A report on the Gascoyne River catchment following the 2010/11 flood events
PA Waddell, PWE Thomas and PA Findlater
Resource Management Technical Report 382
Cover picture: Gascoyne River mouth sediment plume - 10.00 am, 22 December 2010. Image processed and enhanced by Landgate, Satellite Remote Sensing Services; Erosion cell. Photo: P Waddell; Flooding through Carnarvon horticultural district. Source: Unknown.
While all reasonable care has been taken in the preparation of the information in this document, the Chief Executive Officer of the Department of Agriculture and Food and its officers and the State of Western Australia accept no responsibility for any errors or omissions it may contain, whether caused by negligence or otherwise, or for any loss, however caused, arising from reliance on, or the use or release of, this information or any part of it.
Copyright © Western Australian Agriculture Authority, 2012
Copies of this document are available in alternative formats upon request.
3 Baron-Hay Court, South Perth WA 6151
Tel: (08) 9368 3333
A report on the Gascoyne River catchment following the 2010/11 flood events 1
PA Waddell, PWE Thomas and PA Findlater 1
Resource Management Technical Report 382
ISSN 1039-7205 1
Executive summary 1
Context and scope 1
Definition of groundcover and catchment condition 2
Key findings and issues 3
Section 1 Introduction 6
1.1 Purpose 6
1.2 Objectives 7
1.3 Background 7
Section 2 Impact of catchment condition on flooding and erosion 11
2.1 Methodology 11
2.2 Results and analysis 16
Section 3 Effect of soil and vegetation condition on landscape stability 40
3.1 Physiographic regions of the Gascoyne River catchment 41
3.2 Landscape organisation and function 49
3.3 Gascoyne River catchment condition summary 60
Section 4 Discussion 86
Section 5 Conclusions 88
Appendix 1 — Acronyms 92
Appendix 2 — Data sets and sources of data 92
Appendix 3 — Vegetation condition assessment and summary (2002-2009) 93
Appendix 4 — Soil infiltration and profile data 96
Appendix 5 — Peak river heights at Nine Mile Bridge 1960 to 2011 97
Appendix 6 — Average soil surface attributes for capture and shedding zones in 2011 101
Appendix 7 — Physiographic regions of the Gascoyne River catchment 102
The authors, PA Waddell, PWE Thomas and PA Findlater from Land and Water Assessment, Department of Agriculture and Food, Western Australia, wish to thank our department colleagues, particularly the soil surveyors and rangeland advisers, who assisted with field work: Dan Carter, Wayne Fletcher, Peter Russell, John Stretch, Peter Tille, Buddy Wheaton and Santino Vitale.
Also special thanks to:
Wayne Fletcher for his efforts to alter his work schedules in an effort to visit all Gascoyne Western Australian Rangeland Monitoring System (WARMS) sites scheduled for 2011 assessment so as to work in with this project's timelines.
Ken Tinley for his advice prior to undertaking field work.
John Stretch for his guidance and intimate local knowledge throughout the field trips.
Brad Cox from the Department of Water, Carnarvon for the history of Nine Mile Bridge river height data.
Further, thanks to colleagues and peers who provided advice on methodology and reviewed this report. Their contributions enhanced the quality of this report and they are: Paul Novelly, Dan Carter, Wayne Fletcher, John Stretch and Peter Tille. Additional thanks to Pam Booker for her efforts in typing and proofing the various draft manuscripts.
All photographs are of the Gascoyne River catchment. Photographs were taken by the authors principally during field work in June and August 2011; where there are exceptions the photographer is acknowledged in the caption. Aerial photography provided by and with the permission of the Western Australian Land Information Authority trading as Landgate.
Finally, the authors appreciate the hospitality extended by pastoralists of the area. Without this logistical support the project would have been all the more challenging.
Context and scope
The Gascoyne River catchment is located in the mid north-west of Western Australia. The catchment covers an area of about 80 400 km2. The native vegetation primarily consists of scattered perennial shrubs of various genera amongst a very scattered acacia overstorey, and supports an extensive pastoral industry. Near the town of Carnarvon on the river levee and floodplain delta there are about 1000 ha of irrigated horticulture.
In December 2010 an extreme tropical storm resulted in widespread flooding at Carnarvon and across the catchment. Another two flood events followed during the summer of 2010-11. At the time of the floods the catchment was considered to have been in poor condition with low vegetative groundcover following an extended period of dry seasons, in combination with a legacy of historic overgrazing and continuous stocking despite seasonal conditions. The flooding resulted in significant soil erosion and damage to infrastructure in the towns of Carnarvon and Gascoyne Junction, as well as the horticulture area. The damage bill was estimated at $90 million.
The rationale for this assessment is to provide illustrative evidence on the role that perennial vegetation groundcover management has in influencing the risk of flooding and soil loss in the catchment. It may be possible that the impact of flooding associated with extreme storm events can be reduced. This report focuses on catchment condition and is not a review of the pastoral industry's economic viability.
In the context of this report groundcover refers to the presence of perennial vegetation. Annual vegetation was not assessed as it is spatially and seasonally variable, and typically does not persist in summer when most flooding is known to occur. Other non-woody cover (cryptogams, dead vegetation and litter) was assessed at the Western Australian Rangeland Monitoring System (WARMS) sites, though this is a relatively small component of groundcover within the Gascoyne River catchment.
Rangeland condition categories (good, fair, poor) are defined in Payne et al. (1987) and Appendix 3.1. Poor condition rangeland in the Gascoyne River catchment typically has low to nil perennial vegetation groundcover and some degree of soil loss, as shown in examples below, and in Section 3.
Poor condition stripped sand sheet
Poor condition duplex surface stripped by sheet flow
This project, which was jointly funded through the Department of Agriculture and Food, Western Australia (DAFWA) and the Australian Government's Caring for Our Country has three objectives:
I. Assessment of the influence of catchment condition or perennial vegetation groundcover on downstream flooding in the Gascoyne River catchment
II. Assessment of the influence catchment condition or perennial vegetation groundcover has on soil erosion in the Gascoyne River catchment
III. Provision of illustrative evidence on the role perennial vegetation groundcover has in reducing the risk of soil loss for different rangeland landscapes.
To context the flooding and erosion events, descriptions of weather events and rainfall records were sourced from the Bureau of Meteorology (BoM). The magnitude and characteristics of the December 2010 flood, along with previous flood events, were compared using hydrographs obtained from the Department of Water (DoW). A qualitative assessment of soil loss was based on plume areal extent at the mouth of the Gascoyne River using MODIS satellite data and estimates of the plume sediment load.
Prior to undertaking field work, DAFWA conducted an analysis of existing information and data sets relevant to the catchment. Landsat and NOAA NDVI satellite imagery was used to assess trends in perennial vegetation cover and pastoral lease inspection traverse information to review past vegetation condition assessments. Between June and August 2011 officers from DAFWA undertook an assessment of the present condition of the mid to upper Gascoyne River catchment, primarily east of Gascoyne Junction. Survey teams collected data at 96 long-term monitoring sites (WARMS) on perennial plant numbers and landscape function. In addition soil infiltration rates were measured at 50 sites, and were used to assess the relative importance of perennial vegetation cover and soil texture on infiltration rates.
Whilst on route to WARMS sites and areas of interest, predetermined using aerial photography, traverse notes and photographs were compiled to formulate an overall perspective of catchment condition and to assess and describe erosional features. Experienced rangeland advisers based the assessment of catchment condition on subjective visual assessments in accordance with condition categories defined in Payne et al. (1987).
Key findings and issues
A record storm and flood resulting In high sediment loss
The tropical storm that crossed over the Gascoyne River catchment between 16 and 19 December 2010 resulted in falls in excess of 250-300 mm over a 24-hour period, the highest on record. The record rainfall brought record flooding with the peak at 7.77 metres at Nine Mile Bridge, near Carnarvon. The previous high was 7.63 metres in 1960, followed by 7.6 metres in 2000 (BoM, DoW).
Soil erosion estimates indicate that the soil loss was substantial; the total mass of suspended solids in the December flood could have been at least 5 625 000 tonnes. The sediment plume for the December 2010 flood was two to seven times larger than other recent flood events. Examples of erosion described in Section 3 of the report, coupled with data from long-term monitoring sites and earlier published reports indicate that accelerated erosion has been occurring in the catchment at least since the 1960s.
The Gascoyne River catchment is in poor condition with reduced groundcover (perennial vegetation)
The Gascoyne River catchment is in poor condition (characterised by loss of cover, few perennial plants and ongoing soil loss) and has been in poor condition at least since the 1960s and possibly the 1930s (Wilcox & McKinnon 1972; Jennings et al. 1979; Williams, Suijdendorp & Wilcox 1980; House et al. 1991; Hopkins, Pringle & Tinley 2006); many areas are continuing to decline.
Over 3.6 million hectares were assessed as being in poor condition for the years 2002 to 2009, with a 15% decline in perennial shrubs in the last five years. This was characterised by a 39% decline in the perennial plant numbers recorded in the above average seasons of 1995 to 2000, reduced resource capture (13% decline overall and 22% decline in mulga groves/run-on sites) and an increase in erosion features. Generally the overall trend in vegetation cover (1989 to 2010) was stable, thus areas that were assessed as poor condition since 1989 are still in poor condition. However, large contiguous areas are declining in cover between the central Gascoyne and Lyons rivers with plant numbers in 2011 declining to 1995 levels.
Poor condition, eroded interpatch
Poor condition fragmented sand bank surrounded by eroded and scalded surfaces
A series of poor seasons were coupled with continuous stocking
Satellite images and rainfall records (BoM) indicate that the seasonal conditions had been poor for four or more years prior to the December 2010 flood. A low greenness index at the time of the flood indicates that the groundcover was also low. Nevertheless, the sequence of poor seasons coupled with the practice of continuous stocking through consecutive dry years (Annual Return of Livestock and Improvement forms, Pastoral Lands Board of Western Australia), in excess of the carrying capacity of the resource (Wilcox & McKinnon 1972; Payne, Curry & Spencer 1987), has contributed to the poor condition of the catchment.
Catchment condition (and perennial groundcover) may impact flooding
Vegetation, groundcover and obstructions are fundamental to reducing sheet flow and erosion (Coles & Moore 1968; Tongway & Ludwig 1996, 1997). However, it is difficult to determine to what degree groundcover and catchment condition contributed to the Gascoyne River 2010-11 summer floods.
Analysis shows that several major floods, with similar characteristics to the December 2010 flood, have been associated with substantial rainfall events since 1960. This suggests the catchment has not changed substantially since the 1960s. As discussed above, the catchment was in poor condition at the time of the December 2010 flood, and it is likely to have been since at least the 1960s. Such is the poor state of the catchment that despite the decline in plant numbers over the last five years it is probable that this decline would have only a minor influence on major flood events.
The catchment is naturally a high water shedding catchment. Infrequent vegetated zones, 'patches', which capture water and nutrients and moderate run-off are sparse and interspersed between sparsely vegetated areas, 'interpatches'. The ratio of interpatch to patch density is estimated at 88:12. The soil infiltration rates in the patches were about fourfold higher than the interpatches. The presence of vegetation had a greater impact on infiltration rates than soil texture alone. Increasing the number of patches through plant abundance and litter obstructions would likely increase infiltration capacity over time, reduce run-off and therefore the likelihood of flooding.
However, irrespective of infiltration rates, the magnitude of the December rainfall event was such that the subsurface and surface storage capacities of the soil would have been exceeded on the interpatches. The soils are generally shallow, frequently less than 30 cm deep, often consist of a sandy loam over clay, hardpan or weathered rock. Assuming a total soil water storage of 60 mm, the December rainfall event exceeded this amount by at least three to five times.
Nevertheless, it is likely that a catchment in better condition (more perennial groundcover) will likely reduce the severity of flooding from minor and moderate storms.
Erosion is associated with loss of perennial groundcover in different landscapes
Hills and ranges, despite their relief, have a lower susceptibility to accelerated erosion due to the protection offered by their abundant stony mantle. However, they do shed a significant volume of water from their surfaces, and thereby still contribute to erosion problems within adjacent landscapes. In comparison, the slopes of mesas and breakaways generally lack a stony armouring and are typically severely degraded. This is due to overgrazing of smaller areas of highly attractive forage within larger less palatable pasture units. This results in these features also contributing to erosion problems in the catchment.
Within the upland areas the drainage flats provide the most valued pastures, occurring as inclusions within less attractive pasture types. Chenopod communities formerly occupied sites of restricted drainage; however, excessive grazing pressure has largely reduced these areas to unpalatable shrubs and seasonally dependent ephemeral species. Along the valley floors and in the drainage foci, where vegetation loss has been considerable, channelisation as rill and gully erosion encourages water shedding. From these drainage areas increased discharge is affecting downstream landforms.
Downslope of the upland areas the landscape is dominated by extensive sheet wash plains. Here, especially during dry periods, it is the vegetation groves and bush clumps that provide sources of browse. Over-utilisation has increased run-off from upper slopes, causing soil instability and disrupting water flow and nutrient cycles. Overgrazing of wanderrie bank communities has reduced the perennial grass component to such an extent that the low strata of many sandy banks now only supports annual grasses such as wind grass (Aristida contorta) and annual wanderrie grass (Eriachne aristidea).
A significant problem within the catchment is the disruption to surface hydrology by infrastructure (e.g. roads, tracks, fence lines). Where vegetation cover is drastically reduced infrastructure initiated erosion problems have a considerable impact on general rangeland condition.
Riparian pasture productivity is highly variable. Initial settlement of the Gascoyne River catchment was along the river, with stock reliant on river pools and natural springs. Consequently, many riparian pastures are overgrazed and degraded. Where buffel grass has become established, it has a significant role in stabilising surfaces and preventing further erosion. In addition, buffel grass colonisation has increased the productivity of some riparian pastures in favourable seasons. However, stock numbers in favourable seasons are often above that which the surrounding native vegetation can support in the absence of buffel grass (Wilcox & McKinnon 1972; Payne, Curry & Spencer 1987). With the onset of dry conditions the protein content of buffel grass declines and livestock seek supplementary forage. Stock migrate upslope and fertile patches become the primary browse source. Without appropriate stocking rates fertile patches are over-utilised, leading not only to their deterioration as a forage source but also their capacity to retain water. This reduces their resource capture role and contributes to escalating erosion downslope.
The reduction in vegetation cover (Section 220.127.116.11) has reduced the landscape's capacity to retain water (Section 2.2.3). Run-off and erosion potential have increased, resulting in erosion cell development. Consequently, the Gascoyne River catchment is locked in the feed-back loop of an erosion cycle. The loss in capacity to retain water drives the desiccation process, reducing vegetation cover. The cycle will continue until new base levels are reached in equilibrium with erosive processes.
Large areas of the catchment are water shedding with very shallow soils with limited storage capacity. However, a major flood would likely have occurred regardless of the catchment condition, perennial vegetation groundcover or infiltration rates as these landscapes would have been overwhelmed by the December 2010 rainfall event.
Erosional features are widespread throughout the Gascoyne River catchment and have increased over the monitoring period but cannot necessarily be attributed to the December flood. It is almost certain that these features have developed as a result of loss of groundcover since European settlement. It is known that vegetative groundcover reduces erosion, and it is clear that erosion would be much less if the catchment was in better condition.
While catchment condition may not have had a significant impact on the December 2010 flooding resulting from a record rainfall event, improving catchment condition (perennial groundcover) is an important aim that will likely reduce the impact of minor and moderate flood events, in particular soil erosion.
If the arid shrublands of the Gascoyne River catchment are to improve significantly in productive value, and be able to sustain ongoing pastoralism, then land surfaces in such an active catchment will need to be restored before landscape systems can efficiently conserve and use rainfall and run-on. This will require long periods with greatly reduced grazing pressure and interventions at critical control points in the landscapes.
Based on the historical and recent review of the Gascoyne River catchment it is likely that future high rainfall events will continue to result in localised flooding, soil loss and damage to infrastructure unless catchment condition is improved.
Gascoyne River catchment assessment