Development and testing of a rapid appraisal wetland condition index in south-eastern Australia

Journal of Environmental Management (1998) 54, 143–159 Article No. ev980212

Development and testing of a rapid appraisal wetland condition index in south-eastern Australia C. Spencer†, A. I. Robertson‡ and A. Curtis† The high costs and lengthy time commitments associated with traditional monitoring, and the remoteness of many wetlands have necessitated the development of techniques for rapid wetland appraisal. A rapid appraisal condition index based on four attributes of wetlands, soil, fringing vegetation, aquatic vegetation and water quality, was developed for assessing the health of permanent floodplain wetlands in the MurrayDarling Basin of south-eastern Australia. The index, composed of 13 indicators related to wetland function, was tested in the field for scientific validity relative to an independent long-term monitoring data set, replicability of indicator scores by different investigators and the responses to the seasonality in wetland processes. Indicator values were based on a mixture of visual estimates and measurements using simple instruments or procedures and all data could be collected in the field in less than 3 h. There was a significant positive correlation between rankings of the condition of 10 wetlands based on an independent long-term monitoring data set and the wetland condition index. There were also highly significant positive correlations between indicator scores collected by different investigators. Indicator scores for physical factors and fringing vegetation did not differ between autumn and winter, but winter rainfall had a significant impact on aquatic vegetation and water quality indicators. The results indicate that the wetland condition index is a valuable and reliable tool for the rapid surveying of the condition of permanent floodplain wetlands.  1998 Academic Press

Keywords: wetland condition index, ecosystem health, rapid assessment, indicators, validation.

Introduction Wetlands are amongst the most degraded of ecosystems and losses have been estimated at 50% of the original global wetland area (Gardiner, 1994; Jones et al., 1995). Significant resources are currently focused on monitoring the status of the wetlands remaining in many countries (McKenzie et al., 1992; Clough, 1993; Henry and Amoros, 1995). Most of the large Government or intergovernment wetland monitoring projects such as the Environmental Monitoring and Assessment Program (EMAP)—wetlands, in the United States (Peterson, 1994; Novitzki, 1995) rely on scientists or management agency staff working within an organized project framework to collect ecological data. These data are then used via Government channels to influence decisions regarding wetland conservation policy (e.g. Anon, 1996, 0301–4797/98/020143+17 $30.00/0

1997). However, for a variety of reasons (e.g. Schrader-Frechette and McCoy, 1994; Christensen et al., 1996; Robertson, 1997) research and highly structured scientific monitoring efforts do not necessarily result in wetland conservation (Robertson, 1992, 1997). There are two important reasons for the lack of transfer of scientific and monitoring data into wetland conservation action. First, natural resource managers often do not have the resources to spend on detailed and lengthy ecosystem monitoring. Secondly, a large proportion of wetlands occur on private property, and despite Government policy, landowners do not have the simple and reliable monitoring tools available to assess the influence of their land-management actions on wetlands. For these and other reasons rapid appraisal methods have been developed for assessing the condition of wetlands and other ecosystems relative to natural conditions (e.g. Morgan and Bray, 1982; Ainslie, 1994; Larson

† School of Environmental and Information Sciences, Charles Sturt University, PO Box 789, Albury, New South Wales, 2640 Australia ‡ The Johnstone Centre, School of Science and Technology, Charles Sturt University, PO Box 588, Wagga Wagga, New South Wales, 2678 Australia Received 21 August 1997; accepted 26 March 1998

 1998 Academic Press

144

C. Spencer et al.

and Mazzarese, 1994). Most rapid assessment methods are conceptually similar, being based on indicators of important wetland function (e.g. Mitsch, 1992). However, little effort has been devoted to evaluating rapid assessment methodologies (Cable et al., 1989; WWF, 1992). The spatial variability in wetland ecosystems, the paucity of information on the natural temporal variance for wetland variables and the scarcity of unimpacted reference sites (Turner et al., 1995) all contribute to concerns about whether rapid appraisal methods can provide scientifically valid measures of wetland condition (e.g. Kantrud and Newton, 1996). In south-eastern Australia, the floodplain wetlands of the Murray-Darling Basin have been severely degraded since European settlement by grazing and land clearing, and by significant river regulation and alterations to the frequency, timing and duration of hydrologic connections between river channels and floodplain habitats (Walker, 1985; Bacon et al., 1994). There has been significant progress made in Government policy on wetland conservation in the region (e.g. MDBMC, 1990; Sharley and Huggan, 1994; Anon, 1996, 1997). However, the wetlands occur over a vast area (106 km2), precluding detailed ecological monitoring and while many of the largest wetlands are now listed as being of international importance and are actively conserved (ANCA, 1996), more than 90% of all wetlands occur on private property where directives regarding wetland conservation may have little value. For these reasons we have developed a wetland condition index for permanent wetlands in the Murray-Darling Basin that is easy to use and is based on a rapid field survey approach. Rapid wetland survey methodologies have been used in Australia to identify wetlands of conservation value for particular taxa (e.g. Pressey, 1984) but have not been developed or tested as tools for assessing wetland condition. Here the choice of ecological indicators used in the wetland condition index is discussed and the performance of the index evaluated in assessing wetland conditions relative to longer-term scientific data collected for a number of wetlands. The authors also assess the replicability of index scores by different investigators and the index responses to the seasonality in wetland processes.

Methodology Definition of condition Extension of the concept of health to ecosystems has not been without controversy (Okey, 1996), and there is a continuing debate regarding all aspects of the concept (e.g. Suter, 1993; Wicklum and Davies, 1995; Rapport et al., 1996). In the context of this study, condition refers to ecosystem health. This incorporates the stability and sustainability of the system to withstand environmental stress (Rapport, 1995). It also includes the capacity of the ecosystem to support a diverse community of organisms and perform functions comparable to that of a local unimpaired site (Karr and Dudley, 1981).

Indicators Breckenridge et al. (1995) defined an indicator as any expression of the environment that estimates the condition of ecological resources, magnitude of stress, exposure of a biological component to stress, or the amount of change in a condition. To be useful in rapid assessment, indicators should possess most of the following attributes: (1) show low natural temporal and spatial variability; (2) be highly responsive to condition change; (3) not be ambiguous in their interpretation; (4) be cost effective and simple to apply; (5) have regional applicability; (6) be biologically relevant, that is be an indicator of ecosystem condition; (7) be a simple or commonly measured parameter—reference data on healthy thresholds is more likely to be available for simple indicators; (8) be non-destructive of the ecosystem; and (9) be able to have results summarized so as to be understood by nonexperts (Kent et al., 1992; Cairns et al., 1993; Breckenridge et al., 1995 Turner et al., 1995;). In order to choose appropriate indicators for floodplain wetlands in the Murray-Darling Basin, the authors focused on indicators of wetland function (Mitsch, 1992; Mitsch and Gosselink, 1993), including the ability of wetlands to act as habitats for terrestrial and aquatic fauna, as ‘filters’ of materials in overland flows and as sources of primary

Evaluating a wetland condition index Table 1. Justification of the indicators used in the present study and their relationship to floodplain wetland attributes and function in the south-east of the Murray-Darling Basin Attribute Soil

Function(s)

Indicators

References

Interception of overland flows; nutrient storage; supports growth of vegetation; habitat for fauna

Bank stability—unstable banks are detrimental to fringing vegetation and water quality Pugging by livestock—reflects damage to soil structure and chemistry Soil organic content—reflects organic inputs to wetlands

Leeper (1964), van de Graaff (1988)

Fringing vegetation

Interception of overland flows; nutrient storage; habitat for fauna; carbon source for aquatic foodchains

Width—reflects the ability of fringing vegetation to act as a faunal habitat and in material filtering Continuity—as above Height diversity—indicates habitat availability for fauna

Campbell (1993), Catterall (1993), Quinn et al. (1993) As above MacArthur and MacArthur (1961), Krebs (1978), Cairns et al. (1993), Debinski and Brussard (1994), Schultz et al. (1995), Read (1995)

Aquatic vegetation

Habitat for fauna; carbon Cover—indicative of water source for aquatic food- clarity and energy available for chains; damping windaquatic food chains driven mixing Spatial heterogeneity—indicates habitat availability for aquatic fauna Attached algae—important food source, indicative of trophic status of water bodies

Saunders (1980), Maitland (1990), Turak and Bichel (1994), Wilcox (1995), Henry and Amoros (1995) As above and Cairns et al. (1993), Peterson (1994), De Jalon et al. (1996) Costa et al. (1992), Rosas et al. (1993), Goldsborough and Robinson (1996)

Water

Habitat for biota; medium for biogeochemical processes

Turbidity—high turbidities can reduce light penetration for autotrophic community

Mackay et al. (1988), Maitland (1990)

Conductivity—increased salinity will affect natural biotic species composition Colour—indicates input of organic matter, can reduce oxygen concentrations Algal bloom frequency—indicates eutrophication

Hart et al. (1991)

production for floodplain river systems (Table 1).

The wetlands All wetlands considered in the study occurred on the floodplains of four rivers in southeastern Australia (Figure 1). In this region, westward flowing rivers fed by spring snow melt and local rainfall drain the Great Dividing Range. All wetlands on the Kiewa,

Meehan and Platts (1978), Braithwaite et al. (1984), Bacon et al. (1994) Leeper and Uren (1993), Jenkins and Briggs (1995)

Mackay et al. (1988), Detenbeck et al. (1996) Steinberg and Hartmann (1988)

Ovens, Murray and Murrumbidgee floodplains were meander cutoffs or depressions, both of which are filled by overbank flows. Soils on the floodplains are black earths and minimal prairie soils developed on alluvial deposits dominated by fine sand, silt and clay. The dominant vegetation on the floodplains of all four rivers is river red gum (Eucalyptus camaldulensis), which formed extensive forests prior to thinning for European agriculture. Most native understorey vegetation in the sections of the rivers surveyed has now been replaced by introduced grasses (Smith

145

146

C. Spencer et al. 145° E

New South Wales

South Australia

Wagga Wagga Murray River

Murrumbidgee River Albury

37° S

Victoria

Murrumbidgee River study area Kiewa River study area Ovens and Murray River study area Great Dividing Range Murray-Darling Basin

Figure 1. Location of the Murrumbidgee, Kiewa and Ovens and Murray floodplain wetland survey regions in the Murray-Darling Basin of south-eastern Australia.

and Smith, 1990) and the floodplains are now used to support beef cattle and sheep as well as some cropping (Crabb, 1997).

most degraded condition (Table 2). Below field and laboratory measurement procedures have been outlined along with the basis for score allocation.

Measurement and the scoring of indicator values Soil attributes The values obtained from field and laboratory measurements of each indicator were normalized by allocating scores based on: (1) published information, or (2) the authors’ own observations, on relatively undisturbed reference wetlands and a range of damaged wetlands in the south-east of Australia. Scores were allocated on a scale from 0 to 4, and in such a way that highest scores reflected best conditions and lowest scores the

Bank stability. Assessment of the degree of bank erosion was made following a walk around the complete wetland. Scores were allocated as follows (DCNR, 1995): 5= Stable—stable banks protected by good vegetation cover; 4=Good—some minor spot erosion occurring; 3=Moderate—some erosion occurring, spot erosion linked, structural and vegetative damage occurring due to little

Evaluating a wetland condition index Table 2. Score allocation for indicator values in the south-eastern Murray-Darling Basin Indicator Bank stability Degree of pugging (mean pugs per m2) Organic content of soil (mean % organic matter) Width of fringing vegetation strip (mean width m) Continuity of fringing vegetationa (%) Height diversity Cover of aquatic vegetationb (%) Attached algae Spatial heterogeneity (no of layers present) Turbidity (NTU) Conductivity (ls cm−1) Colour (hazen units) Frequency of algal blooms

4

3

Score 2

1

0

Stable 0

Good 1–6

Moderate 7–12

Poor 13–19

Unstable >20

20–30

6–19

<5 and >30

>30

8–30

3–8

0·5–3

<0·5

>90

66–90

36–65

10–35

<10

>25

17–24 26–75

1–8 >95

0 <5

1

Abundant 0

121–160 2500–5833 141–300 Every 2 yrs

>160 >5833 >300 Annually

Little >4

3

9–16 5–25 76–95 Medium 2

0–40 0–292 <15 Never

41–80 292–833 16–80 <5 Yearly

81–120 833–2500 81–140 2–5 Yearly

See text for discussion. a Sum the score for three layers (trees, rushes/sedges and grasses), sum and divide by three to obtain a score between zero and four). b Multiply by 4/3 to obtain a score out of four.

erosion; 2=Poor—significant erosion occurring, little vegetation present; 1=Unstable—extensive erosion occurring unchecked, bare banks, steep banks may be present due to erosion (Table 2). Degree of pugging. Pugging of soil by livestock was measured as the mean of the number of cattle hoof marks in 10 1-m2 quadrants placed randomly on the sediment at the water’s edge. Through pugging, livestock cause soil compaction, erosion, lowered water infiltration rates and a reduction in the water storage capacity of floodplain soils (Bacon et al., 1994). Observations on a variety of wetlands throughout the Murray-Darling Basin (authors’ unpub. data) indicated that severely degraded wetlands had more than 20 hoof marks per square metre (Table 2). Organic content of soil. Soil samples were collected from five random sites at the land– water interface around the perimeter of each wetland. The mean percentage of organic matter (loss on ignition at 500°C for 2 l) was then calculated for the five samples. Wetland soil organic matter content will naturally vary with soil type, degree of flooding and input of primary production. With the

exception of regions with peat accumulations, where organic contents of soil may be up to 75%, most floodplain soils will have organic matter contents greater than 5% (Mitsch and Gosselink, 1993). More frequently flooded soils will have higher levels of organic matter (Leeper and Uren, 1993). Based on this, and observations in relatively pristine floodplain regions in the region of the study scores were allocated for organic matter content as outlined in Table 2.

Fringing vegetation Width of fringing vegetation strip. The mean width of the native vegetation fringing was based on visual estimates of the riparian strip made at the four major compass points at each wetland. In the case of wetlands where sides differed in steepness, the maximum flood height in the wetland was used as a guide to determining the boundary between the wetland riparian strip and other floodplain flora. Recent reviews of riparian buffers for watercourses have indicated that there are no general minimum acceptable widths (e.g. Bren, 1993; Castelle et al., 1994; Spackman and Hughes, 1995). However, it appears

147

148

C. Spencer et al.

that buffer strips less than 5 m provide minimal protection to aquatic resources under most conditions (Castelle et al., 1994). Bren (1993) indicated that buffer strips greater than 20 m in width were most frequently recommended as providing the best protection for the physical, chemical and biological components of wetlands (also see Castelle et al., 1994). In Australia, the most common minimum buffer widths recommended are 20–30 m (Riding and Carter, 1992). Using this information as a guide, scores were allocated for fringing vegetation width as outlined in Table 2. Continuity of fringing vegetation. Continuity of the fringing vegetation around the wetland perimeter was estimated by eye for each of the main vegetation layers including trees, rushes/sedges and grasses. Following Pressey (1987), a score was given for the continuity of trees, emergents and grass around the wetland perimeter (Table 2). These scores were then combined to give an overall score on the continuity of fringing vegetation, i.e. continuity of fringing vegetation=(continuity of trees+continuity or rushes/sedges+ continuity of grasses) /3. Height diversity of fringing vegetation. A measure of height diversity was based on the sum of the scores for the continuity of each of seven recognized vegetation layers in the fringing vegetation strip. Three overstorey layers were recorded, following Gillison and Anderson (1981); viz, trees greater than 30 m, trees 10–30 m, trees less than 10 m. There were four possible understorey layers (Smith and Smith, 1990; Sainty and Jacobs, 1981); viz reeds, rushes/ sedges, reeds, grasses and herbs (lichens, mosses, ground covers). The scores out of four for the continuity of each of the seven vegetation layers were then added to obtain a total score for height diversity. The maximum possible score for height diversity was therefore 7×4=28. All height diversity measures obtained from summing the scores for individual layers of vegetation were then normalized to scores out of four as shown in Table 2. Aquatic vegetation Cover of aquatic vegetation. For each wetland, the percentage of the water surface

that was covered with aquatic vegetation was estimated by eye. Aquatic plants included emergents, submerged and floating plants (see below). Score for the aquatic vegetation cover (Table 2) were adapted from Pressey (1987). A wetland totally covered by aquatic vegetation, and thus without open water, may result from nutrient enrichment or by overgrowth of exotic taxa (Sainty and Jacobs, 1981; Henry and Amoros, 1995). Such wetlands are considered to be in poor condition (Mitchell, 1990) and were thus allocated a low score (Table 2). An estimate of cover between 25 and 75% was allocated the highest score under this scheme. Attached algae. Most surfaces within wetlands support a film of microalgae, bacteria, other micro-organisms and organic matter (= biofilm). While biofilms and their components can be used as useful indicators in fresh waters (Burns et al., 1994), they are not appropriate for a rapid appraisal index. The presence and relative abundance of macroalgae was used to indicate the trophic status of wetlands (Goldsborough and Robinson, 1996). Following a circuit of the wetland the amount of macroalgae was recorded according to the following categories: little= no obvious macroalgae present (score of 1, Table 2); medium=clumps of significant macroalgae present (score of 2, Table 2); and abundant=macroalgae present over at least one-third of water area (score of 3, Table 2). Spatial heterogeneity of aquatic vegetation. The number of layers of aquatic vegetation occurring in each wetland were noted following Williams (1983) who identified the following five layers of aquatic vegetation: free-floating at surface; freefloating beneath surface; emergent; in substrate with floating leaves; and submerged (anchored in substrate). Scores were based on the number of layers observed (Table 2).

Water quality Turbidity. Mean turbidity was calculated based on readings taken at four positions around the perimeter of the wetlands. Turbidity was measured in natural turbidity units (NTU) using a turbidity tube developed by Water Ecoscience of Australia. The highest

Evaluating a wetland condition index

score for turbidity (Table 2) was based on data collected from five billabongs along the Murray River from 1977–1982, considered to be in excellent condition and which had turbidities of less than 40 NTU (Hillman, 1986). Lowest scores were allocated based on data from oxbow lakes subject to severe disturbance by stock and introduced carp (Cyprinus carpio) (King et al., 1997). Conductivity. Mean conductivity was based on measurements taken with a Hanna pocket conductivity meter at four positions in each wetland. The conductivity ratings are based on the classes used by the Victorian Soil Conservation Authority (1980) following their conversion from mgl−1 to electrical conductivity using the relationship, TDS ppm= 0·6×EC in microhms cm−1 (GHD, 1992). The scores allocated to conductivity measurements were based on data for pristine wetlands (Hillman, 1986) and known ranges of salinity for wetlands affected by humanderived salinity (Crabb, 1997). Colour. Water samples were collected from the upper half metre of water, approximately 2 m from the edge at four locations in each wetland. The true colour of each sample was then determined by placing the filtered sample in a Loveybond Colorimeter visual comparator, using distilled water as a reference. The mean of the four colour measures was determined. Colour scores (Table 2) were based on Hillman’s (1986) observations on relatively pristine wetlands in the region and on recommendations by NHMRC (1994). Frequency of algal blooms. Although algal blooms occurred naturally in the river systems of the Murray-Darling Basin in the early phase of European settlement (Codd et al., 1994) they now occur more frequently throughout the Basin. A variety of factors have resulted in increased nutrient loadings in floodplain wetlands in the region (Cullen et al., 1993). The frequency of algal bloom occurrence (green slicks covering at least 50% of the water surface) was therefore considered as a good indicator of anthropogenic disturbance in these wetland types. The frequency of algal blooms was determined through questioning of the owner of the

wetland and scores allocated as shown in Table 2.

Calculating index scores After completion of field and laboratory measurements of indicators and allocation of scores, sub-index values for soil, fringing vegetation, aquatic vegetation and water attributes were calculated as shown in Table 3. Note that each sub-index value was normalized to produce a score out of 10. In the case of soil, fringing vegetation and aquatic vegetation for which there were three indicators, the maximum sum of indicator scores was 12. This was normalized to an overall score out of 10 by dividing by a factor of 1·2. In the case of water attributes where there were four indicators, the maximum total score for the sub-index was 16 and this was normalized to a score out of 10 by dividing by a factor of 1·6. For each of the four subindices, the range of scores was 0=extremely poor condition, to 10=excellent condition. The total index score was the average of the four sub-index scores (Table 3).

Evaluating the condition index Agreement with long-term monitoring data A major issue with rapid-appraisal indices based on indicators is how well they might reflect differences relative to ‘natural’ habitats and conclusions based on long-term, highly structured monitoring data (Hart, 1988; WWF, 1992). The authors chose to test the agreement between the wetland condition index and monitoring data available from an independent, 18 month long study of the effects of farming and river regulation on the ecology of over 30 wetlands in the Ovens and Murray River floodplains (Figure 1) conducted by Ogden (1996). Ten of the 30 wetlands studied by Ogden (1996) were ranked by Ogden (pers. comm.) from best to worst condition based on water quality, vegetation and zooplankton data. The rankings were based on total phosphorus, total nitrogen, turbidity and the species diversity of cladoceran skeletal remains in surface sediments. These chemical parameters

149

150

C. Spencer et al. Table 3. Using the index to determine wetland condition; an example from a wetland on the Kiewa Valley floodplain (from Spencer, 1996) Attribute and indicator

Measure

Score

Sub-index score

Soil Bank stability Degree of pugging Percent organic matter Soil sub-index

Stable 4·9 per sq metre 15·9

4 3 2 9/1·2

7·5

Fringing vegetation Mean width of fringing vegetation 3·75 m Continuity of fringing Trees—50% vegetation Rushes—90% Grasses—100% Height diversity of fringing vegetation

2 2 3 4 (2+3+4)/3=3

Trees >30 m 50%—2 Trees <10 m 10%—1 Rushes/sedges 90%—3 Grass 100%—4 Herbs 100%—4 Total—14

Fringing vegetation sub-index Aquatic vegetation Percent cover of aquatic vegetation Attached algae Spatial heterogeneity of aquatic vegetation

20%

Water Turbidity Conductivity Colour Algal bloom frequency Water quality sub-index

36 NTU 75 ls cm−1 62·5 hazen units Never

Total Index

are commonly employed in long-term water quality monitoring programmes (ANZECC, 1992), and the use of biota to assess the condition of aquatic systems has received widespread acceptance (e.g. Norris et al., 1995). The skeletal remains of cladocera were used to assess differences in the diversity of zooplankton because the use of recent skeletal remains damps the short-term variation observed in net zooplankton data. Cladocera are valuable monitoring taxa because they are responsive to changes in water quality (Anderson-Carnahan et al., 1995) and also reflect alteration in submerged aquatic vegetation because chydorid cladocerans are

5·8

2·7

Little Layers present: Free floating at surface Emergent Submerged

Aquatic vegetation sub-index

2 7/1·2

4 3

9·7/1·2

8·0

4 4 3 4 15/1·6

9·3

(7·5+5·8+8·0+9·3)/4

7·7

littoral zone inhabitants and are phytal associates (Frey, 1986). Total phosphorus, total nitrogen and turbidity measures were all standardized to 100 cm to remove the effect of wetland depth. The turbidity and cladoceran diversity data was given double the weighting of the nutrient data because nutrient data are naturally highly variable at many time scales. Billabongs which were ranked equal based on the above data, were separated using the data on aquatic macrophyte cover for each wetland. Wetlands without macrophytes were ranked lower than those with macrophytes (R. Ogden, pers. comm.). Without prior knowledge of the rankings based on the long-term data, the 10 wetlands were

Evaluating a wetland condition index 12 10

reserves and stock routes, so it was not possible to obtain information on the frequency of algal blooms from land owners. Information on colour was not available owing to a mix-up with samples.

rs = 0.709 (P < 0.05)

Index

8 6 4 2

Results 0

2

4 6 8 Long-term data

10

12

Figure 2. Relationship between the ranks of the condition of 10 wetlands on the Ovens and Murray floodplains established using long-term monitoring data and the index of wetland condition. The line represents the case of absolute agreement. r is the Spearman rank correlation coefficient.

surveyed using the wetland condition index over a 3 day period in 1996.

Influence of the season on index values Ten wetlands on the floodplain of the Kiewa Valley (Figure 2) were surveyed in autumn (April to May). Although the Kiewa Valley is a relatively wet area (710 mm annual average rainfall), the autumn surveys were done in the peak of the dry period in 1996. During the wettest part of the year (August to October) the floodplain is frequently flooded. To determine whether index scores differed significantly in the two season extremes, the same 10 wetlands were re-surveyed in the period 12–14 August 1996.

Influence of investigator on index values Rapid appraisal indices need to be robust to be used by different investigators (Kent et al., 1992). To determine the effect of the investigator on the index, a set of 11 wetlands on the Murrumbidgee River near Wagga Wagga (Figure 3) were surveyed by two independent groups. Both groups worked independently, and obtained index scores without knowledge of the rankings obtained by each other. All surveys were completed between 26 June and 23 July 1996. The index applied to the Murrumbidgee wetlands did not include the frequency of algal bloom or colour indicators. Many of the Murrumbidgee wetlands were located on Government-owned

The index and long-term monitoring data The rapid index results compared very favourably with the rankings of the 10 wetlands based on long-term monitoring data (Figure 2), suggesting the results obtained by the wetland condition index provide a valid estimation of the health of wetlands. Many of the wetlands used in this part of the study were located in reserves and were thus relatively unaffected by floodplain development. For this reason the range of wetland condition index scores of the wetlands was small (6·3–8·3).

The influence of season on index scores Soil Scores for the soil attribute sub-index (consisting of the indicators: bank stability, degree of pugging and organic content of soil) were very similar in the two seasons [Figure 3(a)], with slightly higher scores recorded in winter. In the Kiewa Valley, the floodplain is subjected to severe flooding during winter and grazing herds do not always use the river flats. The cattle hoof marks that exist around wetlands in autumn are covered by water in winter, as are sections of wetland banks that were visible in autumn. The result is a lower average number of pug marks per square metre and a subsequent higher score for this indicator. Seven of the 10 wetlands surveyed in winter had fewer pugs per square metre than in autumn.

Fringing vegetation The fringing vegetation sub-index scores were very similar across the two seasons [Figure

151

C. Spencer et al. 12

Soil sub-index

12

rs = 0.759 (P < 0.05)

10

8

Autumn

Autumn

10

2

6 3

4 2

12 10

2

4

6 Winter

8

10

8 6 4

0

12

Aquatic vegtation sub-index

12

rs = 0.463 ns

8

10 2

6 4 2 0

Fringing vegetation sub-index rs = 0.938 (P < 0.001)

2

Autumn

0

Autumn

2

4

6 8 Winter

10

12

Water quality sub-index rs = 0.156 ns

2 2

8 6 4 2

2

4

6 Winter 12 10 Autumn

152

8

10

0

2

4

6 8 Winter

10

12

12

6 8 Winter

10

12

Total index rs = 0.675 (P < 0.05)

8 6 4 2 0

2

4

Figure 3. Relationship between condition sub-index scores for wetlands on the Kiewa floodplain assessed in autumn and winter. The lines represent the case of absolute agreement. r values are Spearman correlation coefficients.

3(b)]. Slightly higher scores were obtained in the autumn survey because increased water levels in water cover portions of the fringing vegetation, resulting in lower scores for the width of fringing vegetation.

Aquatic vegetation There was little agreement between scores for the aquatic vegetation sub-index which were much lower in the winter survey [Figure 3(c)]. Local rainfall and winter floods increased water levels in the wetlands, removing floating vegetation. Of the 10 wetlands surveyed, seven showed a decreased score for vegetation cover and also a decreased score for spatial heterogeneity.

Water quality There was no correlation between water quality sub-index scores obtained from wetlands in the two seasons [Figure 3(d)]. However, the sub-index scores all fell within a very narrow range and were mostly very high, indicating flood water quality regardless of time of year.

Total index Despite seasonal differences in the soil (high scores in winter) and aquatic vegetation (lower scores in winter) sub-indices, the total index scores from the two seasons were very similar [Figure 3(d)].

Evaluating a wetland condition index

10

rs = 0.855 (P < 0.01)

8

Investigator 2

2

6 4 2 0

10 Investigator 2

Soil sub-index

2

4 6 Investigator 1

Aquatic vegetaion sub-index

6 2

4 2 0

2

4 6 Investigator 1

Investigator 2

10

8

rs = 0.842 (P < 0.01)

8 6 4 2

10

rs = 0.682 (P < 0.05)

8

Fringing vegetation sub-index

0

10

8

Investigator 2

Investigator 2

10

2

8

10

8

10

Water quality sub-index

8

rs = 0.616 (P < 0.05)

6 3

4 2 010

10

4 6 Investigator 1

2

4 6 Investigator 1

Total index rs = 0.918 (P < 0.001)

8 6 4 2 0

2

4 6 Investigator 1

8

10

Figure 4. Relationship between condition sub-index scores obtained for wetlands on the Murrumbidgee floodplain. The lines represent the case of absolute agreement. r values are Spearman rank correlation coefficients.

Robustness of the index to the influence of investigators Soil There were minor differences in soil subindex scores obtained by different investigators [Figure 4(a)]. When broken down into indicator scores, the main differences resulted from assessments of bank stability. This is the most subjective of the three indicators used for soil. Variations in bank stability scores occurred for five of the 11 wetlands, though this was only one scoring bracket difference in all cases.

Fringing vegetation Despite the high correlation between scores obtained by different investigators [Figure 4(b)], scores from investigator one were often one unit higher than those from investigator two. All three of the indicators used to assess fringing vegetation are based on visual assessments and are subject to some investigator error. Inspection of the raw data indicated that the width of fringing vegetation was the indicator that caused most variation in sub-index scores given to the wetlands by the two independent groups.

153

154

C. Spencer et al.

Aquatic vegetation There was little evidence of consistent differences between the two investigators in obtaining aquatic sub-index scores [Figure 4(c)]. However, the slightly lower correlation coefficient indicates some variation in the scoring of indicators. The scores given for the spatial heterogeneity indicator were consistent in all but one case. Scores given for cover of aquatic vegetation varied in four of the 11 wetlands. Several wetlands had a significant cover of the floating plant Azolla pinnata and variation in the spread of Azolla with wind may have affected results, as the surveys by the two groups were not performed on the same day. Scores for attached algae differed by one unit in seven of the 11 wetlands surveyed by the two investigators.

Water quality The water quality sub-index scores obtained by the two groups showed a strong correlation [Figure 4(d)]. This was anticipated as the sub-index scores were based solely on direct measurements of conductivity and turbidity for these surveys. However, two of the wetlands showed differences in turbidity ratings, and this was likely due to rainfall that occurred in the 3 weeks between visits by the two investigators.

Total index There was a very high correlation in total index scores obtained by the two investigators [Figure 4(e)]. However, in seven of the 11 wetlands, investigator one obtained scores that were slightly higher than those obtained by investigator two.

Discussion The monitoring of aquatic systems has evolved in three directions. Ambitious programmes based on formal ecological principles have been developed by national and international agencies to assess the condition of aquatic systems over large temporal and spatial scales (e.g. Novitski, 1995; Resh et al.,

1995). Such programmes depend on rigorous and time-consuming data collection and analyses by scientists and target professional resource managers and policy makers (Hart, 1988). More inexpensive and rapid assessment tools, which integrate condition over various temporal and spatial scales, have been developed to enable smaller teams of scientists to perform assessments of wetlands and waterways (e.g. Karr, 1981; Kent et al., 1992; DCNR, 1995). For instance, the index of biological integrity (Karr, 1981), which uses aspects of fish community structure to assess aquatic ecosystem health, has been tested and used successfully in the United States (e.g. Steedman, 1988) and modified for use elsewhere (e.g. Harris, 1995). Such approaches also target professional resource managers and policy makers. In many countries a variety of simple and rapid monitoring tools have also been developed for use by management agencies to obtain information on wetland conservation status (e.g. Pressey, 1984; Cable et al., 1989). In addition, simple tools are available to the general human community to collect information on the health of aquatic systems. These programmes often target communitybased conservation efforts (Courtemanch, 1988). In the case of rapid assessment tools it is essential that methods are validated and that results from these methods reflect real changes in ecological condition (Hart, 1988; Courtemanch, 1988; Resh et al., 1995). Recent reviews of rapid appraisal methods for wetlands indicate that the validity of these methods is rarely assessed (Cable et al., 1989; WWF, 1992). The results of the present study have shown that the wetland condition index is a rapid and easily used assessment tool which produced results on wetland condition that were similar to those based on longer-term ecological data. It was not surprising that there was not 100% agreement between the two sets of condition ratings for the wetlands on the Ovens and Murray floodplains. Ogden (1996) completed his monitoring of the wetlands 1 year before they were assessed using the wetland condition index. Factors such as water quality, fringing vegetation biomass and the degree of soil disturbance can change

Evaluating a wetland condition index

rapidly in the region in response to alterations in grazing pressure by domestic stock and introduced fish (e.g. Bacon et al., 1994; King et al., unpubl.). The wetlands used for the comparisons of long-term data with the condition index were all relatively protected from severe degradation by farming practices (R. Ogden, pers. comm.). The total condition scores for these wetlands fell within a narrow and relatively high range (6·3–8·3). A closer relationship between the rankings based on rapid index scores and long-term data would have been expected over a greater range of wetland conditions. Parts of the wetland condition index are sensitive to seasonal changes in environmental factors. In the Kiewa Valley, the major environmental variable which changed between autumn and winter was flooding, which increased water levels in the floodplain wetlands and reduced the access of stock to floodplain habitats. These changes were reflected in the higher winter scores for the soil sub-index resulting from a reduction in soil pugging by cattle, and the lower winter score for the aquatic vegetation sub-index resulting from the removal of floating vegetation by floodwater. While such sensitivity is a prerequisite for indicators of ecosystem health (e.g. Kent et al., 1992; Breckenridge et al., 1995), it means that natural seasonal changes will need to be accounted for when using the wetland condition index to assess changes in condition following alterations to land or water management practices. An important consideration in the use of any wetland condition index is access to reference wetlands that are in pristine or relatively pristine condition (Turner et al., 1995; Brinson and Reinhardt, 1996). Reference wetlands will be required in each major bioregion where the wetland condition index is used to ensure that the scoring of indicator data reflects regional variation. For instance, indicators or ‘metric’ scoring has been adjusted against regional reference data for other assessment tools, such as the index of biological integrity, so that it provides a relevant measure of stream health (e.g. Steedman, 1988). In addition, if the wetland condition index is to be used for long-term monitoring, rather than as a tool to assess the relative condition of a number of wetlands at any one time, then the index will need to be used in a

study design which employs several reference wetlands as control sites. In this way natural variation in condition can be separated from that arising from land and water management practices (e.g. Turner et al., 1995). The wetland condition index appears to be relatively robust to use by different investigators, and thus should be useful for comparisons between sites assessed by different groups. However, there are potential problems with the allocation of scores for some of the more qualitative indicators (e.g. attached algae), which were responsible for most of the variation in scoring between groups. This implies that some joint training may be necessary to acquaint investigators with the variations in indicator measurements before the condition index is used by a number of different people. A rapid, easily used, and scientifically valid wetland condition index should have wide application in the south-eastern Murray-Darling Basin of Australia. More than 90% of wetlands in the Basin are on private property and the index will allow property managers and management agency staff to assess the condition of wetlands in a cost-effective manner. The exact version of the index described here may only be applicable in the wetlands used in this study, i.e. floodplain meander cutoffs. Extension to other regions of the Basin, and other floodplain wetlands, may be possible if reference wetlands are available to adjust indicator scores to reflect local natural conditions.

Acknowledgements The authors would like to acknowledge Dr Terry Hillman, Dr Sue Briggs, Dr Ralph Ogden, Mr Greg Raisan and Ms Sue Charles-Jones for their involvement during the development of the condition index and Mr Michael Healey for help in the field. The project was supported by the NSW EPA and Charles Sturt University.

References Ainslie, W. B. (1994). Rapid wetland functional assessment: its role and utility in the regulatory arena. Water Air and Soil Pollution 77, 433–444. ANCA. (1996). A Director of Important Wetlands in Australia. Canberra: Biodiversity Group, Australian Nature Conservation Agency.

155

156

C. Spencer et al. Anderson, S. H. (1991). Managing our Wildlife Resources. 2nd ed. New Jersey: Prentice Hall. Anderson-Carnahan, L., Foster, S., Thomas, M., Korth, W. and Bowmer, K. H. (1995). Selection of suitable cladoceran species for toxicity testing in turbid waters. Australian Journal of Ecology 20, 28–33. Anon. (1996). The NSW Wetlands Management Policy, 32pp. NSW Department of Land and Water Conservation. Anon. (1997). Wetlands Policy of the Commonwealth Government of Australia, 48pp. Environment Australia. ANZECC. (1992). National water quality management strategy. Australian Water Quality Guidelines for Fresh and Marine Waters. Australian and New Zealand Environment and Conservation Council. Bacon, P., Ward, K., Craven, P., Harper, M. and Bone, B. (1994). Floodplain land-use issues in the Murray-Darling Basin. In Murray-Darling Basin Floodplain Wetlands Management (T. Sharley and C. Huggan, eds). Canberra: Murray-Darling Basin Commission. Braithwaite, R. W., Dudzinski, M. L., Ridpath, M. G. and Parker, B. S. (1984). The impact of water buffalo on the monsoon forest ecosystem in Kakadu National Park. Australian Journal of Ecology 9, 303–322. Breckenridge, R. P., Kepner, W. G. and Mouat, D. A. (1995). A process for selecting indicators for monitoring conditions of rangeland health. Environmental Monitoring and Assessment 36, 45–60. Bren, L. J. (1993). Riparian zone, stream, and floodplain issues: a review. Journal of Hydrology 150, 277–299. Brinson, M. M. and Rheinhardt, R. (1996). The role of reference wetlands in functional assessment and mitigation. Ecological Applications 6(1), 69–76. Burns, A., Sheldon, F. and Walker, K. F. (1994). Managing rivers from the bottom up; the importance of littoral biofilms. In Environmental Flows Seminar Proceedings. Australian Water and Wastewater Association Incorporated, New South Wales. Cable, T. T., Brack, V. and Holmes, V. R. (1989). Simplified method for wetland habitat assessment. Environmental Management 13(2), 207– 213. Cairns, J., McCormick, P. V. and Niederlehner, B. R. (1993). A proposed framework for developing indicators of ecosystem health. Hydrobiologia 263, 1–44. Campbell, I. C. (1993). Riparian stream linkages: An Australian perspective on in-stream issues. In Ecology and Management of Riparian Zones in Australia (S. E. Bunn, B. J. Pusey and P. Price, eds). LWRRDC Occasional Paper Series No: 05/93. Canberra: Land and Water Resources Research and Development Corporation and Centre for Catchment and In-stream Research. Castelle, A. J., Johnson, A. W. and Conolly, C. (1994). Wetland and stream buffer size

requirements – a review. Journal of Environmental Quality 23, 878–882. Catterall, C. P. (1993). The importance of riparian zones to terrestrial wildlife. In Ecology and Management of Riparian Zones in Australia (S. E. Bunn, B. J. Pusey and P. Price, eds). LWRRDC Occasional Paper Series No: 05/93. Canberra: Land and Water Resources Research and Development Corporation and Centre for Catchment and In-stream Research. Christensen, N. L., Bartuska, A. M., Brown, J. H., Carpenter, S., D’Antonio, C., Francis, R., Franklin, J. F., McMahon, J. A., Noss, R. F., Partson, D. J., Peterson, C. H., Turner, M. C. and Woodmansee, R. G. (1996). The report of the Ecological Society of America committee on the scientific basis for ecosystem management. Ecological Applications 6, 665–691. Clough, B. F. (1993). The economic and environmental values of mangrove forests and their present state of conservation in the southeast Asia/Pacific region. International Society for Mangrove Ecosystems, 199pp. International Tropical Timber Organisation and Japan International Association for Mangroves. Codd, G. A., Steffensen, D. A., Burch, M. D. and Baker, P. D. (1994). Toxic blooms of cyanobacteria in Lake Alexandrina, South Australia – learning from history. Australian Journal of Marine and Freshwater Research 45, 731–736. Costa, J. E., Howes, B. L., Giblin, A. E. and Valiela, I. (1992). Monitoring nitrogen and indicators of nitrogen loading to support management action in Buzzards Bay. In Ecological Indicators Volume 1 (D. H. McKenzie, D. E. Hyatt and V. J. McDonald, eds). New York: Elsevier Applied Science. Courtemanch, D. L. (1988). Briding the old and new science of biological monitoring. Journal of the North American Benthological Society 13, 117–121. Crabb, P. (1997). Murray-Darling Basin Resources. Canberra: Murray-Darling Basin Commission. Cullen, P., Croome, R., Harris, G., McComb, A., Smalls, I., Steffensen, D. and Tyler, P. (1993). Algal ecology and triggers of algal blooms. In Algal Management Strategy. Technical Advisory Group Report, pp. 1–6. Canberra: Murray-Darling Basin Commission. DCNR. (1995). Development of an Index of Stream Condition. Melbourne: Department of Conservation and Natural Resources Waterways Unit. Debinski, D. M. and Brussard, P. F. (1994). Using biodiversity data to assess species-habitat relationships in Glacier National Park, Montana. Ecological Applications 4(4), 833–843. DeJalon, D. G., Mayo, M. and Molles, M. C. (1996). Characterisation of Spanish pyreneam stream habitat: Relationships between fish communities and their habitat. Regulated Rivers: Research and Management 12, 305–316. Detenbeck, N. E., Taylor, D. L., Lima, A. and Hagley, C. (1996). Temporal and spatial variability in water quality of wetlands in the

Evaluating a wetland condition index Minneapolis/St Paul, MN metropolitan area: Implications for monitoring strategies and designs. Environmental Monitoring and Assessment 40, 11–40. Frey, D. G. (1986). Cladocera analysis. In Handbook of Holocene Palaeoecology and Palaeohydrology (B. E. Berglund, ed.), pp. 667–692. New York: John Wiley and Sons. Gardiner, J. (1994). Pressures on wetlands. In Wetland Management (R. A. Falconer and P. Goodwin, eds). London: Thomas Telford Services. GDH (Gutteridge, Haskins and Davey). (1992). An Investigation of Nutrient Pollution in the Murray-Darling River System, Report Summary. Canberra: MDBC. Gillison, A. N. and Anderson, D. J. (1981). Vegetation Classification in Australia. Canberra: CSIRO and Australian National University Press. Goldsborough, L. G. and Robinson, G. C. C. (1996). Patterns in wetlands. In Algal Ecology (R. J. Stevenson, M. L. Bothwell and R. L. Lowe, eds), pp. 78–117. San Diego: Academic Press. Harris, J. H. (1995). The use of fish in ecological assessments. Australian Journal of Ecology 20, 65–80. Hart, B. T., Bailey, P., Edwards, R., Hortle, K., James, K., McMahon, A., Meredith, C. and Swadling, K. (1991). A review of the salt sensitivity of the Australian freshwater biota. Hydrobiologia 210, 105–144. Hart, D. D. (1988). Building a stronger partnership between ecological research and biological monitoring. Journal of the North American Benthological Society 13, 110–116. Henry, C. P. and Amoros, C. (1995). Restoration ecology and riverine wetlands: a scientific basis. Environmental Management 19(6), 891–902. Hillman, T. J. (1986). Billabongs. In Limnology in Australia (P. De Deckker and W. D. Williams, eds). The Netherlands: CSIRO and Dr W. Junk Publishers. Jenkins, K. M. and Briggs, S. V. (1995). Ecological Management of Lakebed Cropping on the Lakes of the Great Anabranch of the Darling River. National Parks and Wildlife Service, NSW. Jones, D., Cocklin, C. and Cutting, M. (1995). Institutional and landowner perspectives on wetland management in New Zealand. Journal of Environmental Management 45, 143–161. Kantrud, H. A. and Newton, W. E. (1996). A test of vegetation-related indicators of wetland quality in the prairie pothole region. Journal of Aquatic Ecosystem Health 5, 177–191. Karr, J. R. (1981). Assessment of biotic integrity using fish communities. Fisheries (Bethesda) 6, 21–27. Karr, J. R. and Dudley, D. R. (1981). Ecological perspectives on water quality goals. Environmental Management 5, 55–68. Kent, D. M., Reimold, R. J., Kelly, J. M. and Tammi, C. E. (1992). Coupling wetlands structure and function: Developing a condition index for wetlands monitoring. In Ecological Indicators, Volume 1 (D. H. McKenzie, D. E. Hyatt

and V. J. McDonald, eds). New York: Elsevier Applied Science. King, A. J., Robertson, A. I. and Healey, M. R. (1997). Experimental manipulations of the biomass of carp (Cyprinus carpio) in billabongs. 1. Impacts on watercolumn properties. Marine and Freshwater Research 48, 435–443. Krebs, C. J. (1978). Ecology. The Experimental Analysis of Distribution and Abundance, 2nd ed. New York: Harper and Row. Larson, J. S. and Mazzarese, D. B. (1994). Rapid assessment of wetlands: History and application to management. In Global Wetlands: Old World and New (W. J. Mitsch, ed.). The Netherlands: Elsevier Science. Leeper, G. W. (1964). Introduction to Soil Science, 253pp. Parkville: Melbourne University Press. Leeper, G. W. and Uren, N. C. (1993). Soil Science: An Introduction. Melbourne: Melbourne University Press. MacArthur, R. H. and MacArthur, J. W. (1961). On bird species diversity. Ecology 42(3), 594–598. Mackay, N., Hillman, T. and Rolls, J. (1988). Water Quality of the River Murray. Review of Monitoring 1978–1986. Canberra: Murray-Darling Basin Commission. Maitland, P. S. (1990). Biology of Freshwaters, 2nd ed. London: Blackie and Son. McKenzie, D. H., Hyatt, D. E. and McDonald, V. J. (1992). Ecological Indicators. New York: Elsevier Applied Science. Meehan, W. R. and Platts, W. S. (1978). Livestock grazing and the aquatic environment. Journal of Soil and Water Conservation. 33, 274–278. Mitchell, P. (1990). The Environmental Condition of Victorian Streams. Victoria: Department of Water Resources. Mitsch, W. J. (1992). Ecological indicators for ecological engineering. In Ecological Indicators (D. H. McKenzie, E. C. Hyatt and McDonald, V. J. eds). New York: Elsevier Applied Science. Mitsch, W. J. and Gosselink, J. G. (1993). Wetlands. USA: Van Nostrand Rheinhold. Morgan, N. C. and Bray, B. (1982). An ecological survey of standing waters in North West Africa: Rapid survey and classification. Biological Conservation 24(1), 5–44. Murray-Darling Basin Ministerial Council. (1990). Natural Resource Management Strategy – Towards a Sustainable Future, 24pp. Canberra: Murray-Darling Basin Commission. NHMRC (National Health and Medical Research Council and Agricultural and Resource Management Council of Australia and New Zealand). (1994). Australian Drinking Water Guidelines. Canberra: AGPS. Norris, R. H., Hart, B. T., Finlayson, M. and Norris, K. R. (eds). (1995). Use of biota to assess water quality. Australian Journal of Ecology 20(1), 1–227. Novitzki, R. P. (1995). EMAP-Wetlands: A sampling design with global applications. Vegetatio 118(1–2), 171–184. Ogden, R. (1996). The Effects of Farming and River Regulation on Billabongs of the Eastern Murray

157

158

C. Spencer et al. River and Lower Ovens River. Unpublished PhD thesis, Australian National University. Okey, B. W. (1996). Systems approaches and properties, and agroecosystem health. Journal of Environmental Management 48, 187–199. Peterson, S. A. (1994). The environmental monitoring and assessment program (EMAP): Its objectives, approach and status relative to wetlands. In Monitoring of Ecological Change in Wetlands of Middle Europe (G. Aubrecht, G. Dick and C. Prentice, eds). United Kingdom: Proceedings International Workshop, Linz Austria. Pressey, R. L. (1984). A method for description and evaluation of coastal floodplain wetlands. In Survey Methods for Nature Conservation (K. Myers, C. R. Margules and I. Musto, eds). Canberra: CSIRO. Pressey, R. L. (1987). A Survey of Wetlands of the Lower Macleay Floodplain, New South Wales. NSW: National Parks and Wildlife Service. Quinn, J. M., Cooper, A. B. and Williamson, R. B. (1993). Riparian zones as buffer strips: A New Zealand perspective. In Ecology and Management of Riparian Zones in Australia. LWRRDC Occasional Paper Series No 05/93 (S. E. Bunn, B. J. Pusey and P. Price, eds). Canberra: Land and Water Resources Research and Development Corporation and Centre for Catchment and In-stream Research. Rapport, D. J. (1995). Ecosystem services and management options as blanket indicators of ecosystem health. Journal of Aquatic Ecosystem Health 4, 97–105. Rapport, D. J., Gaudet, C. L. and Calow, P. (1996). Evaluating and Monitoring the Health of LargeScale Ecosystems. Berlin: Springer-Verlag. Read, J. L. (1995). Subhabitat variability: A key to the high reptile diversity in chenopod shrublands. Australian Journal of Ecology 20(4), 494– 501. Resh, V. H., Norris, R. H. and Barbour, M. T. (1995). Design and implementation of rapid assessment approaches for water resource monitoring using benthic macroinvertebrates. Australian Journal of Ecology 20, 108–121. Riding, T. and Carter, R. (1992). The Importance of the Riparian Zone in Water Resource Management. A Literature Review. New South Wales: Department of Water Resources. Robertson, A. I. (1992). Concluding remarks: Research and mangrove conservation. In Tropical Mangrove Ecosystems (A. I. Robertson and D. M. Alongi, eds). pp. 327–329. Washington: American Geophysical Union Press. Robertson, A. I. (1997). Estuaries. In Saving our National Heritage? The Role of Science in Managing Australian Ecosystems (C. Copeland and D. Lewis, eds), pp. 127–141. Sydney: Halstead Press. Rosas, I., Velasco, A., Belmont, R., Baez, A. and Martinez, A. (1993). The algal community as an indicator of the trophic status of Lake Patzcuaro, Mexico. Environmental Pollution 80, 255–264. Sainty, G. R. and Jacobs, S. W. L. (1981). Waterplant of New South Wales. New South Wales:

Water Resources Commission New South Wales. Saunders, G. W. (1980). Organic matter and decomposers. In The Functioning of Freshwater Ecosystems (E. D. Le Cren and R. H. LoweMcConnell, eds). Cambridge: Cambridge University Press. Schrader-Frechette, K. S. and McCoy, E. D. (1994). What ecology can do for environmental management. Journal of Environmental Management 41, 293–308. Schultz, R. C., Collettii, J. P., Isenhart, T. M., Simpkins, W. W., Mize, C. W. and Thompson, M. L. (1995). Design and placement of a multispecies riparian buffer strip system. Agroforestry Systems 29, 201–226. Sharley, T. and Huggan, C. (eds). (1994). MurrayDarling Basin Floodplain Wetland Management, 112pp. Canberra: Murray-Darling Basin Commission. Smith, P. and Smith, J. (1990). Floodplain vegetation. In The Murray (N. Mackay and D. Eastburn, eds). Canberra: Murray-Darling Basin Commission. Spackman, S. C. and Hughes, J. W. (1995). Assessment of minimum stream corridor width for biological conservation: Species richness and distribution along mid-order streams in Vermont, USA. Biological Conservation 71, 325– 332. Spencer, C. (1996). Rapid Appraisal Index of Wetland Condition: Development, Testing and Application to Permanent Wetlands on Private Properties. B. App. Sc. (Hons) Thesis, Charles Sturt University. Steedman, R. J. (1988). Modification and assessment of an index of biotic integrity to quantify stream quality in southern Ontario. Canadian Journal of Fisheries and Aquatic Sciences 45, 492–501. Steinberg, C. E. W. and Hartmann, H. M. (1988). Plankton bloom-forming cyanobacteria and the eutrophication of lakes and rivers. Freshwater Biology 20, 279–287. Suter, G. W. II. (1993). A critique of ecosystem health concepts and indexes. Environmental Toxicology and Chemistry 12, 1533–1539. Turak, E. and Bichel, D. (1994). Using Biological Indicators to Monitor the Health of Freshwater Ecosystems. NSW EPA Technical Report 94/69, EPA Sydney. Turner, R. E., Swenson, E. M. and Summers, J. K. (1995). Coastal Wetlands Indicator Study: EMAP - Estuaries Louisianian Province—1991. US Environmental Protection Agency, Office of Research and Development, Environmental Research Laboratory, Florida EPA/620/R-95/005. van de Graaff, R. H. M. (1988). Land evaluation. In Australian Soil and Land Survey Handbook: Guidelines for Conducting Surveys (R. H. Gunn, J. A. Beattie, R. E. Reid and R. H. M. van de Graaff, eds), pp. 258–281. Melbourne: Inkata Press. Victorian Soil Conservation Authority (1980). Quality aspects of farm water supplies. A report of the Victorian Irrigation Research and Advisory Services Committee. Victorian Soil Conservation Authority, Melbourne.

Evaluating a wetland condition index Walker, K. F. (1985). A review of the ecological effects of river regulation in Australia. Hydrobiologia 125, 111–129. Wicklum, D. and Davies, R. W. (1995). Ecosystem health and integrity? Canadian Journal of Botany 73, 997–1000. Wilcox, D. A. (1995). Wetland and aquatic macrophytes as indicators of anthropogenic

hydrologic disturbance. Natural Areas Journal 15(3), 240–248. Williams, W. D. (1983). Life in Inland Waters. Melbourne: Blackwell Scientific Publications. World Wildlife Fund. (1992). Statewide Wetland Strategies: A Guide to Protecting and Managing the Resource. Washington: Island Press.

159