Patterns of Erosion and Sediment and Nutrient Transport in the Douglas Shire Catchments (Daintree, Saltwater, Mossman and Mowbray), Queensland: A …

CSIRO LAND and WATER

Patterns of Erosion and Sediment and

Nutrient Transport in the Herbert River

Catchment, Queensland

R.Bartley , A.Henderson , I.P.Prosser , A.O.Hughes ,

L.McKergow , H.Lu , J.Brodie , Z.Bainbridge , and C.H.Roth .

(1) (2) (3) (3)

(3) (3) (4) (4) (2)

(1)CSIRO Land and Water, Atherton

(2)CSIRO Land and Water, Townsville

(3)CSIRO Land and Water, Canberra

(4)James Cook University, Townsville

Consultancy Report

CSIRO Land and Water, August 2003

2

Copyright © 2003 CSIRO and the Herbert River Catchment Group (Inc). This work is copyright. It may be reproduced subject to the inclusion of an acknowledgement of the source. Important Disclaimer CSIRO Land and Water and the Herbert River Catchment Group (Inc) advises that the information contained in this publication comprises general statements based on scientific research. The reader is advised and needs to be aware that such information may be incomplete or unable to be used in any specific situation. No reliance or actions must therefore be made on that information without seeking prior expert professional, scientific and technical advice. To the extent permitted by law, CSIRO Land and Water and the Herbert River Catchment Group (Inc) (including its employees and consultants) excludes all liability to any person for any consequences, including but not limited to all losses, damages, costs, expenses and any other compensation, arising directly or indirectly from using this publication (in part or in whole) and any information or material contained in it.

3

Table of Contents

Executive Summary____________________________________________________________6

Acknowledgements ____________________________________________________________8

1 Introduction ______________________________________________________________9

2 Purpose and outline of report _______________________________________________11

3 Summary of previous research relating to water quality in the Herbert Catchment ____11

3.1 Previous research in the Herbert Catchment relating to sediment ____________12

3.2 Previous research in the Herbert Catchment relating to nutrients ____________13

3.3 Summary ___________________________________________________________14

4 Sediment and Nutrient Budgets – A background to SedNet and associated models ____15

4.1 Sediment delivery through the river network _____________________________16 4.1.1 Course sediment transport __________________________________________17 4.1.2 Fine (suspended) sediment transport __________________________________17

4.2 Nutrient delivery through the river network______________________________19

5 Model Inputs ____________________________________________________________20

5.1 River Hydrology and Channel Form ____________________________________20

5.2 Hillslope erosion hazard_______________________________________________22

5.3 Gully Erosion Hazard ________________________________________________23

5.4 River Bank Erosion Hazard ___________________________________________24

5.5 Mine site sources_____________________________________________________26

5.6 Nutrient Sources _____________________________________________________27

6 Results _________________________________________________________________29

6.1 Sediment sources to the stream network _________________________________29 6.1.1 Overview _______________________________________________________29 6.1.2 Hillslope erosion__________________________________________________30 6.1.3 Gully erosion ____________________________________________________32 6.1.4 Riverbank erosion_________________________________________________33

6.2 Summary of sediment results __________________________________________34 6.2.1 Contribution of sediment to the coast__________________________________34 6.2.2 Contribution of sediment from major sub catchments _____________________36 6.2.3 Comparison of modelled sediment results with measured data ______________38 6.2.4 Effect of improved modelling on results (comparison with NLWRA) ________39

6.3 Nutrient sources to the stream network __________________________________40 6.3.1 Contribution of nutrients to the coast __________________________________40 6.3.2 Comparison of modelled nutrient results with measured data _______________41

7 Scenario Modelling _______________________________________________________43

7.1 Background _________________________________________________________43

7.2 Methods ____________________________________________________________43

7.3 Findings ____________________________________________________________45

8 Discussion and Conclusion _________________________________________________50

9 References ______________________________________________________________52

4

List of Figures

Figure 1: Location of Herbert River Catchment showing major towns and roads. Note that all subsequent maps are orientated with North. .................................................................................................................................................10 Figure 2: Landuse areas within the Herbert (Source: CSIRO Sustainable Ecosystems) ............................................10 Figure 3: Rainfall distribution in the Herbert Catchment (Source: ANUCLIM data set)...........................................12 Figure 4: A river network showing links, nodes, Shreve magnitude of each link (Shreve, 1966) and internal catchment area of a magnitude one and a magnitude four link..........................................................16 Figure 5: Conceptual diagram of the bedload sediment budget for a river link. STC is the sediment transport capacity of the river link, determined by Equation 1. Hillslope erosion does not contribute to the course sediment budget. .........................................................................................................................................................................17 Figure 6: Conceptual diagram for the suspended sediment budget of a river link. HSDR is hillslope sediment delivery ratio. The equation is for the amount of sediment deposited on the floodplain (t/y), where Ix is the sediment load input to the link, Qfx/Qtx is the proportion of flow that goes over-bank, Afx/Qfx is the ratio of floodplain area to floodplain discharge and ν is the sediment settling velocity. ......................................................................................18 Figure 7: Floodplain width in the Herbert River Catchment ......................................................................................18 Figure 8: Distribution of average channel widths in relation to upslope contributing area for 30 surveyed sites (a = 6.22 and b =0.36) ........................................................................................................................................................22 Figure 9: Example of gully erosion observed in the Upper Herbert ...........................................................................23 Figure 10: Location of aerial photographs analysed for gully mapping ....................................................................23 Figure 11: Example of bank erosion in the Upper Herbert (on Nettle Creek) ............................................................25 Figure 12: Riparian Vegetation Mapping for the Herbert River Catchment (source: State of River Data Base, NR&M, Brisbane)........................................................................................................................................................25 Figure 13: Typical mine site in the Upper Herbert (photo taken from a light plane) .................................................26 Figure 14: Predicted Total Hillslope Erosion for the Herbert River Catchment (RKLSC is the gross hillslope erosion not including the sediment delivery ratio calculations)...............................................................................................31 Figure 15: Contribution of Hillslope erosion to streams (including SDR estimates)..................................................31 Figure 16: Predicted density of gully erosion for the Herbert River Catchment ........................................................33 Figure 17: Bank erosion in centimetres per year in the Herbert River Catchment.....................................................34 Figure 18: Total fine sediment contribution to the coast (t/ha/yr) ..............................................................................35 Figure 19: Predicted bed-load deposition in the Herbert River Catchment ...............................................................36 Figure 20: Subdivision of 17 sub-catchments in the Herbert River Catchment ..........................................................37 Figure 21: Total Suspended Sediment Rating curve for the Herbert River at Ingham................................................38 Figure 22: Pattern of N delivered to streams from each watershed within the Herbert River catchment ..................42 Figure 23: Pattern of P delivered to streams from each watershed within the Herbert River catchment...................43 Figure 24: Predicted pre-1850 contribution of sediment to the coast for Herbert River Catchment. For comparison, Figure 18 shows such contributions for current conditions. .......................................................................................46 Figure 25: Predicted Hillslope Erosion for the Herbert River Catchment based on an average of 20% cover in the open eucalypt woodland grazing lands (RKLSC is the gross hillslope erosion which does not include sediment delivery ratio calculations). For comparison, Figure 14 shows erosion under current grazing conditions (60% cover)...........................................................................................................................................................................48 Figure 26: Predicted Hillslope Erosion for the Herbert River Catchment based on an average of 40% cover in the open eucalypt dominated grazing lands (RKLSC is the gross hillslope erosion which does not include sediment delivery ratio calculations). For comparison, Figure 14 shows erosion under current grazing conditions (60% cover)...........................................................................................................................................................................48 Figure 27: Predicted Hillslope Erosion for the Herbert River Catchment based on an average of 70% cover in the open eucalypt dominated grazing lands (RKLSC is the gross hillslope erosion which does not include sediment delivery ratio calculations). For comparison, Figure 14 shows erosion under current grazing conditions (60% cover)...........................................................................................................................................................................49 Figure 28: Predicted pattern of N delivered to streams from each watershed within the Herbert River catchment for reduced fertiliser use. For comparison, Figure 22 shows current fertiliser use. ........................................................50

5

List of Tables

Table 1: Characteristics of the Herbert River Catchment (adapted from Brodie et al., 2001) ...................................11 Table 2: Estimated average concentrations of DIN, DON, FRP and DOP in runoff from land uses Wet Tropics Regions. These values were used for the nutrient modelling in the Herbert River Catchment. ..................................28 Table 3: Summary of sediment budget for the Herbert River Catchment....................................................................29 Table 4: Contribution of hillslope erosion from each of the major landuse areas with the Herbert River Catchment. (1) The forest and other reserves category contains a range of forest types from open eucalypt woodland to rainforest and melaleuca species. ...............................................................................................................................32 Table 5: Contribution of gully erosion from each of the major landuse areas with the Herbert River Catchment. (1) The forest and other reserves category contains a range of forest type from open eucalypt woodland to rainforest and melaleuca species. ................................................................................................................................................33 Table 6: Number and name of the 17 sub-catchments in the Herbert River with the total contribution from each area. The areas are also ranked in order of highest to lowest sediment loss in t/yr and t/ha/yr.................................37 Table 7: Comparison of the sediment budget between this project and the NLWRA for the Herbert River Catchment.....................................................................................................................................................................................40 Table 8: Components of the nutrient (N and P) budget for the Herbert River Catchment. .........................................40 Table 9: Current 2003 sediment budget compared with the pre-1850 budget. ...........................................................45 Table 10: Comparison of predicted hillslope erosion yields for the different cover levels on ‘open eucalypt woodland’ grazing areas.............................................................................................................................................47 Table 11: Current 2003 sediment budget compared with the 70% cover on grazing lands budget ............................47 Table 12: Comparison of current nitrogen budget with the reduced fertiliser budget for the Herbert River Catchment.....................................................................................................................................................................................50

6

Executive Summary This project was carried out to identify the major processes involved in the delivery of sediment and nutrients to rivers within the Herbert River Catchment. The loss of sediment and nutrients from the land can degrade freshwater and marine ecosystems by reducing water quality and degrading aquatic habitat. It is also acknowledged that a loss of sediment and nutrients from agricultural landscapes can lead to a decline in the productivity of local farming systems. For individual catchments, a spatially-explicit understanding of both the magnitude of these sediment and nutrient losses from different parts of the landscape, and the erosional and hydrological processes controlling them, is essential to plan, implement and monitor improved land management practices with reduced impacts. With support from the Herbert River Catchment Group (HRCG), this project was carried out to: (1) identify the major sources from which sediment and nutrients in rivers and streams of the

Herbert River Catchment were generated; (2) determine the total sediment and nutrient loads for the Herbert River Catchment; and, (3) predict likely changes in these inputs associated with a number of scenarios in which

changes to current land management practice were implemented. The Herbert River Catchment covers an area of approximately 10,000km2 and is characterised by complex patterns of geology, terrain, soils, vegetation type and land use, together with steep climatic gradients (especially for rainfall). There have been numerous studies investigating the movement of sediment and nutrients in the Herbert River Catchment, and a review of the relevant aspects of this previous work is presented in the report. However, only a limited number of these studies have examined the major sources of sediment and nutrients, and fewer still have assessed these from a ‘whole of catchment’ perspective. We used spatial modelling of the erosion, deposition, and transport processes that move sediment and nutrients within landscapes and streams to produce regional budgets for the Herbert River Catchment. These budgets map the main sediment and nutrient sources, as well as the patterns of deposition within the catchment. The resultant mean annual loads are calculated for streams across the catchment. The river budgets are modelled using a set of GIS-based programs known as SedNet. The sources of sediment considered are soil erosion by surface (hillslope) processes, gully erosion and riverbank erosion. These erosion processes, together with dissolved sources in diffuse run-off (e.g. fertiliser), are also the sources of phosphorus and nitrogen in the modelling. The model outputs are averages only, and are considered to represent the range of conditions experienced over a 100-year climate record. The use of averages constrains the relevance of the modelling to larger spatial scales (e.g. sub-catchments) and means that the results should not be used to extrapolate to sediment and nutrient loads at the individual property or farm scale. It is also important to note that the use of average figures for variables such as (ground) cover, means that the management practices of individual landholders are not taken into consideration. This is particularly relevant for the grazing areas where ground cover can vary significantly between properties and between years. Unfortunately, suitable spatial and temporal data describing the level of cover on individual properties or sub-catchments does not currently exist. Work on obtaining such values is in progress, in the mean-time, however, average values will be used. Subsequently, the results of this work must be applied with these limitations in mind. The results of the modelling in the Herbert River Catchment demonstrate that hillslope erosion is the dominant source of sediment, contributing 52% of the total sediment load at the river mouth. Gully and stream bank erosion contribute equally (~24%) to make up the remaining sediment load. The predominant sources of nutrients are in dissolved form for nitrogen (N) and particulate form for phosphorus (P). Both the N and P budgets are dominated by losses from cultivated cane lands on the floodplain, followed by the open forests and cultivated areas of the middle and upper catchment. The grazing dominated areas of the south west of the upper catchment are predicted to contribute low levels of both N and P at the river mouth.

7

Sediment sources from the different land uses were estimated for the catchment. Grazing lands and open forests and woodlands on steeply incised landscapes were the greatest contributors to the overall sediment load, both producing ~ 200,000 t/yr. These contributions reflect the spatial dominance of these land uses in the catchment, with grazing and open forests covering 60% and 30% of the catchment, respectively. On a unit area basis, sediment losses were highest from abandoned mine sites followed by open forests on steeply incised land, and then cultivated agriculture. The catchment was also split into 17 sub-catchments that were differentiated to represent the major tributaries of the river system. Based on this analysis, the area with the highest soil loss (per/ha) is the Ripple Creek/Seymour River area with 2.07 t/ha/yr, followed by streams in the Mount Garnet region with 1.90 t/ha/yr and then Yamanie Creek with 1.76 t/ha/yr. To validate the reliability of the sediment results from this study at the ‘whole of catchment’ scale, we compared our findings with the significant water quality data that have been collected over a number of years by various organisations and researchers in the vicinity of the John Row/Gairloch Bridge (near Ingham). Overall the data collected by researchers from the Australian Institute of Marine Science (AIMS) is the most suitable for comparison with our modelled sediment results due to the substantial length of that data record (which included wet and dry years). The modelled results from SedNet predict that there is approximately 600,000 t/yr of fine sediment (not including bed-load) delivered to the area in the vicinity of the Gairloch bridge. This compares very favourably with the ~ 540,000 t/yr estimated by AIMS researchers from a sampling location at that site (M. Furnas, In Press). Therefore there is good agreement in the average modelled sediment values estimated for the entire catchment, as they are within 10% of the average results based on directly measured data. As part of this project, we ran a number of scenarios using the SedNet programs to investigate the relative effectiveness of different management strategies on long-term sediment and nutrient yields from the river network. These scenarios included (1) the pre-1850 sediment loads, (2) the sediment loads from both increased and decreased cover on grazing lands and (3) reduced fertiliser application on cane lands. The results from the pre-1850 scenario suggest that there has been a 6 fold increase in the amount of sediment exported from the Herbert River Catchment. The results of the changing cover levels on grazing lands suggest that if ground cover levels on all of the open eucalypt woodland grazing areas are increased by just 10%, this would result in the catchment being half way to achieving the sediment load targets set by GBRMPA (Brodie et al., 2001). Conversely, reducing cover levels could result in a massive increase in hillslope erosion in the Upper Catchment. The nutrient scenario results suggests that a reduction in fertiliser rates on cane lands from an average of 200 kg/ha/yr to 130 kg/ha/yr can produce a decrease of ~27% in the dissolved inorganic nitrogen (DIN) levels and a 10% decrease to the overall N budget. This is a relatively good outcome considering that sugar cane occupies only 7% of the land area.

8

Acknowledgements This study was commissioned by the Herbert River Catchment Group (Inc.) using funds obtained through Natural Heritage Trust (NHT) grants. It further tests the application of sediment modeling approaches through the use of regional data sources. We acknowledge the input and support of staff from the Queensland Department of Natural Resources and Mines (NRM), Lex Cogle, Peter Wilson, Georgina Pitt, Ian Webb and John Armour (Mareeba). Geoff Pocock, Neale Searle and Vince Manly from NR&M hydrology section supplied the data from gauging stations, and Graham Hammermeister (Brisbane) assisted with the aerial photography. We also acknowledge the help of Raymond DeLai and Anna Forest of the Herbert River Information Centre in Ingham. There were a number of staff from the Queensland Department of Primary Industries (Mareeba) that provided important input to the grazing management side of the study, in particular Joe Rolfe, Jim Kernot and Bernie English. Thanks also to Ross Hogan (Hinchinbrook Shire) and Mark Jempson (WBM Oceanics) for access to data sets from the Herbert River catchment. There were also a number of people from CSIRO that provided technical input with many parts of the study and we would like to thank Scott Wilkinson, Ron DeRose and Bill Young, of CSIRO Land and Water in Canberra. Paul Reddell (CSIRO Atherton) also provided valuable comments on an earlier draft of this document. Finally we would like to thank the members of the catchment group, and in particular the Herbert River Catchment Co-ordinator, Caroline Coppo, for their assistance and participation in the production of this report.

9

1 Introduction The loss of sediments and nutrients from a catchment can have impacts on the industries that use the land for farming and grazing (in terms of loss of productive nutrient rich soil), as well as impacts on downstream freshwater and marine ecosystems, such as the Great Barrier Reef World Heritage Area (GBRWHA). Sediment delivered to streams has several potential downstream impacts. High loads of suspended sediment, the silts and clays that are carried in the flow, degrade water quality in streams, reservoirs and estuaries. This is a result of both the sediment itself and the nutrients that the sediment carries. High concentrations of suspended sediment reduce stream clarity; inhibit respiration and feeding of stream biota; diminish light needed for plant photosynthesis; make water unsuitable for irrigation and require treatment of water for human use. The export of high suspended sediment volumes to coastal areas and fragile marine environments such as in-shore reefs, and tidal flats can cause accelerated deposition, smothering aquatic habitats and increasing turbidity through re-suspension of the sediment. The general impacts of increased sediment and nutrients on stream systems has been well documented (e.g. Alexander and Hansen, 1986; ASCE, 1992) as well as on nearshore environments such as the Great Barrier Reef (e.g. Williams, 2001; McCulloch et al., 2003; Baker, 2003). It is important to note, however, that not all suspended sediment delivered to streams is exported to the coast. Some of it is deposited along the way on floodplains, providing fertile alluvial soils, or it is deposited in reservoirs. The extent of this deposition is highly variable from one river reach to another. Deposition potential must be considered when trying to relate catchment land use to downstream loads of sediment. Understanding and identifying the major processes involved in the delivery of sediment and nutrients to streams, the critical areas of erosion potential, and the major contributors of sediment and nutrients to the coast, are required for sustainable catchment management. This report presents a sediment and nutrient budget for the Herbert River Catchment, North Queensland, Australia. As part of these budgets, the critical erosion sources, pathways and total loads were calculated. The remainder of this report outlines the details of these budgets.

The Herbert River Catchment The Herbert River Catchment is the largest of the Wet Tropics Catchments in North Queensland (Australia) and covers an area of approximately 10,000 km2. The Herbert River Catchment can be broken up into four distinct physiographic sections (Figure 1). The upper catchment occupies ~4735 km2 and is dominated by grazing, but has also been subject to various forms of mining, and has other mixed agriculture (e.g dairy) around the Ravenshoe and Herberton areas. The middle section of the catchment is ~1825 km2 and is predominantly Wet Tropics World Heritage Area, State Forests and timber reserves. It is important to note that the area classified as ‘forest’ (National Park, State Forest or Reserve) in Figure 2 contains a range of different forest types from open eucalypt woodland to rainforest and melaleuca dominated areas. The lower floodplain section is ~3560 km2 and can be divided into two areas; the floodplain area around the Herbert River channel, and the southern coastal section which contains streams that flow directly to the coast. Both these areas are dominated by sugarcane cultivation (Figure 2). The geology, geomorphology, soils, rainfall and landuse characteristics are very different for the four regions (Figure 1); hence, there are differences in the history of erosion and sedimentation for each area. Further bio-geographic information on the Herbert River catchment can be found in Johnson and Murray (1997).

10

Figure 1: Location of Herbert River Catchment showing major towns and roads. Note that all subsequent maps are orientated with North.

Figure 2: Landuse areas within the Herbert (Source: CSIRO Sustainable Ecosystems)

11

2 Purpose and outline of report The main aims of this project were to use the SedNet modelling framework to:

• integrate as much relevant regional data as possible, including resource and environment data layers, water quality monitoring results, river surveys, and flow gauging data;

• identify the major sources of sediment and nutrients in the catchment; • predict river sediment and nutrient loads, and their delivery to the coast; • predict the likely future sediment and nutrient budgets for a set of specified management actions (scenarios) aimed at ameliorating land degradation.

This report presents the background, methods, data and research results relating to aims outlined above. Section 3 summarises previous studies that have been carried out in the Herbert River Catchment with respect to sediment and nutrient research. Section 4 provides a background to SedNet and the associated models used to develop the sediment and nutrient budgets. Section 5 describes the model inputs, how they were derived and applied in the modelling process. Section 6 presents the results of the modelling process with specific sections dedicated to hillslope, gully and bank erosion. The nutrient budget results are also discussed in this section. Section 7 presents the results of the scenario modelling and Section 8 presents the discussion and conclusions.

3 Summary of previous research relating to water quality in the Herbert Catchment

A considerable amount of research has been undertaken investigating the increase in sediment and nutrient exports to the Great Barrier Reef World Heritage Area (GBRWHA), and much of this research is outlined in previous reviews of this topic (see Brodie et al., 2001; Furnas, In Press; Furnas and Brodie, 1996; Mitchell et al., 1996; Rayment, 2002; Wasson, 1997; Williams et al., 2001 and more recently in Baker, 2003). Many of these reports include data from the Herbert River Catchment. Numerous other reports have been published specifically on the Herbert Catchment. These include a presentation of the general characteristics of the Herbert River Catchment (Johnson and Murray, 1997) a catchment wide assessment of stream condition (Moller, 1996) and numerous sediment and nutrient transport studies (see Bramley and Johnson, 1996; Brunskill et al., 2002, Horn et al., 1999a and 1999b; Ladson and Tilleard, 1999; Mitchell and Bramley, 1997; Mitchell et al., 1991; Mitchell et al., 1997; Roth and Olley, In Prep; Wong, 1996; Visser, 2003). Due to the large amount of literature for the Herbert Catchment, this section will present only an overview of the information that is relevant to this study. Section 3.1 summarises most of the research that has been conducted to look at sediment movement in the Herbert River Catchment and Section 3.2 summarises the work on nutrients. To provide a background to the catchment, a description of the rainfall and landuse characteristics of the Herbert River Catchment are given in Table 1 and Figure 3. Table 1: Characteristics of the Herbert River Catchment (adapted from Brodie et al., 2001)

Characteristic Size Area (km2) 9843 % Gauged 87 Rainfall (mm) 1506 Runoff (mm/m2) 407 Runoff/Rainfall Ratio 27 Population 8778 % Cleared 15 Area under Grazing (km2) 7330 Area under Sugar (km2) 691 Area under Horticulture (km2) 35

12

Figure 3: Rainfall distribution in the Herbert Catchment (Source: ANUCLIM data set)

3.1 Previous research in the Herbert Catchment relating to sediment A comprehensive study by Bramley and Roth (2002) that measured N, P and total suspended sediments (TSS) concentrations in streams draining land under sugarcane in the Herbert River catchment, found that sugarcane has had a significant impact on riverine water quality. Almost 20 sites were sampled monthly as well as following rainfall events of greater than 50 mm d-1, between October 1992 and May 1995. Median values of TSS, for each site, were significantly positively correlated with the proportion of land upstream under sugarcane and negatively correlated with the proportion of land upstream under grazing. These results also showed that the proportions of N were greatest in areas dominated by sugarcane, which suggests that inefficient use of fertilizers may be an important contributor of N to streams. The authors suggest that irrespective of the ecological impact of the sediment and nutrient loadings on marine and freshwater systems, there is considerable room for improvement for land managers in this catchment. Ladson and Tilleard (1999) presented a comprehensive review of the history of sedimentation in the lower Herbert dating back to 1870. Based on historical accounts, gauging records and cross-sectional surveys it appears that the lower Herbert has not necessarily aggraded (with coarse sediment) since human settlement of the region. Historical accounts suggest that the (lower) Herbert has been shallow and sandy, at least since the time of European settlement. It may be that the current perception of increased sedimentation also coincides with an increase in water use. If the overall water level has dropped in the lower reaches, then large bar features would be exposed more frequently giving the idea that sedimentation has occurred. It is important to note, however, that this study looked only at bedload (>2 mm) and not suspended load, and therefore this piece of research did not necessarily disagree with research conducted on fine (suspended)

13

sediment which suggest a 3 - 10 fold increase in sediment load. Even though the work by Ladson and Tilleard (1999) shows no evidence for anthropogenic sedimentation in the lower Herbert, results from other catchments around Australia (eg. Rutherfurd and Budahazy, 1996) suggest that the bedload may eventually follow the same pattern as the finer sediment resulting in excessive sediment loads (both fine and coarse) over time. Work using radio-chemical contaminant tracer signals has found little evidence for massive increases in sedimentation rates in Hinchinbrook Channel and Missionary Bay in the last century (Brunskill pers. com in Williams, 2001). Some sediment cores display declines in terrestrial sediment supply rates over the last two decades, and this could be related to green cane harvesting practices in the Herbert River Valley (Brunskill pers. com in Williams, 2001). This study suggests that further research is needed to integrate the total sediment load (bedload and suspended sediment) so that it is possible to accurately estimate exports from the Herbert River mouth. Bed and suspended sediment loads were measured in the Herbert River during Tropical Cyclone Sadie and another flood event in late February 1994 (Wong, 1996). The total sediment export from the two flood events at Ingham were ~ 193,000 tonnes. Of this 113,000 tonnes is derived from Cyclone Sadie (59%) and 80,000 tonnes of sediment was transported from the normal wet season. Mitchell and Bramley (1997) also presented data for the Herbert River collected during the Cyclone Sadie event of 1994. The results suggest that 100,000 tonnes of suspended sediment, 600 tonnes of N and 65 tonnes of P were exported during the 6 day event. Queensland Department of Natural Resources and Mines have sampled the full spectrum of sediment movement (bedload and suspended load) in 11 rivers since 1994 (Horn et al., 1999a; 1999b). Based on initial results, the average bed material load for the Herbert River was found to be ~81,000 tonnes/year. Visser (2003) calculated a sediment budget for an area of the Ripple Creek Catchment of the Herbert River Floodplain and found that this area is a net source of sediment (under particular flow conditions). Visser (2003) suggested that cane fields, which do not have a protective trash cover, were the largest net source of sediment during the 1999-2000 season. There were also sediment input from water furrows, however, these areas also acted as a sediment storage area, and headlands tend to act as sinks. The source or sink function of drains is less clear, but seems to depend on their shape and vegetation cover. The most interesting finding of this piece of research was the observation that the Herbert River floodplain acted as a sediment source; this essentially contradicts the general understanding that floodplains are areas of sediment storage within river catchments. Visser’s explanations for this finding were that sugar cane cultivation has lowered the resistance of the soil surface. In addition, increased drainage has increased the drainage velocity and flood control structures have altered flooding patterns. In summary, almost all of the research conducted in the Herbert River catchment studies show that there has been an increase in suspended sediment concentrations in the Herbert River system. The research into increased bedload has not conclusively shown there to be an increase in the coarse sediment fraction, nor is there considerable evidence in the near shore zone that suggests a massive increase in total sediment yield to the coast. Much of the on-ground research in this area is still in progress.

3.2 Previous research in the Herbert Catchment relating to nutrients The study of nutrient exports on the Herbert River presented in Bramley and Johnson (1996) suggested that nutrient concentrations in the Herbert were generally below ANZECC (1992) target levels for the protection of freshwater ecosystems except in high flow conditions. In high flow events the ANZECC values were generally exceeded. Nutrient loss in the Herbert

14

catchment were found to be predominantly event-based and insignificant outside the wet season months. Although nutrient losses from intensively managed cane lands exceeded those from other land uses, Bramley and Johnson (1996) suggest that these losses might be expected to occur in these areas irrespective of crop type due to the strongly seasonal climate and high rainfall intensities. Mitchell et al., (1991) also presented results from a study of 74 km of the Herbert River from Yamani Falls National Park to John Row Bridge. The main finding from this study was that nitrate concentrations progressively increased in a downstream direction, whereas phosphate concentrations were low and there was little increase downstream.

Furnas and Mitchell (2001) measured nitrogen and phosphorus concentrations at the Upper (Yamani Falls National Park) and Lower (Ingham) ends of the Herbert River floodplain and found that particulate N fluxes at the downstream site are approximately 1.8 times the input flux, though suspended sediment concentrations did not differ greatly between the two sites. The study by Bramley and Roth (2002) found that with respect total N, approximately 31%, 9% and 3% of samples collected from streams predominantly draining cane land, grazing and forestry, respectively, were found to be above the “interim trigger levels for assessing possible risk of adverse effects due to nutrients” (ANZECC, 2000). For total P, approximately 86%, 47% and 33% of the samples from streams draining these land uses were above the trigger value (Bramley and Roth, 2002). In summary, previous research into nutrient loads in the Herbert River Catchment suggest that there has been an increase in nutrient concentration, particularly for N, in a downstream direction. The main time that nutrients are of concern (above ANZECC standards) are during high flow events.

3.3 Summary The many research reports and investigations in the Herbert River catchment have provided valuable data and insights into various aspects of sediment and nutrient movement within the catchment. This has been done using a range of techniques including field monitoring and modelling. The most recent study conducted that has measured the total sediment loads for the Herbert Catchment was the National Land and Water Resources Audit (NLWRA; 2001). The results from this study suggests that there is ~665,000 t/yr of sediment leaving the Herbert River mouth. The most recent studies that have estimated the overall nutrient losses for the Herbert Catchment were by the Australian Institute of Marine Science (Furnas, In Press) and the NLWRA (2001). The estimates from AIMS suggest that between 1,600 t/yr of N and 170 t/yr of P are leaving the catchment. Whereas, the results of the NLWRA suggests loads of between 3415 t/yr and 702 t/yr. Overall these loads place the Herbert River Catchment in the ‘medium risk’ category for sediment and total phosphorus exports and the ‘high risk’ category for total nitrogen (Brodie et al., 2001). Despite having estimates of the total sediment and nutrient loads leaving the Herbert River Catchment, most of the previous studies have been unable to determine where in the catchment the sediment and nutrient sources are derived. Without knowledge of the main source areas for sediments and nutrients within a catchment, strategic catchment management planning is difficult. This report will identify the major sediment and nutrient sources for the Herbert River Catchment.

15

4 Sediment and Nutrient Budgets – A background to SedNet and associated models

The only practical framework to assess the patterns of sediment and nutrient transport across a large complex area such as the Herbert catchment is a spatial modelling framework. There is a large range in climatic, topographic, and land use patterns which can strongly affect erosion and sediment transport. The lack of sampling conducted in the Upper Catchment prevents the use of existing data sets to determine the range of sediment and nutrient sources, and it is unrealistic to initiate sampling programs of these processes now and expect results within a decade. Furthermore, collation and integration of existing data has to be put within an overall assessment framework, and a large-scale spatial model of sediment transport is the most effective use of that data. Sediment and nutrients are derived from a number of processes which include: • Runoff on the land, termed surface wash and rill erosion or alternatively hillslope erosion; • Erosion of gullies formed as a result of land clearing or grazing; and • Erosion of the banks of streams and rivers. • Diffuse dissolved losses of nutrients In many cases one process dominates the others in terms of delivering sediments and nutrients to streams, and the predominant process can vary from one part of a large catchment to another. Hence, the management actions required to reduce sediment and nutrient loss also vary.

The assessment of sediment and nutrient transport is divided into four aspects: hillslope erosion as a source of sediment and attached nutrients; hillslopes as a source of dissolved nutrients; gully erosion as a source of sediment and nutrients; and stream links as a further source, receiver and propagator of the sediment and nutrients. To calculate the supply of sediment and nutrients, their deposition and delivery downstream, we constructed river sediment and nutrient budgets. We calculated budgets for two types of sediment: suspended sediment and coarse sediment (or bed-load). The programs used to model sediment transport are collectively referred to as the SedNet: the Sediment River Network model (outlined in more detail in Section 4.1). The additional program used to model the nutrients model is known as ANNEX (Annual Network Nutrient Export) (outlined in more detail in Section 4.2).

The SedNet model calculates, among other things:

• the mean annual suspended sediment output from each river link; • the depth of sediment accumulated on the river bed in historical times; • the relative supply of sediment from surface wash, gully and bank erosion processes; • the mean annual export of sediment to the coast; and • the contribution of each watershed to that export.

SedNet and ANNEX were developed for the National Land and Water Resources Audit (NLWRA). Details of the NLWRA model and subsequent applications of SedNet to regional catchments in Australia (e.g. the Burdekin, Mary and Goulburn/Broken catchments) are given in Prosser et al., (2001a; 2001b), DeRose et al., (2002; 2003). The methods use in the construction of input data and implementation of the SedNet program itself are described in detail in a number of CSIRO technical reports which are available at http://www.clw.csiro.au/publications/technical2002/ or by contacting the authors. Consequently only a very brief overview of the model is included here and in particular where this regional

16

application differs from the NLWRA work (http://www.nlwra.gov.au/). Readers should therefore reference previous reports for detail on the models (e.g. Prosser et al., 2001b).

4.1 Sediment delivery through the river network The basic unit of calculation for constructing a sediment budget is a link in a river network. A link is the stretch of river between any two stream junctions (or nodes; Figure 3). Each link has an internal watershed, from which sediment is delivered to the stream network by hillslope and gully erosion processes. The internal catchment area is the catchment area added to the link between its upper and lower nodes (Figure 3). For the purpose of the model, the internal catchment area of first order streams is the entire catchment area of the river link. Additional sediment is supplied from bank erosion along the link and from any tributaries to the link. Sediment is processed sequentially through the river network beginning with first order links and terminating at the catchment outlet (commonly the ocean or a major river such as the Murray River). The sediment load output from each link is calculated from the supply of sediment from tributary links and the local watershed, less losses though floodplain deposition (fine sediment), bed deposition (coarse sediment), and reservoir deposition (coarse and fine sediment). The sediment yield at the terminating link constitutes the total yield of the river network.

1

1

1

1

4

3

2

Figure 4: A river network showing links, nodes, Shreve magnitude of each link (Shreve, 1966) and internal catchment area of a magnitude one and a magnitude four link.

The branching network of streams for the Herbert catchment was built from a 100m DEM (digital elevation model) grid. The DEM was constructed using TOPOGRIDTOOL in Arc/INFO® with input elevation and drainage data consisting of 1:50,000 topographic contours (source: Wet Tropics Management Authority) and TOPO250K drainage data (source: AUSLIG). It was a significant improvement on the DEM used in the National Land and Water Audit (Lu et. al., 2001). The upper branches of the network were defined as any stream longer than 5km with a catchment area of 5 km2 or more. These criteria limited the number of links across the assessment area, while providing a good representation of the channel network. The physical stream network extends upstream of the limit in most areas and these areas are treated as part of the internal catchment area contributing material to the river link. Short links, where the catchment area had reached less than 20 km2 by the downstream node, were removed. Internal catchment areas for each link were determined from the DEM.

17

4.1.1 Course sediment transport The coarse sediment (bedload) budget is illustrated in Figure 5. The main aim of the bedload budget is to predict the formation of sand slugs (large increases in coarse bedload material). These are predicted to occur when there is an excess of sediment supply to a river link beyond the capacity of the link to transport sand sized sediment. This is known as the sediment transport capacity (STC) and is based on Yang’s (1973) relationship to unit stream power (Equation 1).

4.0

4.13.186

x

xxx w

QSSTC

ω∑=

Equation 1

Sediment transport capacity is a function of the river width (wx), slope (Sx), discharge (Qx) settling velocity of the bedload particles (ω) and hydraulic roughness of the channel. ΣQx

1.4

represents mean annual sum of daily flows raised to a power of 1.4 (Ml1.4 y-1). The sediment transport capacity (STC) was calculated for each river link (ty-1) using Yang's (1973) equation, and an average value for Manning’s roughness coefficient of 0.025. The value of ω was determined for particles with a mean diameter of 2 mm, being the average size observed for sediment slug deposits (Rutherfurd and Budahazy, 1996).

If loading < capacity • no deposition • yield = loading

Tributary supply (t/y)

Gullyerosion (t/y)

Riverbankerosion (t/y)

Downstream yield (t/y)

STC (t/y)

If loading > capacity • deposit excess • yield = capacity

Figure 5: Conceptual diagram of the bedload sediment budget for a river link. STC is the sediment transport capacity of the river link, determined by Equation 1. Hillslope erosion does not contribute to the course sediment budget.

4.1.2 Fine (suspended) sediment transport The suspended sediment budget is illustrated in Figure 6. The main aim of the this budget is to predict the export of suspended sediment after loss of sediment on floodplains. For the Lower Herbert, floodplain widths were determined from HRIC maps of the 1967 and 1977 flood extents (source – Hinchinbrook Shire Council) and water-body data (areas defined as “swamp” or “subject to inundation”) in the AUSLIG TOPO250K digital datasets. Information on extent of flooding was not available for the Upper Herbert, instead mapping of quaternary alluvium (based on Department of Minerals and Energy 1:250,000 geological mapping) was used. We inferred floodplain extent by establishing a ratio relationship between area of flooding and area quaternary alluvium for the lower Herbert and applying this to the Upper Herbert. The resulting floodplain widths attributed to each stream link are shown in Figure 7. A relatively simple model of floodplain deposition is implemented in SedNet. Floodplain deposition in this case is simply the proportion of sediment that goes over-bank and settles out during a typical flood. It is calculated as the ratio of the median over-bank flow multiplied by the proportion of sediment that would be expected to settle out during over-bank flow (see Figure 6). Particle settling is a function of the residence time of water on the floodplain. The longer that

18

water sits on the floodplain the greater the proportion of the suspended load that is deposited. The residence time of water on floodplains increases with floodplain area and decreases with floodplain discharge. This simple model of floodplain deposition assumes a uniform sediment concentration and that the majority of suspended sediment is transported at times of high river flow.

−

QvA

=

−fx

fx

xx e Q

Q I D 1 tx

fx

Floodplain Af

Tributary supply (t/y)

Hillslopeerosion (t/y)

Riverbankerosion (t/y)

Gullyerosion (t/y)

HSDR

Downstream yield (t/y)

Figure 6: Conceptual diagram for the suspended sediment budget of a river link. HSDR is hillslope sediment delivery ratio. The equation is for the amount of sediment deposited on the floodplain (t/y), where Ix is the sediment load input to the link, Qfx/Qtx is the proportion of flow that goes over-bank, Afx/Qfx is the ratio of floodplain area to floodplain discharge and ν is the sediment settling velocity.

Figure 7: Floodplain width in the Herbert River Catchment

19

Contribution of Sediment to the Coast The differentiation of areas that contribute strongly to total river sediment export is an important aspect of catchment management as this enables catchment managers to target areas for rehabilitation. It is not always possible, or sensible, to implement erosion control works effectively across large areas. Not all suspended sediment delivered to rivers is exported from the catchment as there are extensive opportunities for floodplain deposition along river courses. The contribution that each stream link’s watershed makes to sediment exported at the catchment outlet can be calculated once the mean annual suspended sediment export is known. Each watershed delivers a mean annual load of suspended sediment (LFx) to the river network. This is the sum of gully, hillslope and riverbank erosion. The watershed delivery and tributary loads constitute the load of suspended sediment (TIFx) received by each river link. Each link yields some fraction of that load (YFx); the rest is deposited. The ratio of YFx/TIFx is the proportion of suspended sediment that passes through each link. It can also be viewed as the probability of any individual grain of suspended sediment passing through the link. The suspended load delivered from each watershed will pass through a number of links on route to the catchment mouth. The amount delivered to the mouth is the product of the loading LFx from the watershed and the probability of passing through each river link on the way:

n

n

x

x

x

xxx TIF

YFxx

TIFYF

xTIFYF

xLFCO ......1

1

+

+= Equation 2

where n is the number of links on the route to the outlet. Dividing this by the internal catchment area expresses contribution to outlet export (COx) as an erosion rate (t/ha/y). The proportion of suspended sediment passing through each river link is ≤ 1. A consequence of Equation 2 is that all other factors being equal, the further a watershed is from the mouth, the lower the probability of sediment reaching the mouth. This behaviour is modified though by differences in source erosion rate and deposition intensity between links.

4.2 Nutrient delivery through the river network The nutrient budget model (Annual Network Nutrient Export – ANNEX; Young et al., 2001), of the SedNet set of programs, predicts the average annual loads of phosphorous (P) and nitrogen (N) in each link in a river network in a similar way to the sediment budget model, with which it is run in conjunction. The model considers only the physical (not biological) stores of nutrients in the river system, and is also primarily concerned with the physical nutrient transport processes. It does, however, consider denitrification - a biological process resulting in loss of N to the atmosphere, and phosphorous adsorption-desorption, a physical process influenced by biological activity. ANNEX therefore assumes that at the annual time scale, the changes in biological nutrient stores within a river network link, and the fluxes between river network links due to biological transport processes, are small in comparison to the fluxes due to physical nutrient transport processes and the changes in physical nutrient stores. The main source terms are sediment associated nutrients (from hillslope erosion, gully erosion and riverbank erosion), dissolved organic (DOP and DON) and inorganic loads in runoff water (DIN and DIP), and point sources of DON and DOP. Each form of nitrogen is transported independently in the model, while equilibration between sediment associated P and DIP is allowed in each river link. More detail regarding the nutrient model is given in Section 5.6.

20

5 Model Inputs

5.1 River Hydrology and Channel Form SedNet incorporates a number of hydrological parameters into the calculation of river sediment budgets. These are used for calculations of patterns of deposition, river bank erosion, and denitrification. The parameters need to be predicted (interpolated) for each river link across the river catchment, based upon observations at hydrological gauging stations within the region. The variables used are: • the mean annual flow (MAF); • the median daily flow (Qmd); • the relative daily Q1.4 for calculating mean annual sediment transport capacity; • the bankfull discharge (Qbf); and • a representative flood discharge for floodplain deposition (in this case median over-bank flow – Qob).

Values for these variables were derived from the time series of daily flows for 12 gauging stations (source: Queensland Department of Natural Resources and Mines) within the Herbert catchment. All gauges had at least 32 years of record and were therefore considered appropriate for use in this analysis. The methods are a further development of those undertaken in the NLWRA project (Young et al., 2001). For each gauging station the mean annual runoff coefficient (ROC) is determined from the observed mean annual flow, the area of the catchment and its mean annual rainfall. The ROC is then predicted for all other links of the river network as a function of the ratio of the potential evaporation (PET, mm) to the mean annual rainfall (RF, mm) for the catchment area upstream of each link:

( )( ) ( )( )

RFPET

RFPETROC

bRFPETabRFPETa

−

+=

++ 1

1 Equation 3

where a and b are empirical values fitted by regression. Equation 3 comes from a conceptual rainfall/runoff model based upon the annual water balance (Zhang et al., submitted). For this study the regression gave a = 0.1769, b = 1.7259, with an R2 = 0.46. Once ROC has been predicted for each river link the mean annual flow (MAF, Ml/y) is calculated from the catchment area (A, km2) and mean annual rainfall:

ARFROCMAF ××= . Equation 4 Sediment transport capacity, used in the bedload budget is a function of the discharge raised to the power of 1.4 (Prosser et al., 2001). The mean annual sum of daily contributions to sediment transport capacity is:

365365365

4.1

1

4.1 ×

×=∑

=

MAFRDSQQn n

ii

Equation 5

where RDSQ is the relative contribution of daily flow to sediment transport capacity: 4.1

1

1∑=

=

n

i

i

QQ

nRDSQ

Equation 6

21

where Qi is the measured daily flow (ML/d), Q is mean daily flow for the period of record (ML/d) and n is the number of days of record. Once RDSQ is calculated for each gauging station, it is related empirically to catchment area. That relationship is then used to predict RDSQ for each river link and Equation 5 is used to calculate sediment transport capacity. Similarly, the other hydrological parameters were calculated at gauging stations and related to MAF. Those relationships were then used to predict the parameters for all stream links. For bankfull discharge the empirical relationship is:

( ) fbf MAFeQ = Equation 7

where e = 0.2219, and f = 0.973 (R2 = 0.97). For median over-bank flow:

( )hob MAFgQ = Equation 8

where g = 0.096, h = 0.957 (R2 = 0.9). For median daily flow:

jmed MAFiQ )(= Equation 9

Where i = 0.00019, j = 1.056 (R2 = 0.64). There are no major storage reservoirs in the Herbert River Catchment, and the small farm and mining dams are not considered of sufficient enough size to be incorporated into a regionalisation analysis, therefore it was not necessary to predict regulated flow parameters for stream links below the major storages and weirs. The calculations of bankfull discharge and median over-bank flow are based on the average flood recurrence interval determined from the time series of daily flows recorded at rated gauging stations. A total of 12 cross-sections were examined across the catchment and these indicated that bank full discharge averaged ~ 5 years in the lower catchment, and ~10 years in the Upper Catchment. These figures agreed with bankfull estimates described by local landholders and community members (eg. Pers comm. Pat Williams of Woodleigh Station; Giles Atkinson of Gunnawarra Station; and Ross Hogan from Hinchinbrook Shire). For consistency a recurrence interval of 5 years for bankfull was chosen for the entire catchment. The channel characteristics of width and bank height are used in SedNet for a number of calculations (eg. bank erosion, sediment transport capacity, flood frequency). It is therefore necessary to estimate channel form for all stream links. This is done by regionalizing point measurements of channel width (Figure 8) on the basis of upslope contributing area. To determine the channel widths at bankfull, the 12 gauging station cross-sections were used, as well as an additional 18 points gathered from aerial photo analysis. It was not possible to regionalise bank height as appropriate data does not exist, therefore a constant bank height of 3 metres was assumed. This procedure produces an average estimate which is applied equally to all stream links on the basis of contributing area.

22

y = 6.2185x0.3554

R2 = 0.61

0

50

100

150

200

250

0 2000 4000 6000 8000 10000

Catchment Area (km2)

Cha

nnel

Wid

th (m

)

Figure 8: Distribution of average channel widths in relation to upslope contributing area for 30 surveyed sites (a = 6.22 and b =0.36)

The Herbert River floodplain is essentially a large avulsive coastal plain (Woolfe et al., 2000) and so in flood conditions, a proportion of the over-bank flow is diverted down floodplain distributary channels. It is acknowledged that most of the Herbert River Floodplain becomes inundated in extremely large floods (such as the 1 in 100 year event), however, for the majority of flow events (e.g. from 1 in 2 year to 1 in 50 year events) only a small number of streams (e.g. Palm Creek) receive water from the Herbert River. The other coastal streams (e.g. Trebonne and Cattle Creeks) are predominantly driven by rain in their own catchments and rarely receive flow from the main Herbert River channel (pers. comm. Vince Manly, Hydrographer, NR&M, Ayr). Therefore, without sufficient hydrological data showing the proportion of flow that moves down the coastal streams from the main Herbert River channel, a redistribution of flow was not considered appropriate.

5.2 Hillslope erosion hazard Hillslope erosion from sheet and rill erosion processes is estimated using the Revised Universal Soil Loss Equation (RUSLE; Renard et al., 1997) as applied in the NLWRA (Lu et al., 2001). The RUSLE calculates mean annual soil loss (Y, tonnes ha-1 y-1) as a product of six factors: rainfall erosivity (R), soil erodibility (K), hillslope length (L), hillslope gradient (S), ground cover (C) and land use practice factor (P): Y = RKLSCP Equation 10 Rainfall erosivity (R) is as per the NLWRA project (Lu et al., 2001). Soil erodibility (K) was derived using Rosewell (1997) estimates based on principal profile form (PPF) equivalencies for mapped soils. Soils mapping from various sources was collated into a “best available” map from which an erodibility map was then derived (Cannon et al., 1992; Grundy and Bryde, 1998; Heiner and Grundy, 1994; Laffan, 1988; Northcote et al., 1960; Rogers et al., 1999; Wilson and Baker, 1990). For the slope length factor (L), we took the same approach as in the Burdekin (Prosser et. al., 2002) and set L to a constant (1). Slope factor (S) across the Herbert River catchment was derived directly from the 100 m digital elevation model (DEM). Landcover mapping at 1:50,000 scale was available over the whole catchment current to 1996 (HRIC). We used this mapping, with some improvements (based on Landsat interpretation) to delineate rainforest and eucalypt dominated areas in the middle and upper Herbert. Cover factors were estimated for each landcover type using tables from Rosewell (1993), except for sugarcane which was based on results published in Visser (2003). The delivery of sediment to streams from sheet and rill erosion on hillslopes is modified by the hillslope sediment delivery ratio (HSDR). HSDR is a measure of the proportion of fine particles eroded from the hillslope which remain suspended long enough to reach a stream channel. For

23

this project, HSDR was estimated using methods and data presented in Lu et al., (2003) and Lu et al., (submitted). Sediment produced by sheet and rill erosion is assumed to contribute only to the suspended sediment (not coarse sediment) load of rivers.

5.3 Gully Erosion Hazard The spatial pattern of gullies in the Herbert catchment was determined through aerial photograph interpretation. In this report, a gully was defined as a non-permanent watercourse with steep, actively eroding banks (Figure 9). Gullies were mapped using 50 stereo aerial photographs pairs (~100 photos) spanning different geographic regions, geologic classes and land uses. Figure 10 shows the locations of the aerial photographs selected for gully mapping in the catchment. The aerial photos were scanned into a digital format and georeferenced using ArcInfo® software. Gully density (km/km2) was then mapped from the photos by (1) defining physiographic regions within each photo; (2) calculating total gully length within each region; and (3) dividing by the total area of each region. Cubist data mining software (Rulequest Research, 2001) was used to predict gully densities throughout the unmapped portions of the catchment based on the regions of mapped gully density. Cubist used a range of environmental indicators in predicting gully density. Factors included land use, hillslope inclination, geology, hillslope length, soil properties and climatic indices. The methods used in generating average gully densities from aerial photographs are described in Hughes et al., (2002).

Figure 9: Example of gully erosion observed in the Upper Herbert

Figure 10: Location of aerial photographs analysed for gully mapping

24

Location of selected gullies from the aerial photo surveys were assessed for accuracy in the field in the Innot Hot Springs and Mt Garnet region. In all cases gullies that were identified on the photos were also identified in the field, thus verifying the accuracy of the data. It was noted during the field reconnaissance that many of the gully systems were related either to mining, roads or power line installation. Gullies were not common in grazing lands that had good grass cover. However, the number and severity of gullies did seem to increase moving south west from Innot Hot Springs through to Mt Garnet and the Gunnawarra Road area. This is most likely a function of the changing soil types and decrease in rainfall gradient along this section. It may also be related to stocking densities and ground cover, although no data is available to support this observation. Total sediment supplied to each stream link from gully erosion (GS, kt/y) is calculated in SedNet as the product of gully density (D km/km2), watershed area (C, km2), average gully cross-sectional area (A =10 m2) and average dry bulk density of soil (ρ = 1.5 t/m3), divided by the time over which gullies have developed (t = 100 y):

tDCAGS ρ= Equation 11

In the Herbert catchment the gullies are assumed to have been active since the early 1900’s when many of the gullies developed as a consequence of land use development and forest clearance. Sediment generated from gullies contributes to both suspended and bedload sediment. It was assumed that 50% of the gully erosion produced coarse bedload sediment and 50% produced fine suspended sediment, based upon the typical particle composition of gully materials.

5.4 River Bank Erosion Hazard In most of Australia, including the Herbert River catchment, there are few direct measurements of bank erosion. A global review of river bank migration data (Rutherfurd, 2000) suggested that the best predictor of bank erosion rate (BE m/y) was bankfull discharge (Qbf). The magnitude of bankfull discharge is often equivalent to a 1.58 year recurrence interval. Rutherfurd’s original relationship appears to overestimate bank erosion along the lower reaches of the main rivers when compared with river loads measured at gauging stations. Rutherfurd also found a significant relationship between bank erosion and stream power ( xbf SgQρ , where p is the density of water, g is the acceleration due to gravity, bfQ is bankfull discharge in m3/s and Sx is the energy slope approximated to the channel gradient). It is assumed that the bank erosion rate decreases as the proportion of remnant riparian vegetation (PRx) along the river link increases. This produces negligible bank erosion under fully intact riparian vegetation. The resultant equation for bank erosion is:

)1)(1(00002.0 008.0 xFxbf ePRSQgBE −−−××××= ρ in m per year Equation 12

Where PR is proportion of riparian vegetation and Fx is the floodplain width (see Figure 7). The average proportion of riparian vegetation within each link was determined from data collected for the State of the Rivers (SOR) Report: “An Ecological and Physical Assessment of the Condition of Streams in the Herbert River Catchment” (Moller, 1996). Although locally more accurate surveys exist for selected reaches of the catchment (such as the mapping done for sections of the lower Herbert and Stone Rivers - see http://www.hric.org.au/geoMine/meta_display.asp?MetaID=56) the SOR survey is at present the only catchment wide survey. In the SOR survey a riparian condition rating of between 0 and 100 was assigned according to riparian vegetation width and structure (Figure 12). A rating of 81 to 100 for example represents very good condition while 0 to 20 represents very poor condition. Of

25

the 204 subsections across the catchment, approximately 25 were found to have no assigned rating. These were subsequently assigned a rating according to those of neighboring streams and evaluation of vegetation cover maps. The predicted bank erosion rate is converted into sediment supply (kt y-1) by multiplying BE by channel length (m), bank height (m), average particle density of bank materials (1.5 t m-3) and dividing by a conversion factor of 1000. Sediment generated from bank erosion contributes to both suspended and bedload sediment supply. Given the lack of data describing the sediment size distribution of bank sediments in the Herbert River catchment, estimates were taken from sediments in the Burdekin Catchment (Roth et al., 2003). A 50:50 split was used for the contribution to the fine and coarse sediment budgets. An example of bank erosion in the Upper Herbert (on Nettle Creek) is shown in Figure 11.

Figure 11: Example of bank erosion in the Upper Herbert (on Nettle Creek)

Figure 12: Riparian Vegetation Mapping for the Herbert River Catchment (source: State of River Data Base, NR&M, Brisbane)

26

5.5 Mine site sources Mining has long been considered one of the major sources of sediment in the Herbert River Catchment. Various forms of mining, particularly tin mining, have been extensive in the Upper Catchment. The number of relic mine workings and diggings have been reported to be in the hundreds (pers comm. Herberton Shire Engineer), however, very few of these are actually marked on landuse or topographic maps. There have also been very few documented studies of the status and role of these mine sites in terms of their role as a sediment source in the catchment. The location of the larger mines are shown in Figure 2. Mining in the Upper Herbert was conducted using a number of methods including sluicing and dredging. Sluicing generally involves blasting the tertiary tin deposits with high pressure water, and dredging involves direct mining of the river bed. Both processes essentially speed up the natural erosion rate by several orders of magnitude. These processes have left many of the remnant mining areas with lose, unstructured soil profiles that are extremely vulnerable to erosion (e.g. Figure 13). Over the last 60-70 years, many of the older mining sites have probably evacuated much of their erodible overburden down to bedrock, and are now relatively stable.

Figure 13: Typical mine site in the Upper Herbert (photo taken from a light plane)

To our knowledge, there is no published (or unpublished) data on the rates of soil loss from mined areas in the Upper Herbert Catchment. As a result, both the hillslope erosion and gully erosion factors were derived from knowledge of erosion rates from other mine sites around Australia. There are numerous publications of sediment loss from mine sites, however, very few provided long term real time erosion rates that were suitable for the Upper Herbert conditions. A paper by Carroll et al., (2000) from Curragh, Oaky Creek and Goonyella Mines in Central Queensland measured runoff, sediment and water quality from 0.01 ha plots on three gradients (10, 20 and 30%). For each gradient there were both pasture and tree treatments. The average sediment loss over the six year study for the 20% slopes were between 78 t/ha/yr and 280 t/ha/yr. The slopes of the mine sites in the Herbert are much lower (<5%) so the rates from the central Queensland mine sites (from the 10% gradient) were roughly halved and applied to the Herbert River areas. Hence, a total erosion rate of 25 t/ha/yr (as a combination of hillslope and gully sources) was applied to the mine sites. This is considered realistic given the 100 year modelling time frame. In addition to the hillslope erosion, the mine sites were given a gully density of 7 km/km2. This is based on field observations of existing mine sites and knowledge of mine sites in southern Australia.

27

5.6 Nutrient Sources As described in Section 4.2, the nutrient budget deals with different forms of nitrogen (N) and phosphorus (P). The main source terms for the nutrient budget are sediment associated nutrients (from hillslope erosion, gully erosion and riverbank erosion), dissolved organic (DOP and DON) and inorganic loads in runoff water (DIN and DIP). Note that due to a lack of appropriate data, point sources were not assessed (e.g sewerage treatment plants) in this study.

Sediment associated nutrients The nutrient load from hillslope erosion is calculated as the product of the hillslope sediment yield (hillslope erosion x HSDR) multiplied by the nutrient concentration of this load (NC). The nutrient concentration of the sediment load is determined from the proportion of clay and nutrient concentration of the bulk soil (SC). ANNEX uses a two-part mixing model that assumes all nutrients are associated with the clay fraction. For internal catchment links where the percentage clay is greater than the HSDR, all sediment delivered to the channel is assumed to be clay. The nutrient concentration is then the bulk soil concentration enriched by the proportion of clay (Cp) in the hillslope soil:

For Cp > HSDR, 5.0*CpSCNC =

Equation 13

In the few cases where the proportion of clay is less than the HSDR, only a portion of the delivered sediment is clay and so the nutrient concentration is reduced by the ratio of the proportion of clay to the HSDR. In this project it was found that the nutrient enrichment ratio simulated by the above equation gave nutrient concentrations significantly higher than those observed on river sediments in the region. The enrichment ratios were also larger than those recorded in field experiments. Thus we reduced the effect of nutrient enrichment by half (the 0.5 factor in Equation 13). Data on soil clay proportions and nutrient concentrations for P and N were extracted from the Australian Soil Resource Information System (Henderson et al., 2001). The nutrient loads from riverbank and gully erosion are calculated as the product of their respective sediment yields times the soil nutrient concentration, which for phosphorous was taken to be 0.25 g kg-1 and for nitrogen 1 g kg-1.

Dissolved nutrients Estimation of dissolved loads due to surface and sub-surface runoff differs from that used in the NLWRA project. In this project, the concentrations of DIN (dissolved inorganic nitrogen), DON (dissolved organic nitrogen), FRP (filterable reactive phosphorus) and DOP (dissolved organic phosphorus) were assessed from water quality studies from the Herbert River Catchment, and the north Queensland region, and are summarized in Table 2. Data for the concentrations of dissolved nutrients were derived from water quality studies where one landuse was dominant. In theory, the concentrations occurring in major runoff events i.e. event mean concentrations (EMCs) were derived. However, in many cases, EMCs were not easily calculated from the existing data sets. Where EMC’s were not available, 80 percentile values from the complete data sets were used. Concentrations of dissolved nutrients in event conditions from catchments with multiple land-uses were also used as a check on individual landuse values. The nutrient concentration for the internal catchment of a river link was calculated as the area weighted mean for the land uses in that catchment area. The load delivered to the river link is the product of the mean nutrient concentration and the mean annual volume of runoff generated in the internal catchment area. This is the predicted difference in mean annual flow between the upstream and downstream nodes of the stream link (see hydrology Section 5.1). Any differences in runoff between land uses are not accounted for. No distinction is made between surface and subsurface runoff for nutrient concentration.

28

Table 2: Estimated average concentrations of DIN, DON, FRP and DOP in runoff from land uses Wet Tropics Regions. These values were used for the nutrient modelling in the Herbert River Catchment.

Landuse DIN (µg/L)

DON (µg/L)

FRP (µg/L)

DOP (µg/L)

Source

Rainforest 40 150 10 10 Butler et al., 1996; Cogle et al., 2000; 2002; Devlin et al., 2001; Faithful, 1990; Faithful and Brodie, 1990; Furnas, in press; Hunter et al., 2001

Ungrazed savannah/ woodland

100 100 20 10 Furnas, in press; Jackson and Ash, 2001; Prove and Hicks. 1991; Schmidt and Lamble, 2002; O’Reagain et al., 2001

Grazing 200 250 50 12 Furnas, In Press; Nelson et al., 1996; Noble and Collins, 2000; Prove and Hicks, 1991; O’Reagain et al., 2001; Brodie, unpublished

Sugar cane 1100 350 40 30 Armour et al., 1999; Baskeran et al., 2002; Bauld et al., 1996; Biggs et al., 2000; 2001; Bohl et al., 2000, 2001; Bramley and Muller, 1999; Bramley and Roth, 2002; Brodie et al., 1984; Clayton and Pearson, 1996; Congdon and Lukacs, 1996; Devlin et al., 2001; Eyre and Davies, 1996; Hunter et al., 1996; 2001; Mitchell et al., 1997; 2001; Moody et al., 1996; Rasiah and Armour, 2001; Rasiah et al., 2002; Reghenzani et al., 1996 Verberg et al., 1998; Weier, 1999; Wilhelm, 2001; White et al., 2002

Horticulture

500 200 30 20 Cogle et al., 2000; 2002; Hashim et al., 1997

Bananas 700 250 100 20 Hunter, 1997; Hunter et al., 1996; 2001; McKergow et al., 1999; Mitchell et al., 2001; Moody et al., 1996

Grains 500 300 60 20 Noble et al., 1997; Noble and Collins, 2000; Noble, pers.com.; Dilshad et al., 1996

Forestry 150 150 8 8 Bubb et al., 2001; 2002; Bubb and Croton, 2002

Nutrient sinks ANNEX includes three loss terms: (1) deposition of sediment associated nutrients on floodplains, (2) storage of all forms of nutrients in reservoirs, and (3) denitrification of DIN. The nutrient load deposited with sediment on floodplains and reservoirs is the product of the suspended sediment load (SS). The loss of nutrients to reservoirs was not relevant to this study. Losses of DIN to denitrification in both in channel and flood flows were modelled as an exponential decay process that is a function of the residence time of water in the river channel or associated floodplain. Residence time is the ratio of water surface area (A) to discharge (Q). A temperature and substrate dependent assimilation rate coefficient (k) was used. The DIN yield from a link is:

QkA

inout eDINDIN −= Equation 14

For channel denitrification the channel bed area over which the assimilation occurs is estimated as the product of the link mean bankfull width (w) and the link length (L). In reservoirs, the area term is the reservoir area and the representative flow is the median over-bank flow (Qmo). On floodplains it is the floodplain area and the median over-bank flow. The assimilation rate coefficients (k, m/d) are based on measurement of denitrification in Australia from Ford (pers. comm.): kSAND = 0.0001 * T Equation 15kMUD = 0.0002 * T Equation 16 where, T is the mean annual water temperature (ºC) which was assumed to be reasonably estimated by the mean annual air temperature. The mean annual air temperature for each link was sourced from the ANUCLIM grid surface (Houlder et al., 2000).

29

Phosphorus transformations ANNEX assumes that phosphorus exchange occurs between dissolved phase and sediment-adsorbed phase. The concentration of P in sediment (Pss, g/kg) is a constant ratio of the concentration in water (Pdis, mg/l) determined by the adsorption isotherm (Kd, m3/kg): Pss = Kd * Pdis Equation 17 This, together with inputs of P, is used to determine the filterable reactive phosphorus (or FRP) load dissolved in runoff, and the sediment associated P load for each river link. A Kd value of 40 applied to the whole catchment gave satisfactory results and is consistent with limited observations (Young et al., 2001).

6 Results

6.1 Sediment sources to the stream network

6.1.1 Overview This section presents the results of the sediment budget for the Herbert River Catchment. It is important to stress that the results are not to be interpreted at a property level. The data inputs, and thus the results, are based on ‘average’ estimates for particular conditions, and although the natural variability of the landscape is taken into consideration wherever possible, not all levels of variability can be incorporated. For example, it is acknowledged that the level of cover on an area of grazing land will vary from property to property and from season to season. However, data describing this variability is not available and therefore an ‘average’ cover level for all grazing lands on the same vegetation type (e.g open eucalypt woodland) is used. Hence, the values obtained at the catchment and sub-catchment level can be interpreted as a indicator of the average condition of that area of the catchment, and of course, over time, these values would fluctuate. On the other hand, the variability of other factors such as rainfall, topography etc have been incorporated where suitable data exists. A summary of the total sediment budget for the Herbert River Catchment is shown in Table 3. The most important figure in this table is the amount of sediment shown to be delivered to the estuary. In the Herbert River it is estimated that approximately 680,000 tonnes of sediment is delivered to the estuaries and coastal region each year. A discussion of the different sources of sediment that contribute to this total is discussed in the following sections. A discussion of the accuracy of this sediment budget is also discussed in Section 6.2.3. Table 3: Summary of sediment budget for the Herbert River Catchment

Sediment budget item Predicted mean annual

rate (t y-1)

Hillslope Delivery 490,000

Gully erosion rate 220,000

Riverbank erosion rate 220,000

Total sediment supply 930,000

Total suspended sediment stored 40,000

Total bed sediment stored 210,000

Sediment delivery to the estuary 680,000

Total losses 930,000

30

6.1.2 Hillslope erosion The total amount of sediment derived from hillslope erosion in the Herbert River Catchment is predicted to be approximately 490,000 t/yr. This represents ~52% of the total sediment load delivered to the stream network. Figure 14 shows the gross or total erosion from hillslopes and Figure 15 shows the contribution of hillslope erosion that actually makes it to the stream network. The two figures shows that there is an order of magnitude difference between the amount of erosion and the amount of sediment that actually gets to the river system. This is because, on average, only between 5-10% of sediment moving on hillslopes finds its way to streams. In many cases material eroded on a ridge might end up being deposited in colluvial fans on flatter valley bottoms or on river frontage areas before reaching a stream. This process is taken into consideration by applying a sediment delivery ratio to the data. It is important to note that although not all areas are necessarily contributing to sediment loss to the river network, the area may still be eroding, and this may not be beneficial to agricultural production. An example of this can be found by comparing the south west corner of the upper catchment in Figure 14 and Figure 15. The results shown in Figure 15 suggest that the highest levels of hillslope erosion are found in the mine sites around the upper Herbert (> 10 t/ha/yr), the forested and grazing areas of the middle catchment (0.2 – 1.0 t/ha/yr) and the steeper cultivated sections of the lower catchment (0.2 – 5.0 t/ha/yr). Areas in the north of the upper catchment are also shown to have higher hillslope erosion rates (~0.4 t/ha/yr). Slope is a strong driver of hillslope erosion which is why the forested areas are shown to have higher hillslope erosion levels. It is important to note that the areas classified as ‘forest (National Park, State Forest or Reserve) contain a range of different forest types from open eucalypt, rainforest and melaleuca species. The areas that had the highest erosion rate are generally the open eucalypt forest areas within the Herbert River Gorge region. The rainforest and melaleuca areas had much lower erosion levels. It is important to note that these sites already have dense vegetation cover, and if they were cleared, the hillslope erosion rates would be much higher. Most of the native pasture areas that are used for grazing in the upper catchment had relatively low levels of hillslope erosion. This can be attributed to moderate to good cover levels, and relatively low slopes. It is important to note that the average level of cover used for the grazing areas was 60% and this is considered an upper level cover estimate for many of these areas; many sites maintain much lower cover levels. Therefore it is re-emphasised that the results presented in this report cannot be used at a property level as the actual levels of cover vary considerably. It is important to make sure that these areas continue to maintain appropriate cover levels otherwise they have the potential to become a much greater contributor to hillslope erosion. Evidence of the implications of low cover in grazing lands can easily be found in the neighbouring Burdekin Catchment, and the scenario modelling presented in Section 7 of this report will show the impact of reduced cover levels on hillslope erosion.

The areas in the catchment that have an erosion rate of >5 t/ha/yr only occupy less than 2% of the catchment, but contribute 10% of their sediment to the river mouth. These figures suggest that by reducing the erosion losses of these high erosion areas will proportionally have a greater impact on reducing the overall loads at the mouth of the catchment.

31

Figure 14: Predicted Total Hillslope Erosion for the Herbert River Catchment (RKLSC is the gross hillslope erosion not including the sediment delivery ratio calculations).

Figure 15: Contribution of Hillslope erosion to streams (including SDR estimates)

32

Contribution of hillslope erosion from different landuse types The contribution of soil loss from hillslope erosion for the different landuse areas within the Herbert River catchment is shown in Table 4. The table shows the percentage area of each landuse and the associated erosion rates in t/yr and t/ha/yr. The total loss rates in t/yr suggests that the areas under grazing and forest are the largest contributors to hillslope erosion with a total of ~ 200,000 t/yr. Together these land use’s represent ~ 80% of the total sediment lost from hillslope erosion. The next highest contributor is the mine sites (40,000 t/yr) and then sugar cane (35,000 t/yr). Table 4: Contribution of hillslope erosion from each of the major landuse areas with the Herbert River Catchment. (1) The forest and other reserves category contains a range of forest types from open eucalypt woodland to rainforest and melaleuca species.

Landuse Area (km2)

Proportion of total area (%)

Predicted total Soil Loss (t/yr)

Average erosion rate (t/ha/yr)

Native Pastures 5840 60% 200,000 0.34 Improved Pastures 48 0.5% 1,000 0.10 Forest and Other Reserves(1) 2881 30% 200,000 0.69 Sugar Cane 690 7% 35,000 0.51 Other Agriculture 79 0.8% 5,000 0.64 Residential or Industrial 22 0.2% 0 0.01 Mining 52 0.5% 40,000 7.76 Other (e.g. quarries) 135 1.4% 10,000 0.74 The high hillslope erosion obtained for grazing areas is essentially because grazing occupies ~60% of the catchment. It is interesting to note that the forest and other reserves area contributes the same amount of sediment (200,000 t/yr), however, it only occupies 30% of catchment. This is because most of the forested sites are on extremely steep slopes with high rainfall. The high erosion sites are also dominated by the open eucalypt woodland areas (rather than rainforest) within the Herbert River Gorge. Any other landuse on this type of topography would yield much higher sediment loads, so it is not always suitable to compare loads from lowland or flat landscapes with no vegetation with forested areas on steep slopes. When the data is converted to t/ha/yr the highest per unit contributors are mining (7.76 t/ha/yr), (followed by ‘other’), then ‘forest and other reserves’ (0.69 t/ha/yr), then ‘other agriculture’ (0.64 t/ha/yr) and then areas under sugar cane (0.51 t/ha/yr).

6.1.3 Gully erosion The total amount of sediment eroded from gullies in the Herbert River Catchment is predicted to be ~220,000 t/yr, which represents 24% of the total sediment load at the mouth of the catchment. The gully density pattern within the Herbert River Catchment is shown in Figure 16. Overall the gully densities were relatively low in the Herbert River Catchment (Table 5). The highest gully densities are found in the mine site areas around Mount Garnet (> 5 km/km2) which produce on average 16 t/ha/yr of sediment from gullies. This is followed by the steeper grazing areas in the middle and upper catchment that produce ~ 0.17 t/ha/yr (Table 5). The grazing areas in the south west of the upper catchment have low to moderate gully densities of around 0.1 km/km2. Most of the lower catchment has little or no gully erosion.

33

Figure 16: Predicted density of gully erosion for the Herbert River Catchment

Table 5: Contribution of gully erosion from each of the major landuse areas with the Herbert River Catchment. (1) The forest and other reserves category contains a range of forest type from open eucalypt woodland to rainforest and melaleuca species.

Landuse Area (km2)

Proportion of total area (%)

Predicted total Soil Loss (t/yr)

Average erosion rate (t/ha/yr)

Native Pastures 5983 60% 100,000 0.17 Improved Pastures 49 0.5% 0 0.04 Forest and Other Reserves(1) 2935 30% 30,000 0.10 Sugar Cane 715 7% 5,000 0.06 Other Agriculture 79 0.8% 2,000 0.10 Residential or Industrial 25 0.2% 0 0.10 Mining 52 0.5% 80,000 16.48 Other 215 1.4% 3,000 0.09

6.1.4 Riverbank erosion The total amount of sediment eroded from banks in the Herbert River Catchment is predicted at 220,000 t/yr. Approximately 24% of the total sediment load at the mouth of the catchment is the result of bank erosion. Overall the contribution of sediment from bank erosion is predicted to be the same as for gully erosion. Areas of high bank erosion are not necessarily concentrated in a specific part of the catchment. This is because the rates of bank erosion are controlled largely by the proportion of riparian vegetation that is present (see Figure 12) as well as other features such as channel slope and discharge. The areas with the highest rates of bank erosion are predicted for sections of the Blunder Creek catchment, downstream of the junction of the Blunder and Rudd

34

Creek systems, parts of the Herbert River Gorge, the main channel near Abergowrie, and some sections of the main channel near Ingham (Figure 17).

Figure 17: Bank erosion in centimetres per year in the Herbert River Catchment

There are very little quantitative measured data available that estimate bank erosion in the Herbert River Catchment. There is, however, qualitative estimates of bank erosion from the lower Herbert and Stone River Catchments. Data supplied by Ross Hogan (Hinchinbrook Shire), by on work done by Andrew Petroechevsky in 1996/97, suggests that ~ 75,000 tonnes of sediment is eroded from parts of the lower Herbert, Stone and Ripple Creek streams. This estimate is based on site assessment of less than 25% of the catchment so a total estimate of 220,000 t/yr bank erosion for the entire catchment is definitely plausible.

6.2 Summary of sediment results

6.2.1 Contribution of sediment to the coast

Suspended Sediment Contribution to the coast One of the major strengths of the SedNet model is the ability to show where in the catchment the major sediment sources are derived. Because of the various areas that sediment can be deposited in a landscape, areas that are closer to the coast are more likely to contribute sediment to the coast, even if the actual soil erosion losses (per area) are less than in areas that are further away from the coast. Figure 18 shows the fine contribution to the coast from different parts of the catchment. Despite the fact that the streams around Mount Garnet are a long way from the coast, the much higher than average sediment loads suggest that these areas still contribute between 0.5 and 2.0 t/ha/yr to the mouth of the river. The mine site areas in the Mount Garnet region produce greater

35

than 8 t/ha/yr from hillslope erosion alone, so even after considerable deposition of sediment down the river network, at least 25% of the sediment eroded still makes it to the coast. The other areas that contribute proportionally higher sediment loads to the coast are found in the area of Abergowrie, downstream of the Herbert River Gorge. However, the areas near Abergowrie were also predicted to be relatively high contributors of fine sediment to the coast, prior to major landuse changes (see Section 7.3).

Figure 18: Total fine sediment contribution to the coast (t/ha/yr)

Coarse Sediment Contribution to the coast The total loss of coarse sediment to the coast is predicted to be ~ 11 t/yr for the Herbert River Catchment. The amount of coarse sediment leaving the Herbert River represents only 2 % of the total sediment budget. This is because most of the coarse sediment is stored in various parts of the stream network (Figure 19). The bed-load sediment budget predicts the accumulation of sand and gravel on the bed of rivers as a result of increased rates of gully and bank erosion. We consider that where historical bed deposition is in excess of 30 cm, there is likely to be some impact on bed habitats. This might be through filling of pools, smothering of cobble beds with finer grained sediment or reduced diversity of bed forms. Our results suggest that only a small percentage of the river network length in the Herbert catchment has bed deposition in excess of 30 cm (Figure 19). The most significant areas impacted are adjacent to the mining areas near Mt Garnet, and sections of the main Herbert River channel near Ingham. Localised deposition of bed material of between 0.3 and 2 m is expected in areas where there is either excessive bank and gully erosion (e.g. Mt Garnet) or where local channel bed slopes are low (e.g. around Ingham).

36

Figure 19: Predicted bed-load deposition in the Herbert River Catchment

6.2.2 Contribution of sediment from major sub catchments The Herbert River Catchment was broken up into 17 sub catchments according to the major tributaries within the stream network (Figure 20). The corresponding names of the sub-catchments are given in Table 6. The main objective of this analysis was to identify specific sub-catchments or areas that are contributing greater sediment loads than other parts of the catchment. The data in Table 6 shows the total sediment loss from each sub-catchment, including hillslope, bank and gully sources. Deposition of sediment is not included. The catchments range in size from ~27,000 to 142,000 hectares and therefore the results are interpreted on a per unit basis. The area with the highest soil loss is the Ripple Creek/Seymour River area with 2.07 t/ha/yr, followed by streams in the Mount Garnet area with 1.90 t/ha/yr and then Yamanie Creek with 1.76 t/ha/yr. The average soil loss for the entire catchment is ~ 0.98 t/ha/yr, and each of these sub-catchments has almost twice the catchment average. The higher values found in the Ripple Creek area can be attributed to the intensive agricultural landuse in the area, high rainfall and poorly drained soils (which can increase surface erosion). The Mount Garnet stream erosion rates can be primarily linked to the relic mine sites in the area and Yamanie Creek is mainly forest and National Park with steep slopes and high rainfall. The Rudd and Gunnawarra Creek systems showed the lowest sediment losses with 0.25 t/ha/yr and 0.23 t/ha/yr, respectively. It is interesting to note that the average sediment yields from the grazing areas in the Upper Herbert catchment compare very well with the sediment losses measured for sections the Burdekin Catchment (see Roth et al., 2003). Likewise, the per hectare losses estimated for the Ripple Creek area match well with the estimates measured by Visser (2003).

37

Figure 20: Subdivision of 17 sub-catchments in the Herbert River Catchment

Table 6: Number and name of the 17 sub-catchments in the Herbert River with the total contribution from each area. The areas are also ranked in order of highest to lowest sediment loss in t/yr and t/ha/yr.

Number Name Area (ha) Gully (t/yr)Hillslope

(t/yr) Bank (t/yr)

Total sediment loss (t/yr)

Total sediment loss (t/ha/yr) Rank

1 Rudd Creek 59,000 10,000 4,000 1,000 15,000 0.25 16

2 Gunnawarra Creek 30,000 5,000 2,000 1,000 7,000 0.23 17

3 Tunmirendah/Poison Creek 49,000 37,000 11,000 2,000 50,000 1.02 7

4 Mount Garnet 62,000 61,000 40,000 17,000 118,000 1.90 2

5 Millstream 37,000 4,000 14,000 1,000 19,000 0.51 14

6 Blunder Creek 55,000 8,000 10,000 30,000 49,000 0.89 8

7 Wild River 61,000 11,000 39,000 1,000 51,000 0.84 9

8 Sunday Creek 44,000 10,000 9,000 8,000 27,000 0.61 13

9 Bell Creek 41,000 7,000 2,000 3,000 12,000 0.29 15

10 Cameron Creek and Blencoe Falls 142,000 37,000 41,000 24,000 100,000 0.71 11

11 Waterfall Creek and Smoko Creek 62,000 9,000 71,000 12,000 92,000 1.48 4

12 Yamanie Creek 33,000 2,000 25,000 31,000 58,000 1.76 3

13 Gowrie Creek and Broadwater Creek 63,000 4,000 58,000 30,000 91,000 1.44 5

14 Ripple Creek and Seymour River 27,000 2,000 25,000 29,000 56,000 2.07 1

15 Stony Creek and Henrietta Creek 50,000 2,000 14,000 23,000 39,000 0.78 10

16 Stone River Catchment 62,000 4,000 66,000 6,000 76,000 1.23 6

17 Coastal streams (Cattle, Trebonne, Mountain, Spring Creeks etc) 99,000 6,000 57,000 6,000 68,000 0.69 12

38

6.2.3 Comparison of modelled sediment results with measured data The preceding sections have concentrated on predictions of the river sediment budgets and their implications with little discussion on tests of their accuracy. For the Herbert River Catchment, there are a number of measured data sets that we can directly compare with the results of the SedNet modelling. Queensland Department of Natural Resources and Mines have been recording suspended sediment data at the Ingham Gauge (116001), upstream of the John Row Bridge. At this station sampling has occurred on 119 days between 1984 to 2001. The samples were taken over a range of flow events that ranged from 90 ML/day to 430,000 ML/day. Only the higher flows (> 1000 ML/day) are considered important for estimating suspended sediment load. A log regression relationship (Equation 18) was applied to the total suspended sediment samples taken at flows greater than 1000 ML/day.

baQTSS = Equation 18where a = 0.042, b = 0.66, TSS is suspended sediment concentration in mg/l, and Q is discharge in Ml/d.

y = 0.04x0.66

R2 = 0.88

1

10

100

1000

1000 10000 100000 1000000

Daily Flow (ML/day)

TSS

conc

(mg/

L)

Figure 21: Total Suspended Sediment Rating curve for the Herbert River at Ingham

The rating curve (Equation 18) when applied to the 87 years of daily flow recorded at the Ingham gauge site (116001) results in a total mean annual load of 284,000 t y-1 estimated to flow past this gauging station. It is important to note that this is the suspended load only and this value is seen as a gross underestimation of the true sediment load. This is because the really large flows, that carry a large proportion of the load, were not sampled. Unfortunately, none of the other gauging stations on the Herbert have recorded suspended sediment data. There has, however, been other data collected in the vicinity of the John Row bridge (Ingham). Horn et al., (1998) measured suspended sediment data in four high flow events between 1994-1998 and obtained an average suspended sediment load of 422,000 tonnes/year. All of the flow events sampled in this study were greater than 150,000 ML/day. AIMS also conducted sediment sampling in the Herbert River between 1995 and 2000. Three sites were sampled down the freshwater section of the river from just below the gorge at Abergowrie, and at the John Row bridge, Ingham. Dalrymple Creek, a tributary of the Herbert River was also sampled at upstream and downstream sites in collaboration with the Bureau of Sugar Experiment Stations (BSES) office at Ingham between 1993 and 1995. An AIMS river logger has been deployed on the Gairlock bridge (downstream of Ingham) for the past 6 years to obtain continuous turbidity measurements through each wet season. The summary results from each of the data sets suggests that the sediment loads for the Herbert River range between

39

143,000 t/yr for 1995/96 to 900,000 t/yr for the 1999/2000 data (M. Furnas, unpublished data). The mean value for all data sets is ~ 540,000 t/yr (Productivity Commission Statement, 2003). Overall the AIMS data is the most suitable for comparison with the modelled results due to the length of data record (which included wet and dry years). The modelled results using SedNet suggest that there is approximately 600, 000 t/yr of fine sediment (not including bed-load) delivered to the area in the vicinity of the Gairloch bridge (Ingham). This compares with the ~ 540,000 t/yr estimated from data measured by AIMS in the vicinity of the Gairloch Bridge (M. Furnas, In Press). Therefore the average modelled values estimated for the entire catchment are within 10% accuracy of the average results based on measured data.

6.2.4 Effect of improved modelling on results (comparison with NLWRA) The results of sediment and nutrient exports for the Herbert River Catchment represent significant improvements over those produced for the national scale National Land and Water Audit Project (NLWRA). This is largely a result of the incorporation of better regional data (eg. land use) as well as some minor improvements to the model itself. A comparison of the sediment budgets for this project, with those from the NLWRA, are shown in Table 7. The specific improvements to the results are summarized below. • Using the local NR&M gauging station data from the Herbert River Catchment allowed for improved regionalisation of flow parameters and their prediction throughout the channel network. These in turn improve estimation of sediment transport and deposition.

• Better regionalisation improved estimates of over-bank flood discharge. Existing flood mapping and higher resolution geological mapping allowed more accurate estimates of floodplain extents. The improved estimate of floodplain deposition, significantly lowered results (30,000 t/yr) compared with the NLWRA result (430,000 t/yr).

• The combination of a better land use map, and measurements of slope angle from the high resolution digital elevation model greatly improved estimation of hillslope erosion. This resulted in a much reduced estimate of hillslope erosion of 430,000 t/yr compared with the 840,000 t/yr of the NLWRA project.

• The use of a better regional map of riparian condition, and improved estimation of bank full discharge resulted in estimates of bank erosion more applicable to local conditions. As a result the contribution of bank erosion to the stream network was larger at 220,000 t/yr compared with the 100,000 t/yr of the NLWRA project.

• Better spatial mapping of gully erosion improved the estimate of gully contribution to the river network. Total gully erosion amounted to 210,000 t/yr for this assessment compared with the 170,000 t/yr of the NLWRA project.

40

Table 7: Comparison of the sediment budget between this project and the NLWRA for the Herbert River Catchment Predicted mean annual rate (t y-1) Sediment budget item

This survey NLWRA

Hillslope delivery 430,000 840,000 Gully erosion rate 210,000 170,000

Riverbank erosion rate 220,000 100,000 Total sediment supply 860,000 1,110,000

Total suspended sediment stored 30,000 430,000 Total bed sediment stored 200,000 120,000

Sediment delivery to estuary 620,000 560,000 Total losses 860,000 1,110,000

6.3 Nutrient sources to the stream network

6.3.1 Contribution of nutrients to the coast The predicted mean annual nitrogen (N) and phosphorus (P) loads for the Herbert River catchment are shown in Table 8. The table shows the total N and total P from particulate (erosion) and dissolved (runoff and sub-surface flows) sources. The spatial patterns of total N and P contribution are shown in Figure 22 and Figure 23, respectively. Table 8: Components of the nutrient (N and P) budget for the Herbert River Catchment.

Predicted mean annual rate (t y-1) Nutrient budget item

Total N Total P

Hillslope to stream delivery 750 150

Gully erosion 110 30

Riverbank erosion 110 30

Dissolved runoff 1900 180

Total supply 2870 390

Floodplain storage 50 15

Denitrification 0 -

Dissolved export 1900 85

Particulate export 920 290

Nutrient delivery to the estuary 2820 375

Total losses 2870 390

The total amount of N and P estimated to be exported by the Herbert River is 2820 t/yr and 375 t/yr, respectively. Dissolved N is by far the greatest contributor to the N budget representing ~ 67% of the nitrogen budget out of the Herbert River catchment. This is expected due to the application of fertiliser in the cane lands of the floodplain area. Fertiliser is readily broken down into dissolved fractions. In contrast particulate P (or sediment attached P) dominates the phosphorus budget (77%). Both the N and P budgets suggest there is relatively little storage of N and P on the floodplain with only 50 kg/ha/yr and 15 kg/ha/yr, respectively. It is important to note that hillslope erosion dominates both the N and P budgets in terms of the particulate export, however, in the case of N, these sources are considerably smaller than the dissolved (or fertiliser based) fractions (Table 8). The contribution of N and P from different areas within the Herbert River Catchment show the same general pattern. Essentially the cane lands are the highest contributors of N with most areas producing between 5-10 kg/ha/yr of N and some of the catchments closer to the coast producing between 10-20 kg/ha/yr (Figure 22). The forested and partially cultivated lands of the middle and

41

northern sections of the upper catchment are producing between 1 - 5 kg/ha/yr of N and the grazing dominated areas of the south west of the catchment are predicted to contribute < 1 kg/ha/yr of N. The results for P show that the areas dominated by cane on the floodplain produce between 0.5 –1.0 kg/ha/yr (Figure 23) and the forested areas contribute between 0.2 –0.5 kg/ha/yr depending on the cover and slope conditions. The steeper cultivated sections of the northern upper catchment produce between 0.2 –2.0 kg/ha/yr and the grazing areas of the south west generally < 0.2 kg/ha/yr.

6.3.2 Comparison of modelled nutrient results with measured data The only measured data available for N and P from the Herbert River Catchment are based on data collected by AIMS (Furnas, In Press). This data suggests that there is between 1,600 t/yr of N and 170 t/yr of P leaving the Herbert River Catchment. These average loads are based on data collected between 1990 and 1995 from the vicinity of the John Row Bridge (Ingham) and were then modelled (adjusted) to account for long-term water flow (Furnas, In Press). The loads estimated by ANNEX for the area near the John Row bridge are 2000 t/yr of N and 300 t/yr of P. The modelled estimates appear to over predict the measured estimates by 20% and 40% for N and P, respectively. This result is rather different to the 10% accuracy obtained between the modelled and measured sediment results discussed in Section 6.2.3. However, there are a number of possible reasons that the modelled and measured results are different: • The years in which nutrient data were measured (1990-1995) were much drier than the

years in which the sediment data were measured (1995-2000). The mean daily flow for the 1990-1995 period is only 7260 ML/day whereas the mean daily flow for 1995-2000 is 11,750 ML/day. Therefore the nutrient concentrations measured in the drier years would have been much lower than in other years. Then when the lower concentrations are extrapolated to the entire flow record, the total loads would be lower than for ‘average’ conditions. The SedNet/ANNEX model has predicted the nutrient loads based on the full record of gauging data for the Herbert River which spans 88 years in the case of the Ingham gauge. It appears that the measured nutrient loads estimated by AIMS may be an underestimate due to the flow conditions in which they were measured. The difference in discharge can therefore partially explain the difference in the measured and modelled values;

• The nutrient data has been collected for the main channel only (at John Row bridge) and have not taken into consideration the coastal catchments. The SedNet/ANNEX results are for the entire catchment (including the coastal streams). The coastal catchments that drain directly to coast are largely cultivated by sugar cane, and would contribute proportionally greater nutrient loads than samples taken at the John Row Bridge. At the John Row Bridge the nutrient loads would be diluted by those areas that are not contributing high nutrient loads (e.g. grazing areas).

• It is uncertain as to whether the measured nutrient data captured those events that were linked to the application of fertiliser on grazing lands. It is anticipated that the highest nutrient concentrations would be experienced in flow events occurring immediately following fertiliser application i.e. a first flush problem. If the first events following fertiliser application were not sampled, then the measured loads may be an underestimate of the actual loads leaving cane areas.

• It is also acknowledged that the concentrations of DIN, DON, FRP and DOP in runoff used in the ANNEX modelling were averages only, and in the future these values may be able to be refined with improved data. This would increase the accuracy of the modelled results.

Despite the low agreement between the measured and modelled results, it is acknowledged that both data sets provide an important insight into nutrient transport within the Herbert River catchment. With increased sampling, and improvements in the model inputs, it is expected that the agreement between the data sets will improve in the future. It is important to note that

42

although there is discrepancy in the results, the spatial pattern of nutrient sources are considered to be fairly accurate. Even if the modelled loads of N and P matched exactly to the measured loads, the relative proportions of N and P leaving the catchment would still be very similar. Therefore, in terms of catchment management and prioritisation of areas with high nutrient loads, the areas on the floodplain would still be contributing proportionally greater N (and probably P) loads, regardless of the total loads at the catchment mouth.

Figure 22: Pattern of N delivered to streams from each watershed within the Herbert River catchment

43

Figure 23: Pattern of P delivered to streams from each watershed within the Herbert River catchment

7 Scenario Modelling

7.1 Background As part of the Herbert SedNet project, we ran the model for three different land-use scenarios. The main purpose of these scenarios is to assess what the sediment and nutrient loads would be in the catchment under different landuse conditions. It is acknowledged that it is very difficult to replicate the exact condition of the catchment under any given situation, therefore, these scenarios are meant as a guide only. Nonetheless, the results will provide a good indictor of the relative change that can be expected if landuse conditions are altered. Three landuse scenarios were run for the Herbert River Catchment using the SedNet model:

1. The pre-1850 or pre-agricultural condition of the catchment; 2. A change in ground cover for the area of the catchment under grazing (decreasing to 20%

and 40% and increasing to 70% cover); and 3. A decrease in fertiliser application in the cane lands of the lower Herbert.

A brief description of the methods used to run each of the scenarios along with the resulting sediment and/or nutrient budgets is outlined below.

7.2 Methods

Scenario 1: pre-1850 sediment budget The pre-1850 scenario was run to estimate the sediment budget for the Herbert River catchment prior to agricultural and urban development, roads and mining. To do this, estimates of pre-

44

European vegetation patterns replaced the current agricultural crops and pastures (data sourced from CSIRO Coastal Zone Project, see http://www.hric.org.au/geoMine/meta_display.asp?MetaID=113.). For example, the areas on the floodplain that are currently covered by sugar cane were returned to melaleuca, grass or sedgeland, eucalypt dominated, rainforest patterns, or mangroves, depending on their position in the catchment. Areas currently under grazing that are dominated by open eucalypt woodland kept the same vegetation pattern, but the overall cover, or C factor, was altered. The C factor in the grazing lands for the current condition was based on 60% ground cover and 25% canopy cover (Rosewell, 1993). The C factor in the pre-1850 scenario reflects 85% cover and 25% canopy for the open eucalypt woodland grazing areas. It is acknowledged that in the dry savannah open woodland areas typical of the upper catchment, it would have been rare, even under natural conditions (which included fire), to have had 100% cover on all hillslopes. The sediment delivery ratios remained the same as for the current modelled estimates. These estimates are based on annual cover averages. In addition to adjustments to the hillslope erosion, gully erosion was set to 5% of current gully erosion. For bank erosion there was only one adjustment, and that was to assume a 100% riparian vegetation cover on all banks.

Scenario 2: Changing cover on grazing lands The main aim of the grazing scenario was to assess what happens to the hillslope erosion rate when the average cover level is varied on grazing areas. We assessed four different cover levels - 20%, 40%, 60% and 70%, with the 60% cover being the cover level used for the current catchment condition. The reason we have modelled a range of cover levels is because it is acknowledged that depending on the weather, rainfall and stocking rates, the level of cover can vary significantly between different parts of the catchment even under the same landuse. In an ideal modelling situation, the actual cover levels at a given time of the year for different parts of the grazing landscape should be used as inputs to the hillslope erosion model. This would greatly increase the accuracy of the results. Unfortunately this kind of data does not exist, and therefore average cover levels must be assumed. By using a range of cover levels, it is hoped to demonstrate the impact these average conditions can have on the overall hillslope erosion from grazing lands. To model the changes in cover in the grazing lands, the main factor that was altered was the C factor in the USLE computation. The C factor in the grazing lands for the current condition was based on 60% ground cover and 25% canopy cover (Rosewell, 1993). There were three different C factors used in the grazing cover scenarios to reflect the different cover levels (20%, 40% and 70% cover). All scenarios assumed a 25% canopy cover. It is important to note that these estimates are based on annual cover averages.

Scenario 3: Reduced fertiliser application for cane lands The main aim of the reduced fertiliser scenario was to determine what happens to the nitrogen budget, and in particular, to the dissolved inorganic nitrogen (DIN) budget, when the average amount of fertiliser is reduced in cane lands. Fertilisers are considered to be an important source of DIN in the Herbert River Catchment. DIN is also considered the greatest threat to marine ecosystems for a number of reasons: (1) DIN is completely and immediately bio-available; and (2) DIN (and DON) as dissolved components of the river discharge travel widely in the Great Barrier Reef lagoon (unlike particulate forms of N that are often trapped in the estuary/inshore) (Devlin et al., 2001; pers. comm. Jon Brodie). In this study, the current average application rate of fertiliser was assumed to be ~ 200 kg/ha/yr of nitrogen composed of a combination of fertiliser (160 kg/ha/year), trash mineralisation return (30 kg/ha/year) and mill mud (10 kg/ha/year). It is estimated that 160 kg/ha/year is derived from fertiliser which is the BSES recommendation in the Herbert area (also Schroeder et al., 1998). Approximately 30 kg/ha/year of N is from trash mineralisation, however, this is considered to be a conservative estimate as Robertson and Thorburn (2000) say that trash return of N provides 40 to 100 kg/ha/year in their

45

trials after 6 years of green cane trash burning. A further 10 kg/ha/year from mill mud (the mill mud is 'smeared out' in the sugarcane lands of the Herbert in the scenario). Mill mud application rates are in fact much higher but the mud is only used on a proportion of farms (Barry et al., 1998). For the scenario, the application rate is reduced to ~ 130 kg/ha/yr from a combination of the above components. The 130 kg/ha/yr value is based on crop replacement needs and can be supported by economic arguments (e.g. Mallawaarachchi et al., 2002). The change in fertiliser application was applied to those areas under cane only.

7.3 Findings

Scenario 1: pre-1850 sediment budget As shown in Table 9, the pre-1850 sediment budget estimates a total sediment loss of ~110,000 t/yr at the mouth of the river. The current sediment export of ~680,000 t/yr represents a six fold increase in the amount of sediment getting to the estuary of the Herbert River Catchment. Other studies have reported variations in the level of pre-1850 to current sediment export with Neil et al., (2002) suggesting a 3.5 fold increase in sediment yield in response to land use intensification for the Herbert River, and GBRMPA (Brodie et al., 2001) suggesting there has been an eight fold increase in sediment export since European settlement. Figure 24 shows the main areas that would have contributed to the fine sediment load at the mouth of the Herbert River catchment under pre-1850 land cover conditions. The data suggests that most of the catchment would have contributed less than 0.1 t/ha/yr to the coast. Some other areas, particularly around the steeper gorge sections, may have contributed between to 0.75 and 1.0 t/ha/yr. Again this higher level is expected even under natural conditions because of the steeper slopes and higher rainfall in the area. Table 9: Current 2003 sediment budget compared with the pre-1850 budget.

Predicted mean annual rate (t y-1) Sediment budget item

This survey Pre-1850

Hillslope delivery 490,000 100,000

Gully erosion rate 220,000 10,000

Riverbank erosion rate 220,000 20,000

Total sediment supply 930,000 130,000

Total suspended sediment stored 40,000 0

Total bed sediment stored 210,000 20,000

Sediment delivery to the estuary 680,000 110,000

Total losses 860,000 130,000

46

Figure 24: Predicted pre-1850 contribution of sediment to the coast for Herbert River Catchment. For comparison, Figure 18 shows such contributions for current conditions.

Scenario 2: Changing cover on grazing lands Approximately 60 % of the catchment is covered by open eucalypt woodland and the majority of this area is used for grazing. In this scenario, the ground cover factor for the open eucalypt woodland areas was varied to simulate different cover levels over the grazing areas. Cover levels of 20%, 40% and 70% were modelled. These all deviated from the current conditions which were based on an average cover of 60%. Table 10 summaries the predicted hillslope erosion losses under the three different cover levels. The predicted loss of sediment from hillslope erosion based on the 20% cover suggests that there could be a loss of around 1,780,000 t/yr to the mouth of the Herbert River catchment. Based on this scenario, most of the grazing lands would have an average soil loss of between 5–10 t/ha/yr (Figure 25). The predicted loss of sediment from hillslope erosion based on the 40% cover suggests that there could be a loss of around 950,000 t/yr. Based on this scenario, most of the grazing lands would have an average soil loss of between 2-10 t/ha/yr (Figure 26). The predicted loss of sediment from hillslope erosion based on the 70% cover suggests that there could be a loss of around 360,000 t/yr. This scenario has much lower hillslope erosion losses of between 0.5-4 t/ha/yr (Figure 27).

47

Table 10: Comparison of predicted hillslope erosion yields for the different cover levels on ‘open eucalypt woodland’ grazing areas.

% cover on grazing areas

Predicted contribution from hillslope erosion (t/yr) to the mouth of the

Herbert River 20% 1,780,000 40% 950,000 60% 490,000 70% 360,000

The 70% cover scenario was considered a good estimate of the ‘best practice’ condition for grazing lands in the Herbert catchment (Figure 27). The increase in cover resulted in a total sediment export to the estuaries of 560,000 t/yr which is a reduction of ~ 120,000 t/yr of sediment for just a 10% increase in ground cover (Table 11). This represents a total decrease in sediment load of 18% at the mouth of the river. According the end of catchment targets set by GBRMPA (Brodie et al., 2001), the Herbert River Catchment are to aim for a sediment export of ~ 445,000 t/yr by 2011. Based on the grazing management scenario presented in this report, an increase in ground cover levels of just 10% on all of the open eucalypt woodland grazing areas would mean that the catchment was half way to achieving the targets set by GBRMPA (Brodie et al., 2001). This exercise shows that in the case of the open eucalypt woodland grazing areas, a relatively small change in landuse practices can potentially have dramatic impacts on the end of catchment sediment loads. Table 11: Current 2003 sediment budget compared with the 70% cover on grazing lands budget

Predicted mean annual rate (t y-1) Sediment budget item

This survey 70% grazing cover

Hillslope delivery 490,000 360,000

Gully erosion rate 220,000 220,000

Riverbank erosion rate 220,000 220,000

Total sediment supply 930,000 800,000

Total suspended sediment stored 40,000 30,000

Total bed sediment stored 210,000 210,000

Sediment delivery to the estuary 680,000 560,000

Total losses 860,000 800,000

48

Figure 25: Predicted Hillslope Erosion for the Herbert River Catchment based on an average of 20% cover in the open eucalypt woodland grazing lands (RKLSC is the gross hillslope erosion which does not include sediment delivery ratio calculations). For comparison, Figure 14 shows erosion under current grazing conditions (60% cover).

Figure 26: Predicted Hillslope Erosion for the Herbert River Catchment based on an average of 40% cover in the open eucalypt dominated grazing lands (RKLSC is the gross hillslope erosion which does not include sediment delivery ratio calculations). For comparison, Figure 14 shows erosion under current grazing conditions (60% cover).

49

Figure 27: Predicted Hillslope Erosion for the Herbert River Catchment based on an average of 70% cover in the open eucalypt dominated grazing lands (RKLSC is the gross hillslope erosion which does not include sediment delivery ratio calculations). For comparison, Figure 14 shows erosion under current grazing conditions (60% cover).

Scenario 3: Reduced fertiliser application A comparison of the current nitrogen budget with the reduced fertiliser nitrogen budget for the Herbert River Catchment is shown in Table 12. Essentially all components of the budgets are the same except for the dissolved runoff and dissolved inorganic (DIN) fractions. The reduction in fertiliser input has reduced the dissolved runoff supply to the streams from 1900 t/yr to 1600 t/yr. This has then resulted in a reduced DIN loss from 1000 t/yr to 730 t/yr. This represents ~27% decrease in DIN and an overall decrease of ~10% to the entire N budget. This is a relatively good outcome considering that sugar cane occupies only 7% of the land area. The predicted pattern of N delivered to streams from each watershed for reduced fertiliser use is shown in Figure 28. The main area of difference between Figure 28 (reduced fertiliser) and Figure 22 (current N losses) is the reduced losses of N in kg/ha/yr, from the cane areas of the lower Herbert. The decrease in fertiliser has reduced the per hectare losses from many sections of the lower Herbert from ~ 5 – 20 kg/ha/yr to between 2.5 – 10 kg/ha/yr. This is up to a 50% reduction in N losses for some sections of the lower Herbert. This scenario has shown that for a decrease in fertiliser application (and a relatively small decline in production), sizeable reductions of the amount of dissolved inorganic nitrogen reaching the stream network can be made.

50

Table 12: Comparison of current nitrogen budget with the reduced fertiliser budget for the Herbert River Catchment

Predicted mean annual rate (t y-1) Nutrient budget item

Total N (current) Total N (scenario)

Hillslope to stream delivery 750 750

Gully erosion 110 110

Riverbank erosion 110 110

Dissolved runoff 1900 1600

Total supply 2870 2600

Floodplain and reservoir storage 50 50

Denitrification 0 0

Dissolved inorganic N (DIN) export 1000 730

Dissolved organic N (DON) export 900 900

Particulate export 920 920

Nutrient delivery to estuary 2820 2550

Total losses 2870 2600

Figure 28: Predicted pattern of N delivered to streams from each watershed within the Herbert River catchment for reduced fertiliser use. For comparison, Figure 22 shows current fertiliser use.

8 Discussion and Conclusion Management aimed at reducing sediment and nutrient transport will target each erosion process quite differently. For example, stream bank and gully erosion is best targeted by managing stock access to streams, protecting vegetation cover in areas prone to future gully erosion, revegetating bare banks and reducing sub-surface seepage in areas with erodible sub-soils. Surface wash erosion is best managed by promoting consistent groundcover, maintaining soil structure,

51

promoting nutrient uptake and promoting deposition of eroded sediment before it reaches the stream. Consequently it is quite important to identify the predominant sediment and nutrient delivery process before undertaking catchment remediation or making recommendations for changed grazing practice. The present estimates of sediment and nutrient loads in the Herbert River Catchment represent significant improvements over those that have been previously undertaken. This is a direct result of the incorporation of better regional information. It is clear that all erosion processes contribute significant amounts of sediment to the channel network and owing to the reasonably strong connectivity between source supply and export to the coast, much of this is directly exported from the catchment. All erosion processes are highly variable across the catchment with localised hotspots evident in places. Hillslope erosion is the greatest contributor of sediment to the stream network in the Herbert River Catchment, with the highest overall losses being from grazing lands and forested areas, however, the per hectare losses are greatest from the mine sites, forested areas and agricultural cropping land. It is important to note that the areas covered by ‘forest’ contribute almost the same amount of sediment as grazing, from roughly half the area. This is because most of the forested sites are on extremely steep slopes with much higher rainfall. The main forest type that contributed to the high loads is the open eucalypt dominated woodland areas that occupy the Herbert River Gorge section. The erosive potential in the forested areas is much greater than the flat depositional landscapes of the upper catchment grazing lands. Any other landuse (other than forests) on this type of topography would yield much higher sediment loads, and it is not necessarily appropriate to compare loads from lowland or flat landscapes with no vegetation, with forested areas on steep slopes. Gully erosion also varies significantly across the catchment with the worst areas being around the mine sites near Mt Garnet and the steeper grazing areas upstream of the Herbert River Gorge. Most of the remainder of the catchment does not suffer from significant gully erosion. Riverbank erosion contributes the same amount of sediment to the stream network as gully erosion, yet there is no obvious pattern to bank erosion in the catchment. The main areas that appear to undergo higher than average rates of bank erosion are in the Blunder Creek catchment and various sections of the main Herbert River channel. Bank erosion it generally at its worst in river reaches lacking good riparian zone condition and where stream power is greatest. The main areas that are predicted from the model to suffer significant bed-load accumulation are around the mining streams of Mt Garnet and the low lying reaches near Ingham. The predominant source of nutrients are in dissolved form for nitrogen (N) and particulate form for phosphorus (P). Both the N and P budgets are dominated by losses from cultivated cane lands on the floodplain, followed generally by the forested and cultivated areas of the middle and upper catchment. The grazing dominated areas of the south west of the upper catchment are predicted to contribute low levels of both N and P. The patterns of erosion sources differ suggesting that each process is fairly independent and that in each location an assessment needs to be made of the dominant source of sediment. Given that the highest rates of erosion occur in localised patches, this suggests that erosion control measures targeted to specific areas will be effective in reducing sediment supply to, and loads in, the stream network. The scenario modelling conducted in this study has been able to demonstrate the changes that can be expected in the sediment budget following alterations to landuse management. The results of the grazing scenarios suggest that changing the level of cover on grazing lands can have

52

significant impacts on the amount of sediment lost from hillslope erosion at the mouth of the Herbert River Catchment. In fact, an increase in ground cover levels of just 10% on all of the open eucalypt woodland grazing areas would result in the catchment being half way to achieving the targets set by GBRMPA (Brodie et al., 2001). Conversely, a reduction in cover can dramatically increase the overall sediment yield as well as reduce the productivity of these agricultural landscapes. The nutrient scenario results suggests that a reduction in fertiliser rates from an average of 200 kg/ha/yr to 130 kg/ha/yr can produce a decrease of ~27% to in the dissolved inorganic nitrogen (DIN) levels and a 10% decrease to the overall N budget. This is a relatively good outcome considering that sugar cane occupies only 7% of the land area. Despite the improvements used in this model, there are still areas of further research that would increase the reliability of estimates made in this report. These include:

• on-ground estimates of sediment loads from mining areas in the Upper Herbert; • measurements of bank erosion from a range of sites across the entire catchment; • measured hillslope erosion rates from different cover levels; • measured discharge data from the distributary channels on the coastal plain; • knowledge of actual fertiliser usage rates for cane blocks on the floodplain;

Without this data, further improvements to the results presented in the study will be difficult. The outputs from this research should assist natural resource management agencies and land managers to appropriately target critical areas, so that a comparatively large benefit in reducing sediment and nutrient loads delivered downstream can be achieved with less effort. The scenario testing tool used in this project provides a means to evaluate the relative effectiveness of targeting these areas for rehabilitation. In this way the most cost-effective strategy can be achieved.

9 References Alexander, G.R., Hansen, E.A. (1986) Sand bed load in a brook trout stream. North American Journal of Fisheries Management, 6, 9-23. ANZECC (2000) The Australian and New Zealand Guidelines for Fresh and Marine Water Quality. Australian and New Zealand Environment and Conservation Council and Agriculture and Resource Management Council of Australia and New Zealand. Armour, J., Hunter, H. and Simpson, B. (1999) Nutrient and sediment transport from irrigated sugarcane in Cattle Creek, north Queensland. DNRM unpublished progress report. ASCE (1992) Sediment and Aquatic Habitat in River Systems. Journal of Hydraulic Engineering, 118, 669-687. Baker, J. (Ed) (2003) A Report on the Study of Land-Sourced Pollutants and their Impacts on Water Quality in and Adjacent to the Great Barrier Reef. Brisbane, Queensland, Department of Premier and Cabinet. http://www.premiers.qld.gov.au/about/reefwater.pdf Barry, GA, Price, AM and Lynch, PJ (1998) Some implications of the recycling of sugar industry by-products. PASCT, 20: 52-55. Baskeran, S., Budd, K.L., Larsen, R.M. and Bauld, J. (2002). A groundwater quality assessment of the lower Pioneer Catchment, Queensland. Bureau of Rural Sciences, Department of Agriculture, Fisheries and Forestry, Canberra. Bauld, J., Leach, L.L. and Sandstrom, M.W. (1996) Impact of land use on groundwater quality in the Burdekin River Delta and the Burdekin River Irrigation Area (Abstract only). In: Hunter, H. M., Eyles, A. G., and Rayment, G. E (eds.). Downstream Effects of Land Use. Queensland Department of Natural Resources, Brisbane, Australia. pp. 195-196. Biggs, J., Keating, B. and Thorburn, P. (2000) Time trends of nitrate in groundwaters under intensive agriculture in the Bundaberg region. Proceedings of the Australian Society of Sugar Cane Technologists 22: 296-301.

53

Biggs, J., Thorburn, P., Weier, K.L. and Hopp, M.L. (2001) Nitrate in groundwaters in Mackay and Burdekin regions. Proceedings of the Australian Society of Sugar Cane Technologists 23: 77-83. Bohl, H.P., Mitchell, D.C., Penny, R.S., and Roth, C.H. (2000). Nitrogen losses via subsurface flow from sugar cane on floodplain soils in the Australian wet tropics. Proceedings of the Australian Society of Sugar Cane Technologists 22: 302-307. Bohl, H.P., Roth, C.H., Tetzlaff, D. and Timmer, J. (2001) Estimation of groundwater recharge and nitrogen leaching under sugarcane in the Ripple creek catchment, lower Herbert. Proceedings of the Australian Society of Sugar Cane Technologists 23: 84-89. Bramley, R. G. V. and Johnson, A. K. L. (1996) Land Use Impacts on Nutrient Loading in the Herbert River. In the Proceedings from the conference of the Downstream Effects of Land Use, Hunter, H. M., Eyles, A. G., and Rayment, G. E (Eds). Department of Natural Resources, Brisbane, p93-96. Bramley, R. G. V., and Muller, D. E. (1999). Water Quality in the Lower Herbert River - The CSIRO Dataset. CSIRO Land and Water Technical Report 16/99, CSIRO, Canberra . Bramley, R.G.V. and Roth, C.H. (2002) Land use effects on water quality in an intensively managed catchment in the Australian humid tropics. Australain Journal of Marine and Freshwater Research, 53, 931-940. Brodie, J. (In press). Keeping the wolf from the door: managing land-based threats to the Great Barrier Reef. Proceedings of the 9th Coral Reef Symposium. Brodie, J. Unpublished water quality data from the 2002 Belyando River wet season event flow. Brodie, J.E., Hicks, W.S., Richards, G.N. and Thomas, F.G. (1984). Residues related to agricultural chemicals in the groundwaters of the Burdekin River delta, north Queensland. Environmental Pollution (Series B) 8: 187-215. Brodie, J., Christie, C., Devlin, M., Haynes, D., Morris, S., Ramsay, M., Waterhouse, J. and Yorkston, J. (2001) Catchment management and the Great Barrier Reef. Water Science and Technology, 43, 203-211. Brodie, J., Furnas, M., Ghonim, S., Haynes, D., Mitchell, A., Morris, S., Waterhouse, J., Yorkston, J., Andas, D., Lowe, D and Ryan, M. (2001) Great Barrier Reef Catchment Water Quality Action Plan. Great Barrier Reef Marine Park Authority, Townsville. http://www.gbrmpa.gov.au/corp_site/key_issues/water_quality/action_plan/. Brunskill, G.B., Zagorskis, I. and Pfitzner, J. (2002) Carbon burial rates in sediments and a carbon mass balance of the Herbert River region of the Great Barrier Reef continental shelf, North Queensland. Estuarine, Coastal and Shelf Science, 54, Issue 4, p677-700. Bubb, K.A. and Croton, J.T. (2002) Effects on catchment water balance from the management of Pinus plantations on the coastal lowlands of south-east Queensland, Australia. Hydrological Processes 16: 105-117. Bubb, K.A., Yu, B., Cakurs, U. and Costantini, A. (2000) Australian Forestry 63: 239-245. Bubb, K.A., Frayne, P.F. and Wittmer, T.R. (2001) Australian Forestry 65: 38-46. Butler et al., (1996). Ecotourism, WQ and Wet Tropics Streams. ACTFR Technical Report No. 96/11, Australian Centre for Tropical Freshwater Research, James Cook University, Townsville. Cannon, M.G., Smith, C.D. and Murtha, G.G., (1992). Soils of the Cardwell-Tully Area, North Queensland, CSIRO Division of Soils Divisional Report Number 115 Carroll, C., Merton, L. and Burger, P. (2000) Impact of vegetative cover and slope on runoff, erosion, and water quality for field plots of soil and spoil materials on central Queensland coal mines. Australian Journal of Soil Research, 38, 313-27. Chiew, F.H.S. and McMahon, T.A. (1999) Modelling runoff and diffuse pollution loads in urban areas. Water Science and Technology 39: 241-248. Ciesiolka, C. (1987) Catchment management in the Nogoa watershed. Australian Water Resource Council Research Project 80/129. Queensland Department of Primary Industry, Brisbane, 204 p

54

Clayton, P. and Pearson, R. (1996). Instream effects of dunder application to canefields in the Sarina region, Queensland: Continued monitoring. ACTFR Report No. 96/09, Australian Centre for Tropical Freshwater Research, James Cook University, Townsville. Cogle, A. L., Langford, P. A., Kistle, S. E., Sadler, G., Ryan, T. J., McDougall, A. E., Russell, D. J., and Best, E.(2000) Natural Resources of the Barron River Catchment 2 - Water Quality, land use and land management interactions, Department of Primary Industries, Brisbane.

Cogle, A. L., Langford, P. A., Russell, D. J., and Keating, M. (2002). Sediment and nutrient movement in a tropical catchment with multiple land-use, . In: Dawson, N. and Brodie, J. and others (Eds) Proceedings of the 2nd National Conference on Aquatic Environments: Sustaining our aquatic environments- Implementing solutions. Queensland Department of Natural Resources and Mines, Brisbane, Australia. Cogle, A.L., Rao, K.P.C., Yule, D.F., Smith, G.D., George, P.J., Srinivasan, S.T. and Jangawad, L. (2002). Soil management for Alfisols in the semi-arid tropics: erosion, enrichment ratios and runoff. Soil Use and Management 18(1): 10–17. Congdon, R.A. and Lukacs, G.P. (1996) Water quality aspects of irrigation runoff from the Burdekin River irrigation area. In: Hunter, H.A., Eyles, A.G., Rayment, G.E. (Eds.) Downstream Effects of Land Use. Queensland Department of Natural Resources, Brisbane, pp. 73-76 Devlin, M., Waterhouse, J. and Brodie, J. (2001). Community and connectivity: Summary of a community based monitoring program set up to assess the movement of nutrients and sediments into the Great Barrier Reef during high flow events. Water Science and Technology 43(9): 121-131. DeRose, R.C., Prosser, I.P, Wilkinson, L.J., Hughes, A.O. and Young, W.J. (2002) Regional Patterns of Erosion and Sediment and Nutrient Transport in the Mary River Catchment, Queensland. CSIRO Land and Water Technical Report 37/02, Canberra (http://www.clw.csiro.au/publications/technical2002/tr37-02.pdf) DeRose, R.C., Prosser, I.P, Wilkinson, L.J., Hughes, A.O. and Young, W.J. (2003) Regional Patterns of Erosion and Sediment and Nutrient Transport in the Goulburn and Broken River Catchments, Victoria. CSIRO Land and Water Technical Report 11/03, Canberra (http://www.clw.csiro.au/publications/technical2003/tr11-03.pdf) Dilshad, M., Motha, J.A. and Peel, L.J. (1996). Surface runoff, soil and nutrient losses from farming systems in the Australian semi-arid tropics. Australian Journal of Experimental Agriculture 36: 1003-1012. Eyre, B. and Davies, P. (1996) A preliminary assessment of suspended sediment and nutrient concentrations in three far north Queensland catchments. In: Hunter, H.A., Eyles, A.G., Rayment, G.E. (Eds.) Downstream Effects of Land Use, Queensland Department of Natural Resources, Brisbane, pp. 57–64. Faithful, J. (1990). Tully-Millstream hydroelectric scheme: Water quality sampling and testing program. Dry season monitoring program. ACTFR Technical report No. 90/14, Australian Centre for Tropical Freshwater Research, James Cook University, Townsville. Faithful, J. and Brodie, J. (1990) Tully-Millstream hydroelectric scheme: Water quality sampling and testing program. ACTFR Technical report No. 90/09, Australian Centre for Tropical Freshwater Research, James Cook University, Townsville Furnas, M. In press. Catchments and Corals: Terrestrial Runoff to the Great Barrier Reef. CRC Reef Research, Townsville. Furnas, M. and Mitchell, A. (2001) Runoff of terrestrial sediment and nutrients into the Great Barrier Reef World Heritage Area., 37-52, CRC Press, Florida. Furnas, M. J. and Brodie, J. (1996) Current Status of Nutrient Levels and other Water Quality Parameters in the Great Barrier Reef. Proceedings from the conference of the Downstream Effects of Land Use, Hunter, H. M., Eyles, A. G., and Rayment, G. E (Eds). Department of Natural Resources Brisbane. Department of Natural Resources Brisbane, p9-21. Grundy, M.G. and Bryde, N.J. (1988). Land Resources of the Einasleigh - Atherton Dry Tropics, Queensland. Land Resources Branch, Department of Primary Industries, Q089004. Hashijm, G.M., Ciesiolka, C.A.A., Rose, C.W., Coughlan, K.J. and Fentie, B. (1997). Chapter 7. Loss of chemical nutrients by soil erosion. In: Coughlan, K.J. and Rose, C.W. (eds) A new soil conservation methodology and

55

application to cropping systems in tropical steeplands. Australian Centre for International Agricultural Research, Canberra, pp. 79-100. Haynes, D. (2001) Great Barrier Reef Water Quality: Current Issues. Great Barrier Reef Marine Park Authority, Townsville. Horn, A., Joo, M., and Poplawski, W. (1998) Queensland Riverine Sediment Transport Rates - A Progress Report. Department of Natural Resources Queensland, Brisbane. Heiner, I.J. and Grundy, M.J., (1994). Land Resources of the Ravenshoe - Mt Garnet area north Queensland: Volume 1 - Land Resource Inventory, Department of Primary Industries Queensland. Henderson, B., Bui, E., Moran, C., Simon, D. and Carlile, P. (2001) ASRIS: Continental-scale soil property predictions from point data. Technical Report 28/01, CSIRO Land and Water, Canberra. Horn, A., Joo, M., and Poplawski, W. (1999a) Sand and Gravel Transport Rates in Queensland Rivers. In the Proceedings of the Second Australian Stream Management Conference, Rutherfurd, I. D. and Bartley, R. (Eds). Cooperative Research Centre for Catchment Hydrology, Volume 1,p 335-340. Horn, A., Joo, M., and Poplawski, W. (1999b) Sediment transport rates in highly episodic river systems: a direct preliminary comparison between empirical formulae and direct flood measurements. Water 99 Congress. Department of Natural Resources Queensland. Houlder, D.J., Hutchinson, M.F., Nix, H.A. and McMahon J.P. (2002). ANUCLIM User guide, Version 5.1. Centre for Resource and Environmental Studies, Australian National University, Canberra. Hughes, A.O., and Prosser, I.P. (2002). Gully and Riverbank Erosion Mapping for the Murray-Darling Basin. Technical Reports G and I for the MDBC Basin-wide Sediment Mapping Project, CSIRO Land and Water, Canberra. Hughes, A., Prosser, P., Stevenson, J., Scott, A., Lu, H., Gallant, J., and Moran, C. (2001) Gully erosion mapping for the National Land and Water Audit. Canberra, CSIRO Land and Water. http://www.clw.csiro.au/publications/technical2001/tr26-01.pdf Hunter, H.M. (1997) Nutrients and suspended sediment discharged from the Johnstone River catchment during cyclone Sadie. In: Steven, A. (Ed.) Cyclone Sadie flood plumes in the Great Barrier Reef lagoon: Composition and consequences. Workshop Series No. 22, Great Barrier Reef Marine Park Authority, Townsville, pp. 1-8 Hunter, H.M. and Walton, R.S. (1997). From land to river to reef lagoon. Land use impacts on water quality in the Johnstone Catchment. Queensland Department of Natural Resources, Indooroopilly. 10pp Hunter, H.M., Walton, R.S. Russell, D.J. (1996) Contemporary water quality in the Johnstone River catchment. In: Hunter, H.A., Eyles, A.G., Rayment, G.E. (Eds.) Downstream Effects of Land Use. Queensland Department of Natural Resources, Brisbane, pp. 339-345 Hunter, H., Sologinkin, S., Choy, S., Hooper, A., Allen, W., Raymond, M. and Peeters, J. (2001). Water management in the Johnstone Basin. Queensland Department of Natural Resources and Mines, Brisbane. Jackson, J. and Ash, A.J. (2001) The role of trees in enhancing solid nutrient availability for native perennial grasses in open eucalypt woodlands of north-east Queensland. Australian Journal of Agricultural Research, 52: 377-386. Johnson, A. K. L. and Murray, A. E. (1997) Herbert River Catchment Atlas. CSIRO, Davies Laboratory, Townsville. Ladson, A.R., Tilleard, J.W. (1999) The Herbert River, Queensland, Tropical Australia: Community Perception and River Management. Australian Geographical Studies, 37, 284-299. Keating, B.A., Bauld, J., Hillier, J., Ellis, R., Weier, K.L., Sunners, F. and Connell, D. (1996). Leaching of nutrients and pesticides to Queensland groundwaters. In ‘Downstream effects of land use.’ (Eds. H.M. Hunter, A.G. Eyles and G.E. Rayment). pp. 151-163. (Department of Natural Resources, Brisbane, Australia). Ladson, A.R. and Tilleard, J.W. (1999) The Herbert River, Queensland, Tropical Australia: Community Perception and River Management. Australian Geographical Studies, 37, 284-299.

Laffan, M.D. (1988), Soils and Landuse on the Atherton Tableland, North Queensland, CSIRO Division of Soils Soil and Landuse Series No. 61.

56

Lu, H., Gallant, J., Prosser, I., Moran, C., Moran, C., and Priestly, G. (2001) Prediction of Sheet and Rill Erosion Over the Australian Continent, Incorporating Monthly Soil Loss Distribution. Canberra, CSIRO Land and Water. http://www.clw.csiro.au/publications/technical2001/tr13-01.pdf Lu, H., Moran, C., Prosser, I.P., Raupach M.R., Jon Olley, J., and Petheram, C. (2002). Sheet and Rill Erosion and Sediment Delivery to Stream: A Basin Wide Estimation at hillslope to Medium Catchment Scale". A Report to Project D10012 of Murray Darling Basin Commission: Basin-wide Mapping of Sediment and Nutrient Exports in Dryland Regions of the MDB. Lu, H., I.P. Prosser, C.J. Moran, J. Gallant, G. Priestley and J.G Stevenson (submitted). Predicting sheetwash and rill erosion over the Australian continent, Aust. J. Soil Res. McIvor, J.G., Williams, J. and Gardner, C.J. (1995) Pasture management influences runoff and soil movement in the semi-arid tropics. Australian Journal of Experimental Agriculture 35: 55-65. McCulloch, M., Fallon, S., Wyndham, T., Hendy, E., Lough, J., Barnes, D. (2003) Coral record of increased sediment flux to the inner Great Barrier Reef since European Settlement. Nature, 421, 727-730. McKergow, L., Prosser, I.P. and Heiner, D. (1999) Preliminary results on the effectiveness of riparian buffers in Far North Queensland. Proc. 2nd Australian Stream Management Conf., (Editors) I.D. Rutherfurd and R. Bartley, CRC for Catchment Hydrology, Monash Univ., Clayton Vic., pp. 439-444. McShane, T.J., Reghenzani, J.R., Prove, B.G., and Moody, P.W. (1993). Nutrient balances and transport from sugarcane land – a preliminary report. Proceedings of the Australian Society of Sugar Cane Technologists 15th Conference, 268-275. Mallawaarachchi, T, Moneypenny, R, and Rayment, G. (2002) An integrated strategy to enhance profitability and environmental compliance in the Australian sugar industry. PASCT, 24: 99-106. Mitchell, A., Rasmussen, C., Blake, S., Congdon, R., Regenzani, J., Saffigna, P. and Sturmey, H. (1991) Nutrient concentrations and fluxes in North Queensland coastal rivers and streams. In: Land Use Patterns and Nutrient Loading of the Great Barrier Reef Region, James Cook University, Townsville, p. 108-125 Mitchell, A. W. and Bramley, R. G. V. (1997) Export of nutrients and suspended sediment from the Herbert River catchment during a flood event associated with Cyclone Sadie. In Cyclone Sadie Flood Plumes in the Great Barrier Reef Lagoon: Composition and Consequences, Steven, A. (Ed). Townsville, GBRMPA. Mitchell, A.W., Bramley, R.G.V. and Johnson, A.K.L. (1997) Export nutrients and suspended sediment during a cyclone-mediated flood event in the Herbert River Catchment, Australia. Marine and Freshwater Research, 48, 79-88. Mitchell, A. W., Reghenzani, J. R., Hunter, H. M., and Bramley, R. G. V. (1996) Water Quality and Nutrient Fluxes from River Systems Draining to the Great Barrier Reef Marine Park. Proceedings from the conference of the Downstream Effects of Land Use, Hunter, H. M., Eyles, A. G., and Rayment, G. E (Eds). Department of Natural Resources Brisbane. Department of Natural Resources Brisbane, p.23-33. Mitchell, A. Reghenzani, J.R. and Furnas, M. (2001) Nitrogen levels in the Tully River – a long-term view. Water Science and Technology 43 (9), 99-105. Moller, G. (1996) An Ecological and Physical Assessment of the Condition of Streams in the Herbert River Catchment. Department of Natural Resources, Brisbane. Moody, P.W., Reghenzani, J.R., Armour, J.D., Prove, B.G. and McShane TJ (1996) Nutrient balances and transport at farm scale- Johnstone River catchment. In: Hunter, H.A., Eyles, A.G., Rayment, G.E. (Eds.) Downstream Effects of Land Use. Queensland Department of Natural Resources, Brisbane, pp. 347-351 Neil, D.T., Orpin, A.R., Ridd, P.V., Yu, B. (2002) Sediment yield and impacts from river catchments to the Great Barrier Reef Lagoon. Mar. Freshwater Res., 53, 733-752. NLWRA (2001) Australian Agricultural Assessment. National Land and Water Resources Audit, Canberra, http://www.nlwra.gov.au/ Nelson, P.N., Cotsaris, E. and Oades, J.M. (1996) Nitrogen, phosphorus and organic carbon in streams draining two grazed catchments. Journal of Environmental Quality 25: 1221-1229

57

Noble, R.M., Duivenvoorden, L.J., Rummenie, S.K., Long, P.E. and Fabbro LD (1997) Downstream effects of land use in the Fitzroy catchment. Queensland Department of Natural Resources, Brisbane, 97pp Noble , R. and Collins, C. (2000) Physical and chemical water quality of the Dawson River. In: Noble, R. (ed) River health in the Fitzroy Catchment: Community ownership. Queensland Department of Natural Resources, Brisbane, p 25 – 39. Northcote, K.H., Beckman, G.G., Bettenay, E., Churchward, H.M., van Dijk, D.C., Dimmock, G.M., Hubble, G.D., Isbell, R.F., McArthur, W.M., Murtha, G.G., Nicholls, K.D., Paton, T.R., Thompson, C.H., Webb, A.A., and Wright, M.J. (1960-1968), Atlas of Australian Soils with Explanatory notes, CSIRO and Melbourne University Press, Melbourne. O’Reagain, P., Bushell. J. and Allen, R. (2001). The effect of grazing management on soil and nutrient loss in the tropical savannas of north Queensland. Report to the Great Barrier Reef Marine Park Authority, Department of Primary Industries, Charters Towers, 42pp. Pearson, R. and Clayton, P. (1993). Teemburra Creek dam study. Part A – Hydrological, biological and water quality environmental surveys. ACTFR Technical report No. 93/01, Australian Centre for Tropical Freshwater Research, James Cook University, Townsville Productivity Commission (2003) Industries, Landuse and Water Quality in the Great Barrier Reef Catchment, Research Report, Canberra. http://www.pc.gov.au/study/gbr/index.html Prosser, I., Hughes, A., Rustomji, P., and Moran, C. (2001a) Predictions of the sediment regime of Australian rivers. In the Proceedings of the Third Australian Stream Management Conference, Rutherfurd, I., Brierley, G., Bunn, S., Sheldon, F., and Kenyon, C. (Eds). Cooperative Research Centre for Catchment Hydrology, Melbourne. p. 529-533. Prosser, I., Young, B., Rustomji, P., Moran, C. and Hughes, A. (2001b) Constructing River Basin Sediment Budgets for the National Land and Water Resources Audit. CSIRO Land and Water Technical Report 15/01, Canberra. (http://www.clw.csiro.au/publications/technical2001/tr15-01.pdf) Prosser, I.P, Moran, C.J., Lu, H., Scott, A., Rustomji, P., Stevenson, J., Priestly, G., Roth, C.H. and Post, D. (2002) Regional Patterns of Erosion and Sediment Transport in the Burdekin River Catchment, Queensland. CSIRO Land and Water Technical Report 5/02, Canberra. http://www.clw.csiro.au/publications/technical2002/tr5-02.pdf Prove, B.G. and Hicks, W.S. (1991) Soil and nutrient movements from rural lands of north Queensland. In: Yellowlees, D. (Ed.) Land Use Patterns and Nutrient Loading of the Great Barrier Reef Region. James Cook University of North Queensland, Townsville, pp 67-76. Rasiah, V. and Armour, J.D. (2001) Nitrate accumulation under cropping in the Ferrosols of Far North Queensland wet tropics. Australian Journal of Soil Research 39: 329-341. Rasiah, V., Armour, J.D., Yamamoto, S., Mahendrarajah, S. and Heiner, D.H. (Accepted – 2002) Nitrate dynamics in groundwater under sugarcane in the Australian wet tropics. Water, Air and Soil Pollution Journal.

Rayment, G. E. (Draft) Water quality pressures and status in sugar catchments. Indooroopilly, CRC for Sustainable Sugar. Renard, K.G., Foster G. A., Weesies D. K., McCool, D.K., and Yoder, D.C. (1997). Predicting Soil Erosion by Water: A Guide to Conservation Planning with the Revised Universal Soil Loss Equation. Agriculture Handbook 703, United States Department of Agriculture, Washington DC. Reganzani, J.R., Armour, J.D., Prove, B.G., Moody P.W., and McShane, T. (1996). Nitrogen balances for sugarcane plant and first ratoon crops in the wet tropics. In Sugarcane research towards efficient and sustainable production, ed. J.R. Wilson, D.M. Hogarth, J.A. Campbell and A.L. Garside. CSIRO Division of Tropical Crops and Pasture, Brisbane. pp 275-277. Robertson, FA and Thorburn, PJ (2000) Trash management - consequences for soil carbon and nitrogen. PASCT, 22: 225-229. Rogers, L.G., Cannon, M.G. and Barry, E.V., (1999) Land Resources Of The Dalrymple Shire, Department of Natural Resources, Queensland, Brisbane. Rosewell, C.J. (1993), SOILOSS - a program to assist in the selection of management practices to reduce erosion. Technical Handbook no. 11 (2nd edition) Soil Conservation Service of NSW, Sydney.

58

Rosewell, C.J. (1997). Potential source of sediments and nutrients: sheet and rill erosion and phosphorus sources. Australia: State of the Environment Technical Papers Series (Inland Waters), Department of the Environment, Sports, and Territories, Canberra. Roth, C.H., Prosser, I.P, Post, D,A., Gross, J.E. and Webb, M.J. (2003) Reducing sediment export from the Burdekin Catchment. Final Report to MLA, NAP 3.224. Roth, C. and Olley, J. (Eds) (In Prep). From source to sea - understanding and managing flows of water and materials in the Herbert River catchment and their impacts. Townsville, CSIRO. Rulequest Research, (2001). Rulequest Research Data Mining Tools. http://rulequest.com Rutherfurd, I. (2000) Some Human Impacts on Australian Stream Channel Morphology. John Wiley and Sons, England. p11-50. Rutherfurd, I. D. and Budahazy, M. A (1996) Sand Management Strategy for the Glenelg River and its Tributaries, Western Victoria. Cooperative Research Centre for Catchment Hydrology, Melbourne. Report No. 96/9. Schmidt, S. and Lamble, R.E. (2002) Nutrient dynamics in Queensland savannas: Implications for the sustainability of land clearing for pasture production. Rangelands Journal 24(1): 96-111. Shreve, R.L. (1966). Statistical law of stream numbers. Journal of Geology, 74:17-37. Schroeder,BL, Wood, G and Kingston, G. (1998) Re-evaluation of the basis for fertiliser recommendations in the Australian sugar industry. PASCT, 21: 239-247. Verberg, K., Keating, B.A., Probert, M.E., Bristow, K.L. and Huth, N.I. (1998). Nitrate leaching under sugarcane: Interactions between crop yield, soil type and management strategies. In: Michalk, D.L. and Pratley, J.E. (eds) Agronomy – growing greener future. Proceedings of the 9th Australian Agronomy Conference, Wagga Wagga, NSW. pp 717-720. Visser (2003) Sediment budget for cane land on the Lower Herbert River floodplain, North Queensland, Australia. PhD Thesis, Australian National University, Canberra. Wasson, R. J. (1997) Runoff from the land to the rivers to the sea. In The Great Barrier Reef: Science Use and Management, Turia, N. and Dalliston, C. (Eds), Great Barrier Reef Marine Park Authority, p. 23-31. Weier, K.L. (1999) The quality of groundwater beneath Australian sugarcane fields. Australian Sugarcane 3(2): 26-27. White, I., Brodie, J. and Mitchell, C. (2002) Pioneer River catchment event based water quality sampling. Healthy waterways program, Mackay Whitsundays Regional Strategy Group, Mackay. Wilhelm, G. (2001). Water quality report for catchments containing sugar cane in Queensland. Queensland Department of Natural Resources and Mines, Brisbane. Williams, D. M.(2001) Impacts of Terrestrial Run-off on the Great Barrier Reef World Heritage Area. CRC Reef and Australian Institute of Marine Science, Townsville. http://www.reef.crc.org.au/aboutreef/coastal/waterqualityreview.html Williams, D., Roth, C. H., Reichelt, R., Ridd, P., Rayment, G. E., Larcombe, P., Brodie, J., Pearson, R., Wilkinson, C., Talbot, F., Furnas, M., Fabricus, K., McCook, L., Hughes, T., Hough-Gulberg, O., and Done, T. (2001) The current level of scientific understanding on impacts of terrestrial run-off on the Great Barrier Reef World Heritage Area. CRC Reef and Australian Institute of Marine Science. Wilson, P.R. and Baker, D.E. (1990). Soils and Agricultural Land Suitability of the Wet Tropical Coast North Qld - Ingham Area - Queensland Department of Natural Resources Report QV90001 Wong, W. T. (1996) Abstract: Sediment Transport in the Herbert River During Flood Events. In the Proceedings from the conference of the Downstream Effects of Land Use, Hunter, H. M., Eyles, A. G., and Rayment, G. E (Eds). Department of Natural Resources Brisbane. Department of Natural Resources Brisbane, p.195. Woolfe, K.J., Larcombe, P., Orpin, R. and Purdon, R.G. (2000) Spatial variability in fluvial style and likely response to sea-level change, Herbert River, Queensland. Australian Journal of Earth Sciences, 47, 689-694.

59

Yang, C.T. (1973) Incipient Motion and Sediment Transport. Journal of the hydraulics division, 99, 1679-1704. Young, W. J., Rustomji, P., Hughes, A. O., and Wilkins, D. (2001) Regionalisations of flow variabiles used in modelling riverine material transport in the National Land and Water Resources Audit. Canberra, CSIRO Land and Water. http://www.clw.csiro.au/publications/technical2001/tr36-01.pdf Zhang, L. Hickel, K., Dawes, W.R., Chiew, F.H.S., Western, A.W., Briggs, P.R. (2003 submitted). A rational function approach for estimating mean annual evapotranspiration. Water Resources Research.