Melissa Parsons, Martin Thoms and Richard Norris
Cooperative Research Centre for Freshwater Ecology
University of Canberra
Monitoring River Health Initiative Technical Report Number 22
Environment Australia, 2002
ISSN 1447-1280
ISBN 0 642 54888 9
Identification of areas that are potentially impacted by large scale and local scale activities allows the elimination of these areas as potential sources of reference sites.
Disturbances that may potentially be impacting the river system are examined at a large catchment scale and at a local scale (see Sections 2.5.2.1 and 2.5.2.2). Sources for obtaining this information on potential disturbances include local managers, experience of agency staff, aerial photographs, hydrology records, GIS maps, and previous data collected for programs such as AusRivAS, individual State or Territory projects or the National Land and Water Audit.
Large scale activities are those which have the potential to effect whole catchments within a river system (Table 2.1).
| Activity | Factors to consider |
|---|---|
| Landuse | Percent cover of native vegetation, percent cover of agricultural or grazing land, time since land clearance, degree of impact of land clearance on the downstream river system, percent cover of urban areas, degree of impact of urban areas on the downstream river system, presence of active (<5 years) logging areas, degree of catchment erosion, degree of sedimentation |
| Hydrological regime |
Presence of major impoundments, downstream effects of major impoundments, degree of change to flooding regime including magnitude and timing, degree of change to flow seasonality, water extraction activities, reductions or increases in velocity, reductions or increases in discharge size. It will be difficult to avoid regulated segments of river in some areas, particularly in lower zones. Where it is impossible to avoid regulation in identifying reference conditions, the overall magnitude of impoundment effects should be considered. |
| Current and historical mining activity | Degree of impact of current mining activities on the downstream river system, degree of impact of historical mining activities on river system character |
Local scale activities are those that may cause localised disturbance to rivers (Table 2.2).
| Activity | Factors to consider |
|---|---|
| Riparian zone characteristics | Presence or absence of riparian vegetation, type of riparian vegetation (native or exotic), influence of exotic vegetation on channel character |
| Channel modification | Channel realignment (straightening or widening etc.), historical incision (ie. severe erosion) of channel, historical infilling (ie. sediment build up) of channel, presence of bridges, fords and culverts and the effects of these on channel character, presence of minor weirs and the effects of these on channel character |
| Desnagging and instream vegetation removal | Historical or recent desnagging, removal of other instream vegetation such as macrophytes |
| Floodplain condition | Connectivity between the river and the floodplain, floodplain erosion, floodplain landuse |
| Human access | Density of public access tracks and roads, location of recreational areas such as camp grounds and picnic areas, presence of road crossings |
| Stock access | Extent of stock access to the channel, impact of stock access on bank condition, impact of stock access on bed condition |
| Bank condition | Extent of non-natural bank erosion, presence or absence of riparian vegetation |
| Point source impacts | Presence of discharge pipes, mining, stormwater discharges, construction sites etc. |
This information on large and local scale activities will be used in Step 5 to determine areas of least impaired condition that are potential sources of reference sites. When using this information it is important to consider the different effects of large scale and local scale impacts. For example, significant forestry activities may occur across a wide area, however, a riparian buffer may exist to protect the stream on a local scale. Conversely, stock may have access to localised patches of river within an otherwise least impaired area and thus, reference sites should not be placed in these localised patches.
Sites assessed by AusRivAS as being in good biological condition can be used to indicate areas of river in least impaired condition. It can also be assumed that sites with a healthy biota will have a healthy supporting habitat.
Plot the location of AusRivAS reference sites (ie. those sites used to construct the predictive models) and any Band A test site (ie. those sites assessed in the First National Assessment of River Health). Mark these sites onto topographic maps.
The identification of 'least impaired' areas within each region and zone will highlight river sections where reference sites can be placed.
Least impaired areas are identified using the information collected in Steps 3 and 4. In each region and zone, mark onto topographic maps the sections of river that are least impaired. These areas are the sections of river where reference sites can be placed.
It is important to include least impaired areas from all the zone types present within a region. However, it is recognised that in comparison to the upper zones, the transitional and lower zone types will contain lower numbers of least impaired areas because it is usually these latter zone types that are most subject to impact. Thus, stringency of the criteria for determining least impaired areas may change among zone types. Relaxation of least impaired status in the transitional and lower zones should be done using supplementary information from previous biological, chemical or physical surveys, or using best professional judgement.
Stratification of reference sites equally across regions and zones within regions will ensure coverage of a range of geomorphological river types. In turn, this coverage will improve the analytical robustness of the physical predictive models (see Section 2.1).
The recommended total number of reference sites to be sampled in each State or Territory is given in Section 2.9. Regardless of the total number of reference sites used, sampling effort should be divided equally among regions and then among functional zones, according to the relative proportion of each zone type in each region. An example stratification of sampling effort across regions and zones is given in Table 2.3.
The final selection of reference sites is achieved by allocating the desired number of sites across zone types located within the least impaired areas identified in Step 5. Existing AusRivAS reference sites should be used where possible, however, additional sites may be required in particular zone types that are not adequately represented in the AusRivAS database. Reference sites should also be spread across a range of different rivers within the region.
The number of reference sites required to construct the physical predictive models is roughly the same as that used to construct the AusRivAS predictive models. The larger States (NSW, QLD, WA, VIC) should sample 230-250 reference sites (minimum 230) and the smaller States and Territories (ACT, SA, TAS, NT) should sample 180-200 reference sites (minimum 180). These figures represent the number of sites required to build the final predictive models. However, it may be necessary to sample additional reference sites to account for situations where sites are excluded post-hoc because of unexpected impairment.
As there are no strongly overriding temporal or seasonal aspects to the measurement of most physical and habitat features, each reference site only needs to be sampled once. Predictive models can be constructed after a single visit to each sampling site, and the subsequent collection of additional office based information (see Part 3).
| Region 1 (50 Sites) | ||
| Zone type | % zone type in region | Number of sites in each zone |
| UZA | 20 | 10 |
| UZB | 40 | 20 |
| TZ | 30 | 15 |
| LZ | 10 | 5 |
| Region 2 (50 Sites) | ||
| Zone type | % zone type in region | Number of sites in each zone |
| UZA | 10 | 5 |
| UZB | 10 | 5 |
| TZ | 70 | 35 |
| LZ | 10 | 5 |
| Region 3 (50 Sites) | ||
| Zone type | % zone type in region | Number of sites in each zone |
| UZA | 10 | 5 |
| UZB | 0 | 0 |
| TZ | 30 | 15 |
| LZ | 60 | 30 |
| Region 4 (50 Sites) | ||
| Zone type | % zone type in region | Number of sites in each zone |
| UZA | 0 | 0 |
| UZB | 70 | 35 |
| TZ | 25 | 12 |
| LZ | 5 | 3 |
| Total Sites | 200 | |
Once the predictive models are constructed using the reference site information, it will be necessary to 'validate' assessments of physical stream condition using information collected from a small set of test sites. A test site is defined as any site at which condition is assessed using the predictive models. The larger States (NSW, QLD, WA, VIC) should sample 20-30 test sites (minimum 20) and the smaller States and Territories (ACT, SA, TAS, NT) should sample 15-20 test sites (minimum 15). Test sites should initially be stratified across the different regions and zones. Within these areas, test sites should then be located to represent a range of disturbances that may potentially influence physical stream condition.