Principal Investigators: M. Thoms and R. Norris
Research Officers: M. Parsons and G. Ransom
Cooperative Research Centre for Freshwater Ecology
Monitoring River Health Initiative Technical Report Number 23
Environment Australia, 2002
ISSN 1447 1280
ISBN 0 642 54889 7
The River Habitat Audit Procedure, Index of Stream Condition, River Styles and Habitat Predictive Modelling were designed for slightly different purposes. Subsequently, each method differs in its compatibility with the requirements of a standardised physical and chemical assessment protocol (Table 4.3). Each method performs equally well against criteria such as 'ability to assess stream condition against a desirable reference state', and 'applicability to all stream types within Australia'. However, only one or two methods perform well against criteria such as 'ability to predict physical stream features that should occur in disturbed rivers and streams' and 'outputs of physical condition that are comparable to AusRivAS outputs of biological condition' (Table 4.3). Overall, no one method meets all the requirements for a stand-alone stream assessment protocol. However, each method contains important individual components that can be combined into a comprehensive protocol for assessing stream physical and chemical condition. The components of each method that will be incorporated into the physical and chemical assessment protocol are outlined below:
As mentioned in Appendix 1, three major considerations in the design of the physical and chemical assessment protocol will be scale of focus, determination of reference condition and comparisons of similar stream types. The River Habitat Audit Procedure, Index of Stream Condition, River Styles and Habitat Predictive Modelling methods adopt broadly similar approaches to these three aspects. For example, each method samples at a reach scale, each assesses condition against some desirable reference state, and each contains a mechanism for delineating stream types with similar physical features. However, there are also several differences between methods that may influence the incorporation of these aspects into the physical and chemical assessment protocol. For example, Habitat Predictive Modelling uses classification analysis to group similar stream types, whereas the River Habitat Audit Procedure uses a combination of map examination and field reconnaissance to determine homogeneous stream sections.
The Habitat Assessment Workshop (Appendix 1) identified areas of concern for the development of a standardised physical and chemical assessment protocol. Evaluation of four existing stream assessment methods against specific criteria (Section 4.2.2) then identified the aspects of each method that should be included in the protocol (Section 4.2.3). The following sections bring together the habitat assessment workshop and method evaluation outcomes, to make recommendations about the major components that will be included in a standardised physical assessment protocol.
The overall approach of the physical and chemical assessment protocol will be based on Habitat Predictive Modelling (Davies et al., 2000). Habitat Predictive Modelling confers four main advantages in the assessment of stream condition. Firstly, it is the only existing method that is able to predict the local-scale habitat features that would be expected to occur at a site in the absence of degradation. These features can then be compared against the features that were actually observed at that site, with the deviation between the two being an indication of physical stream condition. Secondly, Habitat Predictive Modelling uses multivariate data analysis, which allows simultaneous consideration of multiple physical factors. Thirdly, Habitat Predictive Modelling can be easily modified to incorporate aspects of other stream assessment methods and in fact, is strengthened by the addition of components from other methods (see Section 4.2.4.2).
The fourth advantage of using Habitat Predictive Modelling as the basis for the physical and chemical assessment protocol is its compatibility with AusRivAS. In the data collection phase, many of the current AusRivAS variables that are deemed relevant to physical assessment will be included in the protocol (see Section 4.2.4.5). Habitat Predictive Modelling also uses the same model building technique as AusRivAS and thus, the outputs of physical stream condition from Habitat Predictive Modelling (observed:expected ratios and probabilities of occurrence) are identical to those of AusRivAS. In addition, it is envisaged that once physical and chemical data is collected and models constructed, the existing AusRivAS predictive model software would require only minor modifications to accommodate a fully functioning interface for State agency users.
Although Habitat Predictive Modelling (Davies et al., 2000) forms a sound basis for the physical and chemical assessment module, it can be strengthened considerably with the addition of aspects from other physical assessment methods. In particular, the River Habitat Audit Procedure (Anderson, 1993), River Styles (Brierley et al., 1996) and Index of Stream Condition (Ladson and White, 1999) incorporate aspects of physical assessment that are not included in Habitat Predictive Modelling (see Section 4.2.3). Thus, the Habitat Predictive Modelling approach of Davies et al. (2000) will be modified to include sampling design, data collection and analytical components from these other stream assessment methods.
It is generally accepted that stream systems are arranged hierarchically and that physical processes operating at larger scales constrain the expression of physical processes at successively smaller scales (Schumm and Lichty, 1965; Frissell et al., 1986). Habitat Predictive Modelling uses this relationship to predict the occurrence of local-scale stream features on the basis of large-scale catchment characteristics (Davies et al., 2000). Similarly, River Styles uses a hierarchical approach that measures physical character and geomorphological behaviour at the catchment, reach and geomorphic unit scales (Brierley et al., 1996). Merging the hierarchical approaches of Habitat Predictive Modelling and River Styles has the potential to improve prediction of stream habitat features by encompassing the scales that represent geomorphological processes, as well as the scales that represent aspects of the physical environment that are important for macroinvertebrates.
The use of a hierarchical approach has implications for several areas of the physical and chemical assessment protocol. Firstly, Habitat Predictive Modelling uses large-scale characteristics to predict local-scale stream features. The use of a hierarchical approach will assist in determining the link between the large-scale 'controlling' variables and the local-scale features that can be predicted using these larger scale variables. Secondly, physical and chemical variables will be measured across the range of scales that represent geomorphological processes, and across the range of scales that represent the physical environment for macroinvertebrates. This will ensure that the factors influencing physical stream condition and the factors important for macroinvertebrates are included in the predictive model. Lastly, the delineation of groups of reference sites with similar local-scale features is a pivotal step in Habitat Predictive Modelling. Given that river systems are organised hierarchically, meaningful grouping of reference sites on the basis of local-scale features requires some knowledge of the characteristics of the constraining levels of the hierarchy within which these local-scale features sit.
A recent development in river assessment has been the use of reference conditions, rather than reliance on single sites as controls. These reference conditions then serve as the control against which test site conditions are compared. Given the caveat that man is part of the landscape, the determination of reference condition is essentially one of best available condition (Norris and Thoms, 1999). The AusRivAS predictive models incorporate the concept of reference condition in two ways. Firstly, sampling sites are selected to represent least impaired biological reference conditions. The characteristics used to discriminate least impaired reference sites include: the degree of urbanisation, forestry and agriculture in the catchment; the influence of major impoundments, extractions or diversions; the influence of point sources of pollution; the degree of channel modification; and, the degree of bank degradation (Davies, 1994). It is assumed that reference sites with minimally disturbed physical characteristics will contain unimpaired macroinvertebrate communities. Secondly, the physical and biological information collected from these reference sites is used to build the predictive models and thus, this information becomes the template against which test sites are compared to assess their ecological condition. Procedures exist to 'weed out' reference sites that do not represent least impaired conditions (Coysh et al., 2000) but ultimately, the definition of reference condition is critical to the biological assessments of stream condition that are made by the AusRivAS predictive models.
Habitat Predictive Modelling uses the same reference condition approach and modelling techniques as AusRivAS (Davies et al., 2000). As with AusRivAS, the definition of geomorphological or physical reference condition will be critical to the assessments of physical stream condition that are made using Habitat Predictive Modelling. However, the geomorphological reference condition used in Habitat Predictive Modelling has different properties from the biological reference condition used in AusRivAS. Thus, it will be necessary to determine the characteristics that represent least impaired physical and geomorphological condition. The Habitat Assessment Workshop (see Section 4) identified several parameters that could be used to guide the selection of appropriate reference sites for geomorphological or physical assessment of stream condition.
These were:
Additionally, the Index of Stream Condition (Ladson and White, 1999) and River Habitat Audit Procedure (Anderson, 1993) each compare sampling sites against some type of reference state, and these reference states may be incorporated into the identification of geomorphological or physical reference condition for Habitat Predictive Modelling. For example, the rating scores used for each indicator in the Index of Stream Condition are derived from knowledge of a continuum of good to poor condition. Similarly, the condition ratings used in the River Habitat Audit Procedure express condition as a percentage of the original values, functions or utilities that are retained at a sampling site. Subsequently, the threshold reference values for individual data components that are used in both of these rating processes can be used to aid the identification of characteristics that represent geomorphological or physical reference condition.
The development of a standardised physical and chemical protocol that is complementary to AusRivAS biological assessments may also require consideration of biological reference condition. Macroinvertebrates are inextricably linked to their physical surroundings (Southwood, 1988; Townsend and Hildrew, 1994) and as a result, the condition of the macroinvertebrate community is directly related to the condition of the physical habitat (Barbour, 1991; Rankin, 1995). However, it is possible for sites that are physically degraded to contain a relatively healthy fauna or conversely, it is possible for sites in good physical or geomorphological condition to contain a degraded fauna, because of water quality influences. Merging an assessment of the ability of the physical habitat to support biota into the determination of geomorphological reference condition may allow distinction between the effects of habitat impairment and the effects of water quality degradation.
The River Habitat Audit Procedure, Index of Stream Condition, River Styles and AusRivAS stream assessment methods measure a range of physical, chemical and habitat variables. The types of variables measured in these methods correspond to the purpose and context of each method. The variables collected in River Styles (Brierley et al., 1996) are important for the geomorphological characterisation of stream systems at the catchment, reach and geomorphic unit scales. The River Habitat Audit Procedure (Anderson, 1993) measures a comprehensive set of physical variables that describe structural components of streams and which indirectly 'summarise' fluvial processes. The Index of Stream Condition (Ladson and White, 1999) incorporates indicator variables that encompass the important influences on stream condition, and which are able to detect changes in stream condition over a 5 year assessment interval. AusRivAS (Coysh et al., 2000) and Habitat Predictive Modelling (Davies et al., 2000) measure physical variables that are important for macroinvertebrates. As such, there are many variables that can potentially be measured in the physical and chemical assessment protocol.
The robustness and power of a predictive model based solely on physical characteristics is directly dependent on the variables that are available to construct the models, and on the physical processes that these variables indicate. Habitat Predictive Modelling uses large-scale characteristics to predict local-scale stream habitat features and is based on the paradigm that large-scale features constrain the expression of small-scale features (see Section 4.2.4.3). Thus, successful prediction of small-scale features from large-scale characteristics relies on the collection of a range of variables that can be linked hierarchically. Similarly, Habitat Predictive Modelling uses the deviation between the predicted occurrence and observed occurrence of small-scale features as a measure of physical stream condition. Thus, the determination of stream condition using Habitat Predictive Modelling relies on the inclusion of small-scale features that represent a range of fluvial processes, and which may indicate the impacts of stream degradation.
To encompass both the hierarchical linkages and the indicators of degradation within a standardised assessment protocol, it will be necessary to incorporate variables from the Index of Stream Condition, River Habitat Audit Procedure and River Styles. However, variables chosen for inclusion in a standardised protocol must conform to the overriding criteria of being rapid and easy to measure, and must also provide information about physical aspects of stream condition (see Section 4.2.2.5). Additionally, the assessment protocol must have biological relevance and thus, AusRivAS variables that represent the environmental requirements of macroinvertebrates will also be included.
Anderson, J.R. (1993) State of the Rivers Project. Report 1. Development and Validation of the Methodology. Department of Primary Industries, Queensland.
Barbour, M.T. (1991) Stream surveys - the importance of the relation between habitat quality and biological condition. Sediment and Stream Water Quality in a Changing Environment: Trends and Explanation. Proceedings of the Vienna Symposium, August 1991. IAIIS Publication No. 203.
Brierley, G., Fryirs, K. and Cohen, T. (1996) Development of a generic geomorphic framework to assess catchment character. Part 1. A geomorphic approach to catchment characterisation. Working Paper 9603, Macquarie University, Graduate School of the Environment.
Coysh, J.L., Nichols, S.J., Simpson, J.C., Norris, R.H., Barmuta, L.A., Chessman, B.C. and Blackman, P. (2000) Australian River Assessment System (AusRivAS). National River Health Program, Predictive Model Manual. Cooperative Research Centre for Freshwater Ecology, University of Canberra, Canberra, Australia. http://ausrivas.canberra.edu.au/ausrivas
Davies, N.M., Norris, R.H. and Thoms, M.C. (2000) Prediction and assessment of local stream habitat features using large-scale catchment characteristics. Freshwater Biology, 45: 343-369.
Davies, P.E. (1994) Monitoring River Health Initiative River Bioassessment Manual. Freshwater Systems, Tasmania.
Frissell, C.A., Liss, W.J., Warren, C.E. and Hurley, M.D. (1986) A hierarchical framework for stream habitat classification: viewing streams in a watershed context. Environmental Management, 10: 199-214.
Ladson, A.R. and White, L.J. (1999) An Index of Stream Condition: Reference Manual. Department of Natural Resources and Environment, Melbourne, April 1999.
Norris, R.H. (1994) Rapid biological assessment, natural variability, and selecting reference sites. In: Uys, M.C. (ed.) Classification of Rivers, and Environmental Health Indicators. Proceedings of a joint South African/Australian workshop, Cape Town, South Africa. Water Research Commission Report No. TT 63/94. pp. 129-166.
Norris, R.H. and Thoms, M.C. (1999) What is river health? Freshwater Biology, 41: 197-209.
Parsons, M., Thoms, M.C. and Norris, R.H. (2000) Review of physical river assessment methods: a biological perspective. Cooperative Research Centre for Freshwater Ecology and Environment Australia. 59pp.
Rankin, E.T. (1995) Habitat indices in water resource quality assessments. In: Davis, W.S. and Simon, T.P. (eds.) Biological Assessment and Criteria: Tools for Water Resource Planning and Decision Making. CRC Press, Boca Raton. pp. 181-208.
Reynoldson, T.B., Norris, R.H., Resh, V.H., Day, K.E. and Rosenberg, D.M. (1997) The reference condition: a comparison of multimetric and multivariate approaches to assess water-quality impairment using benthic macroinvertebrates. Journal of the North American Benthological Society, 16: 833-852.
Schumm, S.A. and Lichty, R.W. (1965) Time, space and causality in geomorphology. American Journal of Science, 263: 110-119.
Southwood, T.R.E. (1988) Tactics, strategies and templets. Oikos, 52: 3-18.
Townsend, C.R. and Hildrew, A.G. (1994) Species traits in relation to a habitat templet for river systems. Freshwater Biology, 31: 265-275.