106 Chapter 5 Groundwater Quality and Hydrodynamics of an Acid Sulfate Soil Backswamp 5.1 Introduction Intensive drainage of coastal floodplains has been the main factor contributing to the oxidation of acid sulfate soils (ASS) (Indraratna, Blunden & Nethery 1999; Sammut, White & Melville 1996). Drainage of backswamps removes surface water, thereby increasing the effect of evapotranspiration on watertable drawdown (Blunden & Indraratna 2000) and transportation of acidic salts to the soil surface through capillary action (Lin, Melville & Hafer 1995; Rosicky et al. 2006). During high rainfall events the oxidation products (i.e. metals and acid) are mobilised into the groundwater, and acid salts at the sediment surface are dissolved directly into the surface water (Green et al. 2006). The high efficiency of drains readily transports the acidic groundwater and surface water into the estuary and can have a detrimental impact on downstream habitats (Cook et al. 2000). The quality of water discharged from a degraded ASS wetland is controlled by the water balance, which over a given time period is: P + I + Li = E + Lo + D + �S (Equation 5.1) where P is precipitation, I is irrigation, Li is the lateral inflow of water, E is evapotranspiration, Lo is surface and subsurface lateral outflow of water, D is the drainage to the watertable and �S is the change in soil moisture storage above the watertable (Indraratna, Tularam & Blunden 2001; Sammut, White & Melville 1996; White et al. 1997). While this equation may give a general overview of sulfidic floodplain hydrology, an understanding of site-specific details is required for effective remediation and management of degraded ASS wetlands. The impact of drainage-induced oxidation and the transfer of products is dependent upon the local geomorphology, the drainage network, watertable dynamics and antecedent conditions (Kinsela & Melville 2004). Furthermore, to effectively manage ASS it is essential to know: i) the depth of the sulfidic layer; ii) watertable dynamics relative to the sulfidic layer and; iii) the effect of drain management and climate on the water balance, watertable dynamics and the export of acid and heavy metals (White et al. 1997). Chapter 5: Groundwater Quality and Hydrodynamics 107 Re-establishing tidal exchange and/or ponding water on the floodplain are common techniques used to rehabilitate degraded floodplain wetlands. Research by Indratana, Tularam and Blunden (2001) on the Shoalhaven floodplain on the south coast of New South Wales (NSW) showed that ponding water via in-drain weirs was effective in the management of ASS. Ponding water within the drain raises the watertable to submerge exposed ASS and prevents further oxidation and acid production. Where the backswamp can also be reflooded and surface water cover maintained, evapotranspiration and drawing of acid salts to the soil surface is also reduced. Restoring tidal exchange may provide an added benefit to the rehabilitation process by neutralising acidic products, as carbonates and bicarbonates naturally present in the estuarine water can act as a buffer (Indraratna, Glamore & Tularam 2002). However, the effectiveness of this process is dependent upon a number of factors including flushing dynamics, groundwater transport and acid production rates (Glamore & Indraratna 2004). The aim of this chapter was to examine the groundwater characteristics of Little Broadwater following rehabilitation. Specific objectives of the study were to: i) investigate spatial and temporal variability of groundwater chemistry; ii) quantify the impact of tidal exchange management and climate on groundwater quality; and iii) determine if there were linkages between groundwater and surface water. 5.2 Materials and methodology 5.2.1 Acid sulfate soils at Little Broadwater The soils at Little Broadwater are typical ASS, characterised by strong acidity, poor drainage, salinity and sodicity (Morand 2001). The average depth of sulfidic sediment at Little Broadwater was 0.39 m, with potential ASS (PASS) below -0.90 m AHD, actual ASS (AASS) present between -0.90 m AHD and -0.40 m AHD, and a newer layer of PASS at -0.19 m AHD to -0.14 m AHD (Wilkinson 2003). It was estimated by Morand (2002) that the soil profile had an existing acidity of 20.77 kg tonne-1 H2SO4. Scalds were present in the north next to Mantons Road and in the south prior to the start, and during the first 6 to 8 months, of the water quality study (Figure 5.1). Iron and salt crystals were present on the soil surface at both scalds prior to the start of the longer-term study in April 2005 with no vegetation cover present. The scalds were submerged from December 2005 onwards. Chapter 5: Groundwater Quality and Hydrodynamics 108 Figure 5.1: Example of scalding at Little Broadwater prior to the start of the study. 5.2.2 Field equipment and monitoring Prior to re-establishing tidal exchange and reflooding the wetland, groundwater was monitored by Clarence River County Council (CRCC, now Clarence Valley Council (CVC)) from mid-June to mid-November 2002, to monitor short-term watertable dynamics and collect data on groundwater quality. A piezometer with a water level logger was located in the southern region of the wetland and another two groundwater sites were located in the northern region (Figure 5.2). All three of these sites are now permanently inundated. Water samples were collected from the southern piezometer in October 2002 and from the northern groundwater sites in November 2002. All samples were analysed for pH, electrical conductivity (EC), chloride (Cl-), sulfate (SO4 2-), soluble iron (Fe) and soluble aluminium (Al) at the Environmental Analysis Laboratory, Southern Cross University, using standard methods according to APHA (1998). Groundwater monitoring post-restoration of tidal exchange was designed to investigate the seasonal and spatial variation of groundwater quality, examine watertable dynamics, and assess interactions between surface water and groundwater. Monitoring of the groundwater sites commenced in April 2005 and continued to February 2007 (referenced as Apr05, Jun 05, etc.), and was conducted in conjunction with surface water monitoring (see Chapter 4). Six piezometer wells were installed around the wetland, with P1 in the north located at the toe of a hill approximately 90 m from the wetland edge, and the remaining five (P2 to P6) located on the edge of the wetland (Figure 5.3). The wetland edge was defined as the high water mark in non-flood periods, which was determined as the boundary between pasture and bare ground Chapter 5: Groundwater Quality and Hydrodynamics 109 from previous inundation (Figure 5.4). The holes were augered to a depth of 3 m (P1-P4) or 5 m (P5 and P6), using a Bunyip auger system fitted with 75 mm auger flights. The depth of the wells was determined by the length of the water level loggers available for the study. The piezometers wells were 50 nb Class 9 pressure PVC pipe at lengths of 4 m (P1-P4) or 6 m (P5 and P6). The bottom metre was drilled on 4 sides with 5 mm holes at 50 mm spacing, and a PVC cap was glued to the bottom of the pipe (Figure 5.5). The pipe was then placed in the well and the hole was backfilled with 7 mm blue metal aggregate up to 600 mm from the surface, and then powdered bentonite mixed with some of the soil from the well was used to fill to the surface. A cement cap was placed around the pipe at the ground surface, and was domed on the top to allow water to runoff the top of the well. One metre of pipe was left above the ground to attach a cattle frame (Figure 5.5). The piezometers were capped when not in use, and surveyed to m AHD at a later date. Figure 5.2: Location of Clarence Valley Council (CVC) groundwater sites prior to restoring tidal exchange. Chapter 5: Groundwater Quality and Hydrodynamics 110 Figure 5.3: Location of groundwater and surface water sites at Little Broadwater. Figure 5.4: Example of location of wetland edge piezometers at the boundary between pasture and the high water mark (bare ground): (a) facing away from the wetland and (b) facing the wetland. Chapter 5: Groundwater Quality and Hydrodynamics 111 Figure 5.5: Design of the piezometer wells installed around Little Broadwater. Continuous monitoring of watertable depth was conducted with Odyssey Capacitive Water Level Recorders, which were installed in each piezometer. The loggers were programmed to record at hourly intervals and were powered by two 3.6V Lithium batteries. Water elevation data were downloaded every 2 months, and the probes were recalibrated every 2 to 4 months. The groundwater elevation was adjusted to m AHD using the surveyed elevations of the piezometers. Groundwater samples were obtained bi-monthly for physical and chemical analysis. The piezometers were initially pumped out at least 12 hours prior to sampling to ensure that stagnant ‘old’ water was not being sampled. An Amazon inline 12V submersible pump attached to 7 m flexible PVC pipe was used to pump groundwater to the surface. Sampling water was pumped into a jug that was rinsed three times with the sample water and then filled. This water was then used to rinse two 125 mL polyethylene bottles three times and fill the bottles. All air was excluded from the bottles before sealing and storing samples. After Chapter 5: Groundwater Quality and Hydrodynamics 112 collection, samples were stored at 4ºC as soon as possible, and frozen within 48 hours of collection. Samples remained frozen until laboratory analysis of soluble basic cations (calcium (Ca), magnesium (Mg), sodium (Na) and potassium (K)), acidic cations (Al, Fe and manganese (Mn)) and anions (Cl-, SO4 2-). Groundwater pH, EC and temperature were measured from the water remaining in the jug. The procedures for field measurement and laboratory analysis of groundwater were the same as those for surface water described in Chapter 4. Surface water data presented in this Chapter is a subset of the data presented previously in Chapter 4. Rainfall data was recorded at Lawrence Post Office, approximately 2 km east of the study site and sourced from Australian Bureau of Meteorology. Evaporation data was sourced from the Grafton Agricultural Research Station, approximately 18 km southwest of the study site. Refer to Chapter 2 for further details regarding climatic data. 5.2.3 Hydraulic conductivity Hydraulic conductivity (Ksat) of the shallow soil horizons was determined using the shallow pit method (Johnston & Slavich 2003). The technique specifies pits of a maximum of 0.6 m deep and an area of between 0.3 x 0.3 m and 0.5 x 0.5 m. The water level in the pit was left to equilibrate with the surrounding groundwater before an average depth measurement was taken. Water was then rapidly bailed from the pit and the recharge rate recorded, giving the Ksat of the soil. The test was repeated at least once in each pit. A spreadsheet with pre-filled formulas is provided by Johnston and Slavich (2003) and only requires the user to enter the dimensions of the pit, units of time for each measurement and raw water depth measurements. The calculated data is then plotted on a graph which indicates approximate Ksat ranges of low (< 1.5 m day -1), moderate (> 1.5 and < 15 m day-1), high (> 15 m day-1) and extreme (~ 100 m day-1). 5.2.4 Wetland bathymetry A bathymetric map of Little Broadwater was developed from a combination of surveys. The majority of the wetland was surveyed to m AHD by A. Fletcher and Associates (2002) prior to reflooding the wetland, however, areas which were inundated at the time were not surveyed. An additional survey of the permanently inundated areas was conducted in 2007 by Chapter 5: Groundwater Quality and Hydrodynamics 113 Adam Smith and the author. Water depth measurements were recorded along a meandering traverse of the wetland and locations recorded with a GPS. The water depth was related back to the known drain water depth and calculated as m AHD. The data from the two surveys was compiled by Adam Smith and Aleks Maric (Department of Environment and Climate Change, Parks and Wildlife Group). The resultant bathymetry map only extended to 0.24 m AHD, as this was the water height at the time of the 2007 survey. 5.3 Results 5.3.1 Pre-rehabilitation watertable dynamics and groundwater quality Groundwater level and water quality data collected by CVC in 2002 are presented here to illustrate the groundwater quality and watertable dynamics prior to reflooding the wetland. The watertable was below the ASS layers for the majority of the monitoring period (Figure 5.6a). The start of this monitoring period coincided with an extended dry phase, where no rainfall was recorded for 44 consecutive days (see Chapter 2, Table 2.3). When the watertable position was very low (< -0.6 m AHD) only rainfall of more than 25 mm in 24- hours was sufficient to cause the watertable position to rise above the ASS layers. However, it often fell below the potential ASS (PASS) layers within a week. When the watertable was higher, as in October 2002, rainfall events of 5-10 mm caused a rise in watertable position to above the lower PASS layer (Figure 5.6a, b). The longest period that the newer PASS and actual ASS (AASS) layers were submerged was when there was consistent rainfall at the end of August 2007, with 61.8 mm of rainfall recorded over a period of a week. The watertable rapidly increased in response to the majority of rainfall events, although the degree to which watertable position changed was different for rainfall events of similar sizes. This may have been due to varied short-term rainfall patterns, i.e. differences in intensity and duration, or the moisture content of soils in the surrounding catchment. For example, in late September 2002 there was 7.4 mm of rainfall in 24 hours which resulted in a 0.36 m increase in watertable height, however, an 8.8 mm event in mid-November 2002 did not cause any response in watertable height (Figure 5.6a, b). Another feature of the groundwater hydrograph was evidence of tidal oscillations from late October (after a high rainfall event of 33 mm in 24 hours) to mid-November 2002 (Figure 5.6a). There are no records to indicate that the floodgates were open at this time, however, the Drain Management Plan for Little Broadwater allowed landholders to open the floodgates when EC in the river was 2.2 dS m-1 or less, or there was localised flooding Chapter 5: Groundwater Quality and Hydrodynamics 114 (Clarence River County Council 2001). Drain water EC at the time of the tidal oscillations was less than 2.2 dS m-1 and the high rainfall just prior to this may have cause localised flooding, suggesting that the floodgates may have been open to drain water from the backswamp, subjecting the wetland to tidal influence. Figure 5.6: (a) Groundwater hydrograph for the piezometer installed by CVC, with (b) daily rainfall indicated. The average depth of potential acid sulfate soil (PASS) and actual acid sulfate soil (AASS) layers are indicated. The groundwater was saline at all three sample sites prior to restoring tidal exchange, with the average EC ranging from 16.1 dS m-1 to 19.8 dS m-1 (Table 5.1). The low Cl-:SO4 2- and high Fe concentrations indicated that there was sulfidic material present at the sites, however, pH Chapter 5: Groundwater Quality and Hydrodynamics 115 values greater than 5 suggest that the saline groundwater was acting as a buffering agent. Aluminium concentrations were relatively low compared to Fe (Table 5.1). Table 5.1: Average water quality of groundwater prior to restoring tidal exchange. The piezometer was sampled on 22 October 2002, and the Groundwater A and B sites were sampled on 14 November 2002. Parameter Piezometer Groundwater A Groundwater B pH 6.13 5.51 5.31 EC (dS m-1) 17.8 19.8 16.1 Cl- (mg L-1) 7572 6419 4893 SO4 2- (mg L-1) 2695 4835 3800 Cl-:SO4 2- 2.8 1.3 1.3 Fe (mg L-1) 18.9 87.9 204.5 Al (mg L-1) <0.001 0.41 0.51 5.3.2 Watertable dynamics post-restored tidal exchange The watertable at Little Broadwater rose quickly after rainfall events (Figure 5.7). Away from the edge of the wetland the watertable generally fell rapidly, often within days of a rainfall event, whereas at the edge of the wetland the watertable decreased slowly over a period of weeks. Watertable fluctuations at the edge of the wetland were smaller in magnitude than that recorded at P1 away from the wetland (Figure 5.7a). The watertable was generally lower during 2005 than in 2006 and was frequently below the upper boundary of the sulfidic sediments (Figure 5.7a). The watertable was lowest at P1 during 2005 and the beginning of 2006, and the surface water level in the drain was often higher than the watertable. This indicated that the groundwater was draining away from the wetland during this period. In the latter half of 2006 the watertable away from the wetland was generally higher than that at the wetland edge and the drain water level, indicating a change in groundwater gradient with net flows towards the wetland. The watertable at P4 in the south was frequently the highest and closest to the ground surface, and therefore more susceptible to evaporative loss. These periods of evaporative fluxes were characterised by small and rapid fluctuations during May to June 2006, September 2006 and January to February 2007 – all during periods of very low rainfall (Figure 5.7). C ha pt er 5 : G ro un dw at er Q ua li ty a nd H yd ro dy na m ic s 11 6 Fi gu re 5 .7 : G ro un dw at er d yn am ic s an d th e re sp on se to r ai nf al l: (a ) w at er ta bl e po si tio n at e ac h pi ez om et er a nd d ra in w at er d ep th ; ( b) d ai ly r ai nf al l a nd ev ap or at io n. Chapter 5: Groundwater Quality and Hydrodynamics 117 5.3.3 Groundwater quality post-restored tidal exchange Groundwater salinity was measured as Cl- concentrations rather than EC due to problems with equipment. Firstly, there was a considerable increase in EC at all edge-of-wetland sites between April and June 2005, which coincided with a change of equipment. However, other parameters (pH and temperature) measured with the same equipment did not appear to be affected by the equipment change. Secondly, equipment failure in April 2006 resulted in no EC data for this sample period, which was characterised by very fresh conditions due to high summer rainfall. Statistical analysis of EC with missing data during a fresh period may bias results. Therefore, Cl- was used as an indicator of groundwater salinity as it is the main ionic component of seawater, and is a conservative species that is not a product of ASS oxidation unlike other ions such as K or Na (Sammut, White & Melville 1996). Groundwater pH values away from the edge of the wetland (P1) were neutral throughout the study, ranging between 6.48 and 8.98, and concentrations of Al and Fe were low in comparison to the other groundwater sites (Figure 5.8a, b, c). Aluminium concentrations at P1 were highest during Dec05 and Feb06 (Figure 5.8b), coinciding with increased rainfall after a dry period (Figure 5.7b). Groundwater at P1 consistently had the lowest Cl- concentration with the exception of Feb07 when P2 had the lowest concentration (Figure 5.8e). The median Cl- concentration at P1 was 421 mg L-1 (Table 5.2). P1 also had the lowest median concentrations of Ca, K, Mg, Na and SO4 2- (Table 5.2). Groundwater quality at the edge of the wetland along the eastern side (P2, P3 and P4) was acidic and had high concentrations of soluble metals. Sites P2 and P3 had a median pH of less than 4 (range of 3.30-5.11 and 3.40-4.69 respectively) and the highest concentrations of Al and Fe (Table 5.2). Fluctuations in concentrations of Al, Fe and Mn generally corresponded with changes in groundwater pH, with metal concentrations increasing as pH decreased (Figure 5.8a, b, c, d). The exceptions were in Jun05 and Oct06, the latter of which showed a large increase in metal concentrations while pH also increased at P2 and remained stable at P3. There was no consistent pattern of changes in groundwater pH or metal concentrations due to rainfall and/or changes in watertable position at these sites. pH decreased and concentrations of Al, Fe and Mn increased between Oct05 and Dec05 (Figure 5.8 a, b, c, d) and corresponded with an increase in watertable due to high rainfall, and was preceded by a dry period during which the watertable decreased and was below the upper sulfidic layer (Figure 5.7). However, groundwater pH had increased by Feb06 and metal concentrations decreased, also corresponding with increased rainfall. Manganese concentrations were very Chapter 5: Groundwater Quality and Hydrodynamics 118 high at P2 and P3, and P2 had a Cl-:SO4 2- of less than 3 throughout the study, indicating the presence and recent oxidation of sulfidic material (Figure 5.8d, f). Figure 5.8: Groundwater quality at Little Broadwater, showing (a) pH, (b) Al, (c) Fe, (d) Mn, (e) Cl- and (f) Cl-:SO4 2-. Site P1 was located away from the edge of the wetland in the north, and all other sites were along the edge of the wetland. C ha pt er 5 : G ro un dw at er Q ua li ty a nd H yd ro dy na m ic s 11 9 T ab le 5 .2 : S um m ar y st at is ti cs o f gr ou nd w at er q ua li ty a t L it tl e B ro ad w at er , A pr il 2 00 5 to F eb ru ar y 20 07 . M = m ed ia n. P 1 P 2 P 3 P 4 P 5 P 6 Pa ra m et er M R an ge M R an ge M R an ge M R an ge M R an ge M R an ge pH 7. 23 6. 48 – 8 .9 8 4. 22 3. 30 – 5 .1 1 4. 01 3. 40 – 4 .6 9 6. 25 5. 75 – 6 .8 7 6. 55 5. 87 – 6 .9 9 6. 52 6. 05 – 7 .1 3 W at er te m p. ° C 21 .4 18 .6 – 2 5. 4 20 .0 18 .8 – 2 4. 3 21 .5 19 .1 – 2 5. 8 22 .2 16 .7 – 2 4. 6 21 .9 19 .5 – 2 4. 9 21 .2 19 .0 – 2 4. 2 A l m g L -1 0. 28 0. 00 – 2 .2 8 40 .3 0 10 .1 3 – 20 1. 56 21 .9 3 5. 80 – 2 35 .8 1 0. 18 0. 00 – 1 .8 9 0. 00 0. 00 – 2 .3 6 0. 00 0. 00 – 2 .4 2 C a m g L -1 14 .5 1. 1 – 64 .5 21 9. 1 29 .4 – 2 98 .9 23 5. 0 19 .9 – 3 65 .0 22 7. 4 11 3. 6 – 39 5. 0 18 0. 3 24 .9 – 4 47 .5 28 9. 8 84 .5 – 5 20 .0 C l- m g L -1 42 1 25 8 – 29 89 15 86 30 1 – 33 48 50 75 46 3 – 92 15 55 35 79 3 – 97 99 60 69 91 9 – 98 83 58 72 70 3 – 91 74 Fe m g L -1 0. 04 0. 00 – 1 .6 4 34 7. 51 43 .7 6 – 87 7 49 3. 51 12 8. 24 – 9 76 1. 27 0. 14 – 1 03 0. 09 0. 00 – 0 .4 6 0. 10 0. 05 – 0 .1 7 K m g L -1 3. 4 0. 9 – 12 .1 20 .7 11 .6 – 4 8. 9 28 .6 8. 8 – 38 .1 84 .8 36 .7 – 2 65 .7 25 .1 10 .1 – 4 5. 9 42 .0 13 .9 – 7 9. 7 M g m g L -1 19 .3 8. 7 – 11 0. 6 49 0. 9 16 4. 8 – 95 3. 3 60 9. 1 15 8. 4 – 98 30 80 1. 4 12 1. 3 – 14 60 57 0. 3 21 2. 2 – 11 13 55 4. 9 22 6. 8 – 88 1 M n m g L -1 0. 07 0. 01 – 0 .6 6 20 .2 9 7. 20 – 4 2. 86 29 .9 9 6. 94 – 5 8. 12 25 .9 6 9. 60 – 4 0. 76 2. 89 0. 36 – 1 1. 43 1. 08 0. 44 – 1 1. 24 N a m g L -1 45 9 16 8 – 13 50 13 77 46 0 – 20 58 21 52 12 52 – 3 53 3 34 44 24 78 – 4 11 0 31 23 97 7 – 46 72 30 85 50 7 – 41 98 SO 42 - m g L -1 58 .5 23 .0 – 1 35 13 15 .5 38 1. 4 – 35 44 66 7. 7 32 0. 3 – 21 90 11 38 .7 64 3. 5 – 17 59 36 2. 4 76 .1 – 1 80 5 56 6. 7 18 2. 8 – 82 7 C l- : SO 42 - 7. 9 3. 8 – 10 1. 3 1. 0 0. 6 – 2. 9 7. 0 1. 0 – 17 .0 4. 2 0. 9 – 9. 1 15 .8 2. 3 – 48 .7 10 .5 1. 1 – 27 .0 Chapter 5: Groundwater Quality and Hydrodynamics 120 Groundwater along the southeastern edge of the wetland (P4) had high concentrations of Mn and increased concentrations of Fe from Oct06 to Feb07 (Figure 5.8c, d). However, the groundwater pH was predominantly above 6. The median Cl-:SO4 2- was less than 5 at P4, and increased Fe and Mn concentrations indicated that there was sulfidic material present at the site (Table 5.2). Groundwater salinity along the eastern side of the wetland was highest at P4, with a median Cl- concentration of 5535 mg L-1, while P2 had the freshest groundwater at the edge of the wetland (Table 5.2). Chloride concentrations at all sites were lowest during Feb06 and Apr06 (Figure 5.8e), which was preceded by at least two months of consistent rainfall (Figure 5.7b). There was an initial increase in Cl- concentrations along the eastern edge of the wetland at the start of the summer rainfall, which was preceded by low watertable position and low rainfall (Figure 5.7). The western edge of the wetland had the most saline groundwater, with Cl- concentrations ranging between 919-9883 mg L-1 at P5 (median 6069 mg L-1) and 703-9174 mg L-1 at P6 (median 5872 mg L-1) (Table 5.2). These sites generally had low concentrations of Al, Fe and Mn, although concentrations of Mn increased during late 2006 and early 2007 (Figure 5.8b, c, d) when the watertable was maintained at a high position by regular rainfall (Figure 5.7). Aluminium concentrations were highest during Apr05 and Feb06 (Figure 5.8b), with the latter period corresponding with high rainfall and a high watertable position (Figure 5.7). 5.3.4 Surface water and groundwater interactions There was little evidence of interactions between groundwater and wetland surface water at Little Broadwater over the study period. Management of the tidal exchange structures (see Chapter 4, Figure 4.4) did not produce any observable changes to the watertable (i.e. tidal fluctuations when tidal exchange was promoted) (Figure 5.7). Furthermore, water quality of surface water sites near the groundwater sites did not reflect the quality of groundwater (Figures 5.9 and 5.10). Groundwater around the wetland was more saline than the surface water for the majority of the study period, with the exception of P2 during Jun05, and P3 and P5 during Oct05 (Figure 5.9a, b, d). Groundwater and surface water pH in the north (P2 and Sites 1 and 2, respectively) had a similar pattern of temporal variation (Figure 5.9a1), although the surface water was near-neutral with a median pH of 5.70 and the groundwater median pH was 4.22. Groundwater at P3 was acidic throughout the study period, however, surface water was generally of much better quality (Figure 5.9b1). The only exceptions were Dec05, Jun06 and Aug06 when surface water was slightly more acidic. Concentrations of Al Chapter 5: Groundwater Quality and Hydrodynamics 121 and Fe were considerably higher in the groundwater at these acidic locations than in the surface water (Figure 5.10a, a1, b, b1). Figure 5.9: Comparison between groundwater and nearby surface water for (a-e) Cl- and (a1-e1) pH. Chapter 5: Groundwater Quality and Hydrodynamics 122 Figure 5.10: Comparison of groundwater and surface water concentrations of (a-e) Al and (a1-e1) Fe. Chapter 5: Groundwater Quality and Hydrodynamics 123 Surface water in the southern region of the wetland was often acidic, although groundwater in this area was neutral (Figure 5.9c1). The western area of the wetland also had extended periods from Dec05 where the surface water was of a lower pH than the groundwater (Figure 5.9d1, e1). Metal concentrations in the surface water and groundwater at P4, P5 and P6 were variable, with surface water concentrations often exceeding that in the groundwater (Figure 5.10c, c1, d, d1, e, e1). 5.3.5 Hydraulic conductivity and aquifer features The Ksat of soil at Little Broadwater was in the moderate range, between 1.5 and 15 m day -1 (Figure 5.11). Conglomerate rock and sandstone was encountered at a depth of approximately 0.3 m near P3 on the eastern side of the wetland, and at a depth of approximately 0.4 m on the western side approximately 150 m south of P6. Figure 5.11: Hydraulic conductivity of groundwater at Little Broadwater, with categories indicated. Chapter 5: Groundwater Quality and Hydrodynamics 124 5.3.6 Little Broadwater bathymetry The bathymetric survey of Little Broadwater clearly indicated areas of lower elevation along the eastern side of the wetland, along with a deep pool in the central region (Figure 5.12). Areas where surface water pooled can also be clearly identified in the southeastern arm and southern region. Figure 5.12: Topographical map of Little Broadwater. There are a number of areas with lower elevation (� -0.1 m AHD) where surface water ponds. Chapter 5: Groundwater Quality and Hydrodynamics 125 5.4 Discussion Restoring tidal exchange and ponding water within the wetland was successful in raising the watertable above the ASS layers, although there was limited success in reducing further oxidation and new acid production along the eastern side of the wetland. Concentrations of Fe and Al were much higher at the eastern piezometers than at the other sites, and had increased from pre-rehabilitation levels. Groundwater in the north also had a considerable increase in Al and Fe concentrations after reflooding of the wetland. A possible explanation for this is that iron in the form of Fe3+ can continue the oxidation of pyrite directly in the absence of oxygen when in solution and under acidic conditions, i.e. pH less than 3.5 (Nordstrum 1982; Ward, Sullivan & Bush 2004). Bacteria such as Thiobacillus ferrooxidans or Ferrobacillus ferrooxidans may have also been present as they strongly catalyse oxidation reactions when the pH is between 1 and 4, and maintain the supply of Fe3+ and thus pyrite oxidation and acid production (Nordstrum 1982; White et al. 1997). In contrast to the eastern groundwater, the southern region groundwater generally had a pH above 6 and low concentrations of Al and Fe, similar to pre-rehabilitation conditions. The western piezometers also showed that the groundwater was much more saline in this area but there was little evidence of ASS oxidation. The spatial variability of groundwater can often be attributed to the spatial heterogeneity of acid sulfate soils (Dent 1986) and differences in microtopography, vegetation cover and watertable position (Husson et al. 2000). The majority of sites at Little Broadwater had similar vegetation cover, and the watertable at all sites was above the average depth of ASS for most of the study. Therefore, the distribution and previous oxidation of acid sulfate soils appeared to be the dominant control on groundwater variability around the wetland. Similarly, groundwater salinity is due to the spatial extent of estuarine sediments at the site (Walker 1972), microtopography and past and present inundation patterns. The soils at Little Broadwater are Sulfuric Oxyaquic Hydrosols which are characterised by strong salinity and acidity and low permeability (Morand 2001, 2002). Inundation patterns have changed over time due to drainage, restriction of tidal exchange and changes in microtopography. Parish maps from 1898 and 1905 indicate that the northern section of Little Broadwater was an open reedy swamp, subject to inundation on spring high tides. As a result of this regular saline inundation, the soils were probably saline with seasonal rainfall flushing salts from the sediments. However, drainage of the wetland and the installation of one-way floodgates prevented tidal inundation and therefore much of the soil salt content has been removed by Chapter 5: Groundwater Quality and Hydrodynamics 126 rainfall. The current management of restricted tidal exchange does not allow saline water to reach the upper section of Little Broadwater and as a result the groundwater is less saline than in other areas of the wetland. Drainage of wetlands also results in shrinkage and cracking of the estuarine clays (White et al. 1997), which changes the microtopography and therefore inundation patterns. It is unknown if the depressions along the eastern side of Little Broadwater are consistent with the original topography of the wetland (prior to drainage), or if these depressions have increased in size or number due to shrinkage of the sediments after initial drainage. When the wetland was reflooded with estuarine water, saline water may have intruded further into the backswamp than prior to drainage and also ponded in areas due to changes in topography. As a result, soil and groundwater salinity may have increased in these regions and contributed to scald formation and limited vegetation growth (Lin et al. 2001; Rosicky et al. 2006). Areas of ponded water are highly susceptible to scald formation during dry periods as limited flushing coupled with evaporation can concentrate salts and lead to the death of vegetation. The scalded sediment surface in the southern region prior to the start and during the first eight months of the study has been attributed to ponded saline water, which evaporated and concentrated salts on the sediment surface, causing the vegetation to die off (Clay & Hirst 2004). The bare surface increases the effect of evaporation on groundwater, and capillary action then transports acidic salts to the upper soil horizon (Green et al. 2006; Rosicky et al. 2006), thus exacerbating the existing scald. Evapotranspiration fluxes were evident in the watertable hydrograph for southern Little Broadwater, and iron and salt crusts (Figure 5.1) were present on the sediment surface, indicating that this process was occurring. When the groundwater is recharged by rainfall the acidic salts are dissolved and often flushed into the surface water (Blunden & Indraratna 2001), rapidly increasing surface water acidity and metal concentrations. This is termed the ‘first flush’ effect (Drever 1997; Callinan et al. 1993 cited in Green et al. 2006: 357) and occurred at Little Broadwater in December 2005. Aluminium and Fe concentrations in the groundwater along the eastern edge increased rapidly during December 2005 after an extended dry period over October 2005, and then quickly decreased by February 2006. Surface water conditions exhibited a similar trend due to oxidation products in the upper soil profile which were mobilised during the summer rainfall. Acidic conditions improved in some areas of the wetland due to dilution and flushing, although poor surface water circulation maintained acidic conditions in the southeast for an extended period of time. Chapter 5: Groundwater Quality and Hydrodynamics 127 Temporal variation in groundwater quality was determined by the antecedent conditions, in particular the watertable position, and the frequency and intensity of rainfall events. These factors contribute to the ‘first flush’ effect mentioned above. The watertable position is also influenced by the frequency, intensity and amount of rainfall and evapotranspiration (Green et al. 2006). As a result, groundwater quality responded differently to similar sized rainfall events between years. High rainfall at the end of June 2005 (183 mm over 5 days) decreased groundwater salinity, however, a similar sized event at the end of August 2006 (202 mm over 5 days) increased the groundwater salinity. The length of time between rainfall events may have contributed to the different responses between years, with the June 2005 event preceded by only 11 days with no rainfall in comparison to 4 weeks of no rainfall prior to the August 2006 event. Rainfall events which result in a ‘first flush’ effect can be followed by a decrease in concentrations of salts, metals and acidity due to dilution effects (Green et al. 2006). This occurred over the summer of 2005/2006 after the December 2005 acidification. Concentrations of Al, Fe, Mn, Cl- and acidity in the groundwater decreased rapidly between December 2005 and February 2006, and remained low during April 2006 due to dilution by freshwater inflows from the catchment. There was little evidence to indicate that groundwater quality was directly influencing surface water quality. Along the eastern and western sides of the wetland, where the surrounding hills were close to the wetland (P3 and P5), conglomerate rock and sandstone were found at shallow depths which may have acted as a semi-confining layer to the deeper groundwater, thereby reducing or preventing surface water-groundwater interactions. However, there may have been some interaction between shallow groundwater, above the semi-confining layer, and wetland surface water. This may have been through direct exchange, although the moderate hydraulic conductivity and lack of macropores would have reduced the amount of surface water-shallow groundwater exchange. No tidal forcing was evident in the groundwater hydrographs, with rainfall and evapotranspiration having the greatest impact on watertable dynamics (both shallow and deep) during the study. As a result of the restricted connectivity between the semi-confined groundwater and wetland surface water, there was often a contrast between groundwater and surface water quality. Groundwater was generally of much poorer quality (more acidic and/or saline) than the nearby surface water. The eastern area of the wetland characterised by acidic groundwater and neutral surface water, while in the southern region groundwater was near-neutral and surface water was generally acidic. As previously discussed, acidic surface water conditions in the Chapter 5: Groundwater Quality and Hydrodynamics 128 southeast were initiated by oxidation of surface sulfidic sediments and the upward transport of acid salts through the soil profile via evapotranspiration. Increased rainfall would have mobilised the surface oxidation products into the surface water either via runoff or through shallow groundwater movement above the confined layer. Therefore, while there was some interaction between shallow groundwater and wetland surface water, this was generally only after periods of high rainfall. Prolonged acidification of the surface water was a result of poor flushing of the wetland possibly due to the localised depressions and low tidal forcing, rather than direct or continuous exchange between surface water and groundwater (both shallow and deep). 5.5 Conclusion There was considerable spatial variation in groundwater at Little Broadwater, which could be attributed to the distribution of sulfidic and estuarine sediments, microtopography and watertable position. Temporal variability of groundwater quality and hydrodynamics was strongly attributed to rainfall, watertable position and antecedent conditions, and thus groundwater response was different between sites. Ponding water in the wetland raised the watertable position above the ASS layers, however, short-term manipulation of the tidal exchange structures appeared to have little impact on the watertable position. Hydraulic conductivity was only moderate, and there was no evidence of macropores in the soil profile, thereby limiting tidal forcing in the groundwater. Instead, watertable fluctuations were dependent on rainfall, evapotranspiration and surface water cover (a function of tidal exchange structure management). Surface water quality was not directly influenced by groundwater quality, indicating that these two hydrological components act as semi-closed systems at Little Broadwater. The limited knowledge of factors such as watertable dynamics, groundwater transport, acid production rates and flushing dynamics prior to wetland rehabilitation makes it difficult to quantify the impact of ponding and tidal exchange management on groundwater quality at Little Broadwater. A more detailed study on groundwater dynamics over a longer time period would be necessary to develop a better understanding of the effect of future climatic events, such as drought or flood, and ongoing tidal exchange management on ASS remediation and groundwater dynamics. 129 Chapter 6 Spatial Variation of Water Quality Among Three Coastal Floodplain Wetlands 6.1 Introduction Changes in microtopography and tidal regime strongly influence the hydrology both within and between tidal wetlands (Hughes, Binning & Willgoose 1998). Other factors which affect tidal wetland hydrology and water quality include seasonal patterns of rainfall and evapotranspiration, vegetation (Hughes, Binning & Willgoose 1998) and sediment properties (Stolt et al. 2000). Water quality is further affected by land use within the catchment (Galbraith & Burns 2007) and in the wetland itself. The hydrology of many tidal floodplain wetlands has been affected by the construction of levees, drains and floodgates to protect the land during floods, and reduce or prevent saline inundation. Increasingly, landholders are realising the agricultural benefits of restoring some form of the natural inundation patterns to these floodplain wetlands through managing flood mitigation structures (i.e. floodgates) to promote the growth of highly nutritious wet pasture species (Clay et al. 2007). Although it is known that reflooding wetlands and managing floodgates improves water quality (e.g. Indraratna, Glamore & Tularam 2002; Johnston, Slavich & Hirst 2005b; White et al. 1997), differences in landscape position, extent and type of modification will ultimately determine the outcomes of the management strategy. Thus, wetlands under similar management regimes may have varying degrees of agricultural and ecosystem improvement due to differences in topography, climate, vegetation and the extent of modification. It is therefore important to establish reference wetlands to which the managed wetland can be compared for assessment of improvements in water quality and ecological functioning. Reference sites must be carefully selected to ensure similarities to the managed wetland in terms of hydrogeologic setting, size, geomorphology, tidal range, position in the landscape, adjacent land use and water quality (Neckles et al. 2002). A large number of reference wetlands may be needed to account for the high spatial and temporal variability of natural wetlands (Neckles et al. 2002; Simenstad & Thom 1996). However, the high degree of wetland modification and degradation which has occurred on many coastal floodplains, Chapter 6: Spatial Variation of Water Quality Among Three Coastal Floodplain Wetlands 130 especially along the east coast of Australia, makes it difficult to establish natural reference sites to which comparisons can be made. The lack of reference sites on the Clarence River floodplain increases the difficulty of hypothesising that the hydrological and water quality dynamics observed at Little Broadwater (a wetland which is undergoing rehabilitation) are similar to natural, undisturbed coastal floodplain wetlands. This lack of reference sites is mainly due to the large-scale drainage of floodplain wetlands along the Clarence River. The conceptual model of wetland hydrology presented in Chapter 7, while developed and targeted primarily for application to management of Little Broadwater, may also have potential use for guiding management of other coastal floodplain wetlands. Therefore, a similar investigation of wetland water quality was conducted in a freshwater wetland and a saline pseudo-wetland system (a large drainage network), on the Clarence River Floodplain, to determine if spatio-temporal patterns of water quality were similar. Although there were a number of differences between the wetlands in terms of characteristics and type, both of the ‘reference’ systems were under active floodgate management, providing at minimum a means of comparing hydrological processes between rehabilitating wetlands. Thus, hydrological similarities between coastal floodplain wetlands along the Clarence River could be determined. Accordingly, the aim of this chapter was to investigate similarities and differences between the three coastal floodplain wetlands. The objectives were to: i) examine spatial and temporal variation of surface water and groundwater quality within each wetland; and ii) identify the factors that influenced water quality variability within and between the wetlands. 6.2 Materials and methodology 6.2.1 Site description Short-term water quality monitoring was conducted at Wooloweyah Lagoon and Colletts Swamp, and compared to a sub-set of results from longer-term monitoring at Little Broadwater. All three wetland systems are located on the Clarence River floodplain (Figure 6.1). The study area at Wooloweyah was in the drainage network to the west of Wooloweyah Lagoon, while the second study site was located in the southern region (Block 92) of Colletts Swamp. A description of the Little Broadwater study site is provided in Chapter 2. Chapter 6: Spatial Variation of Water Quality Among Three Coastal Floodplain Wetlands 131 Wooloweyah The Wooloweyah study site was located in the drainage network located on the floodplain to the west of Wooloweyah Lagoon. The following site description provides information about land use, habitat and history of Wooloweyah Lagoon and the sub-catchment. Wooloweyah Lagoon is a shallow tidal barrier estuary lagoon (Roy & Thom 1981), located approximately 12 km from the mouth of the Clarence River (Figure 6.1). The lagoon is connected to the river by three channels (Palmers, Oyster and Micalo Channels), and is bounded by Yuraygir National Park to the east and agriculture (predominantly sugarcane and grazing with an extensive drainage network) to the west. Land use to the north is a mixture of agriculture, aquaculture and urban development. Wooloweyah Lagoon and its associated channels has been identified as an acid sulfate soil (ASS) priority management area (Tulau 1999). The Wooloweyah Lagoon catchment covers an area of 102 km2, with drainage primarily from the south and west of the lagoon through the extensive drainage networks (Woodhouse 2001). The largest is the Taloumbi Drain network, which consists of 20 km of drains and has a total surface area of approximately 142 ha (Foley & White 2007; Woodhouse 2001). The western shore of Wooloweyah Lagoon is bounded by the Taloumbi Ring Drain (and levee) and is approximately 9.3 km long (Williams 2000). Four radial drains (Taloumbi Radial Drains #1- 4) also connect to the ring drain (Figure 6.2). Water is discharged into Palmers Channel at the northern end of the Ring Drain and directly into the lagoon at the confluence of the Ring and Radial Drains via culverts with one-way floodgates. Taloumbi Radial Drain #2 also discharges into Palmers Channel via Marshs and Carrs Drains to the north. A fifth radial drain discharges directly into the lagoon. The extensive drainage network and clearing for sugarcane on the western flats of Wooloweyah Lagoon has had a severe impact on wetlands in this area. It is estimated that there has been a 95% reduction in wetland area which supported common reed and salt couch, and a 20% reduction in area which supported mangroves and salt couch (Woodhouse 2001). The lagoon shore along the Taloumbi Radial Drain supports dwindling stands of mangrove and saltmarsh. Remaining wetlands to the west of the Ring drain are dominated by sedges and grasses, mainly Paspalum distichum (water couch), Eleocharis spp. (spikerush), Lepironia articula, Arex fasicularis (tassel sedge) and Juncus acutus (spiny rush) (Woodhouse 2001). Chapter 6: Spatial Variation of Water Quality Among Three Coastal Floodplain Wetlands 132 These wetlands are subject to seasonal freshwater inundation (Tulau 1999) and are used for grazing. Figure 6.1: Location of the study sites (indicated in red) – Wooloweyah, Little Broadwater and Colletts Swamp. Chapter 6: Spatial Variation of Water Quality Among Three Coastal Floodplain Wetlands 133 Figure 6.2: Location of surface water and groundwater monitoring locations in the drainage network west of Wooloweyah Lagoon (study site referred to as Wooloweyah). Drainage on Palmers Island and Micalo Island is not shown. W1-W10 = surface water sites, WP1 and WP2 = groundwater sites. Chapter 6: Spatial Variation of Water Quality Among Three Coastal Floodplain Wetlands 134 The flood mitigation works in the Wooloweyah Lagoon catchment have had a number of deleterious impacts to water quality, habitat and agriculture. Decreased rates of floodwater discharge, due to the Ring levee and poor discharge capacity of the floodgates, results in prolonged inundation of low lying areas causing a loss of crops and pasture (Woodhouse 2001). Conversely, desiccation during extended dry periods due to the intensive drainage can lead to acid and salt scalds (Woodhouse 2001). Other impacts of the flood mitigation works include: conversion of saltmarsh to freshwater habitat; a decline in water quality due to increased sediment export, acidic water and black water; and, a decline in estuarine waders and waterbird species (Woodhouse 2001). Although there has been an overall decline in waterbird and wader species at Wooloweyah, a number of species listed as vulnerable and endangered in NSW have been recorded in the area (Woodhouse 2001). A range of migratory species listed under the Japan-Australia Migratory Bird Agreement (JAMBA) and the China-Australian Migratory Bird Agreement (CAMBA) have also been recorded around the lagoon (Woodhouse 2001). While estuarine habitat has been destroyed or degraded by the flood mitigation works, the drains still provide a form of habitat for waterbirds, fish and other aquatic animals (Woodhouse 2001). Colletts Swamp Colletts Swamp is located on the Coldstream River, a tributary of the Clarence River, approximately 60 km from the Clarence River mouth (Figure 6.1). Colletts Swamp is a 180 ha fresh meadow/swamp (Department of Natural Resources n.d.) on freehold land and is used for grazing. The wetland is drained primarily by Kenny-Lloyds Drain, which runs approximately east-west through the centre of the wetland (Figure 6.3). The southern portion of the wetland, where the study was conducted, is also drained by a natural creek which branches into a southern and northern arm, approximately 600 m upstream from the creek mouth. The northern branch has been channelized and deepened slightly (S. Murphy 2008 pers. comm., 2 October). The smaller creeks are connected to the wetland via culverts (one in the south, two in the north) with one-way floodgates on either side. The main creek discharges into the Coldstream River via a large culvert which also has one-way floodgates installed at both ends. All references to Colletts Swamp henceforth refer to the southern section (Block 92). Chapter 6: Spatial Variation of Water Quality Among Three Coastal Floodplain Wetlands 135 Figure 6.3: Location of surface water and groundwater monitoring sites at Colletts Swamp. The fence line in the wetland for Block 92 is indicated, as are the location of culverts and floodgates. C1-C7 = surface water sites, CP1-CP5 = groundwater sites. Vegetation in Colletts Swamp was comprised predominantly of Cynodon dactylon (common couch) above the waterline, spikerush at the waterline and water couch and in the shallow water (C. Johns pers. comm. 7th October 2008). Bird surveys were conducted at Colletts Swamp during 2006 and 2007 by A. Smith (pers. comm. 9th October 2008). Twenty-four species were observed over 16 surveys during the 2-year survey period, with the dominant species being Anas superciliosa (pacific black duck), Cygnus atratus (black swan) and Threskiornis spinicollis (straw-necked ibis). Other species Chapter 6: Spatial Variation of Water Quality Among Three Coastal Floodplain Wetlands 136 observed at the wetland included Grus rubicundus (brolga), Gallinago hardwickii (latham’s snipe) and Platalea regia (royal spoonbill). 6.2.2 Climate The study region has a sub-tropical climate with high summer/autumn rainfall and a dry winter/early spring. The average annual rainfall at Grafton is 1017 mm and 1453 mm at Yamba, and the mean annual temperature range is 14-26 °C inland and 15-23 °C on the coast (Australian Bureau of Meteorology 2007). The coldest months are July and August and the warmest months are from December to March. The study was conducted over the driest months of the year, which is from May to November inland (Colletts Swamp and Little Broadwater, representative rainfall recorded at Grafton and Lawrence respectively) and July to November for the coastal zone (Wooloweyah, representative rainfall recorded at Yamba). The coastal zone consistently received the highest rainfall during the study period (Figure 6.4). Rainfall across the floodplain was well below average during May 2006 with a total of 8 mm recorded at Lawrence, 20 mm at Grafton and 67 mm at Yamba (long-term averages are 89 mm, 73 mm and 158 mm, respectively). Total monthly rainfall in June 2006 was also below average at Grafton and Lawrence, however, Yamba had above average rainfall with a total of 191 mm recorded in comparison to the long- term average of 131 mm. Rainfall during August 2006 was well above average across the floodplain, ranging between a total of 197 mm inland to 230 mm on the coast, more than 155 mm above the long-term average. Total monthly rainfall during September 2006 was also above average, with rainfall along the coastal zone nearly twice the long-term monthly average. Figure 6.4: Daily rainfall at Grafton, Lawrence and Yamba over the study period. The highest rainfall was consistently recorded at the coast. Chapter 6: Spatial Variation of Water Quality Among Three Coastal Floodplain Wetlands 137 6.2.3 Field equipment and monitoring Monitoring at Wooloweyah and Colletts Swamp was designed primarily to investigate spatial variation of surface water and groundwater quality. Short-term temporal variation was also examined. Water quality monitoring commenced in late May 2006 and continued to September 2006 at Wooloweyah, and from June to October 2006 at Colletts Swamp. Monitoring was conducted monthly at both sites. The late-May sampling period (24th May 2006) at Wooloweyah is labelled as June henceforth to maintain consistency between sites, as no monitoring was conducted in June at Wooloweyah and the Colletts Swamp June sampling period was conducted one week later on 1st June. Monitoring at each site was conducted within one day of each other for the remaining sample periods. Water quality monitoring at Little Broadwater was conducted bi-monthly from April 2005 to February 2007 as a longer-term study on spatio-temporal variability of hydrology and the effect of tidal exchange management. The results for sample periods June, August and October 2006 are presented in this Chapter. Refer to Chapter 4 (surface water) and Chapter 5 (groundwater) for a description of the monitoring locations and methodology. Water quality monitoring sites at Wooloweyah were selected based on consultation with Clarence Valley Council, as the data collected was also used in the development of the ‘Management Options for the Wooloweyah Ring Drain and Palmers Channel Drainage Systems’ Report (Foley & White 2007). Monitoring sites were chosen to represent specific areas of the drainage network (Figure 6.2). Ten surface water sites were established in the drains (Sites W2-W6 (Ring Drain), W8 (Radial #1), W9 (Reedy Creek) and W10 (Little Reedy Creek)), Palmers Channel (Site W1) and Lagoon (Site W7). Groundwater quality was monitored via two piezometer wells – WP1 was located near Palmers Channel along the Ring Drain and was installed in 2002 for the Palmers Channel Comparative Acid Neutralisation Demonstration (Davison & Wilson 2003); the second piezometer (Site WP2) was installed in July 2006 near Site 8 (Figure 6.2). Groundwater monitoring was restricted to only two sites due to limited access through the drainage network area. Surface water quality monitoring sites at Colletts Swamp were selected to represent drains, natural channel, backswamp and river habitats (Figure 6.3). These sites were highly variable in terms of the presence of water. Site C1 was located in the Coldstream River and Site C2 was in the creek immediately upstream of the main culvert and floodgate structures behind the levee. Sites C3 and C4 were located in the southern arm of the creek, downstream and Chapter 6: Spatial Variation of Water Quality Among Three Coastal Floodplain Wetlands 138 upstream of the small culvert and floodgate structures, respectively. Similarly, Sites C5 and C6 were in the northern arm upstream and downstream of the culverts and floodgates, respectively. The backswamp environment was represented by Site C7. Five piezometer wells (CP1-CP5) were installed at Colletts Swamp in July 2006 and locations were selected to provide a cross-section of groundwater quality from the river to the backswamp, and between the northern and southern arms of the channel (Figure 6.3). The piezometers used in this study were constructed from PVC pipe approximately 1 m in length and 50 mm in diameter. The bottom 0.5 m was drilled on all sides with 10 mm holes at 15 cm spacing and a PVC cap was placed on the bottom of the pipe. Two layers of shade cloth gauze were tied over the drilled section of the pipe to reduce the amount of sediment entering the piezometer. The wells were augered to a depth of approximately 1 m using a hand auger. The pipe was then placed in the well which was then backfilled with sand to within 30 cm of the ground surface. Dirt and clay augered from the well was used to seal the pipe at ground surface. A maximum of 5 cm of pipe was left above the ground and the piezometer was capped when not in use. Watertable depth at CP1 was too deep to collect water samples and therefore no results are presented for this site. In situ measurements of surface water and groundwater quality was conducted with a TPS (model 90-FL Field Lab Analyser), calibrated with standard solutions, to measure pH, salinity and water temperature. Salinity was later converted to electrical conductivity (EC) using a standard conversion factor (DNRW 2007). Techniques for collection, storage and analysis (of aluminium (Al), calcium (Ca), chloride (Cl-), iron (Fe), potassium (K), magnesium (Mg), manganese (Mn), sodium (Na), sulfate (SO4 2-), total nitrogen (TN) and total phosphorus (TP)) of water samples at Wooloweyah and Colletts Swamp were the same as those described in Chapter 4 for Little Broadwater. 6.2.4 Statistical methods Data were initially tested for normality using the Kolmogorov-Smirnov test with SPSS 16.0, however, the majority of datasets were non-normally distributed (skewed or unsymmetrical). This is often a characteristic of water quality data, along with the presence of outliers, seasonal patterns, autocorrelation and dependence on other variables (Helsel & Hirsch 2002). The datasets were unable to be transformed to fit a normal distribution, hence nonparametric summary statistics were used to analyse data. Nonparametric statistics are not biased by the Chapter 6: Spatial Variation of Water Quality Among Three Coastal Floodplain Wetlands 139 inclusion of outliers and therefore these were retained as they reflected the high variability of the environment. Multivariate techniques were used to identify spatial zones of differing water quality (surface water and groundwater) and to determine which parameters accounted for the highest variation between zones and wetlands. Techniques used were hierarchical cluster analysis (CA), non-metric multidimensional scaling (nMDS), analysis of similarity (ANOSIM) and SIMPER analysis. These analyses were performed on all data from Wooloweyah and Collets Swamp as individual wetlands, and median values for each zone identified at Little Broadwater (see Chapter 4, Figure 4.14). Cluster analysis, nMDS, ANOSIM and SIMPER analysis were then applied to the combined datasets of Colletts Swamp, Wooloweyah Lagoon and Little Broadwater to compare and contrast characteristics between the wetland systems. Missing data were replaced with the median of that sample period, as cases with missing data were excluded by the software. Data were then z-score standardised with SPSS 16.0 prior to multivariate analysis. Refer to Chapter 3 (Section 3.2.4) for a detailed description of the standardisation and CA methods. Non-metric multidimensional scaling (nMDS) was performed to graphically illustrate, or map, clustering observed in the CA using ordination methods. Factors were applied to the data to represent location (classified by position in wetland), water type (surface water or groundwater) and wetland. A stress value of � 0.1 was accepted as valid, based on the definitions of stress values provided in Clarke and Warwick (1994). Significant differences (p < 0.05) between location, source or wetland were then determined by analysis of similarities (ANOSIM). The similarity matrix for both nMDS and ANOSIM was created using the Euclidean distance. The contribution of parameters to the difference between groups (location, source or wetland) was examined by SIMPER analysis. Non-metric multidimensional scaling, ANOSIM and SIMPER were performed with Primer 5. 6.3 Results 6.3.1 Wooloweyah Surface water quality was variable over the study period, with a general trend of decreasing salinity and increasing concentrations of metals and nutrients. High rainfall prior to the September sample period (Figure 6.4) corresponded to a rapid decrease in salts (EC) decreased from 12.8 dS m-1 to 0.7 dS m-1) and a sharp increase in metal concentrations (Fe increased more than four-fold and Al more than six-fold) (Appendix D, Table D1). Chapter 6: Spatial Variation of Water Quality Among Three Coastal Floodplain Wetlands 140 Electrical conductivity of surface water was highest on average at Sites W1 and W7 (median 37.5 dS m-1 and 34.0 dS m-1, respectively), although it was highly variable (Figure 6.5a). Median values and ranges of EC were similar between all Ring Drain Sites (W2-W6) and Sites W9 and W10. The least variability recorded was at Site W8 which had a median EC of 4.8 dS m-1. The concentrations of other salts followed a similar spatial pattern, although there was considerable variation in the range of Cl- concentrations at each of the Ring Drain sites (Figure 6.5b-g). Minimum Cl-:SO4 2- were above 2 and medians were above 7 (Figure 6.5h). Groundwater EC and salt concentrations were much lower at Site WP1 than WP2 (Figure 6.5a-g). Median EC at WP1 was 5.1 dS m-1 (range 2.3-9.6 dS m-1), compared to WP2 which had a median of 24.6 dS m-1 and a range of 23.4-28.5 dS m-1 (Figure 6.5a). Calcium and SO4 2- concentrations at WP2 also exceeded the concentrations recorded at all surface water sites (Figure 6.5c, g). Overall, salt concentrations at WP2 were comparable with surface water Sites W1 and W7. Groundwater Cl-:SO4 2- were above 5 throughout the study (Figure 6.5h). Surface water pH was similar across all sites at Wooloweyah, ranging between 6.5 (W1 August) and 9.1 (W4 July) (Figure 6.6a). Concentrations of Al and Fe were generally low with maximum concentrations at all sites recorded during September (Figure 6.6b, c). Aluminium concentrations were variable at all sites, as were Fe and Mn with the exception of Sites W1 and W9 (Figure 6.6b, c, d). Groundwater pH was lowest at WP2 with a median of 5.6, although it did not vary much over the study period (Figure 6.6a). Iron and Mn concentrations were also highest at this site (medians of 16.57 mg Fe L-1 and 2.29 mg Mn L-1) (Figure 6.6c, d). Median Al concentrations were also highest at WP2 (Figure 6.6b). The median pH at WP1 was lower than the surface water sites, although concentrations of Al and Fe were comparable (Figure 6.6a, b, c). However, Mn concentrations at WP1 were much higher than surface water concentrations (Figure 6.6d). The Cl-:SO4 2- was above 5 at both groundwater sites throughout the study (Figure 6.5h). Chapter 6: Spatial Variation of Water Quality Among Three Coastal Floodplain Wetlands 141 Figure 6.5: Wooloweyah (a) EC and concentrations of (b) Cl-, (c) Ca, (d) K, (e) Mg, (f) Na, (g) SO4 2- and (h) Cl-:SO4 2- for surface water (W1-W10) and groundwater (WP1 and WP2). Filled black circles represent the median, bars indicate the maximum and minimum values. n = 4 W1-W10 and WP1, n = 3 WP2. Chapter 6: Spatial Variation of Water Quality Among Three Coastal Floodplain Wetlands 142 Figure 6.6: Surface water (W1-W10) and groundwater (WP1 and WP2) (a) pH, (b) Al, (c) Fe and (d) Mn at Wooloweyah. Filled black circles represent the median, bars indicate the maximum and minimum values. Iron and Mn are on a log scale due to the large difference in concentrations between WP2 and surface water sites. n = 4 W1-W10 and WP1, n = 3 WP2. There was little variability of surface water median TN concentrations between sites at Wooloweyah (Figure 6.7a). In contrast, median TP concentrations were highest at Sites W8- W10 while other sites had lower concentrations (Figure 6.7b). However, the median TP concentration of groundwater at Site WP1 was the highest for all surface water and groundwater sites. Total nitrogen concentrations in the groundwater were higher than in the surface water, with WP1 having the highest median (3.87 mg L-1) and largest range (2.85- 8.04 mg L-1) (Figure 6.7a). Cluster analysis and nMDS plots showed a grouping of surface water samples from Palmers Channel (W1) and Lagoon (W7) for all months except September (Figure 6.8). All September samples from Wooloweyah, with the exception of W1 (Palmers Channel) were clustered together. Consequently, there was no significant differences between sites at a p < 0.05 significance level, although Palmers Channel and Lagoon were significantly different to the Ring Drain sites (W2-W6) at p < 0.1 (p = 0.068 and 0.098, respectively). This was due to Chapter 6: Spatial Variation of Water Quality Among Three Coastal Floodplain Wetlands 143 differences in salts (43.9% of the variance for Palmers Channel, 56.29% of the variance for Lagoon; Table 6.1). Similarly, salts were the primary parameters accounting for the significant difference (p = 0.057) between Palmers Channel and Radial #1 (W8) (Table 6.1). Figure 6.7: Wooloweyah (a) total nitrogen and (b) total phosphorus concentrations of surface water (W1-W10) and groundwater (WP1-WP2). Filled black circles represent the median, bars indicate the maximum and minimum values. n = 4 W1-W10 and WP1, n = 3 WP2. Figure 6.8: nMDS plot for Wooloweyah surface water sites. Samples from Palmers Channel (W1) and Lagoon (W7) were clustered together, with the exception of September samples. Chapter 6: Spatial Variation of Water Quality Among Three Coastal Floodplain Wetlands 144 Table 6.1: Water quality parameters and the percent contribution accounting for the majority of dissimilarity between spatial zones of Wooloweyah surface water. Palmers Channel (W1) and Ring Drain (W2-W6) Palmers Channel (W1) and Radial #1 (W8) Lagoon (W7) and Ring Drain (W2-W6) Variable % Con- tribution Cumula- tive % Variable % Con- tribution Cumula- tive % Variable % Con- tribution Cumula- tive % K 9.2 9.2 K 9.5 9.5 Ca 8.4 8.4 Ca 9.0 18.2 Mg 9.4 18.9 Na 8.2 16.6 Mg 9.0 27.2 Ca 9.3 28.2 Mg 8.1 24.8 Cl- 8.5 35.7 EC 9.0 37.2 Cl- 8.0 32.8 EC 8.2 43.9 Cl- 8.2 45.4 K 8.0 40.8 pH 7.5 51.4 Na 7.2 52.6 SO4 2- 7.8 48.6 EC 7.6 56.2 Cluster analysis, nMDS plots (Figure 6.9) and ANOSIM indicated that there was a significant difference (p = 0.029) between groundwater at Sites WP1 and WP2. This was largely due to differences in Cl-, EC and pH (each contributing more than 8% and totalling 25.1%), with the next 25% of variance explained by Ca, SO4 2-, Mg and Mn. There was also a significant difference (p = 0.001) between surface water and groundwater (Figure 6.9). The dominant variables which accounted for the difference were Mn and TN (both 10.8%), pH (8.5%), TP (8.4%), SO4 2- (7.8%) and Ca (7.7%). Chapter 6: Spatial Variation of Water Quality Among Three Coastal Floodplain Wetlands 145 Figure 6.9: nMDS plot of Wooloweyah surface water and groundwater sites. Surface water quality was significantly different (p = 0.001) to groundwater quality, and there was a significant difference (p = 0.029) between groundwater sites. 6.3.2 Colletts Swamp Water quality averages for Colletts Swamp need to be considered carefully, as there was often no surface water at some sites. As a result, water samples were not collected from the backswamp (Site C7) during June, July or August. Site C6 was also only sampled once during the study (September) and Sites C4 and C5 only twice (August and September). Surface water quality was variable over the study period, with a trend of increasing temperature and salt concentrations (Appendix D, Table D2). Colletts Swamp was fresh throughout the study with the median EC ranging from 0.8 dS m-1 during June to 3.8 dS m-1 in September and October. Median pH increased slightly and then decreased by more than 1 pH unit to 5.7 during August and September. However, this was not necessarily due to the influence of a larger sample number as pH at Sites C1-C3 also decreased slightly between July and August. Surface water EC and Cl- concentrations were lowest at Site C1 in the Coldstream River and generally increased as distance from the river increased (Figure 6.10a, b). Site C7 in the backswamp had the highest median EC of 7.2 dS m-1 and also the largest range (4.6- 9.8 dS m-1), while median Cl- and K concentrations were highest at Site C5 (Figure 6.10a, b, d). Calcium, Mg, Na and SO4 2- followed a similar pattern to EC (Figure 6.10c, e-g). Sulfate Chapter 6: Spatial Variation of Water Quality Among Three Coastal Floodplain Wetlands 146 concentrations were highest at Site C7 throughout the study and the Cl-:SO4 2- at this site was less than 2 (Figure 6.10g, h). Median groundwater EC was highest in the backswamp at Site CP5 (9.7 dS m-1) and lowest at CP3 (4.6 dS m-1) (Figure 6.10a). The median EC at Sites CP2 and CP3 was less than that of the surface water at Site C7, although groundwater EC throughout the rest of the wetland was higher than surface water EC. Groundwater Cl- concentrations were also, on average, higher than the surface water concentrations (Figure 6.10b). Similar to surface water, concentrations of Ca, Mg, Na and SO4 2- in the groundwater corresponded to EC patterns (Figure 6.10a, c, e- g). Sulfate concentrations in the groundwater at Site CP5 were similar to those recorded at the nearby surface water Site C7 (Figure 6.10g). Groundwater Cl-:SO4 2- was generally less than 2 at Site CP5 (Figure 6.10h). Surface water pH was highest at Sites C1-C3 in terms of medians and variability, with medians ranging between 6.3 and 7.4 (Figure 6.11a). Sites C5 and C7 had the lowest median pH of 4.8. Concentrations of Al, Fe and Mn were highest at Site C7, although Al and Fe were more variable than Mn (Figure 6.11b-d). Aluminium concentrations ranged between 0.19 mg L-1 and 1.02 mg L-1 at Site C7, and Fe ranged between 0.13 mg L-1 and 3.99 mg L-1. Site C5 also had higher concentrations of Fe and Mn than other surface water sites. The median concentration of Al was 0.00 mg L-1 at the majority of sites (Figure 6.11b). There was little spatial variation of groundwater pH at Colletts Swamp, and temporal variation was similar between sites (Figure 6.11a). The median pH ranged from 5.3 (Sites CP3 and CP5) to 5.6 (Site CP4), and the overall range was 4.9-6.1. Groundwater metal concentrations were highly variable (Figure 6.11b-d). Iron and Mn concentrations were highest at Site CP5 and Al was highest at Site CP3. Iron and Mn concentrations in the groundwater at Site CP5 were much higher than other groundwater and surface water sites (Figure 6.11c, d). On average, concentrations of Al in the groundwater were comparable with that recorded in the surface water, with the exception of surface water Site C7 which had considerably higher Al loads than the groundwater (Figure 6.11b). Chapter 6: Spatial Variation of Water Quality Among Three Coastal Floodplain Wetlands 147 Figure 6.10: Colletts Swamp (a) EC and concentrations of (b) Cl-, (c) Ca, (d) K, (e) Mg, (f) Na, (g) SO4 2- (h) Cl-:SO4 2- for surface water (C1-C7) and groundwater (CP2-CP5). Filled black circles represent the median, bars indicate the maximum and minimum values. n = 5 C1-C3; n = 2 C4, C5, C7; n = 1 C6; n = 4 CP2-CP5. Chapter 6: Spatial Variation of Water Quality Among Three Coastal Floodplain Wetlands 148 Figure 6.11: Surface water (C1-C7) and groundwater (CP2-CP5) (a) pH, (b) Al, (c) Fe and (d) Mn at Colletts Swamp. Filled black circles represent the median, bars indicate the maximum and minimum values. n = 5 C1-C3; n = 2 C4, C5, C7; n = 1 C6; n = 4 CP2-CP5. Concentrations of nutrients in the surface water were highly variable spatially and temporally at Colletts Swamp (Figure 6.12). Nutrient loads were lowest at Sites C5 and C6. Concentrations of TN and TP had high variability at Sites C1, C2 and C3, although this may be misleading due to the larger number of data points for these sites. Site C7 had the highest median TN load (1.27 mg L-1) (Figure 6.12a), while TP was highest on average at Site C2 (0.15 mg L-1) (Figure 6.12b). Groundwater nutrient loads were generally similar to concentrations recorded in the surface water (Figure 6.12). The median concentration of groundwater TN was lowest at Site CP2 and highest at Site CP5 (Figure 6.12a). In contrast, the median TP concentration was lowest at Site CP5 (0.05 mg L-1) and was highest at Site CP2 (0.06 mg L-1) (Figure 6.12b). Cluster analysis, nMDS (Figure 6.13) and ANOSIM indicated that Site C7 (backswamp) was significantly different (p < 0.05 for Sites C1-C4 and p < 0.1 for Sites C5 and C6) to other surface water sites at Colletts Swamp. The primary difference between the backswamp site Chapter 6: Spatial Variation of Water Quality Among Three Coastal Floodplain Wetlands 149 and Sites C1-C4 (Coldstream River, Channel and South Arm) was SO4 2-, which accounted for more than 10% of the difference (Table 6.2). Sulfate also contributed 12.2% to the difference between the backswamp and North Arm sites (C5 and C6), although Al accounted for a greater percentage (13.1%; Table 6.2). Other factors which contributed to the variance between the backswamp and other sites were salts (EC, Ca, Na) and metals (Al, Fe) (Table 6.2). Total nitrogen and DO contributed more than 8% to the variance between the backswamp and North Arm sites. Figure 6.12: Concentrations of (a) TN and (b) TP in surface water (C1-C7) and groundwater (CP2-CP5) at Colletts Swamp. Filled black circles represent the median, bars indicate the maximum and minimum values. n = 5 C1-C3; n = 2 C4, C5, C7; n = 1 C6; n = 4 CP2-CP5. Figure 6.13: nMDS plot of surface water samples for Colletts Swamp. Backswamp sites (C7) were clustered together, as were sites from the North Arm (C5 and C6). C ha pt er 6 : S pa tia l V ar ia tio n of W at er Q ua lit y A m on g T hr ee C oa st al F lo od pl ai n W et la nd s 15 0 T ab le 6 .2 : P ri m ar y fa ct or s co nt ri bu ti ng to th e si gn if ic an t d if fe re nc e be tw ee n th e ba ck sw am p si te ( C 7) a nd o th er s ur fa ce w at er s ite s at C ol le ts S w am p. B ac ks w am p (C 7) & C ol ds tr ea m R iv er ( C 1) B ac ks w am p (C 7) & C ha nn el ( C 2) B ac ks w am p (C 7) & S ou th A rm ( C 3- C 4) B ac ks w am p (C 7) & N or th A rm ( C 5- C 6) V ar ia bl e % C on tr ib ut io n C um ul at iv e % V ar ia bl e % C on tr ib ut io n C um ul at iv e % V ar ia bl e % C on tr ib ut io n C um ul at iv e % V ar ia bl e % C on tr ib ut io n C um ul at iv e % SO 42 - 10 .4 10 .4 S O 42 - 11 .3 11 .3 S O 42 - 11 .8 11 .8 A l 13 .1 13 .1 E C 9. 9 20 .3 A l 10 .2 21 .5 A l 11 .4 23 .2 S O 42 - 12 .2 25 .3 C a 9. 0 29 .3 E C 9. 1 30 .6 F e 10 .2 33 .4 F e 11 .6 36 .9 A l 8. 7 38 .0 F e 9. 0 39 .6 E C 8. 4 41 .8 T N 8. 9 45 .8 M g 8. 6 46 .6 M n 7. 8 47 .4 pH 8. 4 50 .2 D O 8. 3 54 .1 N a 8. 5 55 .1 N a 6. 9 54 .3 Chapter 6: Spatial Variation of Water Quality Among Three Coastal Floodplain Wetlands 151 Groundwater in the backswamp (Site CP5) was significantly different (p = 0.029) to other groundwater sites at Colletts Swamp (Figure 6.14). The principal parameters which contributed to the difference between the backswamp and other groundwater sites were salts, Fe and Mn. Cluster analysis did not group sites according to source (i.e. surface water or groundwater; Figure 6.15), although there was a significant difference between source (p = 0.004) based predominantly on salinity. Instead, sites located in the backswamp were clustered together, with the exception of C7 in September, and were significantly different (p < 0.01) to other site locations. This was also predominantly due to salt concentrations, although Fe also contributed highly to the difference between the backswamp and other locations. Figure 6.14: nMDS plot of groundwater samples for Colletts Swamp. Backswamp samples (CP5) were clustered together and were significantly different (p = 0.029) to other groundwater sites. Chapter 6: Spatial Variation of Water Quality Among Three Coastal Floodplain Wetlands 152 Figure 6.15: nMDS plot of surface water and groundwater sites at Colletts Swamp. The sources of water were significantly different (p = 0.004), although backswamp sites (surface water and groundwater) were clustered together. 6.3.3 Little Broadwater Spatial patterns of surface water and groundwater quality were similar to those presented in Chapters 4 and 5. A brief summary of results from June, August and October 2006 sample periods are presented below (also see Appendix D, Figures D1-D6). Refer to Chapter 4, Figure 4.14 for zones within Little Broadwater. Surface water was freshest in the ULB zone (median EC of 2.5 dS m-1) and most saline in the LLB zone (median EC of 5.3 dS m-1). The MLB had the largest range of EC values over the study period. Chloride, Ca, K, Mg and Na concentrations exhibited a similar pattern to EC. Sulfate concentrations were most variable in the LLB zone and had the highest median, followed by MLB and then ULB. Surface water pH at Little Broadwater was highly variable over the study period, ranging between 2.8 and 11.4. However, the median pH of ULB and MLB was approximately 7, whereas the median of the LLB zone was only 3.9, indicating that this area was generally more acidic than the other zones. Associated with the low pH were higher concentrations of Al, Fe and Mn in LLB (medians of 0.18 mg Al L-1, 0.60 mg Fe L-1 and 6.85 mg Mn L-1). LLB also had the largest range of Al and Mn concentrations, however, MLB had the largest range of Fe (0.01-19.28 mg L-1). ULB had the smallest range of all metal concentrations. Chapter 6: Spatial Variation of Water Quality Among Three Coastal Floodplain Wetlands 153 Nutrient loads were most variable in the ULB zone, ranging between 0.09-13.38 mg TN L-1 and 0.05-0.86 mg TP L-1. The LLB zone also had a large range of TN concentrations and the highest median at 7.09 mg L-1, compared to 2.81 mg L-1 in ULB and 2.27 mg L-1 in MLB. Conversely, the median TP concentration was lowest in LLB (0.07 mg L-1) and highest in the ULB zone (0.25 mg L-1). Groundwater salinity at LBP1 was considerably lower than other groundwater sites at Little Broadwater. Median EC at LBP1 was 3.1 dS m-1, in comparison to 12.0 dS m-1 at LBP2 and greater than 20.0 dS m-1 at the other groundwater sites. Concentrations of salts were also highest at Sites LBP3-LBP6. While EC was highest at Site LBP4, the median Cl- concentration was highest at Site LBP3, and Sites LBP5 and LBP6 also had higher medians than LBP4. Spatial variation of SO4 2- was considerably different to other salts, with LBP4 recording the highest concentrations and elevated concentrations measured at LBP2 and LBP3. With the exception of LBP1, groundwater was generally more saline than the wetland surface water. There was a clear contrast of groundwater pH between areas at Little Broadwater – Site LBP1 ranged from 7.2 to 9.1; LBP2 and LBP3 ranged from 3.7 to 5.1 (medians of 4.2 pH and 4.1 pH, respectively); and LBP4, LBP5 and LBP6 ranged between 5.9 and 6.7. Concentrations of metals were related to pH, with high concentrations of Al, Fe and Mn at Sites LBP2 and LBP3 where the groundwater was acidic. Aluminium concentrations ranged between 13.55 mg L-1 and 64.23 mg L-1, while loads of Fe ranged between 84.73 mg L-1 to 976.13 mg L-1. Concentrations of groundwater TN were much more spatially variable than TP. Median concentrations of TN were lowest at LBP6 (0.43 mg L-1) and highest at LBP2 (8.64 mg L-1), while TP was lowest at LBP3 (0.002 mg L-1) and highest at LBP1 (0.088 mg L-1). 6.3.4 Comparison of three wetlands Surface water Cluster analysis did not group surface water samples according to wetland, most likely due to temporal variation related to rainfall patterns. However, Wooloweyah was significantly different (p < 0.05) to both Colletts Swamp and Little Broadwater. The nMDS plot shows the larger spread of data points for Wooloweyah, whereas Little Broadwater and Colletts Swamp Chapter 6: Spatial Variation of Water Quality Among Three Coastal Floodplain Wetlands 154 were clustered together (Figure 6.16). These two wetlands also had some outliers – Site C7 in September at Colletts Swamp and LLB in June and August for Little Broadwater. Figure 6.16: nMDS plot of surface water samples from Colletts Swamp, Wooloweyah and Little Broadwater. Salts at Wooloweyah accounted for the significant difference of this wetland to the other wetlands, contributing more than 30% to the dissimilarity between Wooloweyah and Colletts Swamp and more than 20% to the dissimilarity between Wooloweyah and Little Broadwater (Table 6.3). Electrical conductivity and concentrations of salts were considerably higher at Wooloweyah than the other wetlands (Figure 6.17). pH also accounted for some of the difference between all three wetland (Table 6.3), primarily due to the different median and range of pH values recorded within each wetland (Figure 6.18a). Concentrations of Mn were more variable at Little Broadwater than at either Wooloweyah or Colletts Swamp (Figure 6.18b) and thus accounted for between 8.3% and 14.3% of the dissimilarity (Table 6.3). Total nitrogen concentrations were highest at Little Broadwater and also most variable (Figure 6.18c), accounting for the highest percentage of dissimilarity between Little Broadwater and Wooloweyah (13.3%), and Little Broadwater and Colletts Swamp (20.2%; Table 6.3). Water temperature and DO were also parameters contributing to the difference between Little Broadwater and Wooloweyah (Table 6.3), although median temperature was similar (Figure 6.18d, e). Chapter 6: Spatial Variation of Water Quality Among Three Coastal Floodplain Wetlands 155 Table 6.3: Parameters accounting for the majority of dissimilarity in surface water between Wooloweyah, Colletts Swamp and Little Broadwater. Wooloweyah & Colletts Swamp Wooloweyah & Little Broadwater Colletts Swamp & Little Broadwater Variable % Con- tribution Cumula- tive % Variable % Con- tribution Cumula- tive % Variable % Con- tribution Cumula- tive % pH 8.7 8.7 TN 13.3 13.3 TN 20.2 20.2 Temp. 8.4 17.1 Mn 8.3 21.6 Mn 14.3 34.5 EC 8.1 25.2 pH 8.0 29.6 Temp. 12.5 47.0 DO 7.8 33.0 EC 7.2 36.8 pH 12.0 59.0 Ca 7.5 40.5 Ca 7.0 43.8 Na 7.5 48.0 Mg 6.8 50.6 Mg 7.4 55.4 Figure 6.17: Surface water salinity parameters which accounted for the majority of difference between Colletts Swamp (CS), Wooloweyah (WL) and Little Broadwater (LB): (a) EC, (b) Ca, (c) Mg and (d) Na. The lower boundary of the box indicates the 25th percentile, the upper boundary the 75th percentile and the line within the box the median. Whiskers above and below the box indicate the 90th and 10th percentile, respectively. Filled circles represent outliers. Chapter 6: Spatial Variation of Water Quality Among Three Coastal Floodplain Wetlands 156 Figure 6.18: Surface water (a) pH, (b) Mn concentrations, (c) TN concentrations, (d) water temperature and (e) DO concentrations at Colletts Swamp (CS), Wooloweyah (WL) and Little Broadwater (LB). The lower boundary of the box indicates the 25th percentile, the upper boundary the 75th percentile and the line within the box the median. Whiskers above and below the box indicate the 90th and 10th percentile, respectively. Filled circles represent outliers. Groundwater comparison Groundwater sites at Colletts Swamp were clustered together by CA and nMDS (Figure 6.19), and were significantly different to both Wooloweyah and Little Broadwater (p = 0.001). Cluster analysis and nMDS also grouped Sites LBP2 and LBP3 at Little Broadwater Chapter 6: Spatial Variation of Water Quality Among Three Coastal Floodplain Wetlands 157 (Figure 6.19). However, Site WP1 at Wooloweyah and Site LBP1 at Little Broadwater were also clustered with Colletts Swamp groundwater by CA. Figure 6.19: nMDS plot of groundwater samples from Colletts Swamp, Wooloweyah and Little Broadwater. Groundwater at Colletts Swamp was significantly different to the other wetlands (p = 0.001), and LB2 and LB3 were clustered separate to other Little Broadwater groundwater sites. Salinity (encompassing EC and contributing salts) was the main difference in groundwater between wetlands. The primary water quality parameters contributing to the difference in groundwater between Colletts Swamp and Wooloweyah were TP, K, Na, water temperature and Cl- (Table 6.4). Total phosphorus concentrations of groundwater were much higher at Wooloweyah than both Colletts Swamp and Little Broadwater (Figure 6.20b), and thus accounted for the majority of difference between these sites (13.9% and 11.6%, respectively; Table 6.4). Colletts Swamp had the lowest median and smallest range of EC and salt concentrations (Figure 6.21). In contrast to Wooloweyah, the main differences in groundwater between Colletts Island and Little Broadwater were EC, Mg, pH, Cl-, SO4 2- and Na (Table 6.4). pH was more variable and had a higher median at Little Broadwater (Figure 6.20a), which was also more saline than Colletts Swamp (Figure 6.21). Chapter 6: Spatial Variation of Water Quality Among Three Coastal Floodplain Wetlands 158 Table 6.4: Parameters accounting for the majority of dissimilarity in groundwater between Wooloweyah, Colletts Swamp and Little Broadwater. Wooloweyah & Colletts Swamp Wooloweyah & Little Broadwater Colletts Swamp & Little Broadwater Variable % Con- tribution Cumula- tive % Variable % Con- tribution Cumula- tive % Variable % Con- tribution Cumula- tive % TP 13.9 13.9 TP 11.6 11.6 EC 10.0 10.0 K 11.2 25.1 K 9.1 20.7 Mg 9.0 19.0 Na 10.2 35.23 EC 7.5 28.2 pH 8.9 27.9 Temp. 9.2 44.5 Ca 7.4 35.6 Cl- 8.9 36.8 Cl- 8.6 53.1 Na 7.4 43.0 SO4 2- 8.9 45.7 SO4 2- 7.2 50.2 Na 8.7 54.4 Figure 6.20: Groundwater (a) pH, (b) TN concentrations and (c) water temperature at Colletts Swamp (CS), Wooloweyah (WL) and Little Broadwater (LB). The lower boundary of the box indicates the 25th percentile, the upper boundary the 75th percentile and the line within the box the median. Whiskers above and below the box indicate the 90th and 10th percentile, respectively. Filled circles represent outliers. Chapter 6: Spatial Variation of Water Quality Among Three Coastal Floodplain Wetlands 159 Figure 6.21: Groundwater (a) EC and concentrations of (b) Cl-, (c) Ca, (d) K, (e) Mg, (f) Na and (g) SO4 2- at Colletts Swamp (CS), Wooloweyah (WL) and Little Broadwater (LB). The lower boundary of the box indicates the 25th percentile, the upper boundary the 75th percentile and the line within the box the median. Whiskers above and below the box indicate the 90th and 10th percentile, respectively. Filled circles represent outliers. Chapter 6: Spatial Variation of Water Quality Among Three Coastal Floodplain Wetlands 160 6.4 Discussion Salinity was the primary water quality parameter which exhibited within- and between- wetland spatial and temporal patterns, both for surface water and groundwater. Spatial variation of water quality within each wetland was due to the influence of different factors, although climate was a common cause of short-term temporal variation. Water quality was generally poorer furthest from the main source of exchange (i.e. the creek/river/lagoon) in all three wetland systems. The results indicated that although Wooloweyah and Colletts Swamp were different types of coastal floodplain wetlands to the main study site (Little Broadwater), the water quality patterns observed at Little Broadwater are representative of coastal floodplain wetlands. 6.4.1 Wooloweyah Spatial variation in surface water quality at Wooloweyah was predominantly due to salinity (EC and Ca, Cl-, K, Mg, Na and SO4 2-) and nutrient loads. While Palmers Channel and the lagoon had the highest salinity over the study period, water in the Ring Drain, Little Reedy Creek and Reedy Creek was also generally saline. Increased salinity of the Ring Drain water over the study period was most likely due to leakage of the floodgates and associated structures. Sources of saline leakage into the drain include cracked culverts and floodgates, debris such as sticks and logs jamming the floodgates open (Foley & White 2007), and gates flapping open and closed due to wave action in the lagoon (often observed during the study). Both Reedy Creek and Little Reedy Creek connect to the northern drainage network which discharges into Palmers Channel. The main drains in this area (Marshes Drain and Carrs Drain) are operated by landholders to promote tidal exchange, flush the drains and improve water quality (Foley & White 2007), and therefore may have contributed to the increased salinity of drain water at these locations. Salinity not only defined spatial differences between sites, but also temporal variation within sites. High rainfall in the latter part of the study decreased salinity considerably across the western Wooloweyah Lagoon floodplain. The efficiency of the drainage network to rapidly remove water from the floodplain contributed to the rapid decrease in salinity of water in the drains and also in Palmers Channel and the lagoon itself. Freshwater may accumulate and stagnate in the drains after high rainfall if the water level in the lagoon is elevated, and combined with a lack of riparian vegetation, provides a suitable environment for algae blooms to occur (Foley & White 2007) and aquatic weeds such as Salvinia to establish. Water quality can quickly degrade if these conditions prolong, and a reduction of DO and increased Chapter 6: Spatial Variation of Water Quality Among Three Coastal Floodplain Wetlands 161 concentrations of nutrients were recorded after the large rainfall event in August 2006. If discharged into the lagoon or Palmers Channel, this degraded water can have a considerable impact on these ecosystems. The increased nutrient loads following rainfall suggests that runoff from agriculture may be a major source of nutrients to the drains. Specific sources of nutrients from agriculture may include cattle manure (Lancaster 1990), top-dressing of pasture and fertilisers (Williams 1987) and fertilisers from sugar cane farming (Woodhouse 2001). Total nitrogen concentrations were higher upstream in Reedy Creek and Little Reedy Creek, where the drain water frequently stagnates (Foley & White 2007) and agriculture completely surrounds the drains. In contrast, groundwater nutrient concentrations were lowest at the site (WP2) near Reedy Creek and Little Reedy Creek. The high concentration of nutrients in the groundwater (WP1) near the Ring Drain may have been due to different types of vegetation cover and different inundation patterns between sites. The WP2 groundwater site was located among Casuarina trees, with little grass cover, whereas the WP1 site was open with a thick cover of grass. More frequent inundation of the WP1 site, coupled with the thick grass cover, may have resulted in increased nutrient inputs to the soil (due to a higher occurrence of vegetation decay), and thus may account for the higher concentrations of nutrients in the groundwater. Acid sulfate soils have been identified in the Wooloweyah Lagoon catchment, and much of the western floodplain has been identified as a high risk ASS area (Tulau 1999), especially in the region drained by Little Reedy Creek (Foley & White 2007). The Cl-:SO4 2- at both groundwater sites indicated that there was no sulfidic material present or sulfidic material had not been recently oxidised (Stone, Ahern & Blunden 1998), however, the inland groundwater site was slightly acidic and had very high concentrations of Fe and Mn. Manganese concentrations were also high in the groundwater adjacent to the Ring drain, although pH was neutral. High rainfall during August and September corresponded with a slight increase in concentrations of Al, Fe and Mn and a small decrease in pH, although the Cl-:SO4 2- of drain water still indicated that there was no sulfidic material present. A possible explanation for the contradiction between indicators of ASS (high Cl-:SO4 2- but increased metal concentrations) may be an additional source of Cl-, possibly from surface water seepage into the groundwater. This is supported by Davison and Wilson (2003) who found that at some groundwater sites in the Palmers Channel area the salinity was higher than that in the channel itself. Furthermore, their study determined that drain water pH was not affected by the mildly acidic groundwater, indicating that either there was no interaction between groundwater and surface water, or that Chapter 6: Spatial Variation of Water Quality Among Three Coastal Floodplain Wetlands 162 the groundwater gradient was away from the drain. Lateral seepage of surface drain water into the groundwater due to the groundwater gradient away from the drain accounts for the low concentrations of metals (such as Fe and Mn) in the surface water in comparison to the groundwater observed in the present study. 6.4.2 Colletts Swamp Water quality at Colletts Swamp exhibited high spatial variation due to the different inundation/exchange patterns of monitoring locations. The Coldstream River is a major tributary of the Clarence River and is subject to freshwater and tidal influence. The main, artificially deepened, drainage channel was permanently inundated and subject to infrequent flushing by opening of the floodgates. The south-arm channel had a permanent pond immediately to the east of the small floodgates and an intermittently flooded shallow depression immediately to the west of the structures. In contrast, the north-arm channel was infrequently inundated and was also shallow, while the backswamp was infrequently flushed by rainfall. The lower elevation in the backswamp area resulted in water pooling, the depth of which was mainly determined by patterns of evaporation and rainfall. Little exchange between the backswamp and channels was due to the lack of tidal forcing, channel topography (as indicated by inundation patterns observed during the study) and the culverts with one-way floodgates at either end. The presence of ASS oxidation products and increased salinity in the backswamp region were the dominant factors in spatial variation across Colletts Swamp. Surface water sites in the north-arm and backswamp, along with all groundwater sites, were acidic and had high concentrations of Mn and elevated concentrations of Al and Fe. The backswamp surface water had high concentrations of Al while the backswamp groundwater was characterised by high Fe concentrations. This contrast of metal concentrations between surface water and groundwater in the backswamp may be due to a lack of interaction between water sources, or that the metals were derived from different horizons of the soil profile (i.e. surface water Al from near-surface ASS oxidation and groundwater Fe from deeper layers of ASS). Cl-:SO4 2- ratios indicated the presence of sulfidic material in the backswamp (Stone, Ahern & Blunden 1998), and the low pH suggested that there was little buffering of the water by the saline water. The backswamp area was also characterised as having the highest salinity in Colletts Swamp. High concentrations of Ca, K, Mg and Na, along with the SO4 2-, accounted for the high EC in Chapter 6: Spatial Variation of Water Quality Among Three Coastal Floodplain Wetlands 163 the backswamp. These salts were likely to have been derived from acid hydrolysis of the clay minerals in the sediments (White et al. 1997), as the ratios to Cl- were generally lower in the Coldstream River than the backswamp, especially SO4 2- and Mg. Therefore, an additional source of salts, such as saline sediments, must have been contributing to the high concentrations observed in this region. Evaporation during dry periods transports salts to the sediment surface via capillary action, which are then mobilised into the surface water during rainfall periods, thereby increasing surface water salinity (Green et al. 2006). The lack of regular flushing, and evaporation of surface water, contributes further to increasing surface water salinity. Large rainfall events are required to flush the saline water and concentrated salts in the soil profile from the backswamp area. While salinity was high, the lower pH in the backswamp indicated that the buffering capacity of the water (both surface water and groundwater) had been exceeded. Nutrient loads were variable across the wetland, however, the dominant source of nutrients at Colletts Swamp was from cattle manure, either via direct input in the backswamp area or as runoff into the channels. Limited flushing and therefore stagnation of channel and backswamp water, would have also contributed to the increased concentration of nutrients. While surface water nutrient loads generally increased over the study period (as rainfall increased), groundwater nutrient concentrations generally decreased. This suggests that either dilution of groundwater nutrients was occurring, or they were being transported into the surface water and thus contributing to the increasing nutrient concentrations. 6.4.3 Little Broadwater A longer-term study of surface water and groundwater characteristics of Little Broadwater identified three zones of different surface water quality within the wetland (Chapter 4), and divided groundwater locations into three groups (Chapter 5). The characteristics of surface water zones and groundwater groups observed over June to October 2006 were similar to the longer-term patterns discussed in Chapters 4 and 5. Thus, only a brief discussion of the results is presented for Little Broadwater. Surface water quality was least variable in the MLB zone which was subject to more regular tidal exchange than the ULB and LLB zones, due partly to the topography of the wetland. Flushing of the ULB was mainly through runoff from the upper catchment and was therefore the freshest region. Surface water pH was generally neutral, although acidic water was Chapter 6: Spatial Variation of Water Quality Among Three Coastal Floodplain Wetlands 164 recorded at times due to temporal variation of water cover (and thus temporal patterns of ASS oxidation). In contrast, LLB in the south and southeast was the most saline and acidic zone, and was also characterised by high concentrations of metals and TN. This zone was identified as a ‘sink’ area in the longer-term study, where water pooled and had little exchange with other areas of the wetland. The higher salinity of LLB was attributed to the poor exchange capacity and subsequent concentration of salts by evapotranspiration. The low pH and high metal concentrations of the surface water was also maintained by oxidation of near-surface ASS in the region due to variable surface water cover, and by the lack of exchange and flushing. Groundwater at Little Broadwater had high spatial variation of salts, acidity and metal concentrations. Along the eastern edge of the wetland groundwater was acidic with very high concentrations of Al, Fe and Mn, indicating that oxidation of ASS was still occurring. Groundwater at the edge of the wetland was saline, while the groundwater away from the edge of the wetland neutral with low salinity and metal concentrations. This was due to its position above the saline estuarine sediments and the ASS layers. 6.4.4 A comparison of three wetlands Variation in water quality between the three wetlands was due to a combination of their position on the floodplain (i.e. salinity regime), different wetland types, different types and intensity of modification/degradation and different management regimes. Wooloweyah was considerably different to the other wetland sites as monitoring was conducted in the drains rather than in a floodplain wetland. This was due to the high modification of the western wetland area by intensive drainage networks and sugarcane farms. Only some of the floodgates on the drainage network studied are actively managed, unlike Colletts Swamp and Little Broadwater where all structures are actively managed. Thus there were also differences in the flushing regime of the drains/wetlands. Furthermore, as Wooloweyah is positioned near the mouth of the Clarence River estuary, the remaining wetlands around the Taloumbi Ring Drain are a mixture of saltmarsh, mangroves and wet pasture. In comparison, Colletts Swamp is freshwater wet pasture, while Little Broadwater is an estuarine wetland which is seasonally fresh and has a range of habitats within the system (open water, swamp oak forest and areas dominated by Phragmites or Eleocharis). Chapter 6: Spatial Variation of Water Quality Among Three Coastal Floodplain Wetlands 165 The main water quality parameters accounting for differences between Little Broadwater and Colletts Swamp were similar to those between Little Broadwater and Wooloweyah. Little Broadwater had much higher concentrations of TN and Mn than the other wetlands, while pH was more variable. Water temperature also accounted for a large percentage of the difference between Little Broadwater and Colletts Swamp. Due to the shallow water at Colletts Swamp, water temperature was often quite high. In contrast, salinity accounted for more than 20% of the difference between Little Broadwater and Wooloweyah. Consistent rainfall during winter 2006 in the Sportsmans Creek catchment maintained low salinity levels in the creek and upper estuary, and thus freshwater exchange with Little Broadwater was maintained (refer to Chapter 4). However, the position of Wooloweyah near the mouth of the estuary reduced the effect of increased winter rainfall on salinity levels and therefore it remained saline during the study. Salinity (encompassing EC and basic cations) also contributed highly to the difference in water quality between Wooloweyah and Colletts Swamp, due to the positions of the wetlands on the floodplain (i.e. Wooloweyah near the mouth of the estuary, Colletts Swamp in the tidal freshwater reach of the estuary). However, pH was the main driver of variability (as an individual parameter) between wetlands. Wooloweyah had the highest median pH and was above 6 during the study period. Similarly, the majority of Colletts Swamp had a pH near or above 6, however, the backswamp area was acidic and thus heavily influenced the dissimilarity between the two wetlands. While DO was identified as a difference between Colletts Swamp and Wooloweyah, high temporal variation (hourly and daily) of DO may not indicate long-term patterns; therefore, careful interpretation is required to determine if there was a significant difference of DO between wetlands. The study indicated DO was lower at Colletts Swamp than Wooloweyah and was more variable. Water temperature was higher at Colletts Swamp and may have contributed to the lower DO. However, both wetlands had very low concentrations of less than 5 mg L-1 at times, which may cause stress to fish (ANZECC 2000) if maintained for an extended period of time. Spatial variability at all wetlands was influenced by distance to the creek/river/main body of water, and hence the flushing regime, as both surface water and groundwater quality was poorest furthest from the main exchange source. This was predominantly evident at Little Broadwater and Colletts Swamp, where the surface water in the far backswamp was acidic and salinity and metal concentrations were highest. Similarly, groundwater in these regions was saline and acidic with very high concentrations of metals. These far backswamp areas are Chapter 6: Spatial Variation of Water Quality Among Three Coastal Floodplain Wetlands 166 more susceptible to degradation than other regions of the wetland, as often there is poor tidal exchange or flushing due to microtopography. Surface water tended to pool in the LLB zone at Little Broadwater and the backswamp area of Colletts Swamp, as these regions are depressions in the wetland topography. Fluctuations in the extent of surface water cover and watertable depth in low lying areas was controlled predominantly by patterns of evapotranspiration and rainfall, and may have contributed to poor water quality in the far backswamp areas. Extended dry periods can reduce surface water cover, thereby exposing the shallow groundwater to the effects of evapotranspiration (Rosicky et al. 2006). Watertable drawdown exposes ASS, and oxidation products such as acidic salts are drawn to, and concentrated near, the sediment surface by capillary action (Green et al. 2006; Rosicky et al. 2006). Rainfall mobilises the acid salts into the surface water and groundwater (Blunden & Indraratna 2001), thus contributing to increased salinity, concentrations of metals and low pH. During high rainfall this saline and acidic water may be flushed out of the backswamp area (‘first flush’ effect, discussed in Chapter 5), thus improving water quality. However, small rainfall events are likely to only partially restore surface water cover and the water may remain pooled, hence the cycle of evapotranspiration and salinisation can begin again. 6.5 Conclusion Factors such as position along the estuary, wetland type, degree of modification and different management regimes influenced the water quality at each wetland. Little Broadwater and Colletts Swamp were the most similar wetland sites, although there were some differences between wetlands in terms of nutrients, pH, water temperature and salinity. Within wetland water quality at Little Broadwater and Colletts Swamp was a function of microtopography and management regime. The far backswamp region had the poorest water quality (both surface water and groundwater), as water tended to pond and had limited to no exchange with the surrounding wetland area. These areas of wetlands are more susceptible to degradation and oxidation of ASS. The similarity of spatial water quality variability between different types of coastal floodplain wetlands indicated that the patterns identified in the longer-term study of Little Broadwater were representative of coastal floodplain wetlands along the Clarence River. Although the Wooloweyah site was different to the other two wetlands (due to the fact it was a drainage network acting as a pseudo-wetland), all three wetlands had poorest water quality in the far Chapter 6: Spatial Variation of Water Quality Among Three Coastal Floodplain Wetlands 167 backswamp region. Colletts Swamp was most similar to Little Broadwater due to similarities in terms of connectivity to an adjoining river, and use of backswamp area for grazing only (i.e. these wetlands were relatively undisturbed in comparison to the Wooloweyah site). These results indicate that distance from the main exchange source (e.g. creek, river or lagoon) and the shape of the wetland have a dominant role in determining spatial water quality patterns. While neither of the two comparative wetlands were true reference sites (i.e. undisturbed), they provided a means of comparing hydrological functioning between wetlands of different types. Therefore, common processes of coastal floodplain wetlands could be identified and incorporated into the development of a conceptual model of wetland functioning. 168 Chapter 7 Conceptual Model and Management Implications for a Coastal Floodplain Wetland 7.1 Introduction Wetland modelling is a useful tool for understanding ecosystem functioning, and is effectively a decision support system (DSS) for determining the potential impacts of various management strategies. Due to the variety of wetland types (and thus differences in hydrological regimes), knowledge of individual wetland hydrology is necessary for correct management of the system. Wetland hydrology is a very significant component for understanding wetland functioning as it affects sediments, vegetation and thus habitat (Callaway 2001; Gilvear et al. 1993; Mitsch & Gosselink 2000). Therefore, any model of wetland ecosystem functioning needs to be developed around the hydrological characteristics of the individual wetland. Models of wetland ecosystems can be in two forms – numerical or conceptual. To develop an accurate numerical model of a wetland system, large amounts of complex data collected over a long period of time (i.e. minimum of 5-10 years) are required for model calibration. This can be very expensive and often is impractical (Johnston, Slavich & Hirst 2005b). General numerical models can be generated from short-term studies conducted over a range of conditions, as long as the output is carefully interpreted (Demayo & Steel 1996). Conceptual models are useful for illustrating the ecosystem components of a wetland and understanding linkages between these components (Duever 1988). However, the model contains a number of assumptions due to the simplification of wetland ecosystem processes (Demayo & Steel 1996). The development of a wetland water budget for an individual wetland is often the first step in the generation of the hydrological component of a conceptual model. The water budget is a function of interactions between precipitation, evapotranspiration, runoff, groundwater flows, infiltration and tidal flows (Gilvear et al. 1993). This can be shown in the form of a simple equation (e.g. Mitsch & Gosselink 2000; Thompson & Finlayson 2001), flow diagram (e.g. Beck, Fisher & Bruland 2001; Boumans, Burdick & Dionne 2002), illustration (e.g. Gilvear et al. 1993) or state-rate diagram (e.g. Lee 1993). However, the water balance is often not Chapter 7: Conceptual Model and Management 169 quantified for individual wetlands (Gilvear et al. 1993) and is instead used to illustrate the basic hydrological processes of the wetland. Models (both conceptual and simulation) of wetlands need to take into account the effect of hydrological characteristics, such as circulation and mixing patterns. These parameters influence water chemistry and vegetation within the wetland (Gilvear et al. 1993; Ranasinghe & Pattiaratchi 1999), and in turn factors such as vegetation, water depth and topography affect circulation within a wetland (Hammer & Kadlec 1986). Ecosystem components such as vegetation (type and density) and topography may vary considerably between wetlands; hence the generation of a site-specific model is important for understanding wetland functioning and predicting responses to management strategies. The aim of this chapter was to develop a conceptual model of physical hydrology, water quality dynamics and ecology of an estuarine floodplain wetland (Little Broadwater), to illustrate the complexity of wetland functioning and response to management. The specific objectives were to: i) summarise the hydrological characteristics of the wetland and relate to ecological components; and ii) discuss management options and the potential benefits and/or negative impacts based on past observations and the conceptual model. 7.2 Materials and methods 7.2.1 Model development and software Wetland processes and linkages identified in the conceptual model were derived from the results previously presented in Chapters 3, 4 and 5. Water quality data within the wetland was collected between November 2004 and February 2007, while drain water quality was monitored continuously between March 2002 and February 2007. Short-term water quality monitoring was also conducted at two other wetlands on the Clarence River floodplain, the results from which were used to support the development of the conceptual model for Little Broadwater. Detailed methodology for these studies is provided in Chapters 3, 4, 5 and 6. Whilst in the field observations of vegetation type and fauna at each site were noted. Photographs were also taken at each site to document changes in vegetation and water cover. This information was compared to water quality data and used to aid development of the conceptual model. Chapter 7: Conceptual Model and Management 170 The conceptual model of the functioning of Little Broadwater was developed with STELLA version 9.1 software. This program is designed to aid in the conceptualisation stage of model development and has been utilised by Zhang and Mitsch (2005), Ahn and Mitsch (2002), Spieles and Mitsch (1999) and Lee (1993) to model wetland ecosystem functioning. The Little Broadwater model defines variables as states and rates, from which state-rate diagrams are constructed (Lee 1993). A hydrological state is where water accumulates and a rate is a change in state over a given time period, i.e. the net effect of hydrological processes (Lee 1993). Since hydrological processes are dependent on states, there is a feedback response between states and rates (Lee 1993). Five symbols were used in the conceptual model of Little Broadwater model: stock, flow, converter, connector and decision process diamond (DPD) (Figure 7.1). A stock is defined as an accumulation and can only be changed by flows which fill and drain stocks, while converters change inputs into outputs and have a number of functions, including defining external inputs, holding values for constants and calculating equations (isee Systems Inc. 2007). The DPD is essentially a black-box which indicates the inputs and outputs of the decision making process but not the detail of the process itself (isee Systems Inc. 2007). Connectors join elements of the model and can be in the form of an action connector (solid line) or an information connector (dashed line) (isee Systems Inc. 2007). Figure 7.1: Symbols used in the conceptual model developed with STELLA software. Chapter 7: Conceptual Model and Management 171 7.3 Little Broadwater hydrology and conceptual model It is important to understand the hydrology of a wetland prior to rehabilitation and/or ongoing management in order to predict positive or negative impacts of strategies. Monitoring of wetland hydrology provides data and information which can be used to develop a conceptual model of wetland functioning. Observations in the field can also be used to relate hydrology to spatio-temporal variability of habitat and thus wildlife usage. A summary of the hydrological characteristics of Little Broadwater is given below to provide background information for the development of the conceptual model. 7.3.1 Little Broadwater hydrological characteristics Climate affects wetland hydrology and thus water quality, vegetation, habitat and fauna (Mitsch & Gosselink 2000). Therefore, it was important to have a basic understanding of climatic variability at Little Broadwater and the effect this had during the study. Although evaporation rates were highest during summer, corresponding high rainfall decreased the effect of evaporation. During the winter and spring months, when evaporation exceeded precipitation, the wetland was more susceptible to desiccation. Further compounding this problem was increased salinity in Sportsmans Creek during winter and spring, thus reducing the amount of exchange and maintenance of surface water at the required level of 0.12 m AHD, as set out in the drain management plan (Clarence Valley Council 2008) (discussed further in Section 7.4). The seasonal and inter-annual variability of climate during the study period had a strong influence on tidal exchange, water cover and water quality at Little Broadwater. The reduction in severity and duration of acidic discharge at Little Broadwater did not appear to be primarily due to saline water buffering the drain water. Comparisons to a drained wetland further upstream of Sportsmans Creek indicated that EC was similar between both sites, although the upstream site continued to discharge acidic water (see Chapter 4, Figure 4.6). The reduction in acid discharge at Little Broadwater after restoring tidal exchange could instead be attributed to dilution and regular flushing, rather than tidal buffering. Although acidic discharge still occurred from Little Broadwater (due to desiccation because of low rainfall and management), the duration and severity (i.e. pH value) had greatly improved compared to pre-rehabilitation events due to the increased exchange and dilution. However, increased surface water cover and a raised watertable did not prevent the production of acidic deep groundwater and high concentrations of iron along the eastern boundary of the wetland. Chapter 7: Conceptual Model and Management 172 Results of the groundwater study indicated that the groundwater at Little Broadwater could be divided into two components – a deep semi-confined system and a shallow unconfined system (Figure 7.2). Piezometers were drilled through the semi-confining layer and thus groundwater quality was reflective of the deeper groundwater. There appeared to be little interaction between the deeper groundwater and surface water, as the surface water quality did not reflect the poor groundwater quality at the edge-of-wetland sites. This was especially the case along the eastern edge of the wetland where groundwater was consistently very acidic (average pH of 4) with extremely high concentration of aluminium and iron. However, the nearby surface water was predominantly neutral with low metal concentrations (see Chapter 5, Figures 5.8 and 5.9). In contrast, groundwater in the southern region was predominantly neutral while the surface water was acidic throughout the majority of the study. Figure 7.2: Cross-section of Little Broadwater, illustrating the interaction of the main hydrological components of the system. Interaction between groundwater and surface water was restricted to the shallow groundwater system. This interaction was primarily through evapotranspiration and capillary rise of acidic salts through the soil profile (Figure 7.2), which were then flushed into the surface water during rainfall events. Oxidation of a shallow ASS layer in regions of higher elevation, which were subject to more frequent desiccation and oxidation of sediments, may have been an extra source of acids and metals in the surface water through direct exchange between the shallow groundwater and wetland surface water. Chapter 7: Conceptual Model and Management 173 Temporal variation of both surface water and groundwater quality was dependent on the antecedent conditions, and the frequency and intensity of rainfall events. Groundwater quality was particularly dependent upon watertable position, which was a function of rainfall, runoff and infiltration. Surface water cover and quality was a function of rainfall/runoff patterns, watertable depth, and exchange (and therefore creek salinity which was determined by rainfall patterns in the catchment). Zonation of water quality within Little Broadwater was identified and analysis of the water quality variability within each zone provided information about the flushing characteristics of the wetland (see Chapter 4, Figures 4.14, 4.16 and 4.17). The central region of the wetland had variable salinity and returned to neutral conditions quickly after the acidification event in 2005/2006, indicating that this region was flushed regularly by tidal exchange. The northern region of the wetland was the freshest zone and had variable acidity and concentrations of acidic metals. Surface water cover was also variable in this area as the majority of flushing was through direct rainfall or runoff from the upper catchment. The southern and southeastern region of the wetland was the most saline area and was acidic for the majority of the study. Ponding of water (due to the topography) and the high salinity indicated that there was very little exchange and circulation, and flushing was achieved predominantly by direct rainfall. Topography influenced the exchange and circulation patterns within the wetland. While the majority of the lowest-lying areas generally had good water quality (i.e. neutral and lower salinity), the southeastern corner consistently had poor water quality. The shallowness of this area resulted in the sediment being exposed more frequently than other areas of the wetland, and thus oxidation of ASS occurred more often. Evapotranspiration and capillary rise drew acids and salts (from the saline estuarine sediments) to the sediment surface which were dissolved into the surface water when reflooded during rainfall events. Lack of exchange due to a small mouth, compared to the size of the wetland, meant the region was not well flushed and therefore saline and acidic water ponded. A cycle of evaporation, concentrations of surface water salts, capillary-action drawing more salts to the sediment surface, reflooding and poor flushing promoted saline to hyper-saline conditions in the southeastern corner. High nutrient concentrations in these regions could also be attributed to the poor flushing regime. The pattern of high salinity and low pH in the far reaches of the wetland was also observed at Colletts Swamp (see Chapter 6, Figures 6.10 and 6.11), a freshwater wetland on the Clarence River. The backswamp region at Colletts Swamp was poorly flushed due to the topography Chapter 7: Conceptual Model and Management 174 and channelisation of the wetland area. However, as the wetland was located in the upper tidal reaches of the Clarence River floodplain, and therefore was only flushed with freshwater, increased salinity in the backswamp was attributed to saline estuarine sediments (Walker 1972). The groundwater reflected that the backswamp region was more saline than the higher areas of the wetland. Frequent desiccation of the backswamp would draw salts to the sediment surface and thereby increase surface water salinity after rainfall events. Poor flushing concentrated the salts further, as was observed at Little Broadwater. Similarly, the large drainage network to the west of Wooloweyah Lagoon also had poorest water quality furthest from the main source of flushing. This similarity between wetlands indicates that this may be a common problem between coastal floodplain wetlands. Frequent or improved flushing is part of the solution to improving water quality of these far backswamp areas. 7.3.2 Conceptual model of Little Broadwater A conceptual model of Little Broadwater ecosystem functioning, based on hydrology and water quality dynamics, was developed from the results of water quality monitoring over a 2.5-year period. Components such as vegetation and fauna were also included based on observations in the field. While Little Broadwater was characterised by zones of different water quality due to variations in the dominance of hydrological processes, the conceptual model was developed for the wetland as a whole system. This provides an overview of the functioning and interactions between wetland components. The influence of management strategies is incorporated into the model, to further exemplify the interrelationships between ecosystem components and aid in future management of Little Broadwater. Furthermore, the similarity in spatial variation in water quality at reference wetland sites indicates the applicability of the model to other coastal floodplain wetlands. Although simulation models provide the most accurate predictions for wetland ecosystem response to management regimes, it is not always practical due to financial constraints, the amount and complexity of data (Johnston, Slavich & Hirst 2005b) and the expertise required to develop such a model. Johnston, Slavich and Hirst (2005b) instead suggest the use of a hazard reduction process based on field assessments, topographical analysis, the range of the watertable and climatic information. Essentially, this is how the conceptual model for Little Broadwater has been developed as it is based on simple and easily attainable data, and examination of the linkages indicates the implications of changing management to various ecosystem components. Chapter 7: Conceptual Model and Management 175 The conceptual model of Little Broadwater incorporates an ASS component which has been highly simplified. Modelling of ASS landscapes requires a detailed understanding of the complex processes occurring within the wetland. Wilson, White and Melville (1999) recommend that in attempting to model the discharge of sulfuric acid from sulfidic lowlands, the chemistry of the discharged acid, chemical interactions between water and soil, and acid and estuarine water needs to be known. Chemical modelling of ASS oxidation and acid flux are beyond the scope of this study and detailed studies can be found in Blunden and Indraratna (2001), Johnston, Slavich and Hirst (2004b) and Liu et al. (2007). Surface water model Surface water hydrology is a complex interaction of many components and external factors (Figure 7.3). Water quality is a key component of wetland functioning which is affected by a number of hydrological and external factors, and in turn influences the management decisions which drives the tidal exchange adjustments. A number of feedback responses exist within the model, illustrating the complexity of the system. Surface water volume is predominantly a function of balances between inflows and outflows (the wetland water budget; see Chapter 1, Section 1.3.1). Inputs to the wetland water budget are direct rainfall, runoff (a function of catchment factors such as slope, vegetation, soil type and area; Figure 7.3) and tidal flows, while outputs are evapotranspiration and tidal flows (Equation 7.1). �V = P + R – ET ± T (Equation 7.1) Where: �V = the change in volume of water storage in wetland over a given time P = precipitation R = runoff ET = evapotranspiration T = tidal inflow (+) and outflow (–) All can be expressed in terms of depth or volume. C ha pt er 7 : C on ce pt ua l M od el a nd M an ag em en t 17 6 Fi gu re 7 .3 : C on ce pt ua l m od el o f L it tl e B ro ad w at er . T he m od el h ig hl ig ht s th e co m pl ex it y of p ro ce ss es w it hi n th e w et la nd a nd n um be r of f ac to rs w hi ch in fl ue nc e w at er q ua li ty . S ee F ig ur e 7. 1 fo r ex pl an at io n of s ym bo ls . Chapter 7: Conceptual Model and Management 177 The amount of tidal exchange at Little Broadwater was controlled by management of the in- drain structures and tidal gates, and was also dependent on surface water depth within the wetland (Figure 7.3). The decision-making process for management and adjustment of the tidal exchange structures was based on a number of sources of information – surface water quality within the wetland, water quality in Sportsmans Creek and water depth/volume within the wetland (Figure 7.3). The salinity of water within the wetland and the creek was the primary water quality parameter that influenced the decision for adjustments to tidal exchange structures. Surface water quality within the wetland was in turn affected by tidal exchange, as allowing saline water into the wetland would increase salinity to varying degrees within the wetland (due to spatial differences in exchange and circulation patterns). The surface water depth within the wetland affected tidal exchange in two ways – firstly, if the water depth and creek salinity was low, then tidal exchange was promoted by removing dropboards and opening the tidal gates. Secondly, when water depth within the wetland was high, tidal exchange was restricted due to differences in head pressure, e.g. there was a net outflow of water as the flow of discharge water was stronger than the tidal forcing of inflow water. Furthermore, the inflow period at high tide was restricted due to the higher surface water within the wetland and therefore lower gradient. Observational experiments with fluoroscein dye in the drain at Little Broadwater indicated that the tidal forcing of incoming water from the Creek was restricted to the immediate mouth due to the high water depth within the wetland and the short period of inflow at high tide. The presence of Salvinia within the drain further restricted flow into the wetland and wind was also observed to affect water movement (and thus circulation) considerably. Circulation patterns within Little Broadwater were affected by three main factors: vegetation wind, and topography (Figure 7.3). Vegetation acts as a barrier to flow, with plant stems causing an obstruction at the fine-scale and hummocks acting as coarse-scale obstructions (Hammer & Kadlec 1986). Water is forced to flow through channels that are created by the vegetation hummocks (Figure 7.4) or by waterbirds selectively grazing areas. Vegetation such as Eleocharis equisetina (spikerush) slowed the flow of water through the wetland by acting as a fine-scale obstruction (Figure 7.5a). However, high density cover of spikerush may restrict flow even further (Figure 7.5b). Chapter 7: Conceptual Model and Management 178 Figure 7.4: Casuarina glauca (swamp oak) in the central area of Little Broadwater have created hummocks which act as an obstruction to flow and create channels. Figure 7.5: (a) Low density vegetation will restrict flow less than (b) high density vegetation. Photo dates: (a) February 2006, (b) February 2007. Spatial variation of vegetation species within Little Broadwater, along with topographical features such as channels (either natural or man-made drainage lines) directed flow through the wetland. Maps of surface water salinity and pH (see Chapter 4, Figures 4.8 and 4.9) indicated that exchange flows from Sportsmans Creek were directed primarily through the centre of the wetland, which is dominated by swamp oak. While the swamp oak hummocks obstruct flow, the channels between hummocks are generally open and direct flow through the central portion of the wetland. In contrast, the outer areas of the wetland were dominated by spikerush, which reduced the movement of water through a decrease in tidal forcing from the lack of channelisation. High density cover during extended freshwater periods would have restricted exchange further and increasing the effect of vegetation on water quality. Chapter 7: Conceptual Model and Management 179 The spatial and temporal variation of vegetation is partially dependent on the topography of the wetland (due to the influence on water depth) and water quality (Figure 7.3). Spatially, spikerush was more predominant in the shallower, outer areas of the wetland. The deeper regions of Little Broadwater, especially along the eastern boundary (see Chapter 5, Figure 5.11), often did not have any vegetation due to the deeper water. Observations in the field over the study period indicated that lower water levels and fresh conditions (e.g. December 2005 to April 2006) encouraged the growth of spikerush across most of the wetland. In contrast, the maintenance of higher water levels during 2006 resulted in reducing the spatial extent of spikerush. However, the availability of fresh water throughout most of 2006 resulted in high density cover in the shallower areas of Little Broadwater. The effect of water quality on vegetation is discussed further in the ecology section. Wind-driven circulation was predominant in areas of open water within the wetland where there was little or no vegetation. The strength and duration of winds will determine how much of an influence wind-driven circulation has on the overall circulation patterns within the wetland (Figure 7.3). The importance of wind-driven circulation for flushing and mixing in shallow estuaries and embayments has been discussed by Geyer (1997) and Steedman and Craig (1983). Wind-driven circulation may be equally important at Little Broadwater, as it may be contributing to the pooling of poor quality water in the southeast corner by preventing or reducing exchange. Surface water quality within Little Broadwater was a function of many hydrological and ecological processes (Figure 7.3) and was therefore difficult to model and predict the response to wetland management strategies. The main processes that affected water quality were tidal exchange, freshwater inflows, evapotranspiration, ASS oxidation, circulation, topography, vegetation and cattle. As previously discussed, tidal exchange mainly influences the salinity of water within the wetland, although this process may also dilute acidic water in the region near the drain (see Chapter 4). Circulation patterns within the wetland also affect surface water quality, as this may promote mixing or cause poor quality water to pile up in the southeastern corner of the wetland. Hydrological processes such as freshwater inputs and evapotranspiration may respectively decrease or increase the salinity of surface water (Figure 7.3). Evapotranspiration and direct rainfall may further influence water quality through the process of ASS oxidation and mobilisation of acidic salts and metals (see Chapter 5). The influence of topography and Chapter 7: Conceptual Model and Management 180 pooling of water can increase the effect of evapotranspiration through the increased concentration of salts and recurrent ASS oxidation. Vegetation and cattle may further affect water quality (Figure 7.3) through increased nutrient inputs. Seasonal growth and decay of vegetation provides organic matter to the wetland and increases the nutrients available for uptake by other species. Nutrient inputs to Little Broadwater are also sourced from cattle, through defecation directly into the wetland water or via runoff from the surrounding higher ground. Furthermore, the close proximity of a golf course to the wetland many have contributed nutrients via runoff of fertilisers. The effect of groundwater on surface water appeared to be restricted to the shallow groundwater component at Little Broadwater, in the form of the relationship between surface water cover and evapotranspiration of the shallow groundwater (Figures 7.2 and 7.3). In turn, the effect of evapotranspiration on ASS oxidation and the vertical movement of acid salts upwards through the soil profile affected surface water quality (see Chapter 5). Although hydraulic conductivity was only moderate at Little Broadwater (Chapter 5, Figure 5.10), there may have been some exchange occurring between the shallow groundwater and wetland surface water, which may have transported ASS oxidation products directly into the wetland surface water (Figure 7.2). Although the conceptual model of Little Broadwater was developed as a whole-system model, the identification of surface water zones within Little Broadwater (see Chapter 4, Figure 4.14) illustrated the spatial variation of dominant factors which affect water quality. The northern section of the wetland was mainly influenced by freshwater inputs and tidal exchange had little effect on water quality. Water quality within the central region was largely a function of tidal flushing and the salinity of Sportsmans Creek. In contrast, the poor water quality of the southern and southeastern region was due to a lack of tidal flushing and pooling of water in depressions. The concentration of salts by evapotranspiration was increased by the lack of flushing. These areas were highly dependent on rainfall to dilute and/or flush acidic/saline/eutrophic water from the region. Thus, the model is applicable to the wetland as a whole, or to individual zones, and indicates the varying dominance of individual components. Chapter 7: Conceptual Model and Management 181 Groundwater The groundwater component of wetland hydrology at Little Broadwater consists of a number of discrete components that interact with each other to determine the spatio-temporal variability of groundwater characteristics. Only the shallow groundwater has been modelled (Figure 7.6), as there was little evidence of interactions between the deeper, semi-confined groundwater and surface water or shallow groundwater (Figure 7.2). The main sources of inflows to the shallow groundwater are precipitation and surface water within the wetland (via infiltration), while evapotranspiration and lateral outflow to the wetland surface water are the main outflows (Figure 7.6). However, there are a number of external factors which influence the infiltration and evapotranspiration rates, and thus watertable dynamics. Figure 7.6: Conceptual model of the shallow groundwater component at Little Broadwater. Watertable depth is the primary component which influences other factors. See Figure 7.1 for explanation of symbols. Chapter 7: Conceptual Model and Management 182 The infiltration rate is dependent on the intensity and duration of rainfall, the hydraulic conductivity (Ksat) of the soils and the watertable depth (Figure 7.6). Low intensity rainfall over an extended period will infiltrate more efficiently into the sediment and groundwater than high intensity rainfall over a short period. The Ksat of the soil will also determine how quickly rainfall infiltrates. Furthermore, a shallow watertable would reduce the infiltration rate and thereby increase runoff into the surface water component. The evapotranspiration rate is affected by surface water cover (a function of wetland volume), vegetation type and density, and the watertable depth. If the wetland volume is low, then decreased surface water cover can result in increased evapotranspiration of groundwater. Similarly, high density vegetation increases the transpiration of shallow groundwater sources, although variations in vegetation type will also affect the total transpiration of a given area. Watertable depth also influences the effect of evapotranspiration (i.e. a shallow watertable is more susceptible to drawdown via evapotranspiration), however, this would also be a function of surface water cover and vegetation cover (Figure 7.6). The oxidation of ASS and quality of the groundwater and surface water is dependent on evapotranspiration and is thus also a function of surface water cover and vegetation (Figure 7.6). Increased evapotranspiration and drawdown of the watertable can expose and oxidise ASS. A decrease in watertable depth and subsequent reflooding of the oxidised ASS reduces groundwater quality by increasing acidity and metal concentrations. Evapotranspiration of the watertable below the ASS layer may draw acid salts to the sediment surface (Figure 7.2), which can then be mobilised into the surface water during reflooding/rainfall events and thereby decrease surface water quality (Figure 7.3). Ecology Vegetation, waterbirds and cattle are the main ecological components of the Little Broadwater conceptual model (Figure 7.3). The spatial and temporal distribution of vegetation (and therefore habitat) influenced the bird species that utilise the wetland, and in turn, the types of waterbirds within the wetland can affect the distribution of vegetation. Similarly, increased growth of wetland pasture species provides feed for cattle; however, grazing decreases vegetation distribution and cover, which then alters the habitat. Wetland vegetation is an ecological component that affects both the groundwater and surface water components (Figures 7.3 and 7.6). The root zone is located in the shallow groundwater Chapter 7: Conceptual Model and Management 183 component (Figure 7.2) and can lower the watertable through transpiration, which may also draw acid salts upwards through the soil profile via capillary action (Johnston, Slavich & Hirst 2003). Vegetation in the surface water component can affect water flow rates (Duever 1988) and circulation patterns within the wetland, thus influencing water quality. The quality of surface water in turn directly influences the vegetation species that establish in the wetland (Figure 7.3). Historic drainage of the wetland and conversion to seasonal freshwater inundation resulted in the loss of mangrove species and the encroachment of swamp oak through the central region. To the east of the drain, Phragmites australis and Schoenoplectus litoralis (river club rush) are dominant and reflect the variable salinity and water depths within this area. Vegetation cover over the remainder of the wetland is dependent on fresh conditions in which spikerush and Paspalum distichum (water couch) grow. During acidic conditions, however, spikerush is the dominant species in the outer wetland areas as it is tolerant of acid water. The seasonal fresh conditions within Little Broadwater, along with varying water depths, results in a seasonal variation in habitat. Although management strategies changed between years over the study period (from brackish to fresh water management), and there was increased winter rainfall during the second year of the study, the seasonal changes in vegetation cover were apparent. Observations over the study period indicated that spikerush was predominant in late summer/early autumn during the wet season and generally died over the winter and spring (Figures 7.7, 7.8 and 7.9). However, increased rainfall in late winter/early spring 2006 resulted in early re-growth of spikerush throughout much of the wetland (Figures 7.8 and 7.9). Changes in management regime and climate during the study period, and the effect on vegetation and habitat, resulted in different bird species utilising the wetland. Saline management during 2004/2005 reduced the spatial coverage of spikerush and often there were large areas of mud flats. During this period waders were the dominant birds utilising Little Broadwater, and included species such as Platalea regia (royal spoonbill), Himantopus himantopus (black-winged stilts), Plegadis falcinellus (glossy ibis) and Ephippiorhynchus asiaticus (black-necked stork). Increased growth of spikerush over the 2005/2006 summer changed the available habitat in the wetland and as a result large flocks of swans were observed in the wetland grazing on the spikerush. However, continuous regrowth of spikerush Chapter 7: Conceptual Model and Management 184 over the winter and spring of 2006 resulted in a dense layer of vegetation, which excluded waterbirds from large areas of the wetland. Figure 7.7: Change in understorey vegetation due to variable salinity and water depth, Site 9 facing east: (a) May 2005; (b) October 2005; (c) February 2006; (d) June 2006; (e) October 2006; and (d) February 2007. Chapter 7: Conceptual Model and Management 185 Figure 7.8: Change in cover of spikerush due to variable salinity, Site 31 facing east: (a) October 2005; (b) February 2006; (c) April 2006; (d) June 2006; (e) October 2006; and (d) February 2007. Chapter 7: Conceptual Model and Management 186 Figure 7.9: Temporal variation in surface water cover and vegetation at Site 43 facing east: (a) June 2005; (b) October 2005; (c) April 2006; (d) June 2006; (e) October 2006; and (d) February 2007. Cattle are a further factor that has been incorporated into the conceptual model. Vegetation may decrease due to grazing and trampling, water quality may be degraded through increased nutrients, and compaction and disturbance of sediments may also occur (Eco Logical Australia 2008). The associated decrease of vegetation cover by grazing can open large areas of the wetland and allow access for waterbirds, which were previously excluded due to the density of growth (Figure 7.3). Cattle also provide a means of seed dispersal and thereby help Chapter 7: Conceptual Model and Management 187 to increase the spatial distribution of vegetation. Sustainable grazing within Little Broadwater will reduce the impact of cattle on vegetation while still providing ecological benefits to the wetland, such as seed dispersal. The processes and linkages identified in the Little Broadwater conceptual model are applicable to other coastal floodplain wetlands. This was exemplified by the comparison of Little Broadwater to the two reference sites – Colletts Swamp and Wooloweyah – in Chapter 6. The surface water volume within all three systems (and similarly in other wetlands), is a function of inflows and outflows, which also affect water quality within the wetland (Figure 7.3). Other factors, such as wind, have a reduced effect on circulation patterns in the reference wetlands due to different types of vegetation and surface water depth. While vegetation had less of an influence on ecosystem components at Colletts Swamp and Wooloweyah, the vegetation species present at a site influences the bird species which utilise the wetland (Figure 7.3). Furthermore, the growth rate and distribution of pasture species, and the flooding regime of the wetland, will determine the stocking rates and thus the effect of cattle and (potential) changes in nutrient concentrations. Processes which affect groundwater and oxidation of ASS (Figures 7.3 and 7.6) were also similar between the wetlands. This was particularly evident at Little Broadwater and Colletts Swamp, where poor circulation in combination with ASS oxidation led to areas of poor quality water within the far backswamp areas. Therefore, while some components may have a no or a reduced influence on other ecosystem components, the conceptual model developed for Little Broadwater is representative of coastal floodplain wetland functioning. The conceptual model of Little Broadwater (Figures 7.3 and 7.6) shows the wetland is very complex and many feedback loops exist between ecosystem components. While water quality is influenced by a number of different factors including tidal exchange management, the quality of surface water within the wetland is one of the information sources used to determine tidal exchange management. The conceptual model breaks down the complexity of the wetland functioning into a manageable representation of the system, indicating the implications of changing one or a number of components for future management. 7.4 Management options for Little Broadwater From the results of the longer-term study presented in the previous Chapters, a range of management options can be identified for ongoing maintenance of Little Broadwater in its current condition. Arising from this work (and other studies run in conjunction), two Chapter 7: Conceptual Model and Management 188 documents were developed for ongoing management of Little Broadwater. These were the drain management plan (Clarence Valley Council 2008) and the remediation management plan (NSW DPI 2006). The drain management plan is a day-to-day guide for management of the tidal exchange structures, whereas the remediation management plan provides options and recommendations for long-term, holistic management of the wetland. While the main purpose of the drain and floodgates at Little Broadwater is for flood mitigation, the drain management plan also provides for ecological benefits to the wetland. Adjustment of the structures can be made to improve water quality, provide fish passage, improve wetland habitat values, improve the grazing potential of the wetland, and for weed control (Clarence Valley Council 2008). Under the drain management plan water levels are to be maintained at approximately 0.12 m AHD (dependent on rainfall and water quality in Sportsmans Creek) for ASS management, to maintain the habitat values provided by the wetland and aid in wet pasture management (Clarence Valley Council 2008). While the drain management plan provides a two-fold purpose of providing flood protection and ecological benefits, it is important to note that there is great difficulty in achieving this given current designs of flood mitigation structures. For example, it is difficult to maintain a set water level while providing fish passage through the fish-flap structure. Private ownership of much of Little Broadwater adds an economic factor which must be considered for ongoing management. While ideally the wetland would be returned to a near- natural state of saline inundation during winter months and freshwater management during the summer, the long period of time between the loss of freshwater vegetation and establishment of saline-tolerant species would not be economically beneficial to the landholders. At the same time, grazing needs to be sustainable to reduce the impact on conservation values (Eco Logical Australia 2008). A comprehensive list of management options for Little Broadwater has been provided in the ‘Little Broadwater Remediation Management Plan’ (NSW DPI 2006) and discusses the positive and negative impacts of each. The management options could be summarised into five management objectives: i) reduce acid export; ii) improve fish passage, water quality and waterway health; iii) maintain and improve remnant terrestrial and aquatic ecosystems; iv) maintain and improve agricultural productivity; and v) rehabilitate and restore terrestrial and aquatic ecosystems (NSW DPI 2006). The management plan was developed from previous outcomes of wetland rehabilitation and ASS management projects, and from a Chapter 7: Conceptual Model and Management 189 variety of research conducted at Little Broadwater including studies on hydrology, vegetation, waterbirds and fish access. The hydrological and water quality study presented in this thesis was one component of the research which contributed to the development of the management plan. As such, a summary of three main management options is provided, with predicted outcomes related to linkages identified by the conceptual model presented previously, and examples are given from the water quality study. 7.4.1 Seasonal/adaptive management Seasonal/adaptive management is the preferred management option and involves limiting tidal exchange when Sportsmans Creek is saline. This strategy provides a compromise between the three initial rehabilitation objectives: i) protect and enhance surface water and groundwater ecosystems, ecological processes and industries (particularly fisheries); ii) reduce the severity, duration and frequency of acidic discharge; and iii) improve sustainable agriculture of the wetland (NSW DPI 2006). During fresh periods, exchange is increased and water retained within the wetland. However, during dry periods when salinity in the creek is above the recommended maximum for exchange (as set out in the drain management plan), there is the potential for desiccation of the wetland and subsequent acidification. The dynamic nature of water cover and water quality, and thus vegetation (see Figures 7.7, 7.8 and 7.9) means that seasonal/adaptive management may be perceived as both a success and a failure in terms of ecological and agricultural benefits within a 6-month period (i.e. spring to summer). Benefits of seasonal/adaptive management include increased wet pasture species, fish passage, provision of waterbird habitat (for some species) and a reduction in the occurrence of salt scalds (Figure 7.3). By retaining water within the wetland groundwater levels are also maintained at a higher level, reducing ASS oxidation and acid export from the system (Johnston, Slavich & Hirst 2004b). The benefits of this form of management were observed from December 2005 onwards, which saw considerable growth of wet pasture species such as spikerush and water couch (Figures 7.7c, 7.8c, e, f and 7.9f). A number of potential negative effects are also associated with the seasonal/adaptive management strategy, primarily relating to desiccation and subsequent acidification during dry periods. This occurred at Little Broadwater during late 2005/early 2006 when large areas of the wetland dried out (Figures 7.7b, 7.8a and 7.9a, b), exposing ASS. Ponded water, especially in the southeast, had very high concentrations of salts (Chapter 4, Figure 4.8) and large areas of vegetation also died. Increased rainfall over the summer months led to acidification of the entire wetland (Chapter 4, Figure 4.9) and acidic discharge into Chapter 7: Conceptual Model and Management 190 Sportsmans Creek (Chapter 4, Figure 4.6), although the severity and duration was considerably reduced in comparison to pre-rehabilitation events. While vegetation thrived in the freshwater, acidic water may have serious health implications for cattle, fish and other aquatic animals (e.g. scouring in cattle, red-spot disease and/or death of fish and frogs). A number of other management strategies have been applied in conjunction with the seasonal/adaptive management, including modification of the in-drain weir and associated structures, tree planting, management of cattle stocking rates and ASS remediation. Modification of the in-drain structures included replacing the dropboards with an undershot/overshot weir and installing a lifting device on the fish-flap. While the new structures still perform the same function as the old structures, there is increased accuracy of water level management within Little Broadwater, and thus better management of ASS and vegetation. To increase habitat and biodiversity within the wetland, Melaleuca trees were planted in the southeastern region to provide additional roosting areas. However, research by Johnston, Slavich and Hirst (2003) has shown that groundwater beneath M. quinquinervia had increased concentrations of acidity and metals, and thus there was high potential for movement of these acidic products to the sediment surface and into the surface water (Figure 7.2). Long-term monitoring of the re-planting site may be required to ensure that there is no negative impact associated with the potential positive biodiversity benefits. During the hydrological and water quality study, management of cattle stocking rates and ASS remediation strategies were trialled at Little Broadwater. Cattle trample vegetation and compact the sediments, thereby reducing vegetation cover and increasing the risk of ASS oxidation (NSW DPI 2006), and also contribute to poor water quality in terms of nutrients and ASS products (Figure 7.3). Management of cattle stocking in a small, severely scalded area of the wetland resulted in a considerable increase in vegetation cover. In association with this strategy, liming of part of the area under stock control was conducted and was effective in encouraging germination of the natural seed-bank. However, liming is only effective in the short-term as reapplication is needed and the cost prohibits large-scale use in the wetland. The overall benefits of seasonal/adaptive management of Little Broadwater could be improved by cleaning built-up sediment from the drainage pipes under Manton Road at the northern end of the wetland. This would allow more freshwater inflow to the main body of Chapter 7: Conceptual Model and Management 191 Little Broadwater from the northern section. While the management plan suggests that this may help reduce surface water salinity in the wetland (NSW DPI 2006), current water quality data indicates that the benefits would be limited to the northern section of the wetland where there is little saline influence (see Chapter 4). Rather, the primary benefit would be longer periods of inundation with freshwater which may encourage longer growing periods of wet pasture species and maintain shallow groundwater levels. Thus, benefits would be provided in terms of habitat, pasture and a reduction in ASS exposure. 7.4.2 Tidal flushing The second form of long-term management which could be applied to Little Broadwater is tidal flushing. The extent to which tidal flushing is restored will depend on the use of the wetland and desired outcomes. In general, this strategy would essentially return the wetland to a pre-drainage tidal regime and encourage the establishment of saline-tolerant vegetation (including mangroves). Associated with this would be an increase in biodiversity and the saline water may buffer acidic discharge while reducing oxidation of ASS (NSW DPI 2006). Saline tolerant vegetation would provide a number of benefits, including drought resistance, thereby reducing the occurrence of salt and acid scalds and providing year-round pastoral benefits. The negative impacts associated with this management strategy outweigh the benefits while the wetland is used for grazing. The long period of time required for wetland vegetation to transition from freshwater species to saline-tolerant species would place large financial burdens on the current landholders. If agricultural production and flood protection were still required, then it would only be viable to retain the concrete in-drain weir but remove the fish- flap and undershot/overshot weir (which replaced the dropboards). However, the reduced flushing capabilities of the southeastern region, due to the small amount of exchange and tidal forcing through the wetland mouth, may lead to hypersaline conditions, salt scalds and exposed sediments in this region (as occurred previously in November 2004 and October 2005; see Chapter 3, and Chapters 4 and 5, respectively). To prevent hypersalinisation and associated impacts in the southeastern region, removal of the flood mitigation structures would be required (potentially including the levee) or at minimum, the provision of additional exchange points through the levee between Sportsmans Creek and the southeastern region. However, implementing these actions would only be a viable option if Little Broadwater was purchased by an independent organisation (e.g. Department of Chapter 7: Conceptual Model and Management 192 Environment and Climate Change, WetlandCare Australia, etc.) due to reduced flood protection and the potential for short-term degradation as wetland vegetation adjusts to the salinity and inundation regime. Restoring full tidal flushing, however, would provide environmental benefits such as open fish habitat and increased biodiversity, along with economic benefits to fisheries and tourism (NSW DPI 2006). To implement this strategy, a detailed hydrological simulation model would need to be developed to ensure all risks to the environment and landholders are accounted for. This is due to a number of unknown factors, including the extent of sediment shrinkage and potential extent of saline intrusion, and lack of in-depth studies of circulation and exchange within the wetland. The financial cost for the development of the model would be substantial and therefore makes restoring tidal exchange an unattractive and potentially unviable management option. Management of Little Broadwater as both a freshwater and saline wetland could potentially be achieved through bunding around the core of the wetland. The centre of the wetland would be managed under full tidal exchange, as discussed above, while the outer regions would be managed as a seasonal freshwater wetland. However, implementation of this strategy would require a large amount of funding to model the hydrology accurately and engineer the structures required (NSW DPI 2006). There is also a high potential for the outer wetland areas to become desiccated or hypersaline (due to salts available in the sediments) during dry periods, exposing ASS and essentially returning the area to pre-rehabilitation conditions with salt and acid scalds, loss of vegetation and reduced agricultural productivity. Liming of the outer areas would be required, increasing the financial cost of implementing this strategy. Although this management option appears to benefit both the environment and landholders, the implications of returning the outer part of the wetland to pre-rehabilitation conditions is not recommended due to a number of financial and environmental costs (see Section 7.4.3). 7.4.3 Return to pre-trial management Returning Little Broadwater to pre-trial management is not a recommended option as there would be substantial costs to both the environment and landholders (NSW DPI 2006). By reflooding the wetland, natural reducing conditions which occur in intertidal environments promotes the reformation of sulfidic materials such as pyrite and monosulfides (Johnston et al. 2009). By excluding tidal flows from the wetland again, oxidation of newly formed and existing ASS would rapidly degrade water quality and the environment (Johnston et al. 2009). The impacts of returning Little Broadwater to pre-trial management can be identified through the conceptual model (Figures 7.3 and 7.6), including re-oxidation of ASS, increased Chapter 7: Conceptual Model and Management 193 occurrences of acid scalds, decreased water quality, loss of vegetation, reduced habitat for fish and waterbirds, and decreased grazing benefits. 7.5 Conclusion The results of a 2.5 year study on hydrology and water quality of Little Broadwater, in conjunction with ecological observations, were used to develop a conceptual model of wetland functioning. Although the model was developed for the wetland as a whole, components of the model have variable influence within different regions of the wetland due to spatial zonation. While the model is discussed here with reference to Little Broadwater, studies at reference wetlands indicated that similar processes occur across coastal floodplain wetlands. Therefore, the model can be adapted to examine both small-scale within-wetland processes, as well as comparing hydrological processes between wetlands. The conceptual model indicates linkages between wetland components, and thus the effect of management strategies. Three main management options have been developed for Little Broadwater, each with positive and negative risks associated with them. Based on the results of this study, restoring full tidal exchange and establishing a second exchange point would provide the best outcomes in terms of water quality and habitat. However, continuing monitoring of hydrological and ecological parameters over a long period of time (i.e. 10-20 years) document ecosystem changes. This would provide long-term data which may be used to develop a simulation model of wetland ecosystem functioning, and is recommended as a result of the current study. Furthermore, the information provided from long-term monitoring continues to build our knowledge of coastal floodplain wetland functioning and the complex relationships between the ecosystem components. 194 Chapter 8 General Conclusions The 2.5 year study of water quality characteristics and hydrology of Little Broadwater, undergoing a rehabilitation trial, indicated that the functioning of the wetland system was highly complex. Many factors influenced each ecosystem component, with a number of feed- back responses apparent within the system. Spatio-temporal variability of water quality indicated a changing dominance of factors within spatial zones and/or seasons, and thus increased the complexity of wetland system functioning. Climatic variability, in the form of precipitation and evapotranspiration, was the dominant control on wetland hydrology and water quality. Climate directly influenced the amount of tidal exchange permitted, surface water volumes, water quality, groundwater dynamics, processes associated with acid sulfate soil (ASS) oxidation, and indirectly influenced the biodiversity of the wetland. Spatio-temporal variability within Little Broadwater was primarily due to salinity, which is influenced by climatic conditions, circulation patterns, topography and management of tidal exchange. The comparative study between three coastal floodplain wetlands indicated that spatial variation of water quality was also a function of distance from the main exchange point (e.g. the creek) and the shape of the wetland. The presence of spatial zonation within Little Broadwater presented a problem for management, as the lack of flushing in the southeastern corner often lead to increased salinity within this region. However, spatial zonation of water quality (and the temporal variability within each zone), and thus spatio-temporal variability in vegetation, provided a range of habitats for waterbirds and aquatic animals. Investigation of discharge water quality pre- and post-rehabilitation, and comparisons to an drained and un-rehabilitated wetland (Reedy Creek), indicated that there had been a considerable reduction in the frequency, duration and severity of acid discharge due to rehabilitation. This was attributed primarily to increased flushing and dilution of drainage waters, rather than neutralisation by saline water. A decrease in acid production due to reflooding of sediments, and thus a reduction of ASS oxidation, also contributed to reduced acidic discharge from Little Broadwater. However, the susceptibility of the wetland to desiccation and acid discharge was exemplified during the study due to an extended dry period. Chapter 8: General Conclusions 195 Groundwater quality at Little Broadwater had high spatial variability due to the distribution of sulfidic and estuarine sediments, microtopography and watertable position. Although the watertable was raised by restoring tidal exchange, this did not appear to reduce the production of acid or the mobilisation of metals along the eastern side of the wetland. However, groundwater quality was measured in a deep, semi-confined aquifer (due to the inconsistent presence of shallow groundwater above the confining layer), which had little interaction with the surface water within the wetland. Instead, the shallow groundwater appeared to have an influence on wetland surface water quality, through either direct exchange or dissolution of acidic salts on the sediment surface by rainfall. The development of a conceptual model indicated the linkages between hydrological and ecological components of Little Broadwater. Although the model was developed for ongoing management and understanding of Little Broadwater, similarities between the main study site and two reference coastal floodplain wetlands indicated that spatial patterns of surface water and groundwater quality were similar. Therefore, the model may be a useful tool for comparing hydrological (and ecological) processes between coastal floodplain wetlands. Furthermore, the model can also be applied to individual zones of the wetland, and thus be used to examine small-scale processes. While the aims of the rehabilitation trial at Little Broadwater were successfully met by the implementation of a seasonal/adaptive management strategy (freshwater), the zonation of water quality within Little Broadwater presents a difficulty for continuing management of the wetland. Although the current seasonal/adaptive management strategy has greatly improved discharge water quality, reduced ASS oxidation, and provided ecosystem and agricultural benefits, there is still the potential for desiccation and acidification of the southern and southeastern region to occur. This is due to the restriction of tidal exchange once the salinity in the creek reaches a specified level, as the water quality study indicated that the EC of surface water in the southeastern region could be up to 7 dS m-1 higher than in the creek. The alternative management strategy of restoring full tidal flushing would provide a considerable reduction in ASS oxidation and increase biodiversity of the wetland. However, the sustainability of agriculture in the wetland would decline while vegetation transitioned from freshwater species to salt-tolerant species. To implement this management strategy land acquisition would be required (due to financial losses for landholders), and a second tidal exchange point between the southern region and Sportsmans Creek would need to be established to reduce hypersalinisation of this area. Chapter 8: General Conclusions 196 The results from this study have increased knowledge of coastal floodplain wetland functioning on the Clarence River. However, it is still necessary for further studies to be conducted to continue the successful management of Little Broadwater and expand knowledge in the field of wetland rehabilitation. Ongoing monitoring of salinity, ASS products and nutrients at Little Broadwater is recommended to provide long-term data on water quality variability. This will also ensure that the initial objectives of the rehabilitation trial are being met into the future. Secondly, a detailed study of groundwater dynamics and ASS processes is required to confirm the role of bacterial reduction in maintaining acidic deep groundwater conditions. Lastly, a hydrological simulation model of Little Broadwater should be developed to aid future management and more accurately predict the outcomes of alternative management strategies. This model may have broader applications for rehabilitation of other coastal floodplain wetlands. 197 References A. Fletcher and Associates 2002, Level Survey Plan of the North Eastern Portion of Everlasting Swamp, Lawrence. A. Fletcher and Associates Consulting Surveyors, Grafton. Ahn, C. & Mitsch, W.J. 2002, 'Evaluating the use of recycled coal combustion products in constructed wetlands: an ecologic-economic modeling approach', Ecological Modelling, vol. 150, no. 1-2, pp. 117-140. http://www.sciencedirect.com/science/article/B6VBS-44NM1KM- 2/2/1a85bfde997c6d2fb301d3218f4eb76f Allen, T.R., Tolvanen, H.T., Oertel, G.F. & McLeod, G.M. 2007, 'Spatial characterization of environmental gradients in a coastal lagoon, Chincoteague Bay', Estuaries and Coasts, vol. 36, no. 6, pp. 959-977. Al-Khudhairy, D.H.A., Thompson, J.R., Gavin, H. & Hamm, N.A.S. 1999, 'Hydrological modelling of a drained grazing marsh under agricultural land use and the simulation of restoration management scenarios', Hydrological Sciences Journal, vol. 44, no. 6, pp. 943-971. Anisfeld, S.C. & Benoit, G. 1997, 'Impacts of flow restrictions on salt marshes: An instance of acidification', Environmental Science and Technology, vol. 31, pp. 1650-1657. APHA 1998, Standard Methods for the Examination of Water and Wastewater, 20th edn, American Public Health Association, Washington, DC. Australian and New Zealand Environment and Conservation Council 2000, 'Australian and New Zealand Guidelines for Fresh and Marine Water Quality', in National Water Quality Management Strategy, ANZECC, Agricultural and Resource Management Council of Australia and New Zealand. Atkinson, G. 2000, ‘Acid sulfate soils assessment and management in New South Wales’, Papers of the Xth World Water Congress. Proceedings of the Conference held March 2000, Department of Land and Water Conservation, Melbourne. Australiana Bureau of Meteorology 2007a. Retrieved 2007 from http://www.bom.gov.au/index.shtml Australian Bureau of Meteorology 2007b, (updated 2 February 2007), Climate Statistics for Australian Locations - Definitions for other daily elements. Retrieved 30 April 2008 from http://www.bom.gov.au/climate/cdo/about/definitionsother.shtml References 198 Bawden, T. 1979, The Bawden Lectures (June 1886, July 1886, August 1888): The First Fifty Years of Settlement on the Clarence; With Notes by Mr R.C. Law, Third Edition. Clarence River Historical Society, Grafton, NSW. Beck, N.G., Fisher, A.T. & Bruland, K.W. 2001, 'Modeling water, heat, and oxygen budgets in a tidally dominated estuarine pond', Marine Ecology Progress Series, vol. 217, pp. 43-58. Bedford, B.L. 1996, 'The need to define hydrologic equivalence at the landscape scale for freshwater wetland mitigation', Ecological Applications, vol. 6, no. 1, pp. 57-68. Berner, R.A. 1984, 'Sedimentary pyrite formation: An update', Geochimica et Cosmochimica Acta, vol. 48, pp. 605-615. Blunden, B. & Indraratna, B. 2001, 'Pyrite oxidation model for assessing ground-water management strategies in acid sulfate soils', Journal of Geotechnical and Geoenvironmental Engineering, vol. 127, no. 2, pp. 146-157. http://link.aip.org/link/?QGT/127/146/1 Blunden, B.G. & Indraratna, B. 2000, 'Evaluation of surface and groundwater management strategies for drained sulfidic soil using numerical simulation models', Australian Journal of Soil Research, vol. 38, pp. 569-590. Boumans, R.M.J., Burdick, D.M. & Dionne, M. 2002, 'Modeling habitat change in salt marshes after tidal restoration', Restoration Ecology, vol. 10, no. 3, pp. 543-555. Boyer, J.N., Fourqurean, J.W. & Jones, R.D. 1997, 'Spatial characterization of water quality in Florida Bay and Whitewater Bay by multivariate analyses: Zones of similar influence', Estuaries, vol. 20, no. 4, pp. 743-758. http://links.jstor.org/sici?sici=0160- 8347%28199712%2920%3A4%3C743%3ASCOWQI%3E2.0.CO%3B2-K Bridgham, S.D. & Richardson, C.J. 1993, 'Hydrology and nutrient gradients in North Carolina peatlands', Wetlands, vol. 13, no. 3, pp. 207-218. Brinson, M.M. & Rheinhardt, R. 1996, 'The role of reference wetlands in functional assessment and mitigation', Ecological Applications, vol. 6, no. 1, pp. 69-76. Burton, E.D., Bush, R.T., Sullivan, L.A., Johnston, S.G. & Hocking, R.K. 2008, 'Mobility of arsenic and selected metals during re-flooding or iron- and organic-rich acid-sulfate soil', Chemical geology, vol. 253, pp. 64-73. Bush, R.T., Fyfe, D. & Sullivan, L.A. 2004, 'Occurrence and abundance of monosulfidic black ooze in coastal acid sulfate soil landscapes', Australian Journal of Soil Research, vol. 42, no. 6, pp. 609-616. References 199 Bush, R.T., Sullivan, L.A., Fyfe, D. & Johnston, S. 2004, 'Redistribution of monosulfidic black oozes by floodwaters in a coastal acid sulfate soil floodplain', Australian Journal of Soil Research, vol. 42, no. 6, pp. 603-607. Cairns, J., Jr. 1988, ‘Increasing diversity by restoring damaged ecosystems’, in Biodiversity, E.O. Wilson (ed.), National Academy Press, Washington DC, pp. 333-343. Callaway, J.C. 2001, 'Hydrology and substrate', in Handbook for Restoring Tidal Wetlands, J.B. Zedler (ed.), CRC Press, Boca Raton, Florida, pp. 89-117. Clarence River County Council 2001, Drain Management Plan: Lawrence/Whalans - 81 Drain. Clarence River County Council, Grafton, Australia. Clarence Valley Council 2006a, Floodplain Engineering Services. Retrieved 04/01/09 from http://www.clarence.nsw.gov.au/cmst/cvc009/lp.asp?cat=212 Clarence Valley Council 2006b, Historical Flood Information. Retrieved 19/12/2007 from http://www.clarence.nsw.gov.au/cmst.cvc009/1p.asp?cat=219 Clarence Valley Council 2008, Drain Management Plan: Lawrence Whalans Drain. Clarence Valley Council, Grafton. Clarke, K.R. & Warwick, R.M. 1994, Change in Marine Communities: An Approach to Statistical Analysis and Interpretation, Natural Environment Research Council, UK, Plymoth. Clay, C. & Hirst, P. 2004, Field Investigation Little Broadwater, NSW Department of Primary Industries. Clay, C., Rose, C., Rose, H. & Wilson, P. 2007, ‘Could their grazing potential ultimately save floodplain wetlands?’, The Tide is High - Coastal Pressures, Climate Change, Sustainability: The Way Forward. 16th NSW Coastal Conference. Proceedings of the Conference held 7-9 November 2007, Yamba NSW. Cook, F.J., Hicks, W., Gardner, E.A., Carlin, G.D. & Froggatt, D.W. 2000, 'Export of acidity in drainage water from acid sulphate soils', Marine Pollution Bulletin, vol. 41, no. 7- 12, pp. 319-326. Cowardin, L.M., Carter, V., Golet, F.C. & LaRoe, E.T. 1979, Classification of Wetlands and Deepwater Habitats of the United States, U.S. Department of the Interior, Fish and Wildlife Service, Washington D.C. Retrieved 2 May 2006 from http://www.npwrc.usgs.gov/resource/1998/classwet/classwet.htm Davison, L. & Wilson, P. 2003, Palmers Channel Comparative Acid Neutralisation Demonstration. Grafton: Clarence River County Council Floodplain Management Authority, 2/701. References 200 Demayo, A. & Steel, A. 1996, 'Data handling and presentation', in Water Quality Assessments: A Guide to the Use of Biota, Sediments and Water in Environmental Monitoring, 2nd edn, D. Chapman (ed.), E&FN Spon, London, pp. 511-612. Dent, D. 1986, Acid Sulphate Soils: A Baseline for Research and Development, International Institute for Land Reclamation and Improvement, Wageningen, The Netherlands. Department of Natural Resources n.d., Wetlands - North Coast Region, NSW Department of Natural Resources. Retrieved 2 October 2008 from http://www.naturalresources.nsw.gov.au/water/wetlands_area_ncoast.shtml Department of Natural Resources and Water 2007, Measuring Salinity. DNRW, State of Queensland, Queensland. Drever, J.I. 1997, The Geochemistry of Natural Waters: Surface and Groundwater Environments, 3rd edn, Prentice Hall, Upper Saddle River, New Jersey, USA. Drexler, J.Z., Bedford, B.L., DeGaetano, A.T. & Siegel, D.I. 1999, 'Quantification of the water budget and nutrient loading in a small peatland', Journal of the American Water Resources Association, vol. 35, no. 4, pp. 753-769. Duever, M.J. 1988, 'Hydrologic processes for models of freshwater wetlands', in Wetland Modelling, W.J. Mitsch, M. Straškraba & S.E. Jørgensen (eds), Elsevier, Amsterdam, pp. 9-39. Eco Logical Australia. 2008, Conservation Assessment of Wetlands in the Clarence Lowlands IBRA Subregion. Report Prepared for Australian Department of the Environment, Water, Heritage and the Arts. Coffs Harbour: Department of Environment and Climate Change (NSW). Eertman, R.H.M., Kornman, B.A., Stikvoort, E. & Verbeek, H. 2002, 'Restoration of the Sieperda tidal marsh, in Scheldt Estuary, The Netherlands', Restoration Ecology, vol. 10, no. 3, pp. 438-449. Environment Australia. 2001, A Directory of Important Wetlands in Australia. Canberra: Environment Australia. Finlayson, C.M., Davidson, N.C., Spiers, A.G. & Stevenson, N.J. 1999, 'Global wetland inventory - Current status and future priorities', Marine and Freshwater Research, vol. 50, pp. 717-727. Finlayson, C.M. & Mitchell, D.S. 1999, 'Australian wetlands: The monitoring challenge', Wetlands Ecology and Management, vol. 7, pp. 105-112. Foley, M. & White, N. 2007, Management Options for the Wooloweyah Ring Drain and Palmers Channel Drainage Systems. Grafton: Clarence Valley Council. References 201 Galbraith, L. & Burns, C. 2007, 'Linking land-use, water body type and water quality in southern New Zealand', Landscape Ecology, vol. 22, no. 2, pp. 231-241. http://dx.doi.org/10.1007/s10980-006-9018-x Green, R., Macdonald, B.C.T., Melville, M.D. & Waite, T.D. 2006, 'Hydrochemistry of episodic drainage waters discharged from an acid sulfate soil affected catchment', Journal of Hydrology, vol. 325, pp. 356-375. Geyer, W.R. 1997, 'Influence of wind on dynamics and flushing of shallow estuaries', Estuarine, Coastal and Shelf Science, vol. 44, pp. 713-722. Gilvear, D.J., Andrews, R., Tellam, J.H., Lloyd, J.W. & Lerner, D.N. 1993, 'Quantification of the water balance and hydrogeological processes in the vicinity of a small groundwater-fed wetland, East Anglia, UK', Journal of Hydrology, vol. 144, pp. 311- 334. Glamore, W.C. & Indraratna, B. 2001, ‘The impact of floodgate modifications on water quality in acid sulphate soil terrains’, Coasts and Ports 2001. 15th Australasian Coastal and Ocean Engineering Conference and 8th Australasian Port and Harbour Conference. Proceedings of the Conference held 25-28 September 2001, The Institute of Engineers Australia, Gold Coast, Queensland, pp. 265-270. Glamore, W.C. & Indraratna, B. 2004, 'A two-stage decision support tool for restoring tidal flows to flood mitigation drains affected by acid sulfate soils: Case study of Broughton Creek floodplain, New South Wales, Australia', Australian Journal of Soil Research, vol. 42, pp. 639-648. Glamore, W.C. & Indraratna, B. 2005, ‘A multi-staged management strategy for restoring tidal flushing’, Coasts and Ports: Coastal Living - Living Coast; Australasian Conference 2005. Proceedings of the Conference held 20-23 September 2005, Institution of Engineers, Australia, Adelaide, South Australia, pp. 619-624. Green, R., Macdonald, B.C.T., Melville, M.D. & Waite, T.D. 2006, 'Hydrochemistry of episodic drainage waters discharged from an acid sulfate soil affected catchment', Journal of Hydrology, vol. 325, pp. 356-375. Greenblatt, M.S. & Sobey, R.J. 1999, 'Subsurface flow and transport in tidal wetlands: Marsh plain equations', Journal of Engineering Mechanics, vol. 125, no. 8, pp. 971-974. Hach Environmental 2006, Instruction Sheet: Hach LDOTM Sensor. Hach Company, USA. Hall, K.C., Baldwin, D.S., Rees, G.N. & Richardson, A.J. 2006, 'Distribution of inland wetlands with sulfidic sediments in the Murray-Darling Basin, Australia', Science of The Total Environment, vol. 370, no. 1, pp. 235-244. References 202 http://www.sciencedirect.com/science/article/B6V78-4KPX8PJ- 1/2/83aeceec3b147a9e5c0828cad086e21e Hammer, D.E. & Kadlec, R.H. 1986, 'A model for wetland surface water dynamics', Water Resources Research, vol. 22, no. 13, pp. 1951-1958. Hamming, A.F.J. & van den Eelaart, A.L.J. 1993, ‘Soil permeability, interflow and actual acidity in acid sulphate soils, South Kalimantan, Indonesia’, Selected Papers from the Ho Chi Minh City Symposium on Acid Sulfate Soils. Proceedings of the Conference held March 1992, International Institute for Land Reclamation and Improvement, Wageningen, The Netherlands, Ho Chi Minh City, Vietnam, pp. 155-159. Haskins, P.G. 2000, ‘Active floodgate management’, Proceedings of the Remediation and Assessment of Broadacre Acid Sulfate Soils Workshop. Proceedings of the Conference held 31 August - 2 September 1999, NSW Agriculture, Lismore, NSW, pp. 70-77. Haworth, R.J. & Ollier, C.D. 1992, 'Continental rifting and drainage reversal: The Clarence River of eastern Australia', Earth Surface and Landforms, vol. 17, pp. 387-397. Helena, B., Pardo, R., Vega, M., Barrado, E., Fernandez, J.M. & Fernandez, L. 2000, 'Temporal evolution of groundwater composition in an alluvial aquifer (Pisuerga River, Spain) by principal component analysis', Water Research, vol. 34, no. 3, pp. 807-816. http://www.sciencedirect.com/science/article/B6V73-3Y3PT09- D/2/5719e50914727c781c2488a7707b4b46 Helsel, D.R. & Hirsch, R.M. 2002, Chapter A3: Statistical Methods in Water Resources, U.S. Geological Survey. Retrieved 30 August 2007 from http://water.usgs.gov/pubs/twri/twri4a3/ Hem, J.D. 1985, Study and Interpretation of the Chemical Characteristics of Natural Water. Washington, DC: United States Geological Survey, Water-Supply Paper No. 2254. Henderson, S. & Tulau, M. 2001, ‘Active water management trials for the remediation of acid sulfate soils backswamps, Kinchela Creek, New South Wales, Australia’, 2nd Australian Wetland Conference. Proceedings of the Conference held 15-17 October 2001, Couran Cove Island Resort, South Stradbroke Island, Qld. Hernández-Romero, A.H., Tovilla-Hernández, C., Malo, E.A. & Bello-Mendoza, R. 2004, 'Water quality and presence of pesticides in a tropical coastal wetland in southern Mexico', Marine Pollution Bulletin, vol. 48, no. 11-12, pp. 1130-1141. http://www.sciencedirect.com/science/article/B6V6N-4BRSD34- 5/1/c001f281c663ca57721b1ed68248c249 References 203 Hollis, G.E. & Thompson, J.R. 1998, 'Hydrological data for wetland management', Journal of the Chartered Institution of Water and Environmental Management, vol. 12, pp. 9-17. Horiba 1991, Instruction Manual: U-10 Water Quality Checker. Horiba, Hosomi, M. & Sudo, R. 1986, 'Simultaneous determination of total nitrogen and total phosphorus in freshwater samples using persulfate digestion', International Journal of Environmental Studies, vol. 27, no. 3, pp. 267 - 275. http://www.informaworld.com/10.1080/00207238608710296 Hughes, C.E., Binning, P. & Willgoose, G.R. 1998, 'Characterisation of the hydrology of an estuarine wetland', Journal of Hydrology, vol. 211, no. 1-4, pp. 34-49. Husson, O., Verburg, P.H., Phung, M.T. & van Mensvoort, M.E.F. 2000, 'Spatial variability of acid sulphate soils in the Plain of Reeds, Mekong delta, Vietnam', Geoderma, vol. 97, no. 1-2, pp. 1-19. Indraratna, B., Blunden, B. & Nethery, A.G. 1999, 'Nature and properties of acid sulphate soils in drained coastal lowlands in New South Wales', Australian Geomechanics, vol. 34, pp. 61-78. Indraratna, B., Glamore, W.C. & Tularam, G.A. 2002, 'The effects of tidal buffering on acid sulphate soil environments in coastal areas of New South Wales', Geotechnical and Geological Engineering, vol. 20, no. 3, pp. 181-199. Indraratna, B., Tularam, G.A. & Blunden, B. 2001, 'Reducing the impact of acid sulphate soils at a site in Shoalhaven Floodplain of New South Wales', Quarterly Journal of Engineering and Hydrogeology, vol. 34, pp. 333-346. isee Systems Inc. 2007, iThink/STELLA 9 Technical Documentation: Software Reference Guide, isee Systems Inc. Johnston, C.A. 1993, 'Material fluxes across wetland ecotones in northern landscapes', Ecological Applications, vol. 3, no. 3, pp. 424-440. Johnston, S.G., Bush, R.T., Sullivan, L.A., Burton, E.D., Smith, D., Martens, M.A., McElnea, A.E., Ahern, C.R., Powell, B., Stephens, L.P., Wilbraham, S.T. & van Heel, S. 2009, 'Changes in water quality following tidal inundation of coastal lowland acid sulfate soil landscapes', Estuarine, Coastal and Shelf Science, vol. 81, pp. 257-266. Johnston, S.G. & Slavich, P.G. 2003, Hydraulic Conductivity - A Simple Field Test for Shallow Coastal Acid Sulfate Soils, NSW Agriculture. Retrieved 6 September 2004 from http://www.agric.nsw.gov.au/reader/hydraulic-conductivity/field-test- instructions.pdf?MIvalObj=19717&doctype=document&MItypeObj=application/pdf& name=/field-test-instructions.pdf References 204 Johnston, S. & Slavich, P. 2004, Hydrological Effects of Floodgate Management on Coastal Floodplain Agriculture: Final Report Summary for Project DAN13. Land and Water Australia. Johnston, S.G., Slavich, P.G. & Hirst, P. 2002, ‘Floodgate and drainage system management, opportunities and limitations - An acid export perspective’, Floodgate Design and Modification Workshop. Proceedings of the Conference held 14 August 2002, NSW Fisheries, Ballina, NSW, pp. 59-73. Johnston, S.G., Slavich, P.G. & Hirst, P. 2003, 'Alteration of groundwater and sediment geochemistry in a sulfidic backswamp due to Melaleuca quinquenervia encroachment', Australian Journal of Soil Research, vol. 41, pp. 1343-1367. Johnston, S.G., Slavich, P.G. & Hirst, P. 2004a, 'The acid flux dynamics of two artificial drains in acid sulfate soil backswamps on the Clarence River floodplain, Australia', Australian Journal of Soil Research, vol. 42, pp. 623-627. Johnston, S.G., Slavich, P.G. & Hirst, P. 2004b, 'The effects of a weir on reducing acid flux from a drained coastal acid sulphate soil backswamp', Agricultural Water Management, vol. 69, pp. 43-67. Johnston, S.G., Slavich, P.G. & Hirst, P. 2005a, 'The impact of controlled tidal exchange on drainage water quality in acid sulphate soil backswamps', Agricultural Water Management, vol. 73, pp. 87-111. Johnston, S.G., Slavich, P.G. & Hirst, P. 2005b, 'Opening floodgates in coastal floodplain drains: Effects on tidal forcing and lateral transport of solutes in adjacent groundwater', Agricultural Water Management, vol. 74, no. 1, pp. 23-46. Kench, P.S. 1999, 'Geomorphology of Australian estuaries: Review and perspective', Australian Journal of Ecology, vol. 24, pp. 367-380. Kentula, M.E. 2000, 'Perspectives on setting success criteria for wetland restoration', Ecological Engineering, vol. 15, no. 3-4, pp. 199-209. http://www.sciencedirect.com/science/article/B6VFB-40X8FRG- 5/2/e2cf3fc4df7b5686153674b6c5f62153 Kinsela, A.S. & Melville, M.D. 2004, 'Mechanisms of acid sulfate soil oxidation and leaching under sugarcane cropping', Australian Journal of Soil Research, vol. 42, no. 5-6, pp. 569-578. http://find.galegroup.com/ips/infomark.do?&contentSet=IAC- Documents&type=retrieve&tabID=T002&prodId=IPS&docId=A124008574&source= gale&srcprod=ITOF&userGroupName=dixson&version=1.0 References 205 Kowalkowski, T., Zbytniewski, R., Szpejna, J. & Buszewski, B. 2006, 'Application of chemometrics in river water classification', Water Research, vol. 40, no. 4, pp. 744- 752. http://www.sciencedirect.com/science/article/B6V73-4J44006- 1/2/a783d1cc400efa22f396c66573484745 Lancaster, G. 1990, Physicochemical Conditions and Seasonal Changes in Wooloweyah Lagoon, Yamba, NSW, Bachelor of Applied Science with Honours, University of New England, Northern Rivers. Lee, J. 1993, 'A formal approach to hydrological model conceptualization', Hydrological Sciences Journal, vol. 38, no. 5, pp. 391-401. Lewis, R.R.III. 1989, ‘Wetlands restoration/creation/enhancement technology: Suggestions for standardization’, in Wetland Creation and Restoration: The Status of The Science, Vol. II, J.A. Kusler and M.E. Kentual (eds), Environmental Research Laboratory, Corvallis, OR, pp. 1-7 Lin, C. & Melville, M.D. 1993, 'Control of soil acidification by fluvial sedimentation in an estuarine floodplain, eastern Australia', Sedimentary Geology, vol. 85, pp. 271-284. Lin, C., Melville, M.D. & Hafer, S. 1995, 'Acid sulphate soil-landscape relationships in an undrained, tide-dominated estuarine floodplain, Eastern Australia', CATENA, vol. 24, no. 3, pp. 177-194. http://www.sciencedirect.com/science/article/B6VCG-3YCN43Y- 7/2/b1d2dbe4a8e1b1a657c30bae76fe87ab Lin, C., Rosicky, M.A., McConchie, D., Sullivan, L.A. & Lancaster, G. 2001, 'Coastal land scalding in NSW, Australia: Soil chemical characteristics and their implications for remediation of the scalded land', Land Degradation and Development, vol. 12, pp. 293-303. Liu, C.-W., Jang, C.-S., Chen, C.-P., Lin, C.-N. & Lou, K.-L. 2007, 'Characterization of groundwater quality in Kinmen Island using multivariate analysis and geochemical modelling', Hydrological Processes, online http://dx.doi.org/10.1002/hyp.6606 Maher, W.A., Cullen, P.W. & Norris, R.H. 1994, 'Framework for designing sampling programs', Environmental Monitoring and Assessment, vol. 30, pp. 139-162. McSwan, E.H. & Switzer, M. 2006, A Thematic History of Maclean Shire, Volume 2, Maclean Shire (former) Community Based Heritage Study. Melville, M.D. & White, I. 2000, 'Acid Sulfate Soils', in Soils: Their Properties and Management, 2nd edn, P.E.V. Charman & B.W. Murphy (eds), Oxford University Press, Melbourne, pp. 249-252. References 206 Milford, S.N. & Church, J.A. 1977, 'Simplified circulation and mixing models of Moreton Bay, Queensland', Australian Journal of Marine and Freshwater Research, vol. 28, no. 1, pp. 23-34. Mitsch, W.J. & Gosselink, J.G. 2000, Wetlands, 3rd edn, John Wiley and Sons, New York. Mitsch, W.J. & Reeder, B.C. 1991, 'Modelling nutrient retention of a freshwater coastal wetland: estimating the roles of primary productivity, sedimentation, resuspension and hydrology', Ecological Modelling, vol. 54, no. 3-4, pp. 151-187. http://www.sciencedirect.com/science/article/B6VBS-48XDC0K- 3W/2/6e1f02e1c621ff2093cfff26338b5a9f Morand, D.T. 2001, Soil Landscapes of the Woodburn 1:100 000 Sheet, Department of Land and Water Conservation, Alstonville, NSW. Morand, D.T. 2002, Everlasting Swamp "Hotspot" Study - Soils Report. Department of Land and Water Conservation, Alstonville, NSW. National Research Council (U.S.) & Committee on Restoration of Aquatic Ecosystems - Science, Technology, and Public Policy 1992, Restoration of Aquatic Ecosystems, National Academy Press, Washington D.C. National Wetlands Working Group 1997, The Canadian Wetlands Classification System, 2nd edn, NWWG, Wetlands Research Centre, University of Waterloo, Waterloo, Ontario. Neckles, H.A., Dionne, M., Burdick, D.M., Roman, C.T., Buchsbaum, R. & Hutchins, E. 2002, 'A monitoring protocol to assess tidal restoration of salt marshes on local and regional scales', Restoration Ecology, vol. 10, no. 3, pp. 556-563. Nordstrum, D.K. 1982, 'Aqueous pyrite oxidation and the consequent formation of secondary iron minerals', in Acid Sulfate Weathering, J.A. Kittrick, D.S. Fanning & L.R. Hassner (eds), Soil Science Society of America, Madison, WI, USA, Special Publication 10, pp. 37-56. NSW Department of Natural Resources n.d., Estuaries in NSW: Clarence River. Retrieved 30 April 2008 from http://www.dnr.nsw.gov.au/estuaries/inventory/clarence.shtml NSW Department of Primary Industries. 2006, Little Broadwater Remediation Management Plan. NSW Department of Primary Industries. Nriagu, J.O. 1978, 'Dissolved silica in pore water of Lakes Ontario, Erie and Superior sediments', Limnology and Oceanography, vol. 23, no. 1, pp. 53-67. Ollier, C.D. & Haworth, R.J. 1994, 'Geomorphology of the Clarence-Moreton Basin', in Geology and Petroleum Potential of the Clarence-Moreton Basin, New South Wales References 207 and Queensland, DPIE Bulletin 241 edn, A.T. Wells & P.E. O'Brien (eds), Australian Geological Survey Organisation, Canberra, pp. 291-302. Portnoy, J.W. 1999, 'Salt marsh diking and restoration: Biogeochemical implications of altered wetland hydrology', Environmental Management, vol. 24, no. 1, pp. 111-120. Portnoy, J.W. & Giblin, A.E. 1997a, 'Biogeochemical effects of seawater restoration to diked salt marshes', Ecological Applications, vol. 7, no. 3, pp. 1054-1063. Portnoy, J.W. & Giblin, A.E. 1997b, 'Effects of historic tidal restrictions on salt marsh sediment chemistry', Biogeochemistry, vol. 36, pp. 275-303. Preda, M. & Cox, M.E. 2004, 'Temporal variations of mineral character of acid-producing pyritic coastal sediments, Southeast Queensland, Australia', Science of The Total Environment, vol. 326, no. 1-3, pp. 257-269. Pressey, R.L. 1987, A Survey of Wetlands on the Lower Clarence Floodplain, New South Wales, NSW National Parks and Wildlife Service, Sydney. Pressey, R.L. 1989, 'Wetlands of the Lower Clarence Floodplain, Northern Coastal New South Wales', Proceedings of the Linnean Society of NSW, vol. 111, no. 3, pp. 143- 155. Pressey, R.L. & Middleton, M.J. 1982, 'Impacts of flood mitigation works on coastal wetlands in New South Wales', Wetlands, vol. 2, no. 1, pp. 27-44. Ramsar Convention Secretariat 2006, The Ramsar Convention Manual: A guide to the Convention on Wetlands (Ramsar, Iran, 1971), 4th edn, Ramsar Convention Secretariat, Gland, Switzerland. Ranasinghe, R. & Pattiaratchi, C. 1999, 'Circulation and mixing characteristics of a seasonally open tidal inlet: A field study', Marine and Freshwater Research, vol. 50, no. 4, pp. 281-290. Rayment, G.E. & Higginson, F.R. 1992, Australian Laboratory Handbook of Soil and Water Chemical Methods, Inkata Press, Melbourne. Refsgaard, J.C., Sørensen, H.R., Mucha, I., Rodak, D., Hlavaty, Z., Bansky, L., Klucovska, J., Topolska, J., Takac, J., Kosc, V., Enggrob, H.G., Engesgaard, P., Jensen, J.K., Fiselier, J., Griffioen, J. & Hansen, S. 1998, 'An integrated model for the Danubian Lowland - Methodology and applications', Water Resources Management, vol. 12, pp. 433-465. Restrepo, J.I., Garces, D., Montoya, A.M., Restrepo, N. & Giddings, J. 2006, 'Simulation of water levels and water diversions in a subtropical coastal wetland', Journal of Coastal Research, vol. 22, no. 2, pp. 339-349. References 208 Restrepo, J.I., Montoya, A.M. & Obeysekera, J. 1998, 'A wetland simulation module for the MODFLOW ground water model', Ground Water, vol. 36, no. 5, pp. 764-770. Roman, C.T., Garvine, R.W. & Portnoy, J.W. 1995, 'Hydrologic modeling as a predictive basis for ecological restoration of salt marshes', Environmental Management, vol. 19, no. 4, pp. 559-566. Rosicky, M.A., Slavich, P., Sullivan, L.A. & Hughes, M. 2006, 'Surface and sub-surface salinity in and around acid sulfate soil scalds in the coastal floodplains of New South Wales, Australia', Australian Journal of Soil Research, vol. 44, pp. 17-25. Roy, P.S. 1984, 'New South Wales estuaries: Their origin and evolution', in Coastal Geomorphology in Australia, B.G. Thom (ed.), Academic Press, Sydney, pp. 99-121. Roy, P.S. & Thom, B.G. 1981, 'Late Quaternary marine deposition in New South Wales and southern Queensland - An evolutionary model', Journal of the Geological Society of Australia, vol. 28, pp. 471-489. Roy, P.S., Williams, R.J., Jones, A.R., Yassini, I., Gibbs, P.J., Coates, B., West, R.J., Scanes, P.R., Hudson, J.P. & Nichol, S. 2001, 'Structure and function of south-east Australian estuaries', Estuarine, Coastal and Shelf Science, vol. 53, no. 3, pp. 351-384. Rozsa, R. 1995, (updated updated 21 June 1997), Tidal wetland restoration in Connecticut, The Connecticut College Arboretum. Retrieved 9 September 2004 from http://camel2.conncoll.edu/ccrec/greennet/arbo/publications/34/MAIN.HTM Salas, H.J. & Martino, P. 1991, 'A simplified phosphorus trophic state model for warm-water tropical lakes', Water Research, vol. 25, no. 3, pp. 341-350. http://www.sciencedirect.com/science/article/B6V73-48BCV06- 3W/2/3292e64cfeb2cd569edf7a6175738be2 Sammut, J., White, I. & Melville, M.D. 1996, 'Acidification of an estuarine tributary in eastern Australia due to drainage of acid sulfate soils', Marine and Freshwater Research, vol. 47, pp. 669-684. Shrestha, S. & Kazama, F. 2007, 'Assessment of surface water quality using multivariate statistical techniques: A case study of the Fuji river basin, Japan', Environmental Modelling & Software, vol. 22, no. 4, pp. 464-475. http://www.sciencedirect.com/science/article/B6VHC-4JJ2BVV- 3/2/ba895c12c1bdf1182a00c74e4bb2ecad Simenstad, C.A. & Thom, R.M. 1996, 'Functional equivalency trajectories of the restored Gog-Le-Hi-Te estuarine wetland', Ecological Applications, vol. 6, no. 1, pp. 38-56. References 209 Simeonov, V., Stratis, J.A., Samara, C., Zachariadis, G., Voutsa, D., Anthemidis, A., Sofoniou, M. & Kouimtzis, T. 2003, 'Assessment of the surface water quality in Northern Greece', Water Research, vol. 37, no. 17, pp. 4119-4124. http://www.sciencedirect.com/science/article/B6V73-498TS4S- 3/2/09c29c723d1d29bf9b7507dbf82eb759 Singh, K.P., Malik, A., Mohan, D. & Sinha, S. 2004, 'Multivariate statistical techniques for the evaluation of spatial and temporal variations in water quality of Gomti River (India) - A case study', Water Research, vol. 38, no. 18, pp. 3980-3992. http://www.sciencedirect.com/science/article/B6V73-4D8MHMB- 2/2/cb2683399ca0a86db8c4ba4f5924988f Singh, V.P. (ed.) 1995, Computer Models of Watershed Hydrology, Water Resources Publications. Smith, A.N. in press, 'Waterbirds and their habitat on the Clarence River floodplain: A history', Australian Zoologist. Smith, B. 2002, Coastal Backswamps: Restoring Their Value, NSW Agriculture. Retrieved 6 September 2004 from http://www.agric.nsw.gov.au/reader/extension-material/coastal- backswamps.pdf?MIvalObj=19593&doctype=document&MItypeObj=application/pdf &name=/coastal-backswamps.pdf Spaulding, M.L., Mendelsohn, D.L. & Swanson, J.C. 1999, 'WQMAP: An integrated three- dimensional hydrodynamic and water quality model system for estuarine and coastal applications', Marine Technology Society Journal, vol. 33, no. 3, pp. 38-54. Spieles, D.J. & Mitsch, W.J. 1999, 'The effects of season and hydrologic and chemical loading on nitrate retention in constructed wetlands: a comparison of low- and high- nutrient riverine systems', Ecological Engineering, vol. 14, no. 1-2, pp. 77-91. http://www.sciencedirect.com/science/article/B6VFB-3XDHH8P- 7/2/7438e3e24b01a2fbef10180f8ae6800b SPSS, SPSS 15.0 for Windows, USA. Steedman, R.K. & Craig, P.D. 1983, 'Wind-driven circulation of Cockburn Sound', Australian Journal of Marine and Freshwater Research, vol. 34, no. 1, pp. 187-212. Stolt, M.H., Genthner, M.H., Daniels, W.L., Groover, V.A., Nagle, S. & Haering, K.C. 2000, 'Comparison of soil and other environmental conditions in constructed and adjacent palustrine reference wetlands', Wetlands, vol. 20, no. 4, pp. 671-683. Stone, Y., Ahern, C.R. & Blunden, B. 1998, Acid Sulfate Soils Assessment Guidelines, Acid Sulfate Soil Management Advisory Committee, Wollongbar, NSW, Australia. References 210 Streever, W.J. & Genders, A.J. 1997, 'Effect of improved tidal flushing and competitive interactions at the boundary between salt marsh and pasture', Estuaries, vol. 20, no. 4, pp. 807-818. Thompson, J.R. & Finlayson, C.M. 2001, 'Freshwater wetlands', in Habitat Conservation: Managing the Physical Environment, A. Warren & J.R. French (eds), John Wiley and Sons, Chichester, West Sussex, pp. 147-178. Todd, D.K. & Mays, L.W. 2005, Groundwater Hydrology, 3rd edn, John Wiley and Sons, USA. TPS Pty Ltd 1997, Model WP-81 Conductivity-Salinity-pH-Temp. Meter. TPS Pty Ltd, Springwood, Brisbane. Tularam, G.A. & Glamore, W.C. 2004, 'Salinity intrusion in coastal flood mitigation drains and creeks', in Coastal Environment V: Incorporating Oil Spill Studies, C.A. Brebbia, J.M.S. Perez, L.G. Andion & Y.V. Esteve (eds), pp. 155-162. Tulau, M. & Henderson, S. 2001, ‘Remediation of drained acid sulfate soil backswamps in New South Wales, Australia - Technical and policy procedures’, 2nd Australian Wetland Conference. Proceedings of the Conference held 15-17 October 2001, Couran Cove Island Resort, South Stradbroke Island, Qld. Tulau, M.J. 1999, Acid Sulfate Soil Management Priority Areas in the Lower Clarence Floodplain. Sydney: Department of Land and Water Conservation. Tulau, M.J. 2007, Acid Sulfate Soils Remediation Guidelines for Coastal Floodplains in New South Wales, Department of Environment and Climate Change. Tuong, T.P. 1993, ‘An overview of water management of acid sulphate soils’, Selected Papers from the Ho Chi Minh City Symposium on Acid Sulphate Soils. Proceedings of the Conference held March 1992, International Institute for Land Reclamation and Improvement, Wageningen, The Netherlands, Ho Chi Minh City, Vietnam, pp. 265- 279. Turner, R.E. & Lewis, R.R., III. 1997, 'Hydrologic restoration of coastal wetlands', Wetlands Ecology and Management - Special Issue: Hydrologic Restoration of Coastal Wetlands, vol. 4, no. 2, pp. 65-72. Varian-Techtron 1979, Analytical Methods for Flame Spectroscopy. Varian-Techtron, Publication No. 85-100009-00, Springvale, Australia. Vega, M., Pardo, R., Barrado, E. & Deban, L. 1998, 'Assessment of seasonal and polluting effects on the quality of river water by exploratory data analysis', Water Research, vol. 32, no. 12, pp. 3581-3592. References 211 http://www.sciencedirect.com/science/article/B6V73-3VW2KW6- 9/2/e41a2d2a5fe3235f87bd01508c938e9c Walker, P.H. 1972, 'Seasonal and stratigraphic controls in coastal floodplain soils', Australian Journal of Soil Research, vol. 10, pp. 127-142. Walton, R., Chapman, R.S. & Davis, J.E. 1996, 'Development and applications of the Wetlands Dynamic Water Budget Model', Wetlands, vol. 16, no. 3, pp. 347-357. Ward, N.J., Sullivan, L.A. & Bush, R.T. 2004, 'The response of partially oxidised acid sulfate soil materials to anoxia', Australian Journal of Soil Research, vol. 42, no. 5-6, pp. 515-525. http://find.galegroup.com/ips/infomark.do?&contentSet=IAC- Documents&type=retrieve&tabID=T002&prodId=IPS&docId=A124008569&source= gale&srcprod=ITOF&userGroupName=dixson&version=1.0 Warren, R.S., Fell, P.E., Rozsa, R., Brawley, A.H., Orsted, A.C., Olson, E.T., Swamy, V. & Niering, W.A. 2002, 'Salt marsh restoration in Connecticut: 20 years of science and management', Restoration Ecology, vol. 10, no. 3, pp. 497-513. White, I., Melville, M.D., Wilson, B.P. & Sammut, J. 1997, 'Reducing acidic discharges from coastal wetlands in eastern Australia', Wetlands Ecology and Management, vol. 5, pp. 55-72. Wilkinson, G. 2003, Everlasting Swamp Acid Sulfate Soil Hot Spot Remediation Management Plan: Final Report. Grafton: Clarence River County Council. Williams, K. 2000, Assessment of Floodgated Watercourses and Drains for Management - Clarence River Coastal Floodplain. Grafton: Clarence River County Council. Williams, R. 1987, Water Quality in the Clarence River. Sydney, NSW: State Pollution Control Commission, Northern Rivers Study No. 3. Wilson, B.P., White, I. & Melville, M.D. 1999, 'Floodplain hydrology, acid discharge and change in water quality associated with a drained acid sulfate soil', Marine and Freshwater Research, vol. 50, pp. 149-157. Woodhouse, S. 2001, Wooloweyah Lagoon Yamba, NSW Management Strategy. Volume 1: Report. Grafton, Australia: Clarence River County Council, 2/664(A). Wunderlin, D.A., Diaz, M, Ame, M.V., Pesce, S.F., Hued, A.C. & Bistoni, M. 2001, 'Pattern recognition techniques for the evaluation of spatial and temporal variations in water quality. A case study: Suquia River basin (Cordoba-Argentina)', Water Research, vol. 35, no. 12, pp. 2881-2894. http://www.sciencedirect.com/science/article/B6V73-43CJX46- 9/2/2b9e478d0ddd3d1f5c65b14f66e9758d References 212 Zar, J.H. 1999, Biostatistical Analysis, 4th edn, Prentice Hall, Upper Saddle River, New Jersey, USA. Zhang, L. & Mitsch, W.J. 2005, 'Modelling hydrological processes in created freshwater wetlands: An integrated system approach', Environmental Modelling & Software, vol. 20, no. 7, pp. 935-946. http://www.sciencedirect.com/science/article/B6VHC-4CYR71R- 1/2/d5f54fc846d6a29af0efb6e2778abade Zhou, F., Guo, H., Liu, Y. & Jiang, Y. 2007, 'Chemometrics data analysis of marine water quality and source identification in southern Hong Kong', Marine Pollution Bulletin, vol. 54, no. 6, pp. 745-756. http://www.sciencedirect.com/science/article/B6V6N-4N4431B- 1/2/b8af33c2250a7390a3544cc1dc8cf68f 213 Appendix A Table A1: Results of factor analysis of all cases for the pilot study at Little Broadwater. Six VFs were identified and the principal constituents for each are underlined. Parameter VF1 VF2 VF3 VF4 VF5 VF6 Water temp 0.058 -0.231 -0.011 -0.051 -0.267 0.782 EC 0.209 0.805 -0.237 0.187 0.045 -0.231 pH -0.639 0.168 -0.176 0.083 0.575 -0.081 DO 0.035 0.098 -0.120 -0.096 0.184 0.854 F -0.016 0.568 0.039 -0.112 -0.259 0.282 Cl- 0.550 0.738 0.028 0.069 0.037 -0.161 NO2 - -0.032 0.029 -0.104 0.884 0.007 -0.103 Br 0.915 0.258 0.015 -0.064 -0.060 -0.011 SO4 2- 0.143 0.888 0.138 0.054 0.017 -0.082 NO3 - 0.028 -0.324 -0.025 -0.245 0.280 -0.109 Al 0.859 0.113 0.179 -0.031 -0.237 0.082 As -0.057 -0.164 0.714 -0.047 -0.163 -0.059 Ca 0.826 0.198 0.417 0.072 0.148 0.009 Cd 0.919 0.082 -0.166 -0.011 0.038 -0.014 Co 0.965 0.065 0.113 -0.024 -0.097 0.015 Cr 0.881 0.132 0.100 0.095 0.119 -0.098 Cu 0.600 -0.252 0.059 0.004 -0.429 0.206 Fe 0.979 0.092 0.006 -0.024 -0.072 0.020 K 0.713 0.266 0.419 0.280 0.347 -0.034 Mg 0.432 0.097 0.599 0.343 0.453 -0.002 Mn 0.966 0.066 0.017 -0.030 0.000 -0.023 Na 0.672 0.306 0.491 0.263 0.315 -0.020 Ni 0.936 0.165 0.203 0.074 -0.050 0.056 P -0.004 0.107 0.097 0.908 0.049 -0.060 Pb 0.556 -0.001 0.059 -0.043 0.287 -0.055 S 0.936 0.168 0.154 -0.019 0.008 0.018 Sb -0.076 -0.198 0.091 0.028 0.611 0.031 Se 0.381 0.169 0.684 -0.083 0.202 -0.086 Zn 0.845 -0.082 -0.015 0.017 -0.278 0.126 Eigenvalue 11.92 3.07 2.23 2.04 1.91 1.63 % of Variance 41.1 10.6 7.7 7.0 6.6 5.6 Cumulative % 41.1 51.7 59.4 66.4 73.0 78.6 Appendices 214 Figure A1: Results of cluster analysis of samples from Little Broadwater, November 2004. Figure A2: Results of cluster analysis of samples from Little Broadwater, December 2004. Figure A3: Results of cluster analysis of samples from Little Broadwater, February 2005. A pp en di ce s 21 5 T ab le A 2: F ac to r an al ys is o f ea ch s am pl e pe ri od d ur in g th e pi lo t s tu dy a t L it tl e B ro ad w at er . N ov em be r 20 04 D ec em be r 20 04 Fe br ua ry 2 00 5 P ar am et er V F 1 V F 2 V F 3 V F 4 V F 5 V F 6 V F 1 V F 2 V F 3 V F 4 V F 5 V F 6 V F 1 V F 2 V F 3 V F 4 V F 5 V F 6 W at er te m p. 0. 01 -0 .1 1 0. 69 0. 16 0. 32 -0 .1 9 0. 51 -0 .0 8 0. 18 -0 .4 1 -0 .1 4 0. 35 -0 .4 5 0. 10 -0 .7 2 -0 .0 7 0. 00 0. 26 E C 0. 63 0. 18 0. 67 0. 07 -0 .0 4 -0 .0 9 0. 36 0. 81 0. 35 0. 14 0. 08 0. 11 0. 97 0. 14 -0 .0 1 0. 00 0. 02 0. 02 pH -0 .8 7 0. 17 -0 .0 4 0. 07 0. 35 0. 05 -0 .8 4 -0 .2 5 -0 .3 6 0. 08 -0 .1 4 0. 03 -0 .0 1 -0 .9 4 -0 .0 6 0. 06 -0 .0 3 0. 12 D O 0. 08 0. 48 0. 26 -0 .2 7 0. 72 0. 13 -0 .1 4 0. 56 -0 .0 5 -0 .2 0 -0 .3 4 -0 .4 1 -0 .3 1 0. 10 -0 .5 8 -0 .1 0 0. 55 0. 12 F- -0 .1 3 -0 .0 2 0. 84 -0 .1 2 -0 .1 4 0. 08 0. 95 0. 19 -0 .1 5 0. 08 0. 07 -0 .0 1 -0 .0 1 -0 .1 1 -0 .1 1 0. 93 -0 .2 0 0. 04 C l- 0. 73 -0 .0 4 0. 56 -0 .0 6 0. 00 -0 .0 4 0. 39 0. 82 0. 38 0. 10 0. 01 0. 08 0. 96 0. 13 0. 10 0. 03 -0 .0 9 0. 03 N O 2- -0 .0 3 -0 .0 5 -0 .0 6 0. 90 -0 .0 1 0. 36 -0 .0 5 0. 02 -0 .0 3 0. 89 -0 .0 1 -0 .1 1 B r 0. 98 0. 12 0. 11 -0 .0 8 0. 01 0. 01 0. 85 0. 23 0. 41 -0 .1 2 0. 02 0. 05 0. 70 0. 27 0. 32 -0 .0 8 -0 .2 6 -0 .1 5 SO 42 - 0. 54 0. 07 0. 78 -0 .0 8 0. 12 -0 .0 9 0. 07 0. 80 -0 .3 5 -0 .0 4 0. 19 0. 06 0. 99 0. 00 0. 01 0. 01 0. 06 0. 07 N O 3- -0 .0 7 -0 .2 3 -0 .0 2 0. 03 0. 88 -0 .0 5 0. 08 -0 .4 4 0. 18 -0 .2 0 -0 .2 2 -0 .6 8 -0 .1 1 -0 .1 1 0. 80 -0 .2 8 -0 .0 5 0. 37 A l 0. 99 0. 10 0. 07 -0 .0 3 0. 01 -0 .0 3 0. 95 0. 21 0. 17 -0 .0 3 0. 07 0. 00 -0 .0 8 0. 89 -0 .1 9 -0 .0 6 -0 .1 7 0. 25 A s -0 .0 8 0. 69 -0 .1 0 0. 50 -0 .1 4 -0 .2 6 0. 00 0. 09 -0 .1 5 -0 .1 2 0. 77 0. 21 0. 01 0. 20 0. 05 0. 26 -0 .1 1 0. 80 C a 0. 85 0. 47 0. 05 0. 09 -0 .0 1 0. 16 0. 70 0. 58 0. 37 -0 .0 8 0. 08 0. 01 0. 96 0. 06 0. 21 -0 .0 1 0. 00 -0 .0 2 C d 0. 99 0. 08 0. 08 -0 .0 3 0. 01 -0 .0 4 0. 28 0. 15 0. 94 -0 .0 5 -0 .0 1 0. 06 C o 0. 98 0. 14 0. 08 -0 .0 1 0. 01 -0 .0 2 0. 70 0. 27 0. 66 -0 .0 5 0. 05 0. 04 0. 38 0. 84 0. 15 -0 .1 3 -0 .1 7 -0 .0 7 C r 0. 98 0. 09 0. 08 0. 12 0. 03 -0 .0 6 0. 49 0. 44 0. 59 0. 13 -0 .0 4 0. 17 C u 0. 98 0. 11 0. 07 -0 .0 5 0. 01 0. 01 0. 28 0. 12 0. 93 -0 .0 6 -0 .0 3 0. 04 F e 0. 99 0. 10 0. 07 -0 .0 3 0. 01 -0 .0 2 0. 69 0. 21 0. 69 -0 .0 4 0. 02 0. 03 0. 14 0. 81 -0 .1 1 -0 .0 6 -0 .1 7 0. 20 K 0. 70 0. 62 0. 01 0. 20 0. 01 2 0. 24 6 0. 42 4 0. 76 8 0. 43 3 0. 13 3 0. 07 4 0. 06 5 0. 97 5 0. 00 8 0. 08 1 0. 00 9 0. 04 6 0. 03 6 M g 0. 35 0. 80 -0 .0 6 0. 29 -0 .0 12 0. 33 7 0. 30 3 0. 81 0 0. 28 8 0. 25 4 0. 10 4 0. 02 4 0. 98 5 0. 04 2 0. 01 5 0. 03 2 0. 10 8 0. 02 8 M n 0. 99 0. 13 0. 07 -0 .0 2 0. 01 -0 .0 2 0. 53 0. 25 0. 80 -0 .0 6 0. 03 0. 05 0. 70 0. 29 0. 46 0. 00 0. 07 -0 .3 0 N a 0. 64 0. 67 0. 03 0. 21 -0 .0 1 0. 27 0. 46 0. 81 0. 31 0. 10 0. 06 0. 05 0. 99 0. 04 0. 04 0. 00 0. 05 0. 06 A pp en di ce s 21 6 N ov em be r 20 04 D ec em be r 20 04 Fe br ua ry 2 00 5 P ar am et er V F 1 V F 2 V F 3 V F 4 V F 5 V F 6 V F 1 V F 2 V F 3 V F 4 V F 5 V F 6 V F 1 V F 2 V F 3 V F 4 V F 5 V F 6 N i 0. 93 0. 33 0. 04 0. 08 0. 00 0. 12 0. 81 0. 35 0. 45 -0 .0 3 0. 07 0. 04 0. 43 0. 79 0. 03 0. 03 0. 25 0. 11 P -0 .0 1 0. 31 0. 04 0. 90 -0 .0 4 -0 .1 3 -0 .0 8 0. 13 -0 .0 5 0. 96 0. 05 0. 18 0. 08 -0 .1 9 -0 .0 3 -0 .0 5 0. 86 -0 .1 8 P b 0. 94 0. 15 0. 06 -0 .0 3 0. 03 0. 00 0. 03 0. 09 0. 50 -0 .1 5 -0 .1 0 0. 68 S 0. 97 0. 23 0. 06 -0 .0 4 0. 01 0. 05 0. 86 0. 30 0. 38 -0 .0 8 0. 09 0. 02 0. 87 0. 23 0. 27 -0 .0 7 -0 .0 2 -0 .1 1 Sb -0 .0 8 0. 13 -0 .1 3 0. 09 0. 00 0. 92 -0 .2 2 -0 .1 2 -0 .1 8 -0 .2 1 -0 .7 9 0. 19 -0 .0 3 -0 .1 0 0. 05 0. 85 0. 12 0. 19 S e 0. 18 0. 88 0. 03 -0 .1 0 -0 .0 2 -0 .0 2 0. 31 0. 68 0. 21 -0 .2 3 0. 03 0. 33 0. 42 0. 34 0. 49 0. 22 0. 02 0. 31 Z n 0. 99 0. 10 0. 08 0. 05 0. 00 -0 .0 5 0. 81 0. 21 0. 55 -0 .0 4 0. 05 0. 03 -0 .4 2 0. 47 0. 01 -0 .0 5 -0 .4 3 -0 .2 6 E ig en va lu e 15 .2 3. 7 2. 7 2. 2 1. 6 1. 5 8. 6 6. 1 5. 9 2. 3 1. 5 1. 5 9. 4 4. 4 2. 3 1. 8 1. 6 1. 4 % o f V ar ia nc e 52 .4 12 .8 9. 4 7. 7 5. 4 5. 0 29 .8 21 .0 20 .5 7. 8 5. 3 5. 2 39 .1 18 .2 9. 5 7. 7 6. 4 5. 7 C um ul at iv e % 52 .4 65 .2 74 .6 82 .3 87 .7 92 .7 29 .8 50 .8 71 .3 79 .1 84 .4 89 .6 39 .1 57 .3 66 .8 74 .5 80 .9 86 .6 217 Appendix B Minitab readout for cluster analysis of all cases, April 2005 to February 2007: Cluster Analysis of Observations: pH, Temp, EC, DO, Al, Ca, Cl, Fe, K, Mg, Mn, Na, SO, TN, TP Euclidean Distance, Ward Linkage Amalgamation Steps Step Number of Similarity Distance Clusters New Number of obs. clusters level level joined cluster in new cluster 1 513 99.77 0.046 480 481 480 2 2 512 99.20 0.161 445 446 445 2 3 511 99.09 0.183 487 508 487 2 4 510 99.09 0.184 479 489 479 2 5 509 98.83 0.236 93 107 93 2 6 508 98.77 0.248 212 214 212 2 7 507 98.74 0.254 256 272 256 2 8 506 98.73 0.255 257 259 257 2 9 505 98.73 0.257 393 394 393 2 10 504 98.72 0.258 260 267 260 2 11 503 98.64 0.274 97 101 97 2 12 502 98.64 0.274 477 478 477 2 13 501 98.62 0.279 474 475 474 2 14 500 98.60 0.281 442 443 442 2 15 499 98.58 0.285 404 406 404 2 16 498 98.56 0.289 452 497 452 2 17 497 98.56 0.291 494 498 494 2 18 496 98.55 0.293 398 405 398 2 19 495 98.52 0.298 274 283 274 2 20 494 98.50 0.301 210 211 210 2 21 493 98.50 0.302 428 429 428 2 22 492 98.48 0.307 506 507 506 2 23 491 98.48 0.307 273 275 273 2 24 490 98.35 0.332 341 342 341 2 25 489 98.34 0.334 274 285 274 3 26 488 98.33 0.336 249 250 249 2 27 487 98.33 0.337 206 207 206 2 28 486 98.30 0.342 430 434 430 2 29 485 98.27 0.349 386 387 386 2 30 484 98.26 0.350 280 281 280 2 31 483 98.26 0.350 417 419 417 2 32 482 98.26 0.351 222 227 222 2 33 481 98.22 0.359 93 98 93 3 34 480 98.18 0.367 257 258 257 3 35 479 98.18 0.367 459 460 459 2 36 478 98.17 0.369 407 409 407 2 37 477 98.16 0.370 477 482 477 3 38 476 98.16 0.370 504 505 504 2 39 475 98.15 0.373 397 400 397 2 40 474 98.14 0.375 299 302 299 2 41 473 98.12 0.378 97 106 97 3 42 472 98.08 0.386 212 213 212 3 43 471 98.06 0.391 265 282 265 2 44 470 98.06 0.391 393 401 393 3 45 469 98.02 0.399 112 113 112 2 46 468 98.01 0.401 486 488 486 2 47 467 97.97 0.408 99 109 99 2 48 466 97.95 0.412 391 432 391 2 49 465 97.94 0.415 252 253 252 2 50 464 97.93 0.417 435 436 435 2 51 463 97.93 0.417 273 474 273 4 Appendices 218 52 462 97.90 0.424 484 506 484 3 53 461 97.88 0.427 421 426 421 2 54 460 97.88 0.427 269 490 269 2 55 459 97.83 0.438 223 485 223 2 56 458 97.82 0.439 397 399 397 3 57 457 97.79 0.444 395 451 395 2 58 456 97.78 0.447 261 268 261 2 59 455 97.78 0.447 447 479 447 3 60 454 97.73 0.457 296 310 296 2 61 453 97.73 0.458 255 256 255 3 62 452 97.72 0.459 215 449 215 2 63 451 97.70 0.463 111 112 111 3 64 450 97.67 0.470 295 303 295 2 65 449 97.66 0.471 296 311 296 3 66 448 97.66 0.472 392 393 392 4 67 447 97.66 0.472 326 327 326 2 68 446 97.60 0.484 299 301 299 3 69 445 97.59 0.486 492 504 492 3 70 444 97.57 0.490 404 407 404 4 71 443 97.56 0.491 279 280 279 3 72 442 97.55 0.493 235 236 235 2 73 441 97.54 0.495 93 95 93 4 74 440 97.54 0.495 270 271 270 2 75 439 97.54 0.496 92 100 92 2 76 438 97.54 0.496 286 287 286 2 77 437 97.52 0.499 447 480 447 5 78 436 97.49 0.506 346 347 346 2 79 435 97.48 0.509 395 431 395 3 80 434 97.44 0.517 430 437 430 3 81 433 97.42 0.521 257 260 257 5 82 432 97.39 0.526 216 222 216 3 83 431 97.38 0.528 225 231 225 2 84 430 97.37 0.530 398 487 398 4 85 429 97.36 0.532 125 126 125 2 86 428 97.36 0.533 315 317 315 2 87 427 97.35 0.535 373 374 373 2 88 426 97.29 0.546 206 483 206 3 89 425 97.29 0.547 450 486 450 3 90 424 97.27 0.551 249 251 249 3 91 423 97.26 0.552 97 105 97 4 92 422 97.26 0.552 104 111 104 4 93 421 97.25 0.554 354 355 354 2 94 420 97.24 0.556 122 123 122 2 95 419 97.20 0.563 248 264 248 2 96 418 97.14 0.576 206 208 206 4 97 417 97.13 0.578 391 433 391 3 98 416 97.12 0.580 452 453 452 3 99 415 97.10 0.584 288 292 288 2 100 414 97.07 0.591 438 476 438 2 101 413 97.05 0.594 396 410 396 2 102 412 97.05 0.595 388 390 388 2 103 411 97.03 0.599 99 110 99 3 104 410 97.02 0.600 225 228 225 3 105 409 97.00 0.604 217 262 217 2 106 408 96.98 0.608 377 381 377 2 107 407 96.91 0.622 305 313 305 2 108 406 96.90 0.625 340 341 340 3 109 405 96.89 0.626 212 216 212 6 110 404 96.89 0.627 94 102 94 2 111 403 96.85 0.634 323 325 323 2 112 402 96.85 0.635 395 477 395 6 113 401 96.84 0.636 402 403 402 2 114 400 96.82 0.641 439 442 439 3 115 399 96.79 0.646 263 276 263 2 116 398 96.78 0.648 265 274 265 5 117 397 96.70 0.664 159 203 159 2 118 396 96.64 0.677 413 414 413 2 119 395 96.63 0.679 25 27 25 2 120 394 96.58 0.689 296 298 296 4 121 393 96.50 0.705 210 492 210 5 122 392 96.46 0.712 224 269 224 3 Appendices 219 123 391 96.45 0.714 10 11 10 2 124 390 96.44 0.718 385 386 385 3 125 389 96.39 0.728 439 441 439 4 126 388 96.37 0.732 396 404 396 6 127 387 96.37 0.732 461 491 461 2 128 386 96.34 0.737 103 116 103 2 129 385 96.33 0.739 252 255 252 5 130 384 96.31 0.743 240 241 240 2 131 383 96.31 0.744 417 420 417 3 132 382 96.29 0.748 430 435 430 5 133 381 96.27 0.752 158 159 158 3 134 380 96.19 0.767 229 288 229 3 135 379 96.17 0.772 254 270 254 3 136 378 96.17 0.772 226 230 226 2 137 377 96.17 0.772 163 171 163 2 138 376 96.13 0.779 349 360 349 2 139 375 96.06 0.794 206 215 206 6 140 374 96.03 0.801 48 50 48 2 141 373 96.02 0.802 114 115 114 2 142 372 96.02 0.802 455 503 455 2 143 371 95.97 0.812 314 316 314 2 144 370 95.97 0.813 20 21 20 2 145 369 95.96 0.813 295 304 295 3 146 368 95.96 0.815 459 462 459 3 147 367 95.94 0.817 428 494 428 4 148 366 95.94 0.819 289 291 289 2 149 365 95.93 0.821 266 284 266 2 150 364 95.91 0.825 332 334 332 2 151 363 95.87 0.831 427 470 427 2 152 362 95.81 0.843 388 418 388 3 153 361 95.74 0.857 128 132 128 2 154 360 95.71 0.865 223 450 223 5 155 359 95.69 0.868 331 336 331 2 156 358 95.69 0.869 368 376 368 2 157 357 95.60 0.887 229 278 229 4 158 356 95.57 0.892 351 354 351 3 159 355 95.57 0.893 179 180 179 2 160 354 95.53 0.901 396 398 396 10 161 353 95.53 0.901 346 356 346 3 162 352 95.51 0.905 329 330 329 2 163 351 95.49 0.908 444 461 444 3 164 350 95.48 0.910 395 447 395 11 165 349 95.47 0.912 257 261 257 7 166 348 95.38 0.930 88 90 88 2 167 347 95.36 0.935 56 58 56 2 168 346 95.34 0.938 322 367 322 2 169 345 95.34 0.938 286 509 286 3 170 344 95.30 0.947 245 247 245 2 171 343 95.29 0.948 397 408 397 4 172 342 95.27 0.954 122 124 122 3 173 341 95.26 0.955 195 196 195 2 174 340 95.25 0.957 237 238 237 2 175 339 95.24 0.958 273 428 273 8 176 338 95.22 0.964 163 166 163 3 177 337 95.21 0.964 456 457 456 2 178 336 95.21 0.965 94 103 94 4 179 335 95.13 0.981 118 120 118 2 180 334 95.12 0.983 51 68 51 2 181 333 95.09 0.989 54 69 54 2 182 332 95.08 0.992 22 23 22 2 183 331 95.08 0.992 243 245 243 3 184 330 95.07 0.992 212 452 212 9 185 329 95.06 0.995 232 234 232 2 186 328 95.05 0.996 59 62 59 2 187 327 95.05 0.997 333 335 333 2 188 326 95.02 1.002 4 6 4 2 189 325 95.02 1.003 158 204 158 4 190 324 95.01 1.004 125 127 125 3 191 323 95.00 1.006 136 138 136 2 192 322 94.84 1.039 223 445 223 7 193 321 94.83 1.042 277 289 277 3 Appendices 220 194 320 94.81 1.045 290 463 290 2 195 319 94.81 1.046 210 484 210 8 196 318 94.81 1.046 244 246 244 2 197 317 94.81 1.046 307 308 307 2 198 316 94.79 1.049 108 359 108 2 199 315 94.79 1.049 448 495 448 2 200 314 94.74 1.060 496 500 496 2 201 313 94.73 1.062 343 385 343 4 202 312 94.69 1.070 402 438 402 4 203 311 94.66 1.076 323 324 323 3 204 310 94.64 1.079 233 239 233 2 205 309 94.62 1.083 96 114 96 3 206 308 94.62 1.084 45 46 45 2 207 307 94.59 1.091 10 13 10 3 208 306 94.55 1.098 254 265 254 8 209 305 94.51 1.106 348 350 348 2 210 304 94.48 1.113 331 370 331 3 211 303 94.42 1.123 413 415 413 3 212 302 94.39 1.130 391 392 391 7 213 301 94.31 1.146 319 349 319 3 214 300 94.30 1.149 314 320 314 3 215 299 94.29 1.149 422 423 422 2 216 298 94.29 1.150 197 202 197 2 217 297 94.24 1.161 402 412 402 5 218 296 94.23 1.162 361 364 361 2 219 295 94.22 1.165 384 417 384 4 220 294 94.21 1.166 63 71 63 2 221 293 94.21 1.167 88 89 88 3 222 292 94.19 1.170 266 286 266 5 223 291 94.19 1.170 240 242 240 3 224 290 94.19 1.171 232 235 232 4 225 289 94.17 1.174 24 25 24 3 226 288 94.15 1.178 345 346 345 4 227 287 94.12 1.185 368 372 368 3 228 286 94.07 1.195 223 224 223 10 229 285 94.00 1.208 309 318 309 2 230 284 94.00 1.210 92 93 92 6 231 283 93.96 1.216 295 299 295 6 232 282 93.94 1.221 49 72 49 2 233 281 93.88 1.232 249 252 249 8 234 280 93.88 1.233 92 97 92 10 235 279 93.87 1.235 78 79 78 2 236 278 93.71 1.268 99 104 99 7 237 277 93.69 1.271 4 17 4 3 238 276 93.66 1.277 448 455 448 4 239 275 93.63 1.284 314 365 314 4 240 274 93.62 1.285 373 375 373 3 241 273 93.62 1.285 15 19 15 2 242 272 93.62 1.286 305 312 305 3 243 271 93.61 1.287 2 168 2 2 244 270 93.61 1.288 444 473 444 4 245 269 93.58 1.292 192 193 192 2 246 268 93.54 1.302 161 164 161 2 247 267 93.52 1.306 48 51 48 4 248 266 93.50 1.310 464 468 464 2 249 265 93.50 1.310 80 81 80 2 250 264 93.48 1.312 172 177 172 2 251 263 93.40 1.329 332 369 332 3 252 262 93.39 1.331 135 137 135 2 253 261 93.37 1.335 464 469 464 3 254 260 93.34 1.341 296 306 296 5 255 259 93.28 1.353 118 121 118 3 256 258 93.25 1.361 182 185 182 2 257 257 93.24 1.362 64 65 64 2 258 256 93.22 1.365 300 383 300 2 259 255 93.20 1.369 41 43 41 2 260 254 93.18 1.373 94 108 94 6 261 253 93.02 1.405 229 279 229 7 262 252 93.01 1.408 152 154 152 2 263 251 92.98 1.413 226 237 226 4 264 250 92.91 1.429 340 344 340 4 Appendices 221 265 249 92.88 1.434 332 333 332 5 266 248 92.88 1.435 1 169 1 2 267 247 92.84 1.442 128 130 128 3 268 246 92.84 1.443 20 28 20 3 269 245 92.83 1.445 243 244 243 5 270 244 92.79 1.452 175 186 175 2 271 243 92.79 1.453 294 338 294 2 272 242 92.78 1.454 53 67 53 2 273 241 92.71 1.468 368 371 368 4 274 240 92.71 1.468 178 458 178 2 275 239 92.71 1.469 16 22 16 3 276 238 92.67 1.476 172 176 172 3 277 237 92.52 1.507 212 225 212 12 278 236 92.47 1.517 309 315 309 4 279 235 92.32 1.547 206 430 206 11 280 234 92.17 1.578 12 35 12 2 281 233 92.13 1.585 143 144 143 2 282 232 92.12 1.587 49 57 49 3 283 231 92.12 1.588 427 496 427 4 284 230 92.08 1.596 56 74 56 3 285 229 92.08 1.596 416 421 416 3 286 228 92.02 1.607 219 263 219 3 287 227 92.00 1.612 189 197 189 3 288 226 91.99 1.614 226 277 226 7 289 225 91.96 1.620 465 466 465 2 290 224 91.92 1.629 119 363 119 2 291 223 91.87 1.638 337 353 337 2 292 222 91.87 1.638 183 200 183 2 293 221 91.85 1.643 296 319 296 8 294 220 91.78 1.656 34 36 34 2 295 219 91.75 1.662 129 131 129 2 296 218 91.74 1.665 439 440 439 5 297 217 91.69 1.675 248 249 248 10 298 216 91.68 1.676 217 220 217 3 299 215 91.62 1.688 427 472 427 5 300 214 91.62 1.688 181 194 181 2 301 213 91.60 1.693 198 199 198 2 302 212 91.58 1.696 133 136 133 3 303 211 91.53 1.707 331 377 331 5 304 210 91.50 1.713 82 85 82 2 305 209 91.48 1.716 378 380 378 2 306 208 91.43 1.726 368 373 368 7 307 207 91.39 1.734 14 15 14 3 308 206 91.38 1.737 510 513 510 2 309 205 91.14 1.785 422 425 422 3 310 204 90.84 1.845 140 141 140 2 311 203 90.82 1.849 96 122 96 6 312 202 90.74 1.865 160 218 160 2 313 201 90.73 1.868 158 209 158 5 314 200 90.70 1.873 205 221 205 2 315 199 90.69 1.875 53 59 53 4 316 198 90.63 1.887 388 391 388 10 317 197 90.48 1.918 117 119 117 3 318 196 90.46 1.922 161 163 161 5 319 195 90.39 1.937 48 66 48 5 320 194 90.34 1.946 8 10 8 4 321 193 90.32 1.950 165 178 165 3 322 192 90.31 1.952 328 329 328 3 323 191 90.27 1.960 300 382 300 3 324 190 90.25 1.963 173 175 173 3 325 189 90.15 1.984 40 44 40 2 326 188 90.08 1.999 223 396 223 20 327 187 90.07 2.000 16 26 16 4 328 186 89.97 2.021 257 273 257 15 329 185 89.96 2.023 30 454 30 2 330 184 89.78 2.059 297 389 297 2 331 183 89.77 2.060 309 326 309 6 332 182 89.77 2.061 54 63 54 4 333 181 89.73 2.069 181 190 181 3 334 180 89.65 2.085 76 87 76 2 335 179 89.59 2.097 118 125 118 6 Appendices 222 336 178 89.56 2.103 37 38 37 2 337 177 89.54 2.108 192 233 192 4 338 176 89.38 2.140 149 150 149 2 339 175 89.29 2.158 345 348 345 6 340 174 89.24 2.167 254 266 254 13 341 173 89.24 2.168 133 135 133 5 342 172 89.22 2.171 314 361 314 6 343 171 89.15 2.186 339 368 339 8 344 170 89.13 2.190 183 201 183 3 345 169 88.98 2.221 91 92 91 11 346 168 88.96 2.223 2 172 2 5 347 167 88.87 2.243 55 61 55 2 348 166 88.85 2.245 337 427 337 7 349 165 88.74 2.267 18 148 18 2 350 164 88.58 2.300 54 78 54 6 351 163 88.54 2.308 160 217 160 5 352 162 88.54 2.309 151 152 151 3 353 161 88.46 2.324 206 448 206 15 354 160 88.44 2.329 60 82 60 3 355 159 88.40 2.337 140 147 140 3 356 158 88.26 2.365 502 510 502 3 357 157 88.26 2.366 290 411 290 3 358 156 88.06 2.406 357 358 357 2 359 155 88.00 2.416 402 444 402 9 360 154 88.00 2.418 165 179 165 5 361 153 87.97 2.423 151 153 151 4 362 152 87.91 2.435 424 465 424 3 363 151 87.87 2.443 210 395 210 19 364 150 87.73 2.472 384 413 384 7 365 149 87.70 2.477 1 195 1 4 366 148 87.65 2.488 323 331 323 8 367 147 87.61 2.496 402 459 402 12 368 146 87.59 2.500 54 70 54 7 369 145 87.57 2.504 31 129 31 3 370 144 87.46 2.525 184 189 184 4 371 143 87.30 2.558 32 33 32 2 372 142 87.25 2.568 4 24 4 6 373 141 87.24 2.570 511 512 511 2 374 140 87.12 2.594 322 456 322 4 375 139 87.12 2.595 91 99 91 18 376 138 87.03 2.613 340 343 340 8 377 137 86.73 2.673 12 30 12 4 378 136 86.36 2.748 305 314 305 9 379 135 86.23 2.773 219 254 219 16 380 134 86.18 2.784 378 379 378 3 381 133 86.07 2.805 170 174 170 2 382 132 85.78 2.865 321 328 321 4 383 131 85.67 2.886 37 39 37 3 384 130 85.64 2.892 143 149 143 4 385 129 85.48 2.925 493 499 493 2 386 128 85.45 2.931 167 181 167 4 387 127 85.34 2.953 162 226 162 8 388 126 85.05 3.012 7 9 7 2 389 125 84.60 3.101 88 296 88 11 390 124 84.54 3.114 502 514 502 4 391 123 84.52 3.119 165 240 165 8 392 122 84.48 3.126 14 20 14 6 393 121 84.36 3.151 47 117 47 4 394 120 84.23 3.177 40 42 40 3 395 119 83.91 3.242 45 73 45 3 396 118 83.88 3.248 94 96 94 12 397 117 83.76 3.271 134 206 134 16 398 116 83.76 3.272 397 439 397 9 399 115 83.49 3.326 183 184 183 7 400 114 83.44 3.335 76 83 76 3 401 113 83.39 3.347 133 139 133 6 402 112 83.32 3.361 12 31 12 7 403 111 83.25 3.374 56 75 56 4 404 110 83.18 3.388 384 416 384 10 405 109 83.07 3.411 293 294 293 3 406 108 82.78 3.469 300 388 300 13 Appendices 223 407 107 82.11 3.603 188 191 188 2 408 106 82.10 3.605 3 352 3 2 409 105 82.02 3.622 48 49 48 8 410 104 81.65 3.696 323 332 323 13 411 103 81.37 3.752 1 167 1 8 412 102 81.16 3.796 88 295 88 17 413 101 81.01 3.826 118 128 118 9 414 100 80.92 3.844 182 243 182 7 415 99 80.73 3.882 158 162 158 13 416 98 80.51 3.926 183 188 183 9 417 97 80.22 3.984 210 257 210 34 418 96 80.09 4.010 160 192 160 9 419 95 80.05 4.019 156 157 156 2 420 94 79.94 4.040 145 155 145 2 421 93 79.86 4.057 47 322 47 8 422 92 79.74 4.081 205 501 205 3 423 91 79.72 4.086 34 37 34 5 424 90 79.64 4.101 18 140 18 5 425 89 79.04 4.222 4 16 4 10 426 88 78.73 4.286 229 232 229 11 427 87 78.46 4.339 290 493 290 5 428 86 78.46 4.340 223 402 223 32 429 85 78.17 4.397 161 165 161 13 430 84 78.00 4.433 91 94 91 30 431 83 77.98 4.435 40 41 40 5 432 82 77.93 4.446 422 464 422 6 433 81 77.90 4.451 12 18 12 12 434 80 77.41 4.551 53 56 53 8 435 79 77.34 4.565 8 133 8 10 436 78 77.31 4.572 1 173 1 11 437 77 77.21 4.590 32 52 32 3 438 76 77.18 4.598 187 198 187 3 439 75 76.94 4.645 337 471 337 8 440 74 76.84 4.665 297 300 297 15 441 73 76.21 4.792 205 293 205 6 442 72 76.21 4.793 134 212 134 28 443 71 75.99 4.837 47 309 47 14 444 70 75.90 4.855 60 77 60 4 445 69 75.44 4.948 29 366 29 2 446 68 75.04 5.029 55 146 55 3 447 67 74.96 5.045 5 170 5 3 448 66 74.17 5.204 48 64 48 10 449 65 73.65 5.307 53 54 53 15 450 64 72.82 5.476 143 145 143 6 451 63 72.40 5.560 205 337 205 14 452 62 71.72 5.696 76 84 76 4 453 61 70.94 5.854 158 397 158 22 454 60 70.24 5.994 502 511 502 6 455 59 70.13 6.016 88 340 88 25 456 58 70.13 6.017 1 2 1 16 457 57 69.85 6.073 323 339 323 21 458 56 68.91 6.264 422 424 422 9 459 55 68.74 6.298 29 378 29 5 460 54 68.52 6.342 345 351 345 9 461 53 68.51 6.344 4 14 4 16 462 52 67.49 6.549 219 248 219 26 463 51 67.03 6.642 47 321 47 18 464 50 66.42 6.764 60 80 60 6 465 49 66.35 6.779 86 156 86 3 466 48 66.12 6.824 7 34 7 7 467 47 64.28 7.196 4 12 4 28 468 46 63.14 7.426 47 307 47 20 469 45 62.29 7.597 134 210 134 62 470 44 62.07 7.641 160 161 160 22 471 43 61.45 7.766 118 305 118 18 472 42 60.25 8.008 187 467 187 4 473 41 59.89 8.080 86 142 86 4 474 40 56.73 8.716 158 290 158 27 475 39 56.68 8.727 29 45 29 8 476 38 55.68 8.928 345 384 345 19 477 37 54.97 9.071 160 229 160 33 Appendices 224 478 36 54.85 9.095 88 297 88 40 479 35 48.99 10.275 143 151 143 10 480 34 48.30 10.415 53 55 53 18 481 33 47.58 10.559 5 182 5 10 482 32 45.42 10.995 32 40 32 8 483 31 43.35 11.412 158 223 158 59 484 30 37.83 12.523 134 219 134 88 485 29 35.40 13.013 357 362 357 3 486 28 34.64 13.167 91 345 91 49 487 27 33.62 13.372 3 183 3 11 488 26 33.42 13.412 60 76 60 10 489 25 30.60 13.980 8 48 8 20 490 24 29.48 14.207 29 422 29 17 491 23 26.59 14.789 47 323 47 41 492 22 14.08 17.307 32 143 32 18 493 21 12.49 17.629 88 118 88 58 494 20 -3.83 20.917 3 187 3 15 495 19 -7.07 21.568 5 29 5 27 496 18 -7.91 21.738 4 7 4 35 497 17 -16.04 23.376 8 53 8 38 498 16 -20.62 24.299 32 357 32 21 499 15 -21.91 24.558 1 160 1 49 500 14 -33.55 26.903 32 86 32 25 501 13 -44.38 29.085 4 8 4 73 502 12 -45.72 29.355 88 205 88 72 503 11 -56.10 31.447 32 60 32 35 504 10 -64.29 33.096 134 158 134 147 505 9 -95.22 39.327 5 47 5 68 506 8 -108.66 42.033 88 91 88 121 507 7 -110.86 42.477 1 5 1 117 508 6 -215.68 63.594 3 502 3 21 509 5 -278.90 76.328 4 32 4 108 510 4 -280.40 76.631 1 3 1 138 511 3 -313.02 83.202 88 134 88 268 512 2 -1053.32 232.335 1 88 1 406 513 1 -1735.10 369.676 1 4 1 514 Final Partition Number of clusters: 1 Number of Within cluster Average distance Maximum distance observations sum of squares from centroid from centroid Cluster1 514 7696.615 3.309 17.029 225 Appendix C Table C1: Results of factor analysis of all samples for the long-term study at Little Broadwater. Three VFs were identified and the named according to the principal constituents (underlined). Parameter VF1 – Salts VF2 – ASS VF3 – Nutrients pH 0.462 -0.715 0.009 Temp 0.157 0.386 -0.369 EC 0.902 0.019 0.048 DO 0.689 -0.010 -0.215 Al 0.137 0.641 -0.112 Ca 0.814 0.181 -0.021 Cl 0.928 0.032 -0.015 Fe 0.018 0.764 -0.056 K 0.908 -0.005 0.189 Mg 0.950 0.133 0.005 Mn 0.119 0.838 0.220 Na 0.838 0.056 0.008 SO4 2- 0.605 0.591 -0.082 TN 0.520 0.340 0.500 TP -0.011 -0.059 0.867 Eigenvalue 6.15 2.88 12.9 % of Variation 41.0 19.2 8.6 Cumulative % 41.0 60.2 68.8 Appendices 226 Figure C1: Dendrogram indicating clustering of sample sites, using median value for each parameter April 2005 to February 2007. 227 Appendix D Table D1: Median values of surface water quality for each sample period at Wooloweyah, 2006. n = 10. Parameter June July August September Temp (ºC) 18.3 15.6 18.3 23.6 pH 7.7 8.0 7.7 6.7 EC (dS m-1) 27.6 22.7 12.8 0.7 DO (mg L-1) 7.4 9.3 10.1 3.3 Al (mg L-1) 0.00 0.00 0.03 0.20 Ca (mg L-1) 124.8 92.7 67.4 3.6 Cl- (mg L-1) 1874.3 4596.9 3436.4 47.5 Fe (mg L-1) 0.13 0.16 0.15 0.63 K (mg L-1) 122.0 97.2 57.2 7.4 Mg (mg L-1) 440.2 320.2 215.6 13.7 Mn (mg L-1) 0.12 0.05 0.17 0.27 Na (mg L-1) 3631.7 2635.6 1676.1 91.4 SO4 2- (mg L-1) 294.4 421.9 100.3 14.7 TN (mg L-1) 0.82 0.44 0.54 1.35 TP (mg L-1) 0.02 0.01 0.05 0.32 Appendices 228 Table D2: Median values of surface water quality for each sample period at Colletts Swamp, 2006. n = 3 June and July, n = 5 August, n = 7 September, n = 4 October. Parameter June July August September October Temp (ºC) 13.1 13.9 16.0 24.3 22.5 pH 6.4 7.0 5.7 5.7 6.7 EC (dS m-1) 0.8 1.9 2.5 3.8 3.8 DO (mg L-1) 2.3 6.1 7.7 10.0 4.3 Al (mg L-1) 0.03 0.10 0.00 0.00 0.00 Ca (mg L-1) 8.9 10.2 16.5 20.9 22.2 Cl- (mg L-1) 46.4 163.3 228.1 238.3 278.8 Fe (mg L-1) 0.06 0.01 0.05 0.34 0.07 K (mg L-1) 5.5 5.9 12.5 11.5 11.8 Mg (mg L-1) 17.0 22.2 46.7 56.2 61.6 Mn (mg L-1) 0.10 0.08 0.16 0.91 0.21 Na (mg L-1) 147.8 177.9 344.6 428.7 470.5 SO4 2- (mg L-1) 6.5 12.2 27.0 58.1 35.7 TN (mg L-1) 1.66 0.53 0.66 0.86 1.06 TP (mg L-1) 0.20 0.05 0.03 0.07 0.08 Appendices 229 Figure D1: Surface water (a) electrical conductivity and concentrations of (b) Cl-, (c) Ca, (d) K, (e) Mg, (f) Na and (g) SO4 2- at Little Broadwater, June to October 2006. Filled black circles represent the median, bars indicate the maximum and minimum values. n = 44 ULB, n = 57 MLB, n = 33 LLB. Appendices 230 Figure D2: Surface water (a) pH, (b) Al, (c) Fe and (d) Mn at Little Broadwater, June to October 2006. Filled black circles represent the median, bars indicate the maximum and minimum values. n = 44 ULB, n = 57 MLB, n = 33 LLB. Figure D3: Surface water (a) total nitrogen and (b) total phosphorus concentrations at Little Broadwater, June to October 2006. Filled black circles represent the median, bars indicate the maximum and minimum values. n = 44 ULB, n = 57 MLB, n = 33 LLB. Appendices 231 Figure D4: Groundwater (a) electrical conductivity and concentrations of (b) Cl-, (c) Ca, (d) K, (e) Mg, (f) Na and (g) SO4 2- at Little Broadwater, June to October 2006. Filled black circles represent the median, bars indicate the maximum and minimum values. n = 3 P1-P6. Appendices 232 Figure D5: Groundwater (a) pH, (b) Al, (c) Fe and (d) Mn at Little Broadwater, June to October 2006. Filled black circles represent the median, bars indicate the maximum and minimum values. n = 3 P1-P6. Figure D6: Groundwater (a) total nitrogen and (b) total phosphorus concentrations at Little Broadwater, June to October 2006. Filled black circles represent the median, bars indicate the maximum and minimum values. n = 3 P1-P6.