Midges (Chironomidae: Diptera) in Australian freshwater and upland streams.

by

Ian Alexander Wright M.Sc. (Macquarie)

A thesis submitted in fulfilment of the requirements for the degree Doctor of Philosophy

University of Western Sydney New South Wales

June 2005 Abstract

This thesis revealed Australian lakes to be richer in chironomid (Diptera: Insecta) species than was previously recognised. A methodology for the collection of chironomid exuviae from lakes was developed using a 12-month study of exuviae from a single , Lake McKenzie, Jervis Bay. The method is a rapid and effective way to produce an inventory of species living within a lake.

In addition, a distinct biogeographical pattern was detected for chironomid species from a survey of chironomids from southern and eastern Australian freshwater lakes. Geographical location of the lake was more influential on the distribution of chironomid species than was the type of lake. The majority of lake- dwelling chironomid species in this investigation were restricted to lakes within one of four geographic lake regions; Tasmania, south-eastern Australian mainland, Fraser Island or tropical north Queensland.

A temporal investigation of chironomid exuviae was conducted on a pair of small upland Blue Mountain waterways. Abundance and species richness of exuviae exhibited diurnal patterns. The majority of chironomid exuviae, from the dominant species, were collected at night, particularly during the hours between sunset and midnight.

A chemical and macroinvertebrate survey of zinc and sewage organic waste discharges to upland streams in the Blue Mountains detected marked ecological impairment. Macroinvertebrate families responded in different ways to the two different types of waste discharge. According to family-level results, chironomid larvae responded negatively (reduced abundance) to the zinc pollution and positively (increased abundance) to the sewage pollution.

Another major finding from this thesis was that chironomid species assemblages, in the streams surveyed, were strongly impaired by zinc-contaminated mine drainage and sewage effluent. This differed to the family-level larval results. This thesis provided the first Australian evidence that many chironomid species are intolerant of heavy-metal pollution. This research also revealed further evidence that many chironomids species are intolerant of sewage pollution.

Dedication

This thesis is dedicated to my parents,

Sina and Colin Wright.

iii Table of Contents

ABSTRACT II

LIST OF TABLES VII

TABLE OF FIGURES XI

ACKNOWLEDGEMENTS XVII

STATEMENT OF ORIGINALITY XIX

CHAPTER ONE 1

1.1 General Introduction 1

1.2 Northern Hemisphere chironomid studies 3

1.3 Australian chironomid studies 9

1.4 Scope of this thesis 16

CHAPTER 2 DEVELOPMENT OF A METHOD FOR COLLECTING LAKE CHIRONOMID EXUVIAE 18

2.1 Introduction 18

2.2 The study area 20

2.3 Materials and Methods 21 2.3.1 Exuviae collection – spatial study 21

2.3.2 Within-year temporal investigation 22

2.3.3 Diel exuviae investigation 22

2.4 Results 20

2.4.1 Spatial heterogeneity 24 2.4.2 Within-year exuviae variation 26 2.4.3 Diel exuviae variation 28 2.4.4 Spatial aggregation during 24-hour study 30

2.5 Discussion 32

CHAPTER 3. EASTERN-AUSTRALIAN SURVEY OF LAKE-DWELLING CHIRONOMIDS. 36

3.1 Introduction: 36

3.2 Study locations: freshwater lakes from southern and eastern Australia. 39

iv 3.3 Materials and Methods 48 3.3.1 Exuviae collection 48 3.3.2 Assessment of human disturbance 49 3.3.3 Data analysis 49

3.4 Results 51 3.4.1 Chironomidae and Chaoboridae results 51 3.4.2 Chironomid community structure analysis (NMDS analysis) entire survey 58 3.4.3 Significance testing of lake groups (ANOSIM results) 59 3.4.4 PRIMER BIOENV results 61 3.4.5 Physical and chemical results 62

3.4 Discussion 74

CHAPTER 4 WATER CHEMISTRY AND MACROINVERTEBRATE SURVEY OF THE UPPER GROSE RIVER, NSW: EFFECTS OF ORGANIC EFFLUENT AND ZINC-RICH MINE DRAINAGE. 79

4.1 Introduction 79

4.2 Material and Methods 81 4.2.1 Description of study area 81 4.2.2 Study sites 86 4.2.3 Collection of water samples 88 4.2.4 Data analysis 90 4.2.5 Measurement of pollution affinity 91

4.3 Macroinvertebrate results 92 4.3.1 Macroinvertebrate abundance and taxon richness 92 4.3.2 Multivariate community structure analysis (NMDS) 96 4.3.3 Results of water physical and chemical indicators 100

4.4 Discussion 112

CHAPTER 5 SPATIAL AND TEMPORAL VARIATION OF CHIRONOMID EXUVIAE FROM TWO SMALL UPLAND STREAMS IN THE BLUE MOUNTAINS, NSW. 119

5.1 Introduction: 119

5.2 Material and Methods 121 5.2.1 Study area 121

5.2.2 Sampling sites and sample collection 123 5.2.3 Temperature and rainfall conditions during the study 124

5.2.4 Data Analysis Procedures 125

5.3 Results 126 5.3.1 Diurnal variation results 126 5.3.2 Diurnal patterns in individual species 132

5.4 Discussion 138

CHAPTER 6 CHIRONOMID EXUVIAE SURVEY OF THE GROSE RIVER, A ZINC AND SEWAGE POLLUTED UPLAND-RIVER. 142

6.1 Introduction 142

v 6.2 Materials and Methods 144 6.2.1 Study area 144 6.2.2 Chironomid exuviae collection 146 6.2.3 Data analysis procedures 148

6.3 Results 150 6.3.1 Chironomid exuviae abundance 150 6.3.2 Comparison of samples taken above and below an STP outflow 150 6.3.3 Comparison of samples taken above and below a zinc-rich drainage outflow 155 6.3.4 Multivariate analysis of chironomid exuviae data 158

6.4 Discussion 164

CHAPTER 7 GENERAL DISCUSSION AND CONCLUSIONS 169

7.1 Summary of results 169

7.2 Human impact on high conservation streams and lakes 181

7.3 Recommendations 184

CHAPTER 8 REFERENCES 188

vi List of Tables

Table 2.1 Chironomid and Chaoborid pupal exuviae taxa collected from eight sites around the shoreline perimeter of Lake McKenzie, 28 April 1997 (see Figure 2.1)...... 25 Table 2.2. Chironomidae and Chaoboridae pupal exuviae taxa collected from leeward shore of Lake McKenzie on four occasions over an 11 month period. (X signifies the taxon was present. Surface water temperature on the day of sampling is given)...... 27 Table 2.3 Chironomid and Chaoborid exuviae collected from Lake McKenzie in the 24-hour exuviae study ...... 29 Table 3.1 Lake groups within each of the four geographical region in this study. Tropical north Queensland is ‘TNQ’, Fraser Island is ‘FSI’, south-eastern mainland is ‘SEM’ and Tasmania is ‘TAS’. See Figures 3.1 and 3.2...... 40 Table 3.2 Key Australian lake publications used to select lakes for this study, by lake geographical region...... 45 Table 3.3 Chironomid taxa reported from Australian lake publications 1974-1993. Species / taxa names as reported...... 46 Table 3.4 Geographic regional distribution of chironomid species, categorised according to biogeographic regions, collected from freshwater lakes in this study. X symbol indicates that the species was found from at least one lake in that biogeographic region. TAS = Tasmania, SEM = south-eastern mainland, FRI = Fraser Island, TNQ = tropical north Queensland...... 53 Table 3.5 Geographical locations of the 134 species recorded ...... 56 Table 3.6 Number of species of chironomid in each sub-family and chaoborid species recorded in each of the four lake regions ...... 56 Table 3.7 F-statistics and associated probabilities from analyses of variance of (log X + 1) transformed chironomid species richness data from lakes in six lake groups (see 3.2.4)...... 56 Table 3.8 R-statistics (Clarke, 1993) from two-way crossed ANOSIM for pairwise comparison of chironomid data from geographic regions (Tasmania, south eastern mainland, Fraser Island, and Tropical North Queensland)...... 60 Table 3.9 R-statistics (Clarke 1993) from two-way crossed ANOSIM for pairwise comparison of chironomid data from geographic regions and geomorphic types (Tasmania dune and glacial; south eastern mainland dune, sinkhole and maar; Fraser Island dune, and Tropical North Queensland dune and maar)...... 61 Table 3.10 Summary of BIOENV results for the chironomid data from the Eastern Australian lake survey...... 62 Table 3.11. Physical and chemical field data collected in this study from Tasmanian Lakes. * data from other studies (see Table 3.2)...... 63

vii Table 3.12. Physical and chemical field data collected in this study from South eastern mainland lakes sampled (South Australian, Victorian and Jervis Bay Territory) * data from other studies (see Table 3.2)...... 64 Table 3.13. Physical and chemical field data collected in this study from Fraser Island Lakes. * data from other studies (see Table 3.2)...... 65 Table 3.14. Physical and chemical field data collected in this study from Tropical North Queensland lakes Lakes. * data from other studies (see Table 3.2)...66 Table 3.15 F-statistics and associated probabilities from analyses of variance of pH and (log X + 1) transformed electrical conductivity data from lakes in six lake groups (see 3.2.4) sampled in this study...... 67 Table 4.1 Summary information for each of the sampling sites used in this study ...... 87 Table 4.2 Date of macroinvertebrate and physio-chemical sampling (2003) for each site in the Grose River survey...... 89 Table 4.3 Pollution affinity calculations for taxonomic groups, based on results from Grose River macroinvertebrate survey, autumn 2003. The water pollution sources were sewage effluent and zinc-rich mine drainage...... 91 Table 4.4 List of macroinvertebrates collected from sites in the Upper Grose River ...... 93 Table 4.5 F-statistics and associated probabilities from analyses of variance of (log X + 1) transformed macroinvertebrate taxon richness data and (log X + 1) transformed macroinvertebrate abundance data...... 94 Table 4.6 Zinc and sewage pollution affinity of pollution indicator taxa, according to the Grose River survey (see 4.2.5 for methodology and Table 4.14 for grades)...... 97 Table 4.7 R-statistics (Clarke 1993) from two-way crossed ANOSIM for pairwise comparison of sites for fourth root transformed macroinvertebrate data corrected from reference sites in the Upper Grose River and from below the mine seepage site (GDD). Abbreviations are given in Table 4.1...... 99 Table 4.8 Results of SIMPER breakdown, the most influential macroinvertebrates contributing to the different communities at the reference sites compared with Grose downstream Dalpura Creek (GDD), the site most affected by mine drainage...... 99 Table 4.9 Results of SIMPER breakdown, the most influential macroinvertebrates contributing to the different communities at the reference sites compared with Hat Hill Creek (HHD)downstream STP, the site most affected by STP effluent...... 99 Table 4.10 F-statistics and associated probabilities from analyses of variance of (log X + 1) transformed electrical conductivity data, pH (untransformed) data and (log X + 1) transformed water temperature data from sites sampled in the upper Grose River and its tributaries between April and June 2003...... 103 Table 4.11 F-statistics and associated probabilities from analyses of variance of (log X + 1) transformed hardness data from sites sampled in the upper Grose River and its tributaries between April and June 2003...... 103

viii Table 4.12 F-statistics and associated probabilities from analyses of variance of (log X + 1) transformed total zinc data, (log X + 1) transformed total nitrogen data and (log X + 1) transformed total phosphorus data from sites sampled in the upper Grose River and its tributaries between April and June 2003...... 106 Table 5.1 Abundance of chironomid exuviae collected at Hat Hill Creek. Collected between 12 December 2003 and 19 February 2004...... 126 Table 5.2 Number of species of chironomid exuviae collected at Hat Hill Creek. Collected between 12 December 2003 and 19 February 2004...... 127 Table 5.3 List of chironomid species recorded at each site in this study ...... 128 Table 5.4 Number of chironomid exuviae collected at Creek collected over a 24-hour period on 31 December 2003...... 129 Table 5.5 Number of chironomid species collected at Victoria Creek collected over a 24-hour period on 31 December 2003...... 129 Table 5.6 F-statistics and associated probabilities from analyses of variance of number of exuviae (log X + 1 transformed) and number of species (log X + 1 transformed) collected from Hat Hill Creek on the four sampling occasions...... 129 Table 6.1 Chironomid exuviae sampling sites, number and duration of drift samples collected in this study in the Grose River system. December 2003 to March 2004...... 147 Table 6.2 List of all chironomid species recorded at each Grose system site (X = species found at site) in this exuviae study, December 2003 to February 2004...... 151 Table 6.3 F-statistics and associated probabilities from analyses of variance of number of species of exuviae (log X + 1 transformed) and number of exuviae (log X + 1 transformed) collected from five sampling occasions at Hat Hill Creek (upstream STP) and Hat Hill Creek (downstream STP)...... 153 Table 6.4 Hat Hill Creek Chironomidae species abundance results recorded from each of the five sampling occasions (T1 to T5) from the exuviae survey December 2003 to March 2004...... 154 Table 6.5 F-statistics and associated probabilities from analyses of variance of number of species of exuviae (log X + 1 transformed) and number of exuviae (log X + 1 transformed) , from three sampling occasions at Victoria Creek (reference site) and at the two Grose River sites immediately below Dalpura Creek (mine leachate affected). December 2003 to February 2004...... 156 Table 6.6 Abundance of chironomid species recorded at the reference site (Victoria Creek) and two sites on the Grose River (both subject to zinc-rich mine drainage) below Dalpura Creek, for each of the three sampling occasions (T1, T2 and T3: December 2003 to February 2004)...... 157 Table 6.7 R-statistics (Clarke 1993) from two-way crossed ANOSIM for pairwise comparison of all Grose system sites for square-root transformed chironomid species data. chironomid species data from both reference sites were combined. (NS, P>0.05; * 0.01

ix Table 6.8 The 10 most influential species contributing to the difference between the chironomid assemblages at the reference sites (combined) compared to Hat Hill Creek, downstream of Blackheath STP, according to SIMPER breakdown...... 160 Table 6.9 The 10 most influential species contributing to the difference between the chironomid assemblages at the reference sites (combined) compared to the Grose River, downstream Dalpura Creek (mine leachate), according to SIMPER breakdown...... 160 Table 6.10 Zinc and sewage pollution affinity of pollution indicator chironomid species, according to the Grose River survey (see Table 6.12 for grades). ND = not detected...... 160 Table 6.11 Pollution affinity calculations for taxonomic groups...... 161

x Table of Figures

Figure 2.1 Location of Lake McKenzie, Jervis Bay. Inset a shows sampling sites on lake shoreline. Inset b indicates location of study area within Australia....23 Figure 2.2. The number of chironomid (black bar) exuviae collected from each three hour sample, over a 24-hour period from Lake McKenzie. The number of Chaoboridae exuviae are indicated by the unshaded bar...... 29 Figure 2.3. The number of chironomid (black bar) exuviae collected from each site (see Figure 2.1) over the 24-hour period from Lake McKenzie. The number of Chaoboridae exuviae at each site, also over 24-hours, are indicated by the unshaded bar...... 30 Figure 2.4. Plates 1 to 4. Lake McKenzie and sampling chironomid exuviae. Plate 1 Lake McKenzie. Plate 2 A sample of chironomid exuviae. Plate 3 and 4 shows the author sampling chironomid exuviae from the shoreline with a hand-net...... 31 Figure 3.1. Map showing four lake geographical regions sampled in this study. Tropical north Queensland had two lake groups: Cape Flattery and Atherton Tablelands. South-east mainland had three groups: Jervis Bay, north-east Victoria and Mt Gambier and south-western Victoria. Tasmania had two groups, see map of Tasmania (Figure 3.2). Fraser Island was a single group...... 42 Figure 3.2 Location of lake groups in Tasmania that were sampled as part of this study. Three coastal dune lake groups were sampled. All of the inland areas were lakes of glacial origin and were mostly located within Conservation Areas (CA) or National Parks (NP)...... 43 Figure 3.3 Historic Mean monthly minimum temperatures (unshaded bar) and Historic mean monthly maximum temperatures (black bar), in degrees Celsius, from weather stations within lake regions in this study. Cooktown represents coastal TNQ, Atherton represents high altitude TNQ, Fraser Island has one station, Jervis Bay represents the northern margin of SEM and Mt Gambier the southern and western margin of SEM, Strahan represents coastal TAS and Liawenee represents high altitude TAS. Source of data, www.bom.gov.au, accessed May 2003...... 44 Figure 3.4 Mean number (back-transformed) of species of Chironomidae collected from each lake within each of the eight lake groups sampled in this study. Star symbol indicates that the group had 5 or more replicates and was included in ANOVA...... 57 Figure 3.5. NMDS ordination of chironomid samples taken from lakes from southern and eastern Australia. Triangles represent samples from lakes. Red represents tropical north Queensland lakes, yellow represents Fraser Island lakes, blue represents south-eastern mainland lakes, green represents Tasmanian lakes. Stress = 0.2...... 58 Figure 3.6. Mean pH of each lake group, plus/minus one standard error. Star symbol indicates that the group had 5 or more replicates and was included in ANOVA...... 68

xi Figure 3.7. Back-transformed mean electrical conductivity (µS/cm) of each lake group, plus/minus one standard error. Star symbol indicates that the group had 5 or more replicates and was included in ANOVA...... 68 Figure 3.8. Plates 1 to 4. Tasmanian lakes. Plate 1 Author sampling exuviae from ‘King Solomons Jewels’ (glacial-derived), Walls of Jerusalem National Park. Plate 2 Hanson Lake (glacial-derived), Cradle Mountain National Park. Plates 3 Platypus Tarn (glacial-derived), Mount Field National Park. Plate 4 Lake Garcia (dune lake), within pine plantation, near Strahan, Tasmania...... 69 Figure 3.9. Plates 5 to 8. South-eastern Australian mainland lakes. Plate 5 Lake Little Beetle (dune lake), north-east Victoria Plate 6 , western Victoria (maar lake), Plate 7 Little Blue Lake, south-eastern South Australia (sinkhole lake). Plate 8 Gouldens Waterhole (sinkhole lake), with irrigation pump housing, south-eastern South Australia...... 70 Figure 3.10. Plates 9 to 12. Fraser Island dune lakes. Plate 9 Yankee Jack Lake (maneuvering field sampling punt through fringing macrophytes). Plate 10 Aerial view of Hidden lake, Plate 11 Processing exuviae samples with field microscope immediately after collection on the shore of Lake Jennings. Plate 12 Lake McKenzie, view of sandy beach along one side of lake...... 71 Figure 3.11. Plates 13 to 16. Fraser Island dune lakes. Plate 13 Lake Wabby. Plate 14 Ocean Lake. Plate 15 Lake Boomagin. Plate 16 Lake Birrbeen...... 72 Figure 3.12. Plates 17 and 18 Tropical north Queensland maar lakes, Atherton Tablelands. Plate 17 Lake Barrine source of photo www.rainforestgallery.com.au/ images/barrine.jpg Plate 18 Lake Eacham source of photo http://www.tesag.jcu.edu.au/subjects/ev3200/FIELD.Plate 14. Plate 19 Field work in dune-field of Cape Flattery, ‘Cape Flattery Silica Mine’. Plate 20 Brown tannic waters of unnamed dune lake at Cape Flattery...... 73 Figure 4.1 Map of survey sites (square symbols), waterways and waste discharge points in the upper Grose River. Approximate catchment boundary is dashed line. Inset shows location of study area in south-eastern Australia. Grose River sites include ‘GEN’ at Engineers Track, ‘GBK’ downstream of Koombandah Brook, ‘GDD’ downstream of Dalpura Ck, ‘GBK’ upstream of Victoria Creek and the lowest Grose River site ‘GHU’ at Hungerfords Track. Hat Hill Creek sites include ‘HHU’ above the STP, ‘HHD’ below the STP and ‘HHG’ above the Grose River. ‘DAL’ is the tributary Dalpura Creek and ‘VIC’ is the tributary Victoria Creek...... 83 Figure 4.2. Historic mean monthly rainfall (mm / month), indicated by unshaded bars, for Mount Boyce weather station (Figure 4.1). Actual monthly rainfall during the study, black bars, was recorded at Mount Boyce from August 2002 to July 2003. Source of data: Commonwealth Bureau of Meterology, (http://www.bom.gov.au/). Accessed April 2004...... 85 Figure 4.3. Historic mean monthly daily minimum (black bar) and maximum (unshaded bar) temperatures (degrees Celcius) for the Katoomba weather station. Source of data: Commonwealth Bureau of Meteorology, (http://www.bom.gov.au/). Accessed April 2004...... 85

xii Figure 4.4 Altitude profile of survey sites in the upper Grose River. The Y axis indicates altitude in AHD in metres. Distance from the top of the Grose watershed (left profile) and Hat Hill watershed (right profile) is indicated. Arrows and names on the Grose River profile indicate the tributaries flowing into the Grose River…...... 88 Figure 4.5 Back-transformed mean abundance of macroinvertebrates collected from sites in the upper Grose River and its tributaries, on each of the three sampling occasions (three bars above each site label), together with the + / - standard error for 5 replicates at each site...... 95 Figure 4.6 Back-Transformed mean taxa richness of macroinvertebrates collected from sites in the upper Grose River and its tributaries, on each of the three sampling occasions (three bars above each site label), together with the + / - standard error for 5 replicates at each site...... 95 Figure 4.7 NMDS ordination of macroinvertebrate data. Each coloured square symbol represents a single macroinvertebrate replicate (one of five) from the Grose River and its tributaries on each of three sampling occasions. Thus each site is represented by 15 coloured squares representing 15 samples. VIC = black, GEN = yellow, GDK = light blue, GDD = pink/purple, GBK = brown, GHU = green, HHU = red, HHD = dark blue, HHG = grey (abbreviations given in Table 4.1)...... 98 Figure 4.8 Back-transformed mean electrical conductivity levels (µS/cm) recorded at each site sampled in the upper Grose River for three sampling occasions April to June 2003, (plus/minus one standard error)...... 101 Figure 4.9 Mean pH levels (pH units) recorded at each site sampled in the upper Grose River for three sampling occasions April to June 2003, (plus/minus one standard error)...... 101

Figure 4.10 Back-transformed mean hardness levels (mg/CaCO3) recorded at each site sampled in the upper Grose River for two sampling occasions April to June 2003, (plus/minus one standard error)...... 102 Figure 4.11 Mean water temperature (°C) recorded at each site sampled in the Upper Grose River on each of the three sampling occasions April to June 2003...... 102 Figure 4.12. Mean Total Nitrogen (in µg/L) recorded at each site sampled in the Upper Grose River for three sampling occasions April to June 2003, plus/minus one standard error. ANZECC (2000) ecosystem protection trigger value is indicated...... 105 Figure 4.13. Mean Total Phosphorus (in µg/L) recorded at each site sampled in the Upper Grose River for three sampling occasions April to June 2003, plus/minus one standard error. ANZECC (2000) ecosystem protection trigger value is indicated...... 105 Figure 4.14. Mean Total Zinc (in µg/L) recorded at each site sampled in the upper Grose River for three sampling occasions April to June 2003, plus/minus one standard error. ANZECC (2000) guideline level for ecosystem protection is indicated...... 106

xiii Figure 4.15. Plates 1 to 5. Reference sites in Grose River system. Plate 1 is Grose River at Engineers track (GEN). Plates 2 and 3 are the Grose River below Koombandah Brook (GDK). Plate 4 is Victoria Creek (VIC) and Plate 5 is Hat Hill Creek (HHU)...... 107 Figure 4.16. Plates 6 to 9. Canyon colliery and Dalpura Creek. Plate 6 shows surface workings at the colliery. Plate 7 shows Dalpura Creek (orange colour) entering Grose River. Plate 8 shows Canyon colliery sealed mine adit. Plate 9 shows the author sampling Dalpura Creek, just before it enters the Grose River...... 108 Figure 4.17. Plates 10 to 13. Blackheath STP and Hat Hill Creek. Plate 10 shows Blackheath STP and one of its trickling filters. Plate 11 shows an open channel of STP effluent cascading overland towards Hat Hill Creek. Plate 12 shows the STP effluent entering Hat Hill Creek (looking upstream). Plate 13 shows Hat Hill Creek (looking downstream) at the point that the STP effluent enters the Creek...... 109 Figure 4.18. Plates 14 to 17. Grose River effluent receiving sites. Plate 14 shows the Grose River substrate, and yabbie, below Dalpura Creek (GDD). Plate 15 shows the Grose River at Hungerfords track (GHU). Plate 16 shows the Grose River at Hungerfords track (GHU). Plate 17 shows collecting physio- chemical readings the Grose River at Burra Korrain (GBK)...... 110 Figure 4.19. Plates 18 to 21. Macroinvertebrates from the Grose River system. Plate 18 shows two Simuliidae (Diptera: Insecta). Plate 19 Hydropsychidae (Trichoptera: Insecta) Plate 20 Leptophlebiidae (Ephemeroptera: Insecta). Plate 21 shows author collecting macroinvertebrates from the Grose River, below Dalpura Creek (‘kick-sampling’)...... 111 Figure 5.1 Map of chironomid exuviae collection sites and waterways in the upper Grose River system. Square symbols indicate sampling sites. Approximate catchment boundary is dashed line. Inset shows location of study area in south eastern Australia...... 122 Figure 5.2 Air temperature (º Celsius) recorded every two hours at Victoria Creek (black) and Hat Hill Creek during the diurnal studies (12 December 2003 = red; 18 December 2003 = green; 2 January 2004 = blue line; 19 February 2004 = orange)...... 124 Figure 5.3 Water temperature (º Celsius) recorded every two hours at Victoria Creek (black) and Hat Hill Creek during the diurnal studies (12 December 2003 = red; 18 December 2003 = green; 2 January 2004 = blue line; 19 February 2004 = orange)...... 124 Figure 5.4 Rainfall recorded (in mm / month), in the months prior to and during this study, from the rainfall station in the study area, Mount Boyce (black bar) from September 2003 to February 2004. Historic mean monthly rainfall indicated by unshaded bars. From www.bom.gov.au accessed May 2004. 125 Figure 5.5. Back-transformed mean number of chironomid exuviae (plus/minus standard error) collected from at Hat Hill Creek for each two hour sampling period, from four 24-hour sampling occasions in summer 2003/04. The base of the figure is shaded to indicate the approximate times of full light (white), darkness (black) and dawn / dusk (grey)...... 130

xiv Figure 5.6. Back-transformed mean number of species of chironomid exuviae (plus/minus standard error) collected from Hat Hill Creek for each two hour sampling period, from four 24-hour sampling occasions in summer 2003/04. The base of the figure is shaded to indicate the approximate times of full light (white), darkness (black) and dawn / dusk (grey)...... 130 Figure 5.7 Total number of chironomid exuviae collected from Victoria Creek for each two hour sampling period, from a single 24-hour sampling occasion, conducted on 31 December 2003 to 1 January 2004. The base of the figure is shaded to indicate the approximate times of full light (white), darkness (black) and dawn / dusk (grey)...... 131 Figure 5.8 Total number of species of chironomid exuviae collected fromt Victoria Creek for each two hour sampling period, from a single 24-hour sampling occasion, conducted on 31 December 2003 to 1 January 2004. The base of the figure is shaded to indicate the approximate times of full light (white), darkness (black) and dawn / dusk (grey)...... 131 Figure 5.9 Total number of chironomid exuviae collected from Hat Hill Creek for each two hour sampling period, from a single 24-hour sampling occasion, conducted on 12-13 December 2003. The base of the figure is shaded to indicate the approximate times of full light (white), darkness (black) and dawn/dusk (grey)...... 132 Figure 5.10 Number of Paratanytarsus nr. furvus exuviae collected from each 2 hour sampling period at Hat Hill Creek. Black bar represents 12 December, unshaded December 18, grey is January 2 and cross-hatched February 19. The base of the figure is shaded to indicate the approximate times of full light (white), darkness (black) and dawn/dusk (grey)...... 133 Figure 5.11 Number of Tanytarsus sp. 5 (nr. spinosus) exuviae collected from each 2 hour sampling period at Hat Hill Creek. Black bar represents 12 December, unshaded December 18, grey is January 2 and cross-hatched February 19. The base of the figure is shaded to indicate the approximate times of full light (white), darkness (black) and dawn/dusk (grey)...... 133 Figure 5.12 Number of most numerous three species of exuviae collected from each 2 hour sampling period at Victoria Creek, 31 December 2003 to 1 January 2004. Black bar represents number of ‘M05’, unshaded represents number of Riethia zeylandica and cross-hatched bar represents number of Tanytarsus nr. bispinosis. The base of the figure is shaded to indicate the approximate times of full light (white), darkness (black) and dawn/dusk (grey)...... 134 Figure 5.13 Victoria Creek. Progressive total number of species collected from the single 24-hour occasion. The line represents the cumulative number of different chironomid species collected in two hourly increments over a single 24-hour occasion...... 135 Figure 5.14 Hat Hill Creek. Progressive total number of species collected from the four 24-hour occasions over the austral summer. The line represents the cumulative number of different chironomid species collected in two hourly increments over four 24-hour occasions...... 135

xv Figure 5.15. Plates 1 to 4. Exuviae Tanytarsus sp. 5 (nr. spinosus) from diurnal study. Plate 1 shows the abdominal segments (ventral view). Plate 2 shows the postero-lateral comb. Plate 3 is a ventral view of abdominal segments II, III and IV. Plate 4 shows the thoracic horn (almost transparent)...... 136 Figure 5.16. Plates 5 to 8. Exuviae from diurnal study. Plate 5 is Paratanytarsus nr. furvus showing the abdominal segments II, III and IV (ventral view). Plate 6 shows the postero-lateral comb of Paratanytarsus nr. furvus. Plate 7 is a lateral-dorsal view of Orthocladiinae species ‘MO5’. Plate 8 is a ventral view of the lowest abdominal segments of Orthocladiinae species ‘MO5’...... 137 Figure 6.1 Map of exuviae survey sites, waterways and waste discharge points in the upper Grose River system. Approximate catchment boundary is dashed line. Inset shows location of study area in south eastern Australia. Sampling sites are indicated by square symbols. Sites on the Grose River below coal mine: GDD (Grose River below Dalpura) and GBK (Grose River at Burra Korain). Reference sites: VIC (Victoria Creek) and HHU (Hat Hill Creek). Hat Hill Creek below STP discharge: HHD...... 145 Figure 6.2 Total monthly rainfall (mm) recorded prior to, and during, this study at Blackheath (black bar). Historic mean monthly rainfall indicated by unshaded bars. Source: www.bom.gov.au accessed May 2004...... 148 Figure 6.3 Total number of chironomid exuviae collected from each site in the Grose River survey after 120 hours of netting. Cross-hatching represents the estimated proportion of exuviae in the fine fraction of the HHD sample...... 150 Figure 6.4 Chironomid species richness at Hat Hill creek upstream of STP (Black bar) and Hat Hill creek downstream of STP (unshaded bar) on each of five sampling occasions (T1 to T5), December 2003 to March 2004...... 153 Figure 6.5 Chironomid species richness at Victoria Creek (reference site is shaded black) and Grose River below Dalpura Creek (GDD = zinc polluted is shaded grey) and Grose River at Burra Korain (GBK = zinc polluted is unshaded) on each of three sampling occasions, December 2003 to February 2004...... 155 Figure 6.6 NMDS ordination of chironomid exuviae species data collected from the Grose River and tributaries, December 2003 to February 2004. Each sample is represented by a triangle. Reference sites are indicated by the blue (site HHU) and red (site VIC) colours. Test sites are indicated by the colours: orange (sewage receiving = HHD), green (mine drainage = GDD) and black (mine drainage = GBK)...... 158 Figure 6.7. Plates 1 to 4. Exuviae from Grose River pollution study. Plate 1 is a lateral-dorsal view of abdominal segments of Eukiefferiella insolita. Plate 2 is a dorsal view of abdominal segments of Nanocladius sp. Plate 3 shows a thoracic horn structure of Paramerina sp.. Plate 4 shows a thoracic horn structure of Pentaneura sp...... 162 Figure 6.8. Plates 5 to 8. Exuviae from Grose River pollution study. Plate 5 is a dorsal view of the lowest abdominal segments of Orthocladiinae species ‘SO4’. Plate 6 is a dorsal view Thiennemaniella sp. Plate 7 is a dorsal view of abdominal segments II, III, IV and V of Stempellina ? australliensis. Plate 8 is a dorsal view of the lowest abdominal segments and postero-lateral combs of Stempellina ? australliensis...... 163

xvi Acknowledgements

I have been very fortunate to have the support and encouragement of a large number of people in carrying out this research. Firstly, I thank Associate Professor Shelley Burgin for her enthusiasm, tireless efforts and patient supervision. Shelley was responsible for helping me plan the middle to final stages of the research and in detailed editing numerous drafts of the thesis.

Dr Peter Cranston (ANU and CSIRO Entomology) generously gave me access to his laboratory and chironomid collection and taught me skills in identification of chironomid exuviae. He also helped me design and provided much encouragement and support in the initial plans for the thesis and field work, and the first two chapters. I also acknowledge the outstanding field work and ancillary support that ANU School of Botany and Zoology provided my research for the first two Chapters of this research. I am indebted to CSIRO Division of Entomology for providing me with laboratory space and access to their collection.

National Park authorities in Tasmania, Victoria, South Australia, NSW, Queensland and Jervis Bay Commonwealth Territory were extremely helpful and gave me permission to collect samples, and when requested, all happily provided information and practical field support. It gave me a first-hand appreciation of the challenges of their job. The Hopevale community generously gave me permission to collect samples from their land. The Cape Flattery Silica Mine kindly provided accommodation and access to their mining area.

Thanks to the small army of volunteer field workers that willingly gave up their time to help me collect samples. In particular I thank Janet Pritchard, Stuart Swinton, Peter Swanson, Ben Gunn, Ian Reid, Liz Stravinsky, Jack Wolfenden, Hamish Manzi, Paul Hammond, Helen Nice and Susan Wright. I must single Peter Swanson and Paul Hammond out, for also volunteering their labours for my earlier MSc. research.

Sydney Water’s Analytical Laboratories provided me with field sampling equipment and laboratory facilities for the last three chapters of this research. Particular thanks go to my colleagues Firoza Azim, Cathy Cunningham, Colin Besley, Dr Marcus Scammel, Dr Peter Tate, Mark Hopkins, Louise Ledwich, Dr Helen Nice, and David Holleley for their assistance. Dr Peter Cox was my very

xvii kind and patient manager and I also thank my employer, Sydney Water, for giving me leave to pursue this PhD.

Through my PhD studentship I have come to appreciate the wider support and companionship of my friends and family more and more as time went by. As a part-time student, for much of my PhD, life frequently became very difficult as study, employment and the rest of life competed for my very limited time and efforts. I apologise to all parties that may have suffered from my inability to cope at times.

I thank my parents for their friendship, encouragement and practical assistance. I have come to appreciate how fortunate I am to have such wonderful parents, and thanks for being such a positive influence in all aspects of my life. Thanks mum for reading and commenting on many thesis drafts, and dad for building sampling gear.

I would also like thank some friends that were towers of strength and have frequently provided very practical assistance for my PhD: Dr Helen Nice, Jack Wolfenden and Paul Hammond. Paul and Jack Wolfenden individually helped me find a route into the upper Grose River, in very difficult, steep, and slippery terrain. Without their help and physical exertion I would never have been able to carry out this research. Thanks Helen for being so positive, helpful and being there to read chapters and help with field work. I also am grateful for the many conversations with Kathy Collins about her PhD experience.

Thanks to Dusty, the miracle Kelpie, who provided me with company for many long days that I worked at home on my research, even if you though it was a complete waste of time.

Finally I must express my heartfelt gratitude to my dear wife Susan. We met in the middle of my PhD, and without her, I doubt that I ever would have been able to finish it. Susan helped keep me going in every type of way. She edited much of the thesis, solved computer problems and helped with field work when I was not able to find any other volunteers. I do not know why it always seemed to rain when ever she did field work with me. Thanks Susan for your love, care and support. I love you very much. I owe you.

xviii Statement of Originality

I hereby certify that this work submitted to the University of Western Sydney in fulfilment of the degree of Doctor of Philosophy is my original work except where sources are acknowledged and that I have not submitted it for credit towards a degree at any other institution.

Ian Alexander Wright

xix Chapter One

1.1 General Introduction

Chironomids are true flies in the order Diptera (Class Insecta). They are commonly known as ‘non-biting midges’. Juvenile stages live predominantly in freshwater habitats, often in enormous numbers (Coffman & Ferrington,1984). They are very species-rich with an estimated 15,000 species of chironomids existing worldwide (Armitage et al., 1995; Cranston, 1995a). They are one of the most abundant invertebrates in freshwater, occupying an enormous variety of conditions worldwide, from arctic regions and very high altitudes to all temperate and tropical regions (Cranston, 1995a).

Chironomids have four life stages: egg, larva, pupa and adult. The last two stages are generally short and adults predominantly rely on energy stored as larvae to achieve reproduction (Tokeshi, 1995). Immediately following copulation, the eggs are laid then the female soon dies (Pinder, 1995). Eggs are laid in large batches, often several hundred in masses near or on water, usually attached to solid objects at the water’s edge. The first instar, sometimes called ‘larvula’, is often present in enormous numbers. They are very small (often less than 125 microns) and have a partly pelagic behaviour (Pinder, 1995).

Chironomids have up to five generations per year, and life cycles can take as long as seven years to complete (several authors, reported in Tokeshi, 1995). A common cosmopolitan northern hemisphere species Chironomus plumosus can have up to three generations per year (observed in Japan: Iwakuma, 1986, in Tokeshi, 1995) and has a two year life cycle in a high-altitude lake in France (Laville 1972 in Tokeshi, 1995). Growth rates are related to temperature (Nolte, 1994; Edward, 1986) with higher growth rates and shorter lifecycles associated with warmer weather. Studies have found that the number of generations per year has been observed to be negatively correlated with latitude (Tokeshi, 1995).

The larval chironomid life stages occupy a huge variety of habitats, particularly associated with freshwater. They have been collected from all lotic (running waters) including glacier-fed streams, upland streams, lowland rivers, lentic

1 waters (lakes), thermal springs, temporary streams, temporary rain-pools, water collected in leaf axils, brackish waters and marine (intertidal) habitats (Pinder (1995). Some species have also been found living in terrestrial areas, particularly in wet soil or damp vegetation.

The density of chironomid eggs and larvae has been observed in very large numbers. A study by Iwakuma (1986, in Tokeshi, 1995), illustrated both their extremely high density and the very high inter-generation mortality. The research documented juvenile phases of one species Tokunagayusurika akamusi from Lake Kasumigaura, Japan. It recorded a density of 693,200 (estimated) eggs per square metre, which resulted in 923 adults per square metre, of which 518 were female.

Chironomids are responsible for a large proportion of the turnover of food resources within freshwater ecosystems. The larvae, in particular, are an important food source in most aquatic ecosystems (Tokeshi, 1995) and they are prey to a wide variety of predators including other invertebrates, such as stoneflies, mayflies, caddisflies and other dipterans and of fish, turtles and water birds (Maher & Carpenter, 1984). Chironomid species from the sub-family Tanypodinae are themselves predatory. Their prey includes other chironomids, Ostracods, Cladocerans and Oligochaetes (Tokeshi, 1995). More generally the food-source of chironomids is mostly fine particles, such as detritus and algae as well as coarse detritus, such as wood and leaves (Coffman, 1995). Within lotic systems chironomid communities usually have representative species that are predators, herbivores, detritivores and fungivores (Coffman, 1995).

Chironomids have very high species-richness, often they make up 50 % or more of all aquatic macroinvertebrate species within a waterway (Coffman & Ferrington, 1984). Even higher levels of chironomid dominance have been recorded in Australian waterways (Jolly & Chapman, 1966; Arthington et al., 1982).

Some chironomid species have a well-known ability to tolerate extreme environments through remarkable physiological adaptations, and thus some have a well-known tolerance of highly polluted waters (Hynes, 1960). For example, some larval species are called ‘bloodworms’ due to their red colour from haemoglobin, which gives them an ability to tolerate very low levels of

2 dissolved oxygen (Cranston, 1995a). Others live in small tubes, and with physical movement, are able to counter the microstratification of oxygen depletion at the sediment/lake interface (Lindegaard, 1995). Some species also have the ability to switch to anaerobic respiration combined with dormancy under anoxic conditions (Lindegaard, 1995).

1.2 Northern Hemisphere Chironomid studies

The northern hemisphere has the best known chironomid fauna in the world thanks to a rich heritage of ecological, biogeographical and taxonomic studies of all types of aquatic environments, and not surprisingly, has benefited because the majority of the world’s chironomid taxonomists are European (Cranston, 1995d). A comparison of world-wide national chironomid faunal inventories included 24 publications for countries in the Palaearctic region; the Oriental region has the next greatest number with 12 publications. In comparison, Australia had no national chironomid faunal inventories (Cranston, 1995b).

In addition, the northern hemisphere hosted the origin of the modern science of limnology, in which chironomids soon played an influential role in pollution studies of running waters (Hellawell, 1986). For example, Chironomus larvae were shown to increase in density according to organic loading, and thus became an indicator species in the ‘saprobien system’ (Kolkwitz & Marsson, 1909). This system heralded the emerging branch of science that studied water pollution by describing predictable biological and chemical changes that occurred in flowing waters when organic wastes were discharged. According to the saprobien system, stream biology and chemistry changed in predictable biological zones when organic wastes are discharged (Hellawell, 1986). A huge literature followed concerning pollution studies in northern hemisphere waterways, the majority of which emphasised the important roles played by chironomids as pollution indicator organisms (e.g. Hynes, 1960; Hellawell, 1986).

Chironomids are often included in northern hemisphere biomonitoring methodologies as pollution indicators, particularly when the increased density of Chironomus riparius (Meigen) = Chironomus thummi (Keiffer) is associated with

3 increased organic pollution (Lindegaard, 1995). However, many biomonitoring systems leave chironomids at the family or genus level due to taxonomic difficulties with identification of larvae (Lindegaard, 1995).

In 1964, the Trent Biotic Index provided estimates of the pollution tolerance of many macroinvertebrate groups (Cairns & Pratt, 1993). Chironomids were given a pollution tolerance score of two out of 10, signifying their relatively high level of pollution tolerance (Cairns & Pratt, 1993). Many versions of this index were later developed for different European regions (Cairns & Pratt, 1993). The complex relationship between water quality and species of chironomid was sufficiently well understood in the northern hemisphere, as shown by the publication in 1993 by Rosenburg and Resh. They published water pollution tolerances, to organic enrichment and acidity, of 169 chironomid species.

Several northern hemisphere studies have also documented that heavy-metal pollution of streams and rivers causes a reduction in chironomid species richness (Winner et al. 1980; Armitage & Blackburn, 1985; Sheehan & Knight, 1985; Yasuno et al., 1985; Wilson, 1988; Ruse et al., 2000). The proportion of chironomids in macroinvertebrate community samples was proposed by Winner et al. (1980) as a metal-pollution indicator. In a study of macroinvertebrate communities in a small second-order copper and zinc-polluted stream in California, United States of America (USA), Sheehan and Knight (1985) found that chironomids accounted for about 10 % of the abundance of reference sites, and up to 80 % of the metal-polluted sites. They did note that the abundance of chironomids in metal-polluted sites was variable, but less than at reference sites at times (Sheehan & Knight, 1985).

Europe also hosted the more recent discovery by Wilson and Bright (1973) and Wilson and McGill (1977) that the study of chironomid communities alone, using collections of floating exuviae rather than larvae, showed great promise as a methodology for assessment of running water pollution. Also in northern Europe and early in the twentieth century, pollution studies of freshwater lakes discovered that chironomid species were reflective of lake organic enrichment (Thienemann, 1922). The study of lake pollution, organic enrichment and the study of lake chironomid fauna subsequently became a very popular research area in northern hemisphere lakes, with a very substantial body of literature

4 generated (e.g Brundin, 1949; Brinkhurst, 1974; Sæther, 1979; Lindegaard, 1995; Pinder 1995).

Sæther (1979) established that there were distinct chironomid communities associated with each of 15 different trophic (nutrient) levels. He discovered that each of these communities was correlated with lake chlorophyll-a and phosphorus (Sæther, 1979). The lake classification was a measure of the intensity of primary production of lakes, just as the saprobic system is a measure of the intensity of decomposition of organic material, usually in streams and rivers (Hynes, 1960). Chironomid species have been observed to differ according to lake trophic status, food resources, lake physical characteristics and dissolved oxygen levels. Oligotrophic lakes were one extreme, having high levels of oxygen and scarce food resources. Highly eutrophic lakes were the other extreme, with high levels of food resources and low levels of oxygen (Sæther, 1979).

Such lake classification schemes, as described by Sæther (1979), incorporated detailed knowledge about the chironomid species living in all varieties of lake and lake chemistry. The limited biogeographic distribution of chironomid species meant that the lake classification scheme did not apply equally well in different parts of the world (e.g. Brundin, 1958). Sæther (1975) recorded that early attempts to create a similar scheme in North America had exactly such a problem. He reported that insufficient studies had been undertaken in North America to test the applicability of the classification. It was concluded that ‘with proper consideration of the zoogeographical distribution, temperature regimes and other ecological niche parameters, the species composition of the chironomid community of a lake can help evaluate lake types and the stresses occurring within the lake’ (Sæther, 1975, p3131).

Chironomid species richness of northern European lakes was as high as 144 species in alpine lakes in Germany (Thienneman, 1936 reported in Humphries, 1938) and 86 species from a large German lake, Grosser Plöner See (Humphries, 1938).

The majority of published studies of chironomids (and other benthic communities) in lakes have utilised sampling of benthic sediments (e.g. Aagaard, 1986; Sæther, 1967; Fulton, 1983a, b ; Timms, 1978) and most

5 researchers take replicated samples within a different range of depth zones in lakes. The larval stages of chironomids are the life-stage that is predominantly collected from sediment sampling. They are most commonly sampled with the use of a ‘grab-sampler’ lowered from a boat into the lake benthos, which removes a sample of sediment (e.g Fulton, 1983 a and b; Timms, 1978). Some researchers (Flannagan, 1970; Int Panis, 1995) have cast considerable doubt on the efficacy of this method of sampling. For example, several grab sampling techniques and manual SCUBA (self-contained underwater breathing apparatus) collected sediment cores were compared and manual SCUBA cores were found to be the most reliable and effective method of collection (Flannagan, 1970). It was noted that grab samples often disturbed and lost much of the fine sediment fraction with a corresponding loss of biota.

The collection of pupal exuviae has been advocated as an effective sampling technique for assessing chironomid species in lakes (Thienneman, 1910; Humphries, 1938; Franquet & Pont, 1996) and flowing waters (Wilson & Bright, 1973; Wilson, 1977). Exuviae have better taxonomic features for species level identification, particularly in comparison with early larval instars (Hardwick et al., 1995). Collection of exuviae integrates chironomid species from all waterway habitats that the larvae occupy. This is in contrast to most larval sampling techniques, such as sediment sampling in lakes and surber or kick sampling in streams and rivers. These sampling approaches only represent a sub-set of habitats within a waterway, potentially missing many species (Franquet & Pont,1996). Sampling exuviae can require less effort and can cause less disturbance to a waterway. Their collection also provides information on emergence of adults rather than the larval stages. Exuviae collection also does not kill chironomids in the process.

There have been some disadvantages of exuviae chironomid collections documented, including the observation that emergence patterns may preclude some species at the time of collecting, results may depend upon sampling effort, operator bias, weather conditions and hydrological conditions (Kuijpers et al., 1992).

Although Thiennemann (1910) recognised the collection of chironomid exuviae from flowing water as an excellent technique for quickly and effectively

6 surveying chironomids from streams, it was many decades before this technique was used for conducting chironomid pollution surveys. In 1973, Wilson and Bright demonstrated the value of exclusively collecting chironomids from rivers and streams for pollution studies using pupal exuviae. They assessed several aspects of chironomid exuviae variability, including the impact of sewage, and detected super abundance of Chironomus thummi after the release of ‘poor quality’ sewage effluent wastes.

Wilson and McGill (1977) confirmed the efficiency of using chironomid exuviae in pollution studies. They used the collection of exuviae upstream and downstream, at multiple sites, of a sewage discharge into the River Chew in Avon, United Kingdom (U.K.). They reported that the species Chironomus riparius was dominant immediately downstream of the discharge and this well known observation was used as an indicator in the Saprobien system (Kolkwitz & Marsson, 1909; Hynes, 1960; Hellwell, 1986). They also reported a change in the chironomid assemblages from various sites upstream and downstream, with an increased number of species and abundance immediately below the discharge.

Numerous other international water pollution surveys have demonstrated the value of conducting pollution studies using chironomid exuviae. They include a five-year survey of the River Meuse in Belgium and The Netherlands, (Frantzen, 1992) with pollution of waterways classified using a chironomid pollution index (developed by Wilson & McGill, 1982). Bazerque et al. (1989) studied chironomid exuviae and water quality in the Garonne River (France). They also developed a chironomid biotic index using 26 species, based on one of five pollution tolerance classes. Chironomid exuviae have also been used to assess heavy-metal pollution, in the Arkansas River, in the United States of America (Ruse et al., 2000). The impact of agriculture on water quality was investigated in Ontario, Canada, also using exuviae and larvae (Barton, et al., 1994).

Despite their widespread usage in pollution studies, relatively few studies have assessed the spatial and temporal variability of chironomid exuviae in streams and rivers (Wilson & Bright, 1973; Evrard, 1994; Hardwick et al., 1996). One of the greatest concerns with the survey of chironomid exuviae identified by Wilson and Bright (1973) was the distance that exuviae may drift downstream.

7 They were concerned that such drift may ‘contaminate’ downstream sampling sites. To investigate this problem they undertook an experimental release of thousands of small polystyrene balls and monitored their progress downstream. They observed that most of the balls were collected along the edge of the streams, and that they were only recovered in a clumped distribution 150 metres downstream in summer, and 400 metres downstream in winter. They attributed the summer versus winter difference in distance to the influence of more abundant growth of water-plants in summer, which ‘trapped’ drifting balls. They also tested the durability of exuviae left in a waterway for an extended period of time and concluded that they would remain intact for up to two weeks.

Evrard (1994) collected pupal exuviae from the River Samson, Belgium, in autumn. This research was based on collections over 24-hours and 50 species and 11,600 individual exuviae were collected. While some species were collected continuously over the 24-hour period, most were collected in a peak after sunset, with a smaller dawn peak. This temporal variation of exuviae capture was attributed to diurnal emergence of chironomid adults discarding their pupal exuviae.

Chironomid emergence patterns are strongly influenced by latitude. A small Canadian arctic lake, with a brief summer period, experienced the majority of the annual chironomid emergence in a few weeks of the year (Oliver, 1968). In northern temperate zones many studies, reviewed by Armitage (1995), established that less seasonal emergence patterns exist where emergence takes place over five to ten months. Much fewer data were available to determine seasonal emergence patterns in the southern hemisphere. Boothroyd (1988) discovered that year-round emergence occurred in New Zealand streams. Unpublished Australian data reported from the (Cranston & Hillman, in Armitage, 1995) and southern Queensland streams (Cranston et. al., in Armitage, 1995) also revealed year-round emergence patterns.

Temporal and spatial variability in chironomid emergence has also been revealed in a variety of freshwater studies (Wartinbee, 1979; Learner et al., 1990). Learner et al.(1990) established six temporal emergence categories for individual species. They included continuous 24-hour emergence (some with peaks at different times), other species which exhibit a morning peak

8 emergence, and crepuscular species that have peak emergence at dusk. Numerous species have nocturnal emergence. Several studies have also established environmental cues that control emergence patterns (Armitage, 1995). The most influential factors are light intensity and water temperature (Wartinbee, 1979).

1.3 Australian Chironomid studies

Historically, the chironomids of Australian lakes and rivers have been poorly studied (Edward, 1986), particularly in comparison to northern-hemisphere areas, such as Europe (Cranston, 1995b).

In 1986, Edward concluded ‘Little is known about lotic-water chironomids in Australia’ (Edward, 1986, p162). In a broad review of freshwater ecosystem research in Australia at the time, Lake (1987) concluded that chironomids were the highest priority for further study because of the need to improve knowledge on their taxonomy and distribution. This provides an indication that pre-1980 Australian species-level identification of chironomids needs to be carefully interpreted.

The biota of a modest number of Australian lakes and rivers has been studied to the species level, usually as part of whole macroinvertebrate community collections. Frequently, chironomids have been discarded from samples, or simply identified at the family level. One reason for the limited number of Australian chironomids species-level studies was a lack of comprehensive taxonomic keys for their identification (Edward, 1986; Lake, 1987).

Three key studies provided the earliest evidence that high levels of chironomid species richness existed in Australian streams (Arthington, et al., 1982; Bunn et al., 1986; Pearson et al., 1986). One of Australia’s most detailed studies of water pollution was conducted in a south-eastern Queensland stream, and yielded 42 larval and 26 pupal species (Arthington, et al., 1982). A study of stream biota in Jarrah forest streams in Western Australia detected 32 chironomid species (Bunn et al., 1986), while an investigation of tropical rainforest streams in north Queensland recorded one of the most species-rich chironomid results from Australian waters with 65 larval species (Pearson et al., 1986).

9 More recently, two chironomid taxonomic guidebooks have provided the most detailed Australian taxonomic descriptions and keys to date (Cranston 1996, 2000a), making accurate species-level identification of chironomids possible for freshwater researchers in Australia. A large proportion of the Australian chironomid fauna is still undescribed and sampling methodologies require the use of morphospecies to differentiate the undescribed species. Cranston (2000) presented summary data on chironomid species richness in lakes and rivers throughout several regions within Australia, which was higher than previous studies had identified. He estimated that the regional pool of chironomid species, in any region of Australia, was probably in the range of 100 to 150 species compared to 174 to 246 species in other parts of the world.

Only a few Australian studies (Hardwick et al., 1995; Cranston et al. 1997, Wright & Cranston, 2000, Dimitriadis and Cranston, 2001) have collected Chironomidae using exuviae in flowing or standing waters.

The relationship of chironomids to water quality is more poorly understood in Australia than in Europe and North America (Cranston, 1995b). Most Australian water pollution studies that have included chironomid data did so as part of macroinvertebrate community studies, rather than targeting only chironomids (Jolly & Chapman, 1966; Campbell, 1978; Nicholas & Thomas, 1978; Arthington et al., 1982; Norris et al., 1982; Pearson & Penridge, 1987; Cosser, 1988; Hardwick et al., 1995; Smith & Cranston, 1994; Chessman, 1995; Growns et al.,1995; Ward et al., 1995; Wright, 1994; Wright et al., 1995; Chessman, 2003; Norris & Sloane, 2003). This may result in an underestimate of chironomid species from study locations (Hardwick et al., 1995).

A landmark Australian aquatic biota pollution study by Jolly and Chapman (1966), distinguished ‘at least 11 species’ of chironomids in Farmers Creek and Coxs River, in the NSW Blue Mountains. They also found that chironomids were the most numerically abundant taxonomic group in their study, at sites disturbed by human activities and particularly organically polluted sites (Jolly & Chapman, 1966). Their response to pollution in Australian waterways was described in more detail by Arthington et al. (1982). They showed that species-richness declined in the most polluted river sections, but chironomid abundance also increased in the same sections. The increased chironomid abundance was

10 attributed to the numerical dominance of four very pollution tolerant species. Their study also produced the first Australian evidence that some species of chironomids were intolerant of water pollution. A later study of organic pollution from sugar waste in a tropical Queensland stream by Pearson and Penridge (1987) showed that chironomids dominated the macroinvertebrate community, particularly at the most polluted sites. They also observed that a single species of Chironomus comprised up to 83.5 % of the total community at the most polluted site on one occasion.

Other sewage pollution studies in the Blue Mountains area, New South Wales (NSW), confirmed that chironomid species richness declined and abundance increased in sewage treatment plant (STP) receiving waters (Wright, 1994; Hardwick et al., 1994; Wright et al., 1995). In a species-level macroinvertebrate study the impact of an STP on Blue Mountain Creek, near Wentworth Falls, was investigated (Wright, 1994; Wright et al.,1995). In that larval study 22 chironomid species were found across three pristine stream sites and 12 species were collected at the site polluted by treated sewage effluent (Wright, 1994). In a macroinvertebrate study of 44 sites in the Blue Mountains, many of which received STP effluent, a total of 60 larval chironomid species was recorded and it was found that chironomid abundance was higher at sewage impacted sites than at the non-impacted areas (Hardwick et al., 1994). chironomid species richness in that study did not vary significantly between pristine and STP affected sites (Hardwick et al., 1994).

A popular Australia biotic index for freshwater macroinvertebrates, the SIGNAL (Stream Invertebrate Grade Number Average Level) index, was calibrated for organic pollution in the Hawkesbury-Nepean catchment, NSW (Chessman, 1995). Tolerance scores for macroinvertebrate families were allocated between one and 10 (one = most pollution tolerant, 10 = most pollution sensitive). The index classified the pollution tolerance of chironomids in the most pollution tolerant category (i.e one out of 10). In subsequent versions the SIGNAL index has been refined to include Chironomidae sub-family organic pollution tolerance scores (Chessman, 2003). Aphronteniinae was the most pollution sensitive (eight out of 10) chironomid sub-family. In contrast, the other sub-families Tanypodinae and Orthocladiinae were both moderately pollution tolerant (four out of 10), with Chironominae classified as the most tolerant (three out of 10).

11 In several Australian studies data have been collected on the response of chironomids, and other macroinvertebrate groups, to heavy-metal pollution associated with drainage from existing or abandoned mines. An early macroinvertebrate and chemistry investigation of the heavy-metal polluted Molongolo River, NSW, revealed that chironomids were one of the few taxonomic groups found in the most polluted sections of the river, although they were present at greatly reduced levels of abundance (Nicholas & Thomas, 1978). A later study of the Molongolo River showed that chironomid sub-families (Orthocladiinae, Chironominae and Tanypodinae) were amongst the few taxonomic groups to tolerate the most heavily polluted sections of the river (Sloane & Norris, 2003). It was also reported that in the heavily metal-polluted River Dee in Central Queensland, chironomids were one of the few taxa to tolerate the most highly polluted sites (Mackey,1988). A study of the macroinvertebrates of the metal-polluted South Esk River in Tasmania did not show a reduction in chironomid abundance or species richness at the polluted sites, and the authors concluded that they were tolerant of heavy metals (Norris et al., 1982).

Only one Australian water pollution investigation has focused exclusively on chironomids and it showed inconclusive results (Cranston et al., 1997). This study was conducted in tropical northern Australia. Drifting chironomid exuviae were collected in an investigation of heavy-metal pollution from mine drainage. An increase in chironomid abundance and species richness below the heavy- metal rich acid mine drainage was recorded with less than 10 species upstream and more than 40 species downstream of the mine discharge. This appears to be the only study in the world that has recorded a positive relationship between chironomid species richness and exposure to heavy-metal pollution. An earlier study of the same waterway had similar results. It recorded 17 larval species above the mine drainage and 20 to 26 species at the sites below the mine (Smith & Cranston, 1993). That study also detected a greater chironomid percentage contribution to the macroinvertebrate community at metal polluted sites than at those that were less polluted (Smith & Cranston, 1993).

A study of heavy metal polluted stream biota in New Zealand (Hickey & Clements, 1998) also found that chironomid abundance was low at some metal polluted sites, while very high at others.

12 Relatively few Australian pollution studies have explored the chironomid response at the species-level, and only a single study has focused exclusively on the response of chironomids to water pollution (Cranston et al. 1997). For such a widespread, abundant and speciose group, species-level studies are required to understand the relationship fully, but it has been historically considered that there are numerous difficulties in conducting species level sampling and identification (Edward, 1986; Lake, 1987).

As previously indicated, few Australian studies have been specifically conducted on freshwater lake chironomids to the species level (Wright & Cranston, 2000; Dimitriadis & Cranston, 2001; Cranston & Dimitriadis, 2004). While many Australian inland lakes are saline, studies have revealed that their biota comprises few chironomid species and they are often highly abundant (Patterson & Walker, 1974; Williams, 1981; Timms, 1981; Edward, 1983). Timms has undertaken widespread collections of lake biota from Australian freshwater lakes over three decades (many publications, see Table 3.2). Some of his studies have included species assessment of chironomids (Timms, 1985) along with other macroinvertebrate and zooplankton groups. Timms (1978) conducted a detailed study of sediment biota from seven large glacial derived lakes in Tasmania. Fulton (1983 a and b) also conducted a detailed examination of lake benthos of three large Tasmanian glacial-derived lakes, which included an assessment of chironomid species. These studies yielded a total of 21 chironomid species across the three lakes (Fulton, 1983 a and b), while Timms recorded 19 species from his study of seven large lakes (Timms, 1978). The greatest number of chironomid species recorded from a single lake in Australia was 30 from Lake McKenzie (Chapter 2; Wright & Cranston, 2000) and the next largest was 28 species from Lake Eacham (Cranston & Dimitriadis, 2004). Both studies were based on collections of exuviae.

Some chironomid studies have been undertaken in , near coastal townships of Western Australia, where periodic emergence of chironomids in swarms has become a nuisance to humans in the Perth area (Edward 1986; Pinder, 1991). In other countries similar massed emergence has caused medical problems for humans, due to inhalation of adult chironomids and subsequent allergic reactions (e.g. Cranston, 1995c).

13 Fossilised chironomid larval remains in the sediments of lakes have been studied throughout the world for paleoenvironmental research. Other forms of biota that can also be fossilised in lake sediment are pollen, ostracods, diatoms and cladocera (Walker 1995). Changes in chironomid assemblages over time can be recorded in fossil sediments and these data have assisted researchers in interpreting previous lake environmental conditions. A small number of Australian researchers have studied chironomid sub-fossils in lake sediments (Patterson & Walker, 1974; Dimitriadis & Cranston, 2001). One of the most recent studies (Dimitriadis & Cranston, 2001), was a study of past climatic conditions based on chironomids in the sediments of Lake Barrine, tropical north Queensland.

Timms (1985) reported widespread collections of lake benthos from Australian lakes, and concluded that these lakes have few species of chironomids, and other macroinvertebrate groups, compared to the northern hemisphere. He also commented that Tasmanian lakes had fewer species (including chironomids) than northern hemisphere lakes (Timms 1978). Fulton (1983 a and b) disputed this by making species-richness comparisons with certain northern hemisphere lakes.

Timms (1985) reported that he believed three main factors contributing to his observed low species richness in macroinvertebrates (including chironomids) in Australian lakes were: firstly, that Australia has few freshwater lakes, and only Tasmania has natural deep ones. He suggested that this may have denied macroinvertebrates opportunities for speciation that exist in the lake-rich lands in the northern hemisphere. Secondly, he commented that freshwater lakes, in Australia, are often very harsh environments. This is because some lakes are saline, and others, such as dune lakes are very acidic, with a resulting low level of primary production. Thirdly, Timms (1985) hypothesised that Australian macroinvertebrate and chironomid species occupy broader ecological niches within lake zones, and thus may be regarded as ‘generalists’ in comparison to northern hemisphere ‘specialists’.

Timms (1985) suggested that, in general, the benthic fauna, and apparent depauperate status of biota in Australian lakes compared to northern hemisphere lakes was a result of Australia having few large natural lakes; most

14 Australian lakes are geologically young, giving insufficient time for widespread speciation; and zoogeographic isolation has hampered distribution of lake inhabiting species. He reported that there was higher endemism in Tasmanian lakes than in the rest of Australia.

There has been some debate about the existence of a lake chironomid classification scheme in the southern hemisphere, such as the northern hemisphere scheme reported by Sæther (1979). This was investigated by Lars Brundin (1958) in extensive field work in South America in 1953/54, particularly in Chile and Argentina, using profundal chironomids. He found that a similar classification system was valid, with different endemic southern hemisphere species closely following their northern hemisphere counterparts. Despite this observation doubts about the applicability of lake classification in the southern hemisphere have come from Forsyth (1975) after studies of New Zealand lakes. He concluded that the benthos of New Zealand lakes is generally poor compared to that in the literature from other biogeographical regions. He observed that Odonata and Chironomidae, in particular, were poorly represented. As a result of this he doubted the application of lake classification schemes using chironomids in New Zealand lakes. He concluded that indices of lake population density, total biomass and distribution may have to be used, rather than relying on composition of chironomid species alone.

Only a single Australian study has investigated the spatial and temporal variation in chironomid exuviae from rivers and streams. It was conducted in tropical Australian streams that flowed seasonally according to monsoonal rainfall (Hardwick, et al., 1995). They found that net location at different points across the stream channel influenced exuviae species and abundance collected. They also attributed the distinct diurnal patterns of exuviae species abundance to variations in adult chironomid emergence from the waterways upstream of their nets. This study also conducted one of the two studies of chironomid emergence from Australian waters (Hardwick et al., 1986). The other was a study on the outskirts of Perth, Western Australia, which presented results of adult chironomid emergence of three dominant species, all of which had a post dusk emergence (Pinder et al., 1991). Both Australian studies found that peak of emergence occurred in the evening and night hours.

15 1.4 Scope of this thesis

A chironomid sampling strategy, using pupal-exuviae, is investigated in Chapter 2. This was conducted at Lake McKenzie (Jervis Bay, Commonwealth Territory). Spatial and temporal variation of chironomid and Chaoborid exuviae are investigated. The most appropriate method for the collection of exuviae is determined, particularly as a biogeographical survey method for rapid assessment of lake-dwelling chironomid species.

Chironomid species data from nearly 70 lakes, comprising four lake-types, across the main freshwater lake districts of southern and eastern Australia, is reported and analysed in Chapter 3. The species richness and biogeographic variation of chironomid species assemblages in southern and eastern Australian lakes is evaluated.

A macroinvertebrate community pollution survey is presented in Chapter 4 to evaluate any impacts that two point-source waste discharges are causing to otherwise pristine upland streams. The nature of each waste discharges is different. One is organic (treated sewage) and the other is heavy-metal (mine drainage). Water chemistry indicators are analysed to evaluate the influence and nature of each waste discharge on the water quality of the waterways. The response of major taxonomic groups within the community is examined to determine the relative level of pollution sensitivity of each group to each type of waste. Of particular interest is the response of chironomid larvae to each pollution type in comparison to other taxonomic groups.

Spatial and temporal variation in chironomid exuviae from two pristine upland streams is evaluated in Chapter 5. Drifting exuviae are collected over 24-hour periods and any diurnal variability of exuviae is described to determine if time of day influences exuviae results. Several 24-hour sampling events are described at one waterway, and tested, on one 24-hour occasion, at another to see if the patterns change with different waterways. The influence of climate and flow regime is examined in relation to chironomid results. Results from this study are used to determine an appropriate sampling methodology for the purposes of conducting chironomid exuviae surveys from similar upland streams.

16 Results from a pollution survey, using chironomid exuviae, are presented and analysed in Chapter 6. Data is collected from the same waterways investigated in Chapter 4. The chironomid community and individual species response to sewage and heavy-metal pollution is examined. Chironomid exuviae species results are compared with larvae results collected in Chapter 4, and other macroinvertebrate families. This chapter evaluates the use of chironomid pollution surveys using the collection of exuviae.

The last chapter (Chapter 7) presents a synthesis of results and considers the current state of information on chironomids in Australian lakes and rivers. Recommendations are made about the future use of chironomids as biological indicators of pollution in running water, particularly using the collection of exuviae. Comment is also made about the conservation value and impact of human activities on the waterways assessed in this study.

17 Chapter 2 Development of a method for collecting lake Chironomid exuviae

2.1 Introduction

Determining the species of chironomids living within a lake is problematic. Chironomid larvae occupy a very diverse, and often cryptic, array of habitats, the sampling of which creates several practical difficulties. Thiennemann (1910), one of the founding chironomid researchers, advocated the collection of their exuviae as part of his many studies into lentic chironomids. The value of using exuviae was further reinforced by Lars Brundin, as part of world-wide research into the systematics and zoogeography of lentic and lotic chironomids (Brundin, 1966; Fittkau, 1995). Historically, larval collections involved laborious sediment collection and sample processing, and numerous problems were encountered by limnologists because of inherent difficulties in accurately identifying larvae to species level. This is because the larvae of many species lack suitably distinct morphological features to enable species-level determination, particularly for early instars (Coffman & Ferrington, 1984).

Renewed interest in the collection of exuviae followed the work undertaken by researchers in the United Kingdom, who explored the utility of collecting chironomid exuviae from flowing waters, as part of water pollution studies (Wilson & Bright, 1973; Wilson & McGill, 1977). Few studies of lentic systems have included the collection of exuviae (Humphries, 1938; Kujipers et al., 1992; Pinder & Morley, 1995; Franquet & Pont, 1996). Indeed lentic chironomid researchers generally collect chironomids using larval methods and chironomid lake classification schemes assume that larval collections are undertaken (Sæther 1979). The majority of Australian data on lake-dwelling chironomids has been derived through benthic samples and larvae (e.g. Timms, 1978; Fulton 1983 a and b; Arthington et al., 1986; Norris et al. 1993). The most detailed long-term (five year) Australian study of lake benthos, Lake McKenzie in Jervis Bay Commonwealth Territory, used benthic sediment sampling and assessed chironomids using benthic larvae.

Temporal variation (seasonal and diurnal) of exuviae abundance, and species richness, has been demonstrated, in flowing water studies, to be an issue that

18 exuviae-based collections of chironomids should consider (Wilson & Bright, 1973; Hardwick et al., 1995, Cranston et al., 1997). Many chironomid studies have established that emergence of many chironomid species follow distinct diurnal patterns (Learner et al., 1990). Emergence patterns, and thus time of sampling, may be an important sampling issue as exuviae may only remain afloat for a short time. For example, Wilson and McGill (1973) reported that some exuviae remain on the surface or at the edge of streams for only a few hours.

Only one study (Franquet & Pont, 1996) has been published that attempts to describe and measure the spatial and temporal variability of lentic chironomids using exuviae. It was demonstrated that towing a sampling net, by boat, across the surface of a lentic water body collected more species, and provided a more spatially and temporally heterogeneous pattern, than did ‘emergence traps’ that had been placed over shallow littoral areas. While Franquet and Pont (1996) explored some of the factors that Wilson and McGill (1973) had assessed for lotic systems they did not consider the issue of spatial variation or factors influencing spatial patterns of exuviae. Franquet and Pont (1996) discussed the problem of extrapolating exuviae dynamics from the results of net tows and emergence traps to provide a picture of the whole water body. They emphasised the problems encountered in terms of microhabitats, and implications for a stratified random sampling program. Although they stated that ‘numerous and variable factors could act on the duration of floating of pupal exuviae in open waters’, they did not elaborate further.

Chironomid exuviae have been observed accumulating in very high densities in wind-ward (lee shore) lake provinces (personal observation). The tendency for large numbers of exuviae to aggregate along the lee shoreline of lakes was also observed by Pinder and Morley (1995), although they did not present data related to this phenomenon. Areas of accumulation tended to be associated with irregularities along the edge of lakes, such as small bays, fallen timber and amongst macrophytes (personal observation). Confirmation that exuviae do aggregate along lake shorelines has important implications for chironomid lake studies. Collection of exuviae from locations of shoreline accretion offers an opportunity to collect evidence of species that have emerged from the full spectrum of larval lake habitats.

19 To assess the value of collecting chironomid exuviae from lakes, I sought to investigate temporal and spatial variation of lake-dwelling chironomid exuviae from a single lake. The aim of my investigation was to develop an exuviae sampling methodology suitable for conducting chironomid species surveys from lakes and other standing water bodies.

2.2 The study area

Lake McKenzie (Figure 2.4, Plate 1) (35º 09´ S 150º 41´E) is located in Jervis Bay Commonwealth Territory, south-eastern Australia (Figure 2.1). It is a small (eight hectare) permanent freshwater sand dune lake, about 1.5 kilometres from the ocean. Lake McKenzie is one of the more intensively studied lakes in Australia, it has previously been the subject of a multi-year physio-chemical and benthological investigation (Norris et al., 1993). The lake is mildly acidic (surface pH 5.5-6.5), dilute (160-230 µS/cm), has low primary production and is usually thermally stratified with an anoxic hypolimnion (Norris et al., 1993). It has fewer emergent or floating macrophytes than most Australian dune lakes (Timms, 1997).

The area around Lake McKenzie mainly consists of Pleistocene sand deposits and contains a large amount of groundwater in an aquifer (Jacobson & Schuett, 1984). They also reported that the lake level is subject to considerable variation and suggested that it was formed by drifting sand dunes that blocked a stream.

The climate of the area is mild, with the warmest mean daily maximum temperature in February of 23.9°C and the coolest mean daily minimum temperature in July of 9.2 °C. (www.bom.gov.au/climate/averages/tables/cw_068034.shtml accessed May 2004). Rainfall in the area is usually slightly higher in the months of March to June and mean total annual rainfall is 1245 mm. (www.bom.gov.au/climate/averages/tables/cw_068034.shtml accessed May 2004).

The lake is located within botanic gardens, and is used as an irrigation water supply. It is likely that it receives fertiliser-enriched runoff from the garden area (personal observation).

20 2.3 Materials and Methods

2.3.1 Exuviae collection – spatial study Chironomid and Chaoborid exuviae were collected from eight sites around the margin of Lake McKenzie on 28 April 1997. The sites were selected at approximately eight equal 45° intervals around the lake perimeter to determine the impact of spatial variation (see Figure 2.1).

At each sampling site, approximately 20 metres of shoreline was thoroughly sampled using a 330 µm mesh hand net. The nearshore and shoreline area were gently swept in an arcing motion of about two metres in length. The arc began about 1.5 metres off-shore and the net was swept in curving movement onto the shoreline and back (Figure 2.4, Plates 3 and 4). This sweeping movement was repeated progressively along the length of shoreline sampled, about one sweep per metre of shoreline. This action created turbulence and a small on-shore wave that resuspended beached scum material, often containing exuviae. Following 'sweeps' netted the resuspended material. The sampling technique was repeated with identical sampling effort at each site. Time taken at each site was approximately eight minutes. The entire sampling exercise took about 90 minutes, during which the wind direction remained from the north-east.

At each site the entire contents of the net was removed and immediately preserved in 70 % ethanol and returned to the laboratory. All chironomid and Chaoborid exuviae were later removed from sample detritus, with the aid of a dissecting microscope. Exuviae were grouped (Figure 2.4, Plate 2) under the dissecting microscope (X 40 magnification) and representative specimens were slide mounted in Euparal and identified using a compound microscope (at up to X 400 magnification). Identification was generally to species level, using the latest taxonomic keys for Australian Chironomidae (Cranston, 1996) and Chaoboridae (Colless, 1986). When available, specimens were also compared with reared material in the Australian National Insect Collection, CSIRO Entomology, Canberra.

21 2.3.2 Within-year temporal investigation Samples were collected from Lake McKenzie on three occasions: 9 September 1997, 3 January 1998 and 7 March 1998 to investigate within-year temporal variation. The sampling location depended upon the predominant wind direction and was conducted along the most leeward region of the lake shoreline. A technique equivalent to that explained above was used for sampling, except that about 50 metres of shoreline were netted over approximately 20 minutes of sampling. The contents of the sampling net were again preserved on-site in 70 % ethanol and were later examined in the laboratory. Representative specimens were slide-mounted and identified under a compound microscope, as described previously in section 2.3.1. Taxa were recorded as presence/absence data on the last three occasions. The results of the 28 April 1997 sampling were recast and expressed as taxon presence/absence and included as part of this temporal investigation. Surface water temperature was recorded on the four sampling occasions.

2.3.3 Diel exuviae investigation Exuviae samples were collected from four sites (sites 1, 3, 5 and 7 in Figure 2.1) around the perimeter of Lake McKenzie on one 24-hour occasion (28 April 1997). Samples were collected from the four sites every three hours for the 24 hour period. The sampling method was identical to that described in the previous sections (2.3.1 and 2.3.3) except that each sample involved five minutes of sweeping effort over a distance of about 10 metres and all exuviae were identified and were recorded as quantitative data.

22

b N

Jervis Bay

a 5.

6.

7. 4.

35º 10'S 8.

3. 2. 1.

Tasman Sea 100 m 150º 40'S

Figure 2.1 Location of Lake McKenzie, Jervis Bay. Inset a shows sampling sites on lake shoreline. Inset b indicates location of study area within Australia.

23 2.4 Results

2.4.1 Spatial heterogeneity In total, 2501 exuviae were recorded, comprising 18 species: 17 chironomids and one Chaoborid. Of all exuviae collected, 1123, or 45 %, were from a single species of Chaoborid, Chaoborus vagus (Table 2.1) and 17 of 18 species (95 %) of the chironomid exuviae were found at one of the sites (Figure 2.1, Site 3). No exuviae were collected at two of the sites (Figure 2.1; Sites 7, 8).

The Tanypod Procladius squamiger was the most abundant chironomid, comprising 72 % of all chironomid exuviae collected (Table 2.1). It was about 12 times more abundant than the second most numerous taxon, Cryptochironomus ? griseodorsum (Table 2.1). The eight most abundant taxa contributed 96.5 % of the total number of chironomid exuviae collected (Table 2.1). The nine least numerous taxa contributed 3.5 % of the total (Table 2.1).

24 Table 2.1 Chironomid and Chaoborid pupal exuviae taxa collected from eight sites around the shoreline perimeter of Lake McKenzie, 28 April 1997 (see Figure 2.1).

Taxon Site 1 Site 2 Site 3 Site 4 Site 5 Site 6 Site 7 Site 8 Total Chironomidae Procladius squamiger 990 1 1 992 Cryptochironomus ? 78 1 3 82 griseodorsum Paramerina sp. 1 72 1 74 Stempellina ? 59 3 62 australiensis Cladotanytarsus sp. 39 8 47 Djalmabatista sp. 6 31 37 Xenochironomus 19 1 20 australiensis Tanytarsus “M1” 16 1 17 Coelopynia pruinose 11 1 12 Polypedilum watsoni 1 9 10 Xenochironomus “S1” 6 6 Paracladopelma “M1” 5 5 Riethia stictoptera 4 1 5 Ablabesmyia notabilis 4 4 Tanytarsus nr. dycei) 4 4 Chironomus “south” 1 1 Chironomini genus “K2” 1 1

Sum of chironomids 1 1 1323 4 48 1 0 0 1378

Chaoboridae

Chaoboris vagus 1 0 928 14 124 56 0 0 1123

Sum of exuviae collected 2 1 2251 18 172 57 0 0 2501

25 2.4.2 Within-year exuviae variation Thirty chironomid and one Chaoborid species were recorded from Lake McKenzie in this study (Table 2.2). Eight of the chironomid and the single Chaoborid species were recorded on all of the four sampling occasions (Table 2.2). Thirteen chironomid species were recorded on either two or three of the sampling occasions. Nine species were recorded only once.

Five of the eight chironomid species present on all sampling occasions (Table 2.2) were among the eight most numerous exuviae (Table 2.1) in April 1997 (i.e. exuviae comprising > 1 % of the total from all sites). Three other species (Ablabesmyia notabilis, Riethia stictoptera and Tanytarsus nr. dycei), present on all sampling occasions, were also present in April 1997 but were less abundant (<1 % of the total collected ). Three species (Xenochironomus australiensis, Tanytarus “M1” and Djalmabatista sp.) were abundant in April 1997, but not recorded on all sampling occasions. They were not recorded on the coolest sampling occasion (September 1997). Xenochironomus australiensis and Djalmabatista sp. were also not recorded in January 1998 (Table 2.2).

Of the nine taxa recorded on only one sampling occasion, six were recorded in the warmest months of January or March 1998 (Table 2.2).

26 Table 2.2. Chironomidae and Chaoboridae pupal exuviae taxa collected from leeward shore of Lake McKenzie on four occasions over an 11 month period. (X signifies the taxon was present. Surface water temperature on the day of sampling is given).

Taxon 28 April 13 3 January 7 March 1997 September 1998 1998 1997

Water temperature ( C) 19 17.5 24 23.5

Chironomidae Coelopynia pruinose X X Clinotanypus crux X X Djalmabatista sp. X X Procladius squamiger X X X X Ablabesmyia notabilis X X X X Paramerina sp. X X X X Nanocladius sp. X X Riethia stictoptera X X X X Cladotanytarsus sp. X X X X Stempellina ?australiensis X X X X Paratanytarsus kathleenae X Tanytarsus nr. dycei X X X X Tanytarsus nr. manlyensis X Tanytarsus “M1” X X X Cladopelma “south” X X Chironomus “south” X X Conochironomus australiensis X Cryptochironomus ? griseodorsum X X X X Dicrotendipes lindae X Paracladopelma “M1” X X X Parachironomus nr. K4 X X Polypedilum vespertinus (M2) X X Polypedilum watsoni X X X Kiefferulus martini X Stenochironomus watsoni X Stictochironomus sp. X Xenochironomus “S1” X X Xenochironomus australiensis X X Xenochironomus “K4” X Chironomini genus “K2” X

Chaoboridae Chaoboris vagus X X X X

Taxa per sampling occasion 19 17 18 20 Cumulative number of chironomid taxa 18 23 26 30

27 2.4.3 Diel exuviae variation A total of 888 exuviae was collected, comprising 576 Chaoborid and 312 chironomid specimens (Table 2.3). A single species of Chaoborid was recorded and 14 chironomid species. The total number of exuviae collected was highest in daylight hours, peaking on the first sampling occasion at 1200 hours (Figure 2.2). The total number of exuviae was much lower at night, lowest (approximately 50 exuviae) at 1800 hours, 2100 hours, 2400 hours and then numbers rose to 80 at 0300 hours, more than three hours prior to sunrise (Figure 2.2, Table 2.3).

Nine species were recorded on the first sampling occasion, and again at 0300 and 0600 hours (Table 2.3). The cumulative species total started at nine species at the conclusion of the first sampling occasion and rose with another ‘new’ species on each sampling occasion until 0600 hours (Table 2.3). Only four species (the lowest) were recorded at 0900 hours, and no ‘new’ species were added to the cumulative total.

Chaoborus vagus (the sole species recorded from the Chaoboridae family) was one of the most abundant species recorded during this segment of the study, and had the most distinct diurnal pattern. Numbers were at their lowest at 1800 - 2400 hours (Figure 2.2, Table 2.3). These numbers progressively rose from 0300 hours and peaked at 1200 hours.

28 Table 2.3 Chironomid and Chaoborid exuviae collected from Lake McKenzie in the 24-hour exuviae study

Time of day 1200 1500 1800 2100 2400 0300 0600 0900 Taxon Coelopynia pruinose 1 1 Procladius paludicola 29 18 11 14 13 45 33 18 Procladius villosimanus 1 Paramerina sp. 14 5 8 11 1 1 13 14 Riethia stictoptera 1 1 1 Cladotanytarsus sp. 1 1 Stempellina ?australiensis 6 2 1 1 1 Tanytarsus nr. dycei 2 3 2 1 1 2 10 Tanytarsus “M1” 2 1 1 Cryptochironomus ? griseodorsum 1 2 1 1 3 1 Paracladopelma “M1” 1 Stictochironomus sp. 1 Polypedilum watsoni 1 2 1 2 Xenochironomus “S1” 1 2 Chaoboris vagus 179 77 26 23 31 37 74 129

Species/time 9 8 6 8 8 9 9 4 Progressive sp list 9 10 11 12 13 14 15 15

250

200

150

100

50

0 1200 1500 1800 2100 2400 0300 0600 0900

Figure 2.2. The number of chironomid (black bar) exuviae collected from each three hour sample, over a 24-hour period from Lake McKenzie. The number of Chaoboridae exuviae are indicated by the unshaded bar.

29 Chironomid species displayed a diurnal pattern in exuviae abundance (Figure 2.2, Table 2.3). The most abundant, Procladius paludicola, was recorded throughout the 24-hour cycle, peaking at 0300 hours (Table 2.3). Paramerina sp., the second most abundant chironomid, was most numerous in daylight hours, and lowest at night (Table 2. 3).

One species, Cryptochironomus ? griseodorsum was recorded mostly at night . Other species were less abundant and thus discrimination of their diurnal patterns was not possible.

2.4.4 Spatial aggregation during 24-hour study The greatest numbers of Chaoborid exuviae were recorded at site 5, then at 7, then 3, then 1. In contrast, most chironomid exuviae were recorded at site 1, then at 3, then 5, then 7.

350

300

250

200

150

100

50

0 Site 1 Site 3 Site 5 Site 7

Figure 2.3. The number of chironomid (black bar) exuviae collected from each site (see Figure 2.1) over the 24-hour period from Lake McKenzie. The number of Chaoboridae exuviae at each site, also over 24-hours, are indicated by the unshaded bar.

30 Figure 2.4. Plate 1 Lake McKenzie. Plate 2 A sample of Chironomid exuviae. Plate 3 and 4 shows the author sampling Chironomid exuviae from the shoreline with a hand-net. Plate 2. Plate 1.

Plate 3. Plate 4. 31 2.5 Discussion

Thirty chironomid species were collected from Lake McKenzie using a simple but effective exuviae sampling technique. This represents the largest collection of chironomid species recorded from a single Australian lake. The second largest number (28) was also based on exuviae collection and was recorded from Lake Barrine in the Queensland Atherton tableland, a much larger and deeper lake than Lake Mc Kenzie (Dimitriadis & Cranston, 2000; Cranston & Dimitriadis, 2004). These data indicate that the exuviae technique used in this study enabled a more complete assessment of the chironomid species present than sampling other stages of the lifecycle. For example, in a previous benthic study of Lake McKenzie, one of the most intensive multiple-year studies of lake benthos ever conducted in Australia, researchers detected only 11 chironomids and one Chaoborid taxa (Norris et al., 1993).

In the current study, the chironomid species inventory benefited from the collection of pupal exuviae, rather than larvae. This approach has a number of advantages when collecting lentic pupal exuviae. The most obvious difference is the ease of collection. Exuviae were collected by sweeping the water along the shoreline with a handnet, in contrast Norris et al. (1993) collected chironomid larvae by taking sediment cores and grab samples. Sediment samples require laborious processing to isolate larvae and thus is much more time consuming than sampling exuviae with a net from the shore. Another advantage is that all pupal chironomids inhabiting a lake must swim to the surface before emerging as adults, and discard their pupal exuviae to float on the water’s surface. Thus sampling exuviae from surface waters provides evidence of recently emerged species from all areas within the lake and combines it into a common sample. In contrast, larval techniques rely on direct sampling of larval habitats. As a lake includes multiple sub-habitats, a researcher that relies on larval sampling will probably miss many occupied habitats and thus underestimate the number of species present. For example, I detected more than twice as many species in a single year study as were collected in the same lake over five years using larvae (Norris et al., 1993). This current study also exploited the third advantage of using exuviae which is that they possess excellent species-distinctive taxonomic features. This is in contrast to many larvae, particularly from the earliest instars, which often have inadequate taxonomic features to enable specific identification.

32 The earlier study of Lake McKenzie (Norris et al., 1993) had two additional limitations that may have resulted in an underestimate of chironomid diversity: generic level of identification and the use of a coarse 500 µm sampling mesh. Despite these limitations, the number of larval chironomid taxa recorded by Norris et al. (1993) was equivalent to the highest numbers collected from Australian dune lakes. Arthington et al. (1986) also collected 10 chironomid taxa from Lake Wabby (Fraser Island, Queensland).

In two studies of lake benthic communities in Tasmania (Timms, 1978; Fulton, 1983 a and b) it was recorded that there were slightly more chironomid taxa than collected from the larval study in Lake McKenzie. In the study of profundal benthic communities of Tasmanian glacial lakes, Timms (1978) detected between 4 and 11 taxa from each of seven lakes. In a temporally replicated study by Fulton (1983 a and b) 14 and 15 chironomid taxa were recorded from two large glacial lakes. Fulton sampled benthos from littoral sediments, the macrophyte zone and the profundal sediments in the two lakes. Numerous lake studies have established that more chironomid species live in the littoral zone of lakes than in the deeper profundal sediments (e.g. Humphries, 1938; Aagard, 1978; Hare & Cater, 1987). These Tasmanian lakes were much larger than Lake McKenzie and, should have yielded more species because of the greater number of potential niches. Exuviae sampling of these lakes is, therefore, likely to reveal more species than previously recorded.

Within Lake McKenzie, greatest species richness was observed at site 3 (Table 2.1, Figure 2.1). This site was located in the south-west corner of the lake, and had been subject to winds, and small on-shore waves from the north-east for several hours on the day of sampling at that site. The observed spatial heterogeneity of exuviae at this site indicates the importance of locating the zone of maximum exuvial aggregation along the shoreline. Contributing factors that influence the abundance of exuviae, apart from the wind strength and direction, include shoreline embayments, areas of floating or emergent macrophytes and other shoreline, or nearshore features that entrap 'driftlines' of floating exuviae, which are often associated with other floating scum (personal observation). Lake McKenzie was unusual in having a uniform sandy shoreline with very few isolated macrophytes. Selection of a poor sampling site could yield few exuviae and an incomplete list of chironomid species. For example, six of the eight sampling sites

33 in this current study yielded less than five individual chironomids, the other two sites contributed 99.5 % of the individuals collected. Ideally the entire shoreline of a lake could be sampled, or at multiple points along the shoreline, but for biodiversity assessment purposes, these results establish that the most effective site for collecting exuviae from a lake is along the lee shoreline.

Floating exuviae were intercepted by Franquet and Pont (1996), through their use of a net tow, behind a boat. Such a technique collects exuviae that have recently emerged and are in the process of being transported, under the influence of wind and waves, towards the shoreline. The species collected by the net tow technique probably depends on the spatial and temporal emergence characteristics of the resident chironomid species. In contrast, this study showed that the collection of shoreline exuviae at the zone of maximum abundance and species richness (Table 2.1) indicated that exuviae had accumulated over a period of time.

Several studies have established the diel periodicity of chironomid emergence (e.g. Learner, 1990) and the temporal variability in exuviae transport has suggested that the transport phase is usually less than three hours (Hardwick et al., 1995). Wilson and Bright (1973) estimated that shoreline exuviae probably remain undamaged in the environment for several days, possibly as long as two weeks. Three-hourly sampling of exuviae, over a 24-hour period from Lake McKenzie confirmed that there was a diurnal pattern of emergence (Table 2.3, Figure 2.2). Chaoboridae exuviae dominated the diurnal collections (576 exuviae versus 312 chironomid exuviae) and had a particularly pronounced diurnal pattern, peaking at 1200 hours. Chironomid exuviae had a less pronounced diurnal pattern and had a peak emergence at 0600 hours. This provides the first evidence that Australian chironomid emergence was higher in daylight hours than during the night. Two Australian studies of chironomid exuviae that involved diurnal collections from running waters (Hardwick et al., 1995) both recorded peak exuviae abundance during the night hours.

The spatial aggregation and the diurnal investigations in this study recorded quite different results (Table 2.1, Figure 2.3). In particular, there was less spatial aggregation recorded over the 24-hour period. The weaker and changing wind conditions during the 24-hour study are assumed to have influenced this result.

34 Sampling exuviae techniques need to consider changing wind direction and possibly sample multiple locations along a lake shoreline (Table 2.1).

Sampling across the year revealed a core of eight species that were always present. This indicates that several chironomid and the one Chaoborid species exhibit a pattern of continuous emergence in this lake (Table 2.2). Another three species were recorded on three of the four sampling occasions and, therefore, also exhibited emergence patterns that extended over many months. While the remaining 18 species may have seasonal patterns of emergence, the evidence was not sufficient to determine their seasonality. Seven of the eight species that were recorded on only one occasion were recorded in one of the three warmest months (January, March and April). Three of these species are known to have north Australian distributions: (Tanytarsus nr. manlyensis (b), Dicrotendipes lindae and Conochironomus australiensis (Cranston & Hare, 1997)). Another, Djalmabatista sp., was recorded on only two of the four sampling occasions, in the relatively warm water temperatures of March and April. This is one of the most southerly lentic records of Djalmabatista sp., which is often found in tropical and sub-tropical lakes (Peter Cranston, personal communication). Evidence indicates, therefore, that these species may be present only in the warmer months in Lake McKenzie.

This study demonstrated that collecting chironomid exuviae is an appropriate methodology for collecting evidence of chironomid species living within a lake. It provides a more complete species inventory than traditional larval techniques and is quicker and easier. However, collection of exuviae requires careful consideration of local weather conditions at the time of sampling. Most important is considering the predominant wind and wave direction, and ensuring that the lake shoreline sampled is located in the most down-wind direction. Although repeated samples need to be undertaken throughout the year to gain the most complete inventory of chironomid species, a single sample of exuviae should provide a reasonable representation of species from that lake, for rapid chironomid survey purposes. Depending on the research question, collection of chironomid exuviae from the lake shoreline may be adapted for multiple research purposes, particularly for biodiversity studies.

35 Chapter 3. Eastern-Australian survey of lake-dwelling chironomids.

3.1 Introduction

Carmel Humphries (1938, p535) commented on the state of northern European knowledge of lake-chironomids. She said: ‘Many ecological surveys of lakes have been made, but in nearly every case the chironomid fauna has either been neglected or only inadequately dealt with. From an ecological as well as a systematic point of view, these investigations are incomplete and of little value because the Chironomidae is one of the most, if not the most, important group of freshwater organisms.’ The same statement could apply in Australia today.

Chironomids are one of the most abundant, species-rich and important groups of biota within Australian lakes (Fulton, 1983a, b; Timms, 1985). However, their assessment has only been included in a modest number of biological studies, particularly at the species level (Tables 3.2, 3.3).

One reason so few species-level studies have been undertaken in Australian lakes is that accurate species identification has been difficult. Until recently there were few taxonomic descriptions and keys available for chironomids (Edward, 1986) to aid accurate species identification. The most common method of sampling lake benthos involved collection of sediment samples, usually through grab samples (numerous references, Table 3.2). Benthos sampling of lakes is thought to miss many of the potential chironomid habitats (Franquet & Pont, 1996), and identification of chironomid larvae is time consuming and fraught with difficulties associated with species-level identification (Coffman & Ferrington, 1984). More recently Peter Cranston has greatly improved the taxonomic keys for larvae and particularly for exuviae stages of chironomids in Australia (Cranston 1996, 2000a) and this has made it more realistic to study chironomids in greater detail.

Australian lake studies that have assessed chironomids to the species-level have been restricted to a single lake, or a small group of lakes, usually within a locality. For example, Tasmanian lakes were sampled by Timms (1978) and Fulton (1983a, b). Arthington et al. (1986) sampled a number of Fraser Island lakes in

36 Queensland. Single lakes were sampled at Jervis Bay (south-eastern Australia) by Norris et al. (1993) and Dimitriadis and Cranston (2001) sampled Lake Barrine in Queensland. The research reported in Chapter 2 was based on a single lake, Lake McKenzie at Jervis Bay (Chapter 2; Wright & Cranston, 2000). With the techniques most widely used (ie. larvae, adults), such studies have generally involved the expenditure of considerable sampling effort on a small regional scale.

Comparing chironomid results in Australia from a variety of studies across different geographical localities, frequently involving different methods and coupled with the often doubtful accuracy of chironomid identifications, has made biogeographical comparisons difficult (Cranston, 1995b; Wright & Cranston, 2000). There is thus limited ability to compare lake-chironomid species on a broader scale in Australia.

The importance of assessment of biogeographical differences in chironomid fauna at a very broad scale has been emphasised by several researchers (Brundin, 1958; Sæther, 1979; Walker & Mathews, 1987; Cranston, 1995b). The lack of such information is in contrast to the research in northern-hemisphere where numerous intensive studies have produced a detailed knowledge of regional chironomid species (Cranston, 1995b), and has resulted in the development of models that relate chironomid fauna to nutrient enrichment, temperature and oxygen level (e.g. Sæther, 1979; Rossaro, 1991; Walker, 1995).

After examination of fauna from lakes in the Andes, Chile and Argentina, a pioneer chironomid researcher, Lars Brundin (1958) suggested that the northern- hemisphere model of chironomid species occurring in lakes, influenced by nutrient levels, also applied to the southern hemisphere. Within Australia and New Zealand some researchers have considered the applicability of northern hemisphere models that have associated assemblages of lake chironomids with nutrient, oxygen and temperature regimes (Forsyth, 1975; Fulton, 1983a, b; Timms, 1985). Timms (1985) concluded that the benthic fauna of Australian lakes was quite different from those of the northern hemisphere. In particular, he noted that there were fewer invertebrate species (particularly chironomids) occupying Australian lakes in comparison with similar northern hemisphere lakes.

37 Few researchers have collected chironomid exuviae from lakes (e.g. Humphries, 1938; Franquet & Pont, 1996; Wright & Cranston, 2000; Dimitriadis & Cranston, 2001). However, as indicated in Chapter 2, sampling chironomid exuviae, which are more readily identified, can result in more than double the number of taxa obtained by larvae from sediment samples. Exuviae have excellent taxonomic features that make species-level identification quick and effective (Coffman & Ferrington, 1984). Exuviae collection is also rapid. This enables the collection of species-inventory information from lakes in a matter of hours that may otherwise take days using larval collection techniques involving chironomid sediment sampling. These data also provide a superior outcome to concentrating on the more time consuming collection of larvae (Wright & Cranston, 2000). As a consequence, exuviae provide a more accurate account of the species that make up assemblages in lakes. Therefore, sampling exuviae is a superior method to use as the basis of biogeographical comparisons.

In this chapter I describe a geographically broad survey of chironomid species from Australian freshwater lakes, based on the collection of chironomid exuviae. Species-assemblage patterns are described over a broad scale, and factors that influence the patterns are investigated. Based on collections from four different lake types, the influence of lake type is also assessed. Biogeographical patterns in chironomid species are also investigated. Finally, in this chapter, I critically evaluate the long-held view of Timms (1985) that Australian lakes, in contrast to northern hemisphere lakes, have impoverished levels of chironomid species- richness, based on his grab sample studies collected from the deeper areas of lakes.

38 3.2 Study locations: freshwater lakes from southern and eastern Australia.

Lakes in this study were surveyed in a north-south pattern along the eastern Australian coast including Tasmania. Lakes were grouped into four geographic regions (Figures 3.1, 3.2).

Four types of lakes were sampled from the major freshwater lake districts of eastern and southern Australia (Figures 3.1, 3.2; Tables 3.1, 3.2). Coastal dune lakes are more common and widespread than other lake types (Timms, 1992). These are scattered in pockets along the Australian eastern and southern coastline, particularly on Fraser Island, and at a few locations along the Cape York coastline (Figure 3.1). The second type of lake in this study was glacial derived lakes within Tasmania. Although common in Tasmania, they are rare in mainland Australia and only five exist (Timms, 1980a). The other two lake types were sinkhole lakes, found in some areas of limestone geology of south-east South Australia (Tyler et al. 1983) and volcanic maar lakes of Western Victoria and nearby South Australia, and two lakes in the Atherton tablelands of tropical north Queensland (Timms, 1992).

The most northerly geographic region was in tropical Queensland and comprised the dune lakes in the Cape Flattery dune field (Hawkins et al., 1988) and two maar lakes in the Atherton tablelands (Timms, 1979). This region included the most northern lakes in the study, 15 south of the equator, at Cape Flattery, on the lower Cape York Peninsula (Figure 3.1; Table 3.1). The second geographic region was Fraser Island, which consists entirely of dune lakes (sub-tropical Queensland) on the largest sand island in the world (Bayly, 1964; Bayly et al.,1975; Arthington et al., 1986) (Figure 3.1). The third geographic region was south-eastern Australian mainland from Jervis Bay in the north, north-eastern Victoria, western Victoria and South Australia, near Mt Gambier (Figure 3.1). This region had the most diverse collection of lakes, including dune lakes, sinkhole lakes and volcanic maar lakes. The fourth geographic region was Tasmania which was the most southerly lake group in the study (Figure 3.1, 3.2; Table 3.1). The most common lakes in Tasmania were highland lakes of glacial origin. Less common were coastal dune lakes (Figure 3.2). The most southern individual lake

39 was at 43 36'S, a coastal dune lake on the southern coast of Tasmania, near South East Cape (Figure 3.2; Table 3.1).

Lakes within the four geographic regions were also classified by their lake geomorphic typology (Timms, 1992). This created a total of eight groupings (Table 3.1).

Table 3.1 Lake groups within each of the four geographical region in this study. Tropical north Queensland is ‘TNQ’, Fraser Island is ‘FSI’, south-eastern mainland is ‘SEM’ and Tasmania is ‘TAS’. See Figures 3.1 and 3.2.

Number Geographic Lake Type Location sampled Region

5 TAS Dune lakes South East Cape, Waterhouse Point and Strahan 20 TAS Glacial lakes Multiple highland areas (West Coast Ranges, Cradle Mountain Lake St Clair, Walls of Jerusalem, Central Plateau, Mount Field and Hartz Mountains). 5 SEM Dune lakes Jervis Bay, East , south-western Victoria 2 SEM Maar lakes South-east South Australia and western districts Victoria 5 SEM Sinkhole Mt Gambier lakes 18 FSI Dune lakes Fraser Island 2 TNQ Maar lakes Atherton tablelands 10 TNQ Dune lakes Cape Flattery

Coastal dune lakes were generally at altitudes from 5 to 100 metres Australian Height Datum (AHD). On Fraser Island dune lakes of up to 150 metres AHD were sampled. The highest lakes were in the Tasmanian highlands and these were of glacial origin (Timms, 1992) and up to 1150 metres AHD (Figure 3.2, Table 3.11).

The selection of lake regions was based on information contained in previous scientific publications on various topics associated with Australian lakes (Tables 3.2, 3.3). This provided the general location for the nearly 70 lakes that were eventually sampled in Australia’s main freshwater lake districts (Figures 3.1, 3.2; Tables 3.1 - 3.7). It was intended that as many different types of lake as practical was sampled.

40 Three lake types were deliberately avoided. First human-made lakes (e.g. dams and reservoirs) were not sampled due to concerns that they may not have developed a ‘mature’ lake fauna and uncertainties about the impact of human activities on chironomids. Secondly, inland and coastal saline lakes were avoided. Several Australian studies have assessed their chironomid fauna and the level of salinity was well-understood to strongly affect assemblages (Bayly & Williams, 1964; Timms, 1973b; Paterson & Walker, 1974; Timms, 1983). The third type of lake that was avoided was fluvial derived lakes, such as billabongs (Timms, 1992). They were avoided due to their links with rivers and concerns about the impact of adjoining, mostly agricultural land use.

The climate varied considerably within and amongst the four lake geographic regions (see Figure 3.3). One of the most obvious climatic difference among regions, and in some circumstances within lake geographical regions (e.g. Cooktown vs. Atherton in tropical north Queensland, Figure 3.3), was annual temperature regimes. Annual monthly rainfall also varied according to region, as did the evenness of monthly rainfall.

41 1. Tropical Queensland Lakes Cape Flattery Lakes

15º S

Atherton Tablelands Lakes

2. Fraser Island Lakes 25º S

3. South-eastern mainland Lakes

Jervis Bay 35º S Mt Gambier and S-W Victorian lakes N-E Victorian lakes 4. Tasmanian Lakes 45º S 140º E 150º E

Figure 3.1. Map showing four lake geographical regions sampled in this study. Tropical north Queensland had two lake groups: Cape Flattery and Atherton Tablelands. South-east mainland had three groups: Jervis Bay, north-east Victoria and Mt Gambier and south- western Victoria. Tasmania had two groups, see map of Tasmania (Figure 3.2). Fraser Island was a single group.

42 Figure 3.2 Location of lake groups in Tasmania that were sampled as part of this study. Three coastal dune lake groups were sampled. All of the inland areas were lakes of glacial origin and were mostly located within Conservation Areas (CA) or National Parks (NP).

43 Cooktown Atherton

35 35

30 30

25 25

20 20

15 15

10 10

5 5

0 0 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec -5 -5

Fraser Island Jervis Bay

35 35

30 30

25 25

20 20

15 15

10 10

5 5

0 0 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec -5 -5

Mt Gambier Strahan

35 35

30 30

25 25

20 20

15 15

10 10

5 5

0 0 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec -5 -5

Liawenee

35

30

25

20

15

10

5

0 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec -5

Figure 3.3 Historic mean monthly minimum temperatures (unshaded bar) and historic mean monthly maximum temperatures (black bar), in degrees Celsius, from weather stations within lake regions in this study. Cooktown represents coastal TNQ, Atherton represents high altitude TNQ, Fraser Island has one station, Jervis Bay represents the northern margin of SEM and Mt Gambier the southern and western margin of SEM, Strahan represents coastal TAS and Liawenee represents high altitude TAS. Source of data, www.bom.gov.au accessed May 2003.

44 The majority of lakes were observed to be influenced by current or historic human activity. The selection of lakes was also influenced through discussions with landowners, lake researchers and national park personnel in each state and territory. Logistical and safety considerations also influenced the selection of lakes. In the final selection lakes were randomised from a pool of accessible ‘candidates’ in each lake area.

Table 3.2 Key Australian lake publications used to select lakes for this study, by lake geographical region.

Geographical region, name of Location Author and date lakes

South eastern Australia mainland Blue Lake, Lake Leake, Valley Mt Gambier Bayly & Williams, 1964 Lake, Browne Lake Mumblin, Surprise Western Victoria Timms, 1975 Leake, Edward and Lakes Valley Mt Gambier Timms, 1974 Bullenmerri, Purrumbete, Gnotuk Western Victoria Timms, 1973b, Timms, 1976b Dune lake Western Victoria Timms, 1977 Several South eastern South Australia Tyler et al., 1983 Western Victoria Timms, 1981 and Timms, 1983 Several Victoria Timms, 1973a Lake McKenzie Commonwealth Territory, Jervis Norris et al., 1993 Bay Lake McKenzie Commonwealth Territory, Jervis Wright & Cranston, 2000 Bay Blue Lake NSW Alps Timms, 1980a

Tropical north Queensland Cape Flattery dune lakes Cape Flattery Hawkins et al., 1988 Several locations, dune lakes Cape Flattery Timms, 1986a Barrine, Eacham and smaller Atherton Tablelands Timms, 1976a, 1979 lakes Barrine, Eacham and smaller Atherton Tablelands Russell, 1987 lakes Lake Barrine Atherton Tablelands Dimitriadis & Cranston, 2001

Fraser Island Several Fraser Island Arthington et al.,1986 Several Fraser Island Bayly et al., 1975 Several Fraser Island Bayly, 1964

Tasmania Several coastal dune lakes Tasmania Bowling et al., 1984 7 lakes, Lake St Clair et al. Tasmania Timms, 1978 Several, large glacial lakes Tasmania Fulton, 1983a Several, large glacial lakes Tasmania Fulton, 1983b Crescent and Sorrell Tasmania Cheng & Tyler, 1976

General lake publications General General description Timms, 1985 General General description Timms, 1980b General General description Timms, 1986b

45 Table 3.3 Chironomid taxa reported from Australian lake publications 1974-1993. Species / taxa names as reported.

N Species / taxa Area Author

1 Chironomus nr. bicoloris Fraser Is, Queensland Arthington et al.,1986 2 Chironomus opposites Tasmania Fulton 1983, a and b 3 Chironomus februarius North Queensland Timms, 1979 4 Chironomus nepeanensis North Queensland Timms, 1979 Chironomus duplex Western Victoria Timms, 1983 5 Chironomus duplex Western Victoria Timms, 1983 Cladopelma curtivalva Western Victoria Timms, 1983 6 Cladopelma curtivalva Tasmania Fulton, 1983 a and b 7 Conochironomus sp. A Fraser Is, Queensland Arthington et al.,1986 8 Cryptochironomus curtivalva Western Victoria Timms, 1983 Cryptochironomus griseidorsum Western Victoria Timms, 1983 9 Cryptochironomus griseidorsum Tasmania Fulton, 1983 a and b 10 Cryptocladopelma curtivalva North Queensland Timms, 1979 11 Cryptocladopelma North Queensland Timms, 1979 12 Dicrotendipes conjunctus Mt Gambier, South Australia Timms, 1974 13 ? Harnischia sp. Tasmania Fulton, 1983 a and b 14 ? nr. Harnischia sp. Tasmania Fulton, 1983 a and b 15 Kiefferulus intertinctus Western Victoria Timms, 1983 16 Kiefferulus parbitinctus Fraser Is, Queensland Arthington et al.,1986 17 Kiefferulus ?martini Fraser Is, Queensland Arthington et al.,1986 18 Kiefferulus warfini Jervis Bay, ACT Norris et al., 1993 19 Microspectra sp. Jervis Bay, ACT Norris et al., 1993 20 Nilodorum biroi North Queensland Timms, 1979 21 ? Parachironomus sp. Tasmania Fulton, 1983 a and b 22 Parachironomus ?delinificus Tasmania Fulton, 1983 a and b 23 Parachironomus sp. Jervis Bay, ACT Norris et al., 1993 24 Polypedilum leei (Penta) N Stradboke, Queensland Bensink & Burton, 1975 25 Polypedilum ?tonnoiri N Stradboke, Queensland Bensink & Burton, 1975 Polypedilum nubifer Western Victoria Timms, 1983 26 Polypedilum nubifer Northern New South Wales Timms, 1985 27 Polypedilum nr. oresitrophus Tasmania Fulton, 1983 a and b 28 Polypedilum sp.1. Tasmania Timms, 1978 29 Polypedilum sp.2. Tasmania Timms, 1978 30 Pseudochironomus sp. Jervis Bay, ACT Norris et al., 1993 31 Riethia stictoptera Fraser Is, Queensland Arthington et al.,1986 32 Rheotanytarsus sp. Jervis Bay, ACT Norris et al., 1993 33 ?Tanytarsus Tasmania Fulton, 1983 a and b 34 Tanytarsus sp. 1 Western Victoria Timms, 1977 35 Tanytarsus ?fuscithorax N Stradboke, Queensland Bensink & Burton, 1975 36 Tanytarsus paskervillensis Mt Gambier, South Australia Timms, 1974 37 Tanytarsus sp B Mt Gambier South Australia Timms, 1974 38 Tanytarsus barbitarsus Western Victoria Timms, 1983 39 Xenochironomus australiensis Mt Gambier, South Australia Timms, 1974 Procladius sp. Western Victoria Timms, 1983 40 Procladius villosimanus Western Victoria Timms, 1983 41 Ablabesmyia notabilis N Stradboke, Queensland Bensink & Burton, 1975 42 Clinotanypus ? crux North Queensland Timms, 1979 Coelopynia pruinose Western Victoria Timms, 1983 43 Coelopynia pruinose Tasmania Fulton, 1983 a and b 44 Paramerina sp. Tasmania Fulton, 1983 a and b 45 Pentaneura sp. Jervis Bay, ACT Norris et al., 1993

46

N Taxa Area Author 46 Tanypodinae sp. Jervis Bay, ACT Norris et al., 1993 47 Orthocladiinae sp.1 (nr. Smitta) Tasmania Fulton, 1983 a and b 48 Orthocladiinae sp.2 Tasmania Fulton, 1983 a and b 49 Orthocladiinae sp.3 Tasmania Fulton, 1983 a and b 50 Orthocladiinae sp.4 Tasmania Fulton, 1983 a and b 51 Un ID Orthocladiinae sp.1 Tasmania Timms, 1978 52 Un ID Orthocladiinae sp.2 Tasmania Timms, 1978 53 Un Id Chironomini sp.1 Tasmania Timms, 1978 54 Un Id Chironomini North Queensland Timms, 1979 55 Un Id tanytarsini North Queensland Timms, 1979

Chaoborus sp. Fraser Is, Queensland Arthington et al.,1986

47 3.3 Materials and Methods

3.3.1 Exuviae collection Chapter 2 revealed that a single sample of chironomid exuviae gave a reasonable estimate of the dominant species living within a lake. Thus to maximise the number of lakes sampled, single samples were taken from each lake, for comparison of chironomid species among lakes. The sampling unit was, therefore, an individual lake. Random selection of lakes within lake groups created lake ‘replicates’.

Although seasonality of emergence was not considered to be significant in Lake McKenzie at Jervis Bay (Chapter 2; Wright & Cranston, 2000), southern- Australian lakes were sampled in the warmer months due to their cool to cold winter temperatures (Figure 3.3) that may have influenced chironomid emergence. As a consequence lakes were sampled in Tasmania in February 1997, Victoria and South Australia in December 1996 and Jervis Bay over several months in 1996, 1997 and early 1998. Lakes in Queensland were sampled in May 1997 (Fraser Island) and June 1997 (Cape Flattery) as the warmer winter weather allowed chironomids in these lakes a less seasonal pattern of emergence (Armitage, 1995).

Sampling protocol followed that used in Chapter 2, except that individual samples were larger, sweeping 50 metres of shoreline rather than 20 metre. The leeward shore of lakes was identified and sampled, following the species richness and abundance of exuviae revealed from the leeward area of Lake McKenzie, in Chapter 2. The shoreline was sampled in one continuous length, unless there were physical obstructions along the shore. Wind and wave conditions were assessed and the leeward shore was sampled. The time taken to net a sample from each lake depended on the ease of movement along the lake shoreline, but was generally about 30 minutes.

At each site, the net contents were placed in a tray, and were examined immediately (Figure 3.10, Plate 1). If field conditions were adverse (e.g. inclement weather, isolated locations), the sample was examined at a more suitable location within 12 hours. Multiple representative morphospecies were selected from samples, using a dissecting microscope (X 40 magnification). These

48 morphospecies were preserved in containers in 70 % ethanol. In the laboratory, at least one specimen of each morphospecies was slide-mounted in Euparol, and identified using a compound microscope (up to X 400 magnification). Identification to species, where possible, followed Cranston (1996) for Chironomidae and Colless (1986) for Chaoboridae. Validation, when available, was achieved by comparison to reared material in the Australian National Insect Collection (CSIRO, Canberra, Australia).

Electrical conductivity, pH, air and water temperature were measured at each lake with a field meter (YSI model ‘AI 112’). Where boat access to the water was available, a single depth measurement was made up to a depth of 10 m, in the centre of the lake.

3.3.2 Assessment of human disturbance A subjective assessment of the level of human disturbance was noted during the collection of samples from each lake. Disturbance was rated on a scale from 1 (no human disturbance observed) to 5 (high level of human disturbance observed). The level of disturbance included the catchment of the lake, and its degree of human modification, the abstraction of water from lakes, and the runoff from surrounding lands, recreational pressures on the lakes and their catchment.

3.3.3 Data analysis Some lakes yielded too few ( 3) chironomid exuviae to be amenable to analysis and were therefore excluded. Other species were collected from only one or two lakes, and were thus classed as rare for the purposes of this study.

Analysis of variance (ANOVA) was conducted on data from the six lake groups that had at least five lakes available to act as replicates. All Tasmanian dune lake results were used in the analysis. Five lake results were selected at random from Tasmanian glacial lakes (20 lakes), south-eastern mainland dune lakes (5 lakes), South Australian sinkhole lakes (5 lakes), Fraser Island dune lakes (18 lakes) and Cape Flattery dune lakes (10 lakes). ANOVA was used to determine if electrical conductivity (EC), pH or chironomid species-richness varied significantly due to lake-group (Table 3.1). EC and species richness data was log (x+ 1) transformed to better approximate a normal distribution. Graphical presentation of means (EC and number of species) and error bars were back-transformed.

49 Multivariate analysis has been demonstrated to be a sound technique for evaluation of freshwater macroinvertebrate (Norris et al., 1982, Wright, 1994) and marine studies (Clarke, 1993; Warwick, 1993). Non-metric multidimensional scaling (NMDS) was performed on the similarity matrix which had been computed with presence / absence chironomid data using the Bray-Curtis dissimilarity measure (Clarke, 1993; Warwick, 1993; Wright, 1994). Two-dimensional ordinational plots represented the dissimilarity among samples. All reference sites were grouped to test differences by one way analysis of similarity values (ANOSIM; Clarke, 1993) in ordinations between both the four geographic regions and the lake groups (Table 3.1). Analysis of similarity is a non-parametric randomisation procedure applied to the association matrix to determine the significance of differences among pre-defined groups of collections (in this case among lakes from four geographic regions (see Clarke, 1993, for its rationale). The influence of particular taxa in creating the differences in the ordinations between the groups was quantified using the similarity percentage procedure (SIMPER). The BIOENV procedure (Clarke & Ainsworth, 1993) was used to assess which environmental variables (pH, electrical conductivity, altitude, latitude and longitude) were correlated with chironomid species-assemblages. The above multivariate analyses were performed using the software package PRIMER version 5 (Clarke, 1993).

50 3.4 Results

3.4.1 Chironomidae and Chaoboridae results Overall, 135 species of chironomid and three species of Chaoborid were collected from the 67 lakes surveyed (Table 3.4). A greater number of lakes were visited but did not yield sufficient chironomid exuviae ( 4/lake) to be incorporate into the analyses. Of the 135 species recorded, 37 were collected from a single lake and 23 were found in only two lakes. Thus 60 of chironomid species in this investigation were regarded as uncommon.

Only 5 of the 135 chironomid and one Chaoborid species could be considered cosmopolitan (Tables 3.4, 3.5). They were collected from at least one lake in each of the four geographic regions. Some 70.4 % of species (95) of chironomid were only found in temperate lakes with 28 collected in both temperate and tropical lakes and 13 were exclusively found in tropical north Queensland (Tables 3.4, 3.5).

The greatest number of species recorded from a lake region (69) was collected from 12 lakes in the south-eastern mainland, while 56 species were collected from 25 lakes in Tasmania (Table 3.6). A total of 47 species was collected from 18 lakes in Fraser Island and 41 from 12 lakes in tropical Queensland (Table 3.6). The mean number of chironomid species recorded from each lake group did not vary significantly (Table 3.7; Figure 3.4).

There was a very strong regional distribution pattern of chironomid species. More than half (n=77) of chironomid species were restricted to a single geographic region (Table 3.5). A similar proportion of the species (17.8 %) (n=24) were only found in Tasmania and (17.0 %) (n=23) from the south-eastern mainland geographic region (Table 3.5). A smaller proportion (11.9 %) (n=16) was collected only from the lakes of Fraser Island geographic region, and the smallest proportion (9.6 %) (n=13) was found only in tropical Queensland (Table 3.5).

Strong regional differences were also observed for species from different chironomid sub-families (Table 3.6). There were particularly strong regional differences for species of sub-family Orthocladiinae: 14 species were recorded in Tasmanian lakes, 10 in south eastern mainland lakes, only one from Fraser

51 Island, and none from tropical north Queensland lakes (Table 3.6). This indicates that the group is restricted to lakes in the southern parts of Australia.

Regional differences were also recorded for the Chironominae sub-family: 48 species were collected in south-eastern mainland lakes, 38 from Fraser Island, 34 from Tasmania and 30 from the tropical lakes of Queensland. Regional differences were also recorded for some of the Chaoborid species (Table 3.6).

52 Table 3.4 Geographic regional distribution of chironomid species, categorised according to biogeographic regions, collected from freshwater lakes in this study. X symbol indicates that the species was found from at least one lake in that biogeographic region. TAS = Tasmania, SEM = south-eastern mainland, FRI = Fraser Island, TNQ = tropical north Queensland.

Sub-family Chironomid Species TAS SEM FRI TNQ Tanypodinae Alotanypus sp. X Apsectrotanypus sp. X Coelopynia pruinosa X X X Clinotanypus crux X Djalmabatista sp. X X X Fittkauimyia disparipes X X Procladius paludicola X X X X Procladius small (sim. to palu.) X Procladius sqamiger X X X Procladius villosimanus X X Ablabesmyia hilli X X Ablabesmyia notabilis X X X X Larsia sp. X X X Paramerina X X X X Pentaneuini genus A sp. X X Pentaneuini genus C sp. X Pentaneuini genus E sp. X Orthocladiinae Corynoneura sp. X X Cricotopus parbictinctus X Cricotopus ‘tasmania’ X X Cricotopus 'pic.ponds' X Cricotopus 'Lake Edward' X Cricotopus nr. parbicinctus X X Cricotopus or nr. MO4 X Botryocladius austrolpinus X Botryocladius grapeth X Nanocladius sp. X Parakefferiella sp 1 X Parakefferiella sp 2 X Parakefferiella tas alpine X Paralimnophyes sp. X X Stictocladius sp. 1 X Stictocladius sp 2 X X Thienemanniella sp. X MO5 X Near S03 X Near MO1 X ? Orthocladinae sp. 1 X Chironominae Riethia new species (PC) X Riethia stictoptera X X Riethia plumose X Riethia (pale stictoptera) X Cladotanytarsus sp 1 X X Cladotanytarsus sp2 X X X Neozavrelia new sp. X Paratanytarsus near furvus X

53 (Table 3.4 Continued) Sub-family Chironomid Species TAS SEM FRI TNQ

Chironominae Paratanytarsus kathleenae X X X (continued) Paratanytarsus nr. K4 X Stempellina ? australiensis X X Stempellina ? johni X X Tanytarsus liepae X Tanytarsus nr.dycei X X Tanytarsus nr. B1 X X X Tanytarsus B1 X Tanytarsus B3 X Tanytarsus B4 X Tanytarsus nr. B4 X X Tanytarsus manlyensis X Tanytarsus nr. manlyensis a X X Tanytarsus nr. manlyensis b X X X Tanytarsus nr. K4 X X Tanytarsus K5 X Tanytarsus nr. bispinosus X Tanytarsus nr. belairensus X Tanytarsus near K1 X X Tanytarsus near B2 X X X Tanytarsus M1 X X Chironomus sp 1 X X Chironomus sp 2 X X X Chironomus sp 4 X X Chironomus sp 6 X Chironomus sp 7 X Chironomus sp 8 X Cladopelma sp 1 X X Cladopelma sp 2 X X Cladopelma sp 3 X Cladopelma sp 4 X X Cladopelma sp 5 X Conochironomus australiensis X X Cryptochironomus X X X X Dicrotendipes conjunctus X X Dicrotendipes pseudoconjunctus X X Dicrotendipes sp 1 X Dicrotendipes sp 2 X Dicrotendipes K1 X Dicrotendipes nr. K8b X X Dicrotendipes lindae X X Dicrotendipes jobetus X Kiefferulus martini X Kiefferulus nr. tinctus X Kiefferulus nr. timidus X X Microchironomus nr. K1 X Microchironomus sp 2 X Microtendipes umbrosus X Nilothauma sp. X X Parachironomus sp. 1 X X X Parachironomus K3 X Parachironomus nr. K4 X X

54 (Table 3.4 Continued) Sub-family Chironomid Species TAS SEM FRI TNQ Chironomidae Paracladopelma M1 X X Paracladopelma sp 2 X Paracladopelma sp 3 X Paracladopelma sp 4 X X Paratendipes sp. X Polypedilum nr. convexum 1 X Polypedilum nr. convexum 2 X Polypedilum leei X X X Polypedilum watsoni X X X X Polypedilum nr. seorus X Polypedilum nr. K1 X X Polypedilum vespertinus (M2) X X X Polypedilum 'pic ponds' X Polypedilum nubifer X X Polypedilum small p/l comb X Polypedilum nr. oresitrophus X Polypedilum 'nr. Tassie 2' X ? Polypedilum X Skusella sp. X Stenochironomus sp. 1 X X Stictochironomus nr. SX X Stictochironomus sp. 1 X X Xenochironomus nr. 'K4' X Xenochironomus K4 X X Xenochironomus SX X Xenochironomus australiensis X X Zavreliella sp. X X ? Chironominae sp. 1 X ? Chironominae sp. 2 X ? Chironominae sp. 3 X ? Chironominae sp. 4 X ? Chironominae sp. 5 X ? Chironominae sp. 6 X unknown genus 'SX' X genus 'K2' X indet. genus K4/K7 X

Chaboridae Chaoboris vagus X X X X Chaoboris punctilliger X Chaoboris ornatipennis X

55 Table 3.5 Geographical locations of the 135 species recorded

Single or combined geographic Number of species regions

All four regions 5 TNQ, FRI and SEM 6 TNQ, FRI and TAS 4 FRI and TAS 3 TNQ and FRI 9 FRI and SEM 7 FRI, SEM and FRI 3 TNQ and SEM 4 TAS and SEM 18 TNQ only 13 FRI only 16 SEM only 23 TAS only 24

Table 3.6 Number of species of chironomid in each sub-family and Chaoborid species recorded in each of the four lake regions

Chironomid sub-family / Chaoborid Geographic regions TAS SEM FRI TNQ Tanypodinae 8 11 8 11 Orthocladiinae 14 10 1 - Chironominae 34 48 38 30

Total 56 69 47 41

Chaoboridae species 1 1 1 3

Table 3.7 F-statistics and associated probabilities from analyses of variance of (log X + 1) transformed chironomid species richness data from lakes in six lake groups (see 3.2.4).

Source of Variation Df MS F P

Between lake 5 0.024 1.58 NS groups

Error 24 0.015

Total 29

56

18

16

14

12

10

8

6

4

2

0 TAS dune TAS glac SEM maar SEM dune SEM sink FRI dune TNQ maar TNQ dune

Figure 3.4 Mean number (back-transformed) of species of Chironomidae collected from each lake within each of the eight lake groups sampled in this study. Star symbol indicates that the group had 5 or more replicates and was included in ANOVA.

57 3.4.2 Chironomid community structure analysis (NMDS analysis) entire survey There were distinct geographical groupings of Chironomidae exuviae apparent in the NMDS of all exuviae data from this survey (Figure 3.5). The stress value of 0.20 gave a fair representation of the original data in two dimensions (Clarke, 1993). The NMDS ordination indicated that geographical location strongly influenced chironomid community structure with some overlap of sites (Figure 3.5). The tropical north Queensland lakes (red symbols) are displayed on the left of the plot. The Fraser Island lakes (yellow symbols) cluster to their right, with some overlap with tropical Queensland lakes. On the far right-hand side are the more broadly scattered Tasmanian lakes (green symbols), clearly with the most dissimilar chironomid community structure. Scattered in between the Queensland and the Tasmanian lakes, with some overlap, are the south-eastern mainland lakes (blue symbols).

Figure 3.5. NMDS ordination of chironomid samples taken from lakes from southern and eastern Australia. Triangles represent samples from lakes. Red represents tropical north Queensland lakes, yellow represents Fraser Island lakes, blue represents south-eastern mainland lakes, green represents Tasmanian lakes. Stress = 0.2.

58 3.4.3 Significance testing of lake groups (ANOSIM results) ANOSIM found that the differences between the four lake biogeographic regions differed highly significantly with a Global R-statistic of 0.532 (Table 3.8). This confirms that chironomid community structure in this study varied according to geographic region.

ANOSIM also tested differences between the eight lake groups, based on the different types of lakes in each geographic region (Table 3.9). ANOSIM indicated that the great majority (21 of the 27) of the pairwise comparisons between lake groups were statistically different (Table 3.9). Two of the eight lake groups had only two lakes and this limited the statistical power of the lake group comparisons (Table 3.9).

Geographic location appeared to be a stronger influence on chironomid community structure in lakes than was the geomorphic lake type (Table 3.8, 3.9). For example, Tasmanian dune lakes and glacial lakes were not significantly different (Table 3.9), but Tasmanian dunes lakes were different from dune lakes in other biogeographic regions (Table 3.9).

59 Table 3.8 R-statistics (Clarke, 1993) from one-way ANOSIM for pairwise comparison of chironomid data from geographic regions (Tasmania, south eastern mainland, Fraser Island, and Tropical North Queensland).

Comparison R-statistic Significance Level % TAS vs SEM 0.301 *** TAS vs FRI 0.663 *** TAS vs TNQ 0.745 *** SEM vs FRI 0.604 *** SEM vs TNQ 0.596 *** FRI vs TNQ 0.404 *** NS, P>0.05; * 0.01

60 Table 3.9 R-statistics (Clarke 1993) from one-way ANOSIM for pairwise comparison of chironomid data from geographic regions and geomorphic types (Tasmania dune and glacial; south eastern mainland dune, sinkhole and maar; Fraser Island dune, and Tropical North Queensland dune and maar).

Comparison R-statistic Significance Level % TAS dune vs TAS glacial 0.077 NS TAS dune vs SEM dune 0.676 ** TAS dune vs SEM volcanic 0.445 NS TAS dune vs SEM sinkhole 0.384 * TAS dune vs FRI dune 0.744 *** TAS dune vs TNQ maar 0.973 * TAS dune vs TNQ dune 0.912 *** TAS glacial vs SEM dune 0.461 *** TAS glacial vs SEM maar 0.448 ** TAS glacial vs SEM sinkhole 0.317 *** TAS glacial vs FRI dune 0.758 *** TAS glacial vs TNQ maar 0.743 *** TAS glacial vs TNQ dune 0.786 *** SEM dune vs SEM maar 0.382 NS SEM dune vs SEM sinkhole 0.464 *** SEM dune vs Fraser dune 0.617 *** SEM dune vs TNQ maar 0.509 NS SEM dune vs TNQ dune 0.675 *** SEM maar vs SEM sinkhole 0.309 NS SEM maar vs FRI dune 0.889 *** SEM maar vs TNQ maar 1 NS SEM maar vs TNQ dune 0.985 * SEM Sinkholes vs FRI 0.779 *** SEM Sinkholes vs TNQ maar 0.618 * SEM Sinkholes vs TNQ dune 0.818 *** FRI dune vs TNQ Maar 0.703 *** FRI dune vs TNQ dune 0.384 *** TNQ maar vs TNQ dune 0.437 * NS, P>0.05; * 0.01

3.4.4 PRIMER BIOENV results Five lake physical and chemical variables were investigated, using the PRIMER BIOENV procedure, to measure their influence in terms of shaping lake-dwelling chironomid communities. The variables were pH (variable 1), electrical conductivity (EC) (variable 2), altitude (variable 3), latitude (variable 4) and longitude (variable 5).

The most influential factors were altitude and latitude, then pH, longitude and electrical conductivity (Table 3.10).

61 Table 3.10 Summary of BIOENV results for the chironomid data from the Eastern Australian lake survey.

K 5 pH, EC, Altitude, Latitude, Longitude (0.434) 4 pH, Altitude, Latitude, Longitude (0.479) 3 pH, Altitude, Latitude (0.514) 1 Latitude (0.523) 2 Altitude, Latitude (0.557)

Combination of environmental variables, K at a time, that gives the highest rank correlations between the macroinvertebrate and environmental similarity matrices for each K. Bold type values indicate overall optimum.

3.4.5 Physical and chemical results The physical and chemical results for the lakes surveyed are detailed in Tables 3.11 – 3.14. Most lakes studied were neutral to acidic (Figure 3.6) and mean pH variation was highly significant among lake groups (Table 3.15). There tended to be a clinal trend from low pH in north Australia to more basic in the south (Figure 3.6). The most acid group were the ten dune lakes of tropical Queensland (Cape Flattery; mean pH 4.8). The next most acidic group were the Fraser Island lakes in temperate south- eastern Queensland. Only one of the eight lake groups sampled had a mean alkaline pH, the sinkhole lakes in south-eastern South Australia (mean pH 7.2). Two individual lakes with the highest pH (mean 8.5) were the two north-eastern Tasmanian dune lakes, Lake Little Waterhouse and Blackmans Lagoon (Table 3.11). The Tasmanian dune lake group included a sub-group of three acidic dune lakes, with a mean pH of 5.1 (Table 3.11).

The electrical conductivity (EC) of the eight lake groups also had variation that was highly significant (Table 3.15, Figure 3.7). The trend for electrical conductivity did not show as strong a clinal trend as pH. The lake groups with the highest EC were the western Victoria maar lakes (mean 1320 µS/cm), and the south-eastern South Australia sinkhole lakes (mean 830 µS/cm; Figure 3.7). Two north-eastern Tasmanian dune lakes (Blackmans Lagoon, EC 3850 µS/cm, Little Waterhouse Lake 1930 µS/cm) had the highest EC recorded in the survey from an individual lake (Table 3.11). The waters of most lake-groups were dilute. The three most dilute lake groups in this survey were the Fraser Island dune lakes, with mean EC of 140 µS/cm, then the Atherton maar lakes, mean EC of 80 µS/cm, and the most dilute group was the Tasmanian glacial lakes with mean EC of 32 µS/cm (Figure 3.7).

62 Table 3.11. Physical and chemical field data collected in this study from Tasmanian Lakes. * data from other studies (see Table 3.2).

Code Name Size pH EC Air temp. Water temp. Depth Lat. ( S) Long. ( E) date Al (h.a.) (µS/cm) ( C) ( C) (m) sampled AH TLT Unnamed 1.5 4.4 365 12 13.9 2 43 36’ 146 50’ 24-3-97 TGA Lake Garcia 8 5.5 101 19.6 16 8* 42 06.5’ 145 19’ 26-3-97 TLA Lake Ashwood 12 5.3 183 20 18.1 4* 42 06.5’ 145 17.5’ 26-3-97 TBL Blackmans Lagoon 28 8.7 3850 19.3 18.1 4 40 54.5’ 147 35.5’ 22-3-97 TLW Little Waterhouse Lake 56 8.4 2930 19.3 18.3 4 40 52.5’ 147 36.5’ 22-3-97 THL Hartz Lake 20* 6.4 25 17.5 10.8 32* 43 14.5’ 146 46’ 22-3-97 TES Lake Esperance 10 6.54 36 8.5 11.9 - 43 14’ 146 47’ 9-3-97 TLO Lake Osborne 7 6.63 32 8.8 11.5 - 43 13’ 146 46’ 9-3-97 TLD Lake Dobson 25 6.93 43 10.5 11.8 4 42 41’ 146 35.5’ 12-3-97 1 TBT Beatties Tarn 6.5 7.06 25 4.7 9.5 - 42 40.3’ 146 38.5’ 13-3-97 TPT Platypus Tarn 4.5 6.49 28 15 14.8 - 42 40.4’ 146 35’ 11-3-97 TFO Robert Tarn 2 6.96 51 11.3 10.9 - 42 40.6’ 146 34’ 11-3-97 1 TSO Lake Sorell 5000 6.86 84 13.9 13.4 4 42 07’ 147 10’ 18-3-97 TAD Lake Ada 200 6.91 29 15.8 16 - 42 54’ 146 28’ 16-3-97 1 TBP Unnamed 4.5 6.71 29 13.8 9.3 - 42 53’ 146 25.5’ 15-3-97 1

TSP Unnamed 0.3 7.07 39 11 10.7 - 42 53’ 146 26’ 15-3-97 1 TLK Lake Kay 60 7.15 31 8 9.5 1.5 41 54’ 146 30’ 14-3-97 1 TLF Lake Forgotten 22 6.8 46 8.8 - - 42 06’ 146 07’ 25-3-97 TSL Shadow Lake 22 6.52 34 11.8 12.1 - 42 06’ 146 08’ 19-3-97 TUP Unnamed 0.3 6.22 25 13.5 11 - 42 06’ 146 08.5’ 19-3-97 TKT King Solomons Jewels 20 5.85 16 18.6 18.1 - 41 47.8’ 146 16.5’ 17-3-97 1 TKO King Solomon Jewels 2 1.5 6.03 17 15.5 15.2 - 41 47.5’ 146 16’ 17-3-97 1 TDO Lake Dove 120 4.94 21 10.3 11.6 60 41 38’ 145 57’ 21-3-97 THA Lake Hanson 15 4.71 24 13 12.3 - 41 39’ 145 58’ 20-3-97 TLS Lake Selina 25 6.7 44 13 13.1 - 41 53’ 145 36’ 27-3-97

Table 3.12. Physical and chemical field data collected in this study from South eastern mainland lakes sampled (South Australia Victorian and Jervis Bay Territory) * data from other studies (see Table 3.2).

Code Name Size pH EC Air temp. Water temp. Depth Lat. ( S) Long. ( E) date Alt. (h.a.) (µS/cm) ( C) ( C) (m) sampled AH

VBA 244* 6.75 368 18.8 20.2 8* 37 32’ 149 52’ 16-12-96 1 VLB Little Beetle 2.5 ha 5.86 354 19.9 22 2.65 37 47.2’ 148 25’ 17-12-96 10 VPU Purrumbete 552* 6.35 664 19 16.6 45* 38 17’ 143 13’ 18-12-96 13 VBM Bullen Merri 488* 8.7 1345 17 16.5 66* 38 15’ 143 06’ 18-12-96 16 SLE Lake Edward 29 7.18 2590 18.5 19.4 7* 37 37.6’ 140 36.2’ 22-12-96 11 VSW Swan Lake 12 6.46 393 19.3 18.9 4* 38 12’ 141 19’ 20-12-96 10 VMO Lake Monibenong 50 7.08 941 21.5 19.3 7 38 0.75’ 141 11’ 20-12-96 10 SPP Piccanninnie Ponds 300* 6.9 3590 21.4 16.5 10-90* 38 2.9’ 140 56.3’ 19-12-96 10 SEP Ewans Ponds 5* 6.8 836 15.8 16.2 11* 38 01’ 149 49.5’ 19-12-96 10 SLB Little Blue 1 ha 8.08 439 30 18.8 37 37 55.7’ 140 40.8’ 21-12-96 20 SGW Gouldens Waterhole 0.5 ha 7.28 619 35 17.2 12 37 56.8’ 140 41’ 21-12-96 20 SSS Sisters Sinkhole 0.5 ha 7.38 501 30 21 12 37 54’ 140 8.3’ 21-12-96 20 ALM Lake McKenzie (Jervis 8* 6.00 205 22 20 8 35 09’ 150 41’ 28-4-97 to 25 Bay) 7-3-98

Table 3.13. Physical and chemical field data collected in this study from Fraser Island Lakes. * data from other studies (see Table 3.2).

Code Name Size pH EC Air temp. Water temp. Depth Lat. ( S) Long. ( E) date Alt. (h.a.) (µS/cm) ( C) ( C) (m) sampled AHD

QLM Lake McKenzie 100 4.74 105 20 21.6 8.5 25 27’ 153 03’ 27-5-97 80 QBA Basin Lake 5.5 4.85 78 22.7 22.7 7 25 28.1’ 153 02.4’ 28-5-97 75 QBI Birrabeen 125 4.74 109 21 21 7 25 30.2’ 153 03.4’ 28-5-97 95 QBE Benaroon 75 4.6 107 18 22 3.5 25 31.3’ 153 03.3’ 28-5-97 95 QYJ Yankee Jack 20 5.1 85 24 17.8 1.1 25 34.5’ 152 59.5’ 29-5-97 50 QBO Boomagin 140 4.73 128 21 20.7 3 25 33.4’ 153 04.1’ 29-5-97 75 QGA Garawongerea 20 5.13 96 22 21.1 4 25 20’ 153 09.2’ 30-5-97 140 QHL Hidden Lake 5 4.5 88 19 19 4 25 14.3’ 153 10.2’ 2/6/1997 112 QJE Lake Jennings 10 4.27 100 22 20.2 4 25 29.7’ 153 03.1’ 30-5-97 122 QBG Lake Boomerang 10 4.36 122 19.9 19.9 0.75 25 13.7’ 153 07.8’ 31-5-97 115 (South) QAL Lake Allom 5.5 4.61 81 21.1 21.1 3.5 25 11.9’ 153 12.3’ 31-5-97 145 QWA Lake Wabby 5 5.93 193 19.5 20.4 6 25 28.6’ 153 07.6’ 1-6-1997 60 QOL Ocean Lake 10 6.3 365 20 19.6 2.5 24 55.5’ 153 16.4’ 2-6-1997 10 QCO Coomboo Lake 15 4.34 84 19 19.7 1.5 25 13.5’ 153 10.1’ 2-6-1997 120 QGN Lake Gnanan 20 4.16 166 19 20 0.75 25 06.9’ 153 11.4’ 3-6-1997 15 QWH White Lake 40 4.6 111 19 19 4 25 07.8’ 153 12’ 3-6-1997 15 QBW Bowarrady 100 5.54 115 21 20 2.5 25 09’ 153 12.7’ 4-6-1997 11 QGY Govi 0.5 5.82 164 20 19 1.5 25 35.3’ 153 05.8’ 4-6-1997 20 QGV Garry Lake 6 4.87 98 21 18 1.5 25 37.4’ 152 58.9’ 5-6-1997 50 QCF Freshwater Lake 5 6.03 114 22 18 3 26 03’ 153 08.3’ 9-6-1997 12

Table 3.14. Physical and chemical field data collected in this study from Tropical North Queensland lakes Lakes. * data from other studies (see Table 3.2).

Code Name Size pH EC Air temp. Water temp. Depth Lat. ( S) Long. ( E) date Alt. (m (h.a.) (µS/cm) ( C) ( C) (m) sampled AHD

QBR Lake Barrine 103.5 6.94 69 19 21.3 65* 17 16’ 145 37’ 11/6/1997 750 QEA Lake Eacham 50.3 6.77 59 18.5 21.7 65* 17 18’ 145 36’ 12/6/1997 750 QAP Unnamed (C. Flattery 200 ha 5.61 225 22 23 3 14 58.65’ 145 19.26’ 15-6-97 <20 QL1 Unnamed (C. Flattery) 5 ha 4.3 99 23 23 1 14 57.78’ 145 16.85’ 15-6-97 <20 QL2 Unnamed (C. Flattery) 80 ha 4.33 110 23 24 1.5 14 58.82’ 145 16.29’ 15-6-97 <20 QL3 Unnamed (C. Flattery) 40 ha 4.01 117 23 24 1.75 14 58.92’ 145 15.8’ 15-6-97 <20 QB1 Unnamed (C. Flattery) 0.5 ha 6.05 210 22 22 2.5 14.59 17’ 145 20.01’ 16-6-97 <20 QB2 Unnamed (C. Flattery) 0.5 ha 4.9 65 22 22 2 15 00.22’ 145 19.23’ 16-6-97 <20 QGS Unnamed (C. Flattery) 2 ha 4.2 200 23 24 2.2 14 59.24’ 145 19.37’ 16-6-97 <20 QFD Unnamed (C. Flattery) 0.5 ha 4.55 184 22.5 22.5 0.5 15 00.17’ 145 18.86’ 17-6-97 <20 QWL Unnamed (C. Flattery) 10 ha 4.12 135 22 23 1.5 15 00.17’ 145 18.86’ 17-6-97 <20 QLL Unnamed (C. Flattery) 15 ha 4.1 78 19 22 1.5 15 05’ 145 11’ 18-6-97 <20

Table 3.15 F-statistics and associated probabilities from analyses of variance of pH and (log X + 1) transformed electrical conductivity data from lakes in six lake groups (see 3.2.4) sampled in this study.

Source of Variable Variation pH Electrical conductivity Df MS F P MS F P

Between 5 5.00 5.09 < 0.005 1.47 11.57 < 0.001 Lake Groups

Error 24 0.98 0.13

Total 29

67 8

7

6

5

4 TAS dune Tas glac SEM dune SEM sink SEM maar FRI dune TNQ maar TNQ dune

Figure 3.6 Mean pH of each lake group, plus/minus one standard error. Star symbol indicates that the group had 5 or more replicates and was included in ANOVA.

1500

1200

900

600

300

0 TAS dune Tas glac SEM dune SEM sink SEM maar FRI dune TNQ maar TNQ dune

Figure 3.7 Back-transformed mean electrical conductivity (µS/cm) of each lake group, plus/minus one standard error. Star symbol indicates that the group had 5 or more replicates and was included in ANOVA.

68 Figure 3.8. Plates 1 to 4. Tasmanian lakes. Plate 1 Author sampling exuviae from King Solomons Jewels (glacial-derived), Walls of Jerusalem National Park. Plate 2 Hanson Lake (glacial-derived), Cradle Mountain National Park. Plates 3 Platypus Tarn (glacial-derived), Mount Field National Park. Plate 4 Lake Garcia (dune lake), within pine plantation, near Strahan, Tasmania. Plate 2. Plate 1.

Plate 4. Plate 3. 69 Figure 3.9. Plates 5 to 8. South-eastern Australian mainland lakes. Plate 5 Lake Little Beetle (dune lake), north-east Victoria Plate 6 Lake Bullen Merri, western Victoria (maar lake), Plate 7 Little Blue Lake, south-eastern South Australia (sinkhole lake). Plate 8 Gouldens Waterhole (sinkhole lake), with irrigation pump housing, south-eastern South Australia. Plate 6. Plate 5.

Plate 7. Plate 8. 70 Figure 3.10. Plates 9 to 12. Fraser Island dune lakes. Plate 9 Yankee Jack Lake (maneuvering field sampling punt through fringing macrophytes). Plate 10 Aerial view of Hidden Lake. Plate 11 Processing exuviae samples with field microscope immediately after collection on the shore of Lake Jennings. Plate 12 Lake McKenzie, view of sandy beach along one side of lake. Plate 9. Plate 10.

Plate 11. Plate 12. 71 Figure 3.11. Plates 13 to 16. Fraser Island dune lakes. Plate 13 Lake Wabby. Plate 14 Ocean Lake. Plate 15 Lake Boomagin. Plate 16 Lake Birrbeen. Plate 14. Plate 13.

Plate 15. Plate 16. 72 Figure 3.12. Plates 17 and 18 Tropical north Queensland maar lakes, Atherton Tablelands. Plate 17 Lake Barrine source of photo www.rainforestgallery.com.au/ images/barrine.jpg Plate 18 Lake Eacham source of photo http://www.tesag.jcu.edu.au/subjects/ev3200/FIELD. Plate 14.Plate 19 Field work in dune-field of Cape Flattery, ‘Cape Flattery Silica Mine’. Plate 20 Brown tannic waters of unnamed dune lake at Cape Flattery. Plate 17. Plate 18.

Plate 19. Plate 20. 73 3.4 Discussion

The chironomid fauna of Australian lakes is substantially more species-rich than has been previously recognised (Table 3.3). A total of 134 species was collected (Table 3.3), in contrast to approximately 55 taxa reported across a range of lakes in numerous previous studies (multiple authors, Table 3.3). This finding challenges the long-standing conclusion of Timms (1985) that Australian lakes are depauperate in chironomid species.

Chironomid species distribution differed according to geographic location and only five species were sufficiently widespread (found in all four geographic regions) to be considered cosmopolitan. More generally there were major differences among bioregions, the ANOSIM pairwise difference between species assemblages was highly significant between all geographic regions (Table 3.8). More than half (56 %, n = 75) of the total number of species was restricted to only one of the four biogeographic zones.

One of the most obvious biogeographic patterns was the species richness of the sub-family Orthocladiinae which showed a strong preference for the two temperate geographic regions with 14 species collected from Tasmanian lakes and 10 species from the south-eastern mainland lakes (Table 3.6). In comparison, a single Orthocladiinae species was recorded in Fraser Island lakes and none from tropical north Queensland (Table 3.6). This finding is supported by European northern-hemisphere studies that have found species of Orthocladiinae prefer cool waters (Lindegaard, 1995), often dominating cool water chironomid species assemblages.

Tasmanian lakes had the highest proportion of endemic chironomid species in this study with 41.1 % (n = 56) of all chironomid species of which 24 species were not found in any of the other three geographic zones (Table 3.6). There was a slightly lower level of endemism apparent in tropical Queensland lakes (31.7 %) while the other biogeographic regions had lower levels of endemic species (Table 3.5). Using larvae collected from the lake profundal zone, as a basis for his investigation, Timms (1985) noted higher endemism of benthic macroinvertebrates from Tasmanian lakes, than from other areas of Australia.

74 The factors that most strongly influenced chironomid species assemblages in this study were the latitude and altitude of lakes (Table 3.10). This probably reflects lake temperature regimes (Figure 3.3). This was also observed by Dimitriadis and Cranston (2001) in their analysis of the chironomid and lake physiochemical dataset generated by this study using BIOCLIM (Busby, 1991). They concluded that the distribution of lake chironomid species was strongly associated with climate, particularly temperature and rainfall regimes, lake latitude and altitude. These attributes of climate and its distribution in the landscape are not mutually exclusive. Climate is strongly influenced by latitude and altitude (see figure 3.3). In this investigation these factors were observed to have more influence to chironomid fauna than did lake chemistry (Table 3.10). An example of altitudinal differences in chironomid assemblages was provided within the tropical-north Queensland geographic region. Dune lakes in this geographic region were all at less than 20 metres AHD compared to the two maar lakes that were at 750 metres AHD. Species assemblages were significantly different between these two lake types. Similar significant differences were not found between chironomid species in Tasmanian dune lakes (10 to 90 metres AHD) compared with Tasmanian glacial lakes (530 to 1190 metres AHD) (Table 3.14).

The association of chironomid species assemblages with climate is not unique to Australia. Rossaro (1991) also observed the distinct temperature preferences of chironomid species in a study of 991 sites in Italy. He concluded that many species had a unimodal response to water temperature. Canadian lake researchers (Walker & Mathews, 1987) also showed that climate had a substantial impact on the development of chironomid assemblages in lakes. They concluded that the relationship between the two was not evident in many studies due to the local scale of most researchers’ work, and reported that continental scale can reveal the dominant influence of climate. The research presented here demonstrated that chironomid species assemblages in Australian lakes show the same continental pattern as observed in the northern hemisphere.

Despite the overwhelming influence of geographical location and climate, chironomid species assemblages could be influenced by the type of lake

75 (Table 3.9). For example, within the south-eastern mainland geographic region species assemblages were different in dune lakes versus sinkhole lakes, yet were not different in dune versus maar lakes, or maar versus sinkhole lakes (Table 3.9). At the sub-geographic regional scale lake chemistry may have influenced the species assemblages, with south- eastern mainland dune lakes having contrasting pH and EC to maar and sinkhole lakes (Table 3.6, 3.7). These conclusions are tentative as the comparison is only based on samples collected from two maar lakes in this geographic region.

This study presents chironomid survey results from pupal exuviae, in contrast to almost all previous Australian studies that only presented larval results. For example, Timms (1978) and Fulton (1983a, 1983b) conducted two of the most intensive Australian lake benthic fauna surveys, in Tasmanian glacial-derived highland lakes. In Fulton’s spatially and temporally (multiple year) replicated studies he recorded a total of 19 species of chironomid larvae. He did use a 0.7 mm mesh that may have influenced results. Small larval species, such as species of Orthocladiinae may have escaped detection. Timms (1978) study used 0.4 mm mesh and recorded 17 chironomid species from his spatially replicated collections from seven lakes. In comparison, this current study recorded 54 chironomid species from single samples of chironomid exuviae collected from 25 Tasmanian lakes, comprising 20 glacial-derived lakes and five coastal dune lakes.

The current study also presents the most complete description, to date, of the distribution of the sub-family Orthocladiinae from Australian lakes. This study also found that the Orthocladiinae sub-family is much richer than previous studies revealed, with 14 different Orthocladiinae species from the 25 lakes sampled in Tasmania (Table 3.11). In contrast, Timms (1978) recorded five unidentified Orthocladiinae species and Fulton (1983a) recorded four unidentified Orthocladiinae species. Although the descriptions and keys for identification of chironomids have improved since both of these surveys, accurate species identification of larvae remains a major impediment.

76 Chironomid species richness was equivalent across the four Australian geographic regions (Table 3.7). Chapter 2 studied one south-eastern mainland dune lake in all seasons and recorded 30 species, a level similar to comparable small northern hemisphere lakes (Chapter 2, Wright & Cranston, 2000). A second detailed study of lake chironomids was done on Lake Barrine, Atherton Tablelands by Dimitriadis and Crantson (2001) using a combination of larval and exuvial chironomids, they recorded 28 species.

This current study has increased our understanding of chironomid species richness in Australian lakes through the sampling of many lakes at a low level of sampling effort. It has not studied an individual lake at the level of detail and very high sampling intensity of some international studies that have recorded correspondingly high levels of chironomid species richness. For example, the study by Carmel Humphries (1938) of Grosser Plöner See, a large German lake of about 30 km² included the collection of larvae, pupal exuviae and adults. Her survey involved 57 collections, in all seasons, and recorded a total of 86 chironomid species. A study of chironomids from a single Norwegian lake (Aagaard, 1978) recorded 76 species in a detailed study (spatial and temporal) of chironomid larvae and adults. A multiple-year study of chironomid larvae and pupal exuviae from two Dutch water-storage reservoirs resulted in 58 species. One of the most detailed chironomid studies of an African lake (Hare & Carter, 1987) recorded 80 chironomid species from a Nigerian lake. They collected adults and larvae and suspected that many of the flying adults originated in the terrestrial environs of the lake. They concluded that 46 species of chironomids lived in the lake itself. This current study, and no previous study of fauna from Australian lakes, has approached the level of detail of these international studies. Thus it is not possible to give a true estimate of the chironomid species richness in Australian lakes until they are investigated in similar detail.

My current study expended a more comparable level of sampling effort, and resulted in similar levels of species richness, to three international studies of lake chironomids. For example, an exuviae survey of a group of 25 tarns in the English lakes district (Pinder & Morley, 1995) recorded 106 species, nine Finnish lakes yielded 34 larval species (Kansanen et al., 1984) and finally, a larval study of one Swedish acid lake yielded 19 chironomid species

77 (Wielderholm & Eriksson, 1977). This suggests that internationally lake- dwelling chironomid species-richness depends, in part, on the level of sampling effort, duration and the number of life stages collected in the study.

This is the first continent-length scale study of chironomid species distributions in Australian freshwater lakes. It reveals that the ubiquitous, but poorly understood, lake-dwelling chironomid is much more species rich than was previously recognised in Australian lentic waters. chironomid species assemblages differed significantly across the four bioregions in this survey, and were most strongly influenced by latitude and altitude, suggesting that climate has a strong influence on species distributions.

78 Chapter 4 Water Chemistry and macroinvertebrate survey of the upper Grose River, NSW: effects of organic effluent and zinc-rich mine drainage.

4.1 Introduction

Chapters 2 and 3 revealed that Australian lakes are much more species-rich than their ‘species-poor’ reputation. Chironomids from Australian streams and rivers have also been considered to be generally tolerant of water pollution, in comparison to other freshwater macroinvertebrate families. This chapter investigates the water pollution response of chironomid larvae and other macroinvertebrate groups.

Macroinvertebrates are widely regarded as one of the best biological indicators for assessing the effects of water pollution on rivers and streams (Hellawell, 1986; Rosenberg & Resh, 1993) and they have been used in several Australian studies that have assessed the impact of sewage wastes (Jolly & Chapman, 1966; Cosser, 1988; Growns et al.,1995; Wright et al., 1995); urban landuse (Campbell, 1978; Arthington et al., 1982); pesticides (Ward et al., 1995); organic sugar wastes (Pearson & Penridge, 1987) and metals from mine drainage (Nicholas & Thomas, 1978; Norris et al., 1982; Mackey, 1988; Napier, 1992; Sloane & Norris, 2003).

In Australia, different taxonomic groups of macroinvertebrates have been demonstrated to have different responses to water pollution (Connell, 1981). Some families, such as Leptophlebiidae mayflies and Helicopsyche caddisflies, are widely regarded as being pollution sensitive (Chessman, 1995). Based on many Australian studies (eg. Jolly & Chapman, 1966; Norris et al., 1982; Napier, 1992; Chessman, 1995) the family Chironomidae is reputed to be one of the most pollution tolerant families. In contrast, an organophosphate pesticide study on a Queensland artificial stream concluded that chironomid species responded more sensitively to pollution than did other taxonomic groups (Ward et al., 1995). In addition, an ecological study of an organically polluted urban creek in Queensland (Arthington et al., 1982) revealed that most chironomid species did not

79 tolerate water pollution, but a few tolerant species were found in large numbers.

The relationship of macroinvertebrate groups to different types of water pollution has generally come from studies within a specific area being extrapolated to other areas. For example, Chessman (1995) produced a pollution tolerance index (SIGNAL) for macroinvertebrates from the Hawkesbury-Nepean catchment, and later produced a revised version (SIGNAL-HU97; Chessman et al., 1997) for the Hunter River catchment. These two versions of the SIGNAL index acknowledge that taxonomic groups of macroinvertebrates may exhibit varying relationships to water pollution, due to regional species differences. Although developed for a specific geographic area, both versions of the SIGNAL index have been used very widely throughout Australia to assist in the interpretation of macroinvertebrate data.

Most Australian water pollution studies using macroinvertebrates have focussed on a single type of pollution, often coming from a single point source (e.g. Norris et al., 1982; Cosser, 1988). No Australian upland stream studies have compared the response of macroinvertebrates, spatially and temporally, to two very different types of waste discharges to waterways within close proximity.

In this chapter, the response of macroinvertebrates to different pollution sources is investigated. The aim was to examine the effects of a heavy metal contaminated discharge and an organic waste discharge on stream macroinvertebrates within an otherwise pristine background. In this chapter I focus on the impact of water quality on insect families. The water pollution response of chironomid larvae to the different waste types and in comparison to the pollution response of other taxonomic groups is also examined.

80 4.2 Material and Methods

4.2.1 Description of study area Field work was carried out in the upper Grose River catchment in the Blue

Mountains area of NSW (Figure 4.1), part of the Great Dividing Range in

south-eastern Australia (33 31’S to 33 38’S, 150 14.8’E to 150 20.2’E). Most of the river valley is incorporated into the declared Grose Wilderness Area within the Blue Mountains National Park estate and the Greater Blue Mountains World Heritage Area (National Parks & Wildlife Service, 2001; Blue Mountains City Council, 2002). Whilst most of the study area is undisturbed bushland, a ribbon of land on the outer rim of the upper Grose catchment is subject to human activity via main roads (Bells Line of Road, Darling Causeway and Great Western Highway), power lines, a railway line, and residential and commercial land uses. Townships include Bell, Mount Victoria, and Blackheath (Figure 4.1). The upper Grose River catchment is approximately 8000 hectares. Elevations within the study area range from 380 to1093 metres AHD. The area is very rugged, often with steep gradients, slippery slopes and rocks, dense and tangled vegetation and numerous other site hazards (National Parks & Wildlife Service, 1999; Macqueen, 1997). Much of the area has no vehicle or walking access trails (National Parks & Wildlife Service, 1999).

The Grose River system flows within rugged and deeply dissected sandstone and shale gorges. The river and tributary streams surveyed are medium to small (1-4 metres wide) and shallow (0.1- 1.5 metres deep), fast- flowing waterways (Figure 4.15-4.18). They flow permanently, under the influence of ground water and quickly increase in flow after heavy rainfall. The stream channels are rocky bottomed, sometimes bedrock, frequently dominated by boulders and cobbled stream beds. Most waterways have a very steep gradient (1 in 10 - 50m), particularly the smaller tributaries that flow down the steep sandstone and shale escarpment towards the Grose River. The Grose River flows into the Hawkesbury-Nepean River, a coastal draining river system, one of the largest coastal river systems in south- eastern Australia.

81 Vegetation of the Grose River catchment was described by Keith and Benson (1988). The dominant vegetation in the study area was identified as the ‘Escarpment Complex’. The complex was present in three forms; the most dominant form, in sheltered locations, was ‘tall open forest’. Dominant species include Eucalyptus deanei, Eucalyptus cypellocarpa and Syncarpia glomulifera. The open forest form was found in less sheltered locations where Angophora costata ssp. costata, Eucalyptus piperita, Eucalyptus punctata were dominant. The closed forest form was found in moist, sheltered gullies, and was dominated by the rainforest species Ceratopetalum apetalum and Doryphora sassafras. ‘Blue Mountains Sedge Swamps’ are also found where Gymnoschoenus sphaerocephalus often dominate (Keith & Benson, 1988).

82 Figure 4.1 Map of survey sites (square symbols), waterways and waste discharge points in the upper Grose River. Approximate catchment boundary is dashed line. Inset shows location of study area in south-eastern Australia. Grose River sites include ‘GEN’ at Engineers Track, ‘GDK’ downstream of Koombandah Brook, ‘GDD’ downstream of Dalpura Ck, ‘GBK’ upstream of Victoria Creek and the lowest Grose River site ‘GHU’ at Hungerfords Track. Hat Hill Creek sites include ‘HHU’ above the STP, ‘HHD’ below the STP and ‘HHG’ above the Grose River. ‘DAL’ is the tributary Dalpura Creek and ‘VIC’ is the tributary Victoria Creek.

Coal mine drainage N Bell

Koombandah Brook O

GEN DALDalpura Creek GDK GDD Birrabang Creek

GBK

VIC Grose River GHU HHG 33º 35’ S

Hat Hill Mount Victoria Victoria Creek Creek

Blackheath STP HHD discharge

Mount Boyce HHU 1088 m ASL

2 km 150º20’E Blackheath

83 Two waste sources discharge into tributaries of the upper Grose River. One is coal mine drainage from a disused underground coal mine ‘Canyon Colliery’ (Figure 4.16) which operated in the upper Grose River Valley from the 1920s (Macqueen, 1997) to 1997 (personal observation). The mine drainage discharges into Dalpura Creek, which shortly thereafter flows into the Grose River (Figure 4.1). The second point source flows from the Blackheath Sewerage Treatment Plant (STP), which discharges approximately 0.92 ML/day of secondary treated effluent from a medium sized township, serving 5180 people (Sydney Water, 2004), into Hat Hill Creek. This waterway flows about 5 kilometres below the STP discharge before it’s confluence with the Grose River (Figure 4.1; Figure 4.17). Blackheath STP was built in the 1930s (Sydney Water, 1999).

The upper Blue Mountains generally receives more rainfall in summer and autumn than over the rest of the year (Figure 4.2) and this pattern occurred during this study. A widespread drought-event occurred in south-eastern Australia during 2002/03 with rainfall recorded at Mount Boyce (Figure 4.1) showing below average monthly rainfall during the study and for six months preceding the study (Figure 4.2).

The climate of the study area has cool winter and mild summer temperatures (Figure 4.3).

84 150

100

50

0

r e ry ry ril ly rch ne ob ua a Ap May Ju Ju ember n brua M August t Oct Ja ep Fe S November December

Figure 4.2. Historic mean monthly rainfall (mm / month), indicated by unshaded bars, for Mount Boyce weather station (Figure 4.1). Actual monthly rainfall during the study, black bars, was recorded at Mount Boyce from August 2002 to July 2003. Source of data: Commonwealth Bureau of Meteorology, (http://www.bom.gov.au/). Accessed April 2004.

25

20

15

10

5

0

y r r ry a ne ly e e pril u u b uary A M J J mber m n brua March ugust ctob e e e A Ja F O ec September Nov D

Figure 4.3. Historic mean monthly daily minimum (black bar) and maximum (unshaded bar) temperatures (degrees Celsius) for the Katoomba weather station. Source of data: Commonwealth Bureau of Meteorology, (http://www.bom.gov.au/). Accessed April 2004.

85 4.2.2 Study sites Ten sites were sampled in this study (Figure 4.1). All ten were located on upland streams with steep gradients, plunge pools and all had the occasional small to medium waterfall (Table 4.1). Their stream-beds were mostly comprised of cobble riffle areas, often scattered with large boulders, areas of bedrock, gravel and sand. Altitudes of the study sites ranged from just below 380 to 980 metres AHD (Figure 4.4). The gradients of the two longitudinal study reaches in Hat Hill Creek and the Grose River were both very steep (Figure 4.4). The Grose River ranged from the steepest gradient of 1 metre fall in altitude in 11.1 metres, to the flattest of 1 metre fall in altitude in 38.1 metres. Hat Hill Creek had a steep average gradient of 1 metre fall in altitude in 10.3 metres.

The vegetation that adjoined eight of the ten study sites belonged to different forms (Table 4.1) of the ‘Escarpment Complex’ (Keith & Benson,1988). ‘Blue Mountains Sedge Swamps’ vegetation adjoined the two upper sites on Hat Hill Creek. The site below the STP (HHD) was partly adjoined by hanging swamp and cleared grassland (Figure 4.1).

Four of the sampling sites were considered to be reference sites (GEN, GDK, VIC, HHU) as they were upstream of waste discharges.

86 Table 4.1 Summary information for each of the sampling sites used in this study Site name Site Co-ordinates Width Vegetation (Keith & Stream code Benson, 1988) order Grose River above GEN 33 32.8‘ S, 1 - 2 m Tall open forest form 2nd Engineers track 150 16.5’E

Grose River below GDK 33 32.9‘ S, 2 - 4 m Tall open forest form 3rd Koombanda Brook 150 18.1’E

Grose River below GDD 33 32.9‘ S, 2 - 4 m Tall open forest form 3rd Dalpura Creek 150 18.1’E

Grose River at GBK 33 34‘ S, 2 - 4 m Open forest form 3rd Burra Korrain 150 18.2’E

Grose River at GHU 33 34.7‘ S, 2 - 4 m Open forest form 3rd Hungerfords Track 150 20.2’E

Victoria Creek VIC 33 34‘ S, 1 - 2 m Closed forest form 2nd 150 18.2’E

Dalpura Creek DAL 33 32.9‘ S, 1 - 2 m Tall open forest form 1st 150 18.1’E

Hat Hill Creek HHU 33 37.1‘ S, 1 m Blue Mountains Sedge 1st above STP 150 18’E Swamp discharge

Hat Hill Creek HHD 33 36.9‘ S, 1 m Blue Mountains Sedge 1st below STP 150 18.1’E Swamp and cleared grassland Hat Hill Creek HHG 33 34.7‘ S, 1 - 2 m Closed forest form 1st above Grose River 150 19.5’E

87

1200 Grose River altitude profile Hat Hill Creek altitude profile 1000 2.5 km 2.75 km Koombanda Brook

800 6 km Dalpura Creek 7.5 km Victoria Creek 8.5 km Hat Hill Creek 600 10.5 km 8 km 14.5 km 400

200

0 Grose u/s of Grose River Grose River Grose River Grose @ Hat Hill u/s Hat Hill d/s Hat Hill u/s Engineers d/s of d/s of @ Burra Hungerfords STP STP Grose Track Koombanda Dalpura Ck Korrain Brook

Figure 4.4 Altitude profile of survey sites in the upper Grose River. The Y axis indicates altitude metres AHD. Distance from the top of the Grose watershed (left profile) and Hat Hill watershed (right profile) is indicated. Arrows and names on the Grose River profile indicate the tributaries flowing into the Grose River.

4.2.3 Collection and analysis of water and macroinvertebrate samples A portable field chemistry meter ‘WTW Multiline P4 Universal Meter’ was used to measure stream electrical conductivity, pH and water temperature. The reading was taken from the middle of the waterway. The field meter used to take these samples was calibrated in the laboratory prior to its use. It was then checked for calibration, using reference solutions, and adjusted if required, on each day it was used in the field.

To acquire water for chemical analysis six new and unused plastic 200 mL bottles were filled at each site on each of the three sampling occasions (April, May, June 2003; Table 4.2). Duplicate bottles were collected at all sites and occasions to enable a measure of sample variability. Sampling generally occurred on two or three consecutive days. All water samples were stored on ice and were analysed in the laboratory within 72 hours of collection. Laboratory testing was later conducted on each sample for total zinc (TZn), hardness (two occasions) and alkalinity (one occasion), and the contents of one bottle was used to test for total nitrogen (TN) and total phosphorus (TP).

88 Samples of water from the first water sampling occasion were analysed, using appropriate methods (APHA, 1998), for an extensive suite of metals (Aluminium, Arsenic, Boron, Barium, Cadmium, Chromium, Cobalt, Copper, Iron, Lead, Manganese, Mercury, Molybdenum, Nickel, Selenium, Silver, Tin, Uranium, Zinc) to determine which metals, if any, were found at levels above ecosystem protection guidelines and triggers levels (ANZECC, 1992 and 2000). Zinc was the only such metal and thus only zinc levels were analysed in all subsequent water samples. The only metal data presented from the first sampling round was zinc.

All water samples were analysed within the laboratories of Australian Water Technologies (Sydney Water) according to standard methods of chemical analysis (APHA, 1998).

Table 4.2 Date of macroinvertebrate and physio-chemical sampling (2003) for each site in the Grose River survey.

Site name Site Date of first, second and label third sampling occasions

Grose River at Engineers Track GEN 12 April 9 May 31 May Grose River below Koombanda Brook GDK 12 April 9 May 31 May Grose River below Dalpura Ck GDD 12 April 9 May 31 May Grose River at Burra Korrain GBK 9 April 10 May 31 May Grose River at Hungerfords Track GHU 9 April 10 May 1 June Victoria Creek at Burra Korrain VIC 9 April 10 May 31 May Hat Hill Creek above STP HHU 10 April 11 May 1 June Hat Hill Creek below STP HHD 10 April 11 May 1 June Hat Hill above Grose River HHG 9 April 10 May 1 June (Dalpura Creek) physio-chemical only DAL 12 April 9 May 31 May

Macroinvertebrate samples were collected at the same time as the physio- chemical sample collection. Samples were collected on three occasions during autumn to enable an assessment of within season temporal variation (Table 4.2). They were collected by ‘kick sampling’, using a ‘kick’ net, with 250 micron mesh, and a square 30 X 30 cm net frame, with a handle (Rosenberg & Resh, 1993; Wright, 1994). Samples were collected only from cobble-type riffle zones, except when these were unavailable (ie. at the upper two Hat Hill sites) when bedrock riffle zones were sampled. Sampling was conducted by disturbing the stream bottom by hand for a period of 60 seconds over a 30cm X 30 cm area immediately upstream of the net (Figure

89 4.19, Plate 19). The location of each replicate was randomly selected within a 15 metre stream reach. Five replicates were taken on each of three occasions making a total of 15 replicates from each site.

The net contents, stream detritus and macroinvertebrates, were immediately placed into a storage container, labelled, and were preserved in 70 % ethanol. In the laboratory the sediment below 250 microns was washed from the macroinvertebrate sample. The sample was then sorted under a dissecting microscope (X40) to extract the macroinvertebrates from stream detritus (leaves, sticks, algae clumps, rocks, gravel and sand). Macroinvertebrates in the remaining sorted sample were then removed and identified to the family, using the identification keys recommended by Hawking (1994). All insect groups were identified to the family level, some non-insect groups were not identified to the family level in accordance with the methodology used previously in a previous Blue Mountains study by the author (Wright, 1994; Wright et al., 1995).

4.2.4 Data analysis Multivariate analysis has been demonstrated to be a sound technique to evaluate freshwater pollution in macroinvertebrate community studies (Norris et al., 1982; Wright, 1994; Marchant et al., 1994; Wright et al., 1995) and marine studies (Clarke, 1993; Warwick, 1993). Non-metric multidimensional scaling (NMDS) was performed on the similarity matrix which had been computed with square-root transformed macroinvertebrate taxon abundance data using the Bray-Curtis dissimilarity measure (Clarke, 1993; Warwick, 1993). Two-dimensional ordinational plots represented the dissimilarity among samples. All four reference sites were grouped to test differences by one way analysis of similarity values (ANOSIM: Clarke, 1993) in ordinations between reference sites and test sites. The influence of particular taxa in creating the differences in the ordinations between the groups was quantified using the similarity percentage procedure (SIMPER). The above multivariate analyses were achieved using the software package PRIMER version 5 (Clarke, 1993).

Analysis of variance (ANOVA) was used to determine if electrical conductivity, pH, water temperature, TN, TP, TZn, hardness,

90 macroinvertebrate abundance or macroinvertebrate taxon richness varied significantly according to site and/or time. pH was not transformed, all other data were log (x+ 1) transformed to better approximate a normal distribution. To graphically present the data, the means and error bars were back- transformed.

4.2.5 Measurement of pollution affinity In order to assess the relationship between waste discharges and macroinvertebrate groups, a new methodology was developed for this study. This approach was based on the average abundance of each taxon at the reference sites compared to the taxon abundance at the sites exposed to water pollution. Each taxon was allocated a positive response if their abundance at a pollution affected site was more than double the average abundance at the reference sites. At progressively higher abundances, 500 % and 1000 % abundance of a taxon at reference sites, higher positive rankings were given (Table 4.3). Each taxon was classified as having a negative response if abundance at the polluted site was less than half the average abundance at the reference sites. At progressively lower abundances (20 % and below 1 % that found at reference sites) lower negative rankings were given (Table 4.3).

Table 4.3 Pollution affinity calculations for taxonomic groups, based on results from Grose River macroinvertebrate survey, autumn 2003. The water pollution sources were sewage effluent and zinc-rich mine drainage.

Abundance at polluted site relative Pollution affinity ranking to average at reference sites 200 % to 500 % Mildly positive 500 % to 1000 % Positive > 1000 % Highly positive

20 % to 50 % Mildly negative ✁✁

1 % to 20 % Negative ✁✁✁ < 1 % (or absent) Highly negative

91 4.3 Macroinvertebrate results

4.3.1 Macroinvertebrate abundance and taxon richness A total of 47,872 (54 taxa) macroinvertebrates were collected (Table 4.4). The most family-rich insect orders were Trichoptera (14 families), Diptera (8), Coleoptera (6) and Ephemeroptera (4). Overall, the greatest number of taxa at an individual site (42) was collected at Victoria Creek, a pristine tributary of the Grose River (Table 4.4). In contrast, the lowest taxon richness (23 taxa) was collected from the site immediately below the leachate from the coal mine (GDD). More generally, the number of taxa among sites ranged from 26 to 40 taxa. Ten of the taxa were only collected from a single site. The most abundant orders were Diptera (12,674 individuals), Ephemeropetra (12,254), and Coleoptera (5,336; Table 4.4. The most abundant families were Baetidae (8,457 individuals), Chironomidae (8,302) and Simuliidae (3,339)(Table 4.4).

Macroinvertebrate abundance (Table 4.5; Figure 4.5) and taxon richness (Table 4.5; Figure 4.6) varied highly significantly among sites, and there was a significant site by time interaction in both measures. The variation in abundance per replicate was greatest on the third sampling occasion (range = 39 at GDD to 1047 at GHU; Figure 4.5). Taxon richness per replicate ranged from 8.6 (GDD, 3rd occasion) to 25 taxa (VIC, 2nd occasion; Figure 4.6).

92 Table 4.4 List of macroinvertebrates collected from sites in the Upper Grose River

VIC GEN GDK GDD GBK GHU HHU HHD HHG Total Phylum Class (Order) Family

Plathelminthese Turbellaria (Tricladida) Dugessidae 4 - 1 - 2 83 - 46 38 174 Temnocephalidae ------1 - 1 Nemertea Tetrastemmatidae ------737 - 737 Annelida Oligochaeta 229 698 1023 6 15 171 82 148 197 2569 Mollusca Gastropoda Limpets 8 - 139 - - 1 - 2476 349 2973 (Ancylidae) Gastropoda - - 2 - - - - 271 - 273 (Non-Anclidae) Bivalvia Corbiculidae - - - - - 5 - 106 - 111 Arhtropoda Arachnida 27 19 21 3 15 104 89 4 10 292 (Acariformes) Orobatids 3 1 6 - 1 36 11 45 4 107 Collembola 1 - - - - - 40 - 1 42 Insecta Baetidae 786 101 714 3 40 6202 213 6 392 8457 (Ephemeroptera) Caenidae 376 - 392 - - - - - 118 886 Coloburiscidae 2 ------2 Leptophlebiidae 682 413 378 - 4 85 536 - 811 2909 Insecta (Odonata) Aeshnidae 8 8 19 3 12 37 28 1 14 130 Gomphidae 84 1 10 3 19 35 - - - 152 Insecta (Plecoptera) Amphipteryigidae - - 1 - 4 - - - - 5 Eustheniidae 4 22 4 3 2 1 - - - 36 Gripopterygidae 292 705 489 94 274 78 1011 94 43 3080 Austroperlidae - 3 ------3 Insecta (Megaloptera) Corydalidae 19 3 8 - 4 29 - - - 63 Veliidae 1 3 3 1 - 1 13 - - 22 Insecta (Neuroptera) Neurothidae 1 8 - 4 - - - - - 13 Insecta (Coleoptera) Elmidae larvae 433 474 449 4 14 118 323 75 987 2877 Elmidae adults 131 227 408 22 59 136 72 30 108 1193 Psphenidae 18 329 92 - - 32 8 6 9 494 Hydrophilidae 17 1 12 - 18 31 1 8 - 88 Scirtidae 146 31 83 6 144 200 40 - 2 652 Stratiomyidae - - 1 - - 15 - - - 16 Dytiscidae - - - - 15 - - 1 - 16 Insecta (Diptera) Culicidae - - 3 - - - 2 - - 5 Ceratopogindae 16 68 183 10 8 22 2 34 12 355 Chironomidae 490 800 1043 147 369 1153 1425 2187 688 8302 Simuliidae 100 5 97 11 77 1931 305 746 67 3339 Empididae 57 7 24 46 93 36 44 1 6 314 Tipulidae 49 43 39 2 6 140 48 2 12 341 Dixidae - - 1 ------1 Athericidae 4 2 - 6 5 - - - - 17 Insecta (Trichoptera) Hydrobiosidae 20 9 11 - 2 114 35 13 48 252 Philopotamidae 9 19 80 - 169 27 15 3 30 352 Hydroptilidae 115 77 302 84 217 41 - 1070 4 1910 Hydropyschidae 223 1 48 288 246 156 83 7 321 1373 Ecnomidae 48 35 44 1 12 50 1 - 20 211 Leptoceridae 14 - 1 30 74 231 - - 35 385 Helicopsychidae 141 1 30 - - 23 - - 34 229 Glossomatidae 30 11 14 - - 71 54 - 3 183 Calamoceratidae 39 - - - - 73 - - 1 113 Conoescucidae 82 - 64 1 22 291 - - 27 487 Tasimiidae 2 - 1 - - 24 - - 62 89 Odontceridae 19 ------19 Stenopsychidae 2 ------2 Calocid/Helicophidae - 2 - - - 308 26 - 25 361 Insecta (Lepidoptera) Pyralidae 1 ------1 Insecta (Mecoptera) Nannochoristidae 1 ------1

Insecta Unidentified 203 23 60 57 11 483 21 27 169 1054

Abundance 4936 4150 6300 835 1953 12574 4528 8118 4478 47872 Number of taxa 42 33 40 23 30 37 26 28 32 53

93 Table 4.5 F-statistics and associated probabilities from analyses of variance of (log X + 1) transformed macroinvertebrate taxon richness data and (log X + 1) transformed macroinvertebrate abundance data.

Source of Variable variation Taxon richness Macroinvertebrate abundance Main effects Df MS F P MS F P

Site (S) 8 0.23 0.23 <0.001 1.85 31.47 <0.001

Time (T) 2 0.01 0.01 NS 0.83 14.05 <0.001

Interaction

S & T 16 0.01 0.01 0.039 1.91 1.91 0.03

Error 108 0.01 0.06

94 1500

1200

900

600

300

0 GEN GDK GDD GBK GHU VIC HHU HHD HHG

Figure 4.5 Back-transformed mean abundance of macroinvertebrates collected from sites in the upper Grose River and its tributaries, on each of the three sampling occasions (three bars above each site label), together with plus/minus standard error for 5 replicates at each site.

30

25

20

15

10

5

0 GEN GDK GDD GBK GHU VIC HHU HHD HHG

Figure 4.6 Back-transformed mean taxa richness of macroinvertebrates collected from sites in the upper Grose River and its tributaries, on each of the three sampling occasions (three bars above each site label), together with plus/minus standard error for 5 replicates at each site.

95 4.3.2 Multivariate community structure analysis (NMDS) NMDS ordinations showed that while many of the sites had similar community structure (ordinations were clustered), the site downstream of the STP in Hat Hill Creek (HHD) and the two Grose River sites below Dalpura Creek (GDD, GBK) were well separated from the reference sites (Figure 4.7). Stress values ranged 0.20 to 0.17. This indicated that the MDS is a ‘good to fair’ representation of the original data (Clarke, 1993). Samples from each of the four reference sites obtained on the three different occasions clustered together in the top-right corner of the MDS (Figure 4.7). This indicates that they had a similar community structure. ANOSIM analysis confirmed that the differences in community structure among test and reference sites (grouped) were significant (Table 4.7).

Data from the four reference sites were compared using SIMPER, with the sites downstream of the STP (HHD, Table 4.8) and mine drainage (GDD, Table 4.9). Of the 10 taxa which contributed most to the separation between GDD and the reference sites, most had higher abundance at the reference sites: Leptophlebiidae mayfly, Elmidae beetle and Baetidae mayfly larvae, Oligochaetae worms and Gripopterygidae stonefly larvae. Hydropsychidae caddisfly larvae was the only key taxon that had higher abundance at the test site GDD than at the reference sites. In contrast, the 10 taxa which contributed most to the separation between HHD (downstream of the STP) and reference sites had higher abundance at the test site (HHD) than at the reference sites: Limpet gastropods, proboscis worms, Gastropoda snails, Simuliidae blackfly larvae, Hydroptilidae caddis larvae and Corbiculidae clams. The other key influential taxa were Leptophlebiid mayfly larvae, Baetidae mayfly larvae, Elmidae beetle larvae and Scirtidae beetle larvae, all of which had higher abundance at the reference sites than at HHD (Table 4.9).

Leptophlebiid mayflies were the only taxon that had a highly negative response to both types of pollution and they were the sole taxon to exhibit a strong negative response to sewage pollution (Table 4.6). Elmidae larvae, Baetidae mayflies, Gripoteryigidae stoneflies and Psephenidae beetle larvae had highly negative responses to zinc pollution and lower negative responses to sewage pollution (Table 4.6). Gastropoda limpets exhibited

96 completely opposite affinities. They responded highly positively to sewage pollution and highly negatively to zinc pollution (Table 4.6). Other taxa that demonstrated a negative response to zinc pollution and a positive response to sewage were filter-feeding taxa chironomid midge larvae and Simuliid blackfly larvae (Table 4.6). Hydropsychidae caddis fly larvae was the only abundant group that demonstrated a negative response to sewage pollution and mild positive response to zinc pollution (Table 4.6).

Table 4.6 Zinc and sewage pollution affinity of pollution indicator taxa, according to the Grose River survey (see 4.2.5 for methodology and Table 4.14 for grades).

Family Zinc Sewage SIGNAL-MET Signal 2 pollution pollution Grades grades

(organic)

Leptophlebiidae 8 8

Elmidae larvae 6 7

Baetidae 6 5

Gastropoda ‘limpet’ ✁ ✁✁ 9 4

Class Oligochatea NA 2 X

Chironomidae ✁ 2-6 # 3

Gripopteryigidae 7 8

Hydropyschidae ✁ 6 6

Psephenidae 8 6 Simuliidae ✁ ✁ 4 5

X Oligochatea grade from order grades, all other family grades (Chessman, 2003). # subfamily responses (Bruce Chessman unpublished data in Sloane and Norris, 2003)

97

Figure 4.7 NMDS ordination of macroinvertebrate data. Each coloured square symbol represents a single macroinvertebrate replicate (one of five) from the Grose River and its tributaries on each of three sampling occasions. Thus each site is represented by 15 coloured squares representing 15 samples. VIC = black, GEN = yellow, GDK = light blue, GDD = pink, GBK = brown, GHU = green, HHU = red, HHD = dark blue, HHG = grey (abbreviations given in Table 4.1).

98 Table 4.7 R-statistics (Clarke 1993) from one-way ANOSIM for pairwise comparison of sites for fourth root transformed macroinvertebrate data corrected from reference sites in the Upper Grose River and from below the mine seepage site (GDD). Abbreviations are given in Table 4.1.

Comparison R-statistic Significance level

Ref vs GDD 0.923 *** Ref vs GBK 0.772 *** Ref vs GHU 0.429 *** Ref vs HHD 0.944 *** Ref vs HHG 0.384 *** GDD vs GBK 0.276 ** GDD vs GHU 0.928 *** GDD vs HHD 0.969 *** GDD vs HHG 0.997 *** GBK vs GHU 0.707 *** GBK vs HHD 0.991 *** GBK vs HHG 0.935 *** GHU vs HHD 0.967 *** GHU vs HHG 0.844 *** HHD vs HHG 1.0 *** NS, P>0.05; * 0.01

Table 4.8 Results of SIMPER breakdown, the most influential macroinvertebrates contributing to the different communities at the reference sites compared with Grose downstream Dalpura Creek (GDD), the site most affected by mine drainage.

Taxon Reference sites GDD Contribution (%) Cumulative (%)

Leptophlebiidae 33.48 0 7.70 7.70 Elmidae larvae 27.98 0.27 6.85 14.55 Baetidae 30.23 0.20 5.97 20.51 Oligochaetae 33.87 0.40 5.89 26.40 Gripopteryigidae 41.62 6.27 4.23 30.63 Psephenidae 7.45 0 4.09 34.72 Hydropyschidae 5.92 19.20 3.90 38.62 Chironomidae 62.63 9.80 3.60 42.21 Scirtidae 5.00 0.40 3.37 45.58 Hydroptilidae 8.23 5.60 3.31 48.89

Table 4.9 Results of SIMPER breakdown, the most influential macroinvertebrates contributing to the different communities at the reference sites compared with Hat Hill Creek (HHD) downstream STP, the site most affected by STP effluent.

Taxon Ref. sites HHD Contribution (%) Cumulative (%)

Gastropoda ‘limpet’ 2.45 165.07 8.10 8.10 ‘Proboscis worm’ 0 49.13 6.22 14.32 Leptophlebiidae 33.48 0 5.83 20.15 Gastropoda ‘snails’ 0.03 18.07 5.11 25.27 Simuliidae 8.45 49.73 4.38 29.65 Baetidae 30.23 0.40 4.35 34.00 Elmidae larvae 27.98 5.00 3.82 37.82 Hydroptilidae 8.23 71.33 3.75 41.57 Corbiculidae 0.00 7.07 3.41 44.97 Scirtidae 5.00 0.00 3.17 48.14

99 4.3.3 Results of water physical and chemical indicators Water chemistry in the upper Grose River system changed due to the influence of the two waste discharges, particularly at the sites (GDD and GBK) immediately below the mine drainage, and below the STP, HHD (Figure 4.1). Longitudinal study of sites below both waste discharges indicated that moderate recovery of water quality was evident.

Water quality at the reference sites was generally equivalent to that expected in unpolluted sites in comparison to Australian water quality guidelines for the protection of ecosystems, specifically of upland NSW waterways (ANZECC, 2000). Reference sites had a low to medium level of dissolved salts, indicated by mean electrical conductivity of between 31.3 µS/cm at HHU and 82.6 µS/cm at GDK (Figure 4.8). Mean pH at the reference sites was acidic at two sites (HHU, GEN), close to neutral at VCK and slightly alkaline at one site, GDK (Figure 4.9). The water was of moderate softness at one reference site (GDK: 25 mg/L CaCO3) and was very soft at the other three reference sites (<10 mg/L CaCO3) (Figure 4.10). All four reference sites had low levels of total nitrogen and total phosphorus and additionally had very low levels of total zinc (Figure 4.12 - 4.14).

The electrical conductivity of the water varied highly significantly from site to site (Table 4.10). The STP caused a rise in mean conductivity from 31.3

S/cm HHU (upstream of the STP) to 326 µS/cm at HHD, immediately below the STP (Figure 4.8). In the Grose River electrical conductivity averaged 39.3 µS/cm at the upper site (GEN) and progressively rose at the next two sites (GDK average 82.6 µS/cm; GDD average 151 µS/cm) (Figure 4.8).

100 500

400

300

200

100

0 GEN GDK GDD GBK GHU VIC HHU HHD HHG DAL

Figure 4.8 Back-transformed mean electrical conductivity levels (µS/cm) recorded at each site sampled in the upper Grose River for three sampling occasions April to June 2003, (plus/minus one standard error).

The pH of the water varied highly significantly from site to site (Table 4.10). In association with the STP there was a rise in pH from 6.01 upstream of the STP (HHU), to 7.2 immediately below the STP (HHD; Figure 4.9). In the Grose River it ranged from 6.1 at the upper site (GEN) to 7.3 (GDK) and a similar level was maintained at all sites downstream (Figure 4.9). Coal mine drainage at Dalpura Creek appeared to have no marked affect on the pH (Figure 4.9). The highest recorded was at the lowest site on the Grose River, GHU with an average pH of 7.58 (Figure 4.9).

8

7

6

5

4 GEN GDK GDD GBK GHU VIC HHU HHD HHG DAL

Figure 4.9 Mean pH levels (pH units) recorded at each site sampled in the upper Grose River for three sampling occasions April to June 2003, (plus/minus one standard error).

101 Water temperature varied highly significantly from site to site (Table 4.10).

The lowest water temperature was 8.7 C at GDK in May 2003. The highest

water temperature was 15.3 C at Dalpura Creek in April 2003 (Figure 4.11).

60

50

40

30

20

10

0 GEN GDK GDD GBK GHU VIC HHU HHD HHG DAL

Figure 4.10 Back-transformed mean hardness levels (mg/CaCO3) recorded at each site sampled in the upper Grose River for two sampling occasions April to June 2003, (plus/minus one standard error).

16

12

8

4 GENGDKGDDGBKGHUVICHHUHHDHHGDAL

Figure 4.11 Mean water temperature (°C) recorded at each site sampled in the Upper Grose River on each of the three sampling occasions April to June 2003.

102 Table 4.10 F-statistics and associated probabilities from analyses of variance of (log X + 1) transformed electrical conductivity data, pH (untransformed) data and (log X + 1) transformed water temperature data from sites sampled in the upper Grose River and its tributaries between April and June 2003.

Source Variable of variation Electrical conductivity pH Water temperature Df MS F P MS F P MS F P Main effects

Site 9 1.61 230.4 <0.001 4.09 158.2 <0.001 0.014 3.72 <0.001

Error 140 0.007 0.026 0.004

Table 4.11 F-statistics and associated probabilities from analyses of variance of (log X + 1) transformed hardness data from sites sampled in the upper Grose River and its tributaries between April and June 2003.

Source of Variable variation Hardness Main Df MS F P effects

Site 9 0.647 403.7 <0.001

Error 30 0.001

The hardness of the water varied highly significantly from site to site (Table

4.11. The lowest mean hardness level of 4.2 mg/L CaCO3 was recorded in Hat Hill Creek (HHU), above the STP (Figure 4.10). Below the STP the level rose to approximately 38 mg/L CaCO3. The hardness level in the Grose

River was lowest at the upper-most site (GEN, average 6.2 mg/L CaCO3) and rose progressively at the next two downstream sites (GDK, average

23.9 mg/L CaCO3) to its highest point below Dalpura Creek (GDD, average

54.6 mg/L CaCO3; Figure 4.10).

Total nitrogen (TN) varied highly significantly among sites (Table 4.12). Average TN rose from 102 µg/L at the site above Blackheath STP (HHU) to 14,316 µg/L at the site immediately below the STP (HHD) and 7,533 µg/L at the site above the confluence with the Grose River (HHG; Figure 4.12). Average TN levels at the upper four Grose River sites were very low. The

103 coal mine drainage via Dalpura Creek had no apparent effect on TN levels in the Grose River. The highest average TN level in the upper four Grose River sites was 68.33 µg/L at GDK (Figure 4.12). In the Grose River TN was below detection limits (<50 µg/L) at three of the four upper-most sampling sites on one or more sampling occasions. Further downstream, it rose to an average of 1,680 µg/L at the lowest site (GHU) apparently due to the influence of Hat Hill Creek (Figure 4.12).

100000 Ecosystem Protection ANZECC (2000) Upland river, NSW, trigger value (200 µg/L). 10000

1000

100

10

1 GEN GDK GDD GBK GHU VIC HHU HHD HHG DAL

Figure 4.12. Mean Total Nitrogen (in µg/L) recorded at each site sampled in the Upper Grose River for three sampling occasions April to June 2003, (plus/minus one standard error). ANZECC (2000) ecosystem protection trigger value is indicated.

Total Phosphorus (TP) varied highly significantly from site to site (Table 4.12). The average level in Hat Hill Creek rose from 3.8 µg/L above Blackheath STP (HHU) to 507 µg/L at the site below Blackheath STP (HHD) and to 189 µg/L at the site above the confluence with the Grose River (HHG;

Figure 4.13). In the Grose River TP was 10 µg/L at all sites, apart from the most downstream site (GHU) which had an average level of 40.5 µg/L (Figure 4.13), due to the influence of Hat Hill Creek.

104 1000

Ecosystem Protection ANZECC (2000) Upland river (20 µg/L)., NSW, trigger value

100

10

1 GEN GDK GDD GBK GHU VIC HHU HHD HHG DAL

Figure 4.13. Mean Total Phosphorus (in µg/L) recorded at each site sampled in the Upper Grose River for three sampling occasions April to June 2003, plus/minus one standard error. ANZECC (2000) ecosystem protection trigger value is indicated.

Total zinc (TZn) levels also varied highly significantly from site to site (Table 4.12). It was not detected at the upper-most two sampling sites (GEN and GDK) in the Grose River (detection limit <10 µg/L) but TZn was always detected at the sampling sites in the Grose River below the confluence with Dalpura Creek (Figure 4.14). The highest average level in the Grose River (388 µg/L) was recorded immediately below Dalpura Creek (GDD) and it then dropped downstream (GBK, 261 µg/L; GHU, 71 µg/L). The average level of total zinc in Dalpura Creek was 595 µg/L. Zinc was detected at two sites in Hat Hill Creek but at low levels (Figure 4.14).

105

1000 Ecosystem Protection ANZECC (2000) Guideline value (5 µg/L).

100

10

1 GEN GDK GDD GBK GHU VIC HHU HHD HHG DAL

Figure 4.14. Mean Total Zinc (in µg/L) recorded at each site sampled in the upper Grose River for three sampling occasions April to June 2003, plus/minus one standard error. ANZECC (2000) guideline level for ecosystem protection is indicated.

Given that water hardness was classified as soft (< 60 mg/L CaCO3) at all sites in this study, the recommended guideline for total zinc levels for maintaining ecosystem health (ANZECC, 2000) is 5 µg/L. The three Grose River sites below the mine drainage always exceeded this ecosystem protection guideline.

Table 4.12 F-statistics and associated probabilities from analyses of variance of (log X + 1) transformed total zinc data, (log X + 1) transformed total nitrogen data and (log X + 1) transformed total phosphorus data from sites sampled in the upper Grose River and its tributaries between April and June 2003.

Source Variable of variation Zinc Total nitrogen Total phosphorus Df MS F P Df MS F P Df MS F P Main effects

Site 9 4.74 241 <0.001 9 6.45 241 <0.001 9 3.78 130.7 <0.001

Error 44 0.02 50 0.03 50 0.03

106 Figure 4.15. Plates 1 to 5. Reference sites in Grose River system. Plate 1 is Grose River at Engineers track (GEN). Plates 2 and 3 are the Grose River below Koombandah Brook (GDK). Plate 4 is Victoria Creek (VIC) and Plate 5 is Hat Hill Creek (HHU). Plate 3. Plate 1. Plate 2.

Plate 5. Plate 4. 107 Figure 4.16. Plates 6 to 9. Canyon colliery and Dalpura Creek. Plate 6 shows surface workings at the colliery. Plate 7 shows Dalpura Creek (orange colour) entering Grose River. Plate 8 shows Canyon colliery sealed mine adit. Plate 9 shows the author sampling Dalpura Creek, just before it enters the Grose River.

Plate 7.

Plate 6.

Plate 8. Plate 9. 108 Figure 4.17. Plates 10 to 13. Blackheath STP and Hat Hill Creek. Plate 10 shows Blackheath STP and one of its trickling filters. Plates 11 shows an open channel of STP effluent cascading overland towards Hat Hill Creek. Plate 12 shows the STP effluent entering Hat Hill Creek (looking upstream). Plate 13 shows Hat Hill Creek (looking downstream) at the point that the STP effluent enters the Creek. Plate 10. Plate 11.

Plate 12. Plate 13. 109 Figure 4.18. Plates 14 to 17. Grose River effluent receiving sites. Plate 14 shows the Grose River substrate, and Decapod Crustacean, below Dalpura Creek (GDD). Plate 15 shows the Grose River at Hungerfords track (GHU). Plate 16 shows the Grose River at Hungerfords track (GHU). Plate 17 shows taking physio- chemical readings at the Grose River at Burra Korrain (GBK). Plate 14.

Plate 15.

Plate 17. Plate 16. 110 Figure 4.19. Plates 18 to 21. Macroinvertebrates from the Grose River system. Plate 18 show two Simuliidae (Diptera: Insecta). Plate 19 Hydropsychidae (Trichoptera: Insecta) Plate 20 Leptophlebiidae (Ephemeroptera: Insecta). Plate 21 shows author collecting macroinvertebrates from the Grose River, below Dalpura Creek (‘kick-sampling’). Plate 19. Plate 18.

Plate 20. Plate 21. 111 4.4 Discussion

The effluent from Blackheath STP and zinc-rich mine drainage from the abandoned Canyon Colliery, strongly influenced macroinvertebrate community structure in the Grose River system. Each waste discharge caused different pollution-related changes to the macroinvertebrate communities and water chemistry of the river system. Zinc contamination affected about five kilometres of the Grose River and a combination of zinc and STP-related contamination covered a further one kilometre, below the confluence with Hat Hill Creek. The extent of the impacts from the two waste sources stretched over at least six kilometres of the Grose River. Chemical results at the lowest site sampled on the Grose River still exceeded some ecosystem protection criteria, indicating that the impact probably stretched further down the Grose River and beyond the limits of sampling (ANZECC, 2000; Figures 4.12, 4.13 and 4.14).

Physio-chemical impacts from the abandoned Canyon Colliery caused changes to the water quality downstream (e.g. electrical conductivity, Figure 4.8; pH, Figure 4.9; Hardness, Figure 4.10) and was very likely to be the source of the impacts on the macroinvertebrate communities. Of most biological concern was a massive increase in zinc levels of the Grose River from undetectable levels (< 10 µg/L) above Dalpura Creek, to 380 µg/L below the mine drainage (Figure 4.14). This impact was present for at least 6 kilometres downstream to the lowest site in the study. The elevated zinc levels at the lowest study site (GHU; Figure 4.14) indicated that zinc contamination remained in the Grose River much further downstream than this study extended. The ANZECC (2000) ecosystem protection guideline for zinc in fresh waters was 5 µg/L, in the soft water of the Grose River, thus the levels observed were at toxic levels, up to 80 times the guideline levels. The highest mean level of zinc in the Grose River (380 µg/L) was much lower than maximum zinc levels detected in other metal-polluted Australian rivers. The highest zinc levels recorded in South Esk River, Tasmania, was 3,700 µg/L (Norris, 1982); 61,000 µg/L in the River Dee, Queensland (Mackey, 1988); 38,700 µg/L Daylight Creek, NSW, (Napier, 1992) and 11,700 µg/L in the Molongolo River, NSW (Sloane & Norris, 2003). In

112 contrast, metal pollution in this study was caused by an abandoned coal mine, elevated metal levels from the other studies originated from abandoned metalliferous mines.

Blackheath STP also caused physio-chemical changes to the water quality downstream in Hat Hill Creek. Total nitrogen and phosphorus levels were elevated about 100 times below the STP than they were upstream of the discharge point (Figure 4.12). In addition electrical conductivity and hardness were substantially elevated (Figure 4.8, 4.10), and pH was relatively less affected (Figure 4.9). This impact was not restricted to the tributary of Hat Hill Creek, which directly received the STP outfall. Water quality was also impaired in the Grose River, where the influence of the STP discharge extended five km below the STP. The ANZECC (2000) ecosystem protection trigger level for TN, in upland freshwater streams of NSW is 200 µg/L. Nitrogen levels in the Grose River greatly exceeded these levels (1500 µg/L; Figure 4.12) in an ecosystem where background levels are much lower (<100 µg/L above the Hat Hill Creek confluence with the Grose River). The ecosystem protection trigger level for TP in NSW upland freshwater streams is 20 µg/L (ANZECC, 2000). Background levels in the Grose River catchment were generally below this guideline, but exceeded the recommended level below the STP outfall and at the lowest site in the Grose River which was 40 µg/L (GHU; Figure 4.12). Although not measured in this study, ammonia has been discharged from Blackheath STP into Hat Hill Creek in 1997/98 at mean levels of 4 mg/L (Sydney Water, 1999). According to ANZECC guidelines (1992), such levels of ammonia are almost certainly toxic to stream biota, at the ambient temperatures and pH of Hat Hill Creek.

The Grose River is suffering the effects of both zinc pollution and nutrient enrichment, although only one sampling site (GHU) in this study was impacted by both pollution sources. The elevated zinc (Tzn), phosphorus (TP) and nitrogen (TN) were all significantly above background levels at the lowest sampling site on the Grose River, indicating that there was a sustained impact from both pollution sources. Based on the rate of change in levels within the area studied, these pollutants were likely to remain above guideline levels for many kilometres downstream. Such a combination of metal pollution from mine drainage and STP organic pollution to an

113 otherwise clean river system has not been reported before in Australia for a high conservation value World Heritage waterway.

Chironomid larvae positively responded to the sewage pollution and negatively to zinc pollution. According to this evidence chironomid larvae responded in accordance with their pollution tolerant reputation only to sewage effluent wastes. They were one of four key taxa that displayed a differential response to the two pollution types (Table 4.6, 4.8, 4.9).

A similar increase in abundance of chironomid larvae at sewage polluted sites, compared to unpolluted waters, was also reported (Wright, 1994; Wright et al., 1995) in a study of biological impacts from Wentworth Falls STP (approximately 20 kilometres from the current study area). The current sewage-pollution chironomid findings are supported by the widely used biotic indices, SIGNAL-95 and SIGNAL-HU97, which were developed for the Hawkesbury-Nepean (SIGNAL-95) and Hunter River (SIGNAL-HU97) areas of NSW. Both indices classified chironomids as being ‘pollution tolerant’ for organic pollution (Chessman, 1995; Chessman et.al., 1997). Two other Australian studies (both in south-eastern Queensland) of sewage-polluted streams also recorded increased chironomid larval abundances below the STP discharge point (Arthington, et al., 1982; Cosser, 1988). In contrast, Whitehurst and Lindsey (1990) reported that chironomids showed no particular preference for organic pollution as they were at similar levels of abundance at sewage polluted and clean reference sites in the River Adur, England.

The finding that chironomid larval abundance responded negatively to metal-pollution, yet was the second most abundant taxon at the most metal- polluted site, was consistent with other Australian studies. Chironomids were reported to be dominating the community at one of the most metal polluted sites on the South Esk River in Tasmania (Norris et. al., 1982). Sloane and Norris (2003) also reported a greater level of metal-tolerance by chironomids in the metal polluted Molongolo River, NSW, where three subfamilies Orthocladiinae, Chironominae and Tanypodinae were amongst the most metal-tolerant taxa in their study. An earlier study of biota in the metal polluted Molongolo River by Nicholas and Thomas (1978) reported very few

114 taxa in the most mine drainage polluted site, comprising only chironomids, Hydracarina, Copepoda and Oligochaetae worms. In the heavily metal- polluted River Dee in Central Queensland, chironomids were one of the few taxa to tolerate the most highly polluted sites (Mackey, 1988). They were also the most abundant group at highly metal-polluted reaches of Daylight Creek in the NSW Central Tablelands (Napier, 1992). Similar observations have been made in northern hemisphere freshwater studies. High chironomid densities have been observed to be associated with pollution from mine drainage (Winner et.al., 1980; Kelly, 1991; Gerhadt et.al., 2004). In contrast, a study of heavy metal polluted stream biota in New Zealand (Hickey and Clements, 1998) found that chironomid abundance was low at some metal polluted sites, although it was observed that they were very high at others. These inconsistencies may have been confounded by other factors not identified by the researchers.

The results of heavy metal impacts on non-chironomid macroinvertebrates in this study are broadly similar to previous Australian studies on metal- pollution (Nicholas & Thomas, 1978; Norris et al.,1982; Sloane & Norris, 2003). The metal-polluted Molongolo River, below the Captains Flat mining area (New South Wales), recorded declines of macroinvertebrate taxon richness and abundance, and loss of well-known sensitive taxa such as Leptophlebiidae (Table 4.4) and Baetidae mayfly larvae, compared to unpolluted sites. A study of the zinc-polluted South Esk River in Tasmania by Norris et al. (1982), recorded a similar reduction in taxon richness of macroinvertebrates to mine pollution and loss of Leptophlebiidae mayfly larvae, and a moderate reduction in Baetidae mayfly larvae abundance.

Hydropsychidae caddisfly larvae (Figure 4.6) demonstrated an unusual positive response to zinc-pollution, the only such response in this study where a taxon was more abundant at the zinc polluted site than at unpolluted sites (Table 4.4 and 4.6). In contrast, other Australian metal pollution studies have found that Hydropsychidae caddis larvae responded negatively to metal-pollution in the Molongolo River (NSW) and River Dee (Queensland; Norris, 1986; Mackey, 1988). In Daylight Creek (NSW) Hydropsychidae caddis larvae were abundant at some metal polluted sites (Napier, 1992) and in the South Esk River (Tasmania) Hydropsychidae

115 caddis larvae were abundant at all but one metal polluted site (Norris, 1982). Hydropsychidae caddis larvae were classified as sensitive to metal-pollution, according to a metal-pollution biotic index SIGNAL-MET rating (unpublished data from Bruce Chessman, in Sloane & Norris, 2003). One other Australasian study (Hickey & Clements, 1998) has detected metal-pollution tolerance of Hydropsychidae. This was observed in a study of New Zealand metal-polluted waterways and it was reported that there was a dominance of Hydropsychidae in metal polluted waters.

Some macroinvertebrate taxa collected beyond the initial outfalls of both pollution-sources were indicative of early stages of downstream pollution- recovery (Table 4.4). Taxa indicative of recovery (Scirtidae beetle larvae, Gripoteryigidae stonefly larvae, Simuliidae blackfly larvae, Leptoceridae caddisfly larvae, Philpotamidae caddis larvae) from zinc pollution were first observed in the Grose River when zinc concentrations had dropped from 388 µg/L to 261 µg/L, 1.8 km downstream of the initial mine-affected site (see GDD to GBK in tables 4.5 and 4.14). Similar recovery was also noted five kilometres below the STP discharge (HHG, HHD; Table 4.5). The STP- recovery taxa (Leptophlebidae mayfly larvae, Elmidae beetles, Baetidae mayfly larvae, Hydropsychidae caddis larvae and Caenidae mayfly larvae) increased their abundance downstream where nutrient levels dropped (from 14,300 µg/L to 750 µg/L TN; 500 µg/L to 189 µg/L TP; see Figures 4.12 and 4.13).

This study found that the Grose River is suffering the effects of zinc-pollution from an abandoned coal mine and sewage-pollution from an active STP. The two waste discharges caused adverse ecological changes to the macroinvertebrate fauna, and the water quality, of the Grose River and tributaries for many kilometres below the two discharge points. Based on the rate of recovery within the study area, it is likely that the water pollution continued well below the lowest site sampled on the Grose River (GHU). The two waste discharges affected different waterways within a small (8,000 hectare.) sub-catchment and enabled pollution responses to macroinvertebrate groups to be calculated and compared.

116 The total zinc water quality guideline (for ecosystem protection) of 5 µg/L appeared to be appropriate for protecting stream ecosystems, although it was not possible to determine the full recovery of the Grose River macroinvertebrate community from zinc pollution due to the confounding effects of STP effluent, via Hat Hill Creek. The elevation of total zinc levels was associated with the loss of sensitive macroinvertebrates from the community (Table 4.8), coupled with a reduction in taxon richness (Figure 4.6) and macroinvertebrate abundance (Table 4.5).

Total nitrogen and phosphorus guidelines (ANZECC, 2000) also appeared to be appropriate ecosystem protection guidelines, although the impact of the STP effluent on the macroinvertebrate community of Hat Hill Creek immediately below Blackheath STP was probably caused by one or more constituents not measured in this study. Elevated nitrogen and phosphorus levels (well above guideline levels) also existed in Hat Hill Creek (at HHG) and in the Grose River (at GHU) but the macroinvertebrate community at those sites was very similar to that found in unpolluted sites. The high levels of ammonia that have previously been detected in Blackheath STP effluent (median levels of 4 mg/L reported in Sydney Water, 1998) suggested that ammonia toxicity may occur in Hat Hill Creek. Elevated nutrient levels may, therefore, have influenced the very high abundance of macroinvertebrates recorded in the Grose River (GHU) on two sampling occasions (Figure 4.5), possibly due to the increased primary production stimulated by the artificially hight nutrient levels in the river.

The response of macroinvertebrate groups to STP effluent recorded in this study is broadly similar to the majority of tolerance grades provided by the ‘SIGNAL-2’ macroinvertebrate organic pollution index (Chessman, 2003). Some key taxa did exhibit a different affinity to organic pollution (Table 4.6). Baetidae mayflies were more sensitive to organic pollution than the SIGNAL-2 grade (five of 10). Limpet gastropods and Simuliidae larvae (Figure 4.19) increased in abundance below the STP (Table 4.9) in contrast to predicted responses of the SIGNAL-2 grades (four and five respectively). The taxonomic grades in the unpublished SIGNAL-MET index (Sloane & Norris, 2003) appeared to accurately reflect the zinc pollution affinities recorded in this study (Table 4.6). The one exception was for

117 Hydropsychidae caddis larvae (Figure 4.19), the only taxonomic group that had a higher abundance at the most zinc contaminated site, than at reference sites in this study (Table 4.8). These results indicate that the SIGNAL-MET and SIGNAL 2 (organic) tolerance grades for Hydropsychidae, which are both of six out of 10, were too high for metals and too low for organic pollution.

Chironomid larvae were one of many groups to exhibit a negative response to zinc pollution and were one of the only groups to exhibit a mild positive response to sewage pollution. Although chironomid larvae were less abundant at the zinc-polluted site than at unpolluted sites, they were the second most abundant taxon, after Hydropsychidae caddis larvae, at the zinc-polluted site (Table 4.5). This relative level of chironomid abundance at the zinc-polluted site in this study is similar to other studies. The sewage- pollution tolerance of chironomids, as observed in this study, has also been commonly found in other studies. A Queensland organophosphate study (Ward et al., 1995) found chironomid species to be more sensitive than non- chironomid stream biota. It was concluded that identification of chironomids to species-level was important to gain maximum information on pollution effects.

118 Chapter 5 Spatial and temporal variation of Chironomid exuviae from two small upland streams in the Blue Mountains, NSW.

5.1 Introduction:

Collecting chironomid exuviae from streams, rivers and lakes has long been recognised internationally as an excellent technique for assessing chironomid biodiversity (Thienemann, 1910; Humphries, 1938; Brundin, 1966) and they have also been used as surrogates in water quality studies (Wilson & Bright, 1973, Wilson, 1977). In Australia, however, their usage is in its infancy (Hardwick et al., 1995; Cranston et al., 1997; Wright & Cranston, 2000; Dimitriadis & Cranston, 2001). Collecting chironomid exuviae does require a different approach from that generally used for collecting other freshwater macroinvertebrates. An inadequate sample of exuviae could risk misrepresentation of chironomid species living within a water body.

Adult chironomids emerge from their pupal lifestage on the water surface and discard their exuviae once they have dried their wings (Langton, 1995). Emergence often occurs after sunset and during the night hours (Learner et al., 1990), although some species emerge during daylight (Learner et al., 1990). The exuviae remain afloat for a short period of time. Although not firmly established, researchers (Wilson & Bright, 1973; Hardwick et al., 1995) have suggested that it is only one or two hours. Surface netting is required to capture floating (and near-surface) exuviae. Typically a ‘drift’ net is placed in a waterway for an extended period of time to filter exuviae from the stream flow (e.g. Hardwick et al., 1995; Cranston et al., 1997).

Issues of seasonality, latitude and diurnal variation in chironomid emergence have been reviewed by Armitage (1995). Hardwick et al. (1995) have undertaken the only previous Australian study that explored diurnal variation of exuviae from flowing water. Their study was undertaken in waterways in tropical northern Australia, at least one of which only flowed seasonally. They found that chironomid exuviae numbers and species richness varied in a diurnal pattern and concluded that sampling nets should be in place for at

119 least 24-hours. The same authors subsequently conducted a study that used chironomid exuviae to measure acid mine drainage impacts (Cranston et al., 1997). No similar information has been published on temporal and spatial variability and sampling methodology required to conduct rigorous surveys of chironomid exuviae in temperate upland Australian streams, although differences in climatic conditions between the tropics and temperate Australian uplands may create differences in species distribution and, therefore, potentially exuviae temporal patterns.

In this chapter I investigated chironomid exuviae in two small upland waterways in temperate Australia. Using two known pristine Grose River catchment reference sites (Chapter 4), the suitability of drift net collection of exuviae for surveying chironomid species was evaluated. The diurnal variation of chironomid exuviae species and abundance was investigated. The temporal variation of chironomid exuviae over multiple 24-hour sampling occasions was also addressed. To test whether the emergence patterns varied from waterway to waterway, diurnal patterns of exuviae at a second sampling site on a different waterway in the same area were compared. The results from this study were used to develop a chironomid exuviae sampling methodology for a pollution survey in Grose River catchment waterways (Chapter 6).

120 5.2 Material and Methods

5.2.1 Study area Field work was carried out in two upland waterways, Hat Hill Creek and

Victoria Creek, in the upper Grose River catchment in the Blue Mountains,

NSW, south-eastern Australia (33 35’S, 150 18’E; Figure 5.1; Chapter 4, Study area).

Two sites were sampled: Hat Hill Creek above Blackheath sewerage treatment plant discharge (HHU) and Victoria Creek (VIC) (Figure 5.1). Both sites were pristine reference sites based upon macroinvertebrate and water quality results (Chapter 4). They were both dilute, poorly buffered and low in nutrients. One was a first order stream (HHU) and the other second order (VIC). Elevations of the two sites were 485 metres AHD (VIC) and 965 metres AHD (HHU) respectively. Both streams were permanent small and shallow, fast-flowing waterways. The stream channels were rocky bottomed, sometimes bedrock and frequently dominated by boulders and cobbled stream-beds.

121 Figure 5.1 Map of Chironomid exuviae collection sites and waterways in the upper Grose River system. Square symbols indicate sampling sites. Approximate catchment boundary is dashed line. Inset shows location of study area in south- eastern Australia.

N

Koombandah Brook O

Dalpura Creek

Birrabang Creek

VIC

Grose River

33º 35’ S

Victoria Hat Hill Creek Creek

HHU Blackheath STP discharge

150º 15’ E 2 km

122 5.2.2 Sampling sites and sample collection Sampling was undertaken during the austral summer, between December 2003 and February 2004. Chironomid exuviae were collected from Hat Hill Creek on four 24-hour occasions (11 December 2003, 18 December 2003, 2 January 2004 and 19 February 2004) and from one site on Victoria Creek on one 24-hour occasion (1 January 2004).

A steel-framed drift 250 micron net with an entry aperture of 30 cm2 and a length of 60 cm was placed across a narrow section of the stream. The net was placed in the best position available to capture the majority of stream flow at both sites. At Hat Hill Creek the net was placed in a very narrow location where it intercepted the whole stream flow. Plastic sheeting was placed between the stream banks and the net to ensure that, apart from some minor leakage, the entire stream flow was filtered by the net. Victoria Creek was wider than Hat Hill Creek, and the net was placed at a narrow section that captured about 75 % of creek flow.

The net was removed every two hours, over a 24-hour cycle, and the net contents were flushed into a plastic collecting tray, bottled and then preserved in 70 % ethanol. The net was then replaced in the stream for the next two hours. In the laboratory the samples were sorted from detritus with the aid of a dissecting microscope (X 40 magnification). All exuviae were slide-mounted in temporary mountant and were identified to species level using a compound microscope (up to X 400 magnification) and keys to Australian chironomids pupal exuviae (Cranston, 1996, 2000a).

123 5.2.3 Temperature and rainfall conditions during the study During sampling, the weather conditions were generally warm to hot during the day, and cool to mild at night. The local summer weather conditions were slightly warmer and drier than average (Figure 5.2). Water temperature trends followed air temperature but the amplitude was narrower (Figure 5.2, 5.3). Monthly rainfall was lower than average for four of the six months before and during the study (Figure 5.4).

35

30

25

20

15

10

5

0 11:00 13:00 15:00 17:00 19:00 21:00 23:00 01:00 03:00 05:00 07:00 09:00

Figure 5.2 Air temperature (º Celsius) recorded every two hours at Victoria Creek (black) and Hat Hill Creek during the diurnal studies (12 December 2003 = red; 18 December 2003 = green; 2 January 2004 = blue line; 19 February 2004 = orange).

35

30

25

20

15

10

5

0 11:00 13:00 15:00 17:00 19:00 21:00 23:00 01:00 03:00 05:00 07:00 09:00

Figure 5.3 Water temperature (º Celsius) recorded every two hours at Victoria Creek (black line) and Hat Hill Creek during the diurnal studies (12 December 2003 = red; 18 December 2003 = green; 2 January 2004 = blue line; 19 February 2004 = orange).

124 160

140

120

100

80

60

40

20

0 September October November December January February

Figure 5.4 Rainfall recorded (in mm / month), in the months prior to and during this study, from the rainfall station in the study area, Mount Boyce (black bar) from September 2003 to February 2004. Historic mean monthly rainfall indicated by unshaded bars. Source: www.bom.gov.au accessed May 2004.

5.2.4 Data Analysis Procedures

Numbers of chironomid exuviae and species number were logarithmically transformed (log X + 1) to better approximate a normal distribution. Analysis of variance (ANOVA) was used to determine if numbers of exuviae or species varied according to time of sampling at Hat Hill Creek.

The results for 0700 hours and 0900 hours for 11 December 2003 (Hat Hill Creek) were replaced with the mean values for 0700 hours and 0900 hours for the other three sampling occasions, due to very heavy rainfall, increased stream flow and higher exuviae numbers.

125 5.3 Results

5.3.1 Diurnal variation results A total of 1813 chironomid exuviae was collected and examined from Hat Hill Creek and Victoria Creek: 1411 were collected from Hat Hill Creek over four 24-hour sampling occasions (Table 5.1). Substantially more (615) were collected in early December than on the other occasions (146 – 331) (Table 5.1). Of these 402 were collected and classified from Victoria Creek (Table 5.4). Overall, 54 species were collected across the study (Table 5.3) and of these 24 were only recorded from Hat Hill Creek and 16 species only from Victoria Creek (Table 5.3). Number of exuviae collected from Hat Hill Creek showed a distinct diurnal pattern (Figure 5.5) which was highly significant (Table 5.6). Most chironomids emerged in the dusk and during the night (Table 5.1, Figure 5.5). A mean of less than 15 exuviae per two hours was recorded in daylight hours (Table 5.1). The difference in numbers over a two hour period varied between a mean of <8 exuviae (1100 –1300 hours) compared to a mean of 50 exuviae (1900 – 2100 hours) (Table 5.1).

Species richness also followed a distinct diurnal pattern that was significant (Table 5.5, Figure 5.6). The mean number of species from each two hour sampling block more than doubled from five (1300 – 1500 hours) to 12.75 (2300 – 0100 hours) (Table 5.2, Figure 5.6).

Table 5.1 Abundance of chironomid exuviae collected at Hat Hill Creek. Collected between 12 December 2003 and 19 February 2004.

Date Time of day (Hours) 1100 1300 1500 1700 1900 2100 2300 0100 0300 0500 0700 0900 Total

12/12/03 22 11 15 15 12 81 50 77 34 35 53* 210* 615 18/12/03 10 9 7 11 8 120 46 25 34 21 11 14 316 2/1/04 17 10 9 6 21 12 53 63 52 30 38 23 334 19/2/04 7 1 1 16 28 19 7 18 16 12 16 5 146 Mean 14 7.75 8 12 17.25 53.5 39 45.75 34 24.5 21.7 14 * Not used for calculation of average due to influence of heavy rain

126 Table 5.2 Number of species of chironomid exuviae collected at Hat Hill Creek. Collected between 12 December 2003 and 19 February 2004.

Date Time of day (Hours) 1100 1300 1500 1700 1900 2100 2300 0100 0300 0500 0700 0900

12/12/03 10 7 7 10 5 10 12 18 12 16 14* 22* 18/12/03 7 6 6 6 6 22 13 10 15 10 6 5 2/1/04 7 7 6 5 7 7 10 15 10 10 11 10 12/2/04 6 1 1 10 10 10 4 9 9 6 7 4 Mean 7.5 5.25 5 7.75 7 12.25 10.25 12.75 11.5 10.5 8 6.3 * Not used for calculation of average due to influence of heavy rain

127 Table 5.3 List of chironomid species recorded at each site in this study

Sub-family Species VIC HHU Chironominae Tanytarsus sp. 1 X Tanytarsus sp. 2 (nr. manlyensis) X Tanytarsus sp. 3 (multidir spines) X Tanytarsus sp. 4 (nr. manlyensis) X X Tanytarsus sp. 5 (nr. spinosus) X Tanytarsus liepae X Tanytarsus sp. 6 (nr. liepae) X Tanytarsus nr.bispinosus X Tanytarsus nr. B3 X Cladotanytarsus sp. X Paratanytarsus nr. furvus X Rheotanytarsus flabellatus? X Rheotanytarsus jeffereyi X Stempellina ?australiensis X X Riethia stictoptera X Riethia zeylandica X X Riethia nr. ‘V4’ X Polypedilum nr.'allocasia' X X Polypedilum sp. 1 X Polypedilum sp. 2 X Harrisius sp. X Unidentified chironominae sp.2 X

Orthocladiinae Thiennemanniella sp. 1 X X Parametriocnmeus sp. X Nanocladius sp. X Eukiefferiella insolita X X Rheocricotpous sp. X MO5 X X Stictocladius sp. X Parakiefferiella nr. ‘variegatus’ X X Parakiefferiella sp. 1 X Cricotpus nr. acornis X X Cricotpus sp.2 X X SO1 X SO4 X X Botryocladius grapeth X Botryocladius collessi X Genus Australia A sp. X Paralimnophyes sp. X Nr. ‘SO1’ sp. X Unidentified Orthocladinae sp. 2 X Unidentified Orthocladinae sp. 3 X Unidentified Orthocladinae sp. 5 X

Tanypodinae Pentaneurini genus A sp. X X Pentaneurini genus D sp. X Pentaneura sp. nov. X X Paramerina parva X X Apectrotanypus maculatus X Ablabesmyia sp. X Larsia sp. X Procladius paludicola X

Podonominae Podochlus australiensis X Podomonopsis ? sp. X Podomonopsis discoceros X

Number of species 54 30 38

128 The single 24-hour collection of chironomid exuviae at Victoria Creek showed a similar diurnal pattern of abundance and species richness (Figure 5.7 and 5.8) to that recorded from Hat Hill Creek (Figure 5.5, 5.6): the majority of exuviae were collected in night hours. The lowest number (4) was collected between 1300 - 1500 hours while up to 93 exuviae were collected between 2300 - 0100 hours (Table 5.4). The number of species collected trended upwards from three between 1300 - 1500 hours and peaked in the 2100 - 2300 hours and the 0100 - 0300 hours periods with 15 species, after which there was a downward trend (Table 5.4, Table 5.5).

Table 5.4 Number of chironomid exuviae collected at Victoria Creek collected over a 24-hour period on 31 December 2003.

Time of day (Hours) 1100 1300 1500 1700 1900 2100 2300 0100 0300 0500 0700 0900 Total 1 January 15 11 4 11 20 49 79 93 48 38 21 13 402

Table 5.5 Number of chironomid species collected at Victoria Creek collected over a 24-hour period on 31 December 2003.

Time of day (Hours) 1100 1300 1500 1700 1900 2100 2300 0100 0300 0500 0700 0900 1 January 9 7 3 8 9 14 15 13 15 12 8 6

Table 5.6 F-statistics and associated probabilities from analyses of variance of number of exuviae (log X + 1 transformed) and number of species (log X + 1 transformed) collected from Hat Hill Creek on the four sampling occasions.

Source of Variable variation Number of exuviae Number of species Df MS F P-value Df MS F P-value

Time of 11 8.55 3.07 0.0054 11 0.89 2.48 0.02 sampling

Error 36 2.79 36 0.04

129 100

90

80

70

60

50

40

30

20

10

0 11:00 13:00 15:00 17:00 19:00 21:00 23:00 01:00 03:00 05:00 07:00 09:00

Figure 5.5. Back-transformed mean number of chironomid exuviae (plus/minus standard error) collected from at Hat Hill Creek for each two hour sampling period, from four 24-hour sampling occasions in summer 2003/04. The base of the figure is shaded to indicate the approximate times of full light (white), darkness (black) and dawn / dusk (grey).

18

16

14

12

10

8

6

4

2

0 11:00 13:00 15:00 17:00 19:00 21:00 23:00 01:00 03:00 05:00 07:00 09:00

Figure 5.6. Back-transformed mean number of species of chironomid exuviae (plus/minus standard error) collected from Hat Hill Creek for each two hour sampling period, from four 24-hour sampling occasions in summer 2003/04. The base of the figure is shaded to indicate the approximate times of full light (white), darkness (black) and dawn / dusk (grey).

130 100

75

50

25

0 11:00 13:00 15:00 17:00 19:00 21:00 23:00 01:00 03:00 05:00 07:00 09:00

Figure 5.7 Total number of chironomid exuviae collected from Victoria Creek for each two hour sampling period, from a single 24-hour sampling occasion, conducted on 31 December 2003 to 1 January 2004. The base of the figure is shaded to indicate the approximate times of full light (white), darkness (black) and dawn / dusk (grey).

16

12

8

4

0 11:00 13:00 15:00 17:00 19:00 21:00 23:00 01:00 03:00 05:00 07:00 09:00

Figure 5.8 Total numbers of species of chironomid exuviae collected from Victoria Creek for each two hour sampling period, from a single 24-hour sampling occasion, conducted on 31 December 2003 to 1 January 2004. The base of the figure is shaded to indicate the approximate times of full light (white), darkness (black) and dawn / dusk (grey).

131 At Hat Hill Creek, on the first 24-hour sampling occasion, heavy rain fell from 0600 hours onwards. The level of the creek, the discharge volume and current increased progressively from then until 0900 hours (personal observation). The number of chironomid exuviae collected at 0700 hours and particularly 0900 hours was much higher than was collected at other sampling times (Table 5.1, 5.2, Figure 5.9).

250

200

150

100

50

0 11:00 13:00 15:00 17:00 19:00 21:00 23:00 01:00 03:00 05:00 07:00 09:00

Figure 5.9 Total number of chironomid exuviae collected from Hat Hill Creek for each two hour sampling period, from a single 24-hour sampling occasion conducted on 12-13 December 2003. The base of the figure is shaded to indicate the approximate times of full light (white), darkness (black) and dawn/dusk (grey).

5.3.2 Diurnal patterns in individual species The exuviae of the most numerous species found at Hat Hill Creek, Paratanytarsus nr. furvus (Figure 5.16) and Tanytarsus sp. 5 (nr. spinosus), exhibited diurnal patterns (Figure 5.10, 5.11, 5.15). They were collected in low numbers (<5 per two hours) in daylight hours, except for one notable exception (0700 - 0900 hours, 12 December) after heavy rain and subsequent increases in stream velocity and volume. Abundance was highest for both species at night (1900 hours to 0100 hours). Numbers generally fell steadily beyond this time (Figure 5.10, Figure 5.11).

132 40

30

20

10

0 11:00 13:00 15:00 17:00 19:00 21:00 23:00 01:00 03:00 05:00 07:00 09:00

Figure 5.10 Number of Paratanytarsus nr. furvus exuviae collected from each two hour sampling period at Hat Hill Creek. Black bar represents 12 December, unshaded December 18, grey is January 2 and cross-hatched February 19. The base of the figure is shaded to indicate the approximate times of full light (white), darkness (black) and dawn/dusk (grey).

20

15

10

5

0 11:00 13:00 15:00 17:00 19:00 21:00 23:00 01:00 03:00 05:00 07:00 09:00

Figure 5.11 Number of Tanytarsus sp. 5 (nr. spinosus) exuviae collected from each 2 hour sampling period at Hat Hill Creek. Black bar represents 12 December, unshaded December 18, grey is January 2 and cross-hatched February 19. The base of the figure is shaded to indicate the approximate times of full light (white), darkness (black) and dawn/dusk (grey).

133 60

45

30

15

0 11:00 13:00 15:00 17:00 19:00 21:00 23:00 01:00 03:00 05:00 07:00 09:00

Figure 5.12 Number of most numerous three species of exuviae collected from each 2 hour sampling period at Victoria Creek, 31 December 2003 to 1 January 2004. Black bar represents number of ‘M05’, unshaded represents number of Riethia zeylandica and cross-hatched bar represents number of Tanytarsus nr. bispinosis. The base of the figure is shaded to indicate the approximate times of full light (white), darkness (black) and dawn/dusk (grey).

The most numerous species at Victoria Creek were Riethia zeylandica and Tanytarsus nr. bispinosus. Both exhibited a strong diurnal pattern (Figure 5.12). They were collected in very low numbers in daylight hours, abundance rising after sunset, and were recorded at the highest levels between 2300 hours to 0100 hours.

The species ‘MO5’ was collected from both sites (Table 5.3; Figure 5.16). It exhibited a post-dusk peak and a smaller dawn peak (Victoria Ck, Figure 5.12) and was found in small numbers during daylight.

Each additional two-hour sample of exuviae added more species of chironomids at both Hat Hill Creek and Victoria Creek (Figure 5.13 and 5.14). The rate of cumulative increase in species number rose quickly during the first 24-hours of sampling (Figure 5.13 and 5.14), and then the rate of increase diminished over the following three 24-hour occasions at Hat Hill Creek (Figure 5.14).

134 40

35

30

25

20

15

10

5

0

Figure 5.13 Victoria Creek. Progressive total number of species collected from the single 24-hour occasion. The line represents the cumulative number of different chironomid species collected in two hourly increments over a single 24-hour occasion.

40

35

30

25

20

15

10

5

12 December 18 December 2 January 19 February 0 1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49

Figure 5.14 Hat Hill Creek. Progressive total number of species collected from the four 24-hour occasions over the austral summer. The line represents the cumulative number of different chironomid species collected in two hourly increments over four 24-hour occasions.

135 Figure 5.15. Plates 1 to 4. Exuviae Tanytarsus sp. 5 (nr. spinosus) from diurnal study. Plate 1 shows the abdominal segments (ventral view). Plate 2 is a view of the postero-lateral comb. Plate 3 is a ventral view of abdominal segments II, III and IV. Plate 4 is a view of the thoracic horn (almost transparent) . Plate 1. Plate 2.

Plate 3. Plate 4. 136 Figure 5.16. Plates 5 to 8. Exuviae from diurnal study. Plate 5 is Paratanytarsus nr. furvus showing the abdominal segments II, III and IV (ventral view). Plate 6 shows the postero-lateral comb of Paratanytarsus nr. furvus. Plate 7 is a lateral-dorsal view of Orthocladiinae species ‘MO5’. Plate 8 is a ventral view of the lowest abdominal segments of Orthocladiinae species ‘MO5’. Plate 5. Plate 6.

Plate 7. Plate 8. 137 5.4 Discussion

It was observed that overall chironomid exuviae abundance and species richness varied diurnally in the two upland streams studied. The highest exuviae abundance and diversity was encountered during the night, particularly between sunset and 0100 hours. These results provide evidence that most common chironomid species exhibit nocturnal emergence (Figure 5.5). This pattern was also demonstrated for the most abundant species e.g. Paratanytarsus nr. furvus, Tanytarsus spinosus, Riethia zeylandica and Tanytarsus nr. bispinosus (Figure 5.10, 5.11 and 5.12). The only comparable Australian study (Hardwick et al., 1995) was undertaken in tropical Australian waterways and a similar nocturnal emergence pattern was also observed. Despite being scanty, these data taken from two very different areas in Australia and showing the same emergence patterns indicate that many common Australian species have a nocturnal emergence pattern. Despite this overall pattern, many uncommon species were collected, often sporadically, at different times of the day. This implies that there are differences in emergence times among species although since the majority of these species were collected in small numbers the patterns are difficult to interpret. Previous international studies have reported many different diurnal emergence patterns (Wartinbee, 1979; Leaner et al., 1990; Brittain & Eikeland, 1988), particularly from cold climates (Danks & Oliver, 1972).

The diel periodicity of exuviae observed has important implications for the sampling methodologies used for collecting Australian chironomid exuviae. To maximize sampling outcomes, the drift sampling net should be put in place for at least one 24-hour cycle. If the sampling net were to be placed in a stream for less than 24-hours, the dusk to dawn period would be the best time to obtain the most abundant species. This was the same conclusion reached by Hardwick et al. (1995).

It was also observed that numbers of new species continued to increase beyond a single day. For example, the species inventory from Hat Hill Creek rose steeply during the first 24-hours of collection and more species were progressively collected on subsequent days, although at a decreasing rate.

138 New species were still being collected at the end of the fourth 24-hour diurnal collection day. This suggests that a great deal of sampling effort would have to be expended to determine the total chironomid inventory from a waterbody.

Heavy rainfall and changed stream flow conditions influenced the abundance and species richness of exuviae collected on one sampling occasion. This incident occurred during the first 24-hour sampling event (12- 13 December, 2003) at Hat Hill Creek. Rain began to fall in the early morning (0500 hours) and became heavier between 0700 hours and 0900 hours. Flow velocity, turbidity, creek height and discharge volume increased steadily from 0600 hours onwards. In parallel with these observed changes, the abundance of exuviae and species-richness increased, particularly between 0500 hours and 0900 hours (Figure 5.9, Table 5.1). The results for these hours contrasted with those recorded at the same times during dry weather and normal flow conditions (Table 5.1). Drought also influenced exuviae results. The rainfall recorded in the study area during the austral summer was lower than average (Figure 5.4), and was particularly dry during the late February 2004 24-hour sampling event. Heavy rain fell later in the month, after samples were taken on February 19. The height of Hat Hill Creek was much lower than on previous sampling occasions and it was estimated that the stream flow had fallen by more than 50 % after a period of particularly low incidence of rainfall of approximately 6 weeks (Figure 5.4). Exuviae abundance and species richness during this period (19 February 2004, Table 5.1) was much lower than was recorded during the previous three sampling occasions (Table 5.1). A similar drought phenomenon was recorded in a tropical Australian pollution study (Crantson et al., 1995) where low stream flows on one sampling occasion resulted in greatly reduced exuviae abundance and species richness.

The relationship between emergence of adult chironomids and environmental factors has been the subject of much study world-wide (e.g. Learner et al., 1990; Armitage, 1995). Most species exhibit a distinct diurnal, crepuscular or nocturnal emergence pattern (Learner et al., 1990). The most important environmental cues for emergence appear to be levels of light intensity and water temperature (Learner et al., 1990; Armitage, 1995). No

139 data in the literature was discovered that suggested a link between the environmental factors of rainfall, or variation in stream flow patterns and resulting changes in chironomid emergence, however, these factors clearly influenced capture patterns during this study.

The flow-related changes in exuviae numbers observed could be explained by variations in stream energy. Higher flow regimes have greater energy and result in entrapped exuviae being removed from stream bank vegetation and other locations. In low-flow conditions, exuviae are more easily removed from the flow, due to the weaker energy of the flow being less able to resist entrapment or adherence along stream bank objects. Similarly exuviae may more readily sink in low energy flow regimes, whereas in higher energy conditions they may remain suspended and/or submerged exuviae may be re-suspended. Wilson and Bright (1973) supported this explanation after they conducted an experiment to simulate the transport of exuviae in a river. They released a large numbers of polystyrene balls (size of 2-4 mm) and found that most balls were recovered in a clumped distribution within 400 metres of the release point in winter and 150 metres in summer. They observed that a large number of balls were trapped in macrophytes along the edge of the waterway, particularly in summer when macrophytes were more abundant.

The experiment by Wilson and Bright (1973) used polystyrene balls as an easily manipulated surrogate for exuviae. They have remarkably different physical characteristics from exuviae and probably understate the entrapment of exuviae in streambank vegetation and other objects. This is because exuviae have a large and highly flexible surface area with a very small mass. Their movement on the surface of the water is strongly influenced by both water flow and wind (Cranston et al., 1997). When they encounter a solid object, they have a tendency to adhere, wrap around sticks and leaves and are thus largely immobilised. Perhaps this is less likely in larger, wider rivers, but certainly is the case in small, narrow creeks with thickly overhanging vegetation, such as Hat Hill and Victoria Creeks.

I am well aware of the tendency for exuviae to adhere to any solid object that it comes into contact with through my laboratory and field manipulation

140 of thousands of exuviae. I have observed that the safest way to manipulate them is to lightly touch individual exuviae with forceps or a probe. Exuviae are very fragile and are easily damaged. I avoided holding them by forceps in a pincer movement. Exuviae usually gently adhere to the forceps or probe, the bond of which is easily broken when it is placed below the surface of liquid.

These data indicated that when sampling chironomid exuviae from flowing waters, climatic conditions should be considered, particularly during conditions, such as extreme rainfall or drought, that may alter the typical flow regime. The response of exuviae to flow regimes is probably more apparent for small streams that are highly responsive to rainfall, such as those in the uplands sampled in this study. Collecting exuviae from larger lowland rivers, such as the exuviae survey of the Thames River, U.K., by Ruse and Wilson (1994), is probably less susceptible to such major variations. Depending upon the objectives of the study, the variation in numbers of chironomid exuviae collected during, or immediately after, heavy rain, or during abnormally low flow conditions, should be considered.

141 Chapter 6 Chironomid exuviae survey of the Grose River, a zinc and sewage polluted upland-river.

6.1 Introduction

Chironomids are often the most species rich and abundant taxonomic group of macroinvertebrates living within freshwater ecosystems (Coffman & Ferrington 1984; Cranston, 1995a). It has been established that individual chironomid species have a wide range of responses to water pollution (Johnson et al., 1993). Consequently, surveys of freshwater chironomids have been demonstrated to be an effective biological monitoring methodology for assessment of water pollution, particularly in the northern hemisphere (eg. Sæther, 1979; Johnson, 1994; Lindegaard, 1995).

There is less evidence that Australian chironomid species have similar differential sensitivities to water pollution and thus biological monitoring using chironomid surveys may not be as effective in Australian streams as they are in the northern hemisphere. The single published Australian study that used chironomids exclusively in a water pollution biomonitoring survey (Cranston et al., 1997) was undertaken in tropical northern Australia. They found that chironomid species assemblages did not exhibit an adverse response to heavy metal pollution.

The majority of Australian macroinvertebrate studies involving water pollution, that have collected species-level chironomid data, produced evidence that chironomid species were sensitive to pollution (Arthington et al., 1982; Hardwick, et al., 1995; Ward et al., 1995; Wright, 1994; Wright et al.,1995). All of these studies involved organic pollution, except for Ward et al. (1995) who examined pollution by an organophosphate pesticide. Conversely, the four macroinvertebrate studies that included species-level chironomid data, of heavy metal drainage from derelict mines (Norris et al., 1982; Norris, 1986; Cranston et al., 1997; Smith & Cranston, 1995) all found that chironomid species were tolerant of metal pollution.

In Australia, most pollution studies of freshwater macroinvertebrates use methodologies such as AUSRIVAS and SIGNAL and identify chironomids

142 only to the family or sub-family level (Anon, 1994; Chessman, 1995; Schofield & Davies, 1996, Sloane & Norris, 2003; Chapter 4). Chironomids have generally developed a reputation as a pollution tolerant group (Chessman, 1995). However, at the sub-family level, it has been observed that some chironomids have different levels of organic pollution sensitivity (Chessman, 2003) and metal pollution (Bruce Chessman, unpublished data, in Sloane & Norris, 2003).

Collection of pupal exuviae has been found to be an effective technique for assessing chironomid species living within waterbodies (Thienemann, 1910; Humphries, 1938; Brundin, 1966; Franquet & Pont, 1996; Cranston 2000b, Wright & Cranston, 2000; Chapter 2, 3 and 5). Exuviae have also been demonstrated to be an effective basis for conducting chironomid water pollution surveys (Wilson & Bright, 1973; Ruse & Wilson, 1994) and their use in Australian studies is gradually gaining recognition (Hardwick et al., 1995; Cranston et al., 1997; Wright & Cranston, 2000; Dimitriadis & Cranston, 2001). Only one Australian pollution survey has exclusively relied on chironomid exuviae and this was from tropical northern Australia waterways (Craston et al., 1997). There is a lack of data for elsewhere in Australia.

In this chapter the value of conducting chironomid species pollution surveys was assessed. Two types of pollution were investigated: treated sewage and zinc-rich mine leachate, and a comparison of species and family level chironomid response (Chapter 4) to these pollution sources was also investigated.

143 6.2 Materials and Methods

6.2.1 Study area

Field work was carried out in the upper Grose River catchment in the Blue

Mountains NSW, south-eastern Australia (33 35’S, 150 15’E). The location of the study area is presented in Figure 6.1 and sampling sites are described in Chapter 4.

Sampling sites were chosen based on the results presented in Chapter 4. Macroinvertebrate analysis had demonstrated that downstream of the treated sewage outflow and of the zinc-enriched mine drainage site, both the macroinvertebrate communities and the water quality were strongly impaired (Chapter 4). Sites impacted in this way, together with unpolluted areas, were used in the investigation. In total five sites were sampled: two on the Upper Grose River (GDD, GBK) to measure the effects of the zinc polluted discharge and three on its tributaries. One site on Hat Hill Creek (HHD) was sewage affected from the Blackheath STP. The study included two reference sites that were not exposed to waste discharges, Hat Hill Creek above the STP discharge (HHU) and Victoria Creek VIC (Figure 6.1).

The two reference sites were slightly acid to neutral, dilute, poorly buffered and low in nutrients (Section 4.3.3). Mine polluted drainage sites were downstream of an abandoned coal mine discharged via a tributary (Dalpura Creek) into Grose River (Figure 6.1). This drainage was rich in heavy metals, particularly zinc, and also had elevated electrical conductivity (Figures 4.8 and 4.14). Zinc in the Grose River at this point was well above ANZECC (2000) guideline levels for ecosystem protection (Figure 4.14). Blackheath STP discharged secondary treated effluent from a medium-sized township to Hat Hill Creek, which flowed approximately five kilometres down a steep escarpment into the Grose River (Figure 6.1). This discharge caused very high levels of phosphorus and nitrogen in Hat Hill Creek and the Grose River that were also well above ANZECC (2000) guideline levels for ecosystem protection (Figures 4.12, 4.13).

144 Figure 6.1 Map of exuviae survey sites, waterways and waste discharge points in the upper Grose River system. Approximate catchment boundary is dashed line. Inset shows location of study area in south-eastern Australia. Sampling sites are indicated by square symbols. Sites on the Grose River below coal mine: GDD (Grose River below Dalpura) and GBK (Grose River at Burra Korain). Reference sites: VIC (Victoria Ck) and HHU (Hat Hill Ck). Hat Hill Creek below STP discharge: HHD.

145 6.2.2 Chironomid exuviae collection To sample chironomid pupal exuviae, a steel-framed drift 250 micron net with an entry aperture of 30 cm2 and a length of 60 cm was placed across a narrow section of the streams. The net was placed in the best position available to capture the majority of stream flow at each of the sites. At HHU the net was placed in a very narrow location where it intercepted the whole stream flow, apart from some minor leakage. At HHD about 60 % of the stream flow was captured and VIC captured about 75 % of creek flow. The Grose River was a larger waterway and the nets on the two Grose River sites, GDD and GBK, captured about 40 % of the river flow. When only a sub-section of the flow could be captured, nets were placed in the fastest flowing section of the channel.

In Chapter 5, the spatial and temporal variation of drifting chironomid exuviae was investigated at two upland streams in the study area. It was observed that a single 24-hour drift netting of a waterway gave a reasonable species estimate, although repeated 24-hour sampling provided more complete species inventories (Figure 5.13, 5.14). It was also discovered that most species emerged at night, generating higher chironomid species richness and abundance of exuviae in night hours. It was considered that five 24-hour periods would give a thorough estimate of the chironomid species and their relative abundance at each site. This involved a total of 120 hours of netting at each site. Nets were left in place for a minimum of 24-hours, however, because of the poor quality (or lack) of walking tracks, very steep gradients and other site hazards, nets were left in place for two 48-hour periods at GDD, GBK and VIC (Table 6.1).

Rainfall and flow conditions were also observed to influence capture rates (Table 5.1). Increased numbers of exuviae were collected during periods of heavy rainfall in parallel with increased creek flow. Conversely, in dry weather, the stream flow was lower and reduced numbers of exuviae were collected (Chapter 5). Based on these observations, extremes of flow were avoided.

146 Upon removal from the water, the net contents were flushed from the net into a plastic collecting tray and immediately bottled and preserved in 70 % ethanol.

In the laboratory the net contents were sorted from detritus (leaves, algae, non-chironomid exuviae invertebrates) with the aid of a dissecting microscope (X 40). At one site (HHD) there was a significant abundance of very fine detritus (<1 mm particle size). The fine fraction was sieved from the coarser fraction of the sample (i.e. >1 mm) and water was added to the fine fraction, to bring the sample to 1000 mL. This was then placed in a sealed conical large funnel in a retort stand. Compressed air was carefully released from the lowest point of the funnel, in order to thoroughly homogenise the sample. After this process, 10 x 20 mL samples were extracted, each comprised 2 % of the fine fraction of the sample. Exuviae from each of these sub-samples were processed and identified and the results summed and multiplied by five to create an estimate of the exuviae from the full fine fraction. All of the coarse fraction was processed.

All exuviae were slide-mounted in temporary mountant (Figure 6.10, 6.11) and were identified using a high resolution compound microscope (up to X 400) using keys to chironomid exuviae (Cranston, 1996, 2000a).

Table 6.1 Chironomid exuviae sampling sites, number and duration of drift samples collected in this study in the Grose River system. December 2003 to March 2004.

Sites Label Number and duration of drift samples

Hat Hill upstream of STP HHU 5 X 24-hours (120 hours)

Hat Hill downstream of STP HHD 5 X 24-hours (120 hours)

Victoria Creek VIC 1 X 24-hours, 2 X 48 hours (120 hours)

Grose River downstream of Dalpura GDD 1 X 24-hours, 2 X 48 hours (120 hours) Creek

Grose River at Burra Korrain GBK 1 X 24-hours, 2 X 48 hours (120 hours)

The study was undertaken in the austral summer and early autumn, from

December 2003 to March 2004. Weather conditions were generally warm to

hot during the day (18 – 32 C), and cool to mild (9 – 16 C) at night. Water temperatures ranged from 15 to 23 C (Chapter 5). Rainfall was below the

147 long-term average, with five of the seven months prior to and during the study recording less than the average monthly rainfall (Figure 6.2).

160

140

120

100

80

mm/month 60

40

20

0 September October November December January February March 2003 - 2004

Figure 6.2 Total monthly rainfall (mm) recorded prior to, and during, this study at Blackheath (black bar). Historic mean monthly rainfall indicated by unshaded bars. Source: www.bom.gov.au accessed May 2004.

6.2.3 Data analysis procedures To ensure a closer approximation to a normal distribution, chironomid species richness and abundance data were logarithmically transformed (X +1). Analysis of variance (ANOVA) was used to determine if these two parameters were affected by site location.

Several authors (Norris et al., 1982; Marchant et al., 1994; Wright, 1994; Wright et al., 1995) have demonstrated that multivariate analysis is a highly effective method for evaluating freshwater pollution patterns in macroinvertebrate studies. Non-metric multidimensional scaling (NMDS) was performed on the similarity matrix which had been computed with square-root transformed chironomid abundance data using the Bray-Curtis dissimilarity measure (Clarke 1993; Warwick 1993). Square-root transformation was selected in order to moderate the influence of highly abundant species in the analysis (Wright, 1994; Wright et al., 1995). Two- dimensional ordinational plots represented the dissimilarity among samples. The two reference sites (VIC and HHU) were grouped to test differences by

148 one way analysis of similarity values (ANOSIM, see Clarke, 1993) in ordinations between reference sites (HHU, VIC) and ‘test’ sites (GDD, GBK, HHD). The influence of particular species in creating the differences in the ordinations between the reference sites, and the test sites were quantified using the similarity percentage procedure (SIMPER). The above multivariate analyses were undertaken using the software package PRIMER version 5 (Clarke, 1993).

149 6.3 Results

6.3.1 Chironomid exuviae abundance A total of 5058 chironomid exuviae from 80 species were collected, examined and identified, often as morphospecies (Figure 6.3, Table 6.2. Table 6.4, Table 6.6).

2000

1500

1000

500

0 HHU HHD VIC GDD GBK

Figure 6.3 Total number of chironomid exuviae collected from each site in the Grose River survey after 120 hours of netting. Cross-hatching represents the estimated proportion of exuviae in the fine fraction of the HHD sample.

6.3.2 Comparison of samples taken above and below an STP outflow Most exuviae (3627, 44 species) were collected from Hat Hill Creek (Figure 6.3, Figure 6.4, Table 6.2). Chironomid species richness varied highly significantly among sites (Table 6.3). Upstream of Blackheath STP discharge point, 1745 exuviae (41 species) were recorded and 1882 exuviae from ten species of chironomid were recorded in Hat Hill Creek immediately below the Blackheath STP discharge point (Figure 6.3 and 6.4, Table 6.4). Exuviae abundance did not vary significantly among sites (Table 6.3).

150 Table 6.2 List of all chironomid species recorded at each Grose River survey site (X = species found at site) in this exuviae study, December 2003 to February 2004.

VIC HHU GDD GBK HHD Sub-family Species Chironominae Tanytarsus sp. 1 X Tanytarsus sp. 2 (nr. manlyensis) X Tanytarsus sp. 3 (multidir spines) X Tanytarsus sp. 4 (nr. manlyensis) X X X X Tanytarsus sp. 5 (nr. spinosus) X Tanytarsus liepae X Tanytarsus sp. 6 (nr. liepae) X Tanytarsus nr. dycei X Tanytarsus nr.bispinosus X Tanytarsus bispinosus X Tanytarsus nr. B3 X X X Cladotanytarsus sp. X Paratanytarsus nr. furvus X Paratanytarsus kathleenae X X Paratanytarsus jeffereyi X Rheotanytarsus christinae? X X Rheotanytarsus flabellatus? X Rheotanytarsus jeffereyi X Rheotanytarsus juliae? X Stempellina ?australiensis X X Riethia stictoptera X Riethia zeylandica X X Riethia nr. ‘V4’ X Riethia ‘divided row’ X Polypedilum nr.'allocasia' X X X X Polypedilum nr. ‘S1’ X Polypedilum sp. 1 X Polypedilum oresitrophus X Polypedilum sp. 2 X Demicryptochironomus sp. X Harrisius sp. X Zavriella sp. 1 X X Zavriella sp. 2 X ? Dicrotendipes sp. X Unidentified chironominae sp.1 X Unidentified chironominae sp.2 X

Orthocladiinae Cardicladius australiensis X Thiennemanniella sp. 1 X X X X X Thiennemanniella sp.2 X Parametriocnmeus sp. X X Nanocladius sp. X X Eukiefferiella insolita X X X X X Rheocricotpous sp. X MO5 X X X X Stictocladius sp. X X Parakiefferiella nr. ‘variegatus’ X X X X X

151

(Table 6.2 continued) VIC HHU GDD GBK HHD Parakiefferiella sp. 1 X Cricotpus nr. acornis X X X X X Cricotpus sp.2 X X SO1 X SO4 X X Botryocladius grapeth X X Botryocladius collessi X Genus Australia sp. X Paralimnophyes sp. X Parametriocnmeus sp. X Coryneura sp. X Nr. ‘SO1’ sp. X Unidentified Orthocladinae sp. 1 X Unidentified Orthocladinae sp. 2 X Unidentified Orthocladinae sp. 3 X Unidentified Orthocladinae sp. 4 Unidentified Orthocladinae sp. 5 X Unidentified Orthocladinae sp. 6 X

Tanypodinae Pentaneurini genus A sp. X X X X X Pentaneurini genus B sp. X Pentaneurini genus C sp. X X Pentaneurini genus D sp. X X X Pentaneurini genus E sp. X Pentaneura sp. nov. X X Paramerina parva X X X X X Apectrotanypus maculatus X Ablabesmyia sp. X Larsia sp. X Procladius paludicola X X Procladius nr. paludicola X Nilotanypus sp. X X

Podonominae Podochlus australiensis X Podomonopsis ? sp. X Podomonopsis discoceros X

152

35

30

25

20

15

10

5

0 T1 T2 T3 T4 T5 T1 T2 T3 T4 T5

HHU HHU HHU HHU HHU HHD HHD HHD HHD HHD

Figure 6.4 Chironomid species richness at Hat Hill creek upstream of STP (Black bar) and Hat Hill creek downstream of STP (unshaded bar) on each of five sampling occasions (T1 to T5), December 2003 to March 2004.

Table 6.3 F-statistics and associated probabilities from analyses of variance of number of species of exuviae (log X + 1 transformed) and number of exuviae (log X + 1 transformed) collected from five sampling occasions at Hat Hill Creek (upstream STP) and Hat Hill Creek (downstream STP).

Source of Variable variation Number of species Number of exuviae Df MS F P-value Df MS F P-value

Site 1 0.74 91.32 <0.001 1 0.06 0.55 NS

Error 8 0.01 8 0.11

153 Table 6.4 Hat Hill Creek Chironomidae species abundance results recorded from each of the five sampling occasions (T1 to T5) from the exuviae survey December 2003 to March 2004.

T1 T2 T3 T4 T5 T1 T2 T3 T4 T5 Sub-family Taxon Hat Hill upstream Hat Hill downstream

Chironominae Tanytarsus sp.1 4 ------Tanytarsus sp.3 ( multidir spines) 1 1 1 2 8 - - - - - Tanytarsus sp. 4 (nr. manlyensis) 3 1 2 21 5 - - - - - Tanytarsus liepae 18 3 11 10 28 - - - - - Tanytarsus sp. 5 (nr. spinosus) 66 39 41 16 16 - - - - - Tanytarsus sp. 6 (nr. liepae) - - 2 ------Cladotanytarsus sp. - - 2 ------Paratanytarsus nr. furvus 139 84 71 13 59 - - - - - Rheotanytarsus christinae? 33 8 12 12 15 - - - - - Rheotanytarsus juliae? ------16 - 2 Stempellina ?australiensis 9 2 7 13 ------Riethia nr. ‘V4’ 3 1 3 ------Riethia zeylandica 7 - - 2 5 - - - - - Riethia ‘divided row’ - - - - 1 - - - - - Polypedilum nr.'allocasia' 3 1 4 ------Polypedilum nr. ‘S1’ - - - - 3 - - - - - Harrisius sp. 1 ------Zavriella sp.2 - - - 3 1 - - - - - Unidentified Chironominae sp.2 3 - 1 ------

Orthocladiinae Cardicladius australiensis - - - - - 132 20 49 16 8 Thiennemanniella sp.1 4 2 1 11 2 853 169 150 73 106 Parametriocnmeus sp. - - - - 1 - - - - - Nanocladius sp. 17 2 9 1 5 - - - - - Eukiefferiella insolita 57 35 39 - 1 324 11 22 7 6 MO5 119 32 44 30 104 - - - - - Stictocladius sp. - - 3 ------Parakiefferiella nr. ‘variegatus’ 50 68 16 9 28 - - - - 10 Cricotpus nr. Acornis 8 14 - 2 4 11 - 1 - - Cricotpus sp.2 8 - 3 ------SO1 - 1 ------SO4 38 16 24 - 29 - - - - - Botryocladius grapeth 2 1 1 - 3 70 1 7 1 - Botryocladius collessi 1 ------Genus Australia 1 - - - 1 - - - - - Paralimnophyes sp.1 - 1 ------Coryneura sp. - - - - 3 - - - - - Unidentified Orthocladiinae sp.2 1 ------Unidentified Orthocladiinae sp. 6 - - - - - 5 - - -

Tanypodinae Pentaneurini genus A sp. 7 3 4 - 4 - - - - 5 Pentaneura sp. nov 4 8 11 - 4 - - - - - Paramerina parva 1 2 - 1 5 3 8 9 12 5 Pentaneurini genus D sp. 1 2 ------Apectrotanypus maculatus 2 - 1 - 4 - - - - -

Podonominae Podochlus australiensis 3 5 - - 1 - - - - -

Total number of species 31 24 24 15 26 6 5 7 5 7 Total number of exuviae 614 332 313 146 340 1261 189 205 93 134 Number of species/ site 41 10

154 6.3.3 Comparison of samples taken above and below a zinc-rich drainage outflow A total of 1431 chironomid exuviae from 48 species were collected from three sites: Victoria Creek, a reference site unaffected by mine drainage, and two sites on the Grose River below the mine drainage point source (Figure 6.5, Table 6.6).

Chironomid species richness was significantly different between the zinc- polluted sites and the reference site (Table 6.5). A total of 44 chironomid species were recorded in Victoria Creek (average 29.7 species/occasion) and 12 (average 9.6 species/occasion) were collected in the Grose River downstream of Dalpura Creek (Figure 6.5). Downstream of this site (Burra Korrain, Grose River) 18 species were recorded (average 11.3 species/occasion) (Figure 6.5).

Abundance did not vary significantly among sites (Table 6.5). A total of 781 exuviae were collected from Victoria Creek, while at the two Grose River sites substantially fewer (GDD, 264, GBK, 386) were recorded (Table 6.6).

35

30

25

20

15

10

5

0 T1 T2 T3 T1 T2 T3 T1 T2 T3 VIC VIC VIC GDD GDD GDD GBK GBK GBK

Figure 6.5 Chironomid species richness at Victoria Creek (reference site is shaded black) and Grose River below Dalpura Creek (GDD = zinc polluted is shaded grey) and Grose River at Burra Korain (GBK = zinc polluted is unshaded) on each of three sampling occasions, December 2003 to February 2004.

155 Table 6.5 F-statistics and associated probabilities from analyses of variance of number of species of exuviae (log X + 1 transformed) and number of exuviae (log X + 1 transformed), from three sampling occasions at Victoria Creek (reference site) and at the two Grose River sites immediately below Dalpura Creek (mine leachate affected). December 2003 to March 2004.

Source of Variable variation Number of species Number of exuviae Df MS F P-value Df MS F P-value

Site 2 0.25 14.94 <0.0046 2 0.16 1.44 NS

Error 6 0.02 6 0.11

156 Table 6.6 Abundance of chironomid species recorded at the reference site (Victoria Creek) and two sites on the Grose River (both subject to zinc- rich mine drainage) below Dalpura Creek, for each of the three sampling occasions (T1, T2 and T3: December 2003 to February 2004).

VIC GDD GBK Sub-family Taxon T1 T2 T3 T1 T2 T3 T1 T2 T3 Chironominae Tanytarsus sp. 2 (nr. manlyensis) 11 49 4 ------Tanytarsus sp. 4 (nr. manlyensis) 5 2 4 1 - - 3 - - Tanytarsus bispinosus - 4 1 ------Tanytarsus nr. bispinosus 97 2 7 ------Tanytarsus nr. B3 2 14 5 1 1 7 29 4 9 Tanytarsus nr. dycei - - - 1 - - - - - Paratanytarsus kathleenae - 5 - - - - 5 - 2 Paratanytarsus jeffereyi - 1 ------Rheotanytarsus christinae? 3 ------Rheotanytarsus flabellatus? 11 18 1 ------Rheotanytarsus jeffereyi 1 ------Stempellina ?australiensis 3 8 1 ------Riethia stictoptera 3 3 ------Riethia zeylandica 126 29 2 ------Polypedilum nr. 'allocasia' 1 1 - 9 3 3 50 5 15 Polypedilum sp.1 4 3 2 ------Polypedilum sp.2 1 - 1 ------Polypedilum oresitrophus - 1 ------Demicryptochironomus - 1 ------

Orthocladiinae Nanocladius - 1 1 ------Eukiefferiella insolita 5 1 - 22 2 4 - 5 - Rheocricotpous 3 ------Stictocladius - 23 3 ------Parakiefferiella "variegatus" 6 41 3 2 - - 6 - - Cricotpus acornis 19 10 1 2 3 13 13 4 4 Cricotpus 'Hat Hill' 1 - 1 ------Thiennemanniella sp.1 20 16 17 9 - - 20 - 2 Thiennemanniella sp.2 ------1 - - Parametriocnmeus sp. 5 6 1 - - - 3 12 4 MO5 37 15 3 99 28 43 106 24 49 SO4 - 2 2 ------Nr. SO4 sp. - 1 ------Zavriella 6 - 3 - - - 1 - - Unidentified Orthocladiinae sp.1 1 ------

Tanypodinae Pentaneurini genus A sp. 2 5 7 1 - - - - 1 Pentaneurini genus B sp. ------1 Pentaneurini genus C sp. - - - 1 - - - - - Pentaneurini genus D sp. - 3 1 - - - 1 - - Pentaneurini genus E sp. - 2 ------Pentaneura sp nov 6 10 3 ------Paramerina parva 12 6 6 3 1 5 1 2 - Ablabesmyia sp. 1 13 ------Larsia sp. 2 ------Procladius paludicola - 4 - - - - 1 - - Nilotanypus sp. - 2 - - - - 2 - 1 Procladius nr. paludicola 2 - 1 ------

Podonominae Podomonopsis discoceros 1 ------Podomonopsis ? 1 ------

Number of exuviae / time 398 302 81 151 38 75 242 56 88

Number of species / time 31 33 25 12 6 6 15 7 10

Number of species / site 44 12 18

157 6.3.4 Multivariate analysis of chironomid exuviae data Using NMDS ordination, it was evident that chironomid assemblages from the sewage polluted site on Hat Hill Creek (HHD = orange symbol) and the zinc-polluted sites on the Grose River (GDD = green symbol, GBK = black symbol) clustered separately from the two pristine reference sites on Victoria Creek (VIC = red symbol) and Hat Hill Creek (HHU = blue symbol) (Figure 6.6). The two zinc-polluted sites (GDD, GBK) formed a separate cluster, and the STP affected site (HHD) was well separated from all other sites. The NMDS stress value of 0.12 indicated that, in two dimensions, the ordinations were ‘fair’ estimates of the original data (Clarke 1993).

ANOSIM confirmed that the differences between all pairs of reference sites (grouped) and test sites (HHD, GDD, GBK) were statistically significant (Table 6.7). The ANOSIM Global R value of 0.848 indicated that the sites were well separated and a significance level of of 0.1 % indicated that the difference was highly significant. The two zinc-polluted sites on the Grose River (GDD, GBK) were not statistically different from other sites.

v

Figure 6.6 NMDS ordination of chironomid exuviae species data collected from the Grose River and tributaries, December 2003 to February 2004. Each sample is represented by a triangle. Reference sites are indicated by the blue (site HHU) and red (site VIC) colours. Test sites are indicated by the colours: orange (sewage receiving = HHD), green (mine drainage = GDD) and black (mine drainage = GBK).

158 The 10 chironomid species that contributed most (combined contribution to 90 % of dissimilarity) to the separation between the communities at reference sites and the sewage polluted site (HHD) are listed in Table 6.8. Three of these species (Thiennemanniella sp. 1, Eukiefferiella insolita, Cardiocladius australiensis) had higher abundance at the sewage polluted site than at the reference sites. In contrast, seven species (MO5, Paratanytarsus nr. furvus, Parakiefferiella nr. variegatus, Riethia zeylandica, Tanytarsus sp. 2 (nr. manlyensis), Stempellina ? australiensis and Cricotpus nr. acornis) had higher abundances at the reference sites than at the sewage-polluted site.

Ten chironomid species also dominated the difference between communities at reference sites and the zinc polluted site (GDD). They are listed in Table 6.9. One species (MO5) had higher abundances at the zinc- polluted site than at the reference sites. Conversely, eight species (Paratanytarsus nr. furvus, Eukiefferiella insolita, Tanytarsus sp. 5 (nr. spinosus), Tanytarsus sp. 5 (nr. manlyensis), Thiennemanniella sp.1, Stempellina ? australiensis, Cricotopus nr. acornis and Tanytarsus sp. 2 (nr. manlyensis)) had higher abundances at the reference sites than at the zinc- polluted site.

According to the methodology developed in Chapter 4 (section 4.2.5), pollution affinities were calculated according to abundance of key chironomid species at reference sites and polluted sites (Table 6.10, 6.11).

Table 6.7 R-statistics (Clarke 1993) from two-way crossed ANOSIM for pairwise comparison of all Grose system sites for square-root transformed chironomid species data. chironomid species data from both reference sites were combined. (NS, P>0.05; * 0.01

Comparison R-statistic Significance level

Reference vs HHD 0.995 *** Reference vs GDD 0.801 ** Reference vs GBK 0.793 ** HHD vs GDD 1 * HHD vs GBK 1 * GDD vs GBK 0.296 NS

159 Table 6.8 The 10 most influential species contributing to the difference between the chironomid assemblages at the reference sites (combined) compared to Hat Hill Creek, downstream of Blackheath STP, according to SIMPER breakdown.

Taxon Reference Hat Hill ds Contribution Cumulative sites STP (%) (%)

Thiennemanniella sp.1 9.45 270.20 34.77 34.77 MO5 39.18 0 7.07 41.84 Eukiefferiella insolita 13.0 74.0 7.04 48.89 Cardiocladius australiensis 0 45.0 6.39 55.27 Paratanytarsus nr.furvus 33.27 0 5.44 60.71 Parakiefferiella variegatus 22.91 2 4.25 64.96 Riethia zeylandica 16.45 0 2.88 67.84 Tanytarsus sp. 2( manlyensis) 16.55 0 2.87 70.71 Stempellina ? australiensis 8.64 0 2.40 73.11 Cricotopus acornis 8.73 0 2.04 75.15

Table 6.9 The 10 most influential species contributing to the difference between the chironomid assemblages at the reference sites (combined) compared to the Grose River, downstream Dalpura Creek (mine leachate), according to SIMPER breakdown.

Taxon Reference Grose Contribution Cumulative sites ds (%) (%) Dalpura

MO5 39.18 56.67 15.41 15.41 Paratanytarsus nr.furvus 33.27 0 8.90 24.31 Eukiefferiella insolita 22.91 0.67 7.58 31.89 Riethia zeylandica 13 9.33 4.86 36.75 Tanytarsus sp. 5 (nr. spinosus) 16.45 0 4.84 41.59 Tanytarsus sp. 2 (nr. manlyensis) 16.55 0 4.81 46.40 Thiennemanniella sp. 1 8.64 0.33 4.61 51.01 Stempellina ? australiensis 9.45 3 3.96 54.97 Cricotopus nr. acornis 8.73 0 3.75 58.72 Tanytarsus sp. 4 (nr. manlyensis) 9.73 6 3.16 61.88

Table 6.10 Zinc and sewage pollution affinity of pollution indicator chironomid species, according to the Grose River survey (see Table 6.12 for grades). ND = not detected.

Taxon STP Zinc

Thiennemanniella sp.1 XX MO5 XXX Eukiefferiella insolita XX Cardiocladius australiensis ND Paratanytarsus nr.furvus XXX XXX Parakiefferiella variegatus XX ND Riethia zeylandica XXX - Tanytarsus sp. 2 (nr. manlyensis) XXX XXX Stempellina ? australiensis XXX X Cricotopus nr. acornis XXX XXX Tanytarsus sp. 4 (nr. manlyensis) ND -

160 Table 6.11 Pollution affinity calculations for taxonomic groups.

Abundance at polluted site relative Pollution affinity ranking to average at reference sites 200 % to 500 % Mildly positive 500 % to 1000 % Positive > 1000 % Highly positive

50 % to 200 % - Neutral ✁

20 % to 50 % Mildly negative ✁✁

1 % to 20 % Negative ✁✁✁ < 1 % (or absent) Highly negative

161 Figure 6.7. Plates 1 to 4. Exuviae from Grose River pollution study. Plate 1 is a lateral-dorsal view of abdominal segments of Eukiefferiella insolita. Plate 2 is a dorsal view of abdominal segments of Nanocladius sp. Plate 3 shows thoracic horn structure of Paramerina sp. Plate 4 shows thoracic horn structure of Pentaneura sp.. Plate 2. Plate 1.

Plate 4. Plate 3. 162 Figure 6.8. Plates 5 to 8. Exuviae from Grose River pollution study. Plate 5 is a dorsal view of the lowest abdominal segments of Orthocladiinae species ‘SO4’. Plate 6 is a dorsal view Thiennemaniella sp. Plate 7 is a dorsal view of abdominal segments II, III, IV and V of Stempellina ? australliensis. Plate 8 is a dorsal view of the lowest abdominal segments and postero-lateral combs of Stempellina ? australliensis. Plate 6. Plate 5.

Plate 8. Plate 7. 163 6.4 Discussion

This study contributes to the growing international evidence that chironomid species are sensitive and robust indicators of freshwater pollution. This study provides the most convincing evidence yet recorded that they respond strongly to water pollution in Australian waterways. Zinc and organic pollution, both strongly influenced chironomid species assemblages in upper Grose River waterways. The chironomid communities downstream of these point source pollution outflows had greatly reduced species richness compared to reference sites. Most chironomid species were observed to be intolerant of pollution, with relatively few species tolerating the polluted conditions.

The structure of the chironomid community at the sewage polluted site (HHD) and at the two zinc polluted sites (GDD, GBK) was very different from the communities found at the reference sites (NMDS: Figure 6.6; ANOSIM: Table 6.8). Fewer chironomid species were recorded downstream of these point sources than upstream. Eleven of the most abundant species exhibited a strong response, negative and/or positive, to water pollution (Table 6.9 - 6.11). These responses varied from greatly increased abundance to complete absence (Table 6.11). Six species had higher abundance at one of the polluted sites than at reference sites, three to zinc pollution and three to STP effluent (Table 6.11). No species responded positively to both pollution types while six species were abundant at reference sites but completely absent at polluted sites (Table 6.9 - 6.11). Based on these data, the species listed in Table 6.10; Thiennemanniella sp. 1, MO5, Eukiefferiella insolita, Cardiocladius australiensis, Paratanytarsus nr.furvus, Parakiefferiella variegates, Riethia zeylandica, Tanytarsus sp. 2 (nr. manlyensis), Stempellina ? australiensis, Cricotopus nr. acornis, Tanytarsus sp. 4 (nr. manlyensis) are considered to be candidates as pollution indicators.

The great majority of the chironomid species recorded were only found at one of the two pristine reference sites (Table 6.2). While the abundance of many species was inadequate to determine their pollution tolerance, it is expected that further examination would reveal that many of these species are appropriate as clean water indicator species. The species level results

164 from this chapter are very different from the family level response reported in Chapter 5, which revealed that chironomids responded positively to organic pollution (increased abundance) and negatively to zinc pollution (reduced abundance). When the focus is at the species level, most chironomid species did not tolerate either types of water pollution. While very few species tolerated the polluted conditions some (e.g. Thiennemanniella sp. 1), were present in large numbers, particularly below the STP, indicating that they thrived in the presence of organic polluted waters in these low nutrient waterways.

This is one of very few Australian studies that found chironomid species were highly sensitive indicators of water pollution. Arthington et al. (1982) conducted one of Australia’s first detailed examinations of the chironomid species-response to water pollution and recorded the larvae of 42 chironomid species in a study of a sewage-polluted urban creek. They found that chironomid species-richness fell below the sewage outfall, at the same time that abundance increased. In their study four species of chironomid became highly abundant in the sewage-polluted section of the study river. They also summarised the sewage-effluent tolerance of 16 chironomid species into five groups. Another study found that abundances of 13, from a total of 19 chironomid larval species, were reduced by the application of an organophosphate chemical (Ward et al., 1995). That study also concluded that chironomids were more sensitive as bioindicators than non-chironomid biota and they strongly advocated species-level assessment of chironomids in biological studies of water pollution. In contrast to the investigations by Arthington et al. (1982) and Ward et al. (1995), this current study recorded the response of many species from the sub-family Orthocladiinae, which are known to prefer cooler waters (Rossaro, 1991; Chapter 3). These results add more evidence that chironomids are highly species-diverse and sensitive water pollution indicators in Australian waterways across a wider climatic range than previously recorded.

The results reported here showed a different response of flowing water chironomids to sewage impacts than did the study by Wilson and McGill (1977). They were one of the first researchers to use chironomid exuviae to investigate the impact of sewage in their studies, conducted in the River

165 Chew in Avon, United Kingdom. In contrast to this study, they recorded more species downstream (15 to 24 species) of the sewage discharge than upstream (11 to 17 species). They also reported a major change in the chironomid species assemblage below the sewage out-fall. They found that one particular species (Chironomus riparius) was dominant at the sites immediately downstream. Unlike the current study, where the reference sites were in a wilderness area that was effectively pristine, above the sewage discharge in the River Chew may have had alternative disturbance or pollution regimes that were not considered and pollution sensitive species may have been absent at their reference sites. Alternatively, the different suite of chironomids adapted to the different conditions of the northern- hemisphere environments may simply respond differently to the Australian species.

This current study has produced results that show that chironomid species respond in different ways to heavy-metal pollution compared to sewage- pollution. A similar project in the United Kingdom studied chironomid species from a stream system that also received zinc-rich mine drainage and sewage effluent (Armitage & Blackburn, 1985). Their mine drainage results were similar to this study, with only one to six chironomid species collected from their most zinc impacted sites compared to ten to 21 species at their unimpacted sites. They were unable to isolate and compare the impact from sewage effluent from the impacts of zinc, as sampling sites above and below the STP inflow were all zinc contaminated.

Results from the current study were also different from an earlier study I was involved with (Wright, 1994; Wright et al., 1995) that used chironomid larvae, together with other macroinvertebrate groups, to assess the impact of an STP on a different waterway, Blue Mountain Creek, near Wentworth Falls (Blue Mountains). In that study fewer species were recorded, with only 22 chironomid species (72 this study) found across three pristine sites and 12 species (10 this study) from the sewage polluted site. The greater number of species recorded in this study is almost certainly due to the greater accuracy of species identification achieved using exuviae. It is probably also due to the collection of exuviae representing species that inhabit a much broader range of larval habitats than direct larval sampling would normally target.

166 Results reported here were also different from an otherwise similar study of chironomids and mine drainage from tropical northern Australia. Cranston et al. (1997) also collected drifting chironomid exuviae and the mine drainage was contaminated by heavy metals. They recorded an increase in chironomid abundance and species richness below the heavy-metal rich acid mine drainage (<10 species upstream, >40 species downstream). However, the authors explained that the difference in results was probably due more to flow regimes rather than water quality. Results obtained in Chapter 5 supported this hypothesis. It was shown that flow regime played a major role in the abundance and diversity of exuviae collected. It is therefore not valid to compare the results of the two studies.

An Australian study on heavy-metal pollution of South Esk River, Tasmania, recorded 10 species (Norris et al., 1982). They found that the chironomid species were tolerant of metal pollution: between two and seven species were recorded at the eight sites they sampled. A second Australian study on heavy metal pollution of Molongolo River, NSW, identified four larval chironomid species (Norris, 1986). They also concluded that the chironomid species assemblages did not respond to the metal drainage (Norris, 1986). In both studies their chironomid species results may have been substantial underestimates of the actual species present due to a legacy of collecting chironomid larval and associated difficulties in distinguishing species.

As shown elsewhere in this thesis (Chapter 2), collection of larvae, rather than exuviae consistently under-estimate the numbers of species present. Conclusions drawn on such a restricted data-base may give erroneous results when some species respond to an impact positively and others negatively.

Internationally the current study is one of the first studies to investigate the response of chironomids to two very different pollution types (organic pollution and zinc-rich mine seepage) within an otherwise undisturbed and small (approximate 8000 hectare) watershed. Species responded to both pollution types differently, although no species had a positive response to both pollution types, many did exhibit a negative response to both pollution types. The species-level investigation of chironomid biota in Australian

167 waterways is supported by this study for the purposes of detecting and measuring water pollution. Their use as stream and river biological indicators of water pollution in Australian freshwaters has great promise.

168 Chapter 7 General discussion and conclusions The overall aim of this thesis was to investigate chironomid species living in Australian freshwater ecosystems, and explore geographical, physical and chemical factors that influence them. The first part of this thesis examined lake-dwelling chironomids in a single lake and consequently surveyed chironomids from a broad range of natural freshwater lakes in southern and eastern Australia. The second part of the thesis investigated chironomids, and other macroinvertebrates, in upland streams. It explored temporal variation of exuviae in upland streams and also pollution impacts on chironomid assemblages from heavy metal and sewage waste discharges.

7.1 Summary of results

This thesis revealed Australian lakes to be richer in chironomid species than was previously recognised (Chapter 2 and 3). The collection of chironomid exuviae, from lakes, provided the ‘key’ to more accurate determination of species (Chapter 2). A total of 134 species was identified (Table 3.10), in contrast to the approximately 55 species reported in numerous previous lake studies (Table 3.3). This finding contributes to a long running debate about the apparent low species richness (chironomids and other macrobenthos) of Australian lakes, particularly by Fulton (1983a, b) and Timms (many references, see Table 3.2). For chironomids, this finding challenges the previous conclusions that Australian lakes are species-poor in benthic biota (Timms, 1985).

A distinct biogeographical pattern was detected for chironomid species from the survey carried out from southern and eastern Australian freshwater lakes. In Australia, this is the first description of the pattern of lake-dwelling chironomid species that revealed species-assemblage variation along a north– south geographical gradient (Chapter 3). This segment of the research indicated that the geographical location of the lakes was more important to the distribution of chironomid species than was the type of lake. Only five of the 134 chironomid species recorded from lakes in this study were found in all of the four broad geographic lake regions of this investigation; Tasmania, south-east Australian mainland, Fraser Island and

169 tropical north Queensland. These five species were considered to be cosmopolitan. In contrast, the majority (n=75) of lake-dwelling chironomid species were found only from a single lake within one of these four geographic lake regions.

Chironomid sub-family affiliations were also evident within the distinct north- south biogeographic pattern in species assemblages (Chapter 3). For example, more species of the sub-family Orthocladiinae were recorded in the two most southern lake biogeographic regions, Tasmania (14 species) and southern mainland (10 species). A single Orthocladiinae species was recorded in the Fraser Island region and none were found in tropical northern Queensland. Orthocladiinae are known to prefer cool water environments (Rosaro, 1991). Very few Orthocladiinae species had been distinguished from Australian lakes prior to this research, and were not distinguished to the genus or species level (Table 3.3). This study has provided the first detailed description of species (n=21) from the sub-family Orthocladiinae from Australian lakes. The results suggest that chironomid species are weak dispersers, a biogeographical factor that has probably strongly influenced their evolution (Cranston, 1995b). Several lakes in this study were subject to intense human activity such as boating, swimming and irrigation. Even in lakes subject to intense activity there did not appear to be a greater presence of cosmopolitan species, which could have resulted from human-related transportation from lake to lake.

Lake latitude and altitude strongly influenced chironomid species assemblages (Chapter 3). To a lesser extent, lake electrical conductivity and pH were also found to be influential (Chapter 3). Dimitriadis and Cranston (2001) used these data (chironomid and physio-chemical data), together with modelled climatic data (Busby, 1991) and concluded that, in addition to the parameters identified here as influential factors; climatic regimes also influenced chironomid species distributions. Over the Australian continent- length scale of this study, latitude and altitude are strongly associated with rainfall and temperature patterns and thus one may be considered a surrogate for the other. In contrast, other studies have indicated that water chemistry of Australian lakes is more strongly influenced by local factors

170 such as hydrology, geology, vegetation and distance to the ocean (e.g Bayly, 1964; Hawkins et al., 1988).

A methodology for the collection of chironomid exuviae from lakes (Chapters 2, 3) was developed through a 12-month study of exuviae from a single lake, Lake McKenzie. The resulting method is a rapid and effective way to collect an inventory of species living within a lake. Testing the method resulted in a total of 30 chironomid species from Lake McKenzie, a small dune lake at Jervis Bay (Chapter 2). This result is the largest chironomid species inventory ever reported from an Australian lake. The previous highest number of species from an Australian lake, of 15 species, was reported by Fulton (1983a and b) from large Tasmanian lakes. In the current study the most common species from Lake McKenzie were detected throughout the year, signifying that they had year-round emergence patterns (at least as far south as this).

The study of chironomids from Lake McKenzie (Chapter 2) revealed a much richer inventory of species (n=30) than had been recorded from any other Australian lake. Previous studies of dune lakes had suggested that these lakes provided a harsh environment for biota and consequently had fewer species present (Timms, 1985 and other authors, see Table 3.2). Internationally, the lakes that have been reported to have the richest chironomid fauna have been subject to detailed studies over many years, frequently using the collection of larvae, pupae and adult life stages (e.g. Humphries, 1938; Aagard, 1978; Hare & Carter, 1987) each resulting in 86, 76 and 80 species respectively. Based on the number of species collected over a relatively short duration (the longest being a one year study in Lake McKenzie), with a focus on only one stage of the lifecycle (exuviae) and a study restricted to a limited range of east coast waterbody types (eg. rivers and adjacent lagoons were not sampled) it is hypothesised that the Australian chironomid fauna is much more diverse than has been identified in this thesis.

Collecting exuviae from lakes was clearly demonstrated to provide an effective method for collecting species-level chironomid data from lakes (Chapter 2, 3). This study also emphasises that careful sampling strategies

171 need to be employed to rigorously sample exuviae from lakes. For example the spatial study of exuviae from eight different points of Lake McKenzie (Chapter 2) revealed a highly spatial heterogeneous distribution along the shoreline. One of the eight sites contained 95 % of the total number of exuviae found within the lake (Table 2.1). Benthic sampling of chironomid larvae from lakes reveals a much more even spatial distribution of larvae, with the notable exception of species assemblage differences according to water depth (e.g. Aagard, 1978). This finding illustrates the difficulties faced if a researcher fails to collect exuviae from the location of maximum exuviae aggregation within a lake. This study concluded that the location of exuviae within lakes is strongly influenced by wind and waves and sampling exuviae from lakes needs to carefully observe such conditions prior to determining the location of sampling sites. From a practical point of view, it is suggested that exuviae are collected from more than one location along a lake shoreline, and that net contents are frequently examined during sampling, to assess sampling performance.

The family-level macroinvertebrate community survey of zinc and sewage waste discharges to otherwise unimpacted upland streams in the Blue Mountains detected marked ecological impairment due to these waste sources (Chapter 4). This segment of the study was unique in Australia. It provided information on pristine waterways and the direct biological and chemical impacts from the release of two very different types of waste material. In such a setting it was possible to clearly and directly attribute the biological and chemical effects of two separate and the resulting pollution effects. Macroinvertebrate families and chironomid larvae responded in different ways to the two different waste discharges. In comparison to pristine streams, chironomids responded negatively (reduced abundance) to the zinc discharge, although compared to other macroinvertebrate taxa they maintained higher levels of abundance (see Table 4.5). This indicated that even under conditions of heavy-metal pollution chironomids remained an important component of the macroinvertebrate community. In contrast to their reduced abundance in the presence of zinc-pollution, chironomids responded positively (increased abundance) to the organic pollution of

172 sewage. They were the only macroinvertebrate family to display such a response.

The heavy-metal pollution in the otherwise cool, clear, slightly acid and relatively dilute Grose River reported in this study is a potentially unique occurrence in an Australian waterway. This situation may be the first reported case of mine drainage from a non-metaliferrous mine (Canyon Colliery was an underground coal mine) causing heavy metal pollution and ecosystem degradation. All other macroinvertebrate investigations of heavy- metal pollution relate to metalliferous mining activity (Nicholas & Thomas, 1978; Norris et al., 1982; Mackey, 1988; Napier, 1992; Norris & Sloane, 2003). The other unusual feature of the pollution was that it only involved zinc at levels associated with toxicity to the aquatic environment. The other metals were at levels below recognised toxicity thresholds (ANZECC, 1992 and 2000). All other investigations of macroinvertebrates and heavy-metal pollution contain a cocktail of metals at toxic levels (Nicholas & Thomas, 1978; Norris et al., 1982; Mackey, 1988; Napier, 1992; Norris & Sloane, 2003). For example, heavy-metals in the pollution investigation of the South Esk River, Tasmania, by Norris et al. (1982) included Cadmium, Copper and Lead at from 35 to 100 times ecosystem protection guideline levels (ANZECC, 1992 and 2000). Consequently this study offered an unusual opportunity to study the response of aquatic invertebrates to the single heavy-metal, zinc, alone at toxic levels.

Drift-netting was used to collect chironomid exuviae from upland temperate streams, a technique that is still in its infancy in Australia. To test and develop an appropriate sampling methodology for small upland streams, I conducted a temporal investigation of chironomid exuviae from a pair of small upland Blue Mountain waterways (Chapter 5). Internationally these waterways were amongst the smallest, narrowest and shallowest yet published for chironomid exuviae. Abundance and species richness of exuviae exhibited diurnal patterns. The majority of chironomid exuviae and species were collected at night, particularly during the hours between sunset and midnight. This implied that the adults in these upland streams had crepuscular and nocturnal emergence patterns. Unlike several overseas studies, none of the abundant species had a daylight emergence pattern. To

173 maximise species diversity for biomonitoring purposes, I concluded that under ‘normal’ flow regimes multiple 24-hour drift net samples were required (Chapter 5).

Different diurnal patterns in chironomid emergence were revealed in the temporal collections of exuviae in flowing waters (Chapter 5) and from Lake McKenzie (Chapter 2). It appeared that the majority of adult chironomids emerged from the upland streams during night hours but during both the day and night from Lake McKenzie.

A relationship between flow regimes and the species richness and abundance of exuviae in upland streams was identified (Chapter 5). Heavy rainfall and increased stream flow increased both. It is unlikely that this was due to an increase in emergence during times of rain and/or high flow as the considerable literature on chironomid emergence has established that the most important environmental cues for emergence are changing light levels and temperature (Learner et al., 1990; Armitage, 1995). During low-flow conditions, due to below average rainfall, the opposite response was apparent, i.e. low abundance and reduced species richness. Internationally, the only similar finding to this was also recorded after drought and low-flow conditions. A similar reduction in exuviae abundance and species richness was observed in a tropical northern Australian waterway (Cranston et al. 1997). A possible explanation for the lower than normal exuviae species- richness and abundance in low-flow conditions may be due to a greater tendency for exuviae to become entrapped along the stream bank and associated vegetation. It is probable that the differences in abundance and species richness in high flow events were due to ‘flushing’ of exuviae that had previously accumulated amongst stream bank debris. Another explanation may be that exuviae variation is correlated with stream discharge. As flow rates increase after heavy rain, a greater volume of water is filtered by the drift nets and exuviae abundance also increases.

Another major finding from this thesis was that chironomid species assemblages were strongly impaired by zinc-contaminated mine drainage and sewage effluent (Chapter 6). This is the first Australian evidence that many chironomid species are intolerant of heavy-metal pollution. These

174 findings are in contrast to those of earlier Australian heavy-metal pollution studies (Norris et al., 1982; Norris, 1986; Cranston et al., 1997). This research also revealed further evidence that many chironomids species are intolerant of sewage pollution. These outcomes are supported by previous organic pollution studies (Arthington et al., 1982; Pearson & Penridge, 1987; Wright, 1994). This constitutes a substantial contribution to our understanding of Australian chironomids and their relationship with stream water quality. The degree of pollution impact on chironomid species assemblages was similar for both heavy-metal and sewage pollution. The number of chironomid species collected at the polluted sites was about a third (9 to 11 species) of that found in pristine areas (27 to 30 species). Species assemblages at the polluted sites were dominated by very few species, some of which were highly abundant. Few chironomid species tolerated both types of water pollution. In contrast to the family level results, which showed that chironomids were one of the dominant macroinvertebrate groups at polluted sites (Chapter 4), the species results (Chapter 6) demonstrated that the majority of chironomid species did not tolerate either type of polluted waters. The apparent ‘tolerance’ of chironomid larvae observed (Chapter 4) can be attributed to the influence of a few tolerant and abundant species (Table 6.4, 6.7). Most species were shown to be intolerant of pollution. This is in contrast to the water pollution tolerances reported by some other Australian researchers (e.g. Norris et al., 1982; Chessman, 1995).

Chapter 6 describes the second only Australian investigation to exclusively survey chironomid exuviae in a flowing water pollution study, and was the first such study in temperate Australia. These results, however, were very different from the previous study of chironomid exuviae and heavy-metal pollution in a tropical Australian stream (Cranston et al., 1997) which revealed higher chironomid species richness below the waste discharge point than upstream. They suggested that this may have been due to the different flow regimes.

Chapter 4 revealed that zinc and organic pollution caused marked ecological impairment to the Grose River system macroinvertebrate community. It was evident that the communities were very different at degraded and reference

175 sites. Multivariate analysis of community structure, using non-metric multidimensional scaling, clearly illustrated that very dissimilar communities were present at reference and polluted sites. Reduction in taxon richness, mostly from family-level data, was accurate for detecting impairment from zinc-pollution but was inaccurate at the STP effluent receiving site where it actually increased. A decline in total abundance was apparent at the zinc- polluted site yet the opposite occurred at the STP affected sites, where total abundance increased.

For management purposes the family level analysis of the macroinvertebrate community (Chapter 4) produced clear evidence of pollution-related ecological impairment. Numerous rapid macroinvertebrate assessment methodologies are widely used in Australia that are quite similar to the techniques and analysis used here (e.g ANZECC, 2000). The biological consequences of disturbance (or pollution) are not fully apparent from family-level results. The species-level study of chironomids (Chapter 6) demonstrated that chironomid species assemblages were strongly impaired by water pollution and that most species were intolerant of both types of pollution. The organic pollution response of chironomids, in particular, appeared to be very different at the species versus family level (Chapter 4 versus Chapter 6). Chironomid larvae appeared to favour the STP-impacted site, where their abundance increased, whereas species data revealed that fewer chironomid species were found at this site than at pristine sites, with those few species present having high abundances. In some ways the family-level analysis of the chironomid response provided misleading results. This was very similar to the majority of pollution studies in Australia that have assessed chironomid larvae only at the family level. These findings contribute to the long running debate about the importance of taxonomic resolution in freshwater macroinvertebrate water pollution studies (e.g Resh & Unzicker, 1975; Cranston, 1990; Wright et al., 1995). Some authors have established that family, or even order, level macroinvertebrate studies are capable of detecting pollution disturbance (e.g. Chapter 4; Chessman, 1995; Growns et al., 1995; Wright et al., 1995). The issue of what level of taxonomic determination is required is critically important and depends on the purpose of the study. Effective management conclusions

176 regarding pollution impacts can be drawn from family-level studies (such as Chapter 4) that do not need species-level information, such as was revealed for chironomids in Chapter 6.

According to many Australian studies, the family Leptophlebiidae is widely reputed to be an obligate clean-water taxonomic group, strongly intolerant of water pollution (Norris, 1986; Cosser, 1988; Wright, 1994; Chessman, 1995; Chessman, 2003). This study provided further support for this reputation. It was observed that the group was strongly and equally impacted by both sewage and zinc pollution. It was the only abundant group that displayed a similar degree of impact to both pollution types investigated (Table 4.6). It has been estimated that 54 species of Leptophlebiidae occur in Australian waters (Hawkings & Smith, 1997). The complete absence of Leptophlebiidae at the most zinc-polluted and organic-polluted sites suggest that no additional information would have been likely to be gained from identification of Leptophlebiidae to the species level, at the most polluted sites. In contrast, larvae from the family Chironomidae displayed a very different response to sewage and zinc pollution (Table 4.6) with a strong adverse response to zinc pollution and a mild positive response to sewage (Chapter 4). Results from Chapter 6 provide solid evidence that the chironomid species-level response is very different from the family-level response. Chironomid species displayed a wide spectrum of response at the species level, with most not tolerant of either type of pollution (Table 6.4 and 6.7). Most previous stream biota studies in Australia did not detect this due to reasons explained previously (e.g. sampling of larvae, taxonomic difficulties), but an organic pollution study in southern Queensland (Arthington et al.,1982) and a pesticide study (Ward, et al., 1995) support this finding. A number of Australian heavy-metal pollution studies have reported tolerance of chironomid species and genera (Nicholas & Thomas, 1978; Norris et al., 1982; Sloane & Norris, 2003). The chironomid findings in this study add further support to the value of undertaking species-level pollution investigations for this ubiquitous, abundant and species-rich family.

An unusual response of Hydroposychidae caddisfly larvae was also uncovered (Chapter 4). It was the only taxon that exhibited a mild positive response to zinc pollution and a strong negative response to sewage

177 pollution (Table 4.6). This finding contrasts to the SIGNAL-MET grade conclusions (Chessman unpublished in Sloane & Norris, 2003). The basis for the different response may be due to a very different assemblage of Hydropsyche species present at the locations in which each pollution study was located, each of which has a different zinc-pollution tolerance. Only one other Australian study recorded a similar degree of metal-pollution tolerance in this caddisfly larvae (Napier, 1993). Napier’s 1993 study was also conducted on NSW upland streams, and was located only about 100 km to the north-west of the current study. It is possible that similar Hydropsyche species are shared between the Daylight Creek study and the waters sampled in the current study, giving a similar ‘tolerance’ to heavy-metal pollution. From this evidence it is likely that pollution studies of Hydropsychidae caddisflies have differential responses at the species level, perhaps similar to that demonstrated by chironomids in Chapter 6. However there are many fewer types of Hydropsychidae caddis species in Australian waters than chironomid species. It was reported by Hawkins and Smith (1997) that 27 species of Hydropsychidae caddis are known from across Australia. In contrast, chironomids are much more species-rich with more than 30 species frequently collected in this study from a single site on each sampling occasion.

This research does provide very strong evidence that the reputation that Australian chironomids are tolerant of water pollution is inaccurate and misleading. This is probably due to false conclusions drawn from family-level studies (e.g Chapter 4), and also due to difficulties of accurate species level identification of larvae. The most polluted zones in the Grose River (Table 6.4, 6.7) both contained greatly reduced chironomid species richness due to the loss of water-pollution sensitive species that were found at adjacent pristine sites. The extent of the loss of sensitive species, in comparison to local pristine sites in the Grose catchment was much greater than had been published in other pollution studies within Australia (Norris et al., 1982; Norris, 1986; Cranston et al., 1997) and worldwide (e.g Wilson & McGill, 1977; Armitage & Blackburn, 1985).

A comparison of the chironomid species collected in the lake survey (Table 3.4) with those recorded in the sewage and zinc pollution study in upland

178 streams (Table 6.2) revealed that some species were present in both lakes and running waters. The communities in the upland Blue Mountains streams were most similar to those found in Tasmanian lakes. For example, species from the sub-family Orthocladiinae were more frequently encountered in Tasmanian lakes (n=15) than the other three biogeographic regions (n = 0- 10). In the pristine Blue Mountains streams a similar number of Orthocladiinae species were found (n=14 -17). Only two of the five Australia- wide cosmopolitan lake species were also recorded in the Blue Mountains streams (Paramerina sp., Procladius paludicola). The Paramerina sp. was also one of the few species that displayed tolerance to both zinc and sewage pollution (Table 6.4, 6.7). This perhaps provides some indication of its adaptability and broad environmental tolerances.

Although many lakes in Tasmania had very dilute water, similar to pristine sites in the Blue Mountains, many species found in Tasmanian lakes also tolerated highly polluted conditions in the Blue Mountains. For example, two species found in Tasmanian lakes (Thiennemanniella sp. and MO5) were also found to display a high level of pollution tolerance in Blue Mountains streams. Only one species, Cardiocladius australiensis, was found exclusively at an organically polluted Blue Mountains site and in no other flowing or lake site in the study. The wider distribution of this species has been reported from fast flowing waterways, including waterfalls, and from organic polluted waterways (Cranston, 1996). The similarities between the number of species shared between Tasmanian lakes and the cool upland Blue Mountains streams in this study adds further evidence to the link between chironomid species and temperature regimes (see Chapter 3).

Although collecting exuviae provided more efficient and accurate identification of chironomid species living in flowing waters than does larvae, the employment of a rigorous sampling strategy is essential, as it was also illustrated for lakes (Chapters 2, 3). For example, it was shown (Chapter 5) that a high degree of temporal variability existed in exuviae collected from Hat Hill and Victoria Creeks over diurnal cycles. If the collection is conducted only in daylight hours, the resulting collection of exuviae will preclude the majority of species that exhibited crepuscular or nocturnal emergence patterns. In addition, the relationship of stream flow regimes and

179 exuviae was also illustrated (Chapter 5, 6). In particular, higher stream flows after heavy rainfall resulted in a ‘flushing’ of exuviae from a stream and a species-rich and abundant collection of exuviae. In contrast, drought- influenced lower flow regimes appeared to result in much lower abundance of exuviae and fewer species collected. This suggested a relationship between abundance and species richness of exuviae collected over time to the energy levels within a stream, proportional to stream flow regime. Although streams in this study were narrow, position of a sampling net across wider streams is potentially another source of sample variability that has been considered by other authors (e.g. Wilson & Bright, 1973; Hardwick et al., 1995).

An ethical consideration that supports the collection of chironomid exuviae is that their collecting provides physical evidence, without killing the individual, that an adult of that species has recently emerged from that waterbody. Almost all other macroinvertebrate, and other biological studies, kill the target biota. In waterways of high conservation value such non-lethal research is desirable.

Just as this study has made a contribution to our knowledge of lake-dwelling chironomid biodiversity, many other groups of invertebrates that inhabit Australia’s freshwater lakes remain poorly studied and warrant further examination. For example, insect groups such as Ephemeroptera mayflies, Odonata dragonflies and Trichoptera caddisflies are frequently found living in standing waters (personal observation) yet species-level studies of such lake fauna are rare in Australia.

Chironomids form a very large part of the biomass and biodiversity of Australian streams and lakes. They have important ecological functions, including the recycling of nutrients and energy. Chironomid larvae, pupae and adults are also a major food source of many types of biota e.g fish, birds, platypus (Maher & Carpenter, 1984; Armitage, 1995). Further study of Australian stream and lake chironomids, possibly using exuviae for species assessment, is warranted to further understand their contribution to ecosystem processes.

180 7.2 Human impact on high conservation streams and lakes

Due to the previous underground mining activity (Macqueen, 1997), the derelict coal mine that caused the heavy metal pollution, reported in Chapter 4 and 6, although closed five years prior to this study, continued to release polluting drainage into a high conservation value stream system, the Grose River. The Blackheath STP has been discharging treated sewage effluent to the Grose River system, for decades. The pollution of this National Park waterway, within a declared Wilderness area, and a World Heritage area, should serve as a reminder of the potential long-term pollution consequences of some human activities, particularly in isolated and sensitive locations. This is one of the first accounts of a derelict coal mine causing zinc contamination and waterway ecosystem impairment in Australia. Such pollution is normally associated with metalliferous mining activity. It is also one of the first accounts of STP effluent discharged into a high conservation value waterway system in Australia. I suggest that aquatic ecosystem monitoring below potentially ecosystem damaging waste discharges be conducted, as a condition of the discharge license, particularly in such high conservation value waterways. Species-level assessment of macroinvertebrate communities is most appropriate. This study provided testimony that chemical assessment of water quality of waste discharges at the point of discharge is not an effective measure of water pollution alone. It was noted that the water quality evidence generated by this study suggested that the discharge of wastes from both the STP (NSW EPA licence number 1712) and the coal mine were within their licence conditions. (www.epa.nsw.gov.au/prpoe/licences/L1712.doc and L558.doc Accessed 18 February 2005). The pollution discharge licence for the coal mine (NSW EPA licence number 558) was surrendered in 2001, presumably due to cessation of the mine’s commercial production of coal in 1997. Although the coal mine is closed and the pollution licence has been surrendered, evidence from this study revealed that pollution of the Grose River is still evident many years later.

In the survey of a selection of Australian lakes (Chapter 2 and 3), I observed that most lakes, or their watersheds, had obvious signs of disturbance due to human activity. Only one of the lakes sampled was identified as being

181 free of any signs of human disturbance, this was a small unnamed lake in the south-east of Tasmania (Labelled ‘TLT’, Table 3.4). This was also the

most southern lake sampled in the entire survey, at longitude 43 36’ South. Sedimentation and infilling of lakes is a natural process that human activity accelerates (Timms, 1992). The majority of the lakes that were visited in this study were clearly modified to some extent by human activity. The most common disturbance was modification to the catchment of the lakes, such as clearing natural vegetation. Some lakes had suffered extensive multiple changes from human activity. For example, many lakes were used as water supplies, others were used for waste disposal and a large number were subject to recreational pressures. Some lakes, such as Lake Barrine and Eacham in tropical north Queensland, Atherton Tablelands, and Lake McKenzie, on Fraser Island, are highly valued as tourist attractions and accommodate a very high level of human visitation, recreation and associated activities. One of Australia’s best examples of a limestone ‘sink- hole’ lake in south-eastern South Australia, Lake ‘Little Blue Lake’, receives road drainage directed from an adjoining road, and on the bottom of this lake is a dumped car. I sought to sample undisturbed lakes, but the search for such lakes promptly revealed that this was impractical. Lakes are probably amongst the most vulnerable of natural environments and present particularly difficult challenges to manage and protect their natural attributes.

This study poses questions about the conservation significance and management of lakes, waterways, their catchments and their biota. The majority of waterways in this study were located within National Park estates, and many were additionally also within declared World Heritage areas. Such listings did not necessarily lead to protection of the waterway from disturbance and pollution. For example the sewage and zinc pollution of the Grose River was located within a National Park, declared Wilderness area and World Heritage area. Perhaps one of the factors that influenced the lack of protection of waterways is a lack of information and expertise to assess the health of aquatic ecosystems.

Another factor may be the lack of available information on conservation significance of aquatic invertebrates. One of the ‘tools’ used to measure the conservation significance of an area of land, or a waterway is the relative

182 level of species present that have a legislated conservation status. For example, species may be listed in Australia under Federal (Federal Endangered Species Protection Act, 1992) or State legislation (e.g. NSW Threatened Species Conservation Act, 1995) as rare, threatened, endangered or vulnerable species. No chironomid species, and very few invertebrate species, are listed under any similar legislation in Australia. This study identified many new species, some of which may satisfy threatened species criteria. For example, this study collected the holotype and only scientific material from the sub-family Orthocladiinae Botryocladius austroalpinus (Table 3.12; Cranston & Edward, 1999) from alpine lakes in Tasmania, of glacial origin. The discovery of this species has implications for understanding the evolutionary and biogeographical natural history of Tasmania and other areas, as several other species in the Botryocladius genus were also found in Chile and Argentina (Cranston & Edward, 1999).

Further examination of chironomids and other invertebrates to determine their taxonomy and life history, distribution, range, and habitats, with a view to determining their conservation status is required. Currently the listing of endangered animal species in Australia is dominated by mammals, birds, fish, amphibians and reptiles (e.g. Burgman & Lindenmayer, 1998). There has been some debate about this issue in Australia and the technical and time consuming nature of assessing the conservation significance of invertebrates is recognised as a major impediment (New, 1984, 1995; Yen & Butcher, 1997; Chessman & Williams, 1999). The identification and listing of endangered invertebrates may assist in improving the conservation status and provide a justification for appropriate protection and management. In addition, similar legislation, such as the NSW Threatened Species Conservation Act, 1995, also has provisions to declare threatened populations, communities and key threatening processes. Many of the disturbances observed, such as water pollution, irrigation, recreational pressure and land clearing may deserve declaration as ‘key threatening processes’ to enable appropriate land and management practices to protect the waterways. Similar themes were also discussed by Chessman and Williams (1999) in their study of conservation status of river macroinvertebrates in western Sydney. They highlighted the threatening

183 processes of water pollution, loss of natural catchment vegetation, erosion and sedimentation, alterations to stream hydrology and habitat disturbance in disturbing and degrading river macroinvertebrate communities. In general they found that all of these factors were associated with urban expansion, which they concluded was the single largest threat to biodiversity of western Sydney. In this study, particularly in the Grose River, the threatening processes, of waste discharges, have been extended well beyond areas of human activity into the most pristine of World Heritage Areas.

7.3 Recommendations

This thesis presents much new information about chironomids in Australian waterways. It provides the basis for further study into their biodiversity, distribution, life histories and response to natural factors and human influences. Such studies should be underpinned by sound taxonomic information and further taxonomic work would be important to underpin these studies. The majority of species recorded in this study are undescribed.

Further examination of the distribution of chironomid species in Australian lakes would benefit from the collection of lakes in other biogeographic regions, such as south-western Australia, and tropical northern Australia. In addition it would be beneficial to study chironomids from more lakes along an altitudinal gradient, particularly in order to conduct a detailed examination into the relationship between species assemblages and the temperature regimes of lakes. Similar studies of running water chironomids along a variety of environmental gradients would also be valuable, and would be easier to achieve, since so few natural lakes are available in Australia.

Further seasonal studies of exuviae from lakes in different biogeographic regions would be valuable. The only such Australian data was from Lake McKenzie (Chapter 2), located in a temperate region. Overseas studies have discovered highly seasonal patterns of chironomid emergence in waterways from cooler northern latitudes (Armitage, 1995). Of particular interest would be to investigate the chironomid fauna in lakes that freeze in winter, such as in the NSW and Tasmanian alpine areas. Such lakes offer

184 unique environments for aquatic biota and would potentially be inhabited by species with particular adaptation for over-wintering as larvae under the ice, and would have a highly seasonal emergence pattern. At the same time comparison could be made of chironomid emergence patterns in lakes from temperate areas.

Further investigation of chironomid adult emergence patterns from lake and running water are warranted. This study made a contribution to our understanding of emergence patterns and environmental triggers of emergence in Australian freshwaters. It provided the first such information from standing and running waters in temperate eastern Australia (Pinder et al., 1991; Hardwick et al., 1995). It is evident from international comparisons of chironomid emergence data that little information is available from Australian waters (Learner et al., 1990). Emergence of chironomid adults is a biological phenomenon that has important implications for all freshwater ecosystems due to their importance as a food source for numerous other types of biota (Tokeshi, 1995). Although chironomids are known to be an important food source in Australian waterways (e.g. Maher & Carpenter, 1984) little Australian information is available about their use as food for other species such as fish, birds, other invertebrate and amphibia (e.g. Armitage, 1995). Adults are highly vulnerable to predation as they gradually emerge from their floating pupal exuviae on the surface of the water, probably the most vulnerable point in their lifecycles (Armitage, 1995). The nocturnal emergence patterns observed in this study may have been to avoid day-time visual predators. Further investigation into chironomids as a food source for other fauna may yield information about why some species emerge at different times of the day.

Results from this study contrast with the chironomid pollution-response from some previous Australian pollution studies (e.g. Norris et al., 1982; Smith & Cranston, 1993; Chessman, 1995; Cranston et al. 1997; Sloane & Norris, 2003). A larger number of Australian studies only considered the family response of chironomids, and as was demonstrated in Chapter 4, this offers little information about the response of chironomid species. The reputation arising from many studies has probably led to incorrect conclusions that chironomids are not responsive to water pollution. Two well-studied

185 Australian waterways that were reported to be strongly impacted by heavy metal pollution, Tasmania’s South Esk River (Norris et al.1982) and NSW’s Molongolo River (Norris, 1986; Sloane & Norris, 2003) could be re- examined, using the exuviae collection methods outlined in Chapters 5 and 6. This would test the hypothesis generated from this study that chironomid species richness was significantly underestimated by previous macroinvertebrate pollution studies, and that chironomid species have a wide range of pollution tolerances.

Chironomids warrant further testing as biomonitors. The current study found that chironomids showed a clear response to two different types of water pollution, sewage effluent (organic) and heavy metal (zinc). There are many other types of water pollution that warrant similar examination, for example pesticides, hydrocarbons, suspended particulate matter and salinity. In addition it would also be very useful to examine combinations of wastes, in a similar manner to the pollution of waterways from multiple waste sources. Examining the response of chironomids to other human pressures that can impair aquatic ecosystems is required. Such activities include the loss of catchment vegetation, changes to natural hydrology patterns, abstraction of water, regulation of flows, degradation of the waterway habitats, invasion of exotic species and other types of disturbances.

Further examination of toxicity of pollutants to chironomid species would be highly beneficial for improved management of wastes. Chironomids have been cultured in the laboratory as ecotoxicology test organisms (e.g. Lindeegard, 1995), particularly in the northern hemisphere. The results of this study provide an indication that Australian species of chironomids may have a valuable future as ecotoxicology test species. It is anticipated that much testing and development would be required to successfully culture species in laboratory conditions. When such technologies are achieved ecotoxicology testing can help reveal much more detailed information on the specific effects of individual pollutants. This can have important implications regarding the environmental requirements for the quality of waste waters prior to their discharge into freshwater environments.

186 This thesis benefited from in-situ field studies that were superior in many ways to laboratory controlled studies. It is currently not possible to culture a wide variety of chironomid species as laboratory ecotoxicology test animals. As an alternative, this study conducted an ecological pollution study of the Grose River system using a rare combination of two different waste sources discharging into an otherwise pristine stream system. Such a rare combination offers an invaluable opportunity to explore different pollution impacts on a specific group of organisms. It would be ideal to create an artificial stream system (e.g. Ward et al., 1995) to conduct further experimental manipulation of wastes and other variables, but the Grose River study area was too rugged and isolated to make it a feasible option.

This thesis provides a substantial contribution to the body of knowledge of chironomids in Australian lakes and rivers. It reveals the range across different lake types on a continental scale having a distinct north-south pattern of species assemblages with most individual species having a limited geographical distribution. It also reveals that geographic location linked to temperature regimes, in particular, influence chironomid distribution. This thesis provides one of Australia’s most detailed species- level investigations into the pollution response of chironomids to sewage effluent and heavy-metal pollution and discovers that species exhibit a wide range of water pollution tolerances. Chironomids demonstrate all of the key attributes required for biological indicators; they are usually found in large numbers, they are species rich, they are easy to collect, they are easy to identify and the individual species have a wide range of responses to pollution. In short, chironomids are excellent environmental indicators.

187 Chapter 8 References

Aagaard, K. (1978) The chironomids of lake Malsjoen. A phenological, diversity, and production study. Norwegan Journal of Entomology. 25: 21- 37.

Aagaard, K. (1986) The chironomid fauna of North Norwegian lakes, with a discussion on methods of community classification. Holarctic Ecology. 9: 1- 12.

Anonymous (1994) River Bioassessment Manual, Version 1.0, National River Processes and Management Program, Monitoring River Health Initiative. Department of the Environment, Sport and Territories; Commonwealth Environment Protection Agency and Land & Water Research & Development Corporation, Canberra.

ANZECC (Australian and New Zealand Environment and Conservation Council). 1992. Australian water quality guidelines for fresh and marine waters. National Water Quality Management Strategy. Australian and New Zealand Environment and Conservation Council, Canberra.

ANZECC (Australian and New Zealand Environment and Conservation Council) and ARMCANZ (Agriculture and Resource Management Council of Australia and New Zealand). 2000. Australian and New Zealand guidelines for fresh and marine waters. National Water Quality Management Strategy Paper No. 4. Australian and New Zealand Environment and Conservation Council/ Agriculture and Resource Management Council of Australia and New Zealand, Canberra.

APHA (1998) Standard Methods for the Examination of Water and Wastewater, 20th edition. American Public Health Association, Washington, DC.

Armitage, P. and Blackburn, J.H. (1985) Chironomidae in a pennine stream system receiving mine drainage and organic enrichment. Hydrobiologia. 121: 165-172.

188 Armitage, P., Cranston, P. and Pinder, C. (1995) Preface, in The Chironomidae: The biology and ecology of non-biting midges, Eds., (Armitage, P., Cranston, P. and Pinder, C.) Chapman & Hall, .

Armitage, P. (1995) Chironomidae as food, in The Chironomidae: The biology and ecology of non-biting midges, Eds., (Armitage, P., Cranston, P. and Pinder, C.) Chapman & Hall, Melbourne.

Arthington, A.H., Conrick, D.L., Connel, D.W., and Outridge, P.M. (1982) The ecology of a polluted urban creek. Australian Water Resources Council. Technical Paper No. 68. Australian Government Printing Service, Canberra.

Arthington, A.H., Burton, H.B., Williams, R.W. and Outridge, P.M. (1986) Ecology of humic and non-humic dune lakes, Fraser Island, with emphasis on the effects of sand infilling in Lake Wabby. Australian Journal of Marine and Freshwater Research. 37: 743-764.

Barton, D.R., Oliver, D.R. and Dillon, M.E. (1994) A comparison of pupal exuviae and larval Chironomidae for biomonitoring of the impacts of agricultural practices on surface water quality, in chironomids: From Genes to Ecosystems, Ed. Peter Cranston, CSIRO, Melbourne.

Bayly, I.A.E. (1964) Chemical and biological studies on some acidic lakes of east Australian sandy coastal lowlands. Australian Journal of Marine and Freshwater Research. 15: 56-72.

Bayly, I.A.E. and Williams, W.D. (1964) Chemical and biological observations on some volcanic lakes in the south-east of South Australia. Australian Journal of Marine and Freshwater Research. 15: 123-132.

Bayly, I.A.E. Ebsworth, E.P. and Hang Fong Wan (1975) Studies on the lakes of Fraser Island, Queensland. Australian Journal of Marine and Freshwater Research. 26: 1-13.

Bazerque, M.F., Laville, H. and Broquet, Y. (1989) Biological quality assessment in two rivers of the Northern Plain of France (Picardie) with special reference to chironomid and diatom indices. Acta Biologica Debrecina Supplementum Oecologica Hungaria. 3: 29-39.

189 Bensink, A.H.A. and Burton, H. (1975) North Stradbroke Island a place for freshwater invertebrates. Proceedings of the Royal Society of Queensland. 86(7): 29-45.

Berg, M.B. (1995) Larval food and feeding behaviour, in The Chironomidae: Biology and ecology of non-biting midges, Eds., (P.D. Armitage, P.S. Cranston and L.C.V. Pinder), Chapman & Hall, Melbourne.

Boothroyd, I.K.G. (1988) Temporal and diel emergence of Chironomidae (Diptera: Insecta) from a New Zealand Stream. Verhandlung der Internationale Vereiningung für Theoetische und Angewandte Limnologie. 23: 1399-1404.

Bowling, L.C. and Tyler, P.A. (1984) Physicochemical differences between lagoons of King and Flinders Islands, . Australian Journal of Marine and Freshwater Research. 39: 535-553.

Blue Mountains City Council, State of the Environment Report 2001/2002, self-published, Katoomba.

Brittain, J.E. and Eikeland, T.J. (1988) Invertebrate drift – a review. Hydrobiologia. 166: 77-93.

Brinkhurst, R.C. (1974) The benthos of lakes. Macmillan Press, London.

Brundin, L. (1949) Chironomiden und andere bodentiere des südschwedischen urgebirgsseen. Ein beitrag zur kenntnis der bodenfaunistischen charakterzüge schwedischen oligotropher seen. Report of the Institute of Freshwater Research. Drottningsholm 30: 1-914.

Brundin, L. (1958) The bottom faunistic lake type system and its application to the southern hemisphere. Moreover a theory of glacial erosion as a factor of productivity in lakes and oceans. Verhandlung der Internationale Vereiningung für Theoetische und Angewandte Limnologie. 12: 288-297.

Brundin, L. (1966) Transantarctic relationships and their significance, as evidenced by chironomid midges. Kungl Svenska Vetenskapakademiens Handlingar. 11: 1-472.

190 Bunn, S.E., Edward, D.H., and Loneragan, N.R. (1986) Spatial and temporal variation in the macroinvertebrate fauna of streams of the northern jarrah forest, Western Australia: community structure. Freshwater Biology. 16: 67- 92.

Burgman, M.A. and Lindenmayer, D.B. (1998) Conservation Biology for the Australian Environment Ed.( Drill, C.) Surrey Beatty & Sons, Sydney.

Busby, J.R. (1991) BIOCLIM – a bioclimatic analysis and predictive system. In: Margules, C.R., Austin, M.P. (Eds.) Nature Conservation: Cost Effective Biological Surveys and Data Analysis. CSIRO, Melbourne.

Cairns, J. Jr. and Pratt, J.R. (1993) A history of biological monitoring using benthic macroinvertebrates, in Freshwater Biomonitoring and Benthic Macroinvertebrates, Eds. (D.M. Rosenberg and V.H. Resh), Chapman and Hall, London and New York.

Campbell, I.C. (1978) A biological investigation of an organically polluted urban stream in Victoria. Australian Journal of Marine and Freshwater Research. 29: 275-291.

Cheng, D.M.H. and Tyler, P.A. (1976) Nutrient economies and trophic status of Lakes Sorell and Crescent, Tasmania. Australian Journal of Marine and Freshwater Research. 27: 151-63.

Chessman, B.C. (1995) Rapid assessment of rivers using macroinvertebrates: A procedure based on habitat-specific sampling, family level identification and a biotic index. Australian Journal of Ecology. 20: 122- 129.

Chessman, B.C., Growns, J.E., and Kotlash, A.R. (1997) Objective derivation of macroinvertebrate family sensitivity grade numbers for the SIGNAL biotic index: application to the Hunter River system, New South Wales. Marine and Freshwater Research 48: 159-172.

Chessman, B.C. and Williams, S.A. (1999) Biodiversity and conservation of river macroinvertebrates. Pacific Conservation Biology. 5: 36-55.

191 Chessman, B.C. (2003) New sensitivity grades for Australian river macroinvertebrates. Marine and Freshwater Research 54: 95-103.

Clarke, K.R. (1993) Non-parametric multivariate analyses of changes in community structure. Australian Journal of Ecology. 18: 117-143.

Clarke, K.R. and Ainsworth. M. (1993) A method of linking multivariate community structure to environmental variables. Marine Ecology Progress Series 92: 205-219.

Coffman, W.P. (1995) Conclusions, in The Chironomidae: The biology and ecology of non-biting midges, Eds, (Armitage, P., Cranston, P. and Pinder, C.) Chapman & Hall, Melbourne.

Coffman, W.P. and Ferrington, L.C. Jr. (1984) Chironomidae, in An Introduction to the Aquatic Insects of North America, 2nd edn, (Eds. R.W. Merritt and K.W. Cummins), Kendall/Hunt, Dubuque.

Colless, D.H. (1986) The Australian Chaoboridae (Diptera). Australian Journal of Zoology, Supplementary Series No. 124.

Connell, D.W. (1981) Water Pollution: Causes and effects in Australian and New Zealand. Second Edition. University of Queensland Press, St Lucia.

Cosser, P.R. (1988) Macroinvertebrate community structure and chemistry of an organically polluted creek in S.E. Queensland. Australian Journal of Marine and Freshwater Research. 39: 671-683.

Cranston, P.S. (1990) Biomonitoring and invertebrate taxonomy. Environmental Monitoring and Assessment. 14: 265-273.

Cranston, P.S. (1995a) Introduction, in The Chironomidae: The biology and ecology of non-biting midges, Eds. (Armitage, P., Cranston, P. and Pinder, C.) Chapman & Hall, Melbourne.

Cranston, P.S. (1995b) Biogeography, in The Chironomidae: The biology and ecology of non-biting midges, Eds. (Armitage, P., Cranston, P. and Pinder, C.) Chapman & Hall, Melbourne.

192 Cranston, P.S. (1995c) Medical significance, in The Chironomidae: The biology and ecology of non-biting midges, Eds. (Armitage, P., Cranston, P. and Pinder, C.) Chapman & Hall, Melbourne.

Cranston, P.S. (1995d) Morphology, in The Chironomidae: The biology and ecology of non-biting midges, Eds. (Armitage, P., Cranston, P. and Pinder, C.) Chapman & Hall, Melbourne.

Cranston, P.S. (1996) Identification guide to the Chironomidae of New South Wales. AWT identification guide No. 1. Australian Water Technologies, West Ryde.

Cranston, P.S. (2000a) Electronic guide to The Chironomidae of Australia. http://entomology.ucdavis.edu/chiropage/index.html accessed January, 2004.

Cranston, P.S. (2000b) August Thienemann’s influence on modern chironomidology – an Australian perspective. Verhandlung der Internationale Vereiningung für Theoetische und Angewandte Limnologie. 27: 278-283.

Cranston, P.S. and Hare, L. (1995) Conochironomus Freeman: an Afro- Australian Chironomini genus revised (Diptera: Chironomidae). Systematic Entomology. 20: 247-264.

Cranston, P.S., Cooper, P.D., Hardwick, R.A., Humphrey, C.L. and Dostine, P.L. (1997) Tropical acid streams – the chironomid (Diptera) response in northern Australia. Freshwater Biology. 37: 473-483.

Cranston, P.S. and Edward, D.H.D. (1999) Botryocladius gen.n: a new transantarctic genus of orthocladiine midge (Diptera: Chironomidae). Systematic Entomology. 24: 305-333.

Cranston, P.S. and Dimitriadis, S. (2004) The Chironomidae (Diptera) larvae of Atherton tableland lakes, North Queensland. Memoirs of the Queensland Museum. 49 (2) 79-94.

Cummins, K.W. (1973) Trophic relations of aquatic insects. Annual review of Entomology. 18: 183-206.

193 Danks, H.V. and Oliver, D.R. (1972) Diel periodicities of emergence of some high arctic Chironomidae (Diptera). Canadian Entomologist. 104: 903-916.

Dimitriadis, S. and Cranston P.S. (2001) An Australian Holocene climate reconstruction using Chironomidae from a tropical volcanic maar lake. Paleaogeography, Paleaoclimatology, Paleaoecology. 176: 109-131.

Edward, D.H.D. (1983) Inland waters of Rottnest Island. Journal of the Royal Society of Western Australia. 66: 41-47.

Edward, D.H.D. (1986) Chironomidae (Diptera) of Australia, in Limnology in Australia, (Eds. De Deckker, P. and Williams, W.D.) CSIRO Dr W. Junk Publishers, Dordrecht.

Evrard, M. (1994) Daily evolution of the drift of pupal exuviae of Chironomidae (Diptera) in a chalk trout stream (the Samson) Belgium. Belgium Journal Zoology. 124: 115-126.

Fittkau, E.J. (1995) In memory of Lars Brundin, in chironomids: From Genes to Ecosystems, Ed. Peter Cranston, CSIRO, Melbourne.

Flannagan, J.F. (1970) Efficiencies of various grabs and corers in sampling freshwater benthos. Journal Fisheries Research Board of Canada. 27 (10): 1691-1700.

Forsyth, D.J. (1975) The benthic fauna, in New Zealand Lakes, (Eds. V.H. Jolly and J.M.A. Brown) Auckland University Press, Oxford University Press.

Franquet, E and Pont, D. (1996) Pupal exuviae as descriptors of the chironomid (Diptera: Nematocera) communities of large rivers. Archiv für Hydrobiologie. 138: (1) 77-98.

Frantzen, N.M.L.H.F. (1992) Water quality changes of the River Meuse assessed by chironomid pupal exuviae. Netherlands Journal of Aquatic Ecology. 26: (2-4) 543-549.

Fulton, W. (1983a) Macrobenthic fauna of Great Lake, Arthurs lake and Lake Sorell, Tasmania. Australian Journal of Marine and Freshwater Research. 34: 775-785.

194 Fulton, W. (1983b) Qualitative and quantitative variation in the macrobenthic fauna of the original lake and new lake areas of Great Lake and Arthur's Lake, Tasmania. Australian Journal of Marine and Freshwater Research. 34: 787-803.

Gerhadt, A., Janssens de Bisthoven, L. and Soares, A.M.V.M. (2003) Macroinvertebrate response to acid mine drainage: community metrics and on-line behavioural toxicity bioassay. Environmental Pollution. 130: 263-274.

Growns, J.E., Chessman, B.C., McEvoy, P.K. and Wright, I.A. (1995) Rapid assessment of rivers using macroinvertebrates: Case studies in the Nepean River and Blue Mountains, NSW. Australian Journal of Ecology. 20: 130- 141.

Hardwick, R.A., Wright, I.A., Jones, H.A., Chessman, B.C., and Holleley, D.A. (1994) Rapid biological assessment of stream in the Blue Mountains, Australia: Characteristics of the Chironomidae fauna. In chironomids - from Genes to Ecosystems Ed. Cranston, P.S. CSIRO Publications, Melbourne, Australia.

Hardwick, R.A., Cooper, P.D., Cranston, P.S., Humphreys, C.L., and Dostine, P.L. (1995) Spatial and temporal distribution patterns of drifting pupal exuviae of Chironomidae (Diptera) in streams of tropical northern Australia. Freshwater Biology. 34: 569-578.

Hare, L. and Carter, J.H.C. (1987) Chironomidae (Diptera: Insecta) from the environs of a natural West African lake. Entomologica scandinavica Supplements 29: 65-74.

Hawking, J.H. (1994) A preliminary guide to keys and zoological information to identify invertebrates from Australian freshwaters. Cooperative Research Centre for Freshwater Ecology Identification Guide No. 2. Cooperative Research Centre for Freshwater Ecology, Albury.

Hawking, J.H. and Smith, F.J. (1997) Colour guide Invertebrates of Australian inland waters. Identification Guide No. 8. Cooperative Research Centre for Freshwater Ecology, Murray-Darling Freshwater Research Centre, Albury.

195 Hawkins, P.R., Taplin, L.E., Duivenvoorden, L.J. and Scott, F. (1988) Limnology of oligotrophic lakes at Cape Flattery, North Queensland. Australian Journal of Marine and Freshwater Research. 39: 535-553.

Hellawell, J.M. (1986) Biological Indicators of Freshwater Pollution and Environmental Management. Elesvier, London.

Hickey, C.W. and Clements, W.H. (1998) Effects of heavy metals on benthic macroinvertebrate communties in New Zealand streams. Environmental Toxicology and Chemistry. 17: (11) 2338-2346.

Humphries, C.F. (1938) The chironomid fauna of the Grosser Plöner See, the relative density of its members and their emergence period. Archiv für Hydrobiologie. 33: 535-548.

Hynes, H.B.N. (1960) The Biology of polluted waters. Liverpool University Press.

Int Panis, L., Goddeeris, B. and Verheyen, R.F. (1995) On the reliability of Ponar grad samples for the quantitative study of benthic invertebrates in pools. Hydrobiologia. 312: 147-152.

Jacobson, G. and Schuett, A.W. (1984) Groundwater seepage and the water balance of a closed freshwater, coastal dune lake: Lake Windamere, Jervis Bay. Australian Journal of Marine and Freshwater Research. 35: 645- 654.

Johnson, R.K., Wiederholm, T. and Rosenberg, D.M. (1993) Freshwater biomonitoring using individual organisms, populations, and species assemblages of benthic macroinvertebrates, in Freshwater Biomonitoring and Benthic Macroinvertebrates, Eds. (D.M. Rosenberg and V.H. Resh), Chapman and Hall, London and New York.

Jolly, V.H. and Chapman, M.A. (1966) A preliminary biological study of the effects of pollution on Farmers’s Creek and Cox’s River, New South Wales. Hydrobiologia. 27:160-192.

196 Kansanen, P.K. Aho, J. and Paasivirta, L. (1984) Testing the benthic lake type concept based on chironomid associations in some Finnish lakes using multivariate statistical methods. Annales Zoologici Fennici. 21: 55-76.

Keith, D.A. and Benson, D.H.(1988) Natural vegetation of the Katoomba area. Cunninghamia. 2: (1): 107-144.

Kelly,M. (1991) Mining and the Freshwater Environment. Elsevier Science Publishers LTD, London.

Kolkwitz, R. and Marsson, M. (1909) Ökologie der tierischen Saprobien. Beiträge zur lehre von der biologischen Gewässerbeurteilung. Internationale Revue der Gesamten Hydrobiologie. 2: 126-152.

Kuijpers, A.M.J.P., Ketelaars, H.A.M., and Breemen, L.W.C.A. (1992) chironomid pupal exuviae and larvae of two storage reservoirs in the Netherlands. Netherlands Journal of Aquatic Ecology. 26 (2-4): 379-383.

Langton, P.H. (1995) The pupa and events leading to eclosion, in The Chironomidae: The biology and ecology of non-biting midges, Eds. (Armitage, P., Cranston, P. and Pinder, C.) Chapman & Hall, Melbourne.

Lake, P.S. (1987) Stream ecosystem research in Australia: problems and strategies. Australian Water Research Advisory Council. National Water Research Program. Canberra.

Learner, M., Wiles, R. and Pickering, J. (1990) Diel emergence patterns of chironomids. Internationale Revue Gesamten Hydrobiologie. 75: 569-581.

Lindegaard, C. (1995) Classification of water-bodies and pollution, in The Chironomidae: The biology and ecology of non-biting midges, Eds. (Armitage, P., Cranston, P. and Pinder, C.) Chapman & Hall, Melbourne.

Mackey, A.P. (1988) The biota of the River Dee (central Queensland, Australia) in relation to the effects of acid mine drainage. Proceedings of the Royal Society of Queensland. 99: 9-19.

Macqueen, A. Back from the Brink. Blue Gum Forest and the Grose Wilderness. (1997) Self-published. 39 Bee Farm Road, Springwood.

197 Maher, M., and Carpenter, S.M. (1984) Benthic studies of waterfowl breeding habitat in south-western New South Wales. II. chironomid populations. Australian Journal of Marine and Freshwater Research. 35: 97- 110.

Marchant, R., Barmutta, L.A. and Chessman, B.C. (1994) Classification of macroinvertebrate communities in Victoria. Freshwater Biology. 41: 253- 268.

Napier, G. M. (1992) Application of laboratory-derived data to natural aquatic ecosystems. Unpublished PhD thesis. Graduate School of the Environment, Macquarie University.

National Parks and Wildlife Service (1999) Blue Mountains National Park, Walking Track Guide. Walking Tracks in the Grose Valley. Self published, Blue Mountains Heritage Centre, Blackheath.

National Parks and Wildlife Service (2001) Blue Mountains National Park Plan of Management. Self published, Blue Mountains Heritage Centre, Blackheath.

New, T. (1984) Insect Conservation – An Australian Perspective. W. Junk, Dordrecht.

New, T. (1984) Focussing on species for invertebrate conservation. Victorian Naturalist. 112: 29-31.

Nicholas, W.L. and Thomas, M. (1978) Biological release and recycling of toxic metals from lake and river sediments. Australian Water Resources Council Technical Paper No. 33. Australian Government Publishing Service, Canberra.

Nolte, U. (1994) From egg to imago in less than seven days: Apedilum elachistus (Chironomidae), in chironomids: From Genes to Ecosystems. Ed. Peter Cranston. CSIRO Publications, Melbourne.

Norris, R.H., Swain, R. and Lake, P.S. (1981) Ecological effects of mine effluents on the South Esk River, North-eastern Tasmania 11. Trace Metals. Australian Journal of Marine and Freshwater Research. 32: 165-173.

198

Norris, R.H., Lake, P.S. and Swain, R. (1982) Ecological effects of mine effluents on the South Esk River, north-eastern Tasmania 111. Benthic macroinvertebrates. Australian Journal of Marine and Freshwater Research. 33: 789-809.

Norris, R.H. (1986) Mine waste pollution of the Molonglo River, New South Wales and the Australian Capital Territory: Effectiveness of remedial works at Captains Flat Mining Area. Australian Journal of Marine and Freshwater Research. 37: 147-157.

Norris, R.H. and Georges, A. (1993) Analysis and interpretation of benthic macroinvertebrate surveys. In Freshwater Biomonitoring and Benthic Macroinvertebrates. (Eds. D.M. Rosenberg and V.H. Resh) Chapman and Hall, New York.

Norris, R.H., Moore, J.L., Maher, W.A. and Wensing, L.P. (1993) Limnological characteristics of two coastal dune lakes, Jervis Bay, south eastern Australia. Australian Journal of Marine and Freshwater Research. 44: 437-458.

Oliver, D.R. (1968) Adaptations of Arctic Chironomidae. Annales Zoologici Fennici, 5:111-118.

Paterson, C.G. and Walker, K.F. (1974) Recent History of Tanytarsus barbitarsus Freeman (Diptera: Chironomidae) in the sediments of a shallow, Saline lake. Australian Journal of Marine and Freshwater Research. 25: 315-25.

Pearson, R.G., Benson, L.J., and Smith, R.E.W. (1986) Diversity and abundance of the fauna in Yuccabine Creek, a tropical rainforest stream. in Limnology in Australia, (Eds. De Deckker, P. and Williams, W.D.) CSIRO Dr W. Junk Publishers, Dordrecht.

Pearson, R.G. and Penridge, L.K. (1987). The effects of pollution by organic sugar mill effluent on the macro-invertebrates of a stream in tropical Queensland, Australia. Journal of Environmental Management. 24: 205-215.

199 Pinder, A.M., Trayler, K.M., and Davis, J.A. (1991) chironomid control in Perth wetlands. Final report and recommendations. School of Biological and Environmental Sciences, Murdoch University.

Pinder, L.C.V., and Morley D.J. (1995) Chironomidae as indicators of water quality - with a comparison of the chironomid faunas of a series of contrasting cumbrian tarns, in Insects in a changing environment, Eds. Harrington, R., and Stork, N.E., Academic Press, London.

Pinder, L.C.V. (1995) The habitats of chironomid larvae, in The Chironomidae: The biology and ecology of non-biting midges, Eds. (Armitage, P., Cranston, P. and Pinder, C.) Chapman & Hall, Melbourne.

Resh, V.H. and Unzicker, J.D. (1975) Water quality monitoring and aquatic organisms: the importance of species identification. Journal Water Pollution Control Federation. 47: 9-19.

Resh, V.H. and Jackson, J.K. (1993) Rapid assessment approaches to biomonitoring using benthic macroinvertebrates. In Freshwater Biomonitoring and Benthic Macroinvertebrates. (Eds. D.M. Rosenberg and V.H. Resh) Chapman and Hall, New York.

Rosaro, B (1991) chironomids and water temperature. Aquatic Insects. 13: Volume 2: 87-98.

Rosenberg, D.M and Resh, V.H. (1993) Freshwater biomonitoring and benthic macroinvertebrates. Chapman & Hall, New York, London.

Ruse, L.P., Herrmann, S.J., and Sublette, J.E. (2000) Chironomidae (Diptera) species distribution related to environmental characteristics of the metal-polluted Arkansas River, Colorado. Western North American Naturalist. 60 (1) 3456-3490.

Ruse, L.P. and Wilson, R.S. (1994) Long-term assessment of water and sediment quality of the River Thames using chironomid pupal skins, in chironomids: From Genes to Ecosystems, Ed. Peter Cranston, CSIRO, Melbourne.

200 Russel, D.J. (1987) Aspects of the limnology of tropical lakes in Queensland - with notes on their suitability for recreational fisheries. Proceeding of the Royal Society of Queensland. 98: 83-91.

Sæther, O.A. (1975) Nearctic chironomids as indicators of lake typology. Verhandlung der Internationale Vereiningung für Theoetische und Angewandte Limnologie. 19: 3127-3133.

Sæther, O.A. (1979) chironomid communities as water quality indicators. Holarctic Ecology. 2: 65-74.

Schofield N.J. and Davies P.E. (1996) Measuring the health of our rivers. Water, May/June, 39-43.

Sheehan, P.J. and Knight, A.W. (1985) A multilevel approach to the assessment of ecotoxicological effects in a heavy metal polluted stream Verhandlung der Internationale Vereiningung für Theoetische und Angewandte Limnologie. 22: 2364-2370.

Sloane, P.I.W. and Norris, R.H. (2003) Relationship of AUSRIVAS-based macroinvertebrate predictive model outputs to a metal pollution gradient. Journal North American Benthological Society. 22 (3): 457-471.

Smith, M and Cranston, P.S. (1995) "Recovery" of an acid mine-waste impacted tropical stream - the chironomid story. In chironomids - from Genes to Ecosystems. Ed. Cranston, P.S. CSIRO Publications, Melbourne.

Sydney Water Corporation (1999) Annual Environmental Indicators Compliance Report 1998 / 1999. Sydney Water Corporation, Sydney.

Sydney Water Corporation (2004) Annual Environmental Indicators Compliance Report 2003 / 2004. Sydney Water Corporation, Sydney.

Thienemann, A. 1910: Das Sammein von Puppenhauten der Chironomiden. Archiv für Hydrobiologie. 6: 213-224.

Thienemann, A. 1922: Die beiden Chironomus-arten der Tiefenfauna der norddeutschen Seen. Ein hydrobiologisches Problem. Archiv für Hydrobiologie. 13: 609-646.

201 Timms, B.V. (1973a) A limnological survey of the freshwater coastal lakes of East Gippsland. Australian Journal of Marine and Freshwater Research. 24: 1-20.

Timms, B.V. (1973b) A comparative study of the limnology of three maar lakes in western Victoria, unpublished PhD thesis Monash University.

Timms, B.V. (1974) Morphology and benthos of three volcanic lakes in the Mt. Gambier district, South Australia. Australian Journal of Marine and Freshwater Research. 25: 287-297.

Timms, B.V. (1975) basic limnology of two crater lakes in Western Victoria. Proceedings of Royal Society of Victoria. 87: 159-165.

Timms, B.V. (1976a) Morphology of Lake Barrine, Eacham and Euramoo, Atherton Tableland, North Queensland. Proceedings of Royal Society of Queensland. 87: 81-84.

Timms, B.V. (1976b) A comparative study of the limnology of three maar lakes in western Victoria I. Physiography and physicochemical features. Australian Journal of Marine and Freshwater Research. 27: 35-60.

Timms, B.V. (1977) A study of some coastal dune lakes in Western Victoria. Proceedings of Royal Society of Victoria. 89: 167-172.

Timms, B.V. (1978) The benthos of seven lakes in Tasmania. Archiv für Hydrobiologie. 81: 422-444.

Timms, B.V. (1979) The Benthos of some Lakes in North-eastern Queensland. Proceedings of Royal Society of Queensland. 90: 57-64.

Timms, B.V. (1980a) The benthos of the Kosciusko glacial lakes. Proceedings of Linnean Society of New South Wales. 104 (2): 119-125.

Timms, B.V. (1980b) The benthos of Australian lakes, in An ecological basis for water resource management, (Ed. WD Williams), Australian National University Press, Canberra.

Timms, B.V. (1981) Animal communities in three Victorian lakes of differing salinity. Hydrobiologia 81: 181-193.

202 Timms, B.V. (1983) A study of the benthic communities in some shallow saline lakes of western Victoria, Australia. Hydrobiologia. 105: 165-177.

Timms, B.V. (1985) The structure of macrobenthic communities of Australian lakes. Proceeding of Ecological Society of Australia. 14: 51-59.

Timms, B.V. (1986a) Reconnaissance limnology of some coastal dune lakes of Cape York Peninsular, Queensland. Australian Journal of Marine and Freshwater Research. 37: 167-176.

Timms, B.V. (1986b) The coastal dune lakes of eastern Australia, in Limnology in Australia, Eds. (De Deckker, P. and Williams, W.D.) CSIRO Australia, Melbourne Dr W. Junk Publishers, Dorrecht.

Timms, B.V. (1992) Lake Geomorphology. Gleneagles, Adelaide.

Timms, B.V. (1997) Study of coastal freshwater lakes in southern New South Wales. Marine and Freshwater Research. 48: 249-256.

Tokeshi, M. (1995) Life cycles and population dynamics, in The Chironomidae: The biology and ecology of non-biting midges, Eds. (Armitage, P., Cranston, P. and Pinder, C.) Chapman & Hall.

Tyler, M.J., Twidale, J.K., Ling, J.K., and Holmes, J.W. (1983) Eds. Natural history of the south east. Occasional Publications of the Royal Society of South Australia.

Walker, I.R. (1995) chironomids as indicators of past environmental change, in The Chironomidae: The biology and ecology of non-biting midges, Eds. (Armitage, P., Cranston, P. and Pinder, C.) Chapman & Hall, Melbourne.

Walker, I.R. and Mathews, R.W. (1987) chironomids, lake trophic status and climate. (reply; letters to the editor) Quaternary Research. 28: 431-437.

Ward, S., Arthington, A.H., and Pusey, B.J. (1995) The effects of a chronic application of Chlorpyriphos on the macroinvertebrate fauna in an outdoor artificial stream system: species responses. Ecotoxicology and Environmental Safety. 30: 2-13.

203 Wartinbee, D.C. (1979) Diel emergence patterns of lotic Chironomidae. Freshwater Biology. 9: 147-156.

Warwick, R.M. (1993) Environmental impact studies on marine communities: pragmatical considerations. Australian Journal of Ecology. 18: 63-80.

Watson, J.A.L., Arthington, A.H., and Conrick, D.L. (1982) Effect of sewage effluent on dragonflies (Odonata) of Bulimba Creek, Brisbane. Australian Journal of Marine and Freshwater Research. 33: 517-528.

Whitehurst, I.T. and Lindsey, B.I. (1990) The impact of organic enrichment on the benthic macroinvertebrate communities of a lowland river. Water Research. 24 (5): 625-630.

Wielderholm, T. and Eriksson, L. (1977) Benthos of an acid lake. Oikos 29: 261-267.

Williams, W.D. (1981) The limnology of saline lakes in western Victoria. Hydrobiologia. 82: 233-259.

Wilson, R.S. (1977) chironomid pupal exuviae in the River Chew. Freshwater Biology. 7: 9-17.

Wilson, R.S. and Bright, P.L. (1973) The use of chironomid pupal exuviae for characterizing streams. Freshwater Biology. 3: 283-302.

Wilson, R.S. and McGill, J.D. (1977) A new method of monitoring water quality in a stream receiving sewage effluent, using chironomid pupal exuviae. Water Research. 11: 959-966.

Wilson, R.S. and McGill, J.D. (1982) A practical key to the genera of pupal exuviae of the British Chironomidae (Diptera, Insecta), University of Bristol Printing Office, Bristol.

Wilson, R.S. and Wilson, S.E. (1983) A reconnaissance of the River Rhine using chironomid pupal exuviae (Insecta: Diptera). Memoirs of the American Entomological Society. 34: 361-385.

204 Winner, R.W., Boesel, M.W., and Farrell (1980) Insect community structure as an index of heavy-metal pollution in lotic ecosystems. Canadian Journal of Fisheries and Aquatic Sciences. 37: 647-655.

Wright, I.A. (1994) The ecological impacts of Wentworth Falls sewage treatment plant on Blue Mountains Creek. Unpublished Master of Science thesis. Graduate School of the Environment, Macquarie University.

Wright, I.A., Chessman, B.C., Fairweather, P.G., and Benson, L.J. (1995) Measuring the impact of sewage effluent on the macroinvertebrate community of an upland stream: The effect of different levels of taxonomic resolution and quantification. Australian Journal of Ecology. 20: 142-149.

Wright, I.A. and Cranston, P.S. (2000) Are Australian lakes different? chironomid and chaoborid exuviae from Lake McKenzie, a coastal temperate dune lake. Verhandlung der Internationale Vereiningung für Theoetische und Angewandte Limnologie. 27: 303-308.

Yasuno, M., Hatakeyama, S. and Sugaya, Y. (1985) Characteristic distribution of chironomids in the rivers polluted by heavy-metals. Verhandlung der Internationale Vereiningung für Theoetische und Angewandte Limnologie. 22: 2371-2377.

Yen, A. and Buthcher, R. (1997). An Overview of the Conservation of Non- marine Invertebrates in Australia. Environment Australia, Canberra.

205