EIS 1483 AA0681 11 Water Quality in the Snowy River Catchment Area

EIS 1483

AA0681 11

Water quality in the Snowy River catchment area : report on

1996/97 data; nutrient loads in the Thredbo river; trend

assessment NSW YEPT PRIMARY IRDUSIRIES

I AA0681 11 I LAND &WATER CONSERVATION I I I I I I

d I I I I I I I I NSW Department of Land and Water Conservation I I

I DEPARTMENT OF LAND & WATER CONSERVATION CENTRE FOR NATURAL RESOURCES

I I I WATER QUALITY IN THE SNOWY I RIVER CATCHMENT AREA I - Report on 1996/97 Data - Nutrient Loads in the Thredbo River I - Trend Assessment I I I I I I I I H I I DEPARTMENT OF LAND & WATER CONSERVATION CENTRE FOR NATURAL RESOURCES

WATER QUALITY IN THE SNOWY RIVER CATCHMENT AREA - Report on 1996/97 Data - Nutrient Loads in the Thredbo River - Trend Assessment

Zenita Acaba, Lee Bowling, Lloyd Flack June 1998 and Hugh Jones CNR 99.005 [s9697co2.Doc]

CENTRE FOR NATURAL RESOURCES

ISBN 0 7347 5023 4 Public Document

Water Quality in the Snowy River Gatchinent Area, 1996197 Report 1 I I I I I I Cologne I In KOln, a town of monks and bones,

I and pavements fang'd with murderous stones I and rags, and hags, and hideous wenches; I counted two and seventy stenches, I All well defined, and several stinks! I Ye Nymphs that reign o'er sewers and sinks, The river Rhine, it is well known, I Doth wash your city of Cologne; But tell me, Nymphs, what power divine

I Shall henceforth wash the river Rhine? Li I Samuel Taylor Coleridge, 1828 I I I 1 Water Quality in the Snowy River Catchment Area, 1996/97 Report I I I ACKNOWLEDGEMENTS The majority of the sampling for this project was undertaken by staff of the Snowy Mountains Hydro-Electric Authority's Hydrographic Office at Jindabyne, chiefly by Messrs. Phil Boreham, I David Whitfield and Jason Venables. The staff of the Hydrographic Office were also principally involved in the setting up of the automatic water quality samplers and flow gauging stations at the I sites within the Thredbo River subcatchment, which enabled event sampling and load estimates for the project.

I We acknowledge the input of Messrs. Ken Jones and John Denham, of the Snowy Mountains Hydro-Electric Authority, for the co-ordination of the SMII-IEA's input into the project, and for I review of this report. We also acknowledge the continued interest in the project, and in the protection and improvement of water quality in the Snowy Mountains area through total catchment management, of Messrs. I Barry Dunn and Tony Gan, of SMHEA's Operations Planning Division. They are also thanked for their review comments on the report.

I Mr Tom Freece of SMHEA is thanked for providing river flow data for the project. I Sample analysis was undertaken by the Department of Land and Water Conservation's Water Environment Laboratory at Arncliffe. I Finally, we wish to thank Mr Brett Miners of DLWC Cooma (Snowy-Genoa Water Resource Manager), who reviewed the initial draft of this report. I I I I I I I I I V Water Quality in the Snowy River Catchment Area, 1996/97 Report I I I SUMMARY The water quality of Lake Jindabyne and the streams within the upper Snowy River catchment I area was generally good in 1996/97. The lake continued to have low to moderate concentrations of nutrients, and low conductivity and turbidity, as in previous years. There was some oxygen depletion of the deeper waters during summer and autumn whilst the lake was thermally stratified. I There was little algal growth within the reservoir.

The Thredbo River continued to have the best water quality of the three major inflows to the lake, I with low nutrient concentrations and turbidity. The Mowamba Aqueduct had the poorest water quality of these inflows in 1996/97, with Wollondibby Creek being only slightly better. Both had total phosphorus concentrations in the moderate range, and slight turbidity. However, in terms of I nutrient concentrations, all three inflows still met ANZECC (1992) water quality guidelines for the protection of aquatic ecosystems. I Nutrient concentrations and turbidity at the six sampling sites in the Thredbo river subcatchment upstream of Paddys Corner were low during base flow periods, but increased during storm events. I Nutrient loads increased with distance downstream from Deadhorse Gap. Areal load estimates indicate high input of nutrients from the catchment area between Bundilla and just downstream of the Little Thredbo River confluence. This area, which comprises only 1.5% of the area of the I Thredbo River catchment upstream of the Little Thredbo River confluence, was estimated to contribute approximately 24% of the total phosphorus load entering the Thredbo River upstream I of this point, and 2 1 % of the total nitrogen load. Nutrient input from the Little Thredbo River subcatchment was the lowest of any part of the Thredbo River catchment. Impacts from the Lake Crackenback resort were indicated as being I negligible, and there was very little difference in the nutrient input from the Thredbo Village and ski slope area compared to other undeveloped areas of the catchment. Unfortunately the frequent I failure of autosampling equipment at Paddys Corner prevented a number of high flow events from being sampled there, leading to unreliable load estimates in comparison to the other sites. For this reason, nutrient inputs to the Thredbo River from the section of catchment between the Little I Thredbo River confluence and Paddys Corner could not be adequately calculated.

The nutrient loads estimated in this study for the entire Thredbo River catchment area were similar I to those calculated in an earlier study by Bowling (1992). Total phosphorus loads exported from the catchment in terms of kilograms per hectare per year were found to be higher than those I reported in the literature for other forested areas of south-east Australia. The high rainfall of this alpine area, much of which falls as snow which captures airborne particles containing phosphorus, has been proposed as the main reason why the Thredbo River annual areal total phosphorus load I coefficients are higher than those measured elsewhere. Areal total nitrogen load export coefficients were also more similar to those reported from urban catchments than for forested catchments.

I Trend assessments were also carried out using nutrient and turbidity data collected at sites where a minimum of at least five years routine water sampling had been undertaken. These sites included Paddys Corner and others upstream on the Thredbo River; Wollondibby Creek at Gunnadoo; I Mowamba Aqueduct at its Lake Jindabyne outfall; Perisher Creek at Blue Cow; and the Snowy River at Dalgety. No trend was apparent in total phosphorus and total nitrogen concentrations at

Water Quality in the Snowy River C'atchment Area, 1996/97 Report 111

I any site, even after the data had been flow adjusted. However, many sites on the Thredbo River indicated significant increasing trends in turbidity over the period 1992 to 1997, although turbidity was still very low. I A range of recommendations have been suggested for catchment management purposes, and for the protection of water quality in the area, which remains amongst the best of anywhere in New South I Wales. Routine monitoring of selected sites should continue, to enable better estimates of long term trends in water quality at these locations. However, further nutrient load studies of the Thredbo I River catchment are not seen as a priority, unless carried out for a specific purpose such as monitoring the impacts of further development in the valley, monitoring the results of remedial action recommended from this study, or determining inputs from the section of catchment between I the Little Thredbo River confluence and Paddys Corner.

Possible nutrient load studies on Wollondibby Creek and the Mowamba River have been mooted, I but a decision needs to be reached on environmental flows in the Snowy River downstream of Jindabyne Dam before it is decided whether these studies should proceed or not. This decision will influence the ultimate destination of Mowamba River water, that is, whether it will continue to be L diverted to Lake Jindabyne, or flow directly to the Snowy River, and will therefore be a factor when applying for future funding of water quality studies of the Mowamba River.

I A closer alignment of the water quality monitoring being undertaken throughout the catchment and the requirements of the Snowy-Genoa Total Catchment Management Committee is urged, with increased input and direction from the TCM process. Better documentation and co-ordination of all water quality studies in the area is also desirable, and the involvement or funding assistance in future studies from all agencies with an interest in the water quality of the area is encouraged. I Only in this way will the good water quality of the area be maintained into the future, and the need for subsequent remedial action avoided. I I I I I I I I iv Water Qualily in the Snowy River Catchment Area, 1996/97 Report I I TABLE OF CONTENTS i. INTRODUCTION...... 1

I 2. BACKGROUND ...... 1

3. OBJECTiVES OF THE STUDY...... 2

I 4. FUNDING ARRANGEMENTS...... 3

5. SAMPLING PROGRAM...... 3

I 6. ANALYSIS OF SAMPLES...... 6 6.1. ANALYSES UNDERTAKEN ...... 6 I 6.2. QUALITYASSURANCE ...... 6 6.3. DATA ANALYSIS AND PRESENTATION...... 7

I 6.3.1. Calculation of loads ...... 7

6.3.2. Estimation of the Error Associated with the Calculation of Loads...... 8

I 6.3.3. TrendAssessment...... 9

7. RESULTS...... 11 I 7.1. HYDROLOGY...... 11 7.2. ROUTINE SAMPLES ...... 14 I 7.2.1. Total phosphorus...... 14

7.2.2. Total nitrogen...... 17 I 7.2.3. Turbidity ...... 20

7.2.4. Additional data for Lake Jindabyne...... 23 7.3. EVENT SAMPLES ...... 26 I 7.4. AmrwL LoADs ...... 30 7.4.1. Total phosphorus Loads...... 30

7.4.2. Total nitrogen Loads ...... 33 7.5. TREND ASSESSMENT ...... 37 I 7.5.1. Total phosphorus...... 37 7.5.2. Total nitrogen...... 40 I 7.5.3. Turbidity...... 43

8. DISCUSSION ...... 46

U 9. CONCLUSIONS...... 50

10. RECOMMENDATIONS...... 53

I Viii Water Quality in the Snowy River Catchment Area, 1996197 Report I List of Tables I

Table 1. Sampling sites for 1995/96 and 1996/97 Water Quality Monitoring Program by subcatchment. (R I = Routine monthly samples, E = High flow event samples) 3 Table 2. Total number of blue-green and other algae at Lake Jindabyne, 1996-97 data. 25 I Table 3. Estimates of total phosphorus loads from 26 July 1995 to 30 June 1997. 31

Table 4. Areal load calculations for total phosphorus. 31

I Table 5. Estimates of total nitrogen loads for 26 July 1995 to 30 June 1997. 33

Table 6. Areal loading calculations for total nitrogen. 34

I Table 7. Percentage of total phosphorus and total nitrogen loads contributed by various sections of the Thredbo River subcatchment passing 1). Site 22210011, the Thredbo River downstream of the I Little Thredbo confluence, and 2). Site 222541, the Thredbo River at Paddys Corner. (Site 222541 is further downstream than Site 22210011). 35 I Table 8. Seasonal Kendall test results for log total phosphorus performed on raw and flow adjusted data. 37 Table 9. Seasonal Kendall results for log total nNitrogen performed on raw and flow adjusted data. 40 I Table 10. Seasonal Kendall results for log turbidity performed on raw and flow adjusted data. 43 I I I I I I I I I IX I Water Quality, in the Snowy River CatchmentArea, 1996/97 Report I d List of Figures i Figure 1. Map of sampling sites in the upper Snowy River catchment area. 5

Figure 2. Mean daily flows at sites within the Thredbo River subcatchment, and at the major inflow points I to Lake Jindabyne. 13 Figure 3. Boxplot of total phosphorus data for river and stream sites, routine sampling during the period I July 1996 to June 1997. 15 Figure 4. Boxplot of total phosphorus data for Lake Jindabyne sites, routine sampling during the period 1 July l996toJune 1997. 15 Figure 5. Boxplots of total phosphorus for river and stream sites, routine sampling comparing data collected during the periods 1995/96 and 1996/97. 16 I Figure 6. Boxplot of total nitrogen for river and stream sites, routine sampling during the period July l996toJune 1997. 18 I Figure 7. Boxplot of total nitrogen data for Lake Jindabyne sites, routine sampling during the period July l996toJune 1997. 18 l Figure 8. Boxplots of total nitrogen for river and stream sites, routine sampling comparing data collected during the period 1995/96 and 1996/97. 19 I Figure Boxplot of turbidity data for river and streams sites, routine sampling during the period July 1996 to June 1997. 21

Figure Boxplot of turbidity data for Lake Jindabyne sites, routine sampling during the period July I 1996 to June 1997. 21

Figure 11. Boxplots of turbidity for river and stream sites, routine sampling comparing data collected I during the periods 1995/96 and 1996/97. 22 Figure 12. Thermal and dissolved oxygen profiles for Dam Wall Station, Lake Jindabyne. 24

I Figure 13. Total phosphorus load (in dotted lines) and hourly flow (in solid lines) during event sampling of September 1996 to October 1996. 28 I Figure 14. Total nitrogen load (in dotted lines) and hourly flow (in solid lines) during event sampling of September 1996 to October 1996. 29 I Figure Map showing the annual total phosphorus and total nitrogen loads for each site in the Thredbo River catchment. 36 I Figure LOESS trend curve fitted to raw total phosphorus data, 1992-1997. 38 Figure 17. LOESS trend curve fitted to flow adjusted total phosphorus data, 1992-1997. 39 I Figure 18. LOESS trend curve fitted to raw total nitrogen data, 1992-1997. 41 Figure 19. LOESS trend curve fitted to flow adjusted total nitrogen data, 1992-1997. 42 I Figure 20. LOESS trend curve fitted to raw turbidity data, 1992-1997. Turbidity values above 100 NTU was not included in the graph. 44

Figure 21. LOESS trend curve fitted to flow adjusted turbidity data, 1992-1997. 45 I 1 x Water Quality in the Snowy River Catchnzent Area, 1996197 Report I I I List of Appendices I Appendix 1. Boxplot defined 58 I Appendix 2. Ratio Method for Calculating Loads 59 Appendix 3. Summary statistics of routine samples, 1996 to 1997 60 I Appendix 4. Summary statistics of event samples. 64 I Appendix 5. Total phosphorus and total nitrogen concentration at the start of an event, by site 66 I I I I I I I I I I I I I Water Quality in the Snowy River Catchment Area, 1996/97 Report I I I GLOSSARY Anoxic. The absence of oxygen in the water.

I Campaign sampling. Special sampling when water samples are taken for specific purposes outside of the routine sampling program, if one exists. Campaign samples may be taken as special I samples in locations where there is no routine sampling program. Cyanobacteria. Blue-green algae. Excessive growth of cyanobacteria is usually a symptom of I high nutrient pollution in lakes, reservoirs and slow-flowing streams. They can produce toxins, form scums, and cause taste and odour problems in drinking water supplies. Their decomposition after a bloom can reduce dissolved oxygen concentrations in water. I Epilimnetic. Referring to the surface layers of water in a lake or reservoir during thermal stratification. These waters overlay the thermocline, and are usually warmer, and have higher I dissolved oxygen concentrations than deeper waters below the thermocline, with which they do not mix.

I Eutrophic. A condition where there are high concentrations of plant nutrients in a water body, often accompanied by excessive algal and macrophyte growth, deoxygenation of the bottom waters due to the decomposition of this growth, and other water quality problems. Waters are often I considered to be eutrophic if total phosphorus concentrations exceed 0.035 mg/L, and total nitrogen concentrations exceed 1.00 mg/L.

Hypolimnetic. Referring to the bottom stratum of water in a lake or reservoir which is thermally stratifled—that is the waters below the thermocline. These waters usually remain cold, and can I suffer oxygen depletion and associated water quality problems as they have no contact with the atmosphere for oxygen replenishment, as they are overlain by waters of the thermocline and I Macrophyte. A large aquatic plant growing in a water body. These may be rooted in the sediments I at the bottom, or free-floating. Mesotrophic. The nutrient status of a stream, lake or reservoir where plant nutrients are present at moderate levels, but are neither scarce nor abundant. The nutrients are usually present at sufficient I concentrations to allow some algal and macrophyte growth, but not excessive growth or algal blooms. Mesotrophic waters are often considered to have total phosphorus concentrations ranging I between 0.010 and 0.035 mg/L, and total nitrogen concentrations ranging between 0.60 and 1.00 mg/L.

I Oligotrophic. The nutrient status of a stream, lake or reservoir where plant nutrients are only present at very low concentrations, and are insufficient to allow much plant or algal growth. Waters are commonly considered to be oligotrophic when total phosphorus concentrations are less I than 0.010 mg/L, and total nitrogen concentrations are less than 0.60 mg/L.

Phytoplankton. Microscopic algae that float within the water column of a lake, reservoir, or I stream. They can be single celled, colonial, or filamentous. I Water Quality in the Snowy River Catchment Area, 1996/97 Report vi I

Thermal stratification. The division of a waterbody into horizontal layers of different densities, I due to the differential uptake of solar radiation by water at different depths. Surface waters heat more than deeper waters, and become less dense. This density difference prevents the complete I mixing of the lake over all depths. Usually three strata are formed: the epilinmion or surface layer; the thermocline where there is greatest change in temperature with depth; and the hypolimnion, towards the bottom. Thermal stratification generally commences in spring and breaks down in 1 autumn.

Thermocline. The stratum of a thermally stratified lake or reservoir over which the maximum I temperature change with depth occurs. It is the layer of water that lies between the epilimnion and the hypolinmion, and is sometimes also called the metalimnion.

I I LI I 1 I I I I I I I I Water Quality in the Snowy River Catchment Area, 1996197 Report Wi I

I 1. INTRODUCTION

Lake Jindabyne and the Snowy River form an important aesthetic, visual and recreational amenity I on the eastern approaches to the Snowy Mountains ski-fields and Kosciusko National Park in south east New South Wales. Increasing tourist development within this area has raised concerns that the water quality of the river and lake, and their tributary streams, may deteriorate.

Investigations to determine the water quality status of Lake Jindabyne, the Snowy River and their U tributary streams, have been under way for a number of years now. Initial discussion of the water quality investigation program took place between what was then the NSW Department of Water Resources (DWR) (now part of the Department of Land and Water Conservation (DLWC)) and I the Snowy Mountains Hydro-Electric Authority (SM}iEA), with some input from the National Parks and Wildlife Service earlier in the program. The major aspects of these discussions were I detailed in a report to the Snowy Mountains Council (Operation Engineers Committee Report No. 145 - Appendix T5, August 1992). This has subsequently been updated in an ongoing water quality management program. Points of agreement between the two Agencies covered the 1 objectives of the program, the sampling program and on joint funding arrangements based on the benefits received by the respective parties.

I This report presents the water quality data collected for the Snowy Mountains Water Quality Program during the period 1995/96 and 1996/97. This report is intended to:

I Provide information on all water quality data collected from the area in 1995/96 and 1996/97 following earlier annual progress reports;

I Report the results of nutrient load studies at sites within the Thredbo River subcatchment during 1995/96 and 1996/97; and

I Assess trends in water quality data for the 3 major inflows to Lake Jindabyne:- Thredbo River at Paddys Corner (Site 222541), Wollondibby Creek at Gunnadoo (Site 222544) and I Lake Jindabyne Mowamba Aqueduct Outflow (Site 22210142). 1 2.. BACKGROUND Concerns of deteriorating water quality in Lake Jindabyne led to a joint SMHEA!DWR study of the lake and its major inflows in 1989 (Bowling 1993). This examined the water quality of Lake I Jindabyne during the period 1989-1992 and calculated a nutrient budget for the lake, including inputs from each of the major inflows. The study found that Lake Jindabyne was in an advanced 1 mesotrophic state in terms of total phosphorus, with concentrations approaching 0.03 mg/L. Total nitrogen concentrations were low, however, and algal presence sparse. While Wollondibby Creek had the highest nutrient concentrations of the major inflows, the Thredbo River contributed the I greatest amount of the total phosphorus load to the lake, in keeping with its high hydraulic loading. Unexpectedly, the Thredbo River sub-catchment also had high annual areal total phosphorus loads. The study concluded that although the water quality of Lake Jindabyne was amongst the best in H New South Wales, there was considerable potential for its sudden deterioration should nutrient I inputs increase and nutrient concentrations within the lake rise. 1 Water Quality in the Snowy River Catchment Area, 1996/97 Report I

Additional water quality studies of the Snowy River catchment area were commenced by the DWR I in 1991. These studies included sampling programs in streams within the Kosciusko National Park, as well as in the Snowy River and tributaries downstream of Jindabyne Dam, in the Dalgety and Berridale areas (Bate 1992). Bate (1992) found generally good water quality during low flow [1 periods in streams of the Kosciusko National Park, but that nutrient concentrations and turbidity in these streams increased during high flow events, especially downstream of resorts. The study also I found nutrient concentrations in the Snowy River at Dalgety to be higher than those in Lake Jindabyne, although turbidity was similar. Both turbidity and nutrient levels were deemed I satisfactory for water quality purposes except during high flows following storni events. The two studies were combined in 1992 as the Snowy River Catchment Water Quality Study, under the OEC Report No. 145 - Appendix T5, August 1992 agreement between SMHEA and I DWR. A number of changes have occurred since then, with greater emphasis on determining nutrient and sediment sources in the Thredbo, Mowamba and Wollondibby subcatchments of Lake El Jindabyne. Since 1994, less emphasis has been given to streams in the Kosciusko National Park and downstream of Jindabyne, with monitoring at only one site in each area. Studies since then have concentrated more on nutrient loads within the Thredbo River. Annual progress reports have I been completed for 1992/93 (Bowling et al. 1993), 1993/94 (Bowling and Acaba 1995), 1994/95 (Kinross and Acaba 1996) and 1995/96 (Maim et al. 1997). Water quality monitoring and nutrient load investigations during the 1995/96 and 1996/97 period are detailed in this report. The program I is still evolving, with water quality concerns downstream of Jindabyne Dam again becoming a higher profile issue.

3. OBJECTIVES OF THE STUDY I The objectives of the study during 1995/96 and 1996/97 were: Upper Snowy Catchment Area (Snowy Catchment above Lake Jindabyne). I Monitor Perisher Creek to detect changes and trends in nutrient concentrations; [1 Middle Snowy Catchment Area (Lake Jindabyne and its catchment). Monitor the water quality in Lake Jindabyne to detect changes in the overall "health" of the reservoir; I Monitor nutrient concentrations in the major inflows of Lake Jindabyne to detect changes and trends in the water quality of these inflows; I Identif' sources of pollution within the Thredbo River sub-catchment. Similar investigations within the Wollondibby, Mowamba and Cobbin subcatcbments are proposed for future years, following completion of the work within the Thredbo I subcatchment. Lower Snowy Catchment Area (Snowy River downstream of Jindabyne Dam). I Monitor the Snowy River at Dalgety to detect changes in water quality I

I 2 Water Quality in the Snowy River Catchment Area, 1996/9 7 Report I I I 4. FUNDING ARRANGEMENTS Funding arrangements for 1996/97 were the same as those for the previous year (Maini et al I 1997). Funding was divided as follows:- DLWC SMHEA

I Upper Snowy Catchment Area 80% 20% Middle Snowy Catchment Area I Lake Jindabyne 10% 90% Major inflow points 40% 60% I Subcatchment streams 40% 60% Lower Snowy Catchment Area 100% 0%

I Of these, studies in the Middle Snowy Catchment Area accounted for approximately 95% of the total budget (of which 10% was for Lake Jindabyne, the remainder for the inflow streams).

I Note: funding for the Lower Snowy River Catchment Area was undertaken entirely by the DLWC, since this part of the study was outside the scope of the OEC agreements.

I 5. SAMPLING PROGRAM I The sampling sites for the 1995/96 and 1996/97 water quality monitoring program are listed in Table 1 and are illustrated in Figure 1.

I Table 1. Sampling sites for 1995/96 and 1996/97 Water Quality Monitoring Program by subcatchment. (R = Routine monthly samples, E = High flow event samples)

I Area Site Site Name Sampling Number. Status Upper Snowy I Perisher Creek 222513 Perisher Creek at Blue Cow R - Middle Snowy Thredbo River 22210019 Thredbo River at Deadhorse Gap R E I 22210016 Thredbo River D/S Thredbo STW R E 22210012 Thredbo River at Bundilla R E 22210017 Little Thredbo River at Alpine Way R E I 22210015 Little Thredbo River at Walking Track Bridge R E 22210011 Thredbo River D/S Little Thredbo River R E 222541 Thredbo River at Paddys Corner R E I Wollondibby 222544 Wollondibby Creek at Gunnadoo R Creek Lake Jindabyne 22210001 Lake Jindabyne at Dam R I 22210002 Lake Jindabyne at Jindabyne East R 22210003 Lake Jindabyne at Hiawatha R Mowamba River 22210142 Lake Jindabyne Mowamba Aqueduct Outflow R - I Lower Snowy Snowy River 222006 Snowy River at Dalgety R - I Water Quality in the Snowy River Catchment Area, 1996197 Report I I

I Water samples are collected at all of the above sites for the analysis of total phosphorus concentration, total nitrogen concentration and turbidity. The samples collected were frozen before being sent to the Department's Water Environmental Laboratory at Arncliffe for analysis. I Nutrients samples were taken from the surface, middle and bottom of Lake Jindabyne. In addition, thermal and dissolved oxygen profiles, electrical conductivity, pH and turbidity were measured in situ in Lake Jindabyne.

The sampling program to monitor water quality at these sites had two major components:

I Routine sampling - Samples were taken at all sites monthly, except for the Snowy River at Dalgety where only eight samples were taken in 1995/96, and 10 in 1996/97. Long term routine I sampling is undertaken to enable statistical analyses to determine trends in the water quality parameters measured at each site over time. The sites of specific interest for trend analysis include the three major iaflows to Lake Jindabyne: Thredbo River at Paddys Corner (Site 222541), Fi Wollondibby Creek at Gunnadoo (Site 222544) and Lake Jindabyne Mowamba Aqueduct Outflow (Site 22210142), as well as Perisher Creek at Blue Cow (Site 222513) and the Snowy River at Dalgety (Site 222006). A minimum of five years of data is needed to undertake trend analyses. To Li date, six years of data have been collected for these sites. This report will incorporate trend analysis for the sites where sufficient data are available.

I Routine monthly samples were taken for two years at the sites within the Thredbo River subcatchment, to provide sufficient temporal coverage of base flow water quality conditions at these ri sites for use in load estimates. Routine sampling was undertaken at the three sites on Lake Jindabyne during the summer of I 1995/96. Samples were also collected monthly at Dam Station on Lake Jindabyne during 1996/97, and at two-monthly intervals at Jindabyne East and Hiawatha Stations. No sampling was undertaken between July 1996 and December 1996, however, as the instrument used to collect I profile data was not working.

Event sampling - Samples were collected during periods of high flows at all seven sites within the I Thredbo River subcatchment in both 1995/96 and 1996/97 using 'Gamet' automatic water samplers. Sampling was triggered when there was a rise in the water level. Six high flow events I were sampled in 1995/96, while more than 10 events were sampled at each site during 1996/97. No campaign sampling was done in the two-year period from 1995 to 1997.

I I I I Water Quality in the Snowy River Catchment Area, 1996/97 Report ------

/ / I

LSO NS xD3 -VALLEY - \ __- -•.s 13ERR?EALE ) h--- 7. __ 1 (\ / .AO - 02 1 SMNS

PERIS HER / - r1 '001 Oil — p142 _..__-_- JINE 0115 012 7 1_7 \ / j( / I

016 i1- f / -fl-1R'OnO -- \ ,2000 0ALGEry 7) oig / .- 9 10 -— KItorr,ofro0 SornIin0 SiIe,

Figure 1. Map of sampling sites in the upper Snowy River catchment area.

Water Quality in the Snowy River Catchment Area, 1996197 Report I

6. ANALYSIS OF SAMPLES 1 6.1. Analyses Undertaken Water samples from the surface, mid-depth and bottom at sites on Lake Jindabyne were analysed for total phosphorus, total nitrogen and turbidity. Additional surface samples from the lake which I had been preserved with Lugols iodine solution were used for the identification and enumeration of algae. i Routine water samples from catchment streams were analysed for turbidity and total phosphorus (TP) following standard methods (APHA 1992), and for total nitrogen (TN) following the method I of D 'Elia et al. (1977). Event samples obtained using the "Gamet" autosamplers were analysed for total phosphorus, total nitrogen and turbidity. I Water samples were analysed at the DLWC's Water Environment Laboratory at Arncliffe.

6.2. Quality Assurance

Documentation of samples using water quality sample log sheets filled out at the time of sampling I allowed easy identification of most samples from sampling, dispatch and laboratory analysis, and assisted subsequent data retrieval. The site, date, time and method of sampling were verified for all data used in this report by referencing to the Sample Log Sheets. Event samples and routine I samples were also identified. If data could not be verified, it was excluded from the data base and not used in the data analyses undertaken for this report (e.g. the absence of a date on the sample I log sheet, and unusually high turbidity due to the sampling probe entering bottom sediments). No data has been rejected due to being unverifiable for the two years reported here.

I The data verification stage also includes checking the sample preservation method and the time taken between preservation and laboratory analysis (see Australian standard "Selection of sample containers and preservation method"). This is done by checking the number of days elapsed using I the fields "Date Sampled" and "Date Analysed" for each record. For frozen total phosphorus (TP) and total nitrogen (TN), standard guidelines allow up to 28 days from the time the water sample is collected to the time the sample is analysed at the laboratory. For the 1996/97 TP and TN data, I almost 40% of samples collected were analysed at least 28 days or more after collection. Instead of rejecting these data, further checking and verification of each record was conducted e.g. these data I were compared to the range of TP and TN values measured in previous years. If the data fell within the range of values for TP and TN for a particular site, these data were utilised in the report. A similar data verification procedure was applied for turbidity. No data was rejected from I the 1996/97 data set.

The laboratory engaged to conduct the chemical and turbidity analyses was the DLWC's Arncliffe I laboratory. It is registered by NATA, the national accreditation agency that ensures high quality analytical results. Analytical errors in total phosphorus, total nitrogen and turbidity analyses are I all less than 5 per cent.

I 6 Water Quality in the Snowy River Catchment Area, 1996197 Report I

The quality assurance procedures undertaken ensured that both the quantity and quality of data I from the Thredbo River and other sites reported below was of the highest standard, and that the adequacy of these data for the statistical analyses performed is beyond doubt. I 6.3. Data Analysis and Presentation I For the purposes of statistical analysis, observations that were below the detection limit for a particular parameter were shown as half the detection limit. For example, the detection limit for total phosphorus is 5 .tg[L, so data observations below this were shown as 2.5 j.tg/L. This method I of treating less than the detection limit (LOD) values by halving the LOD was suggested by Nehis and Aldand (cited by Gilbert 1987, p.lTl). This assumes a uniform distribution of values between zero and the less than detection limit (LOD). This procedure will have a minimal impact on the I estimate of the mean unless the maximum value of the observations above the detection limit is small. This procedure can also result in severe underestimation of the variance if the number of I values less than detection comprise a large proportion of the total data set. Except for Deadhorse Gap (Site 22210019) and Thredbo River DIS STW (Site 22210016), all of the sites had less than 5% of the total number of samples with LOD values for total phosphorus. LOD values for total 1 phosphorus at Deadhorse Gap (Site 2210019) and Thredbo River D/S STW comprised of approximately 20% and 7% of the total samples respectively. This may affect the resulting load estimates at this site. However, as this site had a relatively low total phosphorus load this should I not have a big impact on the overall interpretation of these load estimates.

The use of non-parametric measures of spread like the interquartile range will circumvent 1 underestimation of variances. Trend results should not be affected as the Seasonal Kendall test method is a non-parametric technique for analysing trends.

Previous reports presented values for total phosphorus and total nitrogen in milligrams per litre (mg/L). For this report, these values were converted into micrograms per litre (.tg/L), as this form I of presentation leads to numbers that are easier to comprehend at a glance'; for example, 5 .Lg/L rather than 0.005 mgIL.

I Summary statistics were calculated for each site and for each water quality determinand. Boxplots and time series plots of data were also generated and presented in this report. For a full I explanation of a boxplot, see Appendix 1. 6.3.1. Calculation of loads

I Nutrient loads were calculated for both years to determine the influence of different input sources on the quality of water downstream. Load estimation requires both nutrient concentration data and instantaneous flow data (defined as the flow measured at the date and time the samples were I collected). Reliable load estimates require sampling of both low, or base-flow, and high flow events. Hence, load calculations were only undertaken for those sites that were sampled for both I base flow and storm events. As the instantaneous flow data at the time of sampling were incomplete for some sites, the flow I data for the hour of the day closest to the sampling time was used. Alternatively, if no hourly flow I Water Quality in the Snowy River Catchment Area, 1996/97 Report I

I data were available, the mean daily flow for the day the water quality sample was collected was substituted. I The total annual flow was needed to estimate loads but short gaps were present in the daily discharge record for most sites (Figure 2). The missing daily flows were estimated from the hydrographs of neighbouring stations by regressing the discharge of the neighbouring site with the I discharge of the target site and using this relationship to predict missing stream discharge values. The best models were using time series models based on the logarithms of stream discharge and fitted with autoregressive error terms of up to lag 3. The Autoregressive time series model is used I if the successive values of the error term are not independent that is the value which the error term (pt) assumes in any one period is independent from the value which it assumed in any previous I period. If this assumption of independence is not satisfied, then we say that there is autocorrelation of the error term (Koutsiyannis, 1977). Because we are dealing with time series data such as stream discharge, often the error term is not independent. Durbin-Watson statistics was used to test H for autocorrelation. The plots of partial autocorrelation function (PACF) in SAS software provides information on the number of lags the current time series value are correlated. The number of lags were between 1 and 3 for all sites. More complicated transfer function time series models were not I developed because regression models with autocorrelated errors provided good predictions of stream discharge at neighbouring sites.

I Continuous flow data were available beginning 25 July 1995 for Little Thredbo River at Walking Track Bridge site (Site 2210015). Therefore the starting date for the calculation of loads was at I this date, giving total load estimates for 706 days. Annual loads were calculated by the ratio method (Cochran 1977). Using this method, load is I calculated as the product of the flow-weighted average of the constituent concentration and the total annual discharge for the site. For more information on this method of calculating loads refer to Appendix 2. Some sites were sampled more frequently than others resulting in biased estimates, I particularly for sites with fewer samples during storm events. Also, estimates made by combining base flow data with high flow data tends to result in the overestimation of loads, as very high values measured during flow events can have undue influence on the estimates. To minimise this I bias and to improve the precision of the load estimates, stream discharge was stratified as base flow, the rising limb of the hydrograph, and the falling limb of the hydrograph. The component I loads were then determined for each stratum and summed to provide a total load estimate.

Load values were converted to areal estimates by dividing the load by the area of each local I catchment. The areas of the various sections of the Thredbo River subcatchment between each of the sampling sites were marked on the relevant 1:50,000 scale topographic maps, and their areas I measured using a planimeter. 6.3.2. Estimation of the Error Associated with the Calculation of Loads

I 'Be errors of the load estimates were estimated by a bootstrapping algorithm (Efron and Gong, 1983). Bootstrapping is a non-parametric technique that is appropriate when the sampling distribution is unknown or the calculation of errors is difficult or intractable. A random sample of I size n is taken with replacement from the original sample of n observations. The ratio estimate of I 8 Water Quality in the Snowy River Catchment Area, 1996/97 Report F'j

load is calculated from the bootstrap sample and the resampling procedure is repeated until there I are B bootstrap estimates of the load. I The bootstrap BCa method was used to calculate 95% confidence intervals from B1000 bootstrap replicates (Efron and Tibshirani 1993). The bootstrap mean and standard error of the load estimates were also calculated.

6.3.3. Trend Assessment I The Seasonal Kendall trend test was used to detect trends in total phosphorus, total nitrogen and turbidity. Data collected for the period 1992 to 1997 for the seven sites within the Thredbo River catchment, Wollondibby Creek at Gunnadoo, the Mowamba Aqueduct at the Lake Jindabyne I outfall, Perisher Creek at Blue Cow, and the Snowy River at Dalgety were analysed for trend.

The Seasonal Kendall test is a non-parametric analysis for monotonic trend. Apart from detecting I trend, it calculates a Seasonal Kendall slope estimator which provides an estimate and the probability of the magnitude of the trend. The Seasonal Kendall test is not exact, however, when the data is serially correlated. Statistical significance can be overestimated if a positive correlation I is present, as found by Cunningham and Morton (1983). While the data were not examined for serial correlation, it was considered that this was unlikely to be an issue in small, unregulated, upland streams. Previous experience has shown that monthly sampling intervals are adequate in avoiding serial correlation in samples collected from these types of streams.

The Seasonal Kendall test has been used by several authors to assess trends in water quality, including water quality reports done for the Key Sites Program by Preece et al. (1994) and the Salt I Trends Report by Williamson et al. (1997). To be able to assess trends using the Seasonal Kendall test, the data set for each water quality indicator for each site has to span a period of at least four years. Also, at least 24 data points need I to be present over a 5-year period (the maximum possible is 60 points - that is one data point for each month for 5 years).

I Trend is considered to exist if the Seasonal Kendall test result has a probability (p) of occurring by chance of less than 0.05 (p<0.05). The magnitude of trend is given by the Seasonal Kendall slope, I where negative slope represents an improvement in water quality and positive slope represents deterioration. I In order to make the trend slopes more interpretable, concentrations were log-transformed before applying the Seasonal Kendall test. The resulting log concentration slope can be expressed in percent per year (I-{irsch et al., 1991). The estimated slope expressed as the percentage change I from the beginning of any year to the end of that year will be (eB]) x 100 where B is the estimated slope. The percentage change was converted into original units by multiplying (eB1) with the median concentration. The annual trends are point estimates. The 95% confidence intervals I provide an indication of the precision of the trend slope estimates or of the power of the analysis. Thus, a broad confidence interval shows that the point estimate is poor while a narrow confidence I interval shows that the point estimate is good.

Water Quality in the Snowy River Catchment Area, 1996/97 Report H

I The Seasonal Kendall test was calculated for both raw data and for data from which the effect of flow was removed. The benefits in calculating trend from flow adjusted data are: I 1. Flow may be one of the main factors contributing to variation in water quality through time; 2. Removal of variability associated with flow improves the power and efficiency of the trend I test; and 3. It helps to determine whether a trend observed in the raw water quality data is an artefact of I the prevailing flow pattern or is due to other factors. The process of removing the flow effects from the data involves modelling concentration and flow I in an equation. The LOESS (Locally Weighted Scatterplot Smoothing) technique (for an explanation of LOESS; see below) was used to fit log concentrations with log flow. The residuals (the difference between the concentration value and the predicted value) were then used for the I seasonal Kendall test. Another approach to examining trend is the use of a non-parametric smoothing technique called I LOESS (Cleveland, 1979). This graphical technique was also applied to the water quality data. This technique has been recommended as it complements or provides an alternative to the Seasonal Kendall test, and is more sensitive where the trend is not merely a simple change in one overall I direction. (Esterby, 1992). \Vhile the Seasonal Kendall test is limited to testing a monotonic increase or decrease in trend, LOESS is an invaluable complement by not treating trend as constant, and demonstrates trend variability over time (Williamson, 1997). Another advantage of I using LOESS is that it doesn't have an underlying assumption of linearity and normal distributions of data. I LOESS plots have been produced for all sites with sufficient data. They have been used to illustrate graphically the pattern of both raw data and flow adjusted data over time. I I I I I I I I Water Quality in the Snowy River Catchment Area, 1996/97 Report 10 I

I 7. RESULTS 7.1. Hydrology I Hydrographs for sampling sites within the Thredbo River subcatchment, and for the major inflow sites to Lake Jindabyne, from July 1995 to June 1997, are shown in Figure 2. Flow in the Thredbo I River increased considerably with distance downstream from Deadhorse Gap (Site 22210019) to Bundilla (Site 22210012). There appears to be little difference in flow between the two sites on the Little Thredbo River. Likewise there appears to be little difference in flow between sites on the I Thredbo River from Bundilla downstream. Inflow to the Thredbo River from sources between Bundilla (Site 2210012) and Paddys Corner (Site 222541), including the Little Thredbo River, I Sawpit Creek and other streams must be small in comparison to the flow that is already present in the Thredbo River at Bundilla.

Most of the inflow to Lake Jindabyne comes from the Thredbo River, with inflows from Wollondibby Creek (Site 222544) and the Mowainba Aqueduct (Site 22210142) being small in comparison.

Total inflows to Lake Jindabyne were slightly greater in 1995/96 compared to 1996/97. Total flow in the Thredbo River at Paddys Corner was 205,894 ML in 1995/96, compared to 199,560 ML in 1996/97 (Tom Freece, SMI-IEA, pers. comm). The Mowamba Aqueduct contributed 37,536 ML and 32,135 ML in 1995/96 and 1996/97 respectively, while inflows from Wollondibby Creek totalled 7,368 ML and 4,415 ML in these years. There tended to be a longer period of high flow in the Thredbo River in 1995/96 compared to 1996/97, although the actual high flows were of a smaller volume than in the following year.

1 I I I

I 11 Wafer Quality in the Snowy River Catchment Area, 1996/97 Report 22210019 - Thredbo River at Deadhorse Gap

5000

4000

3000

2000

1000

01JUL95 01JAN96 01JUL96 01JAN97 01JUL97

22210016 - Thredbo River D/S Thredbo STW

5000

4000

3000

2000

1000

01JUL95 01JAN96 01JUL96 01JAN97 01JUL97

22210012 - Thredbo River at Bundilla

5000 > 4000

3000

2000

1000

0•

01JUL95 01JAN96 01JUL96 01JAN97 01JUL97

22210017 - Little Threbro River at Alpine Way

5000

4000

3000

2000

1000

01JUL95 01JAN96 01JUL96 01JAN97 01JUL97

22210015 - Little Thredbo River © Walking Track Bridge

5000

4000

3000

2000

1000

0 II

01JUL95 01JAN96 01JUL96 01JAN97 01JUL97

Figure 2 continued next page

12 Water Qualn'y in the Snowy River Catchment Area, 1996/97 Report I

I 2221 001 1 - Thredbo River D/S Little Thredbo confluence

5000 > I 4000 3000

2000 I 1000 0 II I 01JUL95 01JAN96 01JUL96 01JAN97 01JUL97 222541 - Thredbo River at Raddys Corner

5000 I > 4000

3000 ZI I 2000 1000

I 01JUL95 01JAN96 01JUL96 01JAN97 01JUL97

222544 - Wollondibby Ck at Gunnadoo

I 5000

- 4000 I 3000 2000 I 1000 01JUL95 01JAN96 01JUL96 01JAN97 01JUL97

I 22210142 - Mowamba Aqueduct Outflow at Lake Jindabyne 5000 > 4000 I 3000 2000

1000 I 0

01JAN96 01JUL96 01JAN97 01JUL97 I 01JUL95 I I Figure 2. Mean daily flows at sites within the Thredbo River subcatchment, and at the major 1 inflow points to Lake Jindabyne. I 13 Water Quality in the Snowy River Catchment Area, 1996/97 Report I LI 7.2. Routine Samples

I 7.2.1. Total phosphorus

Note: ANZECC (1992) guidelines for the protection of aquatic ecosystems I . 5-50 ..Lg/L (Lakes and Reservoirs)

I 10-100 g/L (Rivers & Streams)

Total phosphorus (TP) data collected on a routine basis during the period 1996/97 are presented as I boxplots in Figure 3 for the river and stream sites, and in Figure 4 for the Lake Jindabyne sites. Summary statistics of total phosphorus measured at various river and stream sites and the lake I sites are found in Appendix 3A.

The highest median TP value (shown as a dot in the boxplot) was observed at the Mowamba I Aqueduct Outflow site (Site 22210142), with a value of 32.5 g/L, followed by Wollondibby Creek at Gunnadoo (Site 222544), with a median value of 27.5 g/L.

I The highest total phosphorus concentration measured, 210 jtg/L, was observed at the site on the Thredbo River below the Sewerage Treatment Works (Site 22210016). This high value occurred on a routine sampling occasion that also coincided with a storm event, on 12 September 1996. The I lowest maximum concentration from routine sampling of all stream sites was measured at Perisher Creek at Blue Cow (Site 222513) in the Upper Snowy catchment area, with a value of only 30 I .tg/L. Mostriver sites had minimum total phosphorus concentrations below the detection limit of 5 j.tg/L, apart from sites at Alpine Way (Site 22210017), Gunnadoo (Site 222544), and at the Aqueduct outflow (Site 22210142), all of which had minimum values above 10 g/L.

The three sites on Lake Jindabyne had generally low to moderate TP concentrations, ranging from below detection to 20 ig/L. The maximum value (20 j.g/L) was measured at Dam Station (Site 22210001) in both the surface and bottom waters. Median TP values of 15 .tg/L occurred in the 1 mid-depth and bottom waters, while the median TP concentration of the surface waters was below the detection limit. There were insufficient data for the other two sites to enable accurate comparisons to be made between sites (see Appendix 3A). I Most sites showed very little difference between the median TP concentrations measured for I 1995/96 and those measured for 1996/97 (Figure 5), Nevertheless, a few sites did show marked differences. At Deadhorse Gap (Site 22210019), median values increased from of 5 .tg/L in 1995/96 to 12.5 .tg/L in 1996/97 (Appendix 3D). The median concentrations of the Snowy River I at Dalgety (Site 222006) dropped from 38 .tg/L in 1995/96 to 15 pg/L in 1996/97. Of the three major inflows, the highest median value observed in 1996/97 occurred at the Mowamba Aqueduct Outflow (Site 22210142), while in 1995/96 the highest median value was recorded at Wollondibby I Creek at Gunnadoo (Site 222544).

Water Quality in the Snowy River CatchmentArea, 1996/97 Report 14 I I

I 12 12 12 12 12 12 12 12 12 12 10 I

I ------9------0 0 I I ------I

I 222513 22210019 22210016 22210012 22210017 22210015 22210011 222541 222544 22210142 222006 Upper Middle Snowy Lower Snowy Snowy

I Figure 3. Boxplot of total phosphorus data for river and stream sites, routine sampling during I the period July 1996 to June 1997. 1= 5 5 5 2 2 2 3 3 2 I I I ------I I

I 1 TOP MIDDLE BOTTOM TOP MIDDLE BOTTOM TOP MIDDLE BOTTOM I Site 22210001 22210002 22210003 Figure 4. Boxplot of total phosphorus data for Lake Jindabyne sites, routine sampling during I the period July 1996 to June 1997.

I Water Quality in the Snowy River Catchrnent Area, 1996/97 Report 15 I ------

n= 1212 1212 1312 1212 1212 1212 1212 1212 1212 11 12 8 10 1000

-Th —J 100 ---- —p--

A 9 A B A B A B A B A B A B A B A B A B A B 222513 22210019 22210016 22210012 22210017 22210015 22210011 222541 222544 22210142 222006 Upper Middle Snowy Lower Snowy Snowy

Figure S. Boxplots of total phosphorus for river and stream sites, routine sampling comparing data collected during the periods 1995/96 and 1996/97. Note: A and B labels on the x-axis refer to the years 1995/96 and 1996197 respectively.

16 Water Quality in the Snowy River Catchment Area, 199619 7 Report I I 7.22. Total nitrogen Note: ANZECC (1992) guidelines for the protection of aquatic ecosystems

I 100-500 j.tg/L (Lakes and Reservoirs) I 100-750 g/L (Rivers & Streams) Summary statistics for total nitrogen (TN) for the period 1996/97 are presented in Appendix 3B. Data are also presented as boxplots for the river and stream sites in Figure 6 and for the lake sites I in Figure 7.

Minimum and median values were low at all Thredbo River sites, with TN values ranging from 50 I g/Lto 150 g/L. I The highest total nitrogen concentrations during the 1996/97 period were measured at Wollondibby Creek at Gunnadoo (Site 222544), Perisher Creek at Blue Cow (Site 222513) and the Thredbo River D/S of Little Thredbo (Site 22210011), with the maximum value being 1000 I tg/L. Wollondibby Creek at Gunnadoo (Site 222544) had the highest median value of 375 g/L, followed closely by the Mowamba Aqueduct Outflow (Site 22210142), with a median I concentration of 350 g/L. There was little variation in nitrogen concentrations at the three sites on Lake Jindabyne, with values ranging from 150 g/L to 250 j.ig/L. The maximum total nitrogen concentration measured I in the lake was 500 g/L, at the surface of Dam Station (Site 22210001), on the 22 April 1997.

Figure 8 shows the comparisons between total nitrogen concentrations measured in 1995/96 and I those measured in 1996/97, Perisher Creek at Blue Cow (Site 222513), Wollondibby Creek at Gunnadoo (Site 222544) and the Mowamba Aqueduct (Site 22210142) sites had maximum total I nitrogen concentrations above the 750 .tg/L ANZECC guidelines in both years (Appendix 3D). The highest total nitrogen concentration was measured in 1995/96, with a value of 4700 .ig/L for the Thredbo River at Paddys Corner. The explanation for this could be that the routine sampling H coincided with the start of a storm event. This high value occurred on 7 May 96. Although no event samples were taken at Paddys Corner at this time (possibly due to equipment failure), both routine and event samples were taken at another site upstream on this date. Median total nitrogen I concentrations at all three major inflows to Lake Jmdabyne were lower in 1996/97 than in I 1995/96. I I LI

I Water Quality in the Snowy River Catchment Area, 1996/9 7 Report 17 I I

I 12 12 10 12 12 12 12 12 12 12 12 I I -J 1000 01 ------.------3 Cl

I 0 01 0

0 I - 100 ------F- +1 I +

I 10 222513 22210015 22210016 22210012 22210017 22210015 22210011 222541 222544 22210142 222006 Upper Middle Snowy Lower I Snowy Snowy Figure 6. Boxplot of total nitrogen for river and stream sites, routine sampling during the I period July 1996 to June 1997.

5 5 5 2 2 2 3 3 2 I 1e1I1u1s1 I

-J 1000 I 0) 3 I

0 o 100 I F- I

10- I I I I I I I I I I TOP MIDDLE BOTTOM TOP MIDDLE BOTTOM TOP MIDDLE BOTTOM I Site 22210001 22210002 22210003

Figure 7. Boxplot of total nitrogen data for Lake Jindabyne sites, routme sampling during the I period July 1996 to June 1997.

I Water Quality in the Snowy River Catchn7ent Area, 1996/97 Report 18 I — — — — — — — no — — — — — — — — — — — —

10000 - n= 1212 1212 1312 1212 1212 1212 1212 1212 1212 11 12 E11] - -

LI LI LI:3 1000 LI 03 :3 LI LI LI El

kIIIIIiII I

10-L - I I I I I I A B A B A B A B A B A B A B A B A B A B A 9 222513 22210019 22210016 22210012 22210017 22210015 22210011 222541 222544 22210142 222006 Upper Middle Snowy - ----] Lower Snowy Snowy

Figure 8. Boxplots of total nitrogen for river and stream sites, routine sampling comparing data collected during the period 1995/96 and 1996/97. Note: A and B labels on the x-axis refer to the years 1995/96 and 1996/97 respectively.

19 Water Quality in the Snowy River Catchment Area, 199619 7 Report I I 72.3. Turbidity Note: there are no absolute ANZECC (1992) guidelines for turbidity for the protection of aquatic ecosystems. Instead, the DLWC uses the following as "rule of thumb" measures of turbidity (Bek I and Robinson 1991): 1 <5 NTU = Low turbidity S - 50 NTU = Moderate turbidity I >50 NTU = High turbidity

Summary statistics for routine samples of turbidity during the period 1996/97 are presented in I Appendix 3C. Variability in turbidity between each of the sites is presented in Figure 9 to Figure - 11.

During 1996/97, the majority of sites in the rivers and streams had low turbidity. The main exceptions were the Snowy River at Dalgety (Site 222006) and the Mowamba Aqueduct Outflow I (Site 22210142), where maximum turbidities reached 160 NTU and 75 NTU respectively. Median turbidity values were low for most sites, ranging from 0.7 NTU to 9 NTU. The only site where the I median turbidity for routine samples exceeded 5 NTU was Wollondibby Creek at Gunnadoo (Site 222544). The lowest median turbidity was recorded for Perisher Creek at Blue Cow

Turbidity was generally low at all three sites on Lake Jindabyne, and for all depths, on the few I occasions sampled. The highest recorded turbidity was only 6.2 NTU, measured at the bottom of Dam Station (Site 22210001).

I Comparing turbidity between 1995/96 and 1996/97, median turbidity values were generally similar for both years at most sites. However, sites at Thredbo River D/S Little Thredbo River (Site Li 22210011), the Thredbo River at Paddys Corner (Site 222541) and the Snowy River at Dalgety (Site 222006) had slightly lower median turbidity values in 1996/97 compared to 1995/96. I I I I I I I Water Quality in the Snowy River Catchn1ent Area, 1996/9 7 Report 20 I I

12 12 12 12 12 12 12 12 12 12 10 I I.,

I 0 100 0

I 0 0 0 I I I 0

222513 22210019 22210016 22210012 22210017 22210015 22210011 222541 222544 22210142 222006 Upper Middle Snowy i Lower I Snowy Snowy

Figure 9. Boxplot of turbidity data for river and streams sites, routine sampling during the I period July 1996 to June 1997. I 5 5 5 2 2 2 3 3 2 1000 I I Im I 0 I 0 ------ ------I EEE

0 ~ I I II I I I I F TOP MIDDLE BOTTOM TOP MIDDLE BOTTOM TOP MIDDLE BOTTOM Site 22210001 22210002 22210003

1 Figure 10. Boxplot of turbidity data for Lake Jindabyne sites, routine sampling during the period July 1996 to June 1997. I

I Water Quality in the Snowy River CatchmentArea, 1996/97 Report 21 1 - - - - - M------

n= 11 12 1212 1312 1212 1212 1212 1212 1212 1212 11 12 8 10 1000 /

LII

1 00 - LI

222513 22210019 22210016 22210012 22210017 22210015 22210011 222541 222544 22210142 222006 Upper -- -- Middle Snowy HILower Snowy Snowy

Figure 11. Boxplots of turbidity for river and stream sites, routine sampling comparing data collected during the periods 1995/96 and 1996/97.

Note: A and B labels on the x-axis refer to the years 1995/96 and 1996197 respectively.

22 Water Quality in the Snowy River Catchment Area, 1996/97 Report I 1 7.2.4. Additional data for Lake Jindabyne Lake Jindabyne is an impoundment on the Upper Snowy River. It receives inflows mainly from the Thredbo River; from the Mowamba River via the Mowamba Aqueduct; and from some smaller I tributaries. Most flow in the Snowy and Eucumbene Rivers, which would also enter Lake Jindabyne, are diverted by impoundments further upstream on these rivers, although spill from I Island Bend Dam occurs each spring. Thermal and dissolved oxygen measurements.

I Figure 12 shows thermal and dissolved oxygen profiles measured at Dam Station on Lake Jindabyne. Although additional profiles were measured at the two other lake sites (Jindabyne East I and Hiawatha), data for these have not been graphically produced, as there were only a few profiles collected at each location. I Collection of thermal and dissolved oxygen profile data for 1996/97 did not commence until January 1997, as the profiling instrument was not in working condition between July and I December 1996. In January, Lake Jindabyne was thermally stratified, with the water temperature ranging from 23°C at the surface to 11°C at the bottom at Dam Station. The range at Hiawatha was 20°C at the I surface to 11°C at the bottom. Dissolved oxygen concentrations at Dam Wall Station decreased from 7.4 mg/L (87% saturation) at the surface to 5.9 mg/L (53% saturation) at the bottom.

LI The lake remained stratified in March, with a temperature difference between the surface and bottom waters of 9°C at Dam Wall Station, and 8°C at both Jindabyne East and Hiawatha I Stations. The thermocline at Dam Wall Station commenced at a depth of around 11 metres below the surface. Dissolved oxygen concentrations at Dam Wall decreased from 7 mg/L (77% saturation) at the surface to about 4.8 mg//L (48%.saturation) at a depth of 13.7 metres. It then I increased to about 5.9 mg/L (55% saturation) until a depth of 22 metres, below which it decreased again to around 4.5 mg/L (41% saturation) at the bottom of the lake.

I The surface water temperature at Dam Wall Station had dropped to 15°C by April, while the bottom temperature remained at 11°C. Dissolved oxygen concentrations at the surface remained around 7.7 mg/L, (77% saturation), but concentrations at the bottom had decreased further I compared to March, to 2.78 mg/L (25% saturation). 1 Full circulation of Lake Jindabyne commenced in May, with the temperature down the entire water column varying only slightly, from 11.8°C to 11.0°C, at all sampling sites. Dissolved oxygen results for May were much different to those measured on earlier sampling occasions. Examination I of the DO membrane on the profiling instrument revealed that it was ruptured, resulting in unreliable dissolved oxygen measurements.

ii Lake Jindabyne was virtually isothermal in June. The water temperature in the surface and bottom waters at Dam Wall Station differed by only 0.1 T. Dissolved oxygen concentrations showed I slight variations with depths, and ranged between 7 and 6.5 mg/L (65% to 62% saturation).

I Water Quality in the Snowy River Catchment Area, 1996/97 Report 23 I I

I Temperature Dissolved Oxygen 22 January 1997 22 January 1997 0- 01 -10- I U) I -20- - —20 -30-0 —30 -40--40 —40 I I 50-50 —50 60 5 10 15 20 25 0 20 40 60 60 100 I Temperature (deg C) Dissoved Oxygen (15 Sat) 10 March 1997 10 March 1997 0- 0 I —10_10- 10 U) -20--20 - —20 E -30 —30 I —40 —50 —60L —6O-t 5 10 15 20 25 0 20 40 60 80 1OC I Temperature (deg C) Dissoved Oxygen (15 Sat) 22 April 1997 22 April 1997 I 0 —10 —10 - —20 - —20 —30 —30 —40 —40 I CL —50 —60 —60 I 5 10 15 20 25 0 20 40 60 80 100 Temperature (deg C) Dissoved Oxygen (15 Sat) 27 May 1997 27 May 1997 0-f I 0 1 —10 —10 1 —20 - —20 —30 —30 —40- I -40 0 U) I -50-50 —60 _60- I 5 10 15 20 25 -6000 20 40 60 80 100 Temperature (deg C) Dissoved Oxygen (15 Sat) 20 June 1997 20 June 1997 I 0 0 —10 U) -10 - —20 - -20 —3O —3O I -c - 40 0- — 0. -40 5O -50 —60 —60 I 5 10 15 20 25 0 20 40 60 80 100 Temperature (deg C) Dissoved Oxygen (15 Sat)

I Figure 12. Thermal and dissolved oxygen profiles for Dam Wall Station, Lake Jindabyne.

24 I Water Quality in the Snowy River Catchment Area, 1996/97 Report I I I Water Clarity Secchi disc depth measurements at Dam Wall Station on Lake Jindabyne ranged from a minimum of 2.7 metres (in January and March 1997), to a maximum of 4.9 metres (in May 1997). The maximum Secchi disc reading, 5.3 metres, was also recorded in May, at both Jindabyne East and Hiawatha Stations. The lowest Secchi disc measurement, 2.5 metres, occurred at Hiawatha in I January 1997. Electrical Conductivity

Note: ANZECC (1992) guidelines for the protection of aquatic ecosystems

I 0 No more than 1500 p.S/cm Electrical conductivity was very low at all stations, with a maximum value of 33 PS/cm observed I at Dam Wall Station in March 1997. The lowest electrical conductivity value was 20.3 PS/cm measured in January 1997, also at Dam Wall Station.

I Algae Results of algal analysis for all sites on Lake Jindabyne are presented in Table 2. The highest blue- I green algae count was recorded in April 1997, with 2,656 cells/mL of mostly Microc,vstis at Dam Wall Station. Low levels of blue-green algae, mainly Aphanocapsa, were also present at Dam Wall Station and at Hiawatha Station during January 1997. No blue-green algae were detected on I the two sampling occasions at Jindabyne East Station.

Low numbers of other algae were also generally recorded in Lake Jindabyne during 1996/97. The I highest counts were measured in March 1997, when the majority of these algae consisted of the green algal genus Dictyosphaerium (1,876 cells/mL at Dam Wall Station; 2,132 cells/mL at I Jindabyne East Station; and 782 cells/mL at Hiawatha Station). I Table 2: Total number of blue-green and other algae at Lake Jindabyne, 1996-97 data. Site Number Date Total Blue-Green Algae Total Other Algae H (cells/mL) (cells/mL) 22210001 22-Jan-97 650 238 10-Mar-97 0 2160 I 22-Apr-97 2656 925 27-May-97 0 409 20-Jun-97 0 1483 I 22210002 10-Mar-97 0 2245 27-May-97 0 506 22210003 22-Jan-97 497 316 I 10-Mar-97 0 895 I 27-May-97 0 493

I Water Quality in the Snowy River Catchment Area, 1996/97 Report 25 I I 1 7.3. Event Samples A high flow event is when flow in a stream increases markedly above normal base flows (flows that occur the majority of the time), and may occur because of storms and high runoff volumes I from within the catchment, or due to snow melting in spring. Event sampling was undertaken at all sites along the Thredbo and Little Thredbo Rivers from July 1995 to June 1997. Event samples were collected more frequently in 1996/97 compared to 1995/96, because more high flow events I occurred during 1996/97. The duration of high flows in the river ranged from a few days to as long as two months. Up to or more than 10 different storm events were recorded at most sites. In I particular, flow remained high during the months of September to November 1996, following an initial flush which occurred around the first week of September. Decline to base flow level only began in the third week of October. During this event period, more than 65 samples were taken at I some sites (Appendix 4A).

Other storm events occurred in July, August, and November 1996: and in March and May 1997, I but these prevailed for only a few days each.

Events samples generally had higher nutrient concentrations and turbidity compared to routine I samples, although samples varied considerably in terms of these parameters both within events and between events. Samples taken during storm events in March 1997 had higher nutrient I concentrations and turbidity than those taken during other high flow occasions, possibly due to an initial flushing effect following a long period of dry weather. The previous event sampling prior to March 1997 had occurred in November 1996. Turbidity measurements in March 1997 were up to I 68 NTU in the Little Thredbo River at Alpine Way (Site 22210017).

Nutrient concentrations during the prolonged high flow event from early September through to mid I October 1996 were, at most sites, considerably lower than those measured during other high flow periods. This may be due to previous flushing effects of earlier high flows, which would have I already removed stored nutrients from the catch.ment area. An example of flow and nutrient loads during a high flow event is shown in Figures 13 (for total phosphorus) and 14 (for total nitrogen). These data are for the event that occurred between I September and October 1996, with flow being shown by a solid line; and instantaneous nutrient loads (calculated by multiplying nutrient concentrations by instantaneous flow, defined as the flow I at the date and time of sampling) as dotted lines.

These data indicate that stream flow was the main parameter determining the instantaneous I nutrient load, although this was less marked at the two sites furthest upstream, at Deadhorse Gap (Site 22210019) and below the Thredbo sewage treatment works (Site 22210016). Loads generally tend to increase as flows increase, and decline when flow decreases, with maximum loads I occurring when flows peaked.

Summary statistics for total phosphorus loads calculated during the prolonged high flow event I from September and October 1996 (Appendix 4B) indicate increasing total phosphorus loads with distance downstream. Along the Thredbo River, the median phosphorus load was 2.7 kg/day at I Deadhorse Gap (Site 22210019), the most upstream site; 47 kg/day at Bundilla (Site 22210012) midway downstream; and 61 kg/day at Paddys Comer (Site 222541), furthest downstream. Of the I Water Quality in the Snowy River Catchment Area, 1996/97 Report 26 I I

Little Thredbo River sites, total phosphorus loads increased from 2.3 kg/day at the Alpine Way I (Site 22210017) to 3.3 kg/day at the Walking Track Bridge (Site 22210015) further downstream. Similar trends are apparent in minimum and maximum load data. The greatest range in total I phosphorus loads, calculated for Paddys Corner, was from 7 kg/day to 731 kg/day. Mean total nitrogen loads for the September/October 1996 high flow event also generally tended to increase with distance downstream along the Thredbo River, although the median value at Bundilla I was higher than the median values for sites further downstream. The maximum value was calculated for the Thredbo River at Paddys Comer, this being nearly 7,000 kg/day. There was very I little difference in total nitrogen loads between the two sites on the Little Thredbo River. I E I I I I I I I I L I I

27 I Water Quality in the Snowy River Catchment Area, 1996/97 Report I

22210015 - Lithe Thredbo River W Walking Track Bridge 22210019 - Thredbo River a! Deadhoroe Gap I 10000 10CC -1 10000 1000-1

I nov 'a 'N -J I All, '"f'

10 10 I 0100P96:00:00 I 500P9000:Oa 01001860000 1500196:00:00 01 eove&:oo:ao 01SEP96:0000 15S0P9600:00 01Ocl96i0000 150019800:00 01N0v96:00:00 Dateand time Date and time

22210016- Thredbo River 0/9 91W 22210011 - Thredbo River 0/S Lilile Thredbo I 1000-1 1- 10000 1000

118 JMJL 1000 I 1 'a AJJ ko~~'~' I

a-I F IC 01001ge0000 I 500196:00:00 UI e0v96:Oa:00 I Cl 00P95:00:00 1 500pge:ao:oo 0100166:00:00 I 500190:00:00 UI e0v960000 01 nreg6oa0o I 55019&a0:Ce Date and time Date and time

222541 - Thredbo Rivera! Poddys Corner 22210012 - Thredbo River a! Buridilla .uI I 1000 10000 1000 I :118 I 100

10 011 1- 10 0100096:00:00 0r0Ep96:00:00 15OrP9e00:00 v100198:00:00 1000096:00:00 O1II0056:00:00 0100P96:00:00 I 506196:00:00 CI CC1960000 150MUM0 I Date and time Date and time

22210017 - Threbro River a! Alpine Way I 1000-1 10000 118 1000 'a I 'N I ?Mk 100

0106P96:00:00 I 506196:00:00 010C196:00:00 1500196:00:00 DrNOv9e0000 I Date and time Figure 13. Total phosphorus load (in dotted lines) and hourly flow (in solid lines) during event I sampling of September 1996 to October 1996.

28 I Water Quality in the Snowy River Catchment Area, 1996/97 Report I I

22210019 - Thredbo River at Deodhorse Gap 22210015 - Lithe Thredbo River © Walking Track Bridge I 10000 It" 10000 -4 10000

1000 1000 >- 0 a I N N -J 100 -J IOU I IJ L I-I 1- 10 I, 1- 10 I 0100P96:00J00 1500P96:000 010CT96:0000 150C196:00:00 01NOv96;00:00 0100P96J0000 15S0P960000 ol000ee000a 15OCT96:0000 oieoveeoaoo Datetime Datetime

22210016 - Thredbo River 0/S STW 22210011 - Thredbo River D/S Little Thredbo I 10000-1 1- 10000 10000 10000

1000 I a I 100

k 10 I-t i 10 olorpgO0000 1501ree0000 01000960000 15001960000 01vov960000 oIsrPge:oo:oD 150EP9600:00 0I0CT9600:00 150019e:0000 oINovgooaav I Datetime Datetime 22210012 - Thredbo River at BLondillo 222541 - Thredbo River at Poddys Corner I 10000 10000 10000 10000 PM'

a N a I 00 100 z I l 10 0 1-1 I- 10 01 sEPeeaaao I 5SEP9600:D0 01000e60000 I 50C96O0UD 01 60v9600:00 UI OEP960fl0D 10sEP9600:00 DiacngeoUoa I 50C9600UD UI wove000ao I Datetime Datetime 22210017 - Threbro River at Alpine Way 10000-I 10000

I I1' 1000 1- N a a N I 100 100

I k in 0106P960000 150EP990000 01001960000 1000096:00100 01NO'/960000 I Datetime

I Figure 14. Total nitrogen load (in dotted lines) and hourly flow (in solid lines) during event sampling of September 1996 to October 1996.

29 I Water Quality in the Snowy River Catchment Area, 1996/97 Report I I 1 74. Annual Loads 7.41. Total phosphorus Loads I Total phosphorus loads in the Thredbo River increased with distance downstream (Table 3). Estimated total phosphorus loads in the Thredbo River between Deadhorse Gap (Site 22210019) and downstream of the Thredbo sewage treatment works (Site 22210016) increased by 3 tonnes. Between the site downstream of the Thredbo sewage treatment works (Site 22210016) and Bundilla (Site 22210012), the increase was 5.4 tonnes over the period 26 July 1995 and 30 June 1997. The small increase in load between the two sites on the Little Thredbo River was not I significant. Between Bundilla (Site 22210012) and Thredbo River downstream of the Little Thredbo River confluence (Site 22210011), the load difference was 4.1 tonnes of which 1 tonne of I the total phosphorus load originated from the Little Thredbo River.

Therewas very little difference in loads between Site 22210011 and Site 222541 (Paddys Corner). The overlapping confidence intervals for the load estimates at these sites indicate that the difference was not significant. The small difference could also be due to the frequent break down of the autosampling equipment at Paddys Corner, resulting in a number of high flow events not being sampled at Paddys Corner, whereas these were sampled at the sites upstream. (Phil Boreham, SMHEA, personal communication). This has ultimately led to much lower load estimates for this site than would otherwise have been the case. The load data for Paddys Corner should therefore be treated as indicative only, and is in no way comparable to the loads estimated ' for the other six sites further upstream. The area of the catchment between the Little Thredbo River confluence and Paddys Corner comprises 33% of the total Thredbo River catchment, and includes the Sawpit Creek subcatchment and the Gaden Trout Hatchery. Given this, a much greater input of total phosphorus load to the Thredbo River should be expected from this section of the catchment than is indicated by the results of this study.

There was little increase in total phosphorus load between the two sites on the Little Thredbo I River, at the Alpine Way (Site 22210017) and at the Walking Track Bridge (Site 22210015). Total phosphorus load increased by only 0.2 tonnes over the sampling period. With total I phosphorus loads of only about 1 tonne, the Little Thredbo River is a minor contributor to the large increase in load in the Thredbo River downstream of its confluence with the Little Thredbo River.

The calculated average annual loadings for each site along the Thredbo River catchment are shown in Figure 15.

Table 4 shows the contribution to the total phosphorus load from various sections of the catchment. These were measured by subtracting the load calculated for the site(s) immediately I upstream from the next site downstream along the Thredbo River. Much of the total phosphorus load in the Thredbo River originated from the section of the catchment between Deadhorse Gap (Site 22210016) and below the Thredbo sewage treatment works (Site 22210016); the section I between this site and the next site downstream at Bundilla (Site 2221012); and between Bundilla and the site just downstream of the junction of the Little Thredbo River (Site 22210011), but I excluding inputs from the Little Thredbo. The average annual total phosphorus loads exported from these three sections of the catchment were 1500 kg, 2800 kg, and 1700 kg respectively. These

I Water Quality in the Snowy River Catchment Area, 1996197 Report 30

I I

three sections make up 15%, 29%, and just 1.5% of the total area of the entire Thredbo River I catchment, respectively.

I Table 3 Estimates of total phosphorus loads from 26 July 1995 to 30 June 1997. Loads are the bootstrap means

I Total Loads Standard Error 95% Confidence interval Daily Load (tonnes) (kg/day)

I 22210019 1.2 0.129 0.9 - 1.4 1.7 22210016 4.2 0.297 3.6 - 4.8 5.9 I 22210012 9.6 0.679 8.4 - 11.1 13.6 22210017 0.75 0.054 0.7 - 0.9 1.1 22210015 0.9 0.056 0.8 - 1.0 1.3 I 22210011 13,7 0.945 11.6 - 15.4 19,4 222541 14.5 1.100 12.7 - 16.9 20.5 I I Table 4 Areal load calculations for total phosphorus.

I Mean Mean Annual Annual Load per Areal Load I Section of Catchment Area Load estimated as: Section of per Section (ha) Catchment of (kg) Catchment I (kg/halyear) Thredbo River Upstream of 1980 Load at 22210019 620 0.31 I Deadhorse Gap Thredbo River Deadhorse 3615 Load at 22210016 minus load at 1550 0.43 Gap to Downstream of 22210019 I Thredbo Thredbo River Downstream 7035 Load at 22210012 minus load at 2800 0.40 of Thredbo STW to Bundilla 22210016 I Little Thredbo River 2785 Load at 22210017 400 0.14 Upstream of Alpine Way Little Thredbo River Lake 665 Load at 22210015 minus load at 80 0.12 I Crackenback area 22210017 Thredbo River Ski Tube Area 245 Load at 22210011 minus loads at 1650 6.73 22210012 and 22210015 I Thredbo River Downstream 7890 Load at 222541 minus load at 410 0.05 Ski Tube area to Paddys 22210011 I Corner 31 I Water Quality in the Snowy River Catchment Area, 1996197 Report 1 I

I The annual areal export total phosphorus loads for each section of the Thredbo River catchment were calculated by dividing the average annual load exported from each section by the area of that I section (Table 4). The lowest average areal exports occurred from the two sections of catchment along the Little Thredbo River, including not only that part of the catchment upstream of the Alpine Way, but also that part containing the Lake Crackenback resort. Areal loads for both these I sections of the catchment were around 0.14 and 0.12 kg/halyr, respectively. (A smaller areal export of 0.05 kglhalyear was calculated for the section of the catchment between the site just I downstream of the Little Thredbo River junction and Paddys Corner, but this is suspect due to event sampling difficulties encountered at Paddys Corner - see above.) I Areal loads of total phosphorus from the Thredbo River catchment upstream of Bundilla were generally higher than those for the Little Thredbo sub-catchment. Those calculated for the section of catchment between Thredbo and Bundilla and the section centred on Thredbo Village and the ski I fields were slightly higher (0.40 and 0.43 kglhalyr, respectively) compared to the section upstream of Deadhorse Gap (0.31 kg/halyr). By comparison, extremely high areal total phosphorus loadings (6.73 kg/halyr) were recorded from the small section of the catchment surrounding Bullock Flat [1 and the Ski Tube.

LI Contributions to the total phosphorus load in the Thredbo River at the site just below the Little Thredbo River confluence (Site 22210011) from the various sections of the catchment upstream are summarised in Table 7. (Calculations of the contributions to the load at Paddys Corner further I downstream could not be made, due to the unreliability of the load estimates at this site arising from the faulty autosampling equipment located there. For a fuller explanation, see above) The small area around Bullocks Flat and the Ski tube, which comprises just 1.5% of the total I catchment area upstream of Site 22210011, contributed 24% of the total phosphorus load in the Thredbo River at this location. The area that includes Thredbo Village and the ski fields, comprising 22% of the total area upstream of Site 22210011, contributed 21% of the phosphorus load at that site, while the area between the sites downstream of the Thredbo sewage treatment works and Bundilla contributed 39% of this load. This large contribution is anticipated from this I area as it constitutes 43% of the total catchment area upstream of Site 22210011. Contributions from the section of catchment upstream of Deadhorse Gap and from the entire Little Thredbo I River sub-catchment were small, at 8.7% and 6.7% respectively. I I I I [1 32 Water Quality in the Snowy River CatchmentArea, 1996/97Report I I 7.4.2. Total nitrogen Loads

I Estimates of total nitrogen load at each of the sites within the Thredbo River subcatchment are shown in Table 5. These loads, for the period covering 25 July 1995 and 30 June 1996, show a progressive increase with distance downstream along the Thredbo River until the site just I downstream of the confluence with the Little Thredbo River. The estimated load between Deadhorse Gap and just downstream of the Thredbo STW increased by 42 tonnes, and there was I also a marked increase in load of 69 tonnes between the site downstream of STW and Bundilla. Similarly an increase of 52 tonnes occurred over the section of river between Bundilla and downstream of the Little Thredbo River junction, which included a contribution of 11 tonnes I from the Little Thredbo River. Loads in the Little Thredbo River between the Alpine Way and the Walking Track Bridge sites increased by only 2 tonnes. These differences were all significant, as indicated by the clear separation of all the confidence intervals. For average annual total I nitrogen loadings for each site, see Figure 15.

The total nitrogen load results also show an apparent marked decrease in the Thredbo River I between the site immediately below the Little Thredbo inflow (Site 22210011) and the next site downstream, Paddys Corner (Site 222541). This was also due to the frequent break down of the I autosampling equipment at Paddys Corner, so that a number of high flow events were not sampled (Phil Boreham, SMHEA, personal communication) and included in the load calculations. As with the total phosphorus load data for Paddys Corner (see Section 7.4.1, above), I this has led to much lower total nitrogen load calculations than would have otherwise resulted had all the events actually been sampled. Because of this inaccuracy, the total nitrogen load data for Paddys Corner is therefore only indicative, and again not comparable with the loads I calculated for six upstream sites, where all events were recorded and sampled.

Table 5. Estimates of total nitrogen loads for 26 July 1995 to 30 June 1997. Loads are the bootstrap means

I Total Load Standard Error 95% Confidence Limits Daily Load (tonnes) (kg/day)

I 22210019 32 3.09 26.5 - 38.9 45 22210016 74 5.80 63.5 - 85.9 105 1 22210012 143 10.00 127 - 169 203 1 22210017 9 0.52 8.1 - 10.2 13 22210015 11 0.59 10.2 - 12.7 16 I 22210011 195 16.81 167 - 235 278 I 222541 160 9.31 141 - 178 224 I I I

33 Water Quality in the Snowy River Catchinent Area, 1996197 Report I D The highest annual load (Table 6) entering the Thredbo River came from the section between the sites below the Thredbo sewage treatment works (Site 22210016) and Bundilla, of almost 36 000 I kg. This section of catchment comprises 43% of the total catchment area upstream of the Little Thredbo River confluence. The lowest nitrogen loadings came from a small section of catchment I (4.1 % of the total) along the Little Thredbo River between Alpine Way and Walking Track Bridge, which includes the Lake Crackenback resort. The area of each section of catchment is I obviously an important factor in determining the amount of load that it contributes to the river. Annual areal total nitrogen loadings (Table 6) show a similar pattern to those for annual areal total phosphorus loads. Once again, both Little Thredbo River subcatchment areas, including the I section containing Lake Crackenback resort, had lower annual areal total nitrogen loads compared to the rest of the Thredbo River catchment. However, the section of the Thredbo River catchment centred on Thredbo Village had a slightly lower annual areal total nitrogen loading I compared to the section upstream of Deadhorse Gap, while the average annual areal load from the section of catchment between Thredbo and Bundilla had slightly lower areal load coefficients I than the two sections further upstream. Once again, the section around Bullocks Flat and the Ski tube had markedly higher areal total nitrogen loadings than anywhere else in the catchment.

I Table 6. Areal loading calculations for total nitrogen.

Mean Annual Mean Annual I Load per Section Areal Load Section of Catchment Area Load estimated as: of Catchment per Section of (ha) (kg) Catchment I (kg/halyear)

Thredbo River Upstream of 1980 Load at 22210019 16550 8.4 I Deadhorse Gap Thredbo River Deadhorse 3615 Load at 22210016 minus load 21 700 6.0 Gap to Downstream of at 22210019 I Thredbo Thredbo River Downstream 7035 Load at 22210012 minus load 35670 5.1 of Thredbo to Bundilla at 22210016 I Little Thredbo River 2785 Load at 22210017 4650 1.7 Upstream of Alpine Way ILittle Thredbo River Lake 665 Load at 22210015 minus load 1030 1.5 Crackenback area at 22210017 Thredbo River Ski Tube 245 Load at 22210011 minus 21 200 86.5 Area loads at 22210012 and I 22210015 Thredbo River Downstream 7890 Load at 222541 minus load at -18 100 -2.3 Ski Tube area to Paddys 22210011 I Corner

34 Water Quality in the Snowy River Catchment Area, 1996/97 Report I I Percent contributions from different parts of catchment to the total nitrogen load estimated at Site I 22210011 (the Thredbo River just downstream of the Little Thredbo River junction) are given in Table 7. The area containing Thredbo Village and the ski fields contributed around 21.5% of the total nitrogen load over the two years. Contributions from the area between Thredbo Village and I Bundilla were also high, when inputs from this area accounted for 35% of the load at Site 22210011. Once again a large contribution from this part of the catchment is to be expected, I considering that it comprises 43% of the total catchment area upstream of Site 33310011. The area upstream of Deadhorse Gap contributed around 16.5% of the total nitrogen load, while the combined contribution from the two sections of catchment draining to the Little Thredbo River I was 5.6%. Around 21% of the nitrogen load comes from the Bullocks Flat and Ski tube area, which is quite substantial seeing this area makes up only 1.5% of the entire catchment.

I % of the Total Total Phosphorus At Total nitrogen At Thredbo Catchment Area Thredbo River D/S Little River D/S Little Thredbo Section of Catchment upstream of the Little Thredbo Confluence Confluence Thredbo River I confluence I Upstream of Deadhorse Gap 12.1 8.7 16.4 Deadhorse Gap to D/S 22.1 21,1 21.5 Thredbo I D/S Thredbo to Bundilla 43.1 39.4 35.4 Little Thredbo U/S Alpine 17.1 5.6 4.6 Way I Little Thredbo around Lake 4.1 1.1 1.0 Crackenback Bullocks FlatJSki tube 1.5 23.9 21.0 LI D/S Bullocks FlatJSki tube to * * * Paddys Corner

I * = Not calculated due to unreliable results caused by autosampler failures at Paddys Corner. I I d I 35 LI Water Quality in the Snowy River Catchment Area, 1996197 Report I I

I -71 I

I 2225 I II 115 /\ I 480 56th) N

222 17 I 21001 QD I 73820 -

1001 LEGEND 1 17 Staton 38250 TI Phosoms Load Total N0on Load

I 119 I 18050 I Figure 15 Map showing the annual total phosphorus and total nitrogen loads for each site in the I Thredbo River catcbment. I I I I I I 36 I Water Quality in the Snowy River Catchment Area, 1996197 Report I 7.5. Trend Assessment 7.5.1. Total phosphorus

I The Seasonal Kendall test for trend assessment performed on the raw data detected no trend in total phosphorus at any of the sites along the Thredbo River, or at the major inflow sites to Lake Jindabyne, for the period from 1992 to 1997. (Table 8). However, the results were not significant I because of their high p-values. Sites at Perisher Creek at Blue Cow (Site 222513) and the Snowy River at Dalgety (Site 222006) displayed a decreasing trend in total phosphorus, but once again Ii the results were not significant because of high p-values.

Some trends were apparent in the total phosphorus data once these were flow adjusted, though the I trends were again not significant due to high probability values (Table 8). A decreasing trend was indicated at Paddys Corner (Site 222541), while Wollondibby Creek at Gunnadoo (Site 222544) and the Mowamba Aqueduct at the Lake Jindabyne outfall (Site 22210142) both exhibited H increasing trends.

LOESS plots of the raw total phosphorus data are shown in Figure 16. A constant total I phosphorus concentration was apparent at most of the sites from 1992 to about the middle of 1994. Sites on Perisher Creek at Blue Cow (Site 222513), the Thredbo River downstream of I Thredbo sewage treatment works (Site 22210016), at Bundilla (Site 22210012) and at Paddys Corner (Site 222541), and the Little Thredbo River at Alpine Way (Site 22210017) all showed slight increases in total phosphorus concentration starting around mid- 1994. There is no obvious I explanation for these slight increases. However, there were no obvious changes to total phosphorus trend at these sites after this time.

I LOESS curves for flow adjusted data, shown in Figure 17, indicate that total phosphorus concentrations have been fairly constant at the three sites over the period of study. I Table 8. Seasonal Kendall test results for log total phosphorus performed on raw and flow I adjusted data. Site Number Data No. of SK test Probability SK Annual trend Confidence Interval of Type obs statistics Slope (ugfL) annual trend (ugIL) I lower limit upper limit - Site 222513 41 -0.71375 0.48 -0.07 -1.07 -1.26 2.96 22210019 51 0.29047 0.77 0. 0 -0.16 2.25 22210016 48 0.70952 0.48 0 0 0 2.25 I 22210012 54 1.04053 0.30 0 0 0 2.54 22210017 46 0.21865 0.83 0 0 -0.90 2.89 22210015 47 -0.09901 0.92 0 0 -2.68 3.66 22210011 53 1.09393 0.27 0 0 0 3.74 I 222541 52 0.44662 0.66 0.00 0 -1.59 2.79 FA 52 -0.16954 0.87 -0.03 -0.27 -1.74 1.41 222544 54 0.58065 0.56 0 0 -3.04 4.86 I FA 54 0.64258 0.52 0.03 0.94 -2.08 5.03 22210142 51 0.17655 0.86 0 0 -2.20 4.75 FA 50 0.78534 0.43 0.02 0.61 0.61 4.06 I 222006 31 -0.84921 0.40 -0.04 -0.73 -4.78 3.28 Note: blank entry under data type refers to raw data, FA means flow adjusted data. I Water Quality in the Snowy River Catchment Area, 1996/97 Report 37 I I 222513 - Perisher Creek at Blue Cow 22210011 - Thredbo River D/S Little Thredbo

1000 1000

100 100

.c 10 10 I o 01JUL95 01JUL96 01JUL97 01JUL92 01JUL93 01JU04 01JU05 01JUL96 01JUL97 01JUL92 01JUL93 01JUL94

I 22210019 - Thredbo River at Deadhorse Cap 222541 - Thredbo River at Paddys Corner 1000 1000 I 100 100 , 10

01JUL97 I 01JUL92 01JUL93 01JUL94 01JUL95 01JUL96 01JUL97 01JUL92 01JUL93 01JUL94 01JUL95 01JUL96

22210016 - Thredbo River D/S STW 222544 - Wollondibby Creek at Cunnadoo I 1300 1000

100 100

[1 10 .2 10 o Li 01JUL92 01JUL93 01JUL94 01JUL95 01JUL96 01JUL97 01JUL92 01JUL93 01JUL94 01JUL95 01JUL96 01JUL97 22210012 - Thredbo River at Bundilla 22210142 - Lake Jindabyne Mowamba Aqueduct Outf U 1000 1000 100 100

10 .2 io

I o

0IJUL94 01JUL95 01JUL96 01JUL97 01JUL92 01JUL93 01JUL94 01JUL95 01JUL96 01JUL97 I 01JUL92 01JUL93 22210017 - Threbro River at Alpine Way 222006 - Snowy River at Dalgety

1000 1000

Li 100 100

. 10 . 10 I o 01JUL97 01JUL92 01JUL93 01JUL94 01JUL95 01JUL96 01JUL97 01JUL92 01JUL93 01JUL94 01JUL95 01JUL96 I 22210015 - Little Thredbo River @ Walking Track Brid 1000 1 100 .2 ic I 01JUL92 01JUL93 01JUL94 01JUL95 01JUL96 01JUL97 Figure 16. LOESS trend curve fitted to raw total phosphorus data, 1992-1997.

38 Water Qualily in the Snowy River Catchment Area, 1996/97 Report I 1 222541 - Thredbo River at Paddys Corner

0 I -'-J 4 0)

0 2 I 0 0 0 0 0 -c 1 0 0 00 a) 0 0 0 0 I 0 I0 00 -o 000

0 I u —2 I 01JUL92 01JUL93 01JUL94 01JUL95 01JUL96 01JUL97 222544 - Wollondibby Creek at Gunnadoo I ,—'-J 4 0'

2 I 3- I- 0 0 -o 1 0 a) 0 0 0 0 0 0 0 0 00 0 C,, 00 I 0 0 0 0 0 0 0 0 0 0 0 -o 0 0 0 0 I 0 —2 I 01JUL92 01JUL93 01JUL94 01JUL95 01JUL96 01JUL97 22210142 - Lake Jindabye Mowamba Aqueduct Outflow I ,-.-'-J 4

2 I a- H- -o 1 a) 0 0 U, 0 I 0 o C0p o0 00000 0 0 0 0 0 —1 0 0 0 I u —2- 1 I I I 01JUL92 01JUL93 01JUL94 01JUL95 01JUL96 01JUL97 I Figure 17. LOESS trend curve fitted to flow adjusted total phosphorus data, 1992-1997. I Water Quality in the Snowy River Catchment Area, 1996/97 Report I I

I 7.5.2. Total nitrogen Raw total nitrogen concentration data showed significant decreases at two sites over the 5-year period from 1992 to 1997 (Table 9). Even so, these decreases were small. Additionally, the I confidence intervals associated with these trends were wide, indicating a low precision in the estimates of the trend slope. Total nitrogen concentrations decreased by about 9%, or 14 .tg/L per I year in the Little Thredbo River at Walking Track Bridge (Site 22210015) (p0.06). The other site which exhibited a decreasing trend, although with a much lower level of significance (p=0.09), was the Snowy River at Dalgety. This represented a decrease of 32 .tgfL at this site during the period I 1992 to 1997.

All other sites in the Snowy River catchment area displayed no significant trends over this period I for either raw or flow adjusted data.

The time series plots of total nitrogen using LOESS curves showed constant concentrations for [1 most of the sites (Figure 18). A slight increase in concentrations is, however, apparent towards the end of 1994 for Perisher Creek at Blue Cow (Site 222513), though this could be due to data gaps I as few samples were taken between July 1993 and July 1994. These gaps have limited the ability of LOESS to characterise the general pattern of data before 1994. I LOESS curves for flow adjusted total nitrogen data show relatively stable concentrations during the 5-year period for the three sites where this data analysis was possible (Figure 19).

Li Table 9. Seasonal Kendall results for log total nNitrogen performed on raw and flow adjusted data.

I Site Number Data No. of SK test Probability SK Annual trend Confidence Interval of Type obs statistics Slope (ugfL) annual trend (ugfL) I lower limit upper limit 222513 41 -0.99741 0.32 -0.02 -4.74 -29.40 4.04 22210019 51 0.29802 0.77 0 0 0 8.65 22210016 48 -0.38913 0.70 0 0 -23.70 21.68 I 22210012 54 0 1.00 0 0 0 0 22210017 46 -0.90007 0.37 0 0 -20.12 0 22210015 47 -1.8937 0.06 -0.10 -13.73 -23.83 0 I 22210011 53 0.80606 0.42 0 0 0 22.51 222541 52 1.16431 0.24 0.07 11.52 0 33.76 FA 52 1.01783 0.31 0.03 5.19 -9.58 31.03 222544 54 0.92094 0.36 0.04 19.89 -22.96 56.47 FA 54 0.80322 0.42 0.03 14.20 -23.43 41.10 22210142 51 0 1.00 0 0 -44.20 68.19 FA 50 0.26178 0.79 0.03 8.86 -31.08 39.92 i 222006 31 -1.67444 0.09 -0.14 -31.57 -67.74 0

Note: blank entry under data type refers to raw data, FA means flow adjusted data. I I 40 Water Quality in the Snowy River Catchment Area, 1996197 Report I

I 222513 - Perisher Creek at Blue Cow 22210011 - Thredbo River D/S Little Thredbo

10000 10000

I 1000 1000

5 100 100

o Is 10 Li 01JUL92 01JUL93 01JUL94 01JUL95 01JUL96 01JUL97 01JUL92 0IJIJJ93 01JUL94 01JUL95 01JU06 01JUL97

22210019 Thredbo River at Deadhorse Gap 222541 - Thredbo River at Paddys Corner

I 10000 10000

1000 1000

5 ••' . ,,,---8 • I 100 100 C ... ' S

10 10 I 01JUL92 01JUL93 01JUL94 01JUL95 01JUL96 01JUL67 01JUL92 01JUL93 01JUL94 01JUL95 01JUL96 01JUL37

22210016 - Thredbo River D/S STW 222544 - Wollondibby Creek at Cunnadoo I 10000 10000 • 8 003 I 5 100 5 100 10 10 01JU04 01JUL95 01JUL96 01JUL37 I 01JUL92 01JUL90 01JUL94 01JUL95 01JUL96 01JUL97 01JUL92 01JUL93 22210012 - Thredbo River at Bundilla 22210142 - Lake Jindabyne Mowamba Aqueduct Outt

1000 10000

I 100 1000

10 5 100

10 I 01J6L35 01JUL96 01JUL97 01JUL92 01JUL93 01JUL94 01JUL95 01JUL96 01JUL97 01JU02 01JUL93 01JUL94

22210017 - Threbro River at Alpine Way 222006 - Snowy River at Dalgety

10000 10000

1000 1050

I 100 1 100 5

10 10 I 01JUL92 01JUL95 01JU04 01JUL95 01J6L96 01JUL67 01JUL92 01JUL93 01JUL34 01JUL95 01JUL96 01JUL97

22210015 - Little Thredbo River © Walking Track Bridg I 10005 1000 I 5 100 10 01JUL92 01JUL93 01JUL94 01JUL95 01JUL96 01JUL97

Figure 18. LOESS trend curve fitted to raw total nitrogen data, 1992-1997.

L 41 Water Quality in the Snowy River Catchment Area, 1996197 Report I I I 222541 - Thredbo River at Paddys Corner -'-J 4 1 0 O)

0 I z 2 F- 0 0 -o 1 0 0 U) 0 0 0 0 0 0 F') 0 0 0 3 0 I 0 0 00 0 0 00 0 0 00 0 0 0 -o 00 0

0 0 0 0 1 i —2 I I I 01JUL92 01JUL93 01JUL94 01JUL95 01JUL96 01JUL97 222544 - Wollondibby Creek at Gunnadoo I -.'-J 4 0'

1 z 2 H- - 1 0 U) 0 0 F') 0 I 0000 0 0 2 0 0 00000 0 0

0 I L —2 I 01JUL92 01JUL93 01JUL94 01JUL95 01JUL96 01JUL97 22210142 - Lake Jindabye Mowamba Aqueduct Outflow I 2 4

I z 2- F- -o 1- 00 0 0 0 0 -2 0 0 0 0 0 0 0 0 0 I .0 00 00000000 00 -0 00 0 0 0 0 I —2- I I I 01JUL92 01JUL93 01JUL94 01JUL95 01JUL96 01JUL97 I Figure 19. LOESS trend curve fitted to flow adjusted total nitrogen data, 1992-1997. I 42 Water Quality in the Snowy River Carchment Area, 1996/97 Report I P,

7.5.3. Turbidity

Significant increasing trends were detected in raw turbidity data from most sites within the I Thredbo River subcatchment, with the exceptions being the Little Thredbo River at Walking Track Bridge (Site 22210015) and the Thredbo River at Paddys Corner (Site 222541). Annual increases in turbidity were between 22% and 32%. The upstream site on the Little Thredbo River at Alpine I Way (Site 22210017) showed the highest percentage increase per year over the period from 1992 to 1997, with turbidity increasing by 0.4 NTU per year from the 5-year median turbidity value of 1.1 NTU. Along the Thredbo River, the site downstream of Thredbo sewage treatment works (Site I 22210016) indicated the highest rate turbidity increase, of about 0.24 NTU per year. The ranges in the confidence interval values for these sites were, however, large, so that the annual trend I estimates were not significant (Table 10). In reality, the streams of the Thredbo River subcatcbment have very low turbidity, and although increases of between 0.1 NTU and 0.4 NTU per year at most sites may appear large in relation to these, and result in the significant results I indicated by the trend analyses, such increases are in fact only tiny in proportion to the range of turbidity possible in surface freshwater.

I Although Paddys Corner (Site 222541) revealed no significant trends in turbidity, the two other sites on major inflows to Lake Jindabyne, Wollondibby Creek at Gunnadoo (Site 222544) and Mowamba Aqueduct Outflow (Site 22210142) showed significant upwards trends in raw data for I turbidity, at 0.07 and 0.06 probability levels respectively. The probability level increased to less than the 0.05 level once the turbidity data was flow adjusted.

LOESS curves (Figure 20) for raw turbidity data for most sites showed constant to slightly decreasing trend lines until 1994. From then on an increasing trend in turbidity was observed. I LOBS S plots of flow adjusted data show increasing trends in turbidity, beginning in 1994 (Figure 21).

I Table 10. Seasonal Kendall results for log turbidity perfonned on raw and flow adjusted data.

Site Number Data No. of SK test Probability SK Annual trend Confidence Interval of annual I Type obs statistics Slope (ugfL) trend (ug/L) lower limit upper limit

I 222513 40 2.5281 0.01 't'O.13 0.09 0.03 0.33 22210019 50 2.85378 0.00 't'0.23 0.18 0.06 0.40 22210016 47 2.52142 0.01 4'0.28 0.24 0.03 0.52 22210012 54 2.33604 0.02 't'0.20 0.12 0 0.21 I 22210017 46 2.52538 0.01 t'0.27 0.35 0.08 0.86 22210015 46 1.43894 0.15 0.15 0.19 -0.07 0.83 22210011 53 2.54818 0.01 'f.O.14 0.10 0.03 0.30 I 222541 53 1.50876 0.13 0.16 0.16 -0.03 0.48 FA 53 1.1547 0.25 0.13 0.13 -0.11 0.40 222544 55 1.80705 0.07 0.27 1.68 -0.11 3.26 FA 55 2.34978 0.02 4,0.20 1.18 0.40 2.20 I 22210142 49 1.89095 0.06 0.17 0.79 -0.03 2.36 FA 48 2.75783 0.00 't'0.21 1.00 0.21 2.03 I 222006 30 0.17496 0.86 0.07 0.15 -0.27 3.64 Note: blank entry under data type refers to raw data, FA means flow adjusted data. '1' - increasing trend at 0.05 probability level.

43 Water Quality in the Snowy River Catchment Area, 1996/97 Report I

I 2221 0011 - Thredbo River 0/S Little Thredbo 222513 - Perisher Creek at Blue Cow

103 lao

0 I 10.0 10.0

1.0 1 1 .0

0.1 0.1 I 01JUL93 01JU03 01JUL94 01JU05 01JUL96 01JUL97 01JUL92 01JU03 01JUL94 01JUL95 01JUL96 01JUL97

22210019 - Thredbo River at Deadhorse Gap 222541 - Thredbo River at Paddys Corner I 100 100

0 I 10.0 10.0 1.0

0.1 0.1 01JUL92 01JUL93 01JUL94 01JUL95 01JUL96 01JUL97 I 01JUL92 01JUL93 01JUL94 01JUL96 01JUL96 01JUL97

22210016 - Thredbo River 0/S STW 222544 - Wollondibby Creek at Gunnadoo I 100 IOU

10.0 10.0

1.0 1 1.0 0.1 0.1 01JUL92 01JUL93 01JUL94 01JUL95 01JUL96 01JUL97 I 01JUL92 01JUL93 01JUL94 01JUL95 01JUL96 01JUL97 22210012 - Thredbo River at Bundilla 22210142 - Mowomba Aqueduct Outflow

103 IOU

0 I 10.0 10.0

1.0

0.1 I 0.1 01JUL92 01JUL93 01JUL94 01J6L95 01JUL96 01JUL97 01JU02 01JUL93 01JUL94 01JUL95 01JUL96 01JUL97

I 22210017- Threbro River at Alpine Way 222006 - Snowy River at Dalgety 100 100

9 I 10.0 10.0 1.0

0.1 0.1 01JUL92 01JUL93 01JUL94 01JUL95 01JUL96 01JUL97 I 01JUL92 01JUL93 01JUL94 01JUL95 01JUL96 01JUL97

10.0

0.1 01JUL92 01JUL93 01JUL94 01JUL95 01JUL96 01JUL97

I Figure 20. LOESS trend curve fitted to raw turbidity data, 1992-1997. Turbidity values above 100 NTU was not included in the graph.

1 44 Water Quality in the Snowy River CatchmentArea, 1996/97 Report I I 222541 - Thredbo River at Paddys Corner I 3 0

0 0 0 0 0 0 0 0 0 0 0 0 0 0 I 0 0

0 0o 0 0,0 0 0 00 0 0 0 I 0 0 0 0

0 I U- I I I 01JUL97 I 01JUL92 01JUL93 01JUL94 01JUL95 01JUL96 I 222544 - Wollondibby Creek at Gunnadoo 3-1 I I I

I 01JUL92 01JUL93 01JUL94 01JUL95 01JUL96 01JUL97 22210142 - Lake Jindabye Mowamba Aqueduct Outflow I 3 I I I I H I 01JUL92 01JUL93 01JUL94 01JUL95 01JUL96 01JUL97 I Figure 21. LOESS trend curve fitted to flow adjusted turbidity data, 1992-1997. I 45 Water Quality in the Snowy River Catchment Area, 1996/97 Report I I

8. DISCUSSION

Water quality throughout the Snowy River catchment area is generally good. Perisher Creek at Blue Cow (Site 222513) in the upper Snowy Catchment had mostly low nutrient concentrations and turbidity, while the Thredbo River subcatchment displayed similar water quality. There was some variation in water quality between the three major inflow sites to Lake Jindabyne. The Thredbo River at Paddys Corner (Site 222541) had the lowest nutrient concentrations and turbidity levels of these three sites, similar to other locations within the Thredbo River I subcatchment. Wollondibby Creek at Gunnadoo(Site 222544) and the Mowamba Aqueduct inflow at Lake Jindabyne (Site 22210142) had considerably higher total phosphorus concentrations, these being in the upper mesotrophic range, and had higher total nitrogen concentrations and turbidity as well. Lake Jindabyne itself had low nutrient concentrations and low turbidity. Nutrient concentrations in the Snowy River downstream of Lake Jindabyne, at Dalgety (Site 222006), were usually low, although a very high total phosphorus concentration of 115 tg/L was measured on one occasion. Turbidity was generally low most of the time. However, all sites were within the respective ANZECC (1992) guidelines for the protection of aquatic ecosystems for total I phosphorus and total nitrogen, while only Wollondibby Creek at Gunnadoo had a median turbidity value greater than 5 NTU. I The data indicated few differences in nutrient levels for most of the river and stream sites between years 1995/96 and 1996/97. For 1995/96 and 1996/97, median nutrient concentrations remained almost the same. However, one notable difference in nutrient concentrations was observed at I Snowy River at Dalgety, where these, and also turbidity, were much lower in 1996/97 compared to 1995/96. Additionally, the median total phosphorus concentration for the Thredbo River at I Deadhorse Gap (Site 22210019) was higher in 1996/97 compared to previous year. Nutrient concentrations measured at the sites on Lake Jindabyne fell within the low to moderate range, while turbidity was low. The reservoir meets ANZECC (1992) guidelines for the protection I of freshwater ecosystems. The low nutrient concentrations, especially of total nitrogen, may be the reason why very low algal densities have occurred in Lake Jindabyne over the past years. Nutrient I concentrations and turbidity levels at the lake sites were slightly lower in 1996/97 than in 1995/96. However, the water quality of Lake Jindabyne remains similar to that measured in the 1980s and early 1990s by SMHEA (1988) and joint SMHEA/DLWC studies (Bowling 1993, Bowling et al 1993, Bowling and Acaba 1995, Kinross and Acaba 1996, Maim et a! 1997). Slight annual variations do occur, as must be expected in natural systems, especially in minimum dissolved oxygen concentrations in the deeper parts of the reservoir. In some years, these fall to only 10-20% saturation, while in other years the minimum fall is to 3 0-40% saturation.

Lake Jindabyne stratifies thermally each summer. The lake was stratified in January 1997, and this LI persisted until late April 1997. Thermal stratification can cause the bottom waters to become isolated from atmospheric replenishment of oxygen, and microbial decomposition of organic I material within these waters will then reduce the concentration of dissolved oxygen. The lowest dissolved oxygen (DO) concentrations were measured at Dam Station on 22 April 1997, when dissolved oxygen was down to 25% saturation. Since the profile data collected in subsequent I months was not reliable due to instrument failure, there is no means of determining whether I dissolved oxygen concentrations fell lower than this. A dissolved oxygen saturation level of 25% is 46 Water Quality in the Snowy River Catchnient Area, 1996/97 Report I I considered low, but this value is within the range of minimum dissolved oxygen concentrations recorded at Dam Station during thermal stratification from 1989 to 1996. These ranged from 11% to 4 1 % saturation. High algal productivity is a potential major source of organic materials which I leads to oxygen depletion in bottom waters. However, algal count data from Lake Jindabyne in 1996/97 indicate little algal presence to cause oxygen depletion at depth. Other possible sources of organic input may be from the catchment, or from macrophyte plant growth which occurs at times I in the shallower waters of reservoir.

Besides nutrient and profile data, other water quality indicators were also investigated in Lake I Jindabyne. Water clarity remained the same in 1996/97 as 1995/96 and electrical conductivity remained low. There were more blue-green algae detected in 1996/97 compared to 1995/96, 1 although the cell counts were still very low. Blue-green algae were discovered twice at Dam Wall Station in 1996/97, while none were found in 1995/96. Total other algae detected at the sites on I Lake Jindabyne were also higher in 1996/97 than in 1995/96. Considerably more data were collected from event sampling in 1996/97 than in 1995/96. The nutrient concentrations and turbidity of event samples were generally higher than those of routine I samples. During a prolonged high flow period in September and October 1997, nutrient concentrations in Thredbo River tended to increase with distance downstream. Stream flow was I the main parameter detenmning the loads during this event. Annual loading estimates indicate considerable nutrient input to the Thredbo River from much of I its catchment, but in particular from the area around Bullocks Flat and the Ski tube. This small area, which is only 1.5% of the catchment area, contributes 24% of the total phosphorus load carried by the Thredbo River at the site just downstream of the Little Thredbo River confluence. It I also contributes around 21% of the total nitrogen load. The areal export coefficients for this area for total phosphorus and total nitrogen were 6.73 kg/ha/yr and 88 kg/ha/yr. respectively. The data frequently indicated that there were much higher total nitrogen concentrations at this site I (22210011) compared to the other sites along the Thredbo River during high flow events. The first flush at the start of events that followed long dry spells showed very high concentrations at Site 22210011 (Appendix 5). I In comparison, the area around Thredbo Village and the major skiing slopes appears to have only a I minor impact on water quality in the river, and certainly one that is much less than expected from an urbanised area with tourist development and a sewage treatment works. Total phosphorus and total nitrogen loads exported per hectare per year from this area are only slightly greater than the I annual areal loadings of other parts of the Thredbo River catchment where there is no such development, such as the area upstream of Deadhorse Gap, the area between Thredbo and Bundilla, and the Little Thredbo River valley (see Tables 4 and 6). This finding is in keeping with I earlier studies, reviewed by Mathieu and Norris (1995). These past studies have generally found little impact from Thredbo Village and its sewage treatment works on nutrient concentrations and I other physico-chemical parameters in the Thredbo River downstream, and on the macroinvertebrate and benthic algal communities.

Another tourist development within the Thredbo River catchment area, the Lake Crackenback resort, which was also considered to have some potential of elevating nutrient loads in the Little I 47 Water Quality in the Snowy River Catchment Area, 1996/97 Report I

Thredbo and Thredbo Rivers, has also been found to have little or no effect. The total phosphorus I and total nitrogen loads exported per hectare per year from this area are amongst the lowest in the entire Thredbo River catchment. Potential impacts of tourist facilities along the upper sections of I Sawpit Creek, which forms part of the Thredbo catchment area between the Little Thredbo River confluence and Paddys Corner, and the Gaden trout hatchery, could not be determined, due to frequent breakdown of autosampling equipment at Paddys Corner (Phil Boreham, SM}{EA, I personal communication). This meant that a number of high flow events could not be included in load calculations for Paddys Corner, leading to inconclusive data for this site.

I The results of this study indicate that the annual areal total phosphorus loadings for much of the Thredbo River catchment were similar to those reported in earlier studies, of around 0.40 kg/halyr (Bowling, 1993). These earlier areal total phosphorus export coefficients calculated by Bowling I were considered to be high, especially as much of the catchment area is forested, and forms part of Kosciusko National Park. Generally forests have minimal soil erosion, and an ability to conserve U and recycle nutrients. CSIRO (1991) gives annual areal total phosphorus export coefficients for native forested areas of 0.09 ± 0.06 kg/halyr. Bowling et al (1993) further reviewed the literature pertaining to areal total phosphorus loads from forested catchments in south-eastern Australia (see I Section 8.5 of the 1993 report), and found a wide range of values reported, mostly in the range from 0.009 to 0.30 kg/halyr. A number of studies (e.g Cosser, 1989, and Gutteridge, Haskins & Davey, 1992) showed that areal total phosphorus export increases under high rainfall conditions LI compared to low rainfall conditions. Bowling et al (1993) attributed the higher than average annual areal total phosphorus load estimated for the Thredbo River catchment as possibly due to I the high rainfall of the area, much of it in the form of snow. Snow is an efficient scavenger of airborne particles, including those containing phosphorus. The similarity of results obtained from two separate studies of total phosphorus loads in the Thredbo River suggest that these high areal I total phosphorus exports are in fact occurring.

Areal annual export coefficients for total nitrogen from the Thredbo River catchment were also I higher than the few published values for other forested areas in south-east Australia. These range between 1.0 and 5.0 kg/haly (Rosich and Cullen, 1982). In comparison, the typical published range for urban areas is 2 to 8 kg/halyr, although values vary greatly and often lie outside this I range (Gutteridge, Haskins & Davey, 1992). The total nitrogen export coefficients calculated for areas of the Thredbo River catchment were 1.7 in the Little Thredbo River valley, 8.4 kg/halyr in I undeveloped parts of the Thredbo River valley, 5 to 6 kg/halyr in the area containing Thredbo Village, the ski fields and sewage treatment plant discharge, and 88 kg/halyr from Bullocks Flat Ski tube station. The overall annual areal export coefficient for the entire catchment upstream of I Site 22210011 was 6.17 kgfhalyr. Many of these values are closer to the published range for urban areas rather than for forested areas. No previous areal total nitrogen exports coefficients are I available for the Thredbo River catchment area to compare with this study. Bowling (1993) calculated areal export coefficients for oxidised nitrogen (NO-N) and ammoniacal nitrogen (NH3- N) of 0.79 and 0.27 kg/halyr respectively, but these are likely to be only small fractions of the total I nitrogen load exported per hectare. Because of the undeveloped, forested or otherwise vegetated nature of much of the catchment, much of the total nitrogen (and total phosphorus) load is likely to be exported in organic form or bound to mineral particulates, rather than as dissolved inorganic I forms. I Water Quality in the Snowy River Catchment Area, 1996/97 Report 48 I

I Trend assessment of data from 1992 to 1997 showed no significant changes in total phosphorus and total nitrogen concentrations for all sites where there were sufficient data to enable this analysis to be undertaken. Turbidity levels were found be increasing significantly at 0.05 I probability level, particularly at sites in Thredbo River. Turbidity values are, however, still well within the recommended range. I I I I I I I I I I I I I 1 I I 49 Water Quality in the Snowy River Catchment Area, 1996/97 Report I

1 9. CONCLUSIONS In this report, water quality in the Snowy River catchment area during 1995/96 was assessed and I water quality conditions between sites and between major inflows to Lake Jindabyne were compared. An assessment was also made of nutrient loads at sites within the Thredbo River subcatchment between July 1995 and June 1997; and for trends in water quality during the period I 1992 to 1997. I The following conclusions can be made from the results of 1996/97 sampling program: . The water quality in most streams in the Snowy River catchment area is generally good at base flow levels. Nutrient concentrations vary from low to moderate. Most steams have low I turbidity. I . The Thredbo River had the best water quality of the three major inflows entering Lake Jindabyne, with very low nutrient concentrations and turbidity. Mowamba Aqueduct inflows had the poorest water quality, with the highest total phosphorus concentrations and turbidity. I Water quality in Wollondibby Creek was only slightly better than that of the Mowamba Aqueduct inflows.

I Routine sampling showed no change in nutrient concentrations at most sites along Thredbo and Little Thredbo River between the years 1995/96 and 1996/97. An exception was at Deadhorse Gap, where total phosphorus concentrations were higher in 1996/97 than in 1995/96. The site I on the Snowy River downstream of Lake Jindabyne, at Dalgety, had lower nutrient ' concentrations and turbidity levels in 1996/97 compared to 1995/96. All sites meet ANZECC (1992) guideline criteria for nutrient concentrations for the protection of aquatic ecosystems.

. The water quality in Lake Jindabyne remains good, with low to moderate nutrient concentration I levels and low turbidity. The results met the ANZECC (1992) guidelines for the protection of freshwater ecosystems in lakes and reservoirs. Nutrient concentrations and turbidity levels were lower in 1996/97 than in 1995/96. I Lake Jindabyne thermally stratified in summer, and as a result dissolved oxygen concentrations fell to low levels. The factors driving this oxygen depletion are unknown, as there is little algal I or other primary productivity in the reservoir. However, some decrease in DO concentrations over summer are typical in most reservoirs, even those with the best quality. I . The extent of the dissolved oxygen depletion at depth while the reservoir was thermally stratified indicates that there is still a possible risk to the water quality of Lake Jindabyne. I Because of this, better catchment management is recommended to protect the water quality of the reservoir, and to prevent its deterioration. . Nutrient concentrations and turbidity in the Thredbo River were higher during high flow events compared to during baseflow periods. I A n umber of conclusions also arise from the studies to estimate nutrient loads in the Thredbo River catchment: I 50 Water Quality in the Snowy River Catchment Area, 1996/97 Report I

I Load estimates indicate high inputs of nutrients from the catchment area between Bundilla and just below the Little Thredbo River confluence. This area comprises the ski-tube and car park surrounding the station at Bullocks Flat. Inputs from this area, which comprises only 1.5% of I the total catchment, account for approximately 24% of the total phosphorus load entering the Thredbo River from its catchment area upstream of Site 2221011, Just downstream of the Little Thredbo River confluence. The estimates of the total nitrogen load contribution from the I Bullocks Flat area was 21% of the total for whole catchment upstream of Site 22210011.

Inputs from the entire Little Thredbo River sub-catchment area were the lowest of any part of I the entire Thredbo River catchment. Impacts from the Lake Crackenback resort were indicated as being negligible. I Inputs per hectare per year from the catchment area containing Thredbo Village and the ski slopes were only slightly higher than those areas with no ski fields or other tourist development, LI such as upstream of Deadhorse Gap, and the area from downstream of Thredbo Village to Bundilla. This indicates very little impact from the Thredbo Village area, as shown also in I independent studies conducted by the University of Canberra (Mathieu and Norris, 1995). Problems arising from failure of the autosampler at Paddys Corner, leading to a number of high flow events not being sampled, meant that annual nutrient load estimates for the lower part of I the Thredbo River between Site 22210011 and Paddys Corner could not be made with any statistical or scientific rigour. This means that the nutrient input from sources within this lower I part of the catchment still needs to be assessed. However, some individual high flow events when sufficient event samples were obtained, such as the period in September and October 1996, indicate considerable increases in both total phosphorus and total nitrogen load can occur [] over the section of river between these two sites.

With the exception of the Bullocks Flat area, the areal total phosphorus load estimates I calculated for Thredbo River catchment area in this study were similar to those calculated for the same catchment in earlier studies (Bowling 1992). These results are somewhat greater than areal total phosphorus load coefficients calculated for other forested areas in south-east I Australia. The high rainfall of this subalpine area, and possibly the efficient scavenging of airborne particulates containing phosphorus by snow, have been proposed as reasons for the Li higher areal loads measured in the Thredbo River subcatchment. Annual areal total nitrogen load coefficients calculated from the study were also slightly higher I than those published for other forested areas of south-east Australia. Conclusions arising from the assessment of water quality trends between 1992 and 1997 were as I follows:

The trend analyses showed no significant change in total phosphorus concentrations at all sites on Thredbo River and the major inflows entering Lake Jindabyne over at least the five years that sampling was undertaken. There was a slight decrease in total nitrogen concentrations over the five years, indicated at two sites: the Walking Track Bridge on the Little Thredbo River, and the Snowy River at Dalgety. This may indicate slightly improving water quality, but the 0.10 probability level was still high enough to suggest no significant trend. Additionally, a site I Water Quality in the Snowy River Catchment Area, 1996197 Report 51 I

just upstream at Alpine Way indicated no such trend. No trends were apparent at the three major inflow sites to Lake Jindabyne, even when the data were flow adjusted.

I . Most sites on the Thredbo River indicated significant increasing trends in turbidity over the period 1992 to 1997. The increase in turbidity was calculated to be between 22% and 32%, but the actual turbidity is so low that the magnitude of change in relation to the range of turbidity 1 possible in freshwaters makes these trends have little meaning. I I L] I

I 1 I I

I I I I I 52 Water Quality in the Snowy River Catchment Area, 1996/97 Report I

I 10. RECOMMENDATIONS

Although results from the water quality monitoring program in the Snowy River catchment area I and Lake Jindabyne indicate generally good water quality, these data can give us only a partial picture of water quality conditions in the area. Information on the long term effects of land use in the catchment is important if we are to maintain good water quality. There are development 1 pressures within the catchment, especially in terms of increasing tourist use of ski-fields and other alpine areas; development of rural residential areas, and other proposed development in the Lake I Jindabyne area. These changes in land usage may all bring about detrimental change to the water quality if undertaken without consideration of maintaining adequate protection of the area's surface waters. As well as these longer term impacts, there are also short term impacts from I episodic events which cause disturbance within the catchment, and may further impair water quality. Such events could include bush fires, and the effects of construction work within the catchment. There is also a demand from the community for increased environmental flows I downstream of Lake Jindabyne, which may bring changes to water quality in the lower Snowy River.

I The upper Snowy River area is fairly unique to New South Wales in that it continues to have amongst the best water quality of the state. The challenge will be to maintain this water quality for Li the future, because remedial action, once water quality deteriorates, is both lengthy and expensive. Long-term routine monitoring of water quality in the river and streams of the Lake Jindabyne catchment, and also in the lake itself and downstream, will therefore be necessary in order to detect I any changes in water quality over time, particularly if land usage changes and development in the catchment area continues. Early detection of changes to water quality is necessary so that measures to prevent further deterioration can be implemented as quickly as possible. The I information from the water quality monitoring programs is, thus, an important tool for total catchment management use in the Snowy River catchment area.

I Ongoing baseline water quality monitoring is therefore recommended at a number of sites in the upper Snowy River catchment. These recommendations include:

I Water quality monitoring should be based around a regular monthly sampling routine for nutrient concentrations and turbidity. Trend analyses coupled with long-term ongoing routine I monitoring offers a cheap method of detecting impacts or changes in water quality over time If impacts are detected, additional, more expensive loading studies may then be required to find the sources of these impacts. Otherwise loading studies are not a recommended altemative to a I routine monitoring program. Sites for long term trend analysis should include Perisher Creek at Blue Cow, Thredbo River at Paddys Corner, Wollondibby Creek at Gunnadoo, the Mowamba I Aqueduct outfall at Lake Jindabyne, and the Snowy River at Dalgety.. Further sampling at the other sites on the Thredbo and Little Thredbo Rivers is less of a I priority, and could be scaled from that done prior to June 1997, unless additional data on the impacts of the various developments in these parts of the catchment are required in the future. Another use for such data could be to determine the nutrient inputs to the Thredbo River from I the section of its catchment between the Little Thredbo River confluence and Paddys Corner, and to measure the success of any catchment management actions. This will require the I 53 Water Quality in the Snowy River Catchment Area, 199619 7 Report I

installation of reliable autosamplers at Paddys Corner. Measurement of the nutrient I contribution from this section of the catchment, and comparison of these inputs with those from other sections of the catchment further upstream could not be adequately achieved in this study I due to the frequent failure of the equipment at Paddys Corner. The nutrient loading studies of the Thredbo River attributed a considerable input from the small I area surrounding Bullocks Flat and the Ski Tube station. It is recommended that this area be investigated further, with a view to finding ways of managing these inputs, and reducing their impacts on the Thredbo River and their potential for increasing eutrophication in Lake I Jindabyne. These investigations, and any remedial action necessary, should be facilitated through the Total Catchment Management process. Further loading studies of this section of the I Thredbo River may be necessary to determine the success of such actions. Water quality monitoring should continue at sites in Lake Jindabyne. Nutrient samples should I be collected at monthly intervals from the surface, mid-depth and bottom waters at Dam Wall Station during summer, and every second month in winter. Algal presence in the surface six metres of water in the reservoir should also be monitored at the same time. Profile data should I also continue to be collected on these sampling occasions at Dam Wall Station and at approximately quarterly intervals at the other sites.

I Past studies have indicated poorer water quality in the Mowamba River and Wollondibby Creek compared to elsewhere in the catchment area. If further detailed water quality studies, including high flow event sampling for nutrient load estimates, are undertaken in the upper I Snowy River catchment area, they should focus on these two streams. Wollondibby Creek is a priority now because of the high nutrient concentrations of its waters, and it will not be affected J by any environmental flow decisions for the Snowy River downstream of Jindabyne Dam. In contrast, studies of the Mowamba River may need to be delayed until a decision has been reached regarding funding responsibilities for these studies. The ultimate destination of I Mowamba River water rests on environmental flow decisions for the Snowy River downstream of Jindabyne Dam. These will determine whether water from the Mowamba River will continue to be diverted to Lake Jindabyne via the aqueduct, or once more flow direct to the Snowy River. I This choice of destination will be a factor in who needs to be approached to fund future water quality studies of the Mowamba River.

I As well as sampling the Snowy River at Dalgety, other sites such as immediately downstream of Jindabyne Dam, below the Mowamba River junction, and at Burnt Hut Crossing may also Li be useful locations for water quality investigation, to allow estimates of inputs from tributary streams and other sources entering the Snowy River below Lake Jindabyne.

I Of major importance, closer alignment of any monitoring programs with the requirements of the Snowy-Genoa Total Catch_rnent Management Committee is needed. Initially the joint SMHEA!DLWC studies were undertaken under the umbrella of a multi-agency Section 22 sub- I committee, but in more recent years they have taken place with only limited direction from the TCM process, or funding from other sources. This is one aspect that needs to be addressed I before future water quality studies are undertaken in the upper Snowy River catchment. For example, a decision is needed as to whether the proposed nutrient loading and catchment

I 54 Water Quality in the Snowy River Catchment Area, 1996/97 Report I

I management studies of Wollondibby Creek and the Mowamba River are still required. This type of decision should be within the role of the TCM Committee. Only then will the necessary basic data on which to make beneficial catchment management decisions be collected. I Additionally, the involvement of or funding assistance from all agencies with an interest in the water quality of the area should be encouraged for future studies. With the exception of some I early input to the SMHEAIDLWC studies within the catchment from the National Parks and Wildlife Service, the subsequent studies have been undertaken solely by these two agencies. Possibly other work in the area is being undertaken by other parties (e.g NPWS, Canberra I University), but this is unknown to the authors of this report. This highlights the need that if other water quality work is being undertaken in the area, it requires better documentation, and I possibly some co-ordination towards some common objectives. This is a further role for TCM. Other suggested recommendations that may enhance the success of catchment management in the I upper Snowy River catchment are: . Investigations on the control of nutrient and sediment input to the Snowy River between I Dalgety and the Victorian border.

. Interaction between the empirical water quality studies and other programs within the DLWC, I such as the Phosphorus Action Campaign; the algal management programs; and Nutrient Management Planning.

I I I I I I I I

I 55 Water Quality in the Snowy River Catchment Area, 1996/97 Report

I I REFERENCES ANZECC (1992). Australian Water Quality Guidelines for Fresh and Marine Waters. Australian I and New Zealand Environment and Conservation Council. Bate, R. (1992) Water quality monitoring in the upper Snowy Mountains 199 1-1992. Supplementary data report. Report No. TS92.036 of the Technical Services Division, New I South Wales Department of Water Resources, June 1992.

Bek, P (1992) Water Quality Management Plan - Carcoar Reservoir and Catchment, Water I Quality Unit, Department of Water Resources, 28th August 1992, 84pp

Boey, A., Stephens, K., Daly, H. and Lee-Young, S. (1997). Water Quality in the Border Rivers I Basin 1960-1995.

Bowling, L. (1993) Water Quality studies of Lake Jindabyne- Final report to the Snowy I Mountains Hydro-Electric Authority. Report No. TS93.098 of the Technical Services Division, New South Wales Department of Water Resources, January 1993. I Bowling, L. and Acaba, Z. (1995) Water Quality in the Snowy River catchment area, 1993/94. Report No. TS95.079 of the Technical Services Branch, New South Wales Department of Water Resources, May 1995. I Bowling, L., Acaba, Z., and Whalley, P. (1993) Water quality in the Snowy River catchment area, 1992/93. Report No. TS93.079 of the Technical Services Division, New South Wales I Department of Water Resources, November 1993. Cleveland, W.S. (1994). The Elements of Graphing Data. AT&T Bell Laboratories, Murray Hill, I New Jersey. 292 pp. Cochran, W. G. (1977). Sampling Techniques, 3rd edition. John Wiley & Sons, New York

I Cunningham, R. B. and Morton, R. (1983). A statistical method for the estimation of trend in salinity in the River Murray. Aust. J. Soil Res. 21, 123-132. I D'Elia , CF., Steudler, PA., and Corwin, N. (1977) Determination of total nitrogen in aqueous samples using persulfate. Limnology and Oceanography 22, 760-764. I Draper, N. and Smith, H. (1981). Applied Regression Analysis, Second edition. Wiley, Brisbane. Efron, B. and Gong, G. (1983). A leisurely look at the bootstrap, the jackknife, and cross- I validation. The American Statistician 37(1): 36-48. Efron, B. and Tibshirani, R.J. (1993). An Introduction to the Bootstrap. Monographs in Statistics I and Applied Probability 57. Chapman & Hall, New York. Esterby, S .R. (1992). Trend Analysis for Environmental Data. In, Invited Papers International Environmental Biometrics Conference. Sydney Australia. The Statistical Society of I Australia (NSW Branch) and The American Statistical Association.

Ferguson, RI., (1986). River Loads Underestimated by Rating Curves. Water Resources I Research, 22(1), 74-76. I 56 Water Quality in the Snowy River Catchment Area, 1996197 Report

Friendly, M. (1991). SAS' System for Statistical Graphics, First Edition. SAS Institute Inc. Cary, I North Carolina. 697pp.

Gilbert, R. 0. (1987). Statistical Methods for Environmental Pollution Monitoring. Elsevier I Science Publishers, Amsterdam. 552 pp.

Helsel, DR., and Hirsch, R.M. (1992). Trend Analysis. In: Statistical Methods in water I Resources. pp 323-355. Elsevier Science Publishers, The Netherlands.

Houldsworth, B. (1995). Central and North West Regions Water Quality Program. 1994/95 I Report on Nutrients and General Water Quality Monitoring. Department of Land and Water Conservation, TS 95.088.

I Kinross, C. (1996) Water Quality of the Snowy River catchment area, 1994/95- A report to DLWC, Sydney/South Coast Region, Snowy Mountains Hydro-Eleetric Authority, Snowy- Genoa TCM. Report No. TS 96.059 of the Technical Services Division, New South Wales I Department of Water Resources, June 1996. Koutsiyanis, A. (1977). Theory of Econometrics, second edition. McMillan Press Ltd.

I Manly, B.F.J. (1997). Randomisation, Bootstrap and Monte Carlo Methods in Biology. Chapman and Hall.

I Preece R. and Robinson G. (1996). Key Sites Water Quality Monitoring Program - Report on the NSW River Water Quality Trends July 1990 to June 1995. NSW Department of Land and I Water Conservation Report No. TS 95.184. Preece R., Robinson G. and Graice, J. (1994). Key Sites Water Quality Monitoring Program - Report on the NSW River Water Quality Trends July 1989 to June 1994. NSW Department I of Land and Water Conservation Report No. TS 94.125.

Preece,R., Robinson, G. & Currey, M. (1997). Key Sites Program. River Water Quality Trends July 1991 to June 1996. Department of Land and Water Conservation CENTRE FOR NAT97.041.

Preston, C. (1996). Central and North Western Regions Water Quality Program - 1995/96 Report on Nutrients and General Water Quality Monitoring. Department of Land and Water Conservation TS96.049.

Roberts, K.R. (1994). Style Guide, Version 1 (TS 94.050) Document Working Group, Department of Water Resources

I Robinson, G., and Hatfield, E. (1992). Application of Flow Duration Curves in Calculating Nutrient/Pollutant Loadings. Poster paper presented at the Environmental Biometrics I Conference, Sydney, Australia, 14-15 December 1992. Tukey, J.W. (1977). Exploratory Data Analysis. Addison-Wesley, Reading, MA.

I Williamson, DR., Gates, G.W., Robinson, G., Linke, G.K., Seker, M.P. and Evans, W.R. (1997). Historic Trend in Salt Concentration and Saltload of Stream Flow in the Murray Darling I Drainage Division. Dryland Technical Report No. 1, Murray Darling Basin Commission. I 57 Water Quality in the Snowy River Catchment Area, 1996197 Report L I Appendix 1. Boxplot defined

Boxplots provide a useful means of displaying a summary of a group of data, allowing meaningful I comparisons to be made between groups. Boxplots focus attention on five important properties of a group of data:

I typical or central value spread or variability I shape - symmetry or skewness outlying data points I behaviour of the tails. The central box of the boxplot delineates the 25th percentile (lower quartile), the 50th percentile (median) and the 75th percentile (upper quartile). Inner fences are then defined, 1.5 times the E interquartile range (IQR) above and below the box. Whiskers are added to the box, drawn from the top and bottom to the most extreme value inside the fence. All data points outside the inner fence I are individually identified, either as 'outliers' or 'extreme outliers'. The following diagram outlines the principal components and underlying statistics of a boxplot, for a moderately symmetric distribution. MAX Maximum LI 'Extreme outlier' or 'far out point', shown as square, more than 3 IQRs greater than upper I . 1•e. UOF Upper outer fence = UPQ + 3 x IQR

I 0 'Outlier' or 'outside point', shown as diamond, between 1.5 and 3 IQRs greater than upper quartile. n UTF Upper inner fence = UPQ + 1.5 x IQR I m..lueJth. U1F ...... ppqrt1 ..=2.5th..pecentile MED Median = 50th percentile, also shown as filled circle H _•_ LOQ Lower quartile = 25th percentile I LO_W Lower whisker: minimum value greater than LIF

McGill, Tukey and Larsen (1978) suggest making the width of each boxplot proportional to the I square root of the number of observations it represents, based on the fact that standard errors are I inversely proportional to i(n). This technique has been applied throughout this report. I

I 58 Water Quality in the Snowy River Catchment Area, 1996/97 Report I I I Appendix 2. Ratio Method for Calculating Loads I The load estimates were calculated using the ratio method by Cochran (1977). Cochran's ratio I method is expressed as:

rej 7.•~ q1C1 I xQ Load= '='n q. I 1=1 I where: n number of samples Q Total Annual Discharge for station

C1 instantaneous concentration I q1 discharge on sampling occasion i I I I I I I I I I I

I Water Quality in the Snowy River Catchment Area, 1996/97 Report 59 I I

Appendix 3. Summary statistics of routine samples, 1996 to 1997

3A. Total phosphorus (tg/L) Area Site Number Depth N MIN 25th %tiie Median 75th %tilel MAX Upper Snowy I Perisher Creek 222513 12 2.5 8.75 15 20 30 Middle Snowy Thredbo River 22210019 12 2.5 2.5 12.5 15 35 I 22210016 12 2.5 2.5 15 17.5 210 22210012 12 2.5 6.25 10 15 65 22210017 12 10 15 20 22.5 65 I 22210015 12 2.51 12.5 15 30 45 22210011 12 2.5 2.5 10 17.5 100 222541 12 2.5 2.5 15 20 55 F] Wollondibby Creek 222544 12 15 20 27.5 42.5 95 Lake Jindabyne 22210001 Top 5 2.5 2.5 2.5 10 20 Middle 5 2.5 10 15 15 15 I Bottom 5 101 10 15 20 20 22210002 Top 1 2 2.51 2.5 2.5 2.5 2.5 Middle 1 2 2.5 2.5 6.25 10 10 Bottom 2 2.5 2.5 6.25 10 10 H 22210003 Top 3 2.5 2.5 2.5 15 15 Middle 3 2.5 2.5 2.5 15 15 Bottom 2 2.5 2.5 8.75 15 15 I MowambaRiver 22210142 12 15 17.5 32.5 40 110 Lower Snowy I SnonyRiver 222006 10 2.5 15 15 20 115 I I I I I I I I 60 Water Qualn'y in the Snowy River Catchment Area, 1996/97 Report I

I 3B. Total nitrogen (pgIL) Area Site Depth N MIN 25th Median 75th%tile MAX I Number - %tile Upper Snowy Perisher Creek 222513 12 100 150 200 300 1000 I Middle Snowy Thredbo River 22210019 12 50 100 100 225 500 22210016 12 100 100 150 175 650 22210012 12 50 100 100 175 700 I 22210017 12 50 100 125 175 600 22210015 12 501 100 150 200 550 22210011 12 501 100 100 175 1000 I 222541 12 1001 100 150 250 550 Wollondibby Creek 222544 12 100 300 375 650 1000 Lake Jindabyne 22210001 Top 5 150 150 200 200 500 I Vidd 5 150 150 150 200 200 5 150 200 200 200 250 22210002 2 100 100 125 150 150 I 2 150 150 175 200 200 ottom 2 200 200 200 200 200 22210003 Top 3 150 150 150 250 250 LI Middle 3 150 150 150 200 200 Bottom 2 150 150 200 250 250 Mowamba River 22210142 1 12 150 225 350 525 950 Lower Snowy I Snowy River 222006 10 150 150 200 250 650 I I I I I I I I I 61 Water Quality in the Snowy River Catchment Area, 1996/97 Report I

I 3C. Turbidity (NTU) Area Site Depth N MiN 25th Median 75th MAX I Number - %tile %tile Upper Snowy Perisher Creek 222513 121 0.35 0.65 0.7 1.6 7.4 I Middle Snowy Thredbo River 22210019 12 0.35 0.575 1.15 1.65 11 22210016 12 0.15 0.475 1.05 2.75 12 I 22210012 12 0.251 0.5 0.975 2.75 21 22210017 12 0.25 1.1 1.9 4.5 24 22210015 12 0.9 1.05 2.6 6.45 15 I 22210011 12 0.45 0.575 0.7 2.85 23 222541 12 0.2 0.65 1.1 5.8 12 Wollondibby Creek 222544 12 0.551 4.55 9.05 11.5 40 I Lake Jindabyne 22210001 Top 5 0.81 0.9 0.9 1.5 3.1 Middlel 5 0.25 0.75 1.1 1.5 1.8 I Bottom 5 0.75 1 1 1.6 6.2 22210002 Top 2 0.75 0.75 1.375 2 2 Middle 2 0.25 0.25 0.775 1.3 1.3 I Bottom 2 0.951 0.95 1.675 2.4 2.4 22210003 TopI 3 0.25 0.25 0.9 1.1 1.1 Middle 3 0.7 0.7 0.95 1.2 1.2 I Bottom 2 1.1 1.1 1.55 2 2 Mowamba River 22210142 12 0.85 3.4 4.45 8.75 75 Lower Snowy I Snowy River 222006 10 11 1.1 1.8 4 160 I 1 I

I 1 E I Water QualUy in the Snowy River Catchment Area, 1996/97 Report 62 ------

3D.) Summary statistics comparison of 1995/96 and 1996197 data for routine samples.

Upper Snowy Perisher Creek 222513 5 2.5 15 15 30 30 100 100 200 200 1,100 1,000 0.40 0.35 1.1 0.7 5.3 7.4 Middle Snowy Thredbo River 22210019 2.5 2.5 5 125 35 35 50 50 100 100 350 500 0.50 0.35 11 1.2 6.3 11.0 22210016 2.5 25 15 15 25 210 100 100 150 150 1,400 650 0.35 0.1.5 16 1.1 6.0 12.0 22210012 2.5 25 10 10 351 65 501 501 125 100 350 700 0.40 0.25 1.3 1.0 7.0 21.0 22210017 5 10 15 20 50 65 50 50 125 125 500 600 0.60 0.25 1.9 1.9 18.0 24.0 22210015 5 2.5 18 15 50 45 100 50 175 150 500 550 0.75 0.9 1.8 2.6 15.0 15.0 22210011 2.5 2.5 10 10 50 100 50 50 150 100 450 1,000 0.40 0.45 1.5 0.7 8.1 23.0 222541 2.5 2.5 15 15 575 55 100 100 200 150 4,700 550 0.60 0.2 2.7 II 70.0 12.0 Wollondibby 222544 15 15 30 27.5 125 95 300 100 500 375 1,100 1,000 3.90 0.55 8.7 9.1 60.0 40.0 Creek Lake Jindabyne 22210001 Top 5 2.5 10 2.5 15 20 200 150 200 2001 350 5001 0.70 0.8 3.2 0.9 5.6 3.1 Middle 1 5 2.51 13 15 20 15 200 150 225 150 550 200 1.50 025 3.7 1.1 5.8 1.8 Bottom 10 10 20 15 20 20 200 150 250 200 300 250 1.50 0.75 1.9 1.0 2.2 6.2 22210002 Top 5 2.5 10 2.5 20 2.5 150 100 200 125 350 150 0.40 0.75 1.9 1.4 3.3 2.0 Middle 5 2.5 15 6.25 15 10 150 150 200 175 250 200 1.10 0.25 1.6 0.8 2.1 1.3 Bottom 15 2.5 15 6.25 20 10 150 200 250 200 250 200 0.95 0.95 4.5 1.7 8.1 2.4 22210003 Top 5 2.5 10 2.5 15 15 200 150 200 150 250 250 1.10 0.25 1.2 0.9 1.2 1.1 Middle 5 2.5 10 2.5 20 15 150 150 200 150 250 200 1.10 0.7 2.8 1.0 4.5 1.2 Bottom 10 2.5 15 8.75 20 15 200 150 200 200 300 250 1.00 11 3.2 1.6 5.3 2.0 Mowamba River 22210142 20 15 25 32.5 95 110 200 150 500 350 1,100 950 260 0.85 5.6 4.5 40.0 75.0 Lower Snowy I Snowy River 1222006 15 2.5 38 15 200 115F 110 150 425 200 650 650 4.70 1 6.0 1.8 27.0 160.0

63 Water Qualn'y in the Snowy River Catchment Area, 199619 7 Report Appendix 4. Summary statistics of event samples.

I 4A) Event sampling in 1996/97, by site Site Total Phosphorus (ug/L) Total Nitrogen (ug/L) Turbidity (NTU) Duration 01 Event I Number N Min Median Max N Min Median Max N Minl Median Max Date From Date To 22210019 6 15 17.5 25 6 300 475 1200 6 0.95 4.15 8.6 5-Jul-96 7-Jul-96 3 10 20 20 3 250 350 900 3 1.9 2.5 2.9 27-Jul-96 27-Jul-96 5 15 20 25 5 250 350 1000 5 1.3 2.2 3.3 27-Aug-96 30-Aug-96 I 2 15 15 15 2 200 200 200 2 1.2 1.5 1.8 1-Sep-96 2-Sep-96 1 10 10 10 1 400 400 400 1 3.7 3.7 3.7 7-Sep-96 7-Sep-96 53 2.5 10 30 53 150 250 550 53 0.25 1.5 8 10-Sep-96 17-Oct-96 I 8 15 17.5 25 8 250 300 1000 8 2.3 4 7 17-Nov-96 18-Nov-96 6 20 80 115 8 850 1300 3200 6 7.6 22.5 36 2-Mar-97 3-Mar-97 4 35 60 80 4 600 1050 3800 4 7.6 17.5 26 7-May-97 7-May-97 I 22210016 7 15 25 35 7 300 450 2200 7 2.6 4.3 5.1 5-Jul-96 6-Jul-96 2 20 22.5 251 2 850 1075 1300 2 4.3 4.7 5.1 27-Jul-96 27-Jul-96 4 30 30 40 4 400 650 2500 4 4.1 5.35 5.9 28-Aug-96 29-Aug-96 I 2 20 22.5 25 2 400 525 650 2 2.6 2.75 2.9 1-Sep-96 2-Sep-96 63 2.51 15 65 63 150 300 950 63 0.5 3.1 25 9-Sep-96 17-Oct-96 9 15 25 55 9 250 400 500 9 1.2 3.4 14 17-Nov-96 19-Nov-96 I 5 45 105 130 5 850 1500 2100 5 13 24 38 2-Mar-97 2-Mar-97 6 55 102.5 1601 6 900 1450 3000 6 11 17.5 39 7-May-97 7-May-97 22210012 6 20 25 30 6 300 375 1200 6 2 2.9 4.2 5-Jul-96 6-Jul-96 I 1 40 40 40 1 1400 1400 1400 1 4 4 4 27-Jul-96 27-Jul-96 3 20 20 40 3 300 300 1800 31 1.9 2.7 4 28-Aug-96 29-Aug-96 47 2.5 20 80 47 150 350 1000 47 0.55 5.1 35 9-Sep-96 17-Oct-96 3 15 20 25 3 200 450 1200 3 1.4 2 2.5 18-Nov-96 19-Nov-96 I 6 75 132.5 165 6 950 1400 3900 6 10 21.5 38 2-Mar-97 2-Mar-97 6 45 65 105 6 700 1025 6500 6 13 23 28 7-May-97 7-May-97 22210017 2 40 40 40 2 650 675 700 2 6.2 7 7.8 5-Jul-96 6-Jul-96 [1 2 25 27.5 30 2 250 400 550 2 4 4.7 5.4 28-Jul-96 29-Jul-96 59 2.5 25 125 59 200 400 1000 59 1.2 6.2 40 9-Sep-96 17-Oct-96 4 35 45 551 4 550 600 900 4 4.4 6.6 9.3 15-Nov-96 19-Nov-96 I 12 30 70 310 12 350 1100 2400 12 11 19.5 68 2-Mar-97 4-Mar-97 22210015 5 20 25 25 5 250 250 650 5 2.3 3.3 5.2 27-Jul-96 29-Jul-96 9 15 20 25 9 250 300 450 9 1.8 3 7.5 13-Aug-96 31-Aug-96 I 1 20 20 20 1 200 200 200 1 3.1 3.1 3.1 7-Sep-96 7-Sep-96 59 10 25 100 59 200 350 850 59 1.3 5.51 36 9-Sep-96 18-Oct-96 6 20 25 351 6 300 375 1200 6 3.8 5.95 6.8 16-Nov-96 18-Nov-96 I 9 25 45 1751 9 400 700 1300 9 4.7 14 38 2-Mar-97 3-Mar-97 22210011 5 30 50 185 5 350 450 2300 5 2.4 2.7 5.1 6-Jul-96 6-Jul-96 1 285 285 285 1 2300 2300 2300 1 11 11 11 27-Jul-96 27-Jul-96 I 3 80 125 135 3 550 750 900 3 3.6 7.1 21 28-Aug-96 29-Aug-96 67 2.5 25 175 67 1001 300 2700 67 0.3 1.7 28 9-Sep-96 17-Oct-96 41 351 55 115 4 350 475 1700 4 2.9 3.8 4.9 18-Nov-96 19-Nov-96 41 701 117.5 1701 4 650 1150 1500 4 8.2 17.5 28 2-Mar-97 2-Mar-97 I 7800 6 21 28.5 39 7-May-97 7-May-97 6 801 140 615 6 1000 1550 222541 2 25 30 35 2 300 300 300 2 5 5.55 6.1 6-Jul-96 6-Jul-96 68 2.5 20 110 68 100 250 950 68 0.65 2.8 23 9-Sep-96 17-Oct-96 I 5 25 35 45 5 250 350 500 5 3.4 4.3 4.5 18-Nov-96 19-Nov-96 3 75 100 115 3 8501 10001 11001 31 251 30 32 7-May-97 7-May-97 I 64 Water Quality in the Snowy River Catchment Area, 1996/97 Report n

4B.) Total phosphorus and total nitrogen load during event sampling from 9 September I 1996 to October 1996.

I Site Total Phosphorus Load (k /day) Total Nitrogen Load (k /day) Number N Min Median Max N Min Median Max

I 22210019 53 0.18 2.72 30.59 53 25 144 510 22210016 63 1.08 16.07 131.80 63 106 289 1683 I 22210012 47 2.74 47.04 418.61 47 127 1098 4692 22210017 59 0.15 2.28 21.77 59 7 34 207 I 22210015 59 0.67 3.26 19.22 59 8 39 204 22210011 67 4.58 54.04 405.00 67 120 598 4752 I 222541 68 6.87 60.98 731.31 68 150 758 6928 I I I I I I I I I I i I 65 Water Quality in the Snowy River Catchment Area, 1996/97 Report I

I Appendix 5. Total phosphorus and total nitrogen concentration at the start of an event, by site.

I DATE SITENO TIME Total Phosphorus Total Nitrogen FLOWMLD (ug/L) (ug/L) Mliday

I 28-Feb-96 22210019 10:30:00 75 6200 72.576 22210016 10:45:00 80 9400 304.128 22210012 13:45:00 55 6600 633.312 I 22210017 18:10:00 30 900 26.2656 22210015 18:20:00 35 2900 33.8688 22210011 13:55:00 305 14000 681.696

I 11:00:00 90 1500 76.2048 16-Mar-96 22210019 22210016 10:20:00 95 9300 316.224 22210012 13:05:00 145 7800 692.928 I 22210017 12:35:00 55 1200 26.2656 22210015 10:50:00 40 550 33.8688 I 22210011 12:35:00 345 15000 706.752 7-Apr-96 22210019 10:29:00 20 2900 72.576 22210016 11:10:00 75 6700 304.128 I 22210017 16:22:00 50 1 300 26.2656 22210015 15:16:00 25 2800 33.8688 22210011 13:35:00 325 12000 706.752 I 222541 17:30:00 80 850 786.24 5-May-96 22210019 18:50:00 135 3 900 72.576 22210016 19:45:00 150 4400 328.32 I 22210012 22:10:00 125 5200 613.44 22210017 17:00:00 35 900 26.2656 22210015 17:35:00 45 2600 35.3376 I 22210011 22:00:00 470 13000 681.696

I 23-Jun-96 22210019. 30 2000 76.2048 22210016. 50 2900 316.224 22210012 12:25:00 35 2000 756 22210017 19:40:00 25 500 24.2784 I 22210011 12:40:00 260 5500 785.376

28-Aug-96 22210019 5:05:00 25 400 190.944 I 22210016 3:05:00 40 2500 316.224 22210012 5:55:00 40 1 800 652.32 22210015 11:55:00 20 350 50.7168 I 22210011 5:45:00 135 900 706.752 I I I I Water Quality in the Snowy River Catchnient Area, 1996/97 Report I

I Appendix 5 (cont... DATE SITENO TIME Total Phosphorus Total Nitrogen FLOWMLD I (ug/L) (ugfL) ML/day 9-Sep-96 22210016 20:10:00 25 800 328.32 22210012 20:55:00 25 700 652.32 I 22210017 2:10:00 15 250 26.2656 22210015 8:30:00 30 250 33.8688 22210011 21:05:00 175 1600 706.752 I 222541 19:30:00 30 350 482 18-Nov-96 22210019 0:15:00 20 300 293.76 22210016 7:20:00 20 500 725.76 I 22210012 8:00:00 15 200 1710.72 22210017 5:45:00 45 550 62.8992 22210015 5:55:00 30 350 66.528 I 22210011 8:15:00 115 1700 1831.68 222541 7:30:00 35 500 1987

I 2-Mar-97 22210019 5:50:00 115 3200 75.9 22210016 6:15:00 105 2 100 449 22210012 8:25:00 135 3 900 783 I 22210017 7:30:00 60 1100 27.3 22210015 6:40:00 25 800 34.5 22210011 8:10:00 160 1500 719

I 7-May-97 22210019 4:50:00 80 3 800 72.576 22210016 4:30:00 160 3000 457.056 22210012 7:20:00 105 6500 633.312 I 22210017 11:19:00 25 200 95.04 22210015 11:33:00 30 150 0 22210011 7:40:00 615 7800 812.16 I 222541 9:50:00 15 250 93.86 I I I I I I I 67 Water Quality in the Snowy River Catchment Area, 1996/97 Report I