composition Survey 9 9 9 9
Larval Fish Survey 9 9 9 9
Water Quality 9 9 9 9
1. Low priority. Unlikely to be able to establish independent replicate sites
3.4 Other conceptual models and hypotheses of local importance in each
reach
As mentioned previously (§ 2.2) we collated and summarised all the conceptual models and predicted
ecosystem responses associated with the environmental flow recommendations for the eight river
systems that are likely to receive significant environmental flow allocations. A selection of these
conceptual models was further developed for inclusion in the VEFMAP. Of the remaining conceptual
models and hypotheses, some may be of local importance to individual reaches in the Campaspe
River. Table 12 contains a non-exhaustive selection of conceptual models, some of which may be
applicable to the Campaspe River. Like the other models presented in this report, these models are
generic, and only certain parts of each model may be applicable to any given river. Consultants and
the CMA should use river-specific knowledge of the biota and environments present to determine
which parts of the models may apply. The recommendation of which reaches are suitable for the
investigation of individual hypotheses is based solely on the intended flow delivery for that reach.

Monitoring and evaluation of environmental flow releases in the Campaspe River
Table 12. Additional conceptual models and hypotheses which may be of local importance to individual reaches in the Campaspe River.
No.
List of other hypotheses of local relevance in the Campaspe River
Reach 1
Reach 2
Reach 3
Reach 4
1 Biogeochemical process - During periods of flow cessation or summer low flows, exposure of the streambed allows
accumulation and drying of terrestrial organic matter in dry areas of the channel such as benches. Drying and subsequent
rewetting facilitates the decomposition and processing of this organic matter and produces a fresh pool of nutrient and carbon
inputs for the system. Winter-Spring Freshes, High Flows, Bankfull Flows and Overbank Flows can inundate higher portions
of the channel such as benches and banks and entrain organic matter accumulated in the elevated channel features and
terrestrial channel sections. This will provide inputs of dissolved and fine particulate organic matter to maintain nutrient/carbon
cycling inputs to the river. Organic matter and nutrients in this form can be used by macrophytes, algae, microfauna,
zooplankton and microbes. Their influx may result in higher rates of productivity and respiration on benches than in the main
river channel, although it is uncertain how long this effect persists. In the case of Overbank Flows which result in inundation of
the floodplain, there may be significant carbon returns to the river after a period of significant production.
9 9
2 Ecological Process, Drying Disturbance - During periods of flow cessation or summer low flows, the river may contract to a
series of isolated pools, or portions of the channel may dry out. The biota in these pools is likely to be subjected to
physicochemical stresses (e.g. Low DO concentrations, changes in EC and temperature), intensified predation and
competition. Exposure of large areas of the streambed acts as a disturbance mechanism which resets successional
processes for macroinvertebrate and vegetation communities. For instance, by allowing certain plant species to regenerate
on bars and benches. Desiccation disturbance prevents the system from being dominated by any particular group of
organisms. And in particular, any macroinvertebrate species as many macroinvertebrate species would be reduced in their
abundance and distribution over the dry period. Desiccation disturbance maintains aquatic and riparian species characteristic
of dryland river systems where flow cessation and summer low flows naturally occur. The biota in these arid environments
has special physiological or behavioural adaptations that allow them to persist in harsh conditions in locations which they
might otherwise be displaced by dominant but less tolerant species. In the short-term there may be localised extinction of
certain species. And in the long-term, changes in diversity and biomass. Recolonization of “stressed” habitats upon flow
restoration should be feasible provided there are effective refuges for biota during cease-to-flow periods. However, although
drying may be a stress for some species such as obligate aquatics, it also represents an opportunity for transient terrestrial as
well as exotic species to establish or expand.
9 9
80

Monitoring and evaluation of environmental flow releases in the Campaspe River
3 Ecological Process, Mechanical Disturbance - Bankfull flows may act as a disturbance mechanism which resets ecological
processes for both aquatic and riparian flora and fauna communities by drowning and/or sweeping plants and animals
downstream.
4 Biofilms – Winter-Spring Freshes and High Flows provide scouring flows over biofilm habitats, and this acts as a disturbance
mechanism for maintaining species composition and health. Summer Freshes are also expected to perform this function.
5 Aquatic & Riparian Vegetation, Flow Variability - Change in flow variability can take many forms that have different
consequences for vegetation patterns. Prolonged stable water levels allow plants to establish and persist close to the water
line; with the species doing this being more associated with lentic (wetland) environments than lotic (flowing water)
environments. Loss of flow variability may also result in wider zones of terrestrial or flood intolerant plant species and a
shrinking in the width of the zone characterised by flood tolerant species. Freshes provide short-term flow variability and
variation in water levels is important for maintaining species diversity in the emergent and marginal aquatic vegetation
communities. Variation in water levels is the principal driver of zonation patterns across the channel and up the river banks.
Providing Freshes and ensuring the occurrence of flow variability and variation in water levels will help (a)
restore/maintain/increase species diversity in the emergent and marginal aquatic vegetation communities; (b)
maintain/restore distinctive riparian vegetation community and structure with zonation up the bank.
91 91 91 91
6 Aquatic & Riparian Vegetation, Dispersal - High Flows deliver seed from the upper catchment to help maintain/restore
distinctive riparian vegetation community and structure.
7 Invertebrates - High Flows and Bankfull Flows which inundate previously dry sediments in higher portions of the channel
such as benches may provide a stimulus for hatching to invertebrates, with diversity and biomass peaking when inundation
exceeds 2 weeks. Loss of habitat through decreased inundation duration increases the risk of egg mortality, and the loss of
early instars (early life stages) and adults of those species not stimulated to drift. For the others the outcome will depend on
factors such as the availability of alternative habitat and predation pressure.
9
8 Native Fish, Movement - Protection or reinstatement of more natural levels of Winter-Spring baseflows will provide
conditions of sustained water levels in the river which will provide sustained longitudinal connectivity for fish movement,
including the permanent movement of large-bodied fish throughout the river reach in the lead up to the breeding season.
Freshes (which produce a minimum depth of 0.5m over the shallowest point) over the irrigation season (Nov-Apr) provide for
temporary local movement of bigger fish such as Murray Cod and Golden Perch. These Freshes may be important in allowing
upstream/downstream movement of Golden Perch to spawn. Autumn-early winter High Flows/Freshes are important for
transporting larvae of diadromous species such as Australian Grayling, galaxiids and eels to the estuary/sea. Winter-Spring
92 92 92
81

Monitoring and evaluation of environmental flow releases in the Campaspe River
82
High Flows and Freshes and Bankfull Flows may provide the cue which triggers movement and/or migration in some native
fish species, for instance, in Australian Grayling and in Tupong. This migration is associated with spawning and hence may
have an impact on species reproduction and recruitment. Spring-early summer High Flows/Freshes are important for the
upstream movement and recruitment of juvenile diadromous fish. Summer Freshes will increase water depth over low-lying
channel zones such as riffles and increase longitudinal connectivity, thereby temporarily facilitating greater movement of fish
between different instream habitats such as pools.
1. The reaches may contrast each other in terms of flow variability, with greater seasonal variability likely in Reach 1 compared to reaches 2, 3 and 4. This contrast could
provide a useful test of the hypothesis.
1 Conditions may be favourable for summer movement of large bodied fish as described above (due to elevated summer flows). Larval transport and life history cues associated
with winter/spring high flows are unlikely to occur.

Monitoring and evaluation of environmental flow releases in the Campaspe River
4 Implementing the Monitoring Design
4.1 VEFMAP implementation group
The DSE has established an Implementation Group, made up of representatives of the relevant
CMAs, DSE, and the current eWater science project team. This group will provide a forum for
resolving implementation issues, including the provision of advice to CMAs and their consultants for
implementing plans in individual rivers. It is important to establish processes to ensure that
monitoring and evaluation plans are implemented to appropriate standards and in a consistent
manner. It is recommended that quality assurance and control measures are developed to ensure
the collection and management of high quality data and information.
4.2 Quality assurance/quality control
A quality assurance/quality control (QA/QC) plan is recommended as an essential step in collecting
high quality and reliable data and the minimisation of sampling errors to acceptable levels (ANZECC
& ARMCANZ 2000, Baldwin et al. 2005, Cottingham et al. 2005). Such a plan should be based
around four elements:
• Project management;
• Measurement/data acquisition;
• Assessment and oversight; and
• Data validation and usability.
The QA/QC plan should identify important standards to be maintained for the life of the monitoring
program. For example, the plan should state the minimum training standards and qualifications of
staff that collect field and laboratory data, and the format required for the management and reporting
of data, including database structures (e.g. ANZECC & ARMCANZ 2000, Baldwin et al. 2005).
QA/QC considerations are particularly important, as it is the intention of DSE to have the analysis
and interpretation peer reviewed. In addition, the VEFMAP as a whole will be reviewed after three
years.
The collection of high quality data will be critical to the evaluation of environmental flow releases
along the Campaspe River and at State or regional levels. Consistency and repeatability of the
sampling protocol is essential if trends both within and between river systems are to be detected. It
is recommended that the North Central CMA have representation on the proposed Implementation
Group so that the QA/QC plan for the Campaspe complements that developed for other river
systems and the need for reporting at the State level. It is also recommended that the detailed
sampling programs developed from the information in this report are reviewed to assess the
applicability of conceptual model components, variables to be sampled, and methods for specific
variables.
As described in detail by Baldwin et al. (2005), key project management considerations for the
Implementation Group are to confirm:
• The list of the key personnel involved in the project, and their specific roles and responsibilities;
• The problems/questions being addressed in the monitoring program;
83

Monitoring and evaluation of environmental flow releases in the Campaspe River
• The project tasks to be undertaken;
• Quality objectives for measurement data (e.g. statements about the precision, accuracy,
representativeness, completeness, comparability and measurement range of the data);
• Any training and certification requirements for key personnel;
• The documentation required/generated in the project (including copies of all forms used in the
project); and
• Identification of potential occupational, health, safety and environment hazards, risk assessment
and risk minimisation plans.
A QA/QC plan should include a description of the experimental design to be applied, including the
location of sampling sites, sampling frequency, and the sampling methods and protocols to be
applied.
The monitoring variables described in previous sections fall into the following broad categories:
• Physical – channel cross-sections, longitudinal surveys, hydraulic modelling (e.g. HECRAS),
sediment size class distributions, and mapping of habitat elements (e.g. snags, benches, riffles,
pools).
• Physico-chemical – water quality parameters.
• Biological – macroinvertebrates, in-channel and riparian vegetation, fish.
Data collection and sample analysis will require a mixture of field and laboratory measurements and
activities. A QA/QC plan should therefore detail the requirements for:
• Field staff and equipment,
• Field sample collection,
• Field data collection and storage,
• Laboratory staff and equipment,
• Laboratory sample processing,
• Laboratory data storage, and
• Centralized data storage and management
Field staff should be competent in sampling and be able to demonstrate competence in field
procedures, including being able to adhere to established protocols, being able to avoid
contaminating samples, and being able to calibrate field instruments and make field observations.
Where possible, a requirement for formal training and testing of contractor competency should be
built into the monitoring program. Such training includes the EPA course for macroinvertebrate
sampling (L. Metzeling pers. comm.). For fish collection, there are no formal qualifications, but
extensive experience of practitioners is necessary (A.J. King pers. comm.). For sampling by electro-
fishing, contractors should undertake to follow the Australian Code of Electro-fishing Practice (NSW
Fisheries 1997). Similarly, there are no formal qualifications for contractors taking physical
measurements, but as a general guide, geomorphic assessments should only be undertaken by
experience fluvial geomorphologists, and hydraulic surveys should be undertaken by the same
consultant who will do the hydraulic modelling. The choice of consultant for hydraulic modelling
should be based upon their track record of successful model use, evidence of suitable training by
the practitioners, and a demonstrated awareness of potential pitfalls in channel survey methods and
model calibration. All equipment and field instruments should be kept in good working order, with
84

Monitoring and evaluation of environmental flow releases in the Campaspe River
calibrations and preventative maintenance carried out according to the schedule recommended by
manufacturers or other accepted standards.
Sampling protocols for the collections of physico-chemical and biological data are well established
as part of the Victorian Water Quality Monitoring Network (VWQMN) and State Biological Monitoring
Program (AWT 1999). It is recommended that these protocols, methods for sampling and
measurement, and data handling processes be adopted for monitoring environmental flow releases
in the Campaspe River. This will ensure that data collected for the Campaspe can be compared with
that collected in other river systems, and combined with other rivers to inform a Statewide
assessment of ecosystem responses to environmental flows. The VWQMN has detailed
requirements for:
• Sample handling and chain of custody documentation;
• Instrument and equipment QA, including calibration and frequency of maintenance;
• Analytical methods to be used;
• Routine field and laboratory QC activities;
• A description of data acquisition and storage requirements.
Macroinvertebrate community composition has long been used as a measure of river health in
Victoria. EPA Victoria (2003) provides a protocol for rapid biological assessment using
macroinvertebrates. It is recommended that these methods be adopted for collecting
macroinvertebrate data from the Campaspe River.
The sampling of fish populations and collection of fish data should be follow methods appropriate for
the particular circumstances and questions. Data should be recorded and stored in a format that will
allow easy inclusion in the DSE Victorian Fish Database. It is a requirement for fish collection
permits that the data be supplied to DSE for entry into the database in a prompt fashion.
There are no standard approaches to the sampling and collection of vegetation data. It is important
that the Implementation Group agree to the methods to be adopted, and ensure consistency of
methods as far as possible amongst the different CMAs. Furthermore, a vegetation technical expert
should be consulted to ensure consistency across all programs that are a part of the VEFMAP.
The implementation of the monitoring and evaluation plan for the Campaspe River will be the
responsibility of the North Central CMA. The CMA will inform the Implementation Group of progress
as part of their regular reporting on VEFMAP activities to DSE, including any difficulties encountered
and any corrective action required. Baldwin et al. (2005) note a number of factors to consider:
• That data are consistent/compatible with the Australian and New Zealand Land Information
Council National Standard;
• There are agreed protocols to transfer field and laboratory data to electronic data-bases;
• Original data sheets, laboratory records, chain-of-custody documentation and/or QA/QC data
associated with an entry to an electronic database are preserved;
• There are procedures for validation of data entered, including accuracy of transcription and
whether or not any of the data recorded is outside the range expected for that type of system
(cross-checked against QA/QC data associated with a given entry);
85

Monitoring and evaluation of environmental flow releases in the Campaspe River
• There are documented procedures for determining who can enter or change data on the data-
base and appropriate security measures to stop unauthorised access to the database;
• Data bases are flexible enough to accommodate a range of different data types;
• Retrieval of data using a variety of fields (time, place, flow etc.) is relatively straightforward;
• There are agreed procedures for handling chemical data that are below the detection limit (see
ANZECC & ARMCANZ 2000);
• There are agreed protocols for updating the data-base to account for improvements/changes in
software and hardware; and
• Agreed ownership of the data-base and procedures to be followed during organisational
restructuring.
The CMA will also be a partner in the 3-year review of collected data, particularly in terms of
interpreting results given the potential influence of local management and restoration efforts that
may occur across the Campaspe River catchment.
86

Monitoring and evaluation of environmental flow releases in the Campaspe River
5 References
Anderson, J. and QNRM (2003) State of the Rivers Project: Field Manual. J. Anderson and
Queensland Dept. of Natural Resources and Mines, Brisbane.
Annandale, G. W. (1996) Erodibility. Journal of hydraulic research, 33, 471-494.
ANZECC & ARMCANZ (2000) Australian Guidelines for Water Quality Monitoring and Reporting.
Australian and New Zealand Environment and Conservation Council (ANZECC) & Agriculture and
Resource Management Council of Australia and New Zealand (ARMCANZ), Canberra, Australia.
AWT (1999) Victorian Water Quality Monitoring Network and State Biological Monitoring
Programme. Manual of Procedures prepared for Dept. of Natural Resources and Environment,
Report no. WES 182/97, Australian Water Technologies, Melbourne.
Baldwin, D. S., Nielsen, D. L., Bowen, P. M. and Williams, J. (2005) Recommended Methods for
Monitoring Floodplains and Wetlands. Murray-Darling Basin Commission, Canberra.
Barbour, M. T., Gerritsen, J., Snyder, B. D. and Stribling, J. B. (1999) Rapid Bioassessment
Protocols for Use in Streams and Wadeable Rivers: Periphyton, Benthic Macroinvertebrates and
Fish. Second edition, EPA 841-B-99-002. Office of Water, U.S. Environmental Protection Agency,
Washington, D.C.
Bartley, R. and Rutherfurd, I. (2005) Measuring the reach-scale geomorphic diversity of streams:
application to a stream disturbed by a sediment slug. River Research and Applications, 21, 39-59.
Basu, B. K. and Pick, F. R. (1996) Factors regulating phytoplankton and zooplankton biomass in
temperate rivers. Limnology and Oceanography, 41, 1572-1577.
Bond, N. R. and Lake, P. S. (2003) Characterizing fish-habitat associations in streams as the first
step in ecological restoration. Austral Ecology, 28, 611-621.
Bone, Q., Marshall, N. B. and Blaxter, J. H. S. (1995) Biology of Fishes. Blackie Academic and
Professional Publishers, Glasgow.
Boulton, A. J. (2003) Parallels and contrasts in the effects of drought on stream macroinvertebrate
assemblages. Freshwater Biology, 48, 1173-1185.
Brock, M. A. and Casanova, M. T. (1997) Plant life at the edge of wetlands: ecological responses to
wetting and drying patterns.In Frontiers in Ecology. Klomp, N. and Lunt, I., (Ed.). Elsevier Science
Ltd.,Oxford. pp. 181-192.
Burns, A. and Walker, K. F. (2000) Biofilms as food for decapods (Atyidae, Palaemonidae) in the
River Murray, South Australia. Hydrobiologia, 437, 83-90.
Casanova, M. T. and Brock, M. A. (2000) How do depth, duration and frequency of flooding
influence the establishment of wetland plant communities? Plant Ecology, 147, 237-250.
Chessman, B. C. (1995) Rapid assessment of rivers using macroinvertebrates: a procedure based
on habitat-specific sampling, family-level identification and a biotic index. Australian Journal of
Ecology, 20, 122-129.
Chessman, B. and Jones, H. (2001) Integrated Monitoring of Environmental Flows: Design Report.
New South Wales Department of Land and Water Conservation, NSW Government, Parramatta.
Christie, G. and Clarke, S. (1999) Enhanced vegetation survey techniques for riparian management.
Rutherfurd, I. and Bartley, R. (Eds.). Second Australian Stream Management Conference, Adelaide,
South Australia, CRC for Catchment Hydrology. pp. 173-176.
87

Monitoring and evaluation of environmental flow releases in the Campaspe River
Clarke, K. R. (1993) Non-parametric multivariate analyses of changes in community structure.
Australian Journal of Ecology, 18, 117-143.
Cobb, D. G., Galloway, T. D. and Flanagan, J. F. (1992) Effects of discharge and substrate stability
on density and species composition of stream insects. Canadian Journal of Fisheries and Aquatic
Sciences, 49, 1788-1795.
Congdon, P. (2001) Bayesian Statistical Modeling. John Wiley & Sons, Guildford, U.K.
Cottingham, P., Stewardson, M., Roberts, J., Metzeling, L., Humphries, P., Hillman, T. and Hannan,
G. (2001) Report of the Broken River Scientific Panel on the Environmental Conditions and Flow in
the Broken River and Broken Creek. Technical Report 10/2001, CRC for Freshwater Ecology,
Canberra.
Cottingham, P., Crook, D., Hillman, T., Roberts, J., Rutherfurd, I. and Stewardson, M. (2003a) Flow-
related Environmental Issues Related to Goulburn River Below Lake Eildon. Report to Dept. of
Sustainability and Environment, Victoria and Murray-Darling Basin Commission, CRC for Freshwater
Ecology and CRC for Catchment Hydrology, Canberra.
Cottingham, P., Stewardson, M., Crook, D. A., Hillman, T., Roberts, J. and Rutherfurd, I. (2003b)
Environmental Flow Recommendations for the Goulburn River Below Lake Eildon. Technical Report
01/2003, CRC for Freshwater Ecology and CRC for Catchment Hydrology, Canberra.
Cottingham, P., Quinn, G., King, A., Norris, R., Chessman, B. and Marshall, C. (2005a)
Environmental Flows Monitoring and Assessment Framework. CRC for Freshwater Ecology,
Canberra.
Cottingham, P., Stewardson, M. and Webb, A. (2005b) Victorian Environmental Flows Monitoring
and Assessment Program. Stage 1: Statewide Framework. Report to the Dept. of Sustainability and
Environment. CRC for Freshwater Ecology and CRC for Catchment Hydrology, Melbourne.
Crook, D. A. and Robertson, A. I. (1999) Relationships between riverine fish and woody debris:
implications for lowland rivers. Marine and Freshwater Research, 50, 941-953.
Crowder, L. B. and Cooper, W. E. (1982) Habitat structural complexity and the interaction between
bluegills and their prey. Ecology, 63, 1802-1813.
DNRE (2002) Victorian River Health Strategy. Dept. of Natural Resources and Environment,
Melbourne.
Downes, B. J., Hindell, J. S. and Bond, N. R. (2000) What's in a site? Variation in invertebrate
density and diversity in a spatially replicated experiment. Austral Ecology, 25, 128-139.
Downes, B. J., Barmuta, L. A., Fairweather, P. G., Faith, D. P., Keough, M. J., Lake, P. S.,
Mapstone, B. D. and Quinn, G. P. (2002) Monitoring Ecological Impacts: Concepts and Practice in
Flowing Waters. Cambridge University Press, Cambridge.
Downes, B. J., Lake, P. S., Glaister, A. and Bond, N. R. (2006) Effects of sand sedimentation on the
macroinvertebrate fauna of lowland streams: are the effects consistent? Freshwater Biology, 51,
144-160.
DSE (2004) Securing Our Water Future Together. Victorian Government White Paper. Dept. of
Sustainability and Environment, East Melbourne, Victoria.
Dyer, F. and Roberts, J. (2006) Monitoring Vegetation Response to Environmental Flows in the
Wimmera: a Strategic Approach. A Report to the Wimmera Catchment Management Authority,
In2fluve, Canberra.
88

Monitoring and evaluation of environmental flow releases in the Campaspe River
Earth Tech (2003) Thomson River Environmental Flow Requirements & Options to Manage Flow
Stress. Report to West Gippsland CMA, Dept. of Sustainability and Environment, Melbourne Water
Corporation and Southern Rural Water, Earth Tech Engineering Pty Ltd, Melbourne.
EPA (2001a) Policy Background Paper. Nutrient Objectives for Rivers and Streams - Ecosystem
Protection. Draft State Environment Protection Policy (Waters of Victoria). Environment Protection
Authority, Melbourne.
EPA (2001b) Policy Background Paper. Water Quality Objectives for Rivers and Streams -
Ecosystem Protection. Draft State Environment Protection Policy (Waters of Victoria). Environment
Protection Authority, Melbourne.
EPA (2003) Rapid Bioassessment Methodology for Rivers and Streams. Environment Protection
Authority, Melbourne.
Ferrari, I., Farabegoli, A. and Mazzoni, R. (1989) Abundance and diversity of planktonic rotifers in
the Po River. Hydrobiologia, 186/187, 201-208.
Flore, L. and Keckeis, H. (1998) The effects of water current on foraging behaviour of the rheophilic
cyprinid, Chondrostoma nasus, during early ontogeny: evidence of a trade-off between energetic
benefit and swimming costs. Regulated Rivers: Research & Management, 14, 141-154.
Flore, L., Keckeis, H. and Schiemer, F. (2001) Feeding, energetic benefit and swimming capabilities
of 0+ nase (Chondrostoma nasus L.) in flowing water: an integrative laboratory approach. Archiv fur
Hydrobiologie, 135 Suppl, 409-424.
Gelman, A., Rubin, J. B., Stern, H. S. and Rubin, D. B. (2004) Bayesian Data Analysis. Chapman &
Hall/CRC, Boca Raton, FL.
Gippel, C. J., Finlayson, B. L. and O'Neill, I. C. (1996) Distribution and hydraulic significance of large
woody debris in a lowland Australian river. Hydrobiologia, 318, 179-194.
Growns, J. E., Chessman, B. C., McEnvoy, P. K. and Wright, I. A. (1995) Rapid assessment of rivers
using macroinvertebrates: case studies in the Nepean River and Blue Mountains. Australian Journal
of Ecology, 20, 130-141.
Growns, J. E., Chessman, B. C., Jackson, J. E. and Ross, D. G. (1997) Rapid assessment of
Australian rivers using macroinvertebrates: cost and efficiency of six methods of sample processing.
Journal of the North American Benthological Society, 16, 682-693.
Hart, D. D. and Finelli, C. M., (1999) Physical-biological coupling in streams: the pervasive effects of
flow on benthic organisms. Annual Review of Ecology and Systematics 30, 363-395.
Hewlett, R. (2000) Implications of taxonomic resolution and sample habitat for stream classification
at a broad geographic scale. Journal of the North American Benthological Society, 19, 352-361.
Humphries, P. (1996) Aquatic macrophytes, macroinvertebrate associations and water levels in a
lowland Tasmanian river. Hydrobiologia, 321, 219-233.
Humphries, P. and Cook, R. A., (2004) Campaspe Flow Manipulation Final Report. Department of
Environment and Heritage, Canberra.
Humphries, P., King, A. J. and Koehn, J. D. (1999) Fish, flows and flood plains: links between
freshwater fishes and their environment in the Murray-Darling River system, Australia.
Environmental Biology of Fishes, 56, 129-151.
Humphries, P. and Lake, P. S. (2000) Fish larvae and the management of regulated rivers.
Regulated Rivers: Research & Management, 16, 421-432.
89

Monitoring and evaluation of environmental flow releases in the Campaspe River
Humphries, P., Serafini, L. G. and King, A. J. (2002) River regulation and fish larvae: variation
through space and time. Freshwater Biology, 47, 1307-1331.
Humphries, P., Cook, R. A., Richardson, A. J. and Sarafini, L. G. (2006) Creating a disturbance:
manipulating slackwaters in a lowland river. River Research and Applications, in press.
Hurlbert, S. H. (1984) Pseudoreplication and the design of ecological field experiments. Ecological
Monographs, 54, 187-211.
Jobling, M. (1995) Environmental Biology of Fishes. Chapman & Hall Publishers, London.
King, A. J., Humphries, P. and Lake, P. S. (2003a) Fish recruitment on floodplains: the roles of
patterns of flooding and life history characteristics. Canadian Journal of Fisheries and Aquatic
Sciences, 60, 773-786.
King, A., Brooks, J., Quinn, G., Sharpe, A. and McKay, S. (2003b) Monitoring Programs for
Environmental Flows in Australia - A Literature Review. Freshwater Ecology, Arthur Rylah Institute
for Environmental Research, Dept. of Sustainability and Environment, Sinclair Knight Merz, CRC for
Freshwater Ecology, Melbourne, Victoria.
King, A. J. (2004a) Density and distribution of potential prey for larval fish in the main channel of a
floodplain river: pelagic versus epibenthic meiofauna. River Research and Applications, 20, 883-897.
King, A. J. (2004b) Ontogenetic patterns of habitat use by fishes within the main channel of an
Australian floodplain river. Journal of Fish Biology, 65, 1582-1603.
Knighton, D. (1998) Fluvial Forms and Processes. Arnold, London.
Koehn, J. D. and O'Connor, W. G. (1990) Biological Information for Management of Native
Freshwater Fish in Victoria. Victorian Government Printing Office on behalf of Dept. of Conservation
and Environment, Freshwater Fish Management Branch, Arthur Rylah Institute of Environmental
Management, Melbourne, Victoria.
Lake, J. S. (1967). Rearing experiments with five species of Australian freshwater fishes. I.
Inducement to spawning. Australian Journal of Marine and Freshwater Research, 18, 137-153.
Lloyd, N. J. (2001) Spatial Autocorrelation of Benthic Invertebrate Assemblages in two Victorian
Upland Streams. School of Biological Sciences, Monash University, Clayton. pp
LREFSP (2002a) Environmental Flow Determination of the Loddon River Catchment. Part A - Final
Report. Unpublished Report by the Loddon River Environmental Flows Scientific Panel to North
Central CMA and Dept. of Natural Resources and Environment, Melbourne.
LREFSP (2002b) Environmental Flow Determination of the Loddon River Catchment. Part B - Issues
Paper. Unpublished Report by the Loddon River Environmental Flows Scientific Panel to North
Central CMA and Dept. of Natural Resources and Environment, Melbourne.
LREFSP (2005) Review of the Environmental Implications of the Draft Loddon River Bulk
Entitlement. Report by the Loddon River Environmental Flows Scientific Panel to North Central CMA
and Dept. of Sustainability and Environment, Melbourne.
Madsen, J. D., Chambers, P. A., James, W. F., Koch, E. W. and Westlake, D. F. (2001) The
interaction between water movement, sediment dynamics and submersed macrophytes.
Hydrobiologia, 444, 71-84.
Magoulick, D. D. and Kobza, R. M. (2003) The role of refugia for fishes during drought: a review and
synthesis. Freshwater Biology, 48, 1186-1198.
90

Monitoring and evaluation of environmental flow releases in the Campaspe River
Marchant, R., Barmuta, L. A. and Chessman, B. C. (1995) Influence of sample quantification and
taxonomic resolution on the ordination of macroinvertebrate communities from running waters in
Victoria, Australia. Marine and Freshwater Research, 46, 501-506.
Marchant, R., Humphries, P., Rutherfurd, I., Frankenberg, J., McGuckin, J. and Smith, G. (1997)
Scientific Panel Environmental Flow Assessment of the Coliban River Below Malmsbury Reservoir
and the Campaspe River Below Redesdale. Report to the Dept. of Natural Resources and
Environment, Melbourne.
Margulies, D. (1990) Vulnerability of larval white perch, Morone americana, to fish predation.
Environmental Biology of Fishes, 27, 187-200.
Matthews, W. J. (1998) Patterns in Freshwater Fish Ecology. Kluwer, Boston.
McGuckin, J. and Doeg, T. (2000) Investigations of Aquatic Ecosystems of the Loddon Catchment
as part of the Loddon River Health Recovery Plan. Unpublished Report for North Central Catchment
Management Authority.
MDBC (2004) Pilot Audit Technical Reports and Supporting Information - Sustainable Rivers Audit.
Murray-Darling Basin Commission, Canberra.
Minshall, G. W. (1984) Aquatic insect-substratum relationships. In The Ecology of Aquatic Insects.
Resh, V. H. and Rosenberg, D. M., (Ed.). Praeger, New York. pp. 358-400.
Newman, R. M. (1991) Herbivory and detrivory on freshwater macrophytes by invertebrates: a
review. Journal of North American Benthological Society, 10, 89-114.
NSW Fisheries (1997) 'The Australian Code of Electrofishing Practice.' New South Wales Fisheries
Management Publication 1.
Nilsson, C. and Svedmark, M. (2002) Basic principles and ecological consequences of changing
water regimes: riparian plant communities. Environmental Management, 30, 468-480.
O'Connor, N. A. (1991) The effects of habitat complexity on macroinvertebrates colonising stream
substrates in a lowland stream. Oecologia, 85, 504-512.
Pace, M. L., Findlay, S. E. and Lints, D. (1992) Zooplankton in advective environments: the Hudson
River community and a comparative analysis. Canadian Journal of Fisheries and Aquatic Sciences,
49, 1060-1069.
Petts, G. E. and Gurnell, A. M. (2005) Dams and geomorphology: research progress and future
directions. Geomorphology, 71, 27-47.
Poff, N. L., Allan, J. D., Baim, M. B., Karr, J. R. Prestegaard, K. L., Richter, B. D., Sparks, R. E. and
Stromberg, J. C. (1997) The natural flow regime. Bioscience, 47, 769-784.
Pusey, B. J., Arthington, A. H. and Read, M. G. (1993) Spatial and temporal variation in fish
assemblage structure in the Mary River, south-east Queensland: the influence of habitat structure.
Environmental Biology of Fishes, 37, 355-380.
Pusey, B. J., Arthington, A. H. and Read, M. G. (1995) Species richness and spatial variation in fish
assemblage structure in two rivers of the Wet Tropics of northern Queensland, Australia.
Environmental Biology of Fishes, 42, 181-199.
Pusey, B. J. and Kennard, M. J. (1996) Species richness and geographical variation in assemblage
structure of the freshwater fish fauna of the Wet Tropics Region of northern Queensland. Marine and
Freshwater Research, 47, 563-573.
91

Monitoring and evaluation of environmental flow releases in the Campaspe River
Pusey, B. J., Arthington, A. H. and Read, M. G. (1998) Freshwater fishes of the Burdekin River,
Australia: biogeography, history and spatial variation in community structure. Environmental Biology
of Fishes, 53, 303-318.
Pusey, B. J., Kennard, M. J. and Arthington, A. H. (2000) Discharge variability and the development
of predictive models relating stream fish assemblage to habitat in north-eastern Australia. Ecology of
Freshwater Fish, 9, 30-50.
Pusey, B. J. and Arthington, A. H. (2003) Importance of the riparian zone to conservation and
management of freshwater fish: a review. Marine and Freshwater Research, 54, 1-16.
Pusey, B. J., Kennard, M. J. and Arthington, A. H. (2004) Freshwater Fishes of North-eastern
Australia. CSIRO Publishing, Melbourne.
Richardson, A. J., Grownes, J. E. and Cook, R. A. (2004) Distribution and life history of caridean
shrimps in regulated rivers in southern Australia. Marine and Freshwater Research, 55, 295-308.
Roberts, J. and Marston, F. (2000) Water Regime of Wetland and Floodplain Plants in the Murray-
Darling Basin: a Source Book of Ecological Knowledge. Technical Report 30/00, CSIRO Land and
Water, Canberra.
Rosenberg, D. M. and Resh, V. H., Eds. (1993) Freshwater Biomonitoring and Benthic
Macroinvertebrates Chapman and Hall, New York.
Ryan, T., Webb, A., Lennie, R. and Lyon, J. (2001) Status of Cold Water Releases from Victorian
Dams. Report produced for Dept. of Natural Resources and Environment, Melbourne.
Schlosser, I. J. (1982) Fish community structure and function along two habitat gradients in a
headwater stream. Ecological Monographs, 52, 395-414.
Sharpe, A. K. and Quinn, G. P. (2004) Monitoring Environmental Flows in the Wimmera and Glenelg
Rivers. Final Report prepared for Dept. of Sustainability and Environment, Wimmera CMA and
Glenelg-Hopkins CMA, Sinclair Knight Merz and CRC for Freshwater Ecology, Melbourne.
Simpson, J., Norris, R., Barmuta, L. and Blackman, P. (1997) Australian River Assessment System:
National River Health Program Predictive Model Manual.
http://enterprise.canberra.edu.au/Databases/AusRivAS/,
SKM (2001) Environmental Flow Study of the Wimmera River System: Issues Paper. Report to
Wimmera CMA, Sinclair Knight Merz, Melbourne.
SKM (2002) Stressed Rivers Project - Environmental Flows Study. Wimmera River System. Final
Report to Wimmera CMA and Dept. of Natural Resources and Environment, Sinclair Knight Merz,
Melbourne.
SKM (2003a) Macalister River Environmental Flows Assessment. Final Report to West Gippsland
CMA, Sinclair Knight Merz, Melbourne.
SKM (2003b) Stressed Rivers Project - Environmental Flows Study. Glenelg River System. Final
Report to Glenelg-Hopkins CMA and Dept. of Natural Resources and Environment, Sinclair Knight
Merz, Melbourne.
SKM (2003c) Wimmera Bulk Entitlement Conversion Project Environmental Flows Study. Report to
Dept. of Natural Resources and Environment, Sinclair Knight Merz, Melbourne.
SKM (2005) Campaspe River Environmental FLOWS Assessment. Site Paper. Report to North
Central CMA, Sinclair Knight Merz, Melbourne.
SKM (2006a) Campaspe River Environmental FLOWS Assessment. Issues Paper. Report to North
Central CMA, Sinclair Knight Merz, Melbourne.
92

Monitoring and evaluation of environmental flow releases in the Campaspe River
93
SKM (2006b) Campaspe River Environmental FLOWS Assessment. Flow Recommendations.
Report to North Central CMA, Sinclair Knight Merz, Melbourne.
SKM (2006c) Environmental flow monitoring water quality analysis. Data analysis. Report to Glenelg
Hopkins CMA, Sinclair Knight Merz, Melbourne.
Steiger, J., Gurnell, A. M. and Goodson, M. (2003) Quantifying and characterising contemporary
riparian sedimentation. River Research and Applications, 19, 335-352.
Underwood, A. J. (1991) Beyond BACI: experimental designs for detecting human environmental
impacts on temporal variations in natural populations. Australian Journal of Marine and Freshwater
Research, 42, 569-587.
Underwood, A. J. (1992). Beyond BACI: the detection of environmental impacts on populations in
the real, but variable, world. Journal of Experimental Marine Biology and Ecology, 161, 145-178.
Underwood, A. J. (1994) On beyond BACI: sampling designs that might reliably detect
environmental disturbances. Ecological Applications, 4, 3-15.
Vannote, R. L., Minshall, G. W., Cummins, K. W., Sedell, J. R. and Cushing, C. E. (1980) The river
continuum theory. Can. J. Fish. Aquat. Sci., 37, 130-137.
van Wagner, C. E. (1968). The line intersect method in forest fuel sampling. Forest Science, 14, 20-
26.
Vietz, G. J., Stewardson, M. J. and Finlayson, B. L. (2006) Flows that form: the hydromorphology of
concave-bank bench formation in the Ovens River, Australia. IAHS Publication, 306, in press.
Wallace, J. and Benke, A. (1984). Quantification of wood habitat in subtropical coastal plain streams.
Canadian Journal of Fisheries and Aquatic Sciences, 41, 1643-1652.
Warwick, R. M. (1988) Analysis of community attributes of the macrobenthos of
Frierfjord/Langesundfjord at taxonomic levels higher than species. Marine Ecology-Progress Series,
46, 167-170.
Warwick, R. M. (1988) The level of taxonomic discrimination required to detect pollution effects on
marine benthic communities. Marine Pollution Bulletin, 19, 259-268.
Weatherhead, M. A. and James, M. R. (2001) Distribution of macroinvertebrates in relation to
physical and biological variables in the littoral zone of New Zealand lakes. Hydrobiologia, 462, 115-
129.
Werren, G. and Arthington, A. H. (2002) The assessment of riparian vegetation as an indicator of
stream condition, with particular emphasis on the rapid assessment of flow-related impacts. In
Landscape Health of Queensland. Playford, J., Shapcott, A. and Franks, A., (Ed.). Royal Society of
Queensland, Brisbane. pp. 194-222.
Wright, J. F. (1995) Development and use of a system for predicting the macroinvertebrate fauna in
flowing waters. Australian Journal of Ecology, 20, 181-191.

Monitoring and evaluation of environmental flow releases in the Campaspe River
Appendix 1: Environmental Flow Recommendations for
Reaches of the Campaspe River (from SKM 2006b)
Specific objectives for each reach are described in the flows study (SKM 2006b) and are then
referred to in the objective column of tables 13-16. Consultants developing and implementing river
specific monitoring programs will need to refer to the original flow study to ensure that the methods
develop sample the correct response variable (e.g. spawning of a particular fish species in response
to a particular flow component).
Table 13. Environmental flow recommendations for Reach 1, Coliban River, Malmsbury
Reservoir to Lake Eppalock
94

Monitoring and evaluation of environmental flow releases in the Campaspe River
Table 14. Environmental flow recommendations for Reach 2, Lake Eppalock to Campaspe
Weir
95

Monitoring and evaluation of environmental flow releases in the Campaspe River
Table 15. Environmental flow recommendations for Reach 3, Campaspe Weir to Campaspe
Siphon
Table 16. Environmental flow recommendations for Reach 4, Campaspe Siphon to River
Murray
96

Monitoring and evaluation of environmental flow releases in the Campaspe River
Appendix 2: Selection of Conceptual Models
Conceptual models provide a means by which we can represent our understanding and beliefs
about how a particular system functions. This explicit representation can also help to reduce
confusion between various stakeholders, and to highlight knowledge gaps in our understanding of a
system. For this report, conceptual models were developed to make the link between the various
flow components to be supplied and expected ecosystem responses, and from this the appropriate
variables for monitoring were specified. The models were thus developed specifically for the purpose
of identifying variables for monitoring, and were never meant to be a complete or ‘correct’
representation of the system, and were. That is not to say, however, that they were not well-
researched. The models presented in this report were synthesised from the literally hundreds of
conceptual models presented in the various environmental flows reports, and from the published
literature on flow-ecosystem relations. The models were then peer reviewed by experts in the
respective fields, and revised according to these reviews. Further revision is also possible, and
these models should not be considered as a final product. As information is collected through the
monitoring program, it may become apparent that some previously accepted aspect of a conceptual
model is incorrect. Similarly, additional important links between flow and response may become
apparent that can better explain why a particular variable responds to a given flow component. At
this stage, however, there were a large number of conceptual models and potential hypotheses that
could be tested for the Campaspe River and at a statewide level. The information and processes
used to synthesise conceptual models are presented below.
A number of criteria were used to consolidate the conceptual models and hypotheses that would be
addressed, both for the Campaspe River and at the statewide level. Conceptual models and
hypotheses should:
1. Be scientifically ‘sound’ - well-founded and supported by appropriate theoretical or empirical
data from scientific studies and/or expert opinion
2. Involve responses to recommended environmental flows that are detectable - expected
responses must be of sufficient magnitude to be detectable within a useful management
timeframe (nominally 10 years)
3. Address questions that are relevant to the Victorian River Health Strategy (VHRS) (see Box
below)
4. Where possible, have general applicability to multiple reaches within the Victorian Rivers
receiving a EWR.
5. Be realistic, given the quantity of water available for implementing the recommended
environmental flow releases.
6. Include components of the flow regime that can most feasibly be returned to a more natural
pattern using the environmental flow recommendations.
7. Acknowledge potential constraints on ecosystem response because of river-specific
characteristics and/or regulation activities (e.g. cold water releases from large dams).
8. Acknowledge potentially adverse outcomes associated with implementing the recommended
environmental flow releases (e.g. blackwater events).
9. Availability of relevant historical data.
97

Monitoring and evaluation of environmental flow releases in the Campaspe River
Criteria 1-3 were used to develop a preliminary list of priority conceptual models/hypotheses
applicable to all the 8 rivers (process described below).
The Victorian River Health Strategy
The Victorian River Health Strategy (VHRS) is the overarching framework for making decisions
about the management and restoration of Victoria’s rivers (DNRE 2002b). The VHRS is informed
by an ecological understanding of ‘river health’ (Section A), guided by aspirations or a vision for
Victoria’s rivers (Section B), governed by a particular management philosophy towards restoration
(Section C) and assessed with the aid of Statewide targets (Section D). Conceptual models and
hypotheses to be tested must be demonstrably relevant and consistent with the considerations
described in Sections A-D.
Section A VHRS Understanding of ‘river health’ (adapted from DNRE 2002b)
VHRS clearly states that ‘river health’ goes beyond water quality and the flora and fauna present in
the river. It explicitly recommends that proper understanding of ‘river health’ should take into
account the:
1. Diversity of habitats and biota;
2. Effectiveness of linkages; and
3. Maintenance of ecological processes.
(project team emphasis in bold)
Three key processes are highlighted in relation to the ‘maintenance of ecological processes:
1. Energy and nutrient dynamics, including primary production and microbial respiration
which maintain food webs within the entire ecosystem.
2. Processes which maintain animal and plant populations, such as reproduction and
regeneration, dispersal, migration, immigration and emigration.
3. Species interactions, which can affect community structure. These include predator-prey,
host-parasite and competition relationships.
Section B VHRS Vision for Victoria’s rivers (adapted DNRE 2002b)
The VHRS Vision for Victoria’s Rivers is based on ecological sustainability and envisages rivers
that
1. support a diverse array of indigenous plants and animals within their waters and across
their floodplains
2. are flanked by a mostly continuous and broad band of native riparian vegetation
3. have flows that rise and fall with the seasons, inundating floodplains, filling billabongs and
providing a flush of growth and return of essential nutrients back to the river
4. have water quality that sustains critical ecological functions
5. have native fish and other species moving freely along the river and out of the floodplains
and billabongs to feed and breed during inundation
6. replenish productive estuaries or terminal lakes
Section C VHRS Management philosophy towards restoration (adapted from DNRE 2002b)
In the VHRS Strategy Background on the Management Drivers of River Health, a philosophy
towards restoration is outlined and it reflects a realistic, pragmatic approach that
a) postulates that river systems have a tiered number of viable, functioning, self-sustaining
ecological states and proposes that the aim of management may be to prevent transition
from one state to a less desirable one, rather than to restore the system to its original
condition
98

Monitoring and evaluation of environmental flow releases in the Campaspe River
b) considers that all systems have some ecological values; even heavily degraded systems
always have some functioning aspect of the ecology present and some ecological values
c) predicts that the effort that is required to restore a heavily degraded system will be very
much larger than that required to restore a system in a reasonable condition
d) recognizes that the impaired conditions in some heavily degraded systems may be
irreversible
Section D VHRS Statewide Targets (adapted from DNRE 2002b)
In working towards the vision, the following targets for river protection and restoration will be used
to measure progress across the State.
All Heritage Rivers to be maintained at least to their current condition and their Heritage River
values protected.
By 2005:
• An increase in length of river accessible to native fish by an additional 2000 km.
• Significant improvement in floodplain linkages in ten areas of national and/or State significance.
• All rivers with either sustainable catchment limits or negotiated environmental flows in place.
• Report on the second benchmarking of the environmental condition of Victorian rivers.
• A quarter of agricultural production produced from natural resources that are managed within their
capacity. By 2015, this will increase to half of agricultural production (as stated in Victoria’s Salinity
Management Framework).
By 2011: (project team emphasis in bold)
• An improvement in the status of designated freshwater-dependent focal species.
• Significant improvements achieved in environmental flow regimes of 20 high value river
reaches currently flow stressed.
• 4800 km of rivers with improvement of one rating in the measurement of riparian condition.
• An increase of 7000 ha of riparian areas under management agreements.
• 600 km of rivers where in-stream habitat has been reinstated.
• 95% of all highland and upland and 60% of all lowland monitoring sites will meet SEPP
environmental quality objectives.
• 1000 high value public assets provided with appropriate level of protection.
By 2021:
• One major representative river reach in ecologically healthy condition in each major river class.
• An increase of 3000 km in the length of rivers in excellent or good condition.
Progress towards the achievement of these targets will be measured through regular reporting on
river protection and restoration activities, and through regular resource condition monitoring.
Criteria 1 to 3 were used to prioritise a sub-set of these for more detailed consideration as the basis
for the VEFMAP. Each conceptual model/hypotheses was rated against these three criteria. Each
hypothesis was given a score from 1 to 3 (1: weak, 3: strong) for each criteria. Our scores are listed
in Table 17 and were tallied to identify a preliminary list of the conceptual models we would use in
designing the environmental flow monitoring program.
We acknowledge the subjectivity of this scoring procedure. Our intent is to prioritise the
hypotheses/models in a systematic and transparent way. The table of scores can also be used in the
upcoming workshop as a focus for debate around the logic of this selection It should be noted that
this exercise was carried out prior to the first round review of this document by our subject-matter
99

Monitoring and evaluation of environmental flow releases in the Campaspe River
specialists and scientists who had served on Scientific Panels involved in producing the
environmental flow studies for the various river systems.
Conceptual models/hypotheses relating to desiccation disturbance associated with cease to flow
periods (Attribute 3, Table 18), the effects of summer low flows on instream habitats (Attribute 5,
Table 18), and water quality (Attribute 7, Table 18) were excluded from this exercise. These
conceptual models are a component of a concurrent study on the ecology and hydrology of
temporary streams in Victoria (Nick Bond, Monash University). It was decided that the present
project should draw upon the results of that study rather than duplicating the effort.
Table 17. Scores of conceptual models/hypotheses for selection criteria 1-3.
Conceptual Model/Hypotheses Score
for: Total
Releva
nce to
VHRS
Strengt
h of a
priori
support
Expected
magnitud
e
response
within 10
year
timeframe
1 Biogeochemical Processes 3 2 3 8
2 Geomorphic Processes
- channel maintenance
- scouring and removing sediment from infilled
pools
2
2
3
3
1
3
6
8
3 Ecological Process – Disturbance (Bankfull flows) 1.5 2.5 3 7
4 Habitat Processes
- reinstatement of more natural Winter-Spring
baseflows will result in inundation of physical
features and will increase habitat availability
- increase in habitat availability → abundance of
biota (macroinvertebrates, fish)
3
3
3
2
3
2
9
7
5 Habitat Processes
-Freshes, High Flows and Bankfull flows can flush
and remove fine sediments and organic matter from
habitat substrates and improve habitat for biota
2.5
1
2
5.5
6 Biofilms
- High Flows and Winter-Spring-Summer
Freshes provide scouring flows over biofilm
habitats, acting as a disturbance mechanism
that helps maintain species composition and
health
2
3
3
8
7 Aquatic & Riparian Vegetation (Attributes 9 & 10
in Table 18) 3 2.5 3 8.5
8 Invertebrates
-flow connectivity provided by reinstating more
natural Winter-Spring Baseflows and Summer
Freshes will enable invertebrate drift and movement
between different instream habitats
1.5
2
1
4.5
9 Invertebrates
-Flows which inundate sediments in higher portions
of the channel provide a stimulus for invertebrate
hatching
1.5
3
3
7.5
10 Fish – Habitat
-adequate baseflows year round (a) maintains
100

Monitoring and evaluation of environmental flow releases in the Campaspe River
habitat availability and inundates large woody
debris which provides food sources and shelter; (b)
maintains/increases the amount of deepwater
habitat available for large-bodied fish
3 3 1.5 7.5
11 Fish – Movement
-reinstatement of more natural Winter-Spring
Baseflows will provide sustained longitudinal
connectivity for fish movement including permanent
movement of large-bodied fish throughout the river
reach leading up to the breeding season. Summer
Freshes will increase water depth over low-lying
zones and increase longitudinal connectivity,
temporarily facilitating greater movement of fish
between different instream habitats.
- Winter and Spring High Flows and Freshes and
Bankfull Flows may provide the cue which triggers
movement and/or migration in some native fish
species
3
2.5
2
2
1
1
6
5.5
12 Fish – Maturation, Reproduction/Spawning,
Recruitment
- Low Flows may be important for recruitment of
some native fish in lowland rivers because Low
Flows maintain or increase the availability of
slow-water habitats which are important as
refuge and rearing habitats for larval and
juvenile fish. High water velocity over summer-
autumn displace eggs and larvae from
spawning and rearing habitat thus limiting
recruitment.
- Winter-Spring Freshes may provide spawning
cues for freshwater diadromous fish such as
Australian Grayling and Long-finned and Short-
finned Eels.
3
3
3
2
3
3
9
8
13 Fish – Exotic species management
- Summer/Spring Low Flows can expose banks and
beds leading to the drying of Carp eggs and
contributing to exotic fish management. High
summer flows and less annual flow variability also
provides habitat conditions favourable for
introduced species such as Carp and Gambusia.
3
1
3
7
Conceptual models/hypotheses which scored ≥8 in our trial application of the selection criteria are
shown in bold in Table 17. It was felt that although the conceptual model/hypotheses associated with
biofilms rated well, there is a general lack of familiarity and appreciation of the role of biofilms in
riverine systems and this would present difficulties in communicating their relevance to management
stakeholders and the community at large. The surprise finding of invertebrates no being amongst the
top-rated conceptual models/hypotheses turned out to be an artefact of the way the information in
Table 18 is organized, with the conceptual model relating flow regime to habitat availability for
macroinvertebrates being subsumed in the conceptual model for habitat processes. It was therefore
decided that macroinvertebrates should be included in the preliminary shortlist and a more cohesive
conceptual model linking flows, habitat availability, connectivity, drift and stimulus for hatching would
be developed for macroinvertebrates. Hence, our preliminary list included six conceptual models
relating to 1) biogeochemical processes; 2) geomorphic processes; 3) habitat processes; 4) aquatic
and riparian vegetation; 5) macroinvertebrates and 6) fish – spawning and recruitment.
The underlying basis for each of the above conceptual models was reviewed by scientists who are
recognised experts in the area of ecosystem monitoring and evaluation, including a number who
101

Monitoring and evaluation of environmental flow releases in the Campaspe River
102
were on the Scientific Panels that developed flow recommendations for many of the eight rivers in
the program. The models were also explored at a workshop attended by scientific experts, DSE staff
and managers (see acknowledgements). The rationale of the models and how they were to be used
was strongly endorsed at the workshop. It was agreed that the biogeochemical processes model
would not be pursued further, as there was insufficient knowledge of what represents ‘target’
conditions for processes such as production and respiration (making it difficult to interpret any
results). In addition, the macroinvertebrate model was subsumed into the habitat responses model.
A full model of macroinvertebrate responses to flow enhancement was very complex, and unlikely to
be of use in designing a monitoring program. However, we noted that any environment with
sufficient macroinvertebrate habitat would most likely also include the other requirements for
successful macroinvertebrate populations (see § 2.4).
Explanatory notes for Table 18:
1. In examining the flow recommendations for individual rivers, it became evident that there was
some variation (and inconsistency) in the way the FLOWS method flow component terms were
used in the original environmental flow study reports. For instance, High Flows which refers to
the persistent increase in seasonal baseflow that remains in the channel and which does not fill
the channel to bankfull, was sometimes also referred to as ‘Winter Low Flow’ or ‘Winter
Baseflow’. This can lead to some confusion, for example, recommendations to meet the
‘Winter Low Flow’ requirements might be misconstrued as a recommendation that the winter
flow be reduced, when in fact, the intent is to recommend protection or reinstatement of a level
of baseflows appropriate the eight rivers which are southern winter-rainfall dominated systems.
The wording in Table 2 has, in some instances, been amended from that in the original
environmental flow reports to help minimise linguistic uncertainty (e.g. recommendations for
‘winter low flows’ have been amended to ‘protection or reinstatement of more natural levels of
winter baseflows’).
2. Each statement within the description of the conceptual model and generic predictions (column
2) is accompanied by letters in square brackets, [ ], which refer to the respective environmental
flow studies of the various river systems from which the statement is taken. {B – Broken; Go –
Goulburn; T – Thomson; M – Macalister; L – Loddon; C – Campaspe; W – Wimmera; Gl –
Glenelg}
3. References in black are citations provided in the environmental flow study reports for the
various river systems.
4. Statements in blue are comments or additional notes from the literature.
5. References in red sans serif font are potentially relevant references from the literature.
6. An earlier draft of Table 2 was reviewed by subject-matter specialists as well as scientists who
had served on Scientific Panels involved in producing the environmental flow studies for the
various river systems. These reviewers were Mark Kennard, Alison King, Sam Lake, Terry
Hillman, Leon Metzeling and Jane Roberts. Statements in green are comments from the
reviewers. References in violet sans serif font are relevant references suggested by these
reviewers.

Monitoring and evaluation of environmental flow releases in the Campaspe River
Table 18. Summary of conceptual models, generic prediction(s) and references cited for flow relationships with key attributes of riverine
ecosystems.
Attribute Conceptual Model & Generic Prediction(s) Reference(s) Cited
1 Biogeochemic
al Process –
Organic
Matter &
Nutrient
Dynamics
During periods of flow cessation or summer low flows,
exposure of the streambed allows accumulation and drying
of terrestrial organic matter in dry areas of the channel such
as benches. Drying and subsequent rewetting facilitates the
decomposition and processing of this organic matter and
produces a fresh pool of nutrient and carbon inputs for the
system. [D, M]
Freshes can remove organic material from unproductive
areas such as sand beds and encourage it to accumulate in
lateral root zones and woody debris. [T]
Winter/Spring Freshes, High Flows, Bankfull Flows and
Overbank Flows can inundate higher portions of the channel
such as banks and benches and entrain organic matter that
has accumulated in the elevated channel features and
terrestrial channel sections. This will provide inputs of
dissolved and fine particulate organic matter to maintain
nutrient/carbon cycling inputs to the river. [D,M,T,L] Organic
matter and nutrients in this form can be used by microbes,
algae and macrophytes1 [Go, C] And their influx may result in
higher rates of productivity and respiration on benches than
in the main river channel, although it is uncertain how long
this effect persists. [Go]
In the case of Overbank Flows which result in inundation of
the floodplain, there may be significant carbon returns to the
river after a period of significant production. [D]
1Organic matter and nutrients in this form can also be used
by microfauna and zooplankton (Lake, pers.comm.)
No reference cited for the conceptual model. [M,T,C]
Mitchell, A. and Baldwin, D. (1998) Effects of desiccation/oxidation on the
potential of bacterially mediated P-release from sediments. Limnology and
Oceanography, 43, 481-487. [Go]
Baldwin, D.S. and Mitchell, A.M. (2000) The effects of drying and re-
flooding on the sediment and soil nutrient dynamics of lowland river-
floodplain systems: a synthesis. Regulated Rivers: Research &
Management, 16, 457-467. [D, Go]
Nielsen, D.L. and Chick, A.J. (1997) Flood-mediated changes in aquatic
macrophyte community structure. Marine and Freshwater Research, 48,
153-157. [D]
Robertson et al. (1990) – cited in text [D], but not in Reference list.
Briggs, S.V., Maher, M.T. and Carpenter, S.M. (1985) Limnological studies
of waterfowl habitat in south-western New South Wales. I. Water chemistry.
Marine and Freshwater Research, 36, 59-67.
Thoms, M.C. and Sheldon, F. (1997) River channel complexity and
ecosystem processes: the Barwon-Darling River (Australia). In Klomp, N.
and Lunt, I. (Eds.). Frontiers in Ecology. Elsevier Science Ltd., Oxford, UK,
pp. 193-205.
Molles, M.C., Crawford, C.S., Ellis, L.M., Valett, H.M. and Dahm, C.N.
(1998) Managed flooding for riparian ecosystem restoration. BioScience,
48, 749-756.
Ellis, L.M., Molles, M.C. and Crawford, C.S. (1999) Influence of
experimental flooding on litter dynamics in a Rio Grande riparian forest,
New Mexico. Restoration Ecology, 7, 193-204.
Tockner, K., Pennetzdorfer, D., Reiner, N., Schiemer, F. and Ward, J.V.
(1999) Hydrological connectivity and the exchange of organic matter and
nutrients in a dynamic river-floodplain system (Danube, Austria).
Freshwater Biology, 41, 521-535.
Robertson, A.I., Bacon, P. and Heagney, G. (2001) The responses of
floodplain primary production to flood frequency and timing. Journal of
Applied Ecology, 38, 126-136.
103

Monitoring and evaluation of environmental flow releases in the Campaspe River
Hein, T., Baranyi, C., Herndl, G.J., Wanek, W. and Schiemer, F. (2003)
Allochthonous and autochthonous particulate organic matter in floodplains
of the River Danube: the importance of hydrological connectivity.
Freshwater Biology, 48, 220-232.
Pinay, G., Clément, J.C. and Naiman, R.J. (2002) Basic principles and
ecological consequences of changing water regimes on nitrogen cycling in
fluvial systems. Environmental Management, 30, 481-491.
Valett, H.M., Baker, M.A., Morrice, J.A., Crawford, C.S., Molles, M.C.,
Dahm, C.N., Moyer, D.L., Thibault, J.R. and Ellis, L.M. (2005)
Biogeochemical and metabolic responses to the flood pulse in a semiarid
floodplain. Ecology, 86, 220-234.
Ward, J.V., Tockner, K., Arscott, D.B and Claret, C. (2002) Riverine
landscape diversity Freshwater Biology 47: 517-539.
Ward, J.V. 1989. Riverine-wetland interactions. Freshwater Wetlands and
Wildlife, 1989. Conf –8603101, DOE Symposium Series No. 61. Sharitz,
R.R. and Gibbons, J.W. (Eds.). USDOE Office of Scientific and Technical
Information, Oak Ridge, Tennessee, USA pp385-400.
Tockner, K., Malard, F. and Ward, J.V. (2000) An extension of the flood
pulse concept. Hydrological Processes 14: 2861-2883.
Tockner, K., Bunn, S., Gordon, C,. Naiman, R.J., Quinn, G.P. and Stanford,
J.A. (2006) Floodplains: critically threatened ecosystems. In “State of the
Worlds Waters” Polunin, N (Ed.), Cambridge University Press, Cambridge
UK (in press).
Bunn, S.E., Davies, P.M. and Winning, M. (2003) Sources of organic
carbon supporting the food web of an arid zone floodplain river. Freshwater
Biology, 48, 619-635.
Boulton, A.J. and Lake, P.S. (1992) Benthic organic matter and detrivorous
invertebrates in two intermittent streams in south-eastern Australia.
Hydrobiologia, 241, 107-118.
Briggs, S.V., Maher, M.T and Tongway, D.J. (1993) Dissolved and
particulate organic carbon in two wetlands in southwestern New South
Wales, Australia. Hydrobiologia, 264, 13-19.
Baldwin, D.S. (1999) Dissolved organic matter and phosphorus leached
from fresh and ‘terrestrially’ aged river red gum leaves: implication for
assessing river-floodplain interactions. Freshwater Biology 41, 675-685.
104

Monitoring and evaluation of environmental flow releases in the Campaspe River
2 Geomorphic
Process Bankfull Flows are important geomorphologically in shaping
and maintaining river and distributary channels.1 [D]. Bankfull
Flows can reform channels by scouring and sediment
transport and help to maintain/rehabilitate channel form. For
instance, by scouring and effecting the removal of vegetation
which has encroached into the channel [M,T]; or by scouring
and removing sediment from in-filled pools [L]; or by
mobilising sand build-up within channels [Gl]; or by
depositing sediments on benches [T] and constricting sandy
channels [C] (the rationale here is as follows: granite
catchments tend to have sand-dominated sediment loads.
Sand ‘slugs’ and flat sand sheets are unstable, create
inimical conditions for macroinvertebrates and produce
conditions of limited variability in water depth. Hence, in such
rivers, possible enhancements to instream habitat include
measures to store as much as possible of the sand in lateral
benches within the river; to encourage deepening of the
channel, to re-establish a more stable gravel substrate and
to re-establish greater heterogeneity in depth and more
varied pool-riffle morphology. It was proposed in [C] that this
might be achieved by narrowing the channel and
encouraging the development of a sinuous course by
establishing vegetation on point bars.)
1A qualifier should be added to indicate that this applies to
constrained streams (Lake, pers.comm.)
No reference cited for the conceptual model. [M,T,L,Gl,C]
Leopold, L.B. and Maddock, T. (1953) The Hydraulic Geometry of Stream
Channels and Some Physiographic Implications. United States Geological
Survey Professional Paper 252, United States Government Printing Office,
Washington. [D]
Wolman, M.G. and Miller, J.P. (1960) Magnitude and frequency of forces in
geomorphic processes. Journal of Geology, 60, 54-74.
Sigafoos, R.S. (1964) Botanical evidence of floods and flood-plain
deposition. US Geological Survey Professional Paper485-A.
Yanosky, T.M. (1982) Effects of flooding upon woody vegetation along
parts of the Potomac river flood plain. US Geological Survey Professional
Paper 1206.
Osterkamp, W.R. and Costa J.E. (1987) Changes accompanying n
extraordinary flood on a sandbed stream. In Mayer, L. and Nash, D. (Eds.).
Catastrophic Flooding. Allen and Unwin, Boston, M.A., pp 201-224.
Johnson, W.C. (1994) Woodland expansion in the Platte river, Nebraska:
patterns and causes. Ecological Monograph, 64, 45-84.
Petts, G.E. (1996) Water allocation to protect river ecosystems. Regulated
Rivers: Research & Management, 12, 353-365.
Kondolf, G. and Wilcock, P.R. (1996) The flushing flow problem: defining
and evaluating objectives. Water Resources Research, 32, 2589-2599.
Pitlick, J. and Van Steeter, M.M. (1998) Geomorphology and endangered
fish habitats of the upper Colorado River 2. Linking sediment transport to
habitat maintenance. Water Resources Research, 34, 303-316.
Friedman, J.M. and Auble, G.T. (1999) Mortality of riparian box elder from
sediment mobilization and extended inundation. Regulated Rivers:
Research & Management, 15, 463-476.
Bond, N.R. (2004) Spatial variation in fine sediment transport in small
upland streams: the effects of flow regulation and catchment geology. River
Research and Applications 20: 705-717.
Costa, J.E. and O’Connor, J.E. (1995) Geomorphically effective floods. In
Natural and anthropogenic influences in fluvial geomorphology. The
Wolman Volume. Costa, J.E., Miller, A.J., Potter, K.W. and Wilcock, P.R.
(Eds.) American Geophysical Union, Washington, D.C. pp 45-56.
Matthaei, C.D., Peacock, K.A. and Townsend, C.R. (1999) Scour and fill
patterns in a New Zealand stream and potential implications for
invertebrate refugia. Freshwater Biology 42: 41-57.
105

Monitoring and evaluation of environmental flow releases in the Campaspe River
Meidl, E-B and Schönbon, W. (2004) How structure controls assembly in
the hyporheic zone of rivers and streams: colmation as a disturbance. In
Assembly Rules and Restoration Ecology. Bridging the Gap between
Theory and Practice. Temperton, V.M., Hobbs, R.J., Nuttle, T. and Halle, S.
(Eds.) Island Press, Washington, USA. Pp 389-408.
Wood, P.J. and Armitage, P.D. (1997) Biological effects of fine sediment in
the lotic environment. Environmental Management 21: 203-217.
3 Ecological
Process –
Disturbance
(Drying)
During periods of flow cessation or summer low flows, the
river may contract to a series of isolated pools, or portions of
the channel may dry out.
(a) Biota in these pools are likely to be subjected to
intensified predation and physicochemical stresses
(eg. Low DO concentrations).1 [D]
(b) Exposure of large areas of the streambed as a
disturbance mechanism which resets successional
processes for macroinvertebrate and vegetation
communities2 [M]
Desiccation disturbance prevents the system from being
dominated by any particular group of organisms.3 [W, Gl]
And in particular, any macroinvertebrate species as many
macroinvertebrate species would be reduced in their
abundance and distribution over the dry period. [Gl]
Desiccation disturbance maintains aquatic and riparian
species characteristic of dryland river systems where flow
cessation and summer low flows naturally occur. Biota in
these arid environments have special physiological or
behavioural adaptations that allow them to persist in harsh
conditions in locations which they might otherwise be
displaced by dominant but less tolerant species. In the short-
term there may be localised extinction of certain species.
And in the long-term, changes in diversity and biomass.
Recolonisation of “stressed” habitats upon flow restoration
should be feasible provided there are effective refuges for
biota during cease-to-flow periods. [D]
1competition should also be added to the list of stresses
No reference cited for the conceptual model. [W, Gl]
Puckridge, J.T., Walker, K.F. and Costelloe, J.F. (2000) Hydrological
persistence and the ecology of dryland rivers. Regulated Rivers: Research
& Management, 16, 385-402. [D]
Humphries, P. and Lake, P.S. (2000) Fish larvae and the management of
regulated rivers. Regulated Rivers: Research & Management, 16, 421-432.
[D]
Allan, J.D. (1975) The distributional ecology and diversity of benthic insects
in Cement Creek Colorado. Ecology, 55, 1040-1053. [D]
Boulton, A.J. and Lloyd, L.N. (1991) Macroinvertebrate assemblages in
floodplain habitats of the lower River Murray, South Australia. Regulated
Rivers: Research & Management, 6, 183-201. [D]
Townsend, C.R. and Hildrew, A.G. (1976) Field experiments on the drifting,
colonization and continuous redistribution of stream benthos. Journal of
Animal Ecology, 45, 759-772. [D]
Jowett, I.G. and Duncan, M.J. (1990) Flow variability in New Zealand rivers
and its relationship to in-stream habitat and biota. New Zealand Journal of
Marine and Freshwater Research, 24, 305-317. [D]
Cooper et al. (1990) – cited in text [D], but not in Reference list.
Boulton, A.J. and Suter, P.J. (1986) Ecology of temporary streams – an
Australian perspective. In Limnology in Australia. De Dekker, P. and
Wiliams, W.D. (Eds.). Commonwealth Scientific Industrial Research
Organization, Junk, Dordrecht, pp. 313-327.
Resh, V.H., Brown, A.V., Covich, A.P., Gurtz, M.E., Li, H.W., Minshall,
G.W., Reice, S.R., Sheldon, A.L., Wallace, J.B. and Wismaar, R.C. (1988)
The role of disturbance in stream ecology. Journal of the North American
Benthological Society, 7, 433-455.
Lake, P.S. (1990) Disturbing hard and soft-bottom communities: a
comparison of marine and freshwater environments. Australian Journal of
106

Monitoring and evaluation of environmental flow releases in the Campaspe River
(Lake, pers.comm.)
2Drying may be a stress for some species (especially
obligate aquatics) but is an opportunity for transient
terrestrial species to establish or expand, or for certain plant
species to regenerate on bars and benches (Roberts,
pers.comm.)
3Depends on the place and the species involved. For
example, in periods of low flow in south-west US streams,
Tamarix, a non-native invader, proliferates and becomes
dominant. (Shafroth et al. (2005) Control of Tamarix in the
western United States: implications for water salvage, wildlife
use, and riparian restoration. Environmental Management,
35, 231-246. – Lake, pers.comm.)
Ecology, 15, 477-488.
Closs, G.P. and Lake, P.S. (1996) Drought, differential mortality and the
coexistence of a native and introduced fish species in a south east
Australian intermittent stream. Environmental Biology of Fishes, 47, 17-26.
Boulton, A.J. (2003) Parallels and contrasts in the effects of drought on
stream macroinvertebrate assemblages. Freshwater Biology, 48, 1173-
1185.
Bond, N. and Lake, P.S. (2003b). Local habitat restoration in streams:
constraints on the effectiveness of restoration for stream biota. Ecological
Management and Restoration, 4, 193-198.
Lake, P.S. (2003) Ecological effects of perturbation by drought in flowing
waters. Freshwater Biology, 48, 1161-1172.
Bond N. and Lake P.S. (2004). Disturbance regimes and stream
restoration: the importance of restoring refugia. Fourth Australian Stream
Management Conference, Cooperative Research Centre for Catchment
Hydrology, Melbourne, Launceston, Australia, 2004.
Bond N. and Lake P.S. (2005a). Ecological restoration and large-scale
ecological disturbance: the effects of drought on the response by fish to a
habitat restoration experiment. Restoration Ecology, 13, 39-48.
Bond N. and Lake P.S. (2005b). Of sand slugs and fish restoration: the
Granite Creeks Saga. In Lintermans M., Cottingham P. and O’Connor R.
(Eds). Proceedings of the Fish Habitat Restoration Workshop, Albury
February 2004, Murray-Darling Basin Commission, Canberra.
Lake P.S. (2005) Perturbation, restoration and seeking ecological
sustainability in Australian flowing waters. Hydrobiologia, 552, 109-120.
Cowx, I.G., Young, W.O. and Hellawell, J.M. (1984) The influence of
drought on the fish and invertebrate populations of an upland stream in
Wales. Freshwater Biology, 14, 165-177.
Nilsen, E.T., Sharifi, M.R. and Rundel, P.W. (1984) Comparative water
relations of phreatophytes in the Sonoran Desert of California. Ecology, 65,
767-778.
Jenkins, K.M. and Boulton, A.J. (2003) Connectivity in a dryland river:
short-term aquatic microinvertebrate recruitment following floodplain
inundation. Ecology 84: 2708-2723.
Lake, P.S. (2000) Disturbance, patchiness and diversity in streams. Journal
of the North American Benthological Society 19: 573-592.
Stanley, E.H., Fisher, S.G. and Grimm, N.B. (1997) Ecosystem expansion
and contraction in streams. BioScience 47: 427-435.
107

Monitoring and evaluation of environmental flow releases in the Campaspe River
Stanley, E.H, Fisher, S,G, and Jones, J.B. (2004) Effects of water loss on
primary production: a landscape-scale model. Aquatic Sciences 66: 130-
138.
Capon, S.J. and Brock, M.A. (2006) Flooding, soil seed bank dynamics and
vegetation resilience of a hydrologically variable desert floodplain.
Freshwater Biology 51: 206-223.
Beche, L.A., McElravy, E.P. and Resh, V.H. (2006) Long-term seasonal
variation in the biological traits of benthic-macroinvertebrates in two
Mediterranean-climate streams in California , USA. Freshwater Biology 51:
56-75.
Lytle, D.A. and Poff, N.L. (2004) Adaptation to natural flow regimes. Trends
in Ecology and Evolution 19: 94-100.
4 Ecological
Process –
Disturbance
Bankfull flows may act as a disturbance mechanism which
resets ecological processes for both aquatic and riparian
flora and fauna communities by drowning and/or sweeping
plants and animals downstream. [M]
No reference cited for the conceptual model. [M]
Rørslett, B. (1988) Aquatic weed problems in a hydroelectric river: the River
Otra, Norway. Regulated Rivers: Research & Management, 2, 25-37.
Bond, N.R. and Downes, B.J. (2003) The independent and interactive
effects of fine sediment and flow on benthic communities characteristic of
small upland streams. Freshwater Biology, 48, 455-465.
Nilsson, C. and Svedmark, M. (2002) Basic principles and ecological
consequences of changing water regimes: riparian plant communities.
Environmental Management 30: 468-480.
Rood, S.B. et al 2005. Managing river flows to restore floodplain forests.
Frontiers in Ecology and the Environment 3: 193-201.
Stanley, E.H, Fisher, S.G and Grimm, N.B. (1997) Ecosystem expansion
and contraction in streams. BioScience 47: 427-435.
Fisher, S.G., Gray, L.J, Grimm, N.B. and Busch, D.E. (1982) Temporal
succession in a desert stream following flooding. Ecological Monographs
52:93-119.
5 Ecological
Process –
Habitat
Processes
Summer Low Flows maintain minimum water levels. They
provide trickling flows over riffles thereby maintaining riffle
habitat and also maintain adequate depth and (longitudinal)
connectivity between permanent pools [M,G]
In systems with seasonal ‘flow inversion’ (eg. high summer-
autumn flows, reduced winter-spring baseflows due to
regulation activities), constant high water levels can
effectively reduce the area of riffle habitat available for some
invertebrates and fish [Go] This phenomenon may also
No reference cited for the conceptual model. [D,M,T,L,Gl]
Schlosser, I.J. (1982) Fish community structure and function along two
habitat gradients in a headwater stream. Ecological Monographs, 52, 395-
414. [W]
O’Connor, N.A. (1991) The effects of habitat complexity on
macroinvertebrates colonising wood substrates in a lowland stream.
Oecologia, 85, 504-512. [W]
Cobb, D.G., Galloway, T.D. and Flanagan, J.F. (1992) Effects of discharge
and substrate stability on density and species composition of stream
108

Monitoring and evaluation of environmental flow releases in the Campaspe River
reduce the area of shallow water (eg. <0.3m depth) habitat
favoured by some in-channel macrophytes and small-bodied
fish. [Go]
With flow inversion (ie, higher summer-autumn flows) water
is deeper and colder during the growing season and is
expected to result in poor growing conditions for submerged
macrophytes (Roberts, pers.comm.)
Protection or reinstatement of more natural levels of winter-
spring baseflows provide conditions of sustained water levels
in the river. Inundation of physical structures/ features such
as large woody debris, branch-piles, bars,
overhanging/undercut banks and marginal/channel
edge/bankside vegetation and tree roots makes them
available for colonisation and/or attachment and increases
habitat availability for instream flora and fauna, including fish
and macroinvertebrates. [M,T,L,B,W,Gl] Habitat diversity
allows aquatic organisms with different habitat requirements
to coexist. [W] These habitats may be defined by a complex
interaction of ecological requirements (eg. flow, light, shelter
and food). [W] Often, more diverse assemblages of aquatic
biota are present in areas with a relatively high variety of
instream habitats. [W] For example, the complexity of
macroinvertebrate communities and the abundance of
individual species have been correlated with habitat
complexity created by woody debris, macrophytes and
organic debris, coarse substrates and substrate stability. [W]
Low flows during Spring and Summer also provide greater
areas of low or no velocity habitats that produce greater
densities of microinvertebrates due to increased residence
time. (Ferrari et al. 1989, Pace et al. 1992, Basu and Pick
1996, Reckendorfer et al. 1999, Reynolds 2000, King 2004b;
King, pers.comm.)
Protection or reinstatement of more natural levels of Winter-
Spring baseflows also increases the quantity and diversity of
flow velocity habitats by increasing the area of riffle and run
insects. Canadian Journal of Fisheries and Aquatic Sciences, 49, 1788-
1795. [W]
Lake, P.S. (1995) Of floods and droughts: river and stream ecosystems of
Australia. In River and Stream Ecosystems. Cushing, C.E., Cummins, K.W.
and Minshall, G.W. (Eds.)., Elsevier, Amsterdam. pp. 659-694. [W]
Mitchell, B., Rutherford, I., Constable, A., Stagnitti, F. and Merrick, C.
(1996) An Ecological and Environmental Flow Study of the Glenelg River
from Casterton to Rocklands Reservoir. Deakin University, Warrnambool,
Victoria. [W]
Williams, D.D. and Feltmate, B.W. (1992) Aquatic Insects. C.A.B.
International, Wallingford, UK.
Humphries, P. (1996) Aquatic macrophytes, macroinvertebrate
associations and water levels in a lowland Tasmanian river. Hydrobiologia
321, 219-233.
Petts, G.E. (1996) Water allocation to protect river ecosystems. Regulated
Rivers: Research & Management, 12, 353-365.
Armitage, P.D., Lottmann, K., Kneebone, N. and Harris, I. (2001) Bank
profile and structure as determinants of macroinvertebrate assemblages-
seasonal changes and management. Regulated Rivers: Research and
Management 17, 543-556.
Bond, N. R. and Lake, P.S. (2003a) Characterizing fish-habitat associations
in streams as the first step in ecological restoration. Austral Ecology, 28,
611-621.
Ferrari, I., Farabegoli, A and Mazzoni, R. (1989) Abundance and diversity
of planktonic rotifers in the Po River. Hydrobiologia 186/187: 201-208.
Pace, M.L, Findlay, S.E and Lints. D. (1992) Zooplankton in advective
environments: The Hudson River community and a comparative analysis.
Canadian Journal of Fisheries and Aquatic Sciences 49: 1060-1069.
Basu, B.K. and Pick, F.R. (1996) Factors regulating phytoplankton and
zooplankton biomass in temperate rivers. Limnology and Oceanography
41: 1572-1577.
Reckendorfer, W., Heckeis, H., Winkler, G. and Schiemer, F. (1999)
Zooplankton abundance in the River Danube, Austria: the significance of
inshore retention. Freshwater Biology 41: 583-591.
Reynolds, C.S. (2000) Hydroecology of river plankton: the role of variability
in channel flow. Hydrological Processes 14: 3119-3132.
King, A. (2004b) Density and distribution of potential prey for larval fish in
the main channel of a floodplain river: pelagic versus epibenthic meiofauna.
109

Monitoring and evaluation of environmental flow releases in the Campaspe River
habitat. [M] Diversity in flow velocity habitats may be
particularly important for macroinvertebrate community
diversity, which can contain species specialised for high
velocity habitats. [M] Runs provide a site for aquatic
macrophyte growth thus increasing habitat complexity. [W,M]
However, excessive growth of submerged macrophytes
degrades physical habitat quality by decreasing instream
habitat diversity. [W]
Spring Freshes make available in-channel habitat such as
vegetated bars, benches and undercuts and these habitats
may be important for the colonisation of macroinvertebrates
and as spawning sites and refuges for native fish. [W,Gl]
Bankfull Flows enable the recruitment of woody habitat for
the channel [T] and also provide lateral connectivity between
in-channel and floodplain habitats. [W,Gl]
Overbank Flows are critical for maintaining longitudinal and
lateral connectivity between stream channel and floodplain
areas. [D] Overbank Flows restore natural wetland hydrology
and connectivity and help restore biodiversity of floodplain
wetlands. [T]
River Research Applications, 20, 883-897.
Pusey, B.J. and Kennard, M.J. (1996) Species richness and geographical
variation in assemblage structure of the freshwater fish fauna of the Wet
Tropics Region of northern Queensland. Marine and Freshwater Research
47: 563–573.
Pusey, B.J. and Arthington, A.H. (2003) Importance of the riparian zone to
conservation and management of freshwater fish: a review. Marine and
Freshwater Research 54: 1–16.
Pusey, B.J., A.H. Arthington, A.H. and Read, M.G. (1993) Spatial and
temporal variation in fish assemblage structure in the Mary River, south–
east Queensland: the influence of habitat structure. Environmental Biology
of Fishes 37: 355–380.
Pusey, B.J., Arthington, A.H. and Read, M.G. (1995) Species richness and
spatial variation in fish assemblage structure in two rivers of the Wet
Tropics of northern Queensland, Australia. Environmental Biology of Fishes
42: 181–199.
Pusey, B.J., Arthington, A.H. and Read, M.G. (1998) Freshwater fishes of
the Burdekin River, Australia: biogeography, history and spatial variation in
community structure. Environmental Biology of Fishes 53: 303–318.
Pusey, B.J., Kennard, M.J. and Arthington, A.H. (2000) Discharge
variability and the development of predictive models relating stream fish
assemblage structure to habitat in north–eastern Australia. Ecology of
Freshwater Fish 9: 30–50.
Pusey, B.J. Kennard, M.J. and Arthington, A.H. (2004) Freshwater Fishes
of North-eastern Australia. CSIRO Publishing, Collingwood.
6 Ecological
Process –
Habitat
Processes
Freshes, High Flows and Bankfull Flows can help to flush
and remove accumulations of fine sediment and organic
matter from gravel, areas of the streambed such as riffle
areas, and benthic habitats such as large woody debris and
leaf-packs thereby preventing the smothering of these
habitats for biota that utilise them. [D,M,T,L,Gl,W-
winter/spring freshes]
Bankfull Flows may also dislodge and redistribute large
woody debris caught up in lower channel sections. [M]
No reference cited for the conceptual model. [M, T,L,Gl,W]
Chorley, R.J. (1962) Geomorphology and General Systems Theory. United
States Geological Survey Professional Paper, 500-B. [D]
Beschta, R.L. and Jackson, W.L. (1979) The intrusion of fine sediments into
a stable gravel bed. Journal of the Fisheries Research Board of Canada,
36, 207-210.
Sherrard, J.J. and Erskine, W.D. (1991) Complex response of a sand-bed
steam to upstream impoundment. Regulated Rivers: Research and
Management 6, 53-70.
Sear, D.A. (1993) Fine sediment infiltration into gravel spawning beds
within a regulated river experiencing floods: ecological implications for
salmonids. Regulated Rivers: Research and Management 8, 373-390.
Milhous, R.T. (1998) Modeling of instream flow needs: the link between
110

Monitoring and evaluation of environmental flow releases in the Campaspe River
sediment and aquatic habitat. Regulated Rivers: Research and
Management 14, 79-94.
Wood, P.J. and Armitage, P.D. (1999) Sediment deposition in a small
lowland stream-management implications. Regulated Rivers: Research and
Management 15, 199-210.
Wood, P.J. and Armitage, P.D. (1997) Biological effects of fine sediment in
the lotic environment. Environmental Management 21: 203-217.
7 Water Quality
[Provision of ]Low Flows [will] help ameliorate water quality
[D,C] and help minimize the increase in temperature and the
decrease in DO [M] Summer/autumn Low Flows help to slow
the deterioration of water quality that occurs in pools during
summer low flow periods (avoid stagnation). [M, Gl]
Summer Freshes help maintain/improve water quality in
rivers by providing an input of fresh water and mixing and/or
flushing pools which may have stagnated or and/or stratified
after prolonged periods of zero/low flow. [D,M,W,Gl,C]
However, floodplain inundation and connectivity of large
previously dry areas [which results in an influx or organic
matter] can decrease dissolved oxygen concentrations in
waters (particularly during warm summer conditions), and
under some conditions cause fish kills (King, pers.comm.).
No reference cited for the conceptual model. [C, M,Gl]
Mitchell, B., Rutherfurd, I, Constable, A., Stagnitti, F. and Merrick, C. (1996)
An Ecological and Environmental Flow Study of the Glenelg River from
Casterton to Rocklands Reservoir. Aquatic Resource Utilisation and
Management Research Group, Deakin University, Warrnambool. [D]
SKM (1997) Wimmera River Environmental Flow Impact Assessment. S233
Wimmera Stream Salinity Monitoring, Report to Wimmera Mallee Water,
Sinclair Knight Merz. [W, Gl]
Ward J.V. (1985) Thermal characteristics of running waters. Hydrobiologia,
125, 31-46.
Sabo, M.J., Bryan, C.F., Kelso, W.E. and Rutherford, D.A. (1999)
Hydrology and aquatic habitat characteristics of a riverine swamp: I.
Influence of flow on water temperature and chemistry. Regulated Rivers:
Research and Management 15, 505-523.
Western, A. and M. Stewardson (1999). Thermal Stratification in the
Wimmera and Glenelg Rivers. Centre for Environmental Applied Hydrology
(CEAH), The University of Melbourne, Victoria, Australia.
Turner, L. and Erskine, W.D. (2005) Variability in the development,
persistence and breakdown of thermal, oxygen and salt stratification of
regulated rivers of southeastern Australia. Regulated Rivers: Research and
Management 21, 151-168.
8 Biofilms Winter/spring Freshes and High Flows provide scouring
flows over biofilm habitats and this acts as a disturbance
mechanism for maintaining species composition and health.
[T,L] Summer Freshes are also expected to perform this
function [L].
No reference cited for the conceptual model. [T,L]
Horner, R.R. and Welsh, E.B. (1981) Stream periphyton development in
relation to current velocity and nutrients. Canadian Journal of Fisheries and
Aquatic Sciences, 38, 449-457.
Lindström, E.A. and Traaen, T.S. (1984) Influence of current velocity on
periphyton distribution and succession in a Norwegian soft water river.
Hydrobiologia, 22, 1965-1972.
Keithan, E.D. and Lowe, R.L. (1985) Primary productivity and spatial
structure of phytolithic growth in streams in the Great Smokey Mountains
National Park, Tennessee. Hydrobiologia, 123, 59-67.
111

Monitoring and evaluation of environmental flow releases in the Campaspe River
Stevenson, R.J. (1990) Benthic algal community dynamics in a stream
during and after a spate. Journal of the North American Benthological
Society, 16, 248-262.
Ács, É. And Kiss, K.T. (1993) Effects of water discharge on periphyton
abundance and diversity in a large river (River Danube, Hungary).
Hydrobiologia, 249, 125-133.
Biggs, B.J.F. (1995) the contribution of flood disturbance, catchment
geology and land use to the habitat template of periphyton in stream
ecosystems. Freshwater Biology, 33, 419-438.
Peterson, C.G., Weibel, A.C., Grimm, N.B. and Fisher, S.G. (1994)
Mechanisms of benthic algal recovery following spates: comparison of
simulated and natural events. Oecologia, 98, 280-290.
Peterson, C.G. (1996) Mechanisms of lotic microalgae colonization
following space-clearing disturbances acting at different spatial scales.
Oikos, 77, 417-435.
Biggs, B.J.F., Kilroy, C. and Lowe, R.L. (1998) Periphyton development in
three valley segments of a New Zealand grassland river: test of a habitat
matrix conceptual model within a catchment. Archiv für Hydrobiologie, 143,
147-177.
Bourassa, N. and Cattaneo, A. (1998) Control of periphyton biomass in
Laurentian streams, (Québec). Journal of the North American Benthological
Society, 17, 420-429.
Fayolle, S., Cazaubon, A., Comte, K. and Franquet, E. (1998) The
intermediate disturbance hypothesis: application of this concept to the
response of epilithon in a regulated Mediterranean river (Lower-Durance,
south-eastern France). Archiv für Hydrobiologie, 143, 57-77.
Biggs, B.J.F., Smith, R.A. and Duncan, M.J. (1999a) Velocity and sediment
disturbance of periphyton in headwater streams: biomass and metabolism.
Journal of North American Benthological Society, 18, 222-241.
Biggs, B.J.F., Tuchman, N.C., Lowe, R.L. and Stevenson, R.J. (1999b)
Resource stress alters hydrological disturbance effects in a stream
periphyton community. Oikos, 85, 95-108.
Burns, A. and Walker, K. (2000) Effects of water level regulation on algal
biofilms in the River Murray, South Australia. Regulated Rivers: Research
and Management, 16, 433-444.
Burns, A. and Ryder, D.S. (2001) Potential for biofilms as biological
indicators in Australian riverine systems. Ecological Management and
Restoration, 2, 53-63.
112

Monitoring and evaluation of environmental flow releases in the Campaspe River
Ryder, D.S. (2004) Response of epixylic biofilm metabolism to water level
variability in a regulated floodplain river. Journal of the North American
Benthological Society, 23, 214-223.
Watts, R.J., Nye, E.R., Thompson, L.A., Ryder, D.S., Burns, A. and
Lightfoot, K. (2005) Environmental Monitoring of the Mitta Mitta River
Associated with the Transfer of Water Resources from Dartmouth
Reservoir to Hume Reservoir 2004/2005. Report to the Murray-Darling
Basin Commission. Environmental consultancy report No. 97. Johnstone
Centre, Charles Sturt University, Wagga Wagga.
Ryder, D.S., Watts, R.J., Nye, E. and Burns, A. (2006) Can flow velocity
regulate epixylic biofilm structure in a regulated floodplain river? Marine and
Freshwater Research, 57, 29-36.
9 Aquatic &
Riparian
Vegetation
Protection or reinstatement of more natural levels of Winter-
Spring baseflows will provide conditions of sustained water
levels in the river and inundate lower channel portions.
Sustained inundation of riffles, lower benches and channel
margins will maintain shallow water habitat for emergent and
marginal aquatic vegetation during the spring growing
season. [M, Gl] Prolonged winter/spring inundation of lower
channel portions will drown and cause dieback of terrestrial
vegetation (mainly agricultural weeds) which has encroached
down the bank during the low flow period. [M]
Winter/Spring Freshes which inundate vegetation for a
minimum of 4 days will allow regeneration of hydrophilic
species.1 [T]
Summer Low Flows will help to maintain appropriate/
adequate soil moisture conditions for aquatic and riparian
vegetation. [D] Summer Flows will maintain shallow water
(defined as <0.3m) habitat for in-channel macrophytes (eg.
smaller submerged and floating-leafed aquatic macrophytes)
during the latter part of the growing season. [B,Go]
Low Flows also mean low-moderate flow velocities within the
instream environment (eg. mean reach velocities of <0.06
ms/s or even <0.04 m/s) which are suitable for macrophyte
growth. [Go] Fast and very fast velocity flows increase the
risk of mechanical damage to plants of parts breaking off and
of emerging or floating leaves being dragged underwater,
effectively reducing rates of photosynthesis and
No reference cited for the conceptual model. [M,Gl,T,C]
Arthington, A.H., Brizga, S.O., Choy, S.C., Kennard, M.J., Mackay, S.J.,
McCosker, R.O. and Ruffini, J.L. (2000) Environmental Flow Requirements
for the Brisbane River Downstream from Wivenhoe Dam. South-East
Queensland Water Corporation Ltd and Centre for Catchment and Instream
Research, Brisbane. [D]
Madsen, J., Chambers P., James W., Koch E. and Westlake D. (2001) The
interaction between water movement, sediment dynamics and submersed
macrophytes. Hydrobiologia, 444, 71-84. [Go]
Riis, T. and Biggs, B.J.F. (2003) Hydrologic and hydraulic control of
macrophytes establishment and performance in streams. Limnology and
Oceanography, 48, 1488-1497. [Go]
Chambers, P.A., Prepas, E.E., Hamilton, H.R. and Bothwell, M.L. (1991)
Current velocity and its effects on aquatic macrophytes in flowing waters.
Ecological Applications, 1, 249-257.
Casanova, M.T. and Brock, M.A. (2000) How do depth, duration and
frequency of flooding influence the establishment of wetland plant
communities? Plant Ecology, 147, 237-250. [Go]
Pollock, M.M., Naiman, R.J. and Hanley, T.A. (1998) Plant species richness
in riparian wetlands – a test of biodiversity theory. Ecology, 79, 94-105.
Mawhinney, W.A. (2003) Restoring biodiversity in the Gwydir wetlands
through environmental flows. Water Science and Technology, 48, 73-81.
Nicol, J.M., Ganf, G.G. and Pelton, G.A. (2003) Seed banks of a southern
Australian wetland: the influence of water regime on the final floristic
113

Monitoring and evaluation of environmental flow releases in the Campaspe River
consequently growth. [Go]
Summer Freshes wet low-lying channel zones such as riffles
and benches and help to alleviate drought-stress on
emergent and aquatic vegetation that has become exposed
during the low flow period. [M,T]
With flow inversion (eg. higher summer-autumn flows,
reduced winter-spring baseflows due to regulation activities)
water is deeper and colder during the growing season and is
expected to result in poor growing conditions for submerged
macrophytes. Sustained flow inversion also eliminates flood
intolerant plant species from riverbanks. (Roberts,
pers.comm.)
Flows which inundate benches/riparian zone will help to:
(a) maintain riparian vegetation [C]
(b) promote regeneration of native species, including river
red gum2 [C]
(c) control invasion by terrestrial and/or weed species if
sustained for a sufficiently long period [T,M,C,L,Go]
Results from a study of depth, duration and frequency of
flooding on plant recruitment from wetland sediments
suggest that flows which inundate areas for short durations
(eg. <2 weeks) lead to a higher proportion of terrestrial and
introduced plants, while native species are believed to be
favoured by inundation periods of longer duration. [Go]
In addition, [with flow inversion] inundation of benches at a
time when they would naturally have been dry can affect the
mix of aquatic/ terrestrial species and increase the risk of
weed invasion (e.g. willows). [B] For instance, higher than
natural summer flows change the bed environment from an
opportunity for summer-growing annual/perennial
herbaceous forms and short benthic submerged
macrophytes, to one that is too challenging for these weakly-
growing and non-robust species. This may result in a shift in
vegetation on benches from stress-tolerant, but less
competitive to competitive invading species and a loss of
composition. Plant Ecology 168, 191-205.
Reid, M.A. and Quinn, G.P. (2005) Hydrologic regime and macrophyte
assemblages in temporary floodplain wetlands: implications for detecting
responses to environmental water allocations. Wetlands 24, 586-599.
Seablom, E.W., Moloney, K.A. and van der Valk, A.G. (2001) Constraints
on the establishment of plants along a fluctuating water-depth gradient.
Ecology 82, 2216-2232.
Roberts, J., Young, W. and Marston, F. (2000) Estimating the water
Requirements for Plants of Floodplain Wetlands: A Guide. Occasional
Paper 04/00. Land and Water Resources Research and Development
Corporation, Canberra.
Nielsen, D.L. and Chick, A.J. (1997) Flood-mediated changes in aquatic
macrophyte community structure. Marine and Freshwater Research, 48,
153-157.
114

Monitoring and evaluation of environmental flow releases in the Campaspe River
diversity. [B]
In a mesocosm experiments using billabong soils, Nielsen
and Chick (1997) found that prolonged inundation of artificial
billabongs (summer flooding followed by high winter/spring
flows) led to lower plant diversity compared to artificial
billabongs that experienced extended periods of drying
followed by spring flooding. The lower plant diversity was
due to the absence of ephemeral and terrestrial plant taxa.
In developing an approach for monitoring vegetation
response to environmental flows in the Wimmera, Dyer and
Roberts (2006) identified expected responses to
environmental flows recommended for the Wimmera River
system that were site-specific and linked to specific flow
components. They are presented here in generic form
because many of these predictions have not previously been
expressed in the environmental flow reports for the 8 case
study rivers and they represent potentially useful monitoring
endpoints. This list is also relevant to Attribute 10.
Vegetation responses:
1. Increased vigour/improvement in canopy of adjacent
trees and shrubs
2. Growth pulse (eg. in seedlings/saplings);
3. Increased flowering intensity / flowering seed set
4. Germination of riparian trees and shrubs
5. Recruitment of riparian trees and shrubs
6. Slowing of mortality rate in riparian trees
7. Development of ‘recession’ flora in flood runners,
anabranches, depressions
8. Stimulate regrowth of forbs and herbs in flood runners
9. Change in relative abundance of submerged
macrophytes; shift/turnover in species composition
and/or functional types
10. Change in understorey composition and/or functional
groups from grasses and sedges to amphibious plants
11. Increase in overall diversity in channel/channel edge
/bank through establishment of a mix of sedges,
Dyer, F. and Roberts, J. (2006) Monitoring Vegetation Response to
Environmental Flows in the Wimmera: A Strategic Approach. A report to the
Wimmera Catchment Management Authority, Horsham. In2fluve,
O’Connor, Canberra, ACT.
115

Monitoring and evaluation of environmental flow releases in the Campaspe River
grasses, herbs and forbs; diversity expressed over
length of channel
12. More vigorous growth of Phragmites such as increased
stand height. Density, greater flowering intensity of
culms; build up of rhizome starch reserves
13. Colonization and expansion by Phragmites
14. Flushing of senescent plant parts from in-channel
macrophytes
1Not sensible to make a statement that inundating vegetation
for a minimum of 4 days over the Winter-Spring period will
allow regeneration. If there is a factual basis for this, then the
statement most likely refers to a single species or small
group (Roberts, pers.comm.)
2The time for plants to respond to inundation varies very
widely between species (from days to weeks and even
months) and cannot be easily connected to a single attribute
or plant type (ie, hard to predict) (Roberts, pers.comm.)
10 Aquatic &
Riparian
Vegetation
Freshes provide short-term flow variability and variation in
water levels is important for maintaining species diversity in
the emergent and marginal aquatic vegetation communities.
Variation in water levels is the principal driver of zonation
patterns across the channel and up the river banks. [M,T,L]
Providing Freshes and ensuring the occurrence of flow
variability and variation in water levels will help
(a) restore/maintain/increase species diversity in the
emergent and marginal aquatic vegetation communities.
(b) maintain/restore distinctive riparian vegetation community
and structure with zonation up the bank.
Change in flow variability can take many forms which have
different consequences for vegetation patterns. For instance,
prolonged stable water levels allow plants to establish and
persist close to the water line, with the species doing this
being more associated with lentic (wetland) environments
than lotic (flowing water) environments. Loss of flow
variability may also result in wider zones of terrestrial or flood
No reference cited for the conceptual model. [M,T,L]
Rørslett, B. (1988) Aquatic weed problems in a hydroelectric river: the River
Otra, Norway. Regulated Rivers: Research & Management, 2, 25-37.
Rørslett, B., Mjelde, M. and Johansen, S.W. (1989) Effects of hydropower
development on aquatic macrophytes in Norwegian rivers: present state of
knowledge and some case studies. Regulated Rivers: Research &
Management, 3, 19-28.
Biggs, B.J.F. (1996) Hydraulic habitat of plants in streams. Regulated
Rivers: Research & Management, 12, 131-144.
French, T.D. and Chambers, P.A. (1996) Habitat partitioning in riverine
macrophyte communities. Freshwater Biology, 36, 509-520.
Brock, M.A. and Casanova, M.T. (1997) Plant life at the edge of wetlands;
ecological responses to wetting and drying patterns. In Frontiers in
Ecology. Klomp, N. and Lunt, I. (Eds.)., Elsevier Science Ltd, Oxford, UK,
pp. 181-192.
Blanch, S.J., Walker, K.F. and Ganf, G.G. (2000) Water regimes and littoral
plants in four weir pools of the River Murray, Victoria. Regulated Rivers:
Research & Management, 16, 445-456.
Riis, T. and Hawes, I. (2002) Relationships between water level fluctuations
116

Monitoring and evaluation of environmental flow releases in the Campaspe River
intolerant plant species and a shrinking in the width of the
zone characterised by flood tolerant species. (Roberts,
pers.comm.)
Bankfull flows lasting a minimum of 3 days can prevent
encroachment of riparian terrestrial vegetation and maintain
riparian vegetation diversity and structure.1[T] However,
Bankfull flows may also remove aquatic and riparian
vegetation through scouring of the channel bed. [M]
1This statement applies to upland rivers. What is the source
of the number of days? Inference from hydrological analysis?
(Roberts, pers.comm.)
and vegetation diversity in shallow water of New Zealand lakes. Aquatic
Botany, 74, 133-148.
Leyer, I. (2005) Predicting plant species’ responses to river regulation: the
role of water level fluctuations. Journal of Applied Ecology, 42, 239-250.
Van Geest, G.J., Coops, H., Roijackers, R.M.M., Buijse, A.D. and Scheffer,
M. (2005) Succession of aquatic vegetation driven by reduced water-level
fluctuations of floodplain lakes. Journal of Applied Ecology, 42, 251-260.
Stanley, E.H., Fisher, S.G. and Grimm, N.B. (1997) Ecosystem expansion
and contraction in streams. BioScience 47: 427-435.
Ewel, K.C., Cressa, C., Kneib, R.T., Lake, P.S., Levin, L.A., Palmer,M.A.,
Snelgrove, P. and Wall, D.H. (2001) Managing critical transition zones.
Ecosystems 4: 452-460.
Mackay, S.J., Arthington, A.J., Kennard, M.K. and Pusey,B.J. (2003)
Spatial variation in the distribution and abundance of submersed
macrophytes in an Australian subtropical river. Aquatic Botany 77, 169-186.
Lenssen, J., Menting, F., van der Putten, W. and Blom, K. (1999) Control of
plant species richness and zonation of functional groups along a freshwater
flooding gradient. Oikos, 86, 523-534.
11 Aquatic &
Riparian
Vegetation
High Flows deliver seed from the upper catchment to help
maintain/restore distinctive riparian vegetation community
and structure. [T]
No reference cited for the conceptual model. [T]
Andersson, E., Nilsson, C. and Johansson, M.E. (2000) Effects of river
fragmentation on plant dispersal and riparian flora. Regulated Rivers:
Research & Management, 16, 83-89.
Merritt, D.M. and Wohl, E.E. (2006) Plant dispersal along rivers fragmented
by dams. River Research and Applications, 22, 1-26.
Nilsson, C. and Beggren, K. (2000) Alterations of riparian ecosystems
caused by river regulation. BioScience 50, 783-792.
12 Aquatic &
Riparian
Vegetation
Overbank Flows play an important role in vegetation
community maintenance. [D] [Too vague – what’s being
referred to? The instream aquatic, riparian or floodplain
vegetation community? What exactly is meant by ‘vegetation
community maintenance’?]
Nielsen, D.L. and Chick, A.J. (1997) Flood-mediated changes in aquatic
macrophyte community structure. Marine and Freshwater Research, 48,
153-157. [D]
Robertson, A.I., Bacon, P. and Heagney, G. (2001) The responses of
floodplain primary production to flood frequency and timing. Journal of
Applied Ecology, 38, 126-136.
Molles, M.C., Crawford, C.S., Ellis, L.M., Valett, H.M. and Dahm, C.N.
(1998) Managed flooding for riparian ecosystem restoration. BioScience,
48, 749-756.
13 Invertebrates Protection or reinstatement of more natural levels of Winter-
Spring baseflows will provide conditions of sustained water
levels in the river. This will provide sustained longitudinal
No reference cited for the conceptual model. [M,T,L]
Lauters, F., Lavendier, P., Lim P., Sabaton, C. and Belaud, A. (1996)
Influence of hydropeaking on invertebrates and their relationship with fish
117

Monitoring and evaluation of environmental flow releases in the Campaspe River
connectivity for invertebrate drift. [M,T] Provision of Summer
Low Flows maintains minimum water levels maintaining
adequate depth and longitudinal connectivity between
permanent pools. [M,G]
Summer Freshes will increase water depth over low-lying
channel zones such as riffles and increase longitudinal
connectivity, thereby temporarily facilitating greater
movement of macroinvertebrates between different instream
habitats such as pools. [M,L]
In systems with seasonal ‘flow inversion’ (eg. high summer-
autumn flows, reduced winter-spring baseflows due to
regulation activities), increased stream velocity in summer-
autumn, and disruption of the drift patterns of
macroinvertebrates, may make macroinvertebrates more
vulnerable to predation or physical removal.[Go]
Low flows during spring and summer also provide greater
areas of low or no velocity habitats that produce greater
densities of microinvertebrates due to increased residence
time. (King, pers.comm.; see Refs in green)
feeding habits in a Pyrenean river. Regulated Rivers: Research and
Management 12: 563-573. [Go]
Irvine J.R. (1985) Effects of successive flow perturbations on stream
invertebrates. Canadian Journal of Fisheries and Aquatic Sciences, 42,
1922–1927.
Poff N.L. & Ward J.V. (1991) Drift responses of benthic invertebrates to
experimental stream flow variation in a hydrologically stable stream.
Canadian Journal of Fisheries and Aquatic Sciences, 48, 1926–1936.
Bond, N.R. and Downes, B.J. (2003) The independent and interactive
effects of fine sediment and flow on benthic communities characteristic of
small upland streams. Freshwater Biology, 48, 455-465.
King, A. (2004b) Density and distribution of potential prey for larval fish in
the main channel of a floodplain river: pelagic versus epibenthic meiofauna.
River Research Applications, 20, 883-897.
Basu, B.K. and Pick ,F.R. (1996) Factors regulating phytoplankton and
zooplankton biomass in temperate rivers. Limnology and Oceanography
41: 1572-1577.
Ferrari, I., Farabegoli, A and Mazzoni, R. (1989) Abundance and diversity
of planktonic rotifers in the Po River. Hydrobiologia 186/187: 201-208.
Pace, M.L, Findlay, S.E and Lints. D. (1992) Zooplankton in advective
environments: The Hudson River community and a comparative analysis.
Canadian Journal of Fisheries and Aquatic Sciences 49: 1060-1069.
Reckendorfer, W., Heckeis, H., Winkler, G. and Schiemer, F. (1999)
Zooplankton abundance in the River Danube, Austria: the significance of
inshore retention. Freshwater Biology 41: 583-591.
Reynolds, C.S. (2000) Hydroecology of river plankton: the role of variability
in channel flow. Hydrological Processes 14: 3119-3132.
14 Invertebrates High Flows and Bankfull Flows which inundate previously dry
sediments in higher portions of the channel such as benches
may provide a stimulus for hatching to [micro]invertebrates,
with diversity and biomass peaking when inundation exceeds
2 weeks. [Go]
Loss of habitat through decreased inundation duration
increases the risk of egg mortality, and the loss of early
instars (early life stages) and those species not stimulated to
drift. For the others the outcome will depend on factors such
as the availability of alternative habitat and predation
pressure. [Go]
Langley J.M., Shiel R.J., Nielsen D.L., and Green J.D. (2001) Hatching
from the sediment egg-bank, or aerial dispersing? – the use of mesocosms
in assessing rotifer biodiversity. Hydrobiologia, 446/447, 203-211. [Go]
Nielsen D.L., Hillman T.J., Smith F.J. and Shiel R.J. (2002) The influence of
seasonality and duration of flooding on zooplankton in experimental
billabongs. Regulated Rivers: River Research & Applications, 18, 227-237.
[Go]
Ballinger papers???
Hillman, T.J. and Quinn, G.P. (2002) Temporal changes in
macroinvertebrate assemblages following experimental flooding in
permanent and temporary wetlands in an Australian floodplain forest. River
118

Monitoring and evaluation of environmental flow releases in the Campaspe River
Research & Applications 18: 137-154.
Jenkins, K.M. and Boulton, A.J. (2003) Connectivity in a dryland river:
short-term aquatic microinvertebrate recruitment following floodplain
inundation. Ecology 84: 2708-2723.
15 Invertebrates Overbank Flows play an important role in invertebrate
colonisation. [D] [Too vague – invertebrate colonisation of
what exactly? Of instream habitats? Or floodplain wetland
habitats? No details on exactly how overbank flows are
important in this respect.]
Nielsen, D.L., Hillman, T.J. and Smith, F.J. (1999) Effects of hydrological
variation and planktivorous competition on macroinvertebrate community
structure in experimental billabongs. Freshwater Biology, 42, 427-444. [D]
Quinn, G.P., Hillman, T.J. and Cook, R. (2000) The response of
macroinvertebrates to inundation in floodplain wetlands: a possible effect of
river regulation. Regulated Rivers: Research & Management, 16, 469-.477.
[D]
16 Fish - Habitat Low Flows year round help to maintain/enhance native fish
community structure through habitat availability and
inundation of large woody debris which provides food
sources and shelter. [T, L, Gl]
Protection or reinstatement of more natural levels of
baseflows from late Spring through to early Winter under
regulated conditions will help to maintain or increase the
amount of deepwater habitat available for large-bodied fish.
Overseas studies of patterns of fish habitat use have clearly
demonstrated the importance of deepwater habitat in
structuring riverine fish communities and the availability of
deepwater habitats strongly influences the distributions of
large-bodied fish. Research in Australia rivers has also
shown that the adult stages of many larger native species
rely heavily upon the availability of deepwater habitats. [Go]
In systems with seasonal ‘flow inversion’ (eg. high summer-
autumn flows, low winter flows due to regulation activities),
constant high water levels during summer can effectively
reduce the area of riffle habitat available for some fish [Go]
This flow inversion may also reduce the area of shallow
water (eg. <0.3m depth) habitat favoured by some small-
bodied fish. [Go]
No reference cited for the conceptual model. [T, L, Gl]
Gorman, O.T., and Karr, J.R. (1978) Habitat structure and stream fish
communities. Ecology, 59, 507-515. [Go]
Harvey, B.C. and Stewart, A.J. (1991) Fish size and habitat depth
relationships in headwater streams. Oecologia, 87, 336-42. [Go]
Crook D., Robertson A., King A. and Humphries P. (2001) The influence of
spatial scale and habitat arrangement on dual patterns of habitat use by
two lowland river fishes. Oecologia, 129, 525-533. [Go]
Koehn, J.D. and Nicol, S. (1998) Habitat and movement requirements of
fish. In Banens, R.J. and Lehane, R. (Eds). Proceedings of the 1996
Riverine Environment Research Forum, Brisbane, Queensland, October
1996. Murray-Darling Basin Commission, Canberra, ACT, Australia. pp. 1-
6. [Go]
Bond, N. R. and Lake, P.S. (2003a) Characterizing fish-habitat associations
in streams as the first step in ecological restoration. Austral Ecology, 28,
611-621.
Bond, N. and Lake, P.S. (2003b) Local habitat restoration in streams:
constraints on the effectiveness of restoration for stream biota. Ecological
Management and Restoration, 4, 193-198.
Bond N. and Lake P.S. (2004). Disturbance regimes and stream
restoration: the importance of restoring refugia. Fourth Australian Stream
Management Conference, Cooperative Research Centre for Catchment
Hydrology, Melbourne, Launceston, Australia, 2004.
Bond N. and Lake P.S. (2005a) Ecological restoration and large-scale
ecological disturbance: the effects of drought on the response by fish to a
habitat restoration experiment. Restoration Ecology, 13, 39-48.
Bond N. and Lake P.S. (2005b) Of sand slugs and fish restoration: the
119

Monitoring and evaluation of environmental flow releases in the Campaspe River
Granite Creeks Saga. In Lintermans M., Cottingham P. and O’Connor R.
(Eds). Proceedings of the Fish Habitat Restoration Workshop, Albury
February 2004, Murray-Darling Basin Commission, Canberra.
17 Fish -
Movement Protection or reinstatement of more natural levels of Winter-
Spring baseflows will provide conditions of sustained water
levels in the river. This will provide sustained longitudinal
connectivity for fish movement,1[M,T], including the
permanent movement of large-bodied fish throughout the
river reach in the lead up to the breeding season.2 [L]
Summer Freshes will increase water depth over low-lying
channel zones such as riffles and increase longitudinal
connectivity, thereby temporarily facilitating greater
movement of fish between different instream habitats such
as pools. [M,L]
Freshes (which produce a minimum depth of 0.5m over the
shallowest point) over the irrigation season (Nov-Apr) are
important in lower river reaches for temporary local
movement of bigger fish such as Murray Cod and Golden
Perch [L]. These Freshes allow upstream movement of
Golden Perch to spawn [L] Summer (Jan/Feb) Freshes may
act as an attractant flow for Golden Perch from the Kerang
Weir [L]
Winter and Spring High Flows and Freshes and Bankfull
Flows may provide the cue which triggers movement and/or
migration in some native fish species [Gl], for instance, in
Australian Grayling [T] and in Tupong [M]. This migration is
associated with spawning and hence may have an impact on
species reproduction and recruitment. [M]
1Statement is too broad to be useful. Diadromous species
(particularly Australian Grayling , galaxiids and eels) require:
• High flows/flushes for larval transport to sea/estuary
in Autumn/early Winter
• High flows/freshes for juvenile upstream movement
and recruitment during Spring/early Summer
Koehn, J.D. and O’Connor, W.G. (1990) Biological
No reference cited for the conceptual model. [M,T,L]
Fausch, K.D., Torgersen, C.E., Baxter, C.V. and Li, H.W. (2002)
Landscapes to riverscapes: bridging the gap between research and
conservation of stream fishes. BioScience 52: 483-498.
Labbe, T.R. and Fausch, K.D. (2000) Dynamics of intermittent stream
habitat regulate persistence of a threatened fish at multiple scales.
Ecological Applications 10: 1774-1791.
Schlosser, I.J. (1995) Critical landscape attributes that influence fish
population dynamics in headwater streams. Hydrobiologia 303: 71-81.
No reference cited for the conceptual model. [M,W,Gl,C]
No reference cited for the conceptual model. [L]
MacKay, N.J. (1973) Histological changes in the ovaries of the golden
perch, Plectroplites ambigus, associated with the reproductive cycle.
Australian Journal of Marine and Freshwater Research, 24, 95-101.
Reynolds, L. F. (1983). Migration patterns of five fish species in the Murray-
Darling River system. Australian Journal of Marine and Freshwater
Research, 34, 857–871.
O’Connor, J.P., O’Mahoney, J.O and O’Mahoney, J.M. (2005) Movements
of Macquaria ambigua, in the Murray River, south-eastern Australia.
Journal of Fish Biology, 66, 392-403.
120

Monitoring and evaluation of environmental flow releases in the Campaspe River
Information for the Management of Freshwater Fish in
Victoria. Victorian Government Printing Office on behalf of
Department of Conservation and Environment, Freshwater
Fish Management Branch, Arthur Rylah Institute of
Environmental Management, Melbourne, Victoria.
Reasonable evidence exists for galaxiids and eels and this is
current belief for Australian Grayling. (King, pers.com.)
2No data to support this statement, however, is a reasonable
statement to include and what most experts think. (King,
pers.comm.)
18 Fish –
Maturation,
Reproduction/
Spawning,
Recruitment
Low Flows may be important for recruitment of some native
fish in lowland rivers. [D] Low Flows maintain or increase the
availability of slow-water (eg. velocity<0.05 m/s) habitats1
which are important as refuge and rearing habitats for larval
and juvenile fish. [B] (High velocity flows may ‘wash-out’
larvae.)
High water velocity over summer-autumn displace eggs and
larvae from spawning and rearing habitat [B] thus limiting
recruitment.2 [B,Go]
In systems with ‘flow inversion’ (eg, higher summer flows,
lower winter flows due to irrigation schedule) lower winter
flows means that there is less habitat and food available for
native fish at a time that may be critical for reproductive
development prior to spawning. [B]
Winter/Spring Freshes may provide spawning cues for
freshwater diadromous fish such as Australian Grayling and
Long-finned and Short-finned Eels.3 [M].
High Flows and Bankfull Flows have been linked to
requirements for fish breeding and may act as triggers for
breeding in some species. 4[D]
1backwaters/slackwaters (King, pers.comm.)
2recruitment of low flow specialists (King, pers.comm.)
3Diadromous species (such as grayling and eels) require
high flows or freshes in Autumn/early Winter to trigger
Humphries, P., King, A.J. and Koehn, J.D. (1999) Fishes, flows and
floodplains: links between Murray-Darling freshwater fish and their
environment. Environmental Biology of Fishes, 56, 129-151. [D]
Davies, P.E. and Humphries, P. (1996) An Environmental Flow Study of the
Meander, Macquarie and South Esk Rivers, Tasmania. Report to the
Department of Primary Industry and Fisheries, Tasmania. [B]
No reference cited for the conceptual model. [Go]
Heggenes, J. and Traaen, T. (1988) Downstream migration and critical
water velocities in stream channels for fry of four salmonid species. Journal
of Fish Biology, 32, 717-727.
Humphries, P., Serafini, L.G. and King, A.J. (2002) River regulation and fish
larvae: variation through space and time. Freshwater Biology, 47, 1307-
1331.
No reference cited for the conceptual model. [M,W,Gl,C]
Humphries, P. and Lake, P.S. (2000) Fish larvae and the management of
regulated rivers. Regulated Rivers: Research & Management, 16, 421-432.
King A., Humphries P. and Lake P.S. (2003) Fish recruitment on
floodplains: the roles of patterns of flooding and life history strategies.
Canadian Journal of Fisheries and Aquatic Sciences, 60, 773-786.
King A. (2004a) Ontogenetic patterns of habitat use of fish in an Australian
lowland river. Journal of Fish Biology, 65, 1582-1603.
King A. (2004b) Density and distribution of potential prey for larval fish in
the main channel of a floodplain river: pelagic versus epibenthic meiofauna.
River Research Applications, 20, 883-897.
Koehn, J.D. and O’Connor, W.G. (1990) Biological Information for the
Management of Freshwater Fish in Victoria. Victorian Government Printing
Office on behalf of Department of Conservation and Environment,
121

Monitoring and evaluation of environmental flow releases in the Campaspe River
spawning (as well as for larval transport, see Attribute 17)
(King, pers.comm.)
4Statement too broad to be useful. Some species
(particularly Golden Perch and Silver Perch) are thought to
require freshes/high flows/floodplain inundation during spring
and/or early summer to trigger spawning. (Lake 1967a,
Mackay 1973, Cadwallader 1977, King, pers.comm., King
unpub. data). Recent evidence has confirmed increased
abundance of spawning occurs on floods. However,
spawning of these species also occurs (in reduced numbers)
during sustained high flows in the Murray (King et al. 2005;
King, unpub. data). In addition, we also know that Golden
Perch are able to recruit successfully in years where no
floods occur only in-channel rises (Mallen-Cooper and Stuart
2003).
Freshwater Fish Management Branch, Arthur Rylah Institute of
Environmental Management, Melbourne, Victoria. [D]
Humphries, P. (1995) Life history, food and habitat of southern pygmy
perch, Nannoperca australis, in the Macquarie River, Tasmania. Marine
and Freshwater Research, 46, 1159-1169. [D]
Harris, J.H. and Gehrke, P.C. (1997) Fish and Rivers. The NSW Rivers
Survey. NSW Fisheries Research Institute, Cooperative Research Centre
for Freshwater Ecology, Sydney. [D]
O’Conner, W.G. and Koehn, J.D. (1998) Spawning of the broad-finned
galaxias, Galaxias brevipinnis Gunther (Pisces: Galaxiidae) in coastal
streams of southeastern Australia. Ecology of Freshwater Fish, 7, 95-100.
[D]
Lake, J.S. (1967). Rearing experiments with five species of Australian
freshwater fishes. I. Inducement to spawning. Australian Journal of Marine
and Freshwater Research 18, 137-153.
MacKay, N.J. (1973) Histological changes in the ovaries of the golden
perch, Plectroplites ambigus, associated with the reproductive cycle.
Cadwallader, P. L. (1977) J.O. Langtry's 1949-50 Murray River
Investigations. Ministry for Conservation Fisheries and Wildlife Division.
Melbourne, Australia. 70 pp.
King, A.J., Crook, D.A., Koster, W.M, Mahoney, J. and Tonkin, Z. (2005)
Comparison of larval fish drift in the Lower Goulburn and mid-Murray
Rivers. Ecological Management and Restoration, 6, 136-138.
Mallen-Cooper, M. and Stuart, I.G. (2003) Age, growth and non-flood
recruitment of two potadromous fishes in a large semi-arid/temperate river
system. River Research and Applications, 19, 697-719.
19 Fish –
Community
diversity
Overbank Flows play an important role in fish community
diversity. [D] [Too vague – fish community diversity where?
In the instream channel, floodplain wetlands? No details on
exactly how overbank flows are important in this respect.]
Geddes, M.C. and Puckridge, J.T. (1989) Survival and growth of larval and
juvenile native fish: the importance of the floodplain. In Proceedings of the
Native Fish Management workshop, Canberra. Murray-Darling Basin
Commission. [D]
Bayley, P.B. (1991) The flood pulse advantage and the restoration of river-
floodplain systems. Regulated Rivers: Research & Management, 6, 131-
144.
Heiler, G., Hein, T and Schiemer, F. (1995) Hydrological connectivity and
flood pulses as the central aspects for the integrity of a river-floodplain
system. Regulated Rivers: Research & Management, 11, 351-361.
Copp, G.H. (1989) The habitat diversity and fish reproductive function of
floodplain ecosystems. Environmental Biology of Fishes, 26, 1-26.
122

Monitoring and evaluation of environmental flow releases in the Campaspe River
20 Fish – Exotic
species
management
Summer/Spring Low Flows can expose banks and beds
leading to the drying of carp eggs and contributing to exotic
fish management.1 [T]
High summer flows and less annual flow variability also
provides habitat conditions favourable for introduced species
such as carp and Gambusia.2[B]
1Incorrect. A falling limb of a flow may strand carp eggs as
they can be attached. However, carp can spawn in any
conditions. Not a valid statement.
2To some extent, but low flows also enable Gambusia to
successfully spawn and recruit (King et al. 2003, King 2004a
– King, pers.comm.). Less variability in flows (eg. reaches
affected by sustained river regulation) provides suitable
conditions for carp breeding, recruitment and habitat for
adults (King et al. 1995, Stuart and Jones 2002 – King,
pers.comm.). Floodplain inundation events in spring and
early summer also trigger spawning and recruitment events
for carp and represent a potential adverse event.
No reference cited for the conceptual model. [T,B]
Gehrke, P.C. Brown, P., Schiller, C.B., Moffat, D.B. and Bruce A.M. (1995)
River regulation and fish communities in the Murray-Darling River system,
Australia. Regulated Rivers: Research and Management 11, 363-375.
Stuart, I. and Jones, M. (2002) Ecology and Management of Common Carp
in the Barmah-Millewa Forest. Freshwater Ecology, Arthur Rylah Institute
for Environmental Research, Dept of Sustainability and Environment.
Melbourne, Australia. pp 214
21 Waterbirds Overbank Flows are important for waterbirds. [D] [Too vague
- No details on exactly how overbank flows are important in
this respect.]
Kingsford, R.T., Curtin, A.L. and Porter, J. (1999) Water flows on Cooper
Creek in arid Australia determine ‘boom’ and ‘bust’ periods for waterbirds.
Biological Conservation, 88, 231-48.
Kingsford, R..T. and Thomas, R.F. (1995) The Macquarie Marshes in arid
Australia and their waterbirds: a 50-year history of decline. Environmental
Management, 19, 109-127.
Driver, P., Chowdhury, S., Wettin, P. and Jones, H. (in press). Models to
predict the effects of environmental flow releases on wetland inundation
and the success of colonial bird breeding in the Lachlan River, NSW.
Proceedings of the 4th Annual Stream Management Conference.
Launceston, Tasmania 19-22 October 2004.
Roshier, D.A., Robertson, A.I. and Kingsford, R.T. (2002) Responses of
waterbirds to flooding in an arid region of Australia and implications for
conservation. Biological Conservation 106: 399-411.
Leslie, D. (2001) Effect of river management on colonially-nesting
waterbirds in the Barmah-Millewa Forest, south-eastern Australia.
Regulated Rivers: Research & Management 17, 21-36.
123

Monitoring and evaluation of environmental flow releases in the Campaspe River
124
Scott, A. (1997) Relationships between waterbird ecology and river flows in
the Murray Darling Basin. CSIRO Land and Water, Tech. Rep. 5/97,
Canberra.
Briggs, S.V. and Thornton, S.A. (1999) Management of water regimes in
river red gum Eucalyptus camaldulensis wetlands for waterbird breeding.
Aust. Zool. 31,187-197.
BMF (2001) Report on Barmah-Millewa Forest Flood of Spring 2000 and
the Second Use of the Barmah-Millewa Forest Environmental Water
Allocation, Spring Summer 2000/2001. Barmah-Millewa Forum.

Monitoring and evaluation of environmental flow releases in the Campaspe River
Appendix 3: Bayesian Hierarchical Modelling & Illustrative
Case-study
Bayesian hierarchical modelling has been advocated as being applicable to the analysis of the
VEFMAP data. This approach represents a substantial innovation for the analysis of the outcomes of
environmental flow management, and thus was discussed at length in the technical meetings
associated with this project. In order to expand on the technical reasons for the consideration of
Bayesian modelling in this project we have developed the hypothetical case study discussed below.
The simple mock analysis demonstrates the effects of using Bayesian hierarchical modelling in the
analysis of environmental monitoring data. We aim to demonstrate that the hierarchical approach
leads to more precise estimates of site-level parameters, while at the same time having minimal
effects on mean values.
The logical basis of Bayesian hierarchical modelling (drawn largely from Gelman et
al. 1995, Chapter 5)
Many problems involve estimating multiple parameters that may be able to be regarded as related in
some way. The hierarchical analysis framework allows us to formalise this relatedness by setting up
a joint probability model that reflects the dependence among parameter values.
For data concerning the ecological effects of environmental flow augmentations, we will collect data
(e.g. plant germination rate), and estimate parameter values (e.g. relationship coefficient between
germination and fresh frequency) from a number of sites. We can take three positions with regards
to the relatedness of the data from each site. Firstly, we could regard the parameter estimates from
each site as being independent estimates of the same true parameter value. That is, the sites are
considered to respond identically to flow augmentation. In such a case, we could pool data from the
various sites to obtain the best estimate of the true parameter value, acting in effect as if we had
many more data points from the one site. Such a step would be regarded as pseudoreplication by
most ecologists, with the individual sites being the appropriate unit of replication (Hurlbert 1984).
Second, we might believe a priori that the sites are so different to one another that the parameter
estimate at any site tells us nothing about the parameter value at another site. In this case, the data
could be analysed separately for each site, or handled with some form of nested analysis, with the
explicit assumption that the sites are fully independent. Neither of these attitudes to the data seems
particularly satisfying for most ecological questions, where we might believe that a parameter value
at one site should be similar to that at another site (within reason, given the diversity of possible
sites). We would like to use the information contained within the data for other sites to improve
inference, but it cannot simply be pooled. This is achieved by the third possibility; that we regard the
sites neither as identical nor completely independent, but exchangeable. Exchangeability implies
that the parameter estimates from each site will differ, but that they can be thought of as being
drawn from a possible distribution of parameter values. Differences among parameter values exist,
but we cannot a priori predict what these differences might be, as the parameter values themselves
are viewed as samples from a random variable.
Computationally, the above is achieved by specifying a joint probability model for all sites, where the
data for each site are modelled as being conditional on certain parameters (e.g. a regression slope),
which themselves are conditional on higher-level parameters – termed hyperparameters. The
hyperparameter becomes the prior distribution for the site-level parameter, and a prior distribution
must be specified for the hyperparameter.
125

Monitoring and evaluation of environmental flow releases in the Campaspe River
Practically, the consequence of taking this approach is that the means of site-level parameter
estimates will be drawn towards the overall mean of the sites. This property is known as shrinkage.
The means for sites with more data will be less affected than those for sites with fewer data. This
seems intuitively correct, because we will have more confidence in the parameter estimates for sites
with more data. The posterior estimates for the site-level parameters will also be more precise in the
hierarchical model than those for the individual sites models, because the hierarchical model
effectively encodes information from the other sites into the prior distribution, thereby tightening the
posterior estimate.
Hierarchical models tend to avoid over- or under-fitting of large data sets. A non-hierarchical model
cannot fit a large data set very well with few parameters. In the example above, pooling data from
many sites and analysing with a simple model would tend to underfit the data due to site-to-site
differences ignored in the pooled model. Conversely, a non-hierarchical model with many
parameters may overfit the data, in that it will fit the data well, but will be of little use in making
predictions about new or unobserved data. Hierarchical models can have sufficient parameters to fit
the data, while using hyperparameters to model some dependence between parameter estimates,
which avoids over-fitting.
The idea that sites will be exchangeable will be appropriate for some parameters, but may not be for
others. In some aspects, we would expect the response to flow augmentation to differ systematically
between rivers for reasons that we both understand and can measure. In this case, we cannot in
good conscience model the parameter values as being drawn from the same distribution. Nor would
we wish to, as we know something about each site that leads us to believe that the parameter
values will be different. In such a case, the concept of conditional exchangeability may apply. The
parameter estimates are seen as being drawn from the same distribution, conditional upon the value
of one or more covariates. The simplest way of explaining this is to consider a simple linear
regression where the data fit perfectly, and to ignore for the moment any uncertainty. When we
analyse the data, we fit each data point to a model – a straight line described by two parameters: the
slope () and intercept (). Using this model, we can then predict new y values, using the
relationship yi =  + xi, where xi is a value of the independent variable not included in the original
analysis (but usually not beyond either extreme of the previous data). This prediction is only possible
because we are treating the original y1-n values as identical conditional upon the calculated values of
 and  and the measured values x1-n. In the real world, the data will not fit the model perfectly.
There will be uncertainty in  and , meaning that we treat the data as exchangeable conditional on
  and x1-n. Any predicted data point yi will also be uncertain, but the model still uses the
information encoded in the original data to make the prediction. The only difference between this
situation and a Bayesian hierarchical model is that the Bayesian model will use information from the
other data points to inform inference about each individual data point (e.g. its measurement
uncertainty) as well as to inform the estimates of  and  Practically however, the mean calculated
values of  and , from a Bayesian regression will be almost identical to those from a least squares
regression if vague prior distributions are used for these parameters. Conditional relationships
underlie all models where data are fitted to some function. They are not confined to Bayesian
analyses. From the point of view of monitoring ecological effects of environmental flows, conditional
exchangeability may allow us to consider within the same model sites that behave quite differently,
as long as we have a conceptual basis that allows us to build these differences into the model
structure.
126

Monitoring and evaluation of environmental flow releases in the Campaspe River
Case study – Germination of riparian vegetation
The hypothetical (and rather fanciful) case study chosen concerns the rates of germination of
species X of native riparian vegetation. The question is whether or not, at the site level,
environmental flows (measured as the frequency of spring freshes) leads to increased germination
of the species. The measurement endpoint is the density of shoots observed following spring
freshes designed to inundate benches and initiate germination. At each site, we have a single
density estimate taken each spring over a three year period, giving a sample size of three in all. At
this time, we have ignored concerns such as temporal (and also spatial) autocorrelation of data.
When real data are collected, these effects will need to be incorporated in models, a step more
elegantly achieved in Bayesian modelling (e.g. Congdon 2001) than the ad hoc solutions usually
applied to frequentist models (reviewed in Lloyd 2001).
The main driving variable is the proportional achievement each spring of the recommended number
of bench-inundating flows. Due to operational constraints, the recommended number may not be
met each year, and the actual number will differ from year to year. At its simplest, therefore, the
analysis at each site is a linear regression of shoot density versus flow achievement.
We also have a priori reasons to expect that the density of shoots at a site will partly be a function of
the condition of riparian vegetation for a certain distance (say 10 km) upstream of the site. Good
quality riparian zones will supply more propagules of the species of interest, and hence can be
expected to affect the amount by which flow augmentation increases the density of shoots.
Beyond this, we expect that the elevation of the site might also affect flow augmentation of
germination, with species X previously shown to favour slightly higher elevations. In terms of data
for this hypothetical situation, we have three points per site (one each year), with two sites per
reach, two reaches per river, and two rivers.
The most all-inclusive model for the above data is a Bayesian hierarchical model where the
regression slope between flow and germination at each site is modelled as a linear function of
riparian condition for that site within each reach. The reaches are at different elevations, and so the
relationship between riparian condition and flow effects are modelled as a linear function of elevation
within each river. Finally, the rivers are considered as exchangeable entities, in that for this endpoint,
we believe that the flow-germination relationships for each river could be drawn from the same
distribution of parameter values. These assumptions can all be tested by using post-hoc predictive
tests (Gelman et al. 1995) to determine whether the data could have been produced by the model
proposed. We do not perform such tests here, but they should be a mandatory part of any analysis
of real data.
The above analysis considers all the data simultaneously, and uses all available information to come
up with estimates of parameter values. If the sites / reaches / rivers are behaving similarly,
conditional on the covariates already discussed, this will lead to more precise estimates of
parameter values than is possible if a subset of the data is considered. We will also examine a
situation where one site behaves differently to the others, in a way that is not consistent with the
conditional relationships proposed. However, it is not essential that the data be considered together.
For the dataset described above, we present results for analyses conducted at the site level (i.e. the
simple regressions, although conducted as Bayesian analyses), together with analyses of increasing
levels of hierarchies. At the reach level, data from two sites are considered simultaneously, along
with the riparian condition covariate. At the river level, data from four sites from each of two reaches
are considered simultaneously along with the riparian and elevation covariates. All of the results
discussed, however, relate to responses at the site level. The only difference between the analyses
127

Monitoring and evaluation of environmental flow releases in the Campaspe River
is that in the increasingly hierarchical models, the prior distribution for each site-level parameter is
informed by data from more and more sites. For the simplest analyses, we are conducting
regressions with three data points. This will obviously result in low-powered tests. Thus it is
desirable to consider at least multiple sites in the same reach to improve inferential strength.
The models described above were written and run in WinBUGS 1.4.1 software using made up data. We
present results for two parameters of interest. The first is the predicted density of shoots of species X,
given 100% delivery of recommended ‘bench inundating flows’. This parameter facilitates comparison of
the flow-germination relationship between sites that are exposed to different levels of flow
augmentation. As a continuous parameter, this result is presented along with the 95% credible interval
for parameter values. The second parameter is the probability associated with the hypothesis test of
‘benefit of flow on germination’ – simply what is the probability that more bench inundating flows will
lead to more germination, given the data at hand? A high probability (near 1) indicates strong support
for the stated hypothesis. A low probability (near 0) indicates support for the opposite hypothesis, rather
than the absence of an effect. Thus a probability of 0.1 would indicate that flows are leading to reduced
recruitment. A probability near 0.5 supports the absence of an effect – the null hypothesis. In the
hierarchical models, only the slope parameters of the regression were modelled hierarchically. The
intercept parameters were calculated separately for each site / reach / river. This was an arbitrary
decision, and is open to challenge in the future. There are no hard and fast rules about how such
models should be created. All continuous means were assigned vague normal prior probability
distributions, and precision parameters were assigned vague gamma distributions.
The data used in the analysis are reproduced here. Riparian condition was assumed constant over the
three years (maybe a poor assumption), and elevation was assumed equal for the two sites in each
reach (probably more defensible)
Flow Germ Ri
p
elev
0.8 30
0.6 29 0.6
0.9 40
150
0.3 15
0.4 25 0.5
0.8 30
0.7 40
0.6 38 0.7
0.8 41
200
0.5 50
0.7 62 0.9
0.9 66
0.9 35
0.7 27 0.5
0.9 40
170
0.7 20
0.4 12 0.4
0.8 24
0.6 36
0.6 38 0.6
0.8 41
190
0.5 50
0.7 65 0.9
0.9 67
Table 19. Data used in the analysis
128

Monitoring and evaluation of environmental flow releases in the Campaspe River
Results for the 2 parameters are presented graphically on the following pages. The effect on the
median predicted shoot density at 100% flow recommendations is very small. It is worth noting
however, that if more than two sites had been included for each reach, or more than two reaches for
each river, the effect may have been greater if there were any non-linearities in the covariate
relationships. As discussed above, omitting the covariates would also have drawn the means for
various sites closer together, but his would probably be an inappropriate model structure. More
noticeable is the effect of the hierarchical analysis on parameter uncertainty. The 95% credible
intervals are smaller for the analyses conducted at greater hierarchical levels, although the
difference is greatest between the site-level versus reach-level analyses.
Unsurprisingly, the effect of this increased precision is that the hypothesis test is more strongly
supported in the hierarchical analyses. There is one exception to this, for site 7. Although there was
no attempt to make it as such, the data for this site may not support the covariate models as well as
that for other sites.
Data with an ‘odd’ site
The hierarchical analysis works very well for the above data, where the same general relationship is
seen at all sites, and where the effects of riparian vegetation is constant between sites, and that of
elevation is constant between reaches. What if the data for one site are very different to all others?
We altered the data for site seven such that a decline in germination is observed with increased
flow, despite good riparian condition upstream. The new data are shown below. This contradicts
results for all other sites. Perhaps in this case, the flow leads to erosion of benches, and thus loss of
propagules. If such information was known, it may be able to be factored into the larger model, but if
we are ignorant of such effects, we may seek to analyse the data along with the others, using an
inadequate model.
Flow Germ Ri
p
0.6 36
Ori
g
inal Data 0.6380.6
0.8 41
0.6 36
New Data 0.6341.0
0.8 27
Table 20. Data for site seven.
129

Monitoring and evaluation of environmental flow releases in the Campaspe River
Site 1
-50
0
50
100
150
site reac h river st ate
Site 2
-50
0
50
100
150
site reach river state
Site 3
-50
0
50
100
150
site reach river state
Site 4
-50
0
50
100
150
site reach river state
Site 5
-50
0
50
100
150
site reach river state
Site 6
-50
0
50
100
150
site reach river state
Site 7
-50
0
50
100
150
site reach river state
Site 8
-50
0
50
100
150
site reach river state
Table 21. Median and 95% credible intervals for the predicted number of
germinations at each site when the full environmental flow recommendations are
met as calculated by site-level analysis, and reach-, river- and ‘state’- level
hierarchical analyses.
130

Monitoring and evaluation of environmental flow releases in the Campaspe River
Site 1
0.8
0.85
0.9
0.95
1
site reach river st ate
Site 2
0.8
0.85
0.9
0.95
1
site reac h river s tate
Site 3
0.8
0.85
0.9
0.95
1
site reac h river s tate
Site 4
0.8
0.85
0.9
0.95
1
site reac h river s tate
Site 5
0.8
0.85
0.9
0.95
1
site reac h river s tate
Site 6
0.8
0.85
0.9
0.95
1
site reac h river s tate
Site 7
0.8
0.85
0.9
0.95
1
site reac h river s tate
Site 8
0.8
0.85
0.9
0.95
1
site reac h river s tate
Table 22. Probabilities for the hypothesis test of ‘beneficial effects of flow
augmentation on germination rate’ for each site as calculated by site-level analysis,
and reach-, river- and ‘state’- level hierarchical analyses.
131

Monitoring and evaluation of environmental flow releases in the Campaspe River
We explore the effects of including this analogous site on the same two variables from the full
hierarchical model.
Original Hierarchical results Results with new site 7 data
2.5th
% Median 97.5th
% Pr(benefit) 2.5th
% Median 97.5th
% Pr(benefit)
20.7 40.7 61.3 0.91 20.1 40.8 62.1 0.91
2.8 35.6 66.4 0.90 1.9 35.5 67.9 0.89
35.3 44.2 53.2 0.95 36.2 44.2 53.5 0.95
45.7 71.3 95.7 0.94 44.4 71.2 95.3 0.93
28.1 41.7 54.8 0.97 15.1 41.0 57.1 0.80
22.1 29.5 39.1 0.99 21.1 29.5 46.5 0.99
18.7 44.7 63.2 0.86 5.6 20.2 62.4 0.14
40.8 74.3 116.7 0.93 14.6 69.5 96.7 0.72
-13.14 19.01 51.00 0.01
Table 23. Site level analysis of new site 7 data
It is apparent that the change in data has affected the results from other sites, but in general only by
a small amount. The median number of shoots predicted is virtually unaffected by the change in site
7 data, except for site 8, where the dependency among sites within reaches means that the
expected shoot number is decreased by approximately 5. The probabilities are similarly affected; the
conclusion for site 5 is also somewhat weaker than it was with the original data. For the sites where
probability has changed, it is noticeable that the credible intervals are wider. The sites from river 1
(1-4) are virtually unaffected by the change in data for site 7. Comparing the hierarchical result for
the new data set to a site-level analysis of the new data shows that the median is again only slightly
affected by the data from other sites, but that the probability has been ‘dragged’ towards that for the
other sites within the hierarchical analysis.
With only two sites per reach, the effects of one ‘bad’ site will be exacerbated on the other site in
that reach, but the other sites in the analysis will be less affected. If there were more than two sites,
the effects would be noticeable on more sites, but these effects would be smaller.
Practical Implementation
The examples above make a relatively strong argument for the use of hierarchical models. Indeed, it
seems impossible to avoid some form of hierarchical model unless we commit to only analysing data
from the same site within a single analysis. However we do not advocate the blind use of large-scale
hierarchical models in situations where they may be inappropriate. For instance, the reaction to a
given environmental flow regime in say, the Wimmera River, may be very different to that observed
in the Thomson, and we may not be able to explain these differences by the use of covariates in
statistical models (i.e. the rivers are just different, and we don’t know why). In such a case, we
should not analyse the data together in the same model, and would negatively impact on quality of
inference for both rivers. Similarly, if two rivers are experiencing vastly different flow augmentation
programs, and we cannot logically express the differences within the model, their results should be
analysed separately. These types of difference should become apparent during exploratory data
analysis, and the appropriate models chosen. All statistical models should only be used when the
data fit the assumptions of the model, and the use of an inappropriate model is not confined to
Bayesian analyses. Bayesian models do allow us, however, to conduct posterior predictive checks
132

Monitoring and evaluation of environmental flow releases in the Campaspe River
133
to determine whether the data are consistent with the model. Such tests should always be carried
out to determine the validity of any inference.
Conclusions
This analysis has demonstrated that the main advantage of a hierarchical analysis is the increased
precision with which site-level parameters may be estimated. Because the hierarchy exists at many
levels (reach, river, state) it is not the case, as perhaps previously implied, that data must be
analysed at a state-wide level to take advantage of this analysis framework. However, the more
analyses that can be done at this level, the stronger the inference, and commonality of monitoring
programs, where practical, should still be sought so that this option is available when data are
analysed. The ‘odd’ site example shows that results from other sites can be affected by a site that
behaves very differently, but that the effects are probably not that serious. In any case, if we noted
that one site was behaving very differently to all others in preliminary data examinations, we would
probably seek an explanation that could be built into a large scale model or analyse that site’s data
separately.