River Health in the New South Wales Lower North Coast, Hunter and Central Coast Catchments

Caution: archived content

This content may have been superseded, or served a particular purpose at a particular time. It may contain references to activities or policies that have no current application. Many archived documents may link to web pages that have moved or no longer exist, or may refer to other documents that are no longer available.

A report of AusRivAS assessments 1994 - 1999
Grant Hose and Eren Turak
Environment Protection Agency, 2004
ISBN 0 642 55099 9 ISSN 1447-1280

3. Australian River Assessment System (AusRivAS) Explained (continued)

3.3 The use of AusRivAS in assessment of river health

AusRivAS compares the macroinvertebrates found at a test site with those collected from reference sites similar to it. Some understanding of how the predictive models work is essential for the effective use of AusRivAS. The following section provides a brief overview of AusRivAS outputs. Further information can be found in Coysh et al. (2000), Turak and Waddell (2002b), Barmuta et al. 2002, Gray 2004

3.3.1. Understanding AusRivAS outputs

The predictive models in AusRivAS generate probabilities of finding each of the different types of macroinvertebrates at any given site. The term "taxa" (singular taxon) is used to describe the types of invertebrates included in the model, with these relating mostly to macroinvertebrate families, however sometimes other categories are also included such as class, order and sub-family. The basic output of the program for a sample site assessment is a table containing all the taxa used in each of the models and the probabilities of occurrence of each of those taxa at that site. All other outputs of the program are derived from this table.

Group Probability

The first AusRivAS output is the group probability which lists the probability of the test site belonging to each one of the general reference site groups. This output provides information to describe which group (or groups) of reference sites the selected test site is most similar to. The higher the probability of a group, the greater the mathematical weighting of the taxa collected from that group in the predictions made for the site.

3.3.2. O/E Assessment results

For assessments of river health, AusRivAS derives O (Observed) / E (Expected) values. These values compare a summary value of the fauna collected at a site (Observed) with the a summary value of the fauna predicted (Expected) at a site. There are two different types of O/E values presented in this report. These are the O/E-Taxa and O/E-SIGNAL. The O/E-Taxa

This index uses a comparison between observed and expected results. The first step in these assessments is the generation of a list of macroinvertebrate taxa that are likely be found at the site (the "Predicted Taxa"). The "Predicted Taxa" are those which have a 50 % greater probability of being collected in a macroinvertebrate sample (based on the data collected from reference sites). The number of the predicted taxa that were collected at the site is the "observed taxa" (O taxa). The number of the predicted taxa collected at a site is the expected taxa (E taxa). The O/E-Taxa value is obtained by dividing (O) Observed taxa by (E) Expected taxa. O/E-Taxa indicates what proportion of the taxa predicted to be found at the site were actually found there. O/E-Taxa below 1 can be interpreted as having lost some of the taxa. For example a site with an O/E-Taxa value of 0.8 can be said to have lost 20 % of its taxa. O/E-SIGNAL

The SIGNAL results relates to the assessment methodology developed by Chessman (2003) where a pollution tolerance score is assigned to individual Australian macroinvertebrate families. The individual taxa SIGNAL score represents the pollution tolerance rating of the taxa. The E (Expected)SIGNAL score is derived from the average SIGNAL score of the predicted taxa. O-SIGNAL is the average SIGNAL score of the observed taxa.

Where low O/E-SIGNAL values are found it suggests a shift in the fauna types towards those taxa that have a greater tolerance to pollution. The extent to which O/E-SIGNAL falls below 1 could be referred to as % reduction of general sensitivity to pollution.

3.3.3 Bands of Impairment Used for Reporting Results

For ease of interpretation, AusRivAS includes numerical bands that categorise O/E values to reflect the associated level of ecological impairment. An example is shown in Table 1. These bands are determined statistically from the O/E values of reference sites. Band X represents a richer invertebrate community than reference, Band A is considered equivalent to reference; Band B represents sites below reference condition; Band C represents sites well below reference condition; and Band D represents impoverished sites. Further explanation of how indices and bands are calculated can be found in the AusRivAS predictive modelling manual (Coysh et al. 2000). Also see Gray 2004.

Table 1: AusRivAS Reporting Bands for NSW Single Season Eastern Edge Model.
Band Label Upper Limit Band Name Band Description
Band X None More biologically diverse than reference sites. More taxa found than expected. Potential biodiversity hot-spot. Possible mild organic enrichment
Band A 1.17 Reference condition. Most/all of the expected families found. Water quality and/or habitat condition roughly equivalent to reference sites. Impact on water quality and habitat condition does not result in a loss of macroinvertebrate diversity.
Band B 0.81 Significantly impaired. Fewer families than expected. Potential impact either on water quality or habitat quality or both resulting in loss of taxa.
Band C 0.46 Severely impaired. Many fewer families than expected. Loss of macroinvertebrate biodiversity due to substantial impacts on water and/or habitat quality.
Band D 0.11 Extremely impaired. Few of the expected families remain. Extremely poor water and/or habitat quality. Highly degraded.