BETWEEN : SOUTHERN PROPERTIES (WA) PTY LTD (ACN 078 766 659)

EXECUTIVE DIRECTOR OF THE DEPARTMENT OF CONSERVATION AND LAND MANAGEMENT

Torts - Negligence - Duty of care - Public authority - Conflicting statutory powers and duties - Prescribed burning

Torts - Negligence - Breach of duty - Civil Liability Act 2002 (WA) s 5B, s 5W, s 5X

Civil Liability Act 2002 (WA), s 5B, s 5U, s 5W, s 5X

Conservation and Land Management Act 1984 (WA), s 3, s 5, s 6, s 7, s 18, s 19, s 32, s 33, s 33A, s 34, s34A, s 34B, s 36, s 38, s 60, s 132

Crown Suits Act 1947 (WA), s 3

Public Sector Management Act 1994 (WA), s 29, s 64

First Plaintiffs : Mr R I Viner QC & Ms N E Robinson

Second Plaintiffs : Mr R I Viner QC & Ms N E Robinson

Adeels Palace Pty Ltd v Moubarak [2009] HCA 48; (2009) 260 ALR 628

Amaca Pty Ltd (Formerly James Hardie & Co Pty Ltd) v Hannell [2007] WASCA 158; (2007) 34 WAR 109

Bell v The State of Western Australia [2004] WASCA 205

Bridgetown-Greenbushes Friends of the Forest Inc v Executive Director of the Department of Conservation & Land Management (1997) 18 WAR 126

Brodie v Singleton Shire Council [2001] HCA 29; (2001) 206 CLR 512

Burnie Port Authority v General Jones Pty Ltd [1994] HCA 13; (1994) 179 CLR 520

Caledonian Collieries Ltd v Speirs [1957] HCA 14; (1957) 97 CLR 202

Cole v South Tweed Heads Rugby League Football Club Ltd [2004] HCA 29; (2004) 217 CLR 469

Crimmins v Stevedoring Industry Finance Committee [1999] HCA 59; (1999) 200 CLR 1

Department of Housing and Works v Smith [No 2] [2010] WASCA 25

Drexel London (a firm) v Gove (Blackman) [2009] WASCA 181

Hunter Area Heath Service v Presland [2005] NSWCA 33; (2005) 63 NSWLR 22

Maloney v Commissioner for Railways (NSW) (1998) 52 ALJR 292

Modbury Triangle Shopping Centre Pty Ltd v Anzil [2000] HCA 61; (2000) 205 CLR 254

Nada v Knight (1990) Aust Tort Rep 81-032

New South Wales v Lepore [2003] HCA 4; (2003) 212 CLR 511

Penrith Rugby League Club Ltd (t/as Cardiff Panthers) v Elliott [2009] NSWCA 247

Puntoriero v Water Administration Ministerial Corporation [1999] HCA 45; (1999) 199 CLR 575

Pyrenees Shire Council v Day [1998] HCA 3; (1998) 192 CLR 330

Roads and Traffic Authority of NSW v Dederer [2007] HCA 42; (2007) 234 CLR 330

Romeo v Conservation Commission of the Northern Territory [1998] HCA 5; (1998) 192 CLR 431

Rosenberg v Percival [2001] HCA 18; (2001) 205 CLR 434

State of Western Australia v Bond Corporation Holding Ltd (1991) 5 WAR 40

Sullivan v Moody [2001] HCA 59; (2001) 207 CLR 562

Tame v The State of New South Wales [2002] HCA 35; (2002) 211 CLR 317

Vairy v Wyong Shire Council [2005] HCA 62; (2005) 223 CLR 422

Wyong Shire Council v Shirt [1980] HCA 12; (1980) 146 CLR 40

1 Each of the first plaintiffs is the registered proprietor of a parcel of land near Pemberton in the south-west region of Western Australia. The parcels of land share common boundaries. The second plaintiffs are lessees of the land owned by the plaintiffs. The land contains vineyards. The parties in these proceedings did not seek to draw any distinction between the first and second plaintiffs, and referred to the vineyards in question as, collectively, the plaintiffs' vineyards, and the land as the plaintiffs' land. The land is used for the cultivation of grapes for use in wine. The vineyards on the land are only used for grape production; there is no wine producing facility there.

2 The plaintiffs claim that the defendants were negligent in 2004 in conducting a prescribed burn on Crown land, which abutted the vineyards. It is alleged that the prescribed burn caused smoke taint to grapes, which rendered the grapes unfit for wine making, thereby resulting in the plaintiffs suffering economic loss. The burn in question was conducted by the Department of Conservation and Land Management (the Department) and was known as Donnelly Pemberton Hand Burn 8 (DPHB8). The ignition was carried out on 31 March 2004 and 1 April 2004.

3 It is common ground that the grapes on the plaintiffs' land were tainted by smoke emitted by the burn. At trial the defendants admitted that the plaintiffs' grapes were damaged by smoke and that the second defendant caused the smoke to enter the plaintiffs' property. Both counsel agreed that causation was not in issue. The parties also agreed the quantum of damages.

4 The issues at trial were whether the defendants owed a duty of care to the plaintiffs, and if a duty existed, whether it had been breached, and whether, in any event, the defendants had the benefit of statutory immunity from liability under s 132 of the Conservation and Land Management Act 1984 (WA) (CALM Act).

5 By way of general comment on the evidence, there were many witnesses called in this case. I was satisfied that they were all telling the truth to the best of their recollection. In the end, most of the facts were largely not in dispute. Where it seemed to me that there were any material differences in the evidence of the witnesses, I have noted the differences, and made findings accordingly.

The plaintiffs' land and vineyards, and the previous occasion of smoke taint in 2001

6 The first plaintiffs first acquired the land in 1998. Along the whole of the eastern border of the land is State forest. Along the whole of the land's southern border and along nearly half of its western border, lies Warren National Park. In general terms, it is bounded to the east, south and west by National Park or State forest.

7 Vines of a number of grape varieties were planted on the land in 1999 and 2000.

8 Grapes are generally harvested from the vineyards between March and May of each year. The precise timing depends upon a number of variables including the weather and the particular variety of grape. In 2004, certain varieties of the plaintiffs' grapes had been harvested by 31 March; other varieties were in harvest or still to be harvested. Harvest was not complete until 6 May 2004.

9 Prior to harvest, the grapes go through a developmental stage known as veraison. Veraison was described by Associate Professor Gibberd, the Chair of Viticulture and Oenology with Curtin University, as 'a key developmental stage for grapes and is associated with a switch from berry cell development to sugar accumulation and cell expansion'.

10 The plaintiffs led evidence by Mr Muller as to the commencement of veraison, the start of harvest and the finish of harvest for the plaintiffs' grapes in the period 2001 - 2007. He said:

Growers have reported damage to wine grapes from smoke and ash. The most critical periods are reported to be between the start of veraison and the finish of the harvest period. The start and end of these periods varies from year to year according to seasonal conditions. Dates were provided for the period 2001 - 2007 were [sic]:

11 In the period 1998 to late 2004, the vineyards were developed and managed by a firm called Albert Haak & Associates; Mr Campbell-Clause was a partner of that firm. Prior to 24 February 2004, Mr Murray Pearse was employed by Albert Haak & Associates as the vineyard manager. Mr Paul Love held that post on an interim basis between 24 and 28 February 2004. On 1 March 2004, Albert Haak & Associates employed Mr Mike Barry as the vineyard manager.

12 The first vintage was in 2001. The grapes harvested in early 2001 were sold to Palandri Wines. In January 2001 there had been a prescribed burn in the State Forest to the east of the plaintiffs' land. The plaintiffs' vineyards were covered with smoke. In August 2001 the Palandri winemaker complained that the plaintiffs' grapes had been tainted. Later that year, the wine produced from the plaintiffs' grapes was thrown away by Palandri because of the taint. As a consequence, it was around this time that Mr Campbell-Clause, the partner of Albert Haak & Associates responsible for the management of the vineyard, first thought that there was a connection between the burn carried out in January 2001, and the quality of the grapes supplied to Palandri.

13 Mr Pearse, the plaintiffs' then vineyard manager, called the Department's office in early 2002 and expressed concerns about smoke and ash entering the plaintiffs' vineyards.

14 By letter dated 18 February 2002, Mr Pearse wrote to Mr Gillard of the Department. He referred to the damage to the plaintiffs' 2001 vintage and said that whilst he wanted a prescribed burn in the area of DPHB8 to be done, he did not want it to be done until the plaintiffs' fruit was harvested. He also said that he believed that further research was required to determine when it was safe to burn during grape growing.

15 In 2002 or 2003, Mr Banfield, representing the plaintiffs, visited the Department's Manjimup office and relayed concerns which he had about smoke from the burn in 2001 affecting the plaintiffs' vintage that year.

16 There were further communications between the plaintiffs and the Department on the topic of smoke taint to which I refer later in these reasons.

The Warren Region and the land subject to prescribed burn DPHB8

17 There are 926,604 hectares (ha) of National Parks, State Forest, Conservation Reserves and unallocated Crown lands managed by the Department in the Warren Region

18 The Warren Region is in the south-west of Western Australia and includes the towns of Manjimup, Pemberton and Northcliffe.

19 There are approximately 60 vineyards around Manjimup and Pemberton. In many cases, these vineyards abut land, managed by the Department, which is or will be the subject of a prescribed burning programme by the Department (Mr Sneeuwjagt ts 911 - 912, exhibit 33B).

20 The Warren National Park, south-west of Pemberton, is a large national park. It has considerable environmental significance for the State, particularly with respect to its veteran karri forest. It is also a public asset of importance for recreational activities and tourism.

21 The land comprising the burn area was a portion of the Warren National Park comprising 560 ha. The burn area was to the north of Old Vasse Road, and was located about 8 km south-west of Pemberton. The perimeter of the area measured about 12 km. To the north there was private property, including the plaintiffs' land, and an area of State Forest. The Warren National Park extended to the south.

22 The area to the east and to the west was surrounded by private property, with a number of dwellings close to the western boundary.

23 The area of the burn contained a substantial amount of karri forest.

24 The Warren Region, as a whole, is regularly referred to as the karri forest of the State. Within the region, there is more karri forest around Pemberton than around Manjimup. Manjimup has a lot more jarrah forest, according to Mr Muller (ts 334).

25 Karri forest is specific to south-west Western Australia according to Mr Packham, a senior research fellow in the School of Geography and Environmental Sciences, Monash University.

26 Mr Packham quoted the following passage from Griffiths T, Forests of Ash: An Environmental History (2001):

The karri and the mountain ash are unique. Elsewhere in the world there are tall, monospecific temperate forests at similar latitudes and environments - such as the redwood forests of northern California, the kauri and podocarp forests of New Zealand, and the aracarian forests of Chile - but these are all gymnosperms, whereas the karri and mountain ash are angiosperms, flowering plants, and the tallest in the world.

27 The karri forest is a dense, wet forest with deep litter which dries out very slowly compared with other forest types such as jarrah/wandoo, which dry out more rapidly.

28 The nature, density and height of the understorey of a forest will have a significant impact on the rate and spread of fire through the forest. The nature and density of the overstorey canopy is also significant. Dr McCaw, Principal Research Scientist, Science Division of the Department of Environment and Conservation, gave evidence on these matters to the following effect.

29 He said that karri forest has both a dense overstorey canopy, and a dense understorey shrub layer. Litter fuels are therefore slow to dry compared with more open forest types.

30 Attempting to burn karri forest too early in the season can result in partial combustion of the surface fuel layer, which is both ineffective for fuel reduction and poses a risk for re-ignition later in the summer under hot dry conditions.

31 There are three common structural types of karri forest understorey:

• karri types 3 and 6 (K3 and K6) - up to 3.5 m tall, of moderate density, and typical of areas at the margins of the karri forest where jarrah, marri and blackbutt may also occur;
• karri types 4 and 5 (K4 and K5) - up to 5.5 m tall, of moderate density, and typical of upland forest in the main belt of karri occurrence;
• karri types 1 and 2 (K1 and K2) - up to 7 m tall, very dense, and often associated with stream-zones and sheltered southern slopes in the main karri belt.

32 It was common ground amongst the experts that karri forest understorey types 1 and 2 create the densest and wettest karri forest environment. This means they dry out the most slowly and have the fewest number of days available for effective burning in any calendar year (discussed further in [326] - [328], [340] - [341], [344] - [346], [362] below).

33 There was some contention at trial as to whether the understorey on the burn area for DPHB8 was predominantly K1 and K2, or the more moderately dense K4 and K5 type.

34 The Department prepares and maintains records, known as 'Fuel Assessment Records', in relation to the type and characteristics of the vegetation of an area of land to be burned as part of the prescribed burning programme. The Fuel Assessment Records record, inter alia, the litter depth, the type of trees, the type of understorey or scrub, and the height and density of the understorey or scrub. The information is collated with respect to 'sample lines', which are lines of 100 m in length in selected parts of the burn area, along which samples are assessed at every 10 m.

35 The Fuel Assessment Records also contain 'recommended wind ratios'. A wind ratio is the ratio of wind speed of the open wind at canopy level, compared with the wind speed at ground level. The more dense the understorey, the lower the wind speed at ground level and, correspondingly, the higher the wind ratio. A wind ratio of 9:1 is very high.

36 The Fuel Assessment Records for DPHB8 record information taken from sample lines at the four corners of the burn area.

37 In three out of the four sample lines, the forest was described as karri, and the understorey types were shown as K1 or K2. In these areas, the understorey was also tall, and was recorded as ranging from 6 m to 10 m. In the fourth sample line, the vegetation was recorded as blackbutt and teatree, the scrub type was shown as K3 and it was recorded as being 3 m high. In that area, and in the north-west corner where the scrub type was K1, the Fuel Assessment Records noted the presence of 'very volatile fuels'.

38 In the three sample lines of K1 or K2 understorey, the recommended wind ratio was 9:1.

39 A map of the area (TB 2976) was in evidence, upon which the sample lines were noted. The map showed that large areas of the burn area were designated 'KMA' which means predominantly karri, with some marri. Another record showed that 70% of the burn area was karri forest.

40 The burn area for DPHB8 also had a significant gully system, described by Mr Muller, the plaintiffs' expert, as running along the middle of the burn block on an east-west axis from the eastern boundary, across to the middle and then extending to the northern section of the burn area and also extending to the southern section on the south-west side of the burn area. Karri 1 and 2 understorey is generally found in such gully systems.

41 Mr Walker, whose evidence on this point I accept, said that the Fuel Assessment Records for DPHB8 show that the burn area contained predominantly K1 and K2 understorey. Mr Sneeuwjagt gave evidence to the same effect, and although he was cross-examined on the sample lines referred to in the Fuel Assessment Records, his evidence that the area comprised predominantly K1 and K2 was not challenged.

42 The plaintiffs' expert, Mr Muller, also said that the burn prescription (reference to which is made later in these reasons) shows that a significant portion of the burn area was K1 and K2.

43 The plaintiffs' other expert, Dr Tolhurst, said that whilst there was some indication in the evidence which raised a doubt in his mind as to the fuel types on DPHB8, he regarded the compelling argument, to some extent, as being the wind reduction factor, which had a ratio of 9:1. He said that this was appropriate for K1 and K2 understorey.

44 One witness, on the other hand, Mr Simmonds (of the Department), said that the block contained predominantly K4 and K5 understorey with K1 and K2 located in the gully systems. In relation to that evidence, he said that he knew that that was the vegetation on the block from his experience, but he had not carried out an assessment of the area. He also said that prior to the burn, he and Mr Keppel (of the Department) had driven around the boundary of the burn area. In cross-examination, he also accepted a proposition that another document, prepared by Mr Marshall of the Department (who passed away before these proceedings commenced), contained information consistent with K4 and K5 being the predominant understorey.

45 I do not accept Mr Simmonds' evidence on the question of the predominant type of understorey vegetation on the DPHB8 block. I think he is probably mistaken. His evidence is inconsistent with the Department's records (Fuel Assessment Records) prepared for the purpose of recording such information. It was, moreover, given without reference to such records, and it is inconsistent with the evidence of Mr Walker and Mr Sneeuwjagt, and indeed the evidence of the plaintiffs' witnesses, to which I have referred above. I place no weight on his concession with respect to Mr Marshall's document, as it was prefaced with the statement that he could not say what Mr Marshall had intended by it. It is also inherently unlikely to be a correct characterisation of the document because Mr Marshall prepared the Fuel Assessment records and the burn prescription, which indicate that K1 and K2 are the predominant understorey.

46 Mr Simmonds' evidence is also inconsistent with the evidence of Mr Keppel, who was Mr Simmonds' superior officer, and who accompanied Mr Simmonds on his inspection of the boundary. Mr Keppel was a careful and thoughtful witness. He said in evidence-in-chief that from the inspection undertaken, he had formed the view that the area involved primarily K1 and K2 understorey, with parts of K4 and K5, and some lighter forest, including blackbutt. His evidence on this was not challenged in cross-examination. I prefer his evidence to Mr Simmonds' on this.

47 Accordingly, I find that the burn area for DPHB8 contained predominantly K1 and K2 understorey.

Fire behaviour and fuel management - overview

48 Mr Cheney's evidence, which I accept in these matters, was to the following effect.

49 Fire behaviour encompasses the way a fire ignites, builds up or grows; its rate of spread; the characteristics of the flame front; and all other phenomena associated with the moving fire front. Most of the research into fire behaviour has focused on the prediction of fire rate of spread, the dimension of the flames and the likelihood of firebrands being thrown ahead of the fire to start new fires.

50 It has not been possible to apply traditional physical theory to the prediction of fire spread, because fire is a chaotic chemical reaction that moves through a fuel bed that varies spatially in three dimensions, across a variable topography and interacts with a turbulent atmosphere (principally the wind) that varies widely both in space and time. Prediction of fire behaviour at any specific point in space or time can have a wide margin of error because it is difficult to predict the fuel that will be consumed, and the direction and the speed of the wind at that point.

51 However, as fire builds up from ignition it reaches a quasi-steady rate of spread that can be correlated to some mean value or characteristics of the fuel and the mean wind speed measured at a standard location. The error in prediction is reduced by selecting a suitable period of time, often 30 minutes or more, to encompass the inherent variation in wind and fuel.

52 The key factors that influence how fire spreads are the fuel characteristics, the moisture content of the dead fuel, the slope of the ground, and the wind speed and its orientation to the fire. Mostly, predictions are made for the fastest spreading part of the fire called the head fire.

53 The south-west forests of Western Australia are fire prone. The climate, characterised by cool wet winters and hot dry summers, provides the optimum conditions for the build up of litter and debris on the forest floor. Good conditions for forest growth provide ample quantities of leaf bark and twig material to accrete each year. Low temperatures, when the litter is wet in winter and spring, and dry conditions when the litter is warm in summer and autumn, mean that decay rates are slow. As a result, the accumulation of litter on the forest floor can continue for more than 20 years before accretion and decay come to equilibrium.

54 This litter provides the fuel for forest fires. Ignitions from lightning and anthropogenic sources have been present for a very long time and fire has been an integral part of the ecology of the south-west forests. However, the reduction of anthropogenic ignition and the suppression of fires when conditions are mild have resulted in a general increase in fuel load throughout the forest. When heavy fuel is combined with strong, hot, dry wind, uncontrollable high-intensity and damaging forest fires can result.

55 Fuel is the only factor affecting forest fire behaviour that humans can influence to reduce the speed and intensity of fire and make suppression easier. Fire is an integral part of the environment but it needs to be managed to reduce the impact of damaging fires, both within the forest and within the wider community.

56 In all eucalypt forests the spread of fire depends on the debris and litter on the forest floor. Even under the worst conditions a fire cannot spread independently through the crowns of the trees. A crown fire in eucalypt forests is always supported by an intense fire-burning surface fuel. In essence, if there is no surface fuel, there is no fire.

57 The aim of fuel management is to alter the structure of the fuel bed and the load of the available fuel to make fire fighting safer and easier. The cheapest and most ecologically sound way to do this is by prescribed burning.

58 The rate of fire spread is directly proportional to the height and amount of the loosely compacted material on the forest floor. Reducing the surface fuel reduces the rate of spread. This reduction diminishes as fuel accumulates, but there will be some reduction in spread for up to 15 years after a prescribed burn.

59 Reducing the fuel load reduces both the rate of spread and the fire intensity.

60 Fire suppression is itself a hazardous and dangerous activity to those called upon to undertake it on behalf of the community. Hazard reduction by planned burning assists fire suppression in a number of ways. These include:

• Reducing the speed of growth of the fire from its ignition point. All fires take time to build up to their potential rate of spread which is reached when the width of the head fire is in excess of 300 m. Fire in light fuel is frequently checked by bare areas and this tends to keep the fire narrow and spreading below its potential rate.
• Reducing the height of flames and the heat radiated from the flames. The amount of heat radiated depends on the total volume of flame. Flame height is determined by rate of spread and the height of understorey shrubs and the volume is affected by the total fuel load. Fire in light fuels will have smaller and thinner flames and will have a lower effective temperature. This means that less heat is radiated from the flame and fire-fighters can work closer to the fire edge and for longer periods without suffering heat stress.
• Increasing the visual range in the forest. Reducing the height of shrubs allows fire-fighters greater visibility through the forest. This allows them to observe changes in fire behaviour earlier and makes fire-fighting safer.
• Reducing the potential for spot fires. Burning is the only practical way of reducing the flammability of fibrous-barked trees.

61 For prescribed burning to be really effective in reducing the impact of large fires under extreme weather conditions, a mosaic of recently burnt areas has to be established and maintained over large areas.

Sources and conditions for fire, especially fire in south-west Western Australia

62 The following findings are derived principally from the Department's publication entitled 'Fire management on CALM lands in the south-west of Western Australia', October 2000.

63 There are two primary sources of ignition for bushfires - lightning and humans. Both sources have been features of the West Australian landscape for thousands of years. In the period 1989 - 1990 to 1998 - 1999, lightning accounted for between 9% - 13% of wildfires, although in 1991 - 1992 its incidence reached 26%.

64 Lightning is usually associated with thunderstorms, and these occur in every month of the year in the south-west. Most frequently they occur in winter, but they occur with most devastating effect in the summer. As soon as fuels on the forest floor begin to dry out as spring progresses, the chance of bushfires starting from dry lightning strikes rises.

65 Although the frequency of lightning-caused fires varies from year to year, on average, about 35 bushfires are caused by lightning in south Western Australian forests every summer.

66 Humans are responsible for in the order of 90% of all fires in Western Australia. Some fires are accidental (escapes from prescribed burning, camp fires etc) but most are deliberately lit. In recent years, arson has been the most rapidly increasing source of fire in Western Australia, as it is world-wide. Deliberate fire lighting frequently occurs under severe fire weather conditions, and arsonists often select the most difficult areas in which to operate.

67 The amount and rate of heat release, or the intensity of a fire, is an indicator of its damage potential and suppression difficulty. Fuel, consisting of ground litter, twigs, bark, sticks and live and dead vegetation, is the energy source of a bushfire. The amount, or tonnage, of fuel consumed and the rate of combustion directly affect the intensity of the fire. The greater the tonnage of fuel and the higher the rate at which it burns, the more intense the fire will be.

68 Bushland in south-west Western Australia has three important characteristics in relation to its capacity to fuel a fire:

• The vegetation itself is highly flammable. Leaves of most common species of trees and shrubs contain waxes and volatile oils which ignite easily and burn fiercely. As shrubs age, a greater proportion of the foliage dies and comes to consist of flammable, aerated fuel.
• Fine, flammable dry matter, such as leaves, twigs and bark, (collectively known as 'litter') is constantly shed from living plants and accumulates over time. In the absence of fire, forest floor litter continues to accumulate.
• Many tree species (eg, jarrah, marri, tingle, blackbutt) have stringy or fibrous bark which can be carried aloft as burning brands and embers in high winds, setting spot fires ahead of the main fire. 'Spotting' and 'hopovers' are very common features of West Australian forest fires, and are nearly always associated with heavy, dry fuels. They can extend substantial distances from the main fire edge, rendering useless narrow firebreaks, buffers or control lines.

69 The vegetation and litter in a forest becomes fuel when ignited. This fuel can burn in three different ways. One is a mild creeping fire, where the flames are confined to the dry leaves and twigs lying on the forest floor and to ephemeral grasses. The second is a more intense fire with flames up to 8 m in height, which will consume the ground litter and the shrub layer and mildly scorch the tree canopy. The third is a high intensity fire which 'crowns' - ie, the fire leaps into the upper branches of the tall trees, and flames can be up to 100 m in height.

70 A crown fire consumes all green and dry leaves in the tree and shrub layers and the litter, leaving only the large woody trunks standing. Under the same weather conditions, the higher the levels of fuel, the more intense the fire and the greater the likelihood of a fire moving from a creeping ground fire to a crown fire.

71 Crown fires in tall forest country cannot be directly dealt with using current technology.

72 There are also 'hop-over' or spotting fires, in which burning material from, eg, a prescribed burn, is carried into another area and acts as a firebrand, starting a new fire outside the main area. Even under mild prescribed burning conditions, the embers can travel a few hundred metres and start a hop-over fire. Under extreme conditions, they can travel 20 - 30 km.

73 The south-west fire season is hot, dry and windy. Some summers are hotter and drier than others, but in each summer hot, dry conditions occur. Similarly, the onset of the winter rains in about May, which occur every year, signal the end of the fire season for several months.

74 Unlike the south-east of Australia (Victoria and Tasmania) which experience occasional drought, leading to severe summer bushfire problems, the south-west of Western Australia experiences, in effect, a form of drought every summer.

75 There are a number of consistent features of the weather which operate during the south-western fire season, including:

• The regular passage of a series of high pressure systems from west to east across the southern half of the continent. These generate several-day cycles of increasingly hotter weather, with dry continental easterly winds backing into the north and north-west; after a few days these conditions are terminated by a cold front bringing a cool change, and south-westerly and south-easterly winds, and the cycle recommences. This feature is further discussed in the section dealing with the west coast trough (below).
• Thunderstorms occur regularly in the south-west region during the summer months. These frequently generate lightning which starts fires in bush lands and forests. Apart from occasional rain from thunderstorms or light showers associated with a cool change, the summers in the forest areas are generally without rain Lightning has been the source of some of the most destructive fires which have occurred in Western Australian forests, including the worst fires in 1961.
• The vegetation, fallen timber, and the upper soil layers of the bush all progressively dry out as spring and summer progress, culminating at a point in autumn before the first rains. This drought (or cumulative fuel dryness) factor significantly affects fire behaviour - the drier the fuels the more easily they ignite and the more intensely they burn. This factor also affects the difficulty of fire suppression, and overlays the day-to-day factors relevant to fire suppression such as temperature, wind speed and relative humidity.

76 A number of witnesses gave evidence about the effect of a low pressure trough on the western coast of Australia. A report by Mr Reader of the Bureau of Meteorology was also tendered by the defendants by consent (Mr Reader was not required to attend for cross-examination). In his report, Mr Reader described a typical summer cycle of weather involving a low pressure trough, which was referred to by the other witnesses as the 'west coast trough', as follows. A low pressure trough forms over western parts of the State. At the start of the cycle the low pressure trough is pushed eastwards by a new high pressure ridge which forms off the coast. After the high pressure ridge has moved inland and displaced the trough to the east, a new low pressure trough forms or deepens along the coast of Western Australia over the next couple of days. Then, the trough moves eastward again ahead of a new high pressure ridge to complete the sequence. The cycle can take between three to seven days for the sequence to be repeated.

77 A number of witnesses described the winds associated with the west coast trough and their impact on prescribed burning programmes. I found Dr McCaw's evidence lucid and clear. He said that as the high pressure system becomes established in the Great Australian Bight, easterly winds direct dry continental air across the south-west of the State, resulting in progressively higher temperatures over the course of several days. With the formation of a trough down the west coast, the winds shift to the north-east and air temperatures continue to increase. Once winds reach the north-east quadrant, then a change to north-west winds can typically be expected within the next 24 hours.

78 Winds blowing from the north to north-west during summer and autumn are typically hot, gusty, dry and unstable. The transition from north-east to north-west winds often takes place abruptly and in the space of a few hours.

79 There is a well-recognised association between northerly winds and dangerous fire conditions. The strong connection between northerly winds and elevated levels of fire danger has been recognised since at least 1938. Recent examples of major bushfires associated with northerly winds include the Dwellingup fire of 3 February 2007, the Bridgetown and Tenterden fires on 27 December 2003, and the Mount Barker fire of 28 December 2000.

80 Northerly winds can result in extreme fire danger well into the autumn. For example, on 23 March 2005, very hot, dry conditions caused a fire north of Manjimup to burn with unusual severity. A notable feature of that day was the very low relative humidity ( ---I think once the trough starts to move then the forecasts are reasonably reliable but picking the timing of when the trough will begin its movement is quite difficult and uncertain ...

The north-easterly winds are usually associated with a deepening of the trough and that once the wind moves into the north east it rarely stays there for more than a day and sometimes it might be for a much shorter period. It might move from the north-east around to the north and the north-west within a single day.

And you're saying that is difficult to predict?---Yes.

84 To similar effect, was Mr Muller's evidence in cross-examination.

---Northerly winds are less predictable and they have been associated with some fairly severe unforecast winds.

KING, MR: Yes?---And conditions are generally less stable.

...

And the movement of that trough can be very quick or it can be slow?---Yes.

It's not possible really to predict with any certainty how fast that trough is likely to go at any particular time it's there?---That's correct.

That is one of the reasons why northerly winds can be unpredictable?---The trough can move in and out and you can have strong northerly winds persisting for a number of days.

You're going from what to northerly?---From east to north. It usually goes through the north-east to the [north]-west. When that trough develops down the west coast that's a very, very unstable weather pattern. Those charts that you see, I think the met bureau report was accepted yesterday, they are pressure charts. What it's showing is low pressure. Usually there is no inversion layer associated with that trough, okay. While the south-easterly, while the easterly, the high pressure system is there you've got an inversion layer that sits somewhere between three and 5000 feet which essentially keeps a lid on the atmospherics and it's a very stable condition for prescribed burning. Once we get into that west coast trough environment it's very unstable.

Yes?---And you can get some interesting fire behaviour, severe fire behaviour just created by the atmospherics without even getting very strong wind under those conditions.

86 Mr Sneeuwjagt, whose evidence I accept on these matters, said that undertaking a burn usually requires several days for a simple burn and up to several weeks where a large burn contains several different fuel types that dry out at different rates.

87 The DPHB8 prescribed burn involved two days to light up. Several days thereafter are generally allowed for such burns in mopping up and patrolling the boundaries.

88 The burn of the edges of the area to be burned is done by hand. The burning of the remaining bulk of the area is called 'core ignition', which may be done by hand, if the area is small enough, or, in larger areas, may be done by helicopter or plane.

89 The edging burn and the core ignition for DPHB8 had to be done on separate days for the burn to be done safely. The DPHB8 core ignition was done by helicopter.

90 It is important for the edging burn to be deep enough so that when the core ignition is lit, the fire remains in the middle of the burn and does not run out to the boundaries. In the case of DPHB8, an edging depth of at least 100 m was prescribed by the burn prescription.

91 The edging burn and the core ignition create different levels of smoke. Mr Sneeuwjagt said, and I accept:

An edging ... is a burn that is done against the wind. It's really a backfire that is moving backwards at about 10, 15 metres per hour very slowly. To reach 100 metres it takes six to eight hours, so you are actually consuming a small amount of fuel per minute per hour. Therefore, the smoke that comes from edging in that situation is very light as opposed to when you do the core lighting, and I am pointing to the middle of the block where you are obviously burning the bulk of the forest. There are multiple - something like 78, I think, separate lightings of fire spots where you are consuming fuel with head fire and back-fire so it is consuming a lot of fuel at any one time, and the smoke from that is many hundreds, if not a thousand times more than a very light back-burn from the edging (ts 918).

92 Both Mr Sneeuwjagt (ts 945 - 946) and Mr Packham (ts 989), whose evidence I accept in this regard, said that, in relation to the proposed burn of DPHB8, smoke from the edging burn would be expected to lift into the air and blow across the top of the vineyard and be dispersed. Areas continuing to burn overnight would be likely to produce smoke staying closer to ground level by a weather process known as 'inversion'.

93 Mr Sneeuwjagt, whose evidence I accept, described the purpose and objectives of prescribed burns as follows.

94 Wildfires cannot be accepted as an uncontrolled variable in forest management as they represent a real risk to life, property, townships, plantations, young regrowth, tourism, recreation and nature conservation.

95 The most effective and strategic way to reduce the risk of severe wildfire damage is by broadscale fuel reduction by prescribed burning during mild weather conditions. Prescribed burning does not stop fire occurring, nor does it avoid the need to maintain vigilant, organised, well-equipped fire suppression resources. But with less fuel on the forest floor, fires can be more easily suppressed, and the impact of wildfire can be significantly reduced.

96 Fires are an integral and essential part of biological, soil and atmospheric processes that sustain ecosystems. For example, fire has a significant role to play in nutrient cycling and redistribution, plant succession processes, animal habitat maintenance, competition and other inter-actions within ecosystems. Prescribed burning is therefore an essential tool in the maintenance of the ecosystem health and vigour of fire-prone forests.

97 Fire is also used to regenerate forests following timber harvesting through the application of moderately high intensity burns. Habitats for certain animals requiring fire-regenerated thickets are renewed by fires at appropriate intensities and periods between fires.

98 Prescribed burning is the principal land management tool used for the management of accumulated fuels in parks, forest and reserves of south-west Western Australia.

99 The 'fire' experts for the parties also agreed, following an experts' conferral prior to trial, to the following propositions which were recorded and tendered in evidence, which I accept:

(a) Prescribed burning is a necessary part of public land management to meet biodiversity, economic, silviculture, water and community protection objectives.

(b) Prescribed burning is the only effective way of reducing the potential impact of wildfires at a landscape level.

(c) Prescribed burning is complex and requires a high level of skill and planning.

(d) Prescribed burning operations are performed in a way to minimise environmental and economic damage.

(e) The Department has a detailed and thorough prescribed burn planning process.

(f) Prescribed burning prescriptions can be applied with some flexibility and discretion provided that approval to vary the prescription is received at the appropriate level. Professional judgment should be used to assess relevant risk mitigation or other operational factors and the prescribed burn should still meet the burn and operational management objectives.

History and policy of prescribed burning in the south-west of Western Australia

100 The following history of prescribed burning is principally taken from Mr Walker's evidence and the publications to which he referred, and the Department's publication in 2000 entitled 'Fire management of [the Department's] lands in the south-west of Western Australia'. Mr Walker was, at the relevant time, the Director for Regional Services of the Department.

101 In general terms, prescribed burning in Western Australia has been conducted for more than 80 years by the Department and its statutory predecessors.

102 There are two designated fire seasons for prescribed burning in the south-west. One is denoted 'spring', although it covers not only the spring months, but also the summer months of December and January and, sometimes, into February. The other is called 'autumn', which is typically March and April, but may extend into May. Generally, the season of the burn refers to the season in which the core ignition is to take place.

103 Although bushfire ordinances were passed in the mid-1800s, there was little attempt to address the problem of fire in the south-west of Western Australia until after the establishment of the Forests Department in 1919.

104 In the early development of bushland fire policy, conflicting views emerged over the use of fire. Many of the European-trained professional foresters who were appointed to the new Forests Department were opposed to prescribed burning. Field foresters, with long experience in the Australian bush, were in favour of it.

105 A policy of restricting the use of broadscale burning was introduced. This policy, together with improved fire suppression, resulted in heavy fuels accumulating in most forest areas by the 1940s.

106 From the late 1930s onwards, fires had started to become very large and difficult to control. There were major fires in the jarrah forest in 1949 - 1950 and in the jarrah and the karri forests in 1937 and in 1950 - 1951. In the long protected compartments, fires became uncontrollable once they exceeded about a hectare in size, even under quite mild weather conditions.

107 Also at about this time there were large fires in the southern forest national parks, notably the Walpole-Nornalup Park and adjoining areas, where whole hillsides of karri and tingle trees were killed.

108 In 1953, there was a change in Forests Department policy and broadscale prescribed burning for fuel reduction was introduced. Because of the large build-up of fuels in most of the areas to be burnt, implementation of the policy was cautious and slow at first. Most of the initial burning in the northern jarrah forest was done in winter. There were also technical deficiencies, especially lack of fire-behaviour information on which to base burning prescriptions, and a lack of trained staff to undertake the work. Little effective burning could be undertaken in the dense southern forests, principally because of lack of access and problems with predicting fire behaviour in complex karri and karri-tingle fuels.

109 In 1961 large, destructive wildfires occurred in the Dwellingup region, destroying 70 houses and buildings in the town of Dwellingup, and completely destroying the towns of Holyoake, Nanga and Banksiadale.

110 The fires, along with fires in other areas, were the subject of a Royal Commission. One of the terms of reference required an inquiry of the measures necessary or desirable to prevent the outbreak and spread of bushfires in Western Australia and to protect life and private and public property from the consequences of such fires.

111 The report referred to karri forest in these terms:

The karri forest occurs in four or five patches around the coast between Albany and Busselton ... It is a very tall forest with dense undergrowth and is not easy to burn, except in the hottest time of the year when a very damaging fire can develop.

112 One recommendation in the report was that 'the Forests Department make every endeavour to improve and extend the practice of controlled burning to ensure that the forests receive the maximum protection practicable consistent with silvicultural requirements'.

113 This did not represent a redirection of policy for south-west forests. Rather it endorsed the basic policy which had been adopted in 1953, but recommended its improvement and expansion. The Royal Commission's recommendations were adopted in full by the Government of the day, and have continued in operation over the intervening years.

114 In response to this recommendation, the Forests Department commenced a detailed programme of broadscale prescribed burning, including ignition by aircraft, in the early 1960s.

115 Mr Sneeuwjagt's evidence, which I accept, was that there have been few major forest fires in the jarrah forest since 1961 and in karri forest since 1970 after the fuel-reduction burn policy became effective. During this time, no lives of civilians or fire-fighters have been lost from forest fires. There have been comparatively few houses destroyed or damaged from forest fires since the expansion of the fuel-reduction burn policy after the 1961 Dwellingup fires.

116 Prescribed burning has been considered in a number of subsequent government reviews and inquiries into prescribed burning.

117 Commencing in about 1970 and continuing over the next 30 years, forest research scientists in Western Australia investigated and developed a deep understanding of fuel accumulation rates and the effects on fire intensity and rate of spread of different temperatures, wind speeds, relative humidity, fuel dryness and slope for the jarrah and karri forests. This information was incorporated into a fire behaviour prediction system and a prescribed burning guide, used by field staff in planning and implementing prescribed burns. The publication, by Sneeuwjagt and Peet, 1985, is known as the 'Red Book'.

118 The Pemberton National Parks (including the Warren National Park) have had a management plan for fire and environmental protection since 1974.

119 In 1987 the Department prepared an update to the plan entitled 'Interim Guidelines for Necessary Operations'. Clause 4.4 provided:

Due to the fact that there is high recreational use in the Parks areas, the need to maintain the protective burning plan for protection of Park users is essential, as well as to limit the potential for fire spread from recreation sites.

Some areas of the Park need to be fuel reduced to contribute to the system of protective buffers for large areas of karri regeneration in the adjoining State Forest.

120 The guidelines identified fire-protection actions which were required, including burning of strategic buffer zones to reduce the risk of spread of fire in accordance with a 'Master Burn Plan', and to maintain fuel loads below 18 tonnes per hectare in karri stands. Depending on the rate of fuel accumulation, the estimated rotation period was 6 - 7 years. The guidelines noted that 'due to the late drying of karri fuels, burning may need to be undertaken in the prohibited burning season' for the parks. (The Bushfires Act 1954 (WA) prohibited burns being undertaken at certain times in the early part of a calendar year, subject to exceptions being applied for and obtained.)

121 In 1988 - 1989 the Department conducted a prescribed burn on the DPHB8 area. The area had been the subject of earlier burns going back to 1953.

122 In 1993, the Department's policy for prescribed burning was the subject of public review, pursuant to a Ministerial condition which was set when the then Minister approved the Department's 1992 Forest Strategy. The review was conducted by a Fire Review Panel which considered written evidence from over 220 submissions and oral evidence from 125 separate individuals or groups. It reported to the then Minister for Conversation and Land Management on 15 March 1994. The reported endorsed the Department's prescribed burning policy and, further, recommended, inter alia:

• That Fire Management become the fourth programme within the Department organisation, and that this programme receive an additional $3 million per annum and steps be taken to implement this over the next three years.
• That the Government publicly declare that regular prescribed fire is essential to minimise extensive wildfires and publicly recognise that the Department is the appropriate authority to carry out prescribed burning for fuel reduction and habitat management within south-west forests and heathlands.
• That the Government accept that prescribed burning to modify fuel loads is the most cost-effective way to modify fire behaviour and is essential to minimise the potential damage from wildfires.
• That the Department's burning plans should be better integrated with long-term logging plans, wildlife corridor planning, and plans for landscape reserves.
• That the Department investigate greater use of helicopters for prescribed burning.
• That the Department's Wildfire Threat Analysis remains an essential fire management tool and that the ranking of values is appropriate for the Department's areas; that is, human life continues to receive the highest priority.
• That the necessary funding be provided to the prescribed burning programme to increase the diversity of season, intensity, frequency and sequence of fires on any single area to ensure that no species will become threatened or extinct.

123 In 1994, the Department determined that an annual prescribed burning programme of around 200,000 ha each year in the south-west forest regions would be desirable to provide a reasonable level of fuel reduction and hence protection from wildfires across the lands managed by the Department in the south-west.

124 There was bipartisan political support in Western Australia for prescribed burning as the primary fire protection strategy for private and public assets and for rural communities in the south-west.

125 The Department was unable to achieve its 200,000 ha target in the years between 1997 - 1998 and 2002 - 2003 due to a combination of factors. The factors included dry seasons that were unfavourable for safe and effective burning, and other constraints, being:

• the risk of smoke generated from these burns being transported to the Perth metropolitan area;
• lack of available resources due to the occurrence of wildfires in the jarrah forest - wildfires obtain priority for resources, and there is some caution about over-commitment to prescribed burning if wildfires elsewhere are uncontained; and
• planning and implementation difficulties associated with the complex mosaic of land uses and boundary sensitivities in the karri region - timber harvesting, vulnerable regrowth forests, and neighbouring land uses.

126 In 1997 - 1998, the Department achieved prescribed burning of 167,151 ha; in 1998 - 1999 98,117 ha; in 1999 - 2000 174,455 ha; in 2000 - 2001 87,866 ha; in 2001 - 2002 74,739 ha; and in 2002 - 2003 144,835 ha.

127 During these six years, a combined backlog of 452,837 ha of prescribed burns was accumulated. This added significantly to fuel loads across the south-west.

128 In 2001, Mr Walker oversaw an internal review of fire operations which was undertaken by Mr Muller, who was then a project officer with the Department. The review led to a report being published in November 2001.

129 The review made 43 recommendations, including that the size of individual burn units be maximised where possible; that an adaptive approach to burn planning be taken; and that the prescription objectives for prescribed burns identify the reason and the intended outcomes for the burn.

An analysis of the present situation has identified that insufficient burning is being undertaken to maintain an optimal fire distribution to meet either biodiversity or protection objectives. There is currently potential for large areas to be burnt in wildfires that would be uncontrollable during severe weather conditions. Whilst limited high intensity fires are a desirable component of the overall fire distribution, there is currently a significant risk of loss of biodiversity through extensive high intensity wildfires, particularly in the Southern Forest Region. Such wildfires would also pose a threat to community values.

131 In 2002, there was a community meeting at Manjimup Town Hall at which there were expressions of concern about the build-up of fuel loads, and calls for further prescribed burning to be undertaken in the area.

132 2002 - 2003 was one of the worst seasons for bushfires in Western Australia since 1960 - 1961. It was the subject of a statement by the Minister for Environment to the Legislative Assembly on 8 May 2003:

DR J M EDWARDS (Maylands - Minister for the Environment and Heritage) [9.06 am]: Mr Speaker, I draw the attention of the House to the impact the recent bushfire season has had on the State's conservation lands, particularly in the south west. The Department of Conservation and Land Management advises that the season was the worst in more than 40 years in the number of wildfires, their intensity and the area burnt. Since the season began, the department has faced approximately 600 wildfires that have burnt through 754 000 hectares between Geraldton and Esperance. There were three fire events of note. One fire in January at Mt Cooke, about 70 kilometres south east of Armadale, burnt through 18 000 hectares. This was the biggest single fire event in the northern jarrah forest since the disastrous Dwellingup fires in 1961 ... In the proposed Walpole Wilderness Area, a fire that began in February flared in March and ultimately burnt through 35 000 hectares. This was the single biggest forest fire event in the south west since the Boorara fires of 1967. In late March, another group of fires burnt through 25 000 hectares in the D'Entrecasteaux National Park. The cost of suppressing fires on the State's conservation lands this season is in order of $11.5 million. Of the 600 wildfires, approximately 200 were a result of [lightning]. In fact, between 21 and 22 March, the department responded to 30 lightning-caused fires between Yanchep and Walpole. These included five strikes in the D'Entrecasteaux National Park.

... The department has achieved 105 000 hectares of its annual planned burning program of 200 000 hectares. The first part of the program in spring was carried out successfully under difficult and dry conditions. However, the autumn program will also be very problematic because of the pressure to burn before winter rains. The Government supports the use of planned burning to achieve biodiversity and protection outcomes, and has provided additional funds to ensure that the department has sufficient resources to achieve the planned burning program.

133 Mr Gillard, the Donnelly District Fire Manager, said that in the 2002-2003 fire season, in the Donnelly District alone, there were 51 wildfires that burnt more than 5,700 ha. Twenty-four of those fires were caused by lightning. The diversion of resources into fire management and suppression meant there was limited opportunity for prescribed burning in the area in the 2002 - 2003 season.

Fires and inquiries in the Eastern States - 2003

134 In 2002 - 2003, major bushfires also occurred in New South Wales, Victoria and the Australian Capital Territory. About 3 million hectares were burnt, eight people died, 1,180 houses and buildings were destroyed and about 20,000 livestock were killed.

135 During 2003, Mr Walker represented the Department in two major national inquiries relating to those fires.

136 One of the inquiries was conducted by a Select Committee of the House of Representatives of the Australian Parliament. The House of Representatives Select Committee's report was entitled 'A Nation Charred', and was published on 23 October 2003. The report included the following:

3.136 Based on the evidence, the Committee has concluded that the implementation of regimes of prescribed burning is the most environmentally and economically effective method of fuel reduction. To be effective the planning and implementation of prescribed fire regimes require the highest possible level of detail concerning the location and extent of past prescribed and unprescribed burns. The Committee is aware of the possibility of counter-productive consequences flowing from the implementation of an ill conceived regime in which burns are either too frequent, thus increasing the flammability of the environment and degrading local biodiversity, or too infrequent, thus being ineffectual.

3.137 The Committee recommends that the Commonwealth through the National Heritage Trust, offer assistance to the states and the Australian Capital Territory to develop specific prescribed burning guides, at least to the quality of Western Australia, for national parks and state forests through out [sic] the mainland of south eastern Australia.

3.138 The Committee is of the view that the implementation of prescribed burning has fallen significantly behind the levels that are possible and required for the maximum possible protection of life, property and the environment in all areas affected by recent bushfires. It notes that although Tasmania and Western Australia have sustained significant damage through bushfires over recent years, neither state has been subject to a repetition of the catastrophes of their worst fire years, 1967 and 1961 respectively, in more recent experience. This situation stands in stark contrast to the areas burnt out in recent fires that effected the Australian Capital Territory, New South Wales and Victoria.

Recommendation 13

3.139 The Committee recommends that the Commonwealth seek to ensure that the Council of Australian Governments seek agreement from the states and territories on the optimisation and implementation of prescribed burning targets and programs to a degree that is recognised as adequate for the protection of life, property and the environment. The prescribed burning programs should include strategic evaluation of fuel management at the regional level and the results of annual fuel management in each state should be publicly reported and audited. (emphasis added)

137 The other inquiry was conducted by the Council of Australian Governments. The report from that inquiry was published on 31 March 2004. It noted that a reduction in fuel loads in bushland environments will reduce fire intensity and modify fire behaviour, and that prescribed burning is the only cost-effective way to achieve fuel reduction in large areas of landscape. It recommended strategic burning which accommodated operational constraints and established priorities. In its summary in this regard (cl 6.4.4) it concluded that the 'highest priority must be in zones where fuel reduction ... can be most effective in helping to reduce the risk of harm to people, destruction of property and damage to assets' from wildfires.

138 In February 2003, the Minister for the Environment provided the Department with a generic statement to be used in Ministerial correspondence and public statements as follows:

The Government strongly supports prescribed burning as the primary fire protection strategy for biodiversity values, private and public assets and rural communities in the south-west.

139 In late 2003, the Western Australian Government provided the Department with additional resources in order to carry out its prescribed burning programme. This included an extra $3 million in 2003 - 2004.

140 In December 2003 the Government issued a media statement, quoting the Minister for the Environment as saying:

We remain committed to prescribed burning to protect lives, to protect our forests, the biodiversity conservation values of those forests and to protect surrounding property.

141 On 10 December 2003, the Minister for the Environment approved the Forest Management Plan 2004 - 2013.

142 Fire is dealt with in the Forest Management Plan under the heading 'Ecosystem health and vitality'. The Forest Management Plan states:

Objective

An overall aim of the plan is to seek to sustain forest ecosystem health and vitality.

17. Fire

Fire has an impact on forest ecosystems. For example, fire can assist in the regeneration of vegetation. Fire can release nutrients required for plant growth, which might otherwise be locked up within litter and other dead plant materials. However, the excessive use of fire can be detrimental.

Attempts to suppress fire over large areas can result in the gradual build-up of fuel on the forest floor. This will increase the probability of large and intense fires, which are most costly to suppress and result in a greater loss of economic and biological assets. In most vegetated ecosystems, it is difficult to prevent fire for long periods over large areas. To attempt to do so may be detrimental to biodiversity conservation and the protection of the community.

There are strongly held views in the community that more prescribed burning is required to reduce the risk of wildfire and conversely, that less and a different type of burning is required to protect biodiversity. A review of fire management to be undertaken by the Environmental Protection Authority is proposed to address the use of planned fire for biodiversity management while allowing for protection of human life, property and other values.

Objective

The plan proposes the following Actions at the whole of forest and landscape scale for the purpose of seeking to use and respond to fire in a manner that:

• optimises the maintenance of forest ecosystem health and vitality;
• promotes the conservation of biodiversity;
• controls adverse impacts of fire on the social, cultural and economic values of land managed by the Department and adjoining land; and
• minimises the risk of smoke emanating from prescribed burns impacting on population centres and other sensitive areas.

Actions proposed

17.1 The Department will:

17.1.1 maintain a competent fire management, suppression and response capability;

17.1.2 prepare and maintain a fire management plan and smoke management guidelines;

17.1.3 undertake an annual prescribed burning program in a manner that:

• is in accordance with the fire management plan;
• is in accordance with the smoke management guidelines;
• has regard to the Goals for Understorey Structural Diversity referred to in Action 4.1; and
• considers any special vulnerability of fauna and flora known to exist in a particular area to burning in that area; and

17.1.4 consult with stakeholders and interested community members in a manner that seeks to develop community understanding of and support for, and enable constructive discussions and deliberations on, the planning and implementation of prescribed burning and other fire management programs.

143 In relation to the objectives identified under cl 17, three points should be made. First, in its context, the reference to 'fire' in the third dot point is, in my view, on its proper construction, a reference to wildfire, as opposed to prescribed burning. Secondly, on its proper construction, the reference to 'other sensitive areas' in the last dot point is, in my view, a phrase capable of comprehending a reference to, relevantly, areas planted with grapes. Thirdly, also in relation to the last dot point, the phrase 'impacting on population centres' comprehends, in my view, both smoke affecting the amenity of a population centre by the existence of a pall hanging over the centre, and also the potential for smoke to affect the health of the population, including, in particular, the health of those susceptible to respiratory illnesses.

144 Under 'Key Performance Indicators', the Forest Management Plan provided (page 46):

The risk to conservation, life, property and other forest values posed by wildfire.

Target to be determined following the Environment Protection Authority's review of fire management.

The Department to evaluate high-risk areas and incorporate

145 The 'Goals for Understorey Structural Diversity' (referred to in cl 17.1.3) is a document that the Forest Management Plan contemplated would be prepared by the Department for the 'purpose of seeking to conserve biodiversity through a diverse representation and distribution of forest structures and understorey seral stages through time' (cl 4, 4.1 and Appendix 1).

146 By cl 34.1 and Appendix 1, the Department is to maintain a comprehensive suite of operational guidance documents, including its Fire Operations Manual (as to which see [172] - [173] below).

147 The Executive Summary (page 8) to the report, although not forming part of the Forest Management Plan itself, stated:

Ecosystem health and vitality

The plan recognises the proposed review of fire by the Environment Protection Authority and the fact that policy may change as a result of it. However, the broad thrust of the recommendations from the Departmental review of fire management in the south-west forests are adopted, particularly that burn planning be based on diversity/structural models of vegetation units so as to promote the maintenance of biological diversity. This ties in with the proposal in the biodiversity section to set goals for understorey vegetation diversity. The commitment to maintaining an adequate level of protection for south-west communities remains.

148 The references in the Forest Management Plan, both in the Executive Summary and in paragraph 3 under the heading 'Fire' in cl 17, to the proposed review by the Environmental Protection Authority, are both references to the report published in October 2004 (see [150] below).

149 In 2003 - 2004 the Department was able to apply more resources to prescribed burning. The resultant increased levels of prescribed burning were the subject of a further statement to the Legislative Assembly by the Minister for the Environment in these terms on 2 June 2004:

DR J M EDWARDS (Maylands - Minister for the Environment) [12.04 pm]: I will inform the House of the latest situation regarding the Department of Conservation and Land Management's fire management activities in the south west forest areas of the State. Until rain intervened on Monday of this week, [the Department] had achieved prescribed burns over 195 000 hectares in the 2003-04 season. This compares with a total of 145 000 hectares in the previous year. In addition to that, a further 15 000 hectares were burnt in wildfires this year, along with 126 000 hectares in 2002-03.

...

This means that fuel loads have been reduced across 480 000 hectares of land managed by [the Department] in the south west forests in the past two years. This is a very good outcome in protecting property and community assets and conserving biodiversity values. It is significant that [the Department] has now achieved more than 120 000 hectares of prescribed burning in the south west forests this autumn. This is the highest level of autumn burning in decades, and the overall result so far this year is the highest level of prescribed burning for eight years.

Members may remember that last December, the Government announced the allocation of an additional $3 million to [the Department] for fire management in the 2003-04 financial year. This allocation has been a significant factor in allowing [the Department] to achieve its targets through the employment of additional seasonal personnel. That additional funding was provided in response to concerns that the backlog of burning programs in the south west was increasing and that there could be a significant risk to communities and environmental values. In further recognition of the need to protect property, community assets and biological values from the ravages of wildfires, the Government is allocating an additional $24.3 million to [the Department] over the next four years for fire management. Additional funding for 2004-05 is $3.6 million, with extra funding of $5.6 million in 2005-06 and $7.6 million a year in the following years.

...

The additional funding that the Government is providing to [the Department] and the development of new technology will assist [the Department] in placing a greater emphasis on managing fire for ecological purposes, as well as helping to protect human life, private property and community assets.

150 In October 2004 a committee established by the Western Australian Environmental Protection Authority (EPA) provided a report, to the then Minister for the Environment, into a review of the Department's fire management policy and practices in the Swan, South West and Warren administrative regions. The EPA made various recommendations in its report including that burn objectives be made clearer; that community consultation programmes be further developed; that the Conservation Commission audit the burn programme; and that the Department should document its planning and operational processes for prescribed burning.

151 The EPA report endorsed the Department's objectives with respect to prescribed burning. It stated in its conclusions:

The EPA recognises that a fuel reduction programme is a key strategy in reducing the extent and damage to biodiversity and other assets, which might otherwise be caused by wildfires. The value of a prescribed burning programme is not in reducing the number of wildfires, but rather their intensity, to assist in its suppression, and to reduce the damage they may cause to biodiversity.

The EPA supports the objectives expressed in [the Department's] draft policy and finds that it and the attached appendix outlining the rationale and principles for fire management practices are comprehensive. However, despite the fifty or so years of accumulated fire research that has been undertaken by [the Department] and others there are still many gaps in our knowledge and understanding of the complex ecosystems of the south west, for example the increasing effects of fragmentation of the forest areas, and climate change, including declining rainfall. Therefore, an ongoing research programme and an adaptive management approach need to be used by [the Department] to continuously upgrade its fire planning and operations as new information becomes available. [The Department] supports this approach.

With regard to the stated objectives for managing fire the EPA considers that the protection of human life must be a priority and that [the Department's] planning and procedures for life, assets and strategic infrastructure on its managed lands are effective. The EPA considers that practices in place are appropriate for minimising smoke from prescribed burns in the metropolitan area. Further consideration should be given, however, to the effects of this smoke on human health to establish the full significance of this relationship so that it can be balanced against the risks of wildfires to health and property.

With respect to neighbouring properties and assets in peri-urban developments and towns and communities in the south west forest areas, [the Department] has an obligation to manage the risk of wildfire either entering or exiting the lands it manages and applies a risk management approach to discharging its implied fire protection responsibilities to minimise wildfire impacts and potential liabilities. The EPA endorses the concept of buffer zones, in which the amount of flammable vegetation is reduced, both on surrounding [Department]-managed lands and on the adjacent privately-owned portions. The EPA notes, however, that whilst Government should be leading in its planning and provision of subdivisions that are safely located with respect to wildfire, Government will never be able to protect all private property, nor should it be expected to. Land and property owners need to accept responsibility for their own protection from fire, especially if they have chosen to live where the risk of fire is elevated.

The EPA recognises that the impact of prescribed burning on biodiversity is not adequately known or understood and understands that the community has strong and conflicting views on whether present practices address this adequately. While it is explicitly mentioned in [the Department's] draft policy, there is insufficient information on the effects of prescribed burning on WA's south west ecosystems to allow the EPA to come to a conclusive view that optimum outcomes are being achieved. As a consequence a cautionary approach needs to be followed, and adaptive management techniques applied as more information becomes available from fire ecology research. The EPA considers that more research is required in this area as a basis for the introduction of improvements to current fire management practices to facilitate the protection and enhancement of biodiversity.

The set of issues raised and the recommendations made will require ongoing assessment to determine whether they are achieving the desired outcomes and improvements suggested. The EPA believes that this fits into the same timeframe as is used for the review of the Forestry Management Plan and suggests that a formal review be considered in unison with the review of the Forestry Management Plan. (emphasis added)

The Department's planning and processes for prescribed burning

152 The processes and planning for prescribed burning were, to varying degrees, described by a number of witnesses. The following is taken principally from the evidence of Mr Sneeuwjagt, Mr Keppel, Mr Gillard and Dr McCaw of the Department, whose evidence I accept on these matters.

153 The Wildfire Threat Analysis system is a system developed by the Department for analysing the threat from, and planning appropriate responses to, wildfires. It provides a standardised, repeatable framework to analyse quantified data from the significant factors likely to contribute to the threat of wildfire and it consequences.

154 The Wildfire Threat Analysis provides an objective comparison of wildfire threat between different areas with different risks. Thus, it allows, for example, the threat of wildfires in the Wanneroo area (which has its own weather and fuel conditions) to be compared with the threat of wildfires in the Walpole region (with different conditions).

155 In conjunction with the Wildfire Threat Analysis, there is a rating system devised for prescribed burning, which uses the fire protection values and other management values referred to in the Wildfire Threat Analysis. The rating system is used to provide a guide for priorities of burning, and is used to ensure that the highest priority areas are burnt. Clause 1.5 of the rating system deals with the strategic value of the burn. Under this measure, the strategic value of a burn is allocated points between zero and 60, with 60 signifying the highest strategic value. DPHB8 was given the highest strategic value of 60 points.

156 The Wildfire Threat Analysis is a key input into the Department's Master Burn Plan. The Wildfire Threat Analysis does not, of itself, however, indicate the tactical value of any particular prescribed burn in relation to other prescribed burns. There are tactical benefits derived from burning areas in an integrated scheme and as part of an integrated sequence of burns, which are taken into account in the planning of burns by the Department.

157 The Master Burn Plan is a strategic plan designed by the Department to achieve its objectives for fire management. These objectives include the security of communities, properties and natural values from damage by intense wildfires; the sustainability of biodiversity values such as threatened plant and animal communities; and the protection of timber values, recreation assets and water catchment values.

158 The Master Burn Plan contains spatial data that includes all the values that can be affected by fire, as well as the Department's management activities and their associated fire-management requirements.

159 The Master Burn Plan is used to design a strategically-managed fuel age and vegetation/habitat mosaic across the land managed by the Department. The mosaic is developed and managed by a process of recording the historical prescribed and wildfire events, and integrating these with information on the fuel accumulation rates, the known fire responses of vegetation and fauna, the silvicultural requirements of forest stands, and the Department's fire suppression capabilities.

160 The process relies on a computer system, 'Geographic Information Systems', which permits the manipulation of large spatial data sets to allow 'what if' analyses to be undertaken of different conditions and different scenarios. It enables the Department's managers to determine the most appropriate time for future burning to be undertaken in each area to achieve specific management objectives.

161 The Fire Management Services branch of the Department is the Department's custodian of the Master Burn Plan.

162 The Master Burn Plan contains data for each region in Western Australia in which land is managed by the Department. Each Departmental Region develops a Master Burn Plan for the region, which is ultimately submitted to the Department's Regional Services Director, and to the Department's Corporate Executive for approval. Once approved, it is available for community access.

163 The Warren Region's Master Burn Plan has involved an annual burn target of 90,000 ha of prescribed burning. This target is set to maximise opportunities, with the intention of achieving, at least, a probable burn level of 60,000 ha. The higher targeted burn area allows for flexibility in the annual execution of the plan, and takes into account the variety of operational and other constraints, particularly in respect of karri forest, which can emerge in any year to inhibit the scope of the burning programme. Within the Warren Region, the Donnelly District target is 30,000 - 35,000 ha per annum of prescribed burning.

164 The Master Burn Plan comprises planning at four levels. One is a three-year (ie, six-season) indicative burning plan. The second is an annual indicative burning plan. The third involves prescriptions for individual burns on the annual indicative burning plan, and the fourth involves daily schedules in relation to the burning of particular areas on particular days. These aspects of planning are generally sequential, but in practice some may occur concurrently with others.

The indicative three-year plan, and the annual plan

165 The indicative three-year (six-season) burn plan is developed with input from senior regional and district staff taking into account:

• requirements for nature conservation and protection of biodiversity values;
• requirements for community protection around towns, private property and the infrastructure of public services;
• continued provision of strategic low fuel loads to assist in suppressing large, intense summer wildfires;
• fuel ages to determine areas of high fuel loads;
• the principles and information layers of the Wildfire Threat Analysis;
• the requirements for protecting the Department's infrastructure and assets in forests, parks and reserves;
• proposed timber harvesting operations where fire exclusion may need to occur until harvesting is complete; and
• areas of regenerated karri and jarrah forest that require fire exclusion until the new stands are old enough to withstand a prescribed burn.

166 The annual prescribed burn plan for the Warren Region is developed from the indicative three-year (six-season) burn plan for the region by identifying approximately 90,000 ha of burns, which are candidates for burning in the upcoming spring-summer fire season. The annual plan encompasses, in general terms, the first year of the three-year indicative plan.

167 The annual burn plans of the Warren Region are reviewed twice each year. The reviews are carried out after each burning season on or about 15 January and 15 June. The review covers the achievements of the previous season, and plans the proposed programme for the next season. The annual burn plan is updated and approved by senior District, Regional and Fire Management Services Branch staff, before being submitted to the Department's Corporate Executive for approval. If, for operational or other reasons, a burn in the annual programme is not able to be carried out that year, it will generally be 'rolled over' into the burn programme for the following year. Burns may be 'rolled over' from year-to-year for a number of years if constraints in undertaking the burn continue to operate in those years.

168 The EPA report of October 2004 noted (cl 4.1.5):

The implementation of a prescribed burn is the culmination of long and detailed planning processes at both a strategic and operational level. It involves input from other agencies - FESA, the Bureau of Meteorology, the Air Quality Branch of the Department of Environment - and local government. Likewise, the community should be informed.

[The Department] has in place set procedures for all of the above, with proforma sheets to be completed to help ensure nothing is overlooked. The detailed burn prescription dictates what should be happening on the day of the burn.

Despite the extensive planning leading up to the day of a prescribed burn there are variables, for example the weather conditions, which cannot be accurately predicted or controlled. A burn may have to be postponed on the day on which it is intended to be carried out. Even if the burning teams are mobilised they have to be prepared to change tack if weather conditions vary too much from predictions.

169 The annual burn plan is used by district personnel to develop a Prescribed Fire Plan, ie, a 'burn prescription', for each burn area for the upcoming season. The burn prescriptions include the Threat Index derived from a Wildfire Threat Analysis of each burn.

170 The burn prescription is the instrument that specifies the burn objectives, the weather and fuel conditions required to meet these objectives, the lighting sequences and patterns to be adopted to achieve the fire behaviour, burn coverage standards, and smoke management requirements. The burn prescription includes a 'pre-burn checklist', which identifies potential risks and impacts of the burn operations, including smoke impacts, on the values and activities within the burn and on neighbouring lands.

171 The purpose of the checklist is to provide the burn managers with an early opportunity to identify the risks and consequences and to take steps to modify, postpone or cancel the burn if these risks and consequences are significant. For example, where the impact of smoke on population centres, airports, public roads and neighbouring properties is identified in the pre-burn checklist analysis, consideration is to be given to the wind direction, and lighting sequence and patterns that will minimise the risk of smoke impacts on any of these values.

172 Fuel management requirements for jarrah and karri forests are contained in the Fire Operational Guidelines and reflected in the Objectives and Standards that are written into Prescribed Fire Plans.

173 The Department's accepted standard for fuels in karri forest is 15 - 19 tonnes per hectare and the Department's fuel reduction requirements are contained in Fire Operations Guideline 61. In karri forest such as DPHB8 generally the fuel reduction requirement is to reduce ground fuels to less than 8 tonnes per hectare over 60% - 80% of the total burn area.

174 The daily operational scheduling of prescribed burns taken from the approved annual burn plan is done daily via State-wide telephone conferences organised by the Fire Management Services Branch of the Department. Participants in the telephone conference include district, regional and Fire Management Services Branch personnel. Sometimes, the Director of Regional Services attends the telephone conference, if his input is required on key issues affecting the prescribed burning programme for the day.

175 There are two telephone conferences each day. One is at 8.30 am and the other at 4.15 pm.

176 The objectives of the 8.30 am telephone conference are to review the two-day weather forecast and synoptic information provided by the Bureau of Meteorology, to discuss any wildfires occurring overnight, to review the achievements of the previous day's burning, and to consider requests for the upcoming day's prescribed burning activity. Requests for new burning and further ignitions of existing burns are submitted by each region's Fire Duty Officer to the State Operations Officer of Fire Management Services Branch by telephone conference.

177 In the course of the telephone conference, the State Operations Officer agrees on a final burn programme for the day, after considering key factors such as existing fire commitments, the level of resources available and smoke management. Staff, fire crew and aircraft resources to achieve the planned work are also co-ordinated. The Manager, Fire Management Services Branch, approves the draft daily burn programme. The approved programme is submitted to the Director of Regional Services for noting or further endorsement where burns may be politically sensitive.

178 The second telephone conference occurs at 4.15 pm each day. On that occasion, the next four days' weather forecast, synoptic information and prescribed burning programme are considered.

179 In undertaking the annual prescribed burn programme, the Department's Donnelly District office carries out annual community consultation meetings with stakeholders, landowners and community leaders. These meetings normally occur in October of each year. During these meetings, the previous year's prescribed burn programme and achievements are discussed, along with the location, size and causes of wildfires. The indicative three year (six-season) burn plan is also discussed and feedback is sought.

180 The focus of consultation meetings is usually on seasons four to six in the indicative three-year programme, given that it is a rolling programme and seasons one and two would ordinarily have been discussed in previous years, and the planning for the burns in those two seasons would generally be well advanced.

181 The meetings generally involve discussion about the size, location, timing and boundaries of individual burns. Notes are taken and, where practical, the programme is adjusted to take account of community feedback. Copies of the six-season indicative burn plan are also provided for people to take away, circulate and discuss with their group members.

182 The Department's processes in the Warren region for notifying the public of prescribed burning activities include the posting of Operational Lists in the Manjimup Bridgetown Times. The lists include the burn identification number, the location, and the size in hectares of the proposed burn.

183 Prescribed burn notices are also sent by mail to neighbouring properties outlining the Department's intention to carry out a prescribed burn near or adjacent to their property. The notices expire each 28 days, and new notifications are sent if the prescribed burn remains on the annual programme and has not commenced.

184 The Bureau of Meteorology (the Bureau) provides a range of weather services to the Department. These include products available on the Bureau's public website, and specific fire-related products available on a restricted-user website accessed via the Department's Fire Management Services website.

185 Specific fire weather forecasts are provided twice daily for eight locations nominated by the Department. The morning forecast provides details of maximum temperature, minimum relative humidity, dew point, and wind strength and direction during the day at 1100, 1300, 1500, 1700, and 0300 hours. Observations of wind strength and direction at five heights (1,000 - 7,000') above the surface level are also provided from the early morning balloons launched at Perth and Albany. The afternoon forecast provides an outlook for temperature and wind strength and direction over the following four days.

186 Weather observations are recorded by automatic weather stations at each of the nominated forecast locations, except for Pemberton which still uses manual observations. Automatic weather stations offer a number of advantages including recording data at hourly or half-hourly intervals throughout the diurnal cycle, and the ability to remotely access weather observations via the Bureau's website.

187 The Bureau's restricted-user website provides predictions of smoke dispersal for nominated source sites in the South West, Lower West and Perth metropolitan forecast districts. The model provides general guidance as to the likely direction of smoke dispersal, but is limited by inherent uncertainties in weather forecasting, and does not provide an absolute measure of the density or composition of a smoke plume and how this may change over time.

188 Whilst the Bureau provides a high standard of weather forecasting services to the Department in relation to its fire management activities, there are some specific weather phenomena that are particularly difficult to forecast reliably, which can significantly influence fire operations. These include:

• the initiation of the easterly migration of the west coast trough;
• the nature of north-west wind changes and how this affects the direction, strength and duration of winds; and
• conditions leading to subsidence of dry air masses from aloft that can cause abrupt drying at the surface level.

189 The Department had not received any complaints or indications of concern regarding smoke affecting the quality of grapes prior to 2001.

190 Mr Walker, the Director of Regional Policy and Projects in the Department, who had worked for the Department or its predecessor since 1970, and who had been involved in its fire management programmes, was familiar with the prescribed burning programmes carried out by the Department in areas including wine growing regions such as Margaret River. In the 40 years that grapes have been grown in the Margaret River region, he had never received or heard of any complaint by vignerons about smoke from prescribed burning affecting grapes, prior to the events in these proceedings.

191 The Warren Region has, and had over the past 35 years, many sources of wood-smoke emissions each year. Smoke has been emitted from saw mills and other timber processing facilities that burn some of their waste materials. Prescribed burning of bush and heaps of logs on private properties is a common practice in the region. This is done as part of farming operations within the establishment of timber plantations. Every fire season the Department has recorded hundreds of smokes from fires burning on private properties in the region. These smokes are plotted by fire-spotting towers and aircraft and they are recorded in the Department's district office logbooks. In addition, smoke has also been emitted from the Department's own prescribed burnings, and from wildfires that have occurred from time to time across the south-west, over the last 30 - 40 years. Mr Muller, an expert for the plaintiffs, said that smoke can travel long distances and that smoke from prescribed burning of the Warren Region has blanketed Perth. Despite the smoke emissions which have affected the area over that time, the Department was first notified of a potential issue concerning the effect of smoke on wine grapes in 2001, from the plaintiffs.

192 Similarly, as indicated earlier, Mr Campbell-Clause, whose firm had developed and managed the plaintiffs' vineyards, had not been aware of any potential connection between smoke and grape quality prior to 2001. Mr Pearse, the plaintiffs' vineyard manager at that time, also said that he had no knowledge or suspicion that grapes might be affected by smoke until after the 2001 vintage.

193 In late 2001 Mr Marshall, of the Department's Donnelly District, asked Dr McCaw of the Department whether there was any scientific information available about the effect of bushfire smoke on grapes and vineyards. Dr McCaw is and was a principal research scientist with the Department at its Manjimup office. Dr McCaw also had substantial experience in management of prescribed fire programmes and incident control in south-west forests. He had undertaken the role of Regional Duty Officer for the Warren Region for each fire season since 1998, and in 2001 - 2002 he was the Regional Fire Co-ordinator for the Warren Region. He was an honest, knowledgeable and careful witness.

194 Dr McCaw had not come across the question of smoke taint before. He was aware that other winegrowing regions in Australia such as the Barossa Valley, the Hunter Valley and north eastern Victoria, and overseas, such as southern California and southern France, were also subject to bushfire smoke from time to time. He thought there may have been a body of published scientific literature relating to the impact of smoke on grapes. He accordingly made inquiries of the Department of Agriculture around 6 November 2001 for information on the matter of smoke taint of wine grapes. The Department of Agriculture was unable to assist, but suggested that he contact the Australian Wine Research Institute (AWRI).

195 At that stage, Dr McCaw considered that in the absence of viticultural expertise in the Department, it was probably not necessary to seek to delve too deeply into the science of the potential connection between smoke and grape quality, and that the better course would be to deal with vignerons as another stakeholder group in the course of devising burn strategies.

196 Dr McCaw, nevertheless, contacted the AWRI to ascertain the level and extent of scientific information available on the impact of smoke on grape quality. He checked the AWRI website and spoke to Ms Bell of the AWRI on 7 February 2002. The AWRI was unable to identify any relevant literature but provided advice about the testing procedure that would need to be undertaken to establish any link between smoke exposure and taint.

197 Also, shortly prior to this, Mr Marshall prepared a note following a discussion with Mr Pearse. He attached the note to the prescribed burn plan for DPHB8. It said, inter alia:

Murray Pearse ... has only found out ... recently that last years grape crop was a complete loss.

Murray said the grapes were smoke affected from last years aerial burn ... (they had smoke blown across the grape vines for days).

Murray said they could not afford to have this happen again ... he said the critical time would be from early December to mid-April.

Murray is not against prescribed burning - he is only concerned with the smoke problem with the grapes.

I suggest CALM carry out this burn with winds from the East to North east. This will keep the smoke off the grape vines. (emphasis added)

198 There was some dispute at trial as to whether Mr Marshall's note constituted a formal variation to the burn prescription. Mr Muller, for the plaintiff, who had previously worked for the Department, said that Mr Marshall was not authorised to vary the burn prescription. Mr Sneeuwjagt's evidence was to similar effect (ts 954). The District Manager, Mr Gillard, on the other hand, said that Mr Mr Marshall's note was a formal amendment to the burn prescription (ts 496). I find that Mr Marshall's note constituted an amendment to the burn prescription upon, but not prior to, Mr Gillard signing and authorising the burn prescription, with the note attached, on 1 October 2002. The point is not, however, ultimately of significance because the burn prescription was subsequently further amended, in any event, by Mr Gillard's note of 3 February 2004 ([236] below), prior to the burn of DPHB8.

199 Based on information gained from his inquiries, Dr McCaw drafted a response to Mr Pearse's letter dated 18 February 2002. The response was sent on 18 March 2002. It said, inter alia:

In order to be better informed about this issue I have sought advice from the Western Australian Department of Agriculture and the Australian Wine Research Institute. From these discussions it appears that there is very little published information about the potential for smoke and other products of combustion to affect the quality of wine grapes. This is despite the fact that a number of premium wine producing regions are located in areas where smoke from bushfires and other sources may be present from time to time ...

I appreciate the expression of support for the Department's prescribed burning made in the concluding paragraph of your letter. It is clearly in the best interest of the whole community that forests are managed in a way that minimizes the threat of high intensity wildfires to life, property and community assets while at the same time providing for conservation of biodiversity. Prescribed burning is an important tool in achieving this outcome. In undertaking its program of prescribed burning the Department works within a number of constraints imposed by climate and weather factors, which include:

- the need to conduct burning operations when fuels are in the desired range of moisture content. In particular, fuels in Karri forest may only be at suitable levels for prescribed burning for a period of a few weeks in late spring and early summer, and sometimes for a week or two in autumn. In a normal season it would be unlikely that Karri forest would burn before mid November or remain dry enough for prescribed burning beyond mid-April,

- the need to ensure burn security by removing sufficient fuel to minimize the risk of later re-ignition as conditions become warmer and drier;

- avoidance of burning under unstable weather conditions associated with winds in the north-east to north-westerly quadrant, when fire danger may be too high for safe prescribed burning.

With regard to selecting wind directions to disperse the smoke from prescribed burning, the Department takes into account a range of factors including avoidance of smoke accumulation in large population centers, safety and visibility along main roads, and the requirements of land uses on properties adjoining Departmental estate. In the context of the expanding wine industry across the Manjimup-Pemberton region it is not practical to give an undertaking that vineyards will not be affected on some occasions by smoke from prescribed burning or from unplanned wildfires.

There is clearly a need to work through this issue and reach an outcome that is acceptable to the wine industry, the Department and the local community. I look forward to working with you and staff from the Donnelly District office to achieve this.

200 Mr Pearse forwarded Dr McCaw's letter to Mr Campbell-Clause. There was no response to Dr McCaw's letter.

201 Mr Pearse said that later in 2002, Mr Bennett of the Department visited the plaintiffs' property and discussed smoke from prescribed burning. I think Mr Pearse's recollection of the timing of that event is probably incorrect, and that the meeting occurred in December 2003, which is when Mr Bennett says the meeting occurred. (I refer to this meeting further below.)

202 In January 2003, the Canberra bushfires caused approximately 3 million ha of land to be burnt throughout the Australian Capital Territory, New South Wales and Victoria, including over 90% of the land managed for nature conservation by the Australian Capital Territory's relevant department. It became known that the wine grape production industry was impacted by smoke from the bushfires on a regional scale. The issue of smoke taint started to gain currency.

203 In April 2003, the 'Weekend Australian Magazine' published an article on smoke from the Victorian bushfires damaging grape crops. The article was sent by Mr Gillard to Dr McCaw of the Department.

204 During March 2003, the Department conducted prescribed burn DPHB26 in the Gloucester National Park, adjacent to the Gloucester Ridge winery. The Department did not receive any complaints about smoke taint arising from the burn.

205 On 10 April 2003, Salitage Winery in Pemberton sent an email to vignerons on the AWRI list and to the Department (Mr Winfield), in these terms:

Smoke effects on grapes.

Following the major bush fires in the Eastern states the AWRI has done a significant amount of research about the effects of smoke haze on grapes and wine.

Growers in the East have had grapes rejected and wineries have found that batches of fruit, including white destined for sparkling base are unsuitable due to smoke character.

Many of the areas worst affected have been from vineyards where the smoke haze was not dispersed overnight and so the fruit was under a constant smoke fog.

I understand that the AWRI informed [the Department] some three years ago about the possibility of smoke [a]ffecting grapes as part of the controlled burn off process and recommended that burning off be left until after the grape harvest had finished. In the main this appears to be happening.

With the large number of lightning strikes around the district this year some of you may have noticed smoke characters in your ferments. It may be possible to assist the AWRI in their research by the provision of samples. The AWRI do not wish to [be] inundated with samples but are willing to do some investigation.

The smoke character (guiacol) is more notable in lighter style reds than fuller bodied reds, i.e Pinot Noir than Shiraz or Cabernet Sauvignon and the threshold levels consequently alter depending on the weight and body of the wine.

Consequently those of you who have noticed ferments that are notably smoky in character or growers who have had the purchaser of their fruit make comment and would be willing to send samples should contact Patrick at Salitage and he will outline the sampling criteria and conditions needed so as to ensure the samples sent are suitable for investigation purposes and who to forward the sample to.

206 On 11 April 2003, Mr Winfield forwarded the Salitage Winery email to others in the Department, including Messrs Keppel, Gillard, McCaw, and Simmonds of the Department. Mr Keppel asked Mr Simmonds and Dr McCaw to follow it up, and to be prepared to deal with any inquiries.

207 On 8 May 2003, Mr Steve Partridge of Agribusiness Research and Management emailed Mr Keppel of the Department in relation to grapes affected by smoke at the Brown Brothers winery in Victoria, in consequence of the Victorian bushfires. Mr Partridge was a vineyard management consultant in the south-west. On 9 May 2003, Mr Keppel forwarded Mr Partridge's email to other Departmental members and suggested that the Department undertake research and develop a dialogue with the wine industry.

208 On 9 May 2003, Mr Maher of the Department responded to Mr Keppel's email, with copies to other members of the Department, in these terms:

Bob Chandler provided me with an article about the [e]ffect the Vic fires had on the harvest Browns etal There [was] no scientific reasoning provided with this article. If [Dr McCaw] does not still have I can provide a copy. The amount of smoke the Vic grapes were exposed to was significant ie one month plus in some cases.

209 Mr Maher had visited the scene of the Victorian bushfires on behalf of the Department.

210 On 19 May 2003, an email was sent from Mr Winfield of the Department to Dr McCaw, Mr Sneeuwjagt, and others in the Department, in these terms:

Steve Partridge from Agribusiness Management Group just called. He'd been at a function in Adelaide on the weekend and talked to Peter Godden who is leading the research and monitoring of this issue for the Australian Wine Research Institute at Adelaide Uni.

He said that there seemed to be no point [in the Department] seeking further research - the evidence they have from the fires in the east this year is compelling. I've e-mailed Peter Godden and asked if we can be kept informed of AWRI data, papers etc relevant to the topic.

My view is that FMS and or Science Div. should develop a relationship with AWRI to get in the information loop. Then perhaps organise some sort of forum on the topic with WA Wine Industry Association (WIA).

I agree that establishing a dialogue through some type of forum with the industry is the way to go. The sorts of issues that would be useful to know more about include:

- effects of exposure time on grape quality,

- effects of different types of smoke (eg. smouldering smoke with high particulate concentrations from [silvicultural] burns, versus smoke from litter bed, versus wildfire)

- what concentrations of smoke are required to cause detectable changes in grape quality

212 On 20 May 2003, Mr Godden on behalf of AWRI emailed Mr Winfield, and said:

Yes, we have done a lot of work on the bushfire smoke taint issue this year, and have some pretty good data. I am trying to get it written up for publication, but we have many other pressing matters at the moment, so publication is not imminent. However, if you would like to give me a call I would be happy to discuss what we did, and what we found.

213 On 23 October 2003, Mr Simmonds of the Department, in response to an email sent by Ms Green of the Department in relation to another vigneron raising concern about smoke taint from a prescribed burn, said:

At regional level [Dr McCaw] is handling this issue. [Mr Winfield] is our contact with the viticulturists. The long and short of it as I understand it is if burning occurs upwind of a [vineyard] within 2 to 4 weeks of harvest, heavy smoke may settle in low lying areas. If this occurs with a dew smoke particles can collect in the water [droplets on] the grapes. When the droplets dry the smoke particles remain on the skin of the grape. Depending on the number of exposures, smoke density etc. it may affect wine quality. There does not appear [to be] an issue outside of this time frame.

Give [Dr McCaw] a ring.

214 In December 2003, Mr Pearse received a further burn notice from the Department in relation to prescribed burn DPHB8. Mr Pearse, on behalf of the plaintiffs, telephoned Mr Bennett of the Department to advise that he was not happy with the burn of DPHB8 proceeding prior to harvest, and raised the smoke issue from 2001. In response, Mr Bennett went to see Mr Pearse at the vineyard to discuss the proposed burn of DPHB8. Mr Bennett took with him a map and showed Mr Pearse the burn outlines and described how the Department intended to do the burn. Mr Pearse reiterated his concern about the prospect of smoke tainting the grapes. Following this meeting, Mr Bennett spoke to Mr Gillard and Mr Marshall about the issue of smoke and how the Department could implement the burn to minimise the smoke over the plaintiffs' vineyards. The discussion involved how the Department could edge the burn with south-east winds and undertake the core ignition the subsequent day on more easterly winds.

215 On 17 December 2003, Hardy Wine Company Ltd in Pemberton contacted Mr Bennett of the Department to inform him that its grapes around Canberra and Tumbarumba had been damaged by the fires earlier that year, and to request that another prescribed burn, DP33, to be undertaken near Hardy Wine's Pemberton vineyards, be deferred until after harvest. The issue was raised with Mr Gillard who in turn, on 19 December 2003, emailed Mr Simmonds of the Department in these terms:

Can you discuss this issue with [Mr Sneeuwjagt]. I am keen to hear his comments on the e-mail from Hardy Wine Company. I am confident there won't be an issue of SE winds [and we wouldn't commence this burn unless we were at the start of a stable SE cycle]. We will wait for your advice on this before commencing the burn [the bulk of the burn probably won't be dry enough now until after Christmas]. (emphasis added)

216 On 22 December 2003, Mr Walker sent an email to Messrs Sneeuwjagt, Gillard, Keppel and others of the Department and said:

I think we need to communicate to landowners in the wine industry that whilst we will try to avoid smoke over vineyards during prescribed burning, we can't guarantee it.

I don't think we should contemplate deferring a burn such as this, that was scheduled for late spring/summer, if the conditions are right to [complete] the burn.

It can be argued that the fire management regime on [Department]-managed lands was in place before the vineyard was established. The onus is on the landowner to manage around this circumstance, not us!

We also need to consider whether we deal with this issue piecemeal in the future or attempt to have a dialogue with vignerons as a group. If we go the latter way we might finish up with a bigger issue on our hands.

An individual response strategy might be preferable at this stage. Your thoughts please.

217 Mr Sneeuwjagt replied by email on the same day:

I agree with your comments. I believe the onus is on the [neighbouring] [vineyard]/winery to manage around the fire issue. We cannot guarantee that smoke from multi-stage burns that may burn over several days will not affect the [neighbouring] lands. We must not allow this to be yet another constraint on the burn program that is already seriously behind schedule. I think we should attempt to liaise with the vignerons as a group if such groups exist in the southwest. I will seek information on the names/contacts of representatives of these groups.

I trust that Donnelly District will proceed with the Treen Brook burn at the earliest opportunity as soon as fuel moistures are suitable.

218 On 22 December 2003, Mr Sneeuwjagt also asked Mr Winfield to provide a list of contacts for growers in the lower south-west region to enable the commencement of dialogue on the issue of prescribed burning and smoke management.

219 Also, on 22 December 2003, Mr Winfield emailed Mr Sneeuwjagt, referring to the economic value of the wine industry and suggesting the need to talk to AWRI.

220 On 26 December 2003, Mr Keppel sent an email to Mr Maher (and others in the Department) in the following terms:

We need to have a Dept position on the smoke and grapes issue. When we have this it will need to be worked through with the grape growers, industry groups and community.

I believe if we dont take this approach the commencement of [prescribed burns] in the karri forest will continually be frustrated and postponed to a point that the burns [don't] get done. This has occurred in the past 2 years.

[In] 2002 and 2003 I suggested we develop a dialogue with the Wine Industry peak groups to start working this issue through.

Please advise how the Dept proposes to work the issue through so we can manage accordingly.

221 In late 2003, the AWRI published research entitled 'Investigations conducted into the nature and amelioration of taints in grapes and wine caused by smoke resulting from bushfires'. A copy was sent to Mr Gillard of the Department by fax on 27 January 2004. It was circulated to various people in the Department including Dr McCaw, Mr Maher, Mr Keppel and Mr Sneeuwjagt. The AWRI publication reported on the examination of smoke taint in grapes resulting from the bushfires that occurred in Victoria and New South Wales in early 2003. In its conclusions, it said:

The Institute had previously been asked to consider if smoke resulting from controlled burning of bushland might result in a taint occurring in grapes in nearby vineyards, and had provided advice that there was a possibility that such a taint could arise. The results of this investigation provide conclusive evidence that tainting of grapes and wine by bushfire smoke can occur, and potentially have a major economic impact. Many affected white juices, wines and especially sparking base wines, were deemed to be 'unfit for purpose' and were consequently severely downgraded in terms of value. It remains to be seen what the commercial impact of the elevated guaiacol levels in many red wines will be.

As Australian viticulture continues to spread into locations that might be considered as more bushfire-prone than many established grape growing areas, there is a possibility that smoke taint might become a sporadic but more common occurrence in the future. It is evident that few vineyards were actually damaged by the 2003 fires, but damage caused by smoke taint was widespread. It is therefore apparent that, in this situation, insurance cover for smoke damage would have been of greater benefit than for fire damage. However, while it appears that few growers had insurance cover for smoke damage, many do maintain insurance against contamination, and in the Institute's view guaiacol in grapes and wines resulting from the bushfire smoke, is a contaminant. (emphasis added)

222 Notwithstanding the reference to insurance in the AWRI report, there was no debate in this case about the prospect of insurance cover for smoke taint.

223 By January 2004, the Department had received communications from other grape growers in relation to planned prescribed burns in the south-west involving DP33 (a block of Crown land in the Treenbrook State forest located approximately 5 km north-west of Pemberton), and in relation to prescribed burn DP14 (a block of Crown land in the Boorara Conservation Park located approximately 15 km south-east of Northcliffe). A proposed burn of a small section of DP33 was deferred on various grounds, including the smoke taint issue, which Mr Bennett in his email to Mr Gillard said was being 'sorted out from a higher level'. In relation to DP33, Mr Bennett in his email to Mr Gillard of 6 January 2004 summarised the position thus:

I have only spoken to some owners around DP 33.

Below is an outline of discussion and recommendations.

Houghtons - Loc 11938. Kim Verral is the manager of the vineyard here. I spoke to him yesterday and inspected the berries on the vine.

Currently the berries vary in size from 8 - 12 mm. They are still very hard. I told him that the issue is being addressed at a higher level and [the Department] was [still] to complete the burn. We would need to communicate this on the day of the burn if the decision is to go [ahead].

Channybearup Wines - Loc Lot 3 and 4. ... John Rowe is the manager. My discussion with John was very favourable. His belief is that while the berries are small and hard there is no risk to the grapes from smoke. He however did tell me of an incident that occurred in the early 1990's at Smithbrook Wines where [the Department] was conducting a burn while the machine was picking. Smoke was around for 3 days and blowing through the vineyard. This was a case where smoke did affect the wine. He believes that there is a date around the 26th January when the berries begin to soften and then the grapes may be at risk. He tells his workers that farting in the vineyard is out of bounds when picking as a side joke as they have heaps of absorption potential.

His last comment to me was go ahead and burn.

Vineyard at Pimelea - Loc 10080 ... John Cole is the manager. No issues with him. He had much the same story as John Rowe.

Silkwood Wines - Loc 5204. ... John Allen is the manager. John was in agreement with John Rowe. He believes there is only an issue when the grapes are at vintage. He would like to see Giblett Block burnt on the north side of him.

Phillips Estate - Loc 9649. ... Phil Wilkinson is the manager. Left a message to call me but no return at this stage.

Kevin Knight - Loc 8205 - No contact made at this stage.

Mr Bennett recommended that the Department proceed with the burns while the berries were small and hard, providing he made contact with the people to whom he had not yet spoken.

224 DP33 was then carried out on 7, 8 and 9 January 2004. DP33 was burnt on south-east winds, in accordance with the Department's planning prescription. The location and axis of Hardy's vineyard was such that with south-east winds, the smoke was blown away from the vineyard.

225 On 9 January 2004, Mr Simmonds sent an email to Mr Sneeuwjagt which said:

As requested info on DP33 and the grapes. This is a problem that is beginning to face us at every turn and needs a industry and departmental agreement on what are the problems, what are the critical time frames, duration and tolerance and some agreement on a underpinning of science.

Operationally we have in excess of 55 vineyards in the region and just today I was made aware of another being developed along Sevan Day road so it is an escalating issue that demands scarce resources without a long term resolution. I suggest in the short term to assist us in planning that we have vineyards added to our COG maps.

226 On 9 January 2004, Mr Sneeuwjagt replied to the effect that 'it was a good idea to map the vineyards' and requested information as to the latest date that grapes were picked. He said 'I would be concerned if this runs into our autumn burning period'.

227 In late January 2004, the Department became aware of a visit to Western Australia by Mr Peter Godden, the author of the section of the AWRI Annual Report regarding smoke taint of wine grapes, and took steps to arrange an informal meeting on 20 January 2004 at the Department of Agriculture (WA) with the Department's senior fire managers at the State, Regional and District levels including Messrs Gillard, Sneeuwjagt and Simmonds.

228 I accept the evidence of Mr Sneeuwjagt and Mr Gillard to the effect that, at that meeting, Mr Godden reported that there was evidence that wine from Victorian vineyards that had been exposed to heavy smoke for many weeks from the large bushfires in 2003 had been tainted. In response to questions from Mr Sneeuwjagt as to what evidence there was on possible wine taint from prescribed burn smoke, Mr Godden said that there was none. Mr Godden also said that he was not able to indicate whether the short duration of smoke exposure from a day or two of prescribed burns represented a significant risk to wine growers other than to say that it was a possibility.

229 In relation to DP14, on 22 January 2004, Mr Gillard sent an email to Mr Simmonds which said:

Allan Rudd has had a change of heart re smoke over his vineyard [NL 9988]. He is now saying that because the majority of his vineyard is in a gully and he has a mix of white and red varieties he does not want the burn to proceed until after he has harvested his grapes in mid to late March. We initially 'talked him around' but he then went back to his wine maker and rang [Mr Bennett] again last night saying it is a serious issue and he definitely does not want the burn to proceed.

There is no way we can complete the burn without putting smoke over the vineyard [which will last for several days].

On this basis I am not comfortable with us starting the burn [if we get suitable conditions in the next couple of weeks]. Can you discuss this with [Mr Winfield] and [Mr Sneeuwjagt] please and let me know the outcome.

230 In relation to DP14, Mr Gillard emailed Mr Walker on 23 January 2004 and said:

This isn't straight forward. Some members of the Northcliffe Community are opposed to the burn. Northcliffe for Future Generations and Northcliffe Environment Centre. The brigade network and Fire for Life are keen to see the burn go ahead. Of the direct neighbours Max Rudd is supportive but concerned about us burning this time of year, Alan Rudd is supportive of us doing the burn but doesn't want us to start until he has harvested his grapes [other investors have a stake in his vineyard and they are very nervous about the potential impact on wine quality]. Noel Makin is supportive of the burn going ahead. Ken Gwynne is supportive of the burn but concerned about us burning in summer. Eric Pratt is supportive of the burn going ahead. A bit of a mixed bag and unlikely to result in any 'local pressure' on Alan Rudd. I have tried to contact Alan again without success. Our last contact with Alan indicated that he was not prepared to take the economic risk even though he acknowledged he would be seen by some as 'the biggest mongrel around'. Given the potential economic impact and potential for a damages claim I am not prepared to start this burn without Departmental approval.

I look forward to your advice.

231 On 23 January 2004, Mr Sneeuwjagt emailed Mr Gillard advising that prescribed burn DP14 was a critically important strategic burn with which Mr Gillard should proceed with as soon as possible.

232 In relation to DPHB8, Mr Gillard reported to Mr Simmonds also by email dated 23 January 2004 and said:

Peter Godden's visit to WA has obviously created an increased level of interest in the smoke issue with local wine growers. Murray Pearse the manager from Treen Brook Estate is now saying that his whites are entering the 'raison'[?] stage and there is a significant risk of smoke contamination. He still supports us burning but now wants the burn delayed until after harvest. He was not prepared to negotiate on this and estimates the value of this years crop at $1M. The vineyard is owned by the Barwick Wine Group and they have 180 acres planted on NLs 8945 and 8946.

We have also spoken with the owner of Treen Ridge - Barry Scottman - NL 11344 - he has no issues with the burn but has leased his vineyard. We haven't discussed the issue with the current lease holder [message left on answering machine].

We are unable to complete this burn without smoke over the vineyards. Given the potential economic impact and the possibility of a damages claim I believe this burn will also require corporate approval.

233 On 27 January 2004, Mr Keppel, orally and by email, briefed Mr Walker about an increasing level of concern about the potential impact of smoke from prescribed burns on wine grapes. The email stated:

As you are aware we have been dealing with the issue of smoke created from prescribed burns and potential affect of smoke on wine grapes in the Warren Region on an individual burn basis for the past 3 years. During this time we have recommended to FMS that the issue be dealt with at a 'industry level'. The reason for this is that we are unable to reach agreement with individual growers in regard to if and when the smoke from prescribed burns carried out near their properties will adversely affect the quality of their wine grape crops. The lack of clear understanding of this issue is affecting our potential to carry out prescribed burning during the summer and autumn period.

Recent communication by our Donnelly district staff with local wine grape growers highlights that the issue is complex, and that locally the wine grape growers lack the scientific information necessary to make an informed position as to whether our prescribed burning will affect the quality of their wine grapes (and subsequent wine that is produced from the grapes) or not. As a result of this situation wine growers positions and opinions often vacillate between having no concerns of the burn being carried out to [where] they express a concern, request that the area not be burnt and threaten legal action if we go ahead and carry out the burn. I understand that [Mr Sneeuwjagt] is working with representatives from the State industry body in an endeavour to ensure that a more informed objective position is formulated. Additionally we are experiencing the situation where pro and anti prescribed burn sections of the community are using the issue to promote their positions.

Whilst I fully support, understand and accept the Depts position that for the overall protection of the community we are committed to getting on with our burn program I believe the issue has the potential to escalate into a political and legal situation if we go ahead and carry out prescribed burns that put smoke over the vineyards adjacent to the areas being burnt. In Warren Region we have more than 55 vineyards in the Manjimup, Pemberton, Northcliffe and Denmark areas that could be involved in this situation during our summer and autumn burn seasons.

For the above reasons I recommend the Executive Director and Minister be informed of the situation and advised that we will continue to consult with wine grape growers at a local and state industry level. But our opportunity to burn will not be postponed or cancelled to accommodate an individual wine grape growers opposition to a burn being carried out_._

Please advise if my position is acceptable. (original emphasis)

234 Mr Walker instructed Mr Keppel to refuse requests to defer prescribed burns at the request of local wine grape growers.

235 On 30 January 2004, Mr Keppel sent an email to Mr Gillard advising that he had discussed the implementation of prescribed burns with Mr Walker and that the Department's position is that it would not delay the commencement of prescribed burns at the request of local wine grape growers.

February 2004 and Ministerial approval of the Contentious Issue Briefing Note

236 On 3 February 2004, Mr Gillard spoke to Mr Pearse to advise him that the conditions appeared suitable to commence prescribed burn DPHB8 in February 2004 and to advise him of the Department's position on smoke and wine grapes. Mr Pearse reiterated the plaintiffs' concern, particularly in relation to smoke on the western boundary. Burning of a section of pasture on the plaintiffs' land was also discussed. Mr Gillard prepared a file note of his telephone conversation with Mr Pearse and attached it to, and thereby amended, the burn prescription. The note recorded, inter alia:

[Plaintiffs'] Vineyards ... will be an issue. The most susceptible varieties are on the western boundary. We need to complete the burn on ESE - E winds to minimise smoke over the vineyard. Murray Pearse ... is also happy for us to burn a strip of pasture inside the property if it will make the burn easier. Need to inspect in the field with Murray. This burn needs lots of resources with the aim of completing the burn in 2 days - again to minimise impact on the vineyard. (emphasis added)

237 Also on 3 February 2004, Albert Haak & Associates wrote to the Department regarding the Department's intention to proceed with DPHB8. The letter said, inter alia:

Our acting vineyard manager at Pemberton rang me today concerned that [the Department] are preparing to carry out a controlled burn near the vineyards that we manage. We manage three vineyards on Packer Road, Treenbrook, Old Vasse and Barwick Estate totalling nearly 75 hectares.

We have experienced problems in the past with fruit quality following a late summer [Department] burn in 2001. ...

The Cabernet is now at veraison meaning that it is in the fruit ripening stage. Other varieties including Chardonnay and Pinot Noir are more advanced. Harvest will begin next month and will not be complete until late April.

We are extremely concerned that you are contemplating a burn at this stage of the grape season. Any smoke entering the vineyard at this stage of the season will jeopardise fruit quality and fruit sales.

We would like to work together to ensure that the timing of burning does not cause any damage to grape crops.

238 Also at this time Albert Haak & Associates instructed solicitors to write on behalf of the plaintiffs to the Department in relation to the proposed burn.

239 Mr Keppel then sent a fax to Mr Walker in these terms:

Attached is a note from a vineyard manager who manages 3 vineyards adjacent to Hand Burn 8 ...

Hand Burn 8 is a burn of predominantly Karri forest in Warren National Park. This burn is constantly being monitored and considered for ignition. It is likely to be available within the next 2 weeks. It is a very high priority for us.

John Gillard will contact Jim Campbell-Clause and explain to him the Departments and Governments position; including that we intend to go ahead with the burn.

For your information.

240 On 3 February 2004, Mr Gillard met with Mr Campbell-Clause and Mr Pearse and discussed prescribed burn DPHB8. Mr Gillard reported to Mr Keppel that Mr Campbell-Clause and Mr Pearse were 'reasonably comfortable' with the Department's approach to minimise smoke over the vineyard, but remained very concerned about the potential impact on their crop'.

241 On 3 February 2004, Mr Walker drafted a Contentious Issue Briefing Note titled 'Smoke from Prescribed Burning - Wine Grape Quality' (Contentious Issue Briefing Note) and provided it to Mr McNamara, the Executive Director of the Department, for submission to the Minister for the Environment. The briefing note attached the plaintiffs' correspondence of earlier that day in relation to DPHB8.

242 Later that day, Mr McNamara signed the Contentious Issue Briefing Note and submitted it to the Minister for the Environment, the Hon Judy Edwards.

243 The Contentious Issue Briefing Note including the following:

Current status

The Department's current position on this issue for public communication is summarised below:

1. The Government has strongly supported [the Department] to achieve its prescribed burning program. There is generally a very high degree of local community support for this prescribed burning program, and [the Department] has been provided with additional resources to achieve a higher level of prescribed burning in 2003-04. Therefore [the Department] will not be delaying the commencement of prescribed burns to accommodate the concerns of individual wine grape growers.
2. Where practical [the Department] will carry out its prescribed burns in a manner that minimises the amount of smoke (in quantity and duration) that is likely to be in the area where wine grapes are near to harvest.
3. [The Department] is aware of the 'Extract from Australian Wine Industry Report 2003'. The report focuses on the Eastern States 2003 wildfire situation that resulted in wine grapes being exposed to smoke and ash for a long period of time when compared to the short duration of carrying out a prescribed burn. [The Department] has communicated with the Wine Industry Research Institute and will work collaboratively with the wine industry at a peak body level to address the issue.
4. The practice of prescribed burning has been in place on [Department]-managed lands for 40 years (including before vineyards were established) and will continue as an important strategy as part of the Department's and Government's responsibility to protect neighbouring community and conservation values.
5. The Department's position is that it will consult neighbours but the commencement and carrying out of prescribed burns will not be delayed at the request of individual wine grape growers. All prescribed burns will be commenced and carried out when the field conditions are suitable for the burn and resources are available.

• Will the Government request [the Department] to defer burns that may cause degradation to wine grapes?

Response: Not at this time. The Government supports [the Department's] achievement of the prescribed burning program, noting that [the Department] is willing to collaborate with the wine growing industry to conduct research into this issue.

• Will the Government compensate wine growers where grapes are affected by smoke from prescribed burns?

Response: No. Prescribed burning has been a part of land management in the region for four decades. Landowners need to recognise that there is an infrequent potential for some impacts to be caused by smoke from prescribed burning and manage accordingly.

244 On 3 February 2004, Mr McNamara also chaired the Department's Corporate Executive Meeting No 405 (comprising the Executive Director of the Department, seven Directors of the Department, the Executive Officer of the Department, and the Fire Management Services Branch Manager). The autumn 2004 burn programme was endorsed by the Corporate Executive in principle, subject to Directors providing comments on contentious burns. None were received.

245 On 6 February 2004, the plaintiffs' solicitors wrote to the Department advising that the plaintiffs would hold the Department liable if the burn of DPHB8 proceeded and the plaintiffs' grapes were damaged.

246 On 7 February 2004, the Minister for the Environment approved the Department's approach as set out in the Contentious Issue Briefing Note, endorsed her approval on the note and referred it to the Department for action.

247 On 9 February 2004, Mr Keppel forwarded to Mr Walker the plaintiffs' solicitors letter dated 6 February 2004.

248 On 12 February 2004 the approved Contentious Issues Briefing Note was given operational effect by Mr Walker, by him sending the note endorsed by the Minister to relevant fire managers. This meant that DPHB8 could go onto the daily burn approval process and proceed to burning if weather conditions were suitable. On 13 February 2004, Mr Walker provided Mr McNamara with a copy of the Contentious Issue Briefing Note. Mr McNamara noted on the Contentious Issue Briefing Note that the Minister's approval had been given operational effect by Mr Walker on 12 February 2004.

249 On 19 February 2004, Mr Sneeuwjagt sent a letter to the Department of Agriculture seeking the Department's assistance to map existing vineyards in the Pemberton area.

250 On 26 February 2004, Mr Walker replied to the plaintiffs' solicitors in these terms:

In response to your letter of 6 February 2004 to the District Manager of the Department's Donnelly District, I wish to advise the following:

• Prescribed burning has been an integral part of land management of [Department]-managed lands in the Pemberton area for four decades. Prescribed burning in karri forest vegetation occurs mostly during the months of December to April. Landowners in the Pemberton area would be very well aware of this longstanding land management activity.
• Prescribed burning is an important element of the Department's Fire Management Policy. This policy is strongly supported by the Government.
• Prescribed burning is strongly supported by local rural communities and most landowners in the Pemberton area express strong support for the continuation, and indeed often the expansion, of the program undertaken annually by [the Department].
• As the District Manager has stated in his discussions with Mr Campbell-Clause, [the Department] will endeavour to minimise any impact to neighbours caused by prescribed burning. The program will, however, go ahead as scheduled, subject to suitable weather conditions.
• In regard to the possible effects of smoke on wine grapes, [the Department] has indicated that it will cooperate fully with the wine industry in any scientifically conducted research into this issue.

251 At around this time the Department received a complaint from another winery regarding prescribed burning.

252 Mr Sneeuwjagt became a member of the Western Australian Wine Industry's smoke taint working group, and, with Mr Walker, encouraged research to be undertaken on the issue of smoke taint.

253 In 2005 - 2006 the Department was involved in sponsoring research into the question of smoke taint in grapes. The research was conducted as part of a doctorate programme under the supervision of Associate Professor Gibberd. Professor Gibberd prepared two reports for the purpose of these proceedings, referred to below. The research has been time-consuming. Its nature was described by Mr Walker:

What's the nature of this research like? Is it something that can be done in a year, for example?---No, I'd have to describe it as frustratingly slow because the first elements of the research involved what you would call laboratory trials rather than field trials in that a hothouse-type environment was set up and grapes were exposed to the burning of straw, so it's a long way from there to the burning of eucalypt fuels and the sorts of conditions that would occur in a field situation as opposed to a very controlled experiment. I understand why the research has to start at that point, but in terms of getting the answers that Mr Armstrong has indicated we need it would be my view that we are still quite some way off having those answers (ts 857).

254 There was expert evidence adduced by the defendants in relation to the effects of smoke on grapes. The evidence comprised two reports from Associate Professor Gibberd which were tendered by consent. The second report in effect updated the first report. There was also a report of the pre-trial conferral held by the parties' 'wine' experts, which was tendered.

255 The substance of Associate Professor Gibberd's evidence, which I accept, was, relevantly, as follows.

256 In 2003 bushfires in Victoria and New South Wales stimulated concern over the potential of smoke to contaminate grapes. The AWRI documented a series of small studies in their 2003 annual report.

257 Whilst the AWRI Annual Report was released prior to the 2004 vintage, the timing was close, and the level of knowledge on cause and effect linking smoke exposure to smoke taint was scant.

258 No subsequent significant research was undertaken until research activities in Western Australia commenced in late 2005. The results of the ongoing WA-based studies were presented at the Australian Wine Industry Technical Conference in 2007 and the first peer review article was accepted for publication in November 2007. The level of information available in 2004 was very much less than the current level of knowledge.

259 Studies in 2007 and 2008 have shown that smoke exposure of both field-grown grapevines and excised grape berries has the capacity to result in smoke taint of the resultant wine following vinification.

260 Smoke taint-related compounds can be taken up by both the fruit and the vegetative components of the grapevines. If such compounds are taken up by the vegetative components they can be translocated to the developing fruit.

261 There are three apparent phases of risk of smoke damage during the grapevine annual growth cycle. Phase 1 represents a low risk of smoke taint development and occurs between growth stages 12 (shoots 10 cm) and stage 23 (flowering). For practical purposes this phase could be extended to include the vine dormancy period and early shoot emergence. Phase 2 represents a period of increased and variable risk. This period extends from growth stage 31 (berries pea size) to stage 35 (three days post veraison). Phase 3 represents a period of consistently high risk of smoke taint development and extends from growth stage 35 (seven days post-veraison) through to harvest.

262 Regarding the question of what density, type and duration of smoke is required to contaminate wine grapes, the information required definitively to answer that question is still somewhat lacking and is the subject of further research. Most experiments have utilised comparatively high levels of smoke exposure for short periods of time. It appears that any smoke-exposure events following flowering have the potential to contribute to the uptake of smoke taint-related compounds in grapevines. Smoke exposures that occur prior to flowering are also likely to result in uptake by grapevine leaves. However, the taint-related compounds taken up in early exposures are either not translocated to the fruit that later develops or they are diluted by shoot and fruit growth to the point whereby they do not have a measurable influence on the subsequent wine quality. Importantly, smoke exposures are additive and it is therefore reasonable to assume that even small (dilute or short period) exposures may contribute to the final taint observed in wine.

263 There does not appear to be a practical method to prevent smoke uptake by grapevines. Where smoke taint is found to exist in wine there is some potential to reduce the concentration of smoke-derived volatile phenols with reverse osmosis and solid phase adsorption and such a reduction may result in an increase in the consumer acceptance of the wine.

264 The first instance of putative smoke impact on grape quality at one of the vineyards was noted in 2001. Although the suggested link between cause and effect was not immediately apparent, by 2004 there was a degree of awareness of smoke taint within the wine industry due to the impact of the 2003 bushfires in Victoria and New South Wales on the wine industry.

265 The report from the pre-trial conference, which I also accept, indicated that the parties' 'wine' experts agreed:

There was awareness of the risk of smoke contamination of wine grapes at the time of the prescribed burn but the level of knowledge was considerably less than that available today ([the] paragraphs ... below are based on the state of knowledge applicable at the time of the expert conference). There was information from Victoria following the smoke contamination of wine grapes that occurred as a result of smoke from the large wildfires (not prescribed burns) that occurred in 2003.

Smoke from both bushfires and prescribed burns have potential to contaminate wine grapes and the wine made from those grapes.

The process of smoke contamination is absorption of smoke compounds directly into grape berries and into other plant components, particularly leaves followed by translocation to the berries via the phloem of the grapevines.

Information is lacking as to the density, type and duration of smoke required to contaminate wine grapes. It is likely that any smoke exposure event has the potential to contribute to the uptake of smoke taint-related compounds however the level of contribution will be dependent on undefined environmental or physiological constraints.

The evidence as to the stage in the growing season at which wine grapes are capable of being contaminated is variable. Based on information available, the influence of smoke exposure is greatest between veraison and harvest but it is not possible to set a particular time of peak sensitivity within this period. Uptake and translocation are therefore likely to be influenced by factors other than, or in addition to, grapevine phenology (for example, the environmental conditions at the time of exposure).

It is impossible to avoid the uptake of smoke by the grapevine, leaf or berries during the period of smoke exposure. The uptake of smoke tainted compounds by the grapevine is very rapid.

The effect of smoke exposure events separated by time on the level of smoke taint in resultant wine is additive.

Compared with aerosol smoke, ash and burning debris are not significant contributors in terms of smoke taint of wine grapes. Ash has minimal odourific content and studies with respect to pre-washed fruit to remove ash have not demonstrated a sizeable reduction in smoke taint. (Emphasis added)

DPHB8 within the burn programme - up to 30 March 2004

266 DPHB8 was a high priority burn because of its significance in protecting the Warren National Park and the Pemberton community.

267 DPHB8 had last been burnt in 1988 - 1989. It was carrying fuel loads of 73 - 93 tonnes per hectare when the limit at which most wildfires can be contained in karri forest is 15 - 19 tonnes per hectare.

268 DPHB8 also had strategic value. Mr Gillard, in his supplementary statement, explained the strategic importance as follows:

1. Once DPHB8 was completed this provided the Department with an opportunity to utilise the areas of low fuel to safely burn the area to the south of DPHB8.
2. This area is known as DP209 and contains areas that have not been burnt for 12, 20 and 25 years. The prescribed fire plan for DP209 was prepared in 2005 and burn boundary preparation works completed in 2006.
3. To date the Department has not received suitable conditions to complete this burn and it remains on the current burn program.
4. Without low fuel areas adjoining this area it would be more difficult to burn and there would be a greater risk of fire escapes ...
5. If individual burns are delayed or postponed it can have a significant impact on the overall burn program, with the opportunity to utilise the benefits from adjoining low fuel areas being reduced. This can lead to individual burns having to be undertaken against higher fuel loads. This requires increased resources and increases the risk of escape.

269 It was put to Mr Keppel in cross-examination (ts 567), in effect, that because there was a strip of forest immediately south of Old Vasse Road which had been burnt 10 - 12 years previously, it was, in effect, unnecessary for the Department to be concerned about the risk of wildfire to the south of DPHB8 as the strip would provide a protective barrier for the forest to the south. Mr Keppel's evidence, which I accept, was that the strip of karri forest with litter 10 years old, would not provide a significant advantage to stop a severe wildfire to the south, which had fuel loads of 25 years and more.

270 DPHB8 was put on to the six-year indicative burn plan in the 1998 - 1999 fire season, based on Wildlife Threat Analysis principles, fuel ages and the time since last burn. DPHB8 was strategically important to protect life and property values and assets on adjoining private property, visitors and Departmental infrastructure in the Warren National Park and regrowth forest south of the Warren National Park. The Warren National Park south of the burn area had high fuel levels with some areas more than 25 years old and there was a significant risk that a fire escape across the southern boundary could result in an uncontrollable wildfire. Such a fire would threaten conservation values, visitor facilities and visitor safety in the Warren National Park.

271 In 1999, DPHB8 burn prescription preparation commenced. This included doing a field inspection to confirm the burn boundary; measuring and recording fuel loadings on the Fuel Assessment Records; calculating the available fuel weight; completing the Fuel Assessment Summary; and calculating acceptable scorch height, actual rate of spread and equivalent Fire Danger Index from the Forest Fire Behaviour Tables for Western Australia.

272 This information was compiled by the then Fire Operations Officer, the late Denis Marshall. Mr Marshall also completed the Pre-Burn Checklist and developed the Objectives and Standards for the burn.

273 In 2000, the DPHB8 burn prescription was approved for entry into the Annual Burn Programme. The initial burn prescription prepared by Mr Marshall was checked, reviewed and approved for implementation.

274 The District Fire Coordinator then included DPHB8 into the Annual Burn Programme for the District. Burn notification letters for DPHB8 were sent to neighbours on 1 January 2001 and 31 January 2001.

275 The burn boundary preparation works were programmed to be completed in summer or early autumn 2001. The boundary preparation works were, however, delayed and therefore there was no opportunity to safely undertake the burn that season.

276 DPHB8 was then carried over into the 2001 - 2002 burn programme. Mr Marshall reviewed and updated the burn prescription in spring 2001.

277 Mr Gillard reviewed the burn prescription and approved it in December 2001.

278 Under the approved burn prescription for DPHB8, the objectives of the prescribed burn were:

(a) To provide protection against wildfire damage to adjoining private property, the mature forest within the National Park, park facilities and park visitors by creation of fuel reduced areas.

(b) To protect and maintain biological diversity processes and ecological values within the burn area by the application of a low intensity, slow moving prescribed burn that will provide a mosaic of burnt (70% - 90%) and unburnt (10% - 30%) patches.

279 In the core lighting strategy of the burn prescription, it stated 'winds to be from a southerly direction'.

280 Burn notification letters were sent to neighbours in early 2002 and DPHB8 was included in the Warren Region Operations List published in the Manjimup Bridgetown Times.

281 In 2001 - 2002, there were limited opportunities to safely complete burns including DPHB8, with the Donnelly District only completing around 10,200 ha of prescribed burning including around 4,000 ha of silvicultural burns. In that year the Donnelly District was also required to manage 12 wildfires totalling around 7,960 ha.

282 DPHB8 was then carried over into the 2002 - 2003 burn programme. Again, burn notification letters were sent to neighbours and DPHB8 was included in the monthly Warren Region Operations List published in the Manjimup Bridgetown Times.

283 There were no opportunities to safely undertake the burn in the 2002 - 2003 season as Western Australia experienced a severe fire season across the state. The Department had to manage 656 fires burning more than 2 million ha. The Donnelly District managed 51 fires that burnt more than 5,700 ha. Twenty-four of these fires were caused by lightning. This required a significant commitment from the Department's resources and the Donnelly District was only able to achieve around 20,300 ha of prescribed burning, including around 4,500 ha of silvicultural burns.

284 DPHB8 was carried over into the 2003 - 2004 burn programme with increased fuel loads. This raised the priority for the Department to complete DPHB8 due to the increasing fuel loads and wildfire suppression difficulty and the threat this posed to the community and values within and adjoining the burn.

285 In October 2003, the Department conducted community consultation meetings in Pemberton, Manjimup and Northcliffe to discuss the Donnelly District prescribed burning programme.

286 Burn notification letters, including to the plaintiffs, were sent out on 19 November 2003, 17 December 2003, 19 January 2004, 11 February 2004, 5 March 2004 and 26 March 2004. DPHB8 was included in the Monthly Warren Region Operations List published in the Manjimup Bridgetown Times in December 2003, January 2004, February 2004, March 2004 and April 2004.

287 The burn prescription was amended by Mr Gillard's note of 3 February 2004 referred to in [236] above. It was put to Mr Gillard, in effect, in cross-examination, that his note required that the burn of DPHB8 be commenced and completed on east-south-east to easterly winds, and that the Department had acted contrary to the note by commencing the edging burn on more southerly winds (ts 497 - 498). Mr Gillard said, and I accept, that the focus of his note was on the core ignition, and not the edging burn. That is consistent with the plan to conduct the burn in two days, with the core ignition 'c_ompleting_' the ignition of DPHB8 on the second day.

288 In February 2004 the Department continued to monitor the weather and fuel moisture conditions for DPHB8. On 16 February 2004 a significant rainfall event occurred which precluded the possibility of an effective burn of DPHB8 around that time. There was a further 5.2 mm of rain at Pemberton on 8 March 2004.

289 Throughout March 2004 the Department's District Duty Officer also continued to monitor the fuel moisture and forecast weather conditions in relation to the prescribed conditions for prescribed burns DP037, DP09 and DP14. These burns contained similar forest types to DPHB8 and required similar weather and fuel moisture conditions safely to conduct the burns. DP037 also contained some areas of jarrah forest that would be available for burning earlier than the karri types. DP09 and DP14 were located to the east and south-east of Northcliffe.

290 On 24 March 2004 prescribed burn DP037 was undertaken in karri forest approximately 3 km south-east of DPHB8. The burn was completed effectively and within its prescription parameters. The Department's District Officers concluded that areas for other proposed karri burns should be regularly field checked as they would likely become available to burn in the near future.

291 Around this time Mr Keppel and Mr Simmonds conducted a field inspection of the DPHB8 burn area. Mr Keppel said:

1. ... I formed the view that prescribed burn DPHB8 could best be carried out to avoid the risk of escape with southerly winds as initially prescribed but with an easterly component to minimise smoke over the vineyard. It was clear that we could only minimise smoke over the vineyard. Avoiding smoke over the vineyard was not possible considering the proximity of the open farmland sloping downhill immediately adjacent to tall karri forest and the relevant diurnal weather cycle that occurs in the south-west during the prescribed burning weather conditions in autumn including low overnight inversions.
2. I was of the view that it was not safe to attempt to burn the area with north to north west winds as the risk of ignition from embers being blown out of the burn area and into dry heavy forest fuels in the adjoining forested landscape in the Warren National Park was too high. This was [a] fundamental requirement. If you carry out a burn and push it over heavy fuels the risk of escape is increased significantly.
3. In terms of the ignition plan, I was of the view that it would be important to edge the burn along the vineyard and nearby farmland on day one with the core ignition by helicopter to be carried out on day two and sufficient resources would be required to achieve this.

292 These views accorded with the views of Mr Gillard, who had prepared the 3 February 2004 amendment to the burn prescription (see [293] below).

293 On 30 March 2004, Mr Bennett and Mr Marshall discussed the proposed burn of DPHB8 with Mr Gillard. Mr Bennett and Mr Marshall advised that the forecast weather pattern would most likely be suitable to initiate the burn over the next few days, with winds initially in the south-east and then becoming north-east and east. It was agreed that this weather pattern would provide the best opportunity to safely complete the burn and to minimise smoke over the plaintiffs' vineyards. They also discussed the resources needed to complete sufficient edging on the northern and western boundaries on 31 March, and to allow the core ignition to be completed on 1 April. They agreed that the minimum additional resources required were two extra field staff, one extra fire truck and one extra bulldozer. (These additional resources were in fact allocated to the burn on 31 March 2004.)

294 In the afternoon of 30 March 2004, Mr Gillard telephoned Mr Keppel, the Regional Manager, and discussed the four day forecast and field conditions. They discussed the potential for smoke to drift over the vineyard during the edging, and the opportunity provided on 1 April to complete the core ignition while minimising the amount of drift smoke over the vineyard. They also agreed that the Director of Regional Services' endorsement should be obtained to put DPHB8 on the burn programme for Departmental approval on 31 March 2004. The Director's approval was sought and obtained, and Mr Keppel confirmed that the District Duty Officer could request approval to commence the burn of DPHB8 on 31 March 2004. The request was to be considered as part of the Departmental daily burn approval process for 31 March, following receipt of the 0750 weather forecast for that morning.

295 Around 31 March 2004 Mr Maher, as State Burn Coordinator, received a personal briefing from the Bureau's Duty Forecaster. Although Mr Maher did not recall the details of this, both Mr Gillard (statement par 221, ts 523) and Mr Keppel (statement par 46, ts 540) said, and I accept, that they were informed that the Bureau's Duty Forecaster had advised, in effect, that the weather system was moving through quite quickly and that north to north-west winds (and hence, they inferred, severe fire weather) could be expected in the next three to four days.

296 Mr Gillard asked Mr Marshall to make contact with the plaintiffs' vineyard managers in order to advise them that conditions were suitable to commence the burn of DPHB8, and that the District would be seeking Departmental approval to commence the burn on 31 March.

297 On the morning of 31 March 2004 Mr Marshall carried out a review of proposed prescribed burns, including DPHB8. This included an overview of planned burns over the next four to ten-day period, and gathering the necessary information to assess the prescribed fuel moisture and forward rates of headfire spread.

298 There was a meeting at the Department's Pemberton Office at about 7.50 am on 31 March 2004, attended by Mr Marshall, Mr Gillard, Mr Bennett and Mr McKenzie of the Department. They received the forecast issued at 7.50 am on that morning and noted the possibility of morning drizzle, and that south-south-east winds had been forecast for the day. They also noted that the outlook for the next day, 1 April, was for north-easterly winds in the morning, and variable in the afternoon. Mr Marshall advised that he had tried to contact the neighbours adjacent to DPHB8, including the plaintiffs' vineyard managers, the previous evening without success.

299 Mr Marshall prepared a Burn Request Form in respect of DPHB8 which was sent to Mr Keppel, the Regional Duty Officer. Mr Keppel sent the Regional Burn Requests for the day, including the request for DPHB8, to Fire Management Services. The burn requests were discussed at the telephone conference with Fire Management Services at 8.30 am. At this telephone conference Mr Maher endorsed the proposal to burn DPHB8, and faxed the approved form to Mr Sneeuwjagt, the State Duty Officer, for his approval.

300 Mr Sneeuwjagt viewed the smoke trajectory model from the Bureau on the morning of 31 March 2004, which indicated that smoke emanating from a location near Pemberton would be blown slightly west of north-west. This confirmed, in his mind, that the bulk of the smoke from the edge burn on 31 March should be blown away from the plaintiffs' vineyards, located north-east of DPHB8. At about 9.00 am, Mr Sneeuwjagt faxed his approval to the Regional level to start burn DPHB8 that day.

301 Mr Bennett attempted to contact the vineyard by phone without success. Mr Gillard asked the National Park Ranger, Yvette Caruso, to make face to face contact with all the neighbours including the plaintiffs, but her commitments were such that she was unable to do so before 1.00 pm.

302 On 31 March 2004, at 8.30 am, Mr Barry was notified by his neighbour that the Department was going to conduct the prescribed burn that morning. Mr Barry told his neighbour that he was not surprised, as he had seen a letter giving notification of the burn, although the date had not been specified.

303 At 10.00 am Mr Marshall reviewed the 10.00 am forecast to see if there were any forecast changes. There were none. Mr Marshall also estimated fuel and burn parameters and checked those against the burn prescription. Mr Marshall also briefed Mr McKenzie, one of the Department's experienced prescribed burn and wildfire suppression officers. Mr McKenzie was the Operations Officer for the burn.

304 Mr Marshall subsequently gave Mr McKenzie radio approval to start the burn.

305 In addition to Mr McKenzie, there were four section commanders responsible for each of the four designated sections of the burn, a liaison officer, six fire crews and a bulldozer. Each crew comprised an overseer, a gang truck with at least three operators, and a heavy duty truck with two operators. Two of the fire crews were taken from Blackwood District to assist with the burn.

306 Mr McKenzie briefed the operators on site at the Old Vasse Road at about 10.15 am. He outlined the objectives, the sectors into which the burn area would be broken, the allocation of crew and personnel to sectors, and the proposed conduct of the burn over the next two days.

307 The burning of edges near the plaintiffs' vineyards was commenced late morning and prior to lunch.

308 The edge burning of the northern boundary of the burn (which included the area next to the plaintiffs' vineyards), the western boundary and the western third of the southern boundary continued as planned throughout the remainder of 31 March 2004.

309 Mr Barry spoke to officers of the Department in the late morning and around lunch-time, indicating his concern over the amount of smoke around and over the vineyards. He asked why the burn was being conducted at that time, and was told that the winds were favourable for the burn. Mr Barry also saw Ms Caruso of the Department in the afternoon. She provided Mr Barry with information on the burn and gave Mr Barry a copy of the Department's weather report and forecast for that day.

310 There was a hop-over adjacent to the power line in the north-west corner of the burn area, which had to be, and was, dealt with.

311 Mr Marshall was kept informed of the progress of the burn by radio and telephone by Mr McKenzie. Mr Marshall recorded information in a Fire Diary for the burn.

312 On the afternoon of 31 March 2004, Mr Marshall and Mr Gillard also discussed resources for the following day.

313 The edge burning on 31 March 2004 burnt out approximately 120 ha, of which 20 ha covered the area adjacent to the south and east of the plaintiffs' vineyards.

314 It was safe to leave the burn by about 7.00 pm, and crews left the burn area between 7.00 pm and 8.00 pm.

315 On the morning of 1 April 2004, Mr McKenzie and Mr Bennett reported to Mr Marshall from the field and advised that there had been no problems with the prescribed burn. The information was relayed by Mr Marshall to Mr Gillard and Mr Keppel. Mr Keppel was also advised that there was smoke haze over the plaintiffs' vineyard due to the low inversion overnight. Mr Keppel considered, based on the forecast, and from his experience, that the inversion would lift and that the smoke haze would consequently clear, and that the winds, if they remained as forecast, would blow the smoke away from the vineyards.

316 Mr Marshall requested approval from Mr Keppel to complete the core ignition of the prescribed burn.

317 Ms Michelle Widmer of the Department, acting for Mr Maher in his absence, hosted the 8.30 am telephone conference. Mr Keppel advised the participants in the telephone conference that the situation with DPHB8 remained stable overnight. Mr Keppel requested a core ignition by helicopter to commence from 11.00 am onwards. Mr Sneeuwjagt signed the form to approve the burn programme for 1 April 2004. Mr Sneeuwjagt's approval was faxed to the Warren Region and Donnelly District.

318 At 11.00 am, the core ignition of the burn commenced by aerial ignition. The core ignition, and the edge burning on the eastern and remaining part of the southern boundary, were successfully completed by that evening.

319 In the afternoon, Mr Campbell-Clause drove to the Department's Donnelly District Office and saw Mr Gillard. The parties to that discussion gave somewhat conflicting accounts of it at trial. I find that the substance of the discussion at the meeting was recorded in an email sent by Mr Gillard to Mr Keppel following the meeting:

Jim Campbell-Clause called in to see me this afternoon concerned that we did not notify them sooner that we were burning DPHB8. He wasn't happy with confirmation 'on the day'. I explained what had been happening with the drying in the Karri forest and the uncertainty on the timing of the burn. I also explained that we only received a small amount of drizzle out of the last change and until a field check on Tuesday afternoon we could not confirm the timing of the burn. Once a decision was made late on Tuesday we then went through our normal notification process. This included a field visit to all adjoining landowners on the day of the burn. He was not happy with this and had been hoping for a week or so notice [not sure where he had this idea from - it wasn't from us]. He was also very concerned about smoke through the vineyard this afternoon and wanted to register his concerns. I was unable to ascertain what he intends to do. He said he would need to get the winemaker to check the grapes and go from there.

320 By 2 April 2004, the majority of the area of DPHB8 had burnt out. Ms Caruso was the Operations Officer for the burn in the mop-up phase on 2 April. She said that the mop-up phase involved checking the burn boundary to a depth up to 100 m to extinguish any hot fuels or hollow butts in preparation to reach black-out phase. There were essentially no flames during the day, and there was no approach by any neighbouring property owners.

321 Mopping-up operations were completed on 4 April 2004.

322 Mr Barry took a number of photographs of the smoke from the edging burn near the plaintiffs' vineyards on 31 March 2004. The photographs show varying degrees of smoke thickness at or near the plaintiffs' vineyards. Mr Maher (ts 778 - 779) and Mr Sneeuwjagt (ts 946) described the smoke depicted in the photographs as 'light', by which I understood their evidence to mean light in comparison with smoke from a core burn. They also said that the smoke described was the kind of smoke that they would have expected from an edging burn. I accept their evidence. Mr Barry, for the plaintiffs, also gave evidence (ts 248) that the smoke from the core ignition was 'a lot worse' than the smoke from the edging burn.

323 Mr Barry gave evidence in relation to the movement of the smoke in the period 31 March to 5 April 2004. He had prepared a contemporaneous note of this at the time. Having regard to his evidence, I find:

• On 31 March there was smoke over the vineyard by about 12.30 pm, and it stayed over the vineyard for the rest of the day. The smoke at night was not as heavy.
• On 1 April the smoke had blown off the property in the morning by the north-east wind, but returned at about midday, and stayed over the property, and was fairly dense.
• By 2 April the smoke had moved off the vineyards, but returned at about 3.30 pm. Around this time everything was covered in ash.
• On 3 April the sky was fairly clear of smoke, until about 4 pm when smoke again covered the vineyards.
• On 4 April there was smoke over the vineyards, but it was less dense.
• On 5 April there was again smoke over the vineyards, with about the same density as the smoke level on 4 April.

324 The defendants tendered a report of Dr Li Shu, a research scientist with the Department. He reported on satellite imagery of smoke plume location in relation to the burn of DPHB8. Although his report was tendered by consent, and without cross-examination, I prefer the evidence of Mr Barry on the location and extent of the smoke. Dr Shu's report contains much technical discussion which, without elaboration, is difficult to follow. This makes it difficult properly to assess the cogency of his conclusions. I also infer from his report that at times of cloud-cover, the satellite could not detect whether there was smoke over the vineyard. In any event, as best I understand it, his report on smoke presence is not materially different from Mr Barry's evidence, save perhaps in relation to the afternoon of 1 April.

325 The passage of smoke over the vineyards reflects the potential for variance between forecast winds and actual conditions. Although the winds for 1 April to 3 April were forecast to be to the north-east and east (see [420] below), on each of those days, as the evidence from Mr Reader of the Bureau indicated, the actual winds had a southerly component, particularly in the afternoon.

The window of opportunity for prescribed burning in the Warren Region

326 There are limited opportunities in any year safely and effectively to conduct prescribed burns, particularly of karri forest. Dr McCaw summarised the general position as follows:

The window of opportunity for prescribed burning in south-west forests is influenced by:

• the nature of the forest, in particular the density and species composition of the overstorey and understorey layers;
• location, including aspect, elevation, soil type and proximity to the coast;
• prevailing patterns of weather during the fire season.

In general terms, weather and fuel conditions suitable for prescribed burning occur earliest in northern and eastern parts of the jarrah ... forest, as early as September in some years.

In the Warren Region, the prescribed burning season commences in mid-October in most years when fuels in open jarrah forest and coastal shrubland dry to a level where effective ignition and fuel consumption can be achieved.

The window of opportunity for prescribed burning in karri forest commences from late November to mid December in most years, and may extend into April in years where autumn rains are late. This window may be considerably reduced if spring rains persist into December or autumn rains commence in late March. The window of suitable burning conditions is also shorter in areas subject to the influence of coastal showers and milder temperatures.

...

After mid April the reducing day length, lower angle of the sun and increasing frequency of overnight dew restrict the extent of fuel drying possible beneath a dense forest canopy. This means that opportunities for burning become quite limited even in the absence of rain. ... effective prescribed burning [cannot] be conducted in karri forest between May and October, unless the forest has been modified by timber harvesting, partial clearing or removal of the understorey shrub layer.

327 In relation to forest with K1 and K2 understorey, the forest will not usually be dry enough to burn until late December to mid-January (see [340], [344], [362] below).

328 In the report of the pre-trial conferral, the experts all agreed and I accept that, generally, karri forests are too wet to burn in April and May.

329 The Department uses a publication, developed by Mr Sneeuwjagt and Mr Peet, entitled 'The Forest Fire Behaviour Tables for Western Australia' to assist in predicting burning opportunities in the karri forest. The publication is commonly described as the 'Red Book'.

330 The 'fire' experts in these proceedings regarded the Red Book as the best predictive tool for determining suitable prescribed burning conditions in the karri forest. Despite this, one of the plaintiffs' experts, Dr Tolhurst, provided evidence using another publication - McArthur's Control Burning Guide (1962) - to identify windows of opportunity for burning in the karri forest.

331 Evidence, which I accept, from the report from the pre-trial experts' conferral, recorded:

(a) The experts agreed that McArthur's Control Burning Guide (1962) was developed using experimental burns in northern jarrah forest around Dwellingup in WA as well as in dry eucalypt forests in NSW and ACT.

(b) The experts agreed that the Red Book was a result of further research, development and improvement (by Peet, Sneeuwjagt, Burrows, McCaw and others) following on from McArthur's early work.

(c) The experts agreed that the Red Book includes fuel drying tables and fire behaviour tables specifically designed for the karri forest in south-west Western Australia.

(d) The experts agreed that the Red Book would be the best predictive tool for determining suitable prescribed burning conditions in karri forest.

(e) The experts agreed that McArthur's Control Burning Guide (1962) will indicate fine fuel moisture contents drier than those that actually occurring in a mixed-karri forest.

332 The Red Book is used, inter alia, to predict the moisture content of fuel litter.

333 The following findings are derived from Mr Sneeuwjagt's evidence, which I accept.

334 The moisture content of the surface litter and the near-surface fine fuels influences the ignitibility and the rate of fire spread. The Department's research into karri forest fuel moisture and fire behaviour found that the surface fuel moisture content needs to be within the range of 10% to 18% (oven dry weight) to ensure good ignition and mild to moderate fire behaviour (Sneeuwjagt and Peet, 1976). Surface moisture content greater than 18% results in poor ignition and a poor burn result that will require a subsequent ignition to reduce the risk of costly escapes as the weather warms up.

335 Surface moisture content of less than 10% can lead to excessive fire behaviour that can result in fire escapes and damage to the timber and other natural values.

336 Karri litter fuels are relatively deep and the moisture content increases with depth. The amount of litter that is consumed in a burn depends on the 'Profile Moisture Content'. If the Profile Moisture Content is too wet (greater than 60%) the burn will only consume the surface layers of the litter bed, and the remaining unburnt litter may re-ignite in the following days or weeks and result in a high intensity fire that may escape and burn out neighbouring properties.

337 If the Profile Moisture Content is less than 25%, the entire litter profile is available to burn, and in heavy karri fuels this will lead to a high intensity burn that may damage or kill karri trees, and lead to possible escapes.

338 For a fuel reduction burn to be safe and effective, the proportion of the fuel bed that is available to burn needs to be between 30% to 70% of the litter. This proportion, known as the Available Fuel Factor (AFF) occurs through the following combinations of Surface Moisture Content and Profile Moisture Content:

• Surface moisture content 10% - 18%
• Profile Moisture Content 30% - 60%

339 There are only a limited number of days each year when this combination of Surface Moisture Content and Profile Moisture Content occurs in conjunction with the mild conditions (wind, air temperature, relative humidity, atmospheric stability) that are required to conduct an effective and safe fuel-reduction burn.

340 The Department's research into fuel moisture trends demonstrates that in most years, the karri fuels do not dry out sufficiently to achieve a satisfactory fuel-reduction outcome until early to mid-December. The fuels become too dry in late January to mid-March, before they again reach satisfactory levels after light rainfall in early Autumn (late March/early April). After mid to late April, the karri fuels generally become too moist to burn, due to either rainfall or overnight dew.

341 In most years the number of days in a year with suitable fuel moisture (Surface Moisture Content, Profile Moisture Content) within mixed karri forest in the Pemberton area is less than twelve, and can be as few as two days.

342 The method for establishing Surface Moisture Content involves obtaining representative values for Surface Moisture Content for use in headfire rate of spread prediction. Representative samples are first placed in a grinder and the percentage of Surface Moisture Content is determined on the ground material using an electronic moisture metre.

343 The method for establishing Profile Moisture Content involves obtaining a representative value for profile or the deeper fuels to 'mineral earth' using oven dry weight techniques. Samples are collected in sealed tins and weighed, and the samples and tins dried in an oven at 105 degrees centigrade for 12 - 14 hours. Profile Moisture Content figures are then calculated. The Profile Moisture Content represents the moisture content of the entire litter profile expressed as a percentage of the oven dry weight of the sample.

Fuel drying regimes for different forest/vegetation types

344 The rate at which the surface fuels dry out depends on their exposure to sunlight and wind. Open forests such as the jarrah/wandoo dry out most rapidly, whilst dense karri forest types dry very slowly. The deep karri fuels, particularly in lower slopes and gullies, do not dry out until the summer months of December and even into January.

345 The dense forest cover means that the ground fuels are slow to dry after significant rainfall. This is particularly the case in autumn months when the relatively long nights and short days prevent the deep profile litter fuels from drying after rains in April to May.

346 The southern forests consist of a variety of forest types that dry at different rates. In order to achieve an effective burn coverage, a number of ignitions may need to be undertaken as each fuel type becomes available. The period between the first ignition in the open jarrah fuel types, and the last ignition in the dense, moist karri fuel types may be over several drying cycles. This requirement for a series of suitably dry days further restricts the window of opportunity for a safe and effective burn.

347 The fuel moisture levels are only one of several indicators of burning opportunities and wind speed and fire danger levels are also important constraints.

348 The wind speed and the surface fuel moisture content define the suitable fire behaviour required for a safe and effective fuel reduction burn. In most cases, the fire spread for prescribed burns is confined to a range of 15 to 40 m per hour. Faster fires are likely to be difficult to control and may result in an escape and damage to neighbouring properties. If fuel loads are high, as they usually are in karri forests, the fire intensity can result in high scorch and damage to trees.

349 Wind speeds recorded in the open (or above the tree canopy) should not exceed 20 km per hour. Lower wind speeds are required if surface fuels are dry and where a high proportion of the fuel bed is available to burn.

350 The fire behaviour limits are defined in the burn prescription by the Fire Danger Index. The fire danger index is calculated from the Surface Moisture Content, and the wind for each forest type. There are a number of combinations of wind speed and Surface Moisture Content that can result in the same fire danger index. For example, the lower the Surface Moisture Content, the lower the wind speed needs to be, and vice versa.

351 The wind direction is an important limiting factor as it affects the direction of the headfire and the embers that can 'spot over' a burn boundary.

352 The safest burns in the south-west forests are undertaken when the winds are from the south (SE to SW) as these are maritime air lows that have a relatively higher moisture, ie, humidity, content. The most dangerous winds are the north and north-west continental winds that are hot, dry and gusty.

353 The acceptable wind direction is also dictated by the location of adjacent heavy and flammable fuels. DPHB8 was located immediately north of very old heavy karri fuels in the Warren National Park. For this reason also, a burn prescription for DPHB8 on north to north-west winds would be inappropriate.

354 The safest wind directions for a burn of the area of DPHB8 are those that blow from the south, albeit with an easterly component.

355 An objective of prescribed burning under the Forest Management Plan is to minimise smoke impact on population centres and other sensitive areas.

356 The stable weather conditions that are normally suitable for safe fuel-reduction burning in the south-west forests are frequently the same that lead to smoke being blown by southerly winds into Perth, Bunbury and Mandurah. Smoke can be trapped below a blanket of warm air, called an inversion. Smoke can also be trapped under the night-time inversion that often develops during the cool nights in the spring and autumn burning periods.

357 The Department has, for over 20 years, managed smoke from prescribed burns, primarily with a view to preventing smoke and haze from settling on Perth and major regional centres. Other locations sensitive to smoke and ash, such as harnessed water bodies and local towns and communities, have also been taken into consideration in the management of smoke from prescribed fires.

358 Since 2005, a 'decision support system' has been established to reduce the likelihood of smoke impacting on vineyards that adjoin areas of planned burning. In relation to DPHB8, the Department also planned the burn with a view to minimising smoke over the plaintiffs' vineyards (see [236] above regarding Mr Gillard's note).

359 The smoke management strategies applied by the Department have been based on the following steps:

• identification of suitable meteorological conditions for rapid dispersal of smoke;
• maximising the use of winds that blow smoke away from Perth and other population centres;
• monitoring of atmospheric conditions and inversions;
• devising and planning burns so that large burn areas are separated from each other as much as possible; and
• reducing the total area to be burned to days of very stable atmospheric conditions.

360 There are a large number of variable factors influencing smoke accumulation that burn managers must take into consideration before deciding on whether or not to commence a burn. These factors include the long-range weather forecast which can affect burns requiring a sequence of ignitions on numerous days; overnight inversion conditions; smoke plume trajectory at varying altitudes; the locations of the burn relative to smoke sensitive areas; burn size; the burn type and fuel loading; burn duration; smoke volumes and concentrations; existing smoke levels from other burns or wildfires; and the resources available to ignite and control planned burns.

361 The Soil Dryness Index is a numerical value that reflects the dryness of soil, deep forest litter, logs and branches. The Soil Dryness Index ranges from 0, when soils are saturated, to 2000 when soils are bone-dry to a depth of 200 mm. The Soil Dryness Index rises with soil moisture loss from evapotranspiration and falls with moisture gain from effective rainfall. The Soil Dryness Index reflects the potential for logs, branches and tree bark to ignite and burn. The Forest Fire Behaviour Tables contains a table of recommended Soil Dryness Index limits for fire control operations in the forest. Burning outside of these recommended Soil Dryness Index limits increases the risk of fire escape and the cost of control and mop-up of the burn.

362 The dense karri fuel types, K1 and K2, will usually not be sufficiently dry until the Soil Dryness Index exceeds 800. The five-year average of the Soil Dryness Index levels recorded for the Pemberton office over the period 2003 to 2007 indicates that the Soil Dryness Index level of 800 is generally not exceeded until late December to mid January, and falls below 800 in April each year.

363 The Red Book current guidelines recommend that a fall of 400 from the summer maximum Soil Dryness Index will provide conditions that reduce the risk of excessive fire behaviour and escapes. A fall of that kind is typically expected to occur in autumn. However, in deciding whether to burn, this recommended feature in relation to Soil Dryness Index needs to be balanced against the prospect that, in autumn, the rainfall needed to reduce the Soil Dryness Index by 400 may saturate the karri litter so much that it will not dry out sufficiently to burn. The short daytime drying period during the autumn months further reduces the opportunity for the deep karri litter to be dry enough to burn. The risks associated with a high Soil Dryness Index are accepted for deep karri fuels, as there is often a greater risk that heavy rains in autumn will prevent further burning of these fuels.

364 The burn prescription for DPHB8 prescribed a Soil Dryness Index of at least 800. That figure was first reached in the spring/summer of 2003/2004 on 6 January 2004.

365 Dr Tolhurst, one of the plaintiffs' 'fire' experts, prepared two reports. In the first report, based on a number of constraints which he identified, including that the burn should be carried out on northerly winds with a view to blowing smoke away from the plaintiffs' vineyards, Dr Tolhurst found that there was one suitable two-day period in 2006, being early May 2006, to conduct a burn of the DPHB8 area. In the same report, he also found that there were four suitable burning periods in autumn 2003 (between 9/10 April and 10/11 May), 11 suitable periods in spring 2003 (between 9/10 August and 29/30 November), three suitable periods in autumn 2004 (between 19/20 April and 27/28 May) and seven suitable periods in spring 2004 (between 8/9 August and 28/29 November).

366 I observe at this point that the periods he identified as suitable for burning occurred mainly in parts of the year when, according to the preponderance of other expert evidence, karri forest with K1 and K2 understorey is not generally amenable to safe and effective burning.

367 Dr Tolhurst said, in relation to his criterion that the burn be conducted on north winds, that even a northerly wind blowing smoke away from the vineyards 'would not necessarily guarantee that the vineyards would remain unaffected by smoke, as overnight conditions may also result in smoke collecting in the area and potentially affecting the vineyards' (first report, page 15).

368 In his second report, Dr Tolhurst enlarged the period of time under consideration, to cover each year between 1994 and 2006 inclusive. He also added two further constraints, one being to only report on periods of three consecutive days of suitable weather, and the other to find days when the McArthur 'Drought Factor' was '5 or greater to account for days when recent rainfall may have reduced the amount of available fuel to a level below an acceptable level' (page 6). Applying his criteria in his second report, Dr Tolhurst found eight occasions in autumn over the 13 years when conditions would have been suitable for burning DPHB8, and 27 occasions in spring over those 13 years.

369 I did not find Dr Tolhurst's reports helpful on the window of opportunity for prescribed burning, nor did his oral evidence carry much conviction. He conceded that his work was based on the McArthur book and not the Red Book, and that the McArthur book is of inferior application to karri forests (ts 645, 652 and 667). He also said that he had done no research in Western Australia, he had done no operational work in Western Australia, and he had no experience with karri forest. His reports were 'desk top' studies, and did not incorporate information regarding the Soil Dryness Index information (ts 644, 649, 650, 665, 661 and 674). As noted below, Dr Tolhurst's evidence was not supported by the other experts in relation to the window for burning.

370 Dr Tolhurst did agree with the other experts at the pre-trial conference that:

4.1 It was not possible to conduct DPHB8 without putting some smoke over the plaintiffs' vineyard at some time during the prescribed burn. In the stable weather conditions suitable for prescribed burning in Pemberton there can be a smoke inversion at night time and an afternoon sea-breeze

371 In cross-examination, Dr Tolhurst's evidence was (ts 649):

You would agree, I think, that if you burn within that period [veraison to harvest period] then smoke will inevitably go onto the vineyards?---To a greater or lesser extent there will probably be some smoke impact.

372 I accept that evidence too. The veraison to harvest period around Pemberton generally falls within the period January to May.

373 Dr Tolhurst found no support from the plaintiffs' other 'fire' expert, Mr Muller, or from the defendants' witnesses, in relation to the window of opportunity for burning DPHB8. The report from the pre-trial expert conferral concluded:

3.8 Dr Tolhurst was of the opinion that prescribed burning can be conducted by [the Department] in karri forest during the spring months, defined here as September, October and November. The other experts were of the opinion that it would not be practicable, safe and effective for [the Department] to conduct prescribed burning in karri forest during the spring months.

...

4.5 The Defendants' experts were of the opinion that for a large karri burn such as DPHB8 which takes place over several days, it is correct to commence the prescribed burn on south-easterly winds, at the commencement of the wind cycle. They considered it would not have been safe for [the Department] to commence the prescribed burn on northerly winds having regard to the unstable weather and extreme conditions often associated with northerly winds at the relevant time of year, the fire danger posed by the heavy and long-unburnt fuels to the south of DPHB8 and the likely effect of the wind cycle on subsequent days.

4.6 Mr Muller agreed with the Defendants' experts' position as stated in paragraph 4.5 above.

374 Mr Muller considered the window of opportunity for prescribed burning in seven fire seasons between November 2000 and May 2007. In relation to those seven seasons, his conclusions in his report were:

On average there are only between one (for Karri 1&2) and seven (for Southern Jarrah) suitable burning days outside the veraison to harvest period. Even if there were no other constraints and burning could be conducted on each of these days, this is fewer days than the number of burns required to effectively restrict the spread of major summer wildfires.

It is therefore not possible to maintain an effective fuel reduction burning program if burning is restricted to outside the veraison to harvest period.

The potential risk to grapes can be reduced by maximising spring burning, particularly for types other than Karri 1&2, however as many burns are mixed fuel some spring burns will perforce carry-over into the early post veraison period. For biodiversity reasons, it is necessary that some burns be conducted in autumn.

In most years the break in the weather that ends the burning season follows closely after harvest, and there are few autumn post harvest burning opportunities. The break in the season is not predictable, and excluding a potential burning day to delay until post harvest carries a high risk that such a burn will not be able to be carried out in that year.

The threat of smoke affecting grapes can be reduced if a higher priority is given to burning those areas closest to vineyards on the days available prior to veraison where this is possible, and conducting more distant burns later in the season.

375 Mr Muller explained in evidence-in-chief that his conclusions assumed that prescribed burning would not be carried out in the prohibited burning period, although, he said, the prohibited period would, in any event, be within the veraison to harvest period.

376 Mr Muller's overall views were summarised in the following passage in cross-examination (ts 371 - 372):

All right. Perhaps we can just recap your evidence, as I understand it, Mr Muller. Do you accept that DPHB8 was a necessary part of the prescribed - necessary as part of a prescribed burning program?---I believe it was but I'm not sure that my opinion is relevant.

Do you accept that it was inevitable that prescribed burning will cause smoke on vineyards in Pemberton or in the Pemberton region?---Yes.

And that it's not possible to have an effective prescribed burning program in the Pemberton region if you can burn only outside the [veraison] to harvest period?---That's correct.

That it's not possible to burn in winter or early spring?---That's correct, in karri - - -

In karri 1 and 2?--- - - - 1 and 2.

That it's not prudent, if not foolhardy, to burn on northerly winds?---That's correct.

That in autumn there is a high risk that a burn will not take place at all that season if it's deferred?---That's correct.

That for DPHB8 reasonable steps were taken to minimise smoke over the vineyards when the burn was conducted?---At the time the decision was taken I believe that's the case.

And there were no other reasonable steps that could have been taken short of deferring the burn?---I believe that was a reasonable course of action.

You would accept that deferring that burn would have carried a high risk that it would not have been able to be carried out that year?---Yes.

And that there was no reason to believe that a similar situation in relation to that risk would not occur in the following year?---That's correct.

That the fuel levels within DPHB8 were at hazardous levels with a risk of catastrophic results if they were not burnt?---They were at levels that would make control difficult at lower fire danger conditions. Whether it is catastrophic or not depends on what ends up in the way.

Who gets burnt?---Yes.

You would agree that another year would mean an increased fuel load?---There would be some increase but after a while the amount of annual increase tapers off so it would - but (indistinct) a risk for longer.

Yes. You would agree that a prudent fire manager would have taken advantage of the window of opportunity on 31 March 2004?---Yes.

Do you agree in general that a failure to do prescribed burns would lead - that would lead to the risk of bushfire disaster?---Yes.

Yesterday you said someone needs to decide - or words to this effect - if damage to grapes is going to affect the priority given to burns. That's one of the last things you said?---Yes.

You would agree that the question of whether to carry out prescribed burning, given the risk of damage to grapes, is really a policy question rather than a fire management question?---Yes.

Because grapes are going to be damaged if prescribed burning continued?---Yes.

377 The burn of DPHB8 was the last burn of dense karri forest in the Donnelly-Pemberton area for autumn 2004. The Department carried out other burns in April and into early May that year, but they were silvicultural burns or were conducted in dryer areas with jarrah and other more open fuel types (Walker ts 817, 850 - 860).

378 A silvicultural burn follows timber harvesting or logging, where the debris is often pushed into heaps by bulldozers. The arrangement of the fuel allows for burning further into the year, despite the presence of autumn rains. Even, however, in relation to silvicultural burns, the last burn occurred in the Donnelly district on 13 May 2004 (TB 1164).

The defendants, the statutory scheme, and the responsibility for the DPHB8 land

379 The first defendant is the Executive Director of the Department which was, at the relevant time, a statutory body. The second defendant is the State of Western Australia. The relevant statutory scheme, as at 31 March 2004, is discussed below. All references are to the legislation as it stood at the time.

380 By virtue of the definition in s 3 of the CALM Act, the 'Department' means the Department of Conservation and Land Management referred to in s 32.

381 Section 32 provides that, for the purposes of the CALM Act, there shall be established under the Public Sector Management Act 1994 (WA) (PSM Act), as a department of the Public Service of the State, a department to be known as the Department of Conservation and Land Management.

(1) The functions of the Department are, subject to the direction and control of the Minister -

...

(d) to be responsible for the conservation and protection of flora and fauna throughout the State, and in particular to be the instrument by which the administration of the Wildlife Conservation Act 1950 is carried out by the Executive Director pursuant to section 7 of that Act;

(da) to promote and facilitate public recreation, in accordance with this Act, on land to which this Act applies;

...

(3) The management of the land referred to in subsection (1)(a)(i) and the associated forest produce, flora and fauna shall be carried out -

(a) where there is a management plan for the land, in accordance with that plan ...

383 In the absence of a 'management plan', the Department may carry out necessary operations including works for the protection of persons, property, flora and fauna: s 33(3)(b), s 33A.

384 The term 'management plan' is defined in s 3 of the CALM Act to include a management plan approved under s 60 of that Act. Under s 60(4), a management plan comes into operation on the day of publication of the notice of its approval by the Minister.

385 Section 3 defines 'Executive Director' to mean the person referred to in s 36(1).

386 By s 36(1) of the CALM Act, a person is to be appointed, under and subject to Pt 3 of the PSM Act, to be the Executive Director of the Department. By s 36(3) of the CALM Act, the Executive Director is the chief executive officer of the Department.

387 As chief executive officer of the Department, the Executive Director's statutory functions and powers are prescribed in s 29 of the PSM Act. The functions of a chief executive officer are to manage the department, including managing and directing employees in the department (s 29(1)(g) PSM Act).

Subject to this Act and the Public Sector Management Act 1994, the Executive Director has power to do all things that are necessary or convenient to be done for, or in connection with, the performance of the functions of the Department.

389 Section 34A and s 34B provide, in effect, that the powers under s 34 included the power to establish business undertakings (subject to the approval of the Treasurer) and to enter into timber share-farming agreements.

390 Under s 38 of the CALM Act, the Executive Director is a body corporate with the incidental capacities of such a body. The corporate name of the Executive Director is the Executive Director of the Department of Conservation and Land Management.

391 The person who occupied the office of the Executive Director at the relevant time was Mr Keiran McNamara.

392 The Department employees who approved and conducted DPHB8 were employees of the second defendant, the State of Western Australia: s 64 PSM Act.

Management of the DPHB8 land under the CALM Act

393 The Conservation Commission of Western Australia is a body corporate pursuant to s 18(2) of the CALM Act, for the limited purposes of the CALM Act.

394 DPHB8 is a block of Crown land in the Warren National Park, and is a 'national park' within the meaning of s 3 of the CALM Act (see also s 6(3)(a) and s 7(3) of the CALM Act). It is land vested in the Conservation Commission under s 7(3) of the CALM Act. It is 'land to which the Act applies' under s 5(1)(c) of the CALM Act, and hence attracts the operation of s 33(1)(a). (See also s 7(6) and s 19(3) of the CALM Act.)

395 The management of DPHB8, including the prescribed burning of DPHB8, was conducted by the Department, subject to the direction and control of the Minister, under s 33(1)(a)(i) and s 33(3)(a) of the CALM Act.

396 The Forest Management Plan 2004 - 2013 was prepared in accordance with the CALM Act for land vested in the Conservation Commission within, inter alia, the Warren Region. It was approved under s 60 by the Minister for the Environment on 10 December 2003. Notice of the approval was published in the Government Gazette on 19 December 2003.

397 The relevant parts of the Forest Management Plan are set out in [141] - [148] above.

398 The Executive Director, as the Department's Chief Executive Officer, had a legally binding duty to manage DPHB8 in accordance with the Forest Management Plan, but it was a matter for the Executive Director to implement the plan as it saw fit, subject to any direction by the Minister. Apart from a duty to act honestly and in good faith, the Executive Director's discretion was largely unfettered: Bridgetown-Greenbushes Friends of the Forest Inc v Executive Director of the Department of Conservation & Land Management (1997) 18 WAR 126, 153 - 154, 166 - 169, 175.

Responsibility as a Hazard Management Agency

399 The Department also had a duty to take appropriate preventative action to prevent bushfire on the Department-managed lands in its capacity as a Hazard Management Agency.

400 At the time of prescribed burn DPHB8, Western Australia did not have emergency management legislation (see now Emergency Management Act 2005 (WA)).

401 Policy Statement No 7 was the overarching Policy Document for Western Australian Emergency Management Arrangements. It was issued under the authority of the State Emergency Management Committee as approved by State Cabinet on 18 July 1985 to determine policy and develop necessary planning to achieve an effective response to disaster or emergency situations. It was revised as at November 2001.

402 Policy Statement No 7 relevantly identified hazards and Hazard Management Agencies. The Department was a Hazard Management Agency for fires in Western Australia (cl 40, Annexure F).

403 Policy Statement No 7 required the Department, as a Hazard management Agency, to prepare strategic plans called 'Westplans', to manage fire hazards (cl 40, Annexure H(1)(b)). It also required the Department to ensure that emergency management activities in relation to the prevention of fire were undertaken (cl 40, Annexure H(1)(a)).

404 Westplan-Wildfire was issued in October 1999 and amended in 2001 and 2003. It provided that the Department was responsible for risk management of the Department-managed lands, including State forests, Timber Reserves, National Parks, Nature Reserves, Unallocated Crown Lands and Unmanaged Crown Reserves, except where they existed within Gazetted Fire Districts. Section 3 provided for prevention activities to eliminate or reduce the probability of occurrence and impact of bushfire. It identified Wildfire Threat Analysis as a tool to be used by Fire and Emergency Services Authority Fire Services, Local Government, the Department and other stakeholders, to identify the likelihood of wildfire and its consequences.

The Minister, the Executive Director, any person employed in the Department, a member of the Conservation Commission, Marine Authority or Marine Committee, and any person who performs functions under this Act or the Wildlife Conservation Act 1950 is not personally liable for anything done or omitted to be done by him in good faith in, or in connection with, the performance or purported performance of his functions under this Act or the Wildlife Conservation Act 1950.

405 For reasons mentioned later, it has not been necessary to determine the proper construction and scope of this section.

The Department's decision to carry out the burn on 31 March 2004 and not to defer the burn of DPHB8

406 The decision to burn DPHB8 on 31 March 2004 was made in the following broad context:

(a) the area was nearby to a significant town, Pemberton, and formed part of an important National Park. The burn had a high strategic value in the prescribed burning programme;

(b) the area had a fuel load of 73 - 93 tonnes per hectare, when the limit at which wildfire in karri forest can generally be controlled in most weather conditions is under 19 tonnes per hectare;

(c) the area also had tactical value within the overall burning programme in that, once burned, it would provide an area of low fuel from which to burn other very high fuel areas to the south;

(d) it was in dense karri forest with predominantly K1 and K2 understorey, which meant that there was a limited opportunity effectively and safely to carry out a burn. Essentially, its location and vegetation meant that it could not be properly burned at any time in the period May to at least mid-December inclusive, it was likely to become unsuitable for burning at some time in April and even in January to March/mid-April there was no assurance of favourable wind and weather conditions;

(e) it was a burn which had been introduced as part of the Master Burn Plan into the six-season burn programme in 1999, and which had remained unburned for several years because of resource limitations and the absence of favourable burning conditions;

(f) the forecast weather conditions for the period 31 March to 3 April 2004 indicated that the burn could be conducted safely and effectively within that period;

(g) the burn would serve important objectives, in accordance with the Forest Management Plan. The objectives, referred to in the burn prescription, were the protection against wildfire damage to nearby private property, and to the mature forest within the Warren National Park, and to park facilities, and also to visitors who used the park for recreational purposes, and the protection and maintenance of the biodiversity processes and ecological values;

(h) if the burn was delayed, there was no guarantee that a combination of propitious conditions and adequate resources would converge to ensure that the burn occurred in the following year, or even in subsequent years;

(i) there had been a backlog of prescribed burning and the State had recently (in 2003) experienced its worst bushfires for many years and there had also been, at a national level, the experience of the terrible bushfires in Canberra, New South Wales and Victoria in 2002/2003;

(j) the south-west forests are fire-prone, and are at real risk of wildfire from both lightning and anthropogenic causes;

(k) wildfire has potentially devastating consequences, including loss of life and destruction of private and public assets and the environment; and

(l) prescribed burning is the cheapest and most ecologically sound way to reduce fuel loads, and thereby to reduce the risk of uncontrollable high-intensity wildfires.

407 The plaintiffs, in their submissions, contended that the burn of DPHB8 'could have been delayed' or rolled over into the next spring burn season. They say that Mr McNamara accepted that the burn was 'neither necessary nor urgent'. Mr McNamara's evidence was, relevantly (ts 705 - 706):

The view that was taken at the time, that I took at the time, was that the necessity of achieving the overall strategic program was the overriding consideration.

Did you have any written advice that conducting DPHB8 burn in the autumn of 2004 was a matter or urgency?---I don't recall being told that that particular burn, or indeed any other, was a matter of particular urgency, but I would take the view that in most circumstances any individual burn is not necessarily a matter of particular urgency; but if numerous burns are not carried out and postponed on that basis, the cumulative effect is to destroy a strategic burning program.

...

Is it the case that at the time you prepared this briefing note and made the recommendation to the minister, that you had no direction to you that carrying out DPHB8 in autumn 2004 was a matter of necessity or urgency?---I'm not aware of any particular advice to me that that particular burn was a matter of necessity or urgency. What I am aware of is that the Pemberton area in a general sense was regarded as an area in which fuel loads were excessive and of concern to the department, and indeed to the community. If I might just make one other comment, I have been asked on a number of occasions about whether we have had any regard to the effect on the adjacent landowners and indeed in the particular case, but I note that in point number 2 in the briefing note ... we did say that where practical [the Department] will carry out its prescribed burns in a manner that minimises the amount of smoke in quantity and duration that is likely to be in the area where wine grapes are near the harvest, so we very clearly had in mind that our managers are expected to achieve the program, but to achieve it in a way that minimises the effect on grape growers and indeed other people, by targeting the best weather opportunities and so on, but within the limits of trying to achieve the overall program.

408 In my view, the words 'necessary' and 'urgent' are capable of creating a misleading impression, in this context, in that they are apt to convey the notion that prescribed burning is done in response to ad hoc dangers as they arise from time to time, rather than as an integrated strategy designed to deal with the systemic risk of wildfire inherent in the landscape and environment of the State generally, and of the south-west in particular. It is correct to say that there was no evidence that absent a prescribed burn on 31 March 2004, there was an immediate and substantial risk that a wildfire would break out in the area of DPHB8. To that extent, the prescribed burn of DPHB8 might, on one view of it, conceivably be described as being neither 'necessary' nor 'urgent' in early 2004. Nevertheless, that characterisation imports an artificial and narrow focus which obscures the broader context in which the large-scale programme of prescribed burning is designed and implemented in Western Australia generally, and in the south-west of the State in particular.

The decision not to defer the burn - generally

409 The decision not to defer the burn of DPHB8 was made following a request, coupled with a threat of litigation, from the plaintiffs that the burn not be undertaken until the plaintiffs had finished their harvest. That demand had been made in the context of other requests and demands by other vineyard owners to defer the carrying out of prescribed burning in the early part of 2004.

410 Both the issue generally of whether the Department should not proceed with its burning programme in the early part of 2004 where there was a risk of smoke affecting vineyards, and the plaintiffs' demand, in particular, that the burning of DPHB8 be deferred until the plaintiffs had finished harvest, were matters which were considered at the highest level within the Department, and referred to the Minister for consideration.

411 These matters were considered by the Department, and referred to the Minister, in the context of the Department having received additional funding for the purpose of carrying out prescribed burning in order to catch up on the deficit in the burning programme which had accumulated over the years from 1999 onwards, due to budgetary and other constraints. The referral to the Minister was also made in the context of the Minister having raised the need to catch up on prescribed burning, and the provision of funds for that purpose, in her statement to the Legislative Assembly. At a local level, there had also been calls at a community meeting at the Manjimup Town Hall in 2002 for the Department to undertake more prescribed burning in the area.

412 The Department had recommended that prescribed burning continue in the early part of 2004, notwithstanding the requests or demands by vineyard owners to defer burning until the end of the grape harvest that year. The Department's concern was that the deferral of strategic burns at the request of grape growers would cause significant reductions in prescribed burns and increase the risk of uncontrollable wildfire.

413 The Minister endorsed the Department's approach. At a national level, the Department's general policy of prescribed burning had been, at least impliedly, endorsed in the report 'A Nation Charred'.

The decision not to defer the commencement of the burn from 31 March to 1 April 2004

414 The burn prescription originally prescribed that the edging of the burn be carried out on southerly (south-west to south-east) winds, and for the core ignition also to be carried out on southerly winds. This accorded with the Department's standard procedure that, in this area, prescribed burns are most safely conducted on winds from the south.

415 Mr Marshall's note of 1 February 2002 recorded the proposal that the burn be carried out with winds from the east to north-east, with a view to keeping the smoke off the plaintiffs' vines. Mr Marshall's note made no reference to the additional resources which would be required if the burn were conducted on winds in the east to north-east quadrant.

416 Mr Gillard's note of 3 February 2004 amended the burn prescription by providing that the burn be conducted so as to minimise its impact on the plaintiffs' vineyards. It did so by addressing the wind direction on which the burn should be 'completed' and the need for extra resources to be deployed in order to make the burn safe. Mr Gillard's evidence, which I accept, was that he wanted the core ignition to be undertaken on east-south-east to easterly winds to minimise the volume of smoke near the vineyard from the overall burn, without compromising the safety of the burn. By amending the burn prescription so that core ignition would occur on winds within the south to east quadrant, he wished to ensure that the majority was burnt out before the wind had moved from the east, through the north, and to the north-west, and then down to the south. Winds from the north, north-west would have posed a serious risk to burn security. (See Gillard's supplementary statement pars 13, 15 and 17).

417 This matter, and the need for additional resources was referred to by Mr Sneeuwjagt in cross-examination. This evidence, which I accept, was (ts 955):

Wasn't it your opinion that burns should not be carried out on winds with a northerly component?---North-north-westerly, absolutely, yes.

But with a northerly component?---Well, as I said to you, the north-nor'-westerly are the unstable winds.

So north-east and easterly are stable winds in the cycle?---North-east can be unstable, can be, but the forecast for this did not indicate that.

And you had no objection then to a burn being carried out on winds with a northerly component other than north-west. Is that your evidence?---In this case I was happy to approve that but knowing the weather, the moisture, the time of the year et cetera I was quite comfortable, knowing that this was an increased risk to those properties directly to the west of the burn. It must be said the reason we prescribed this burn for a south-west - south-east in the first instance was to minimise the possibility of these houses getting impacted by fire. To take it to east-north-east increased the risk. We therefore put a bulldozer on that site and had extra trucks put on there, because we were attempting to minimise the impact here, realising that we may be increasing the risk to that community.

Well, you actually maximised the impact on the vineyards because of the south-south-east winds blowing smoke directly onto it?---I don't believe so.

That was the effect of your authorising the burn, wasn't it?---I think I explained that the authorisation was on the basis that I expected minimal impact from the edging on that vineyard and to allow this burn to proceed I was prepared to accept north-east winds.

I put it to you, Mr Sneeuwjagt, that that's an after-thought in the light of these proceedings and in order to try to excuse what was done?---I had the benefit of the four-day forecast. I had the benefit of the smoke trajectories. They all fitted in with what I've just said. Which is that the bulk of the smoke would be blowing away from the vineyard, or at least minimal impact.

418 As noted previously, there was, in fact, a hop-over on the western boundary towards the north of that boundary that had to be dealt with.

419 After Mr Gillard's amendment to the burn prescription, the Department continued to monitor the weather and fuel moisture conditions. The four-day forecast on 30 March 2004 indicated that weather conditions would be propitious for the burn of DPHB8. The fire behaviour parameters (surface moisture content, profile moisture content and soil dryness index) were also recorded as being consistent with the burn prescription.

420 The weather forecast on 30 March 2004 predicted light south to south-easterly winds on 31 March 2004, light north-easterly winds on 1 April, and winds moving back to become easterly winds on 2 and 3 April 2004. In other words, the four-day forecast indicated that if the prediction proved accurate, the west coast trough would not arrive and bring hot, gusty winds from the north to west quadrant within the period 31 March to 3 April.

421 There was, however, information from the Bureau of Meteorology duty forecaster which suggested that after or towards the end of that four-day period, there was a risk of the arrival of hot, gusty northerly winds (see [295] above).

422 Mr Gillard was asked why he did not delay another day so that the edging burning commenced on 1 April 2004, when the wind was expected to be from the north-east. He said, and I accept (ts 523):

Why did you not delay for another day?---The information that we had was from the duty forecaster relayed though our various duty officers that there was a trough forming later in the week and that meant that we would have unstable north-side winds and it was important that we completed the burn within the two days which was Wednesday and Thursday then to allow a couple of days for us to complete the ignition component of [the] burn in two days so that the bulk of the burn was burnt out and then allowing a couple of days before the trough formation and movement to ensure that we were able to successfully patrol and extinguish any burning embers, any trees that may be alight [particularly] on the southern boundary which would then be a risk of the burn escape from the northerly winds ...

423 He also referred to another 'minor' consideration being that once the trough came through and the winds moved from the north-west quadrant and down to the south-west, those latter winds would have exacerbated the prospect of smoke being carried over the vineyards.

424 The plaintiffs, in their submissions, drew attention to evidence from a number of the defendants' witnesses to the effect that burning, by core ignition, on forecast light north-easterly to easterly winds was safe and that winds of that kind in fact provided favourable burning conditions. The plaintiffs say that this evidence indicates that the burn of DPHB8 could have commenced on forecast north to north-easterly winds on 1 April, and amounts, in effect, to an admission by the Department that it was negligent in having commenced the burn on forecast south to south-easterly winds on 31 March 2004.

425 In my opinion, the plaintiffs' submissions fail to address the burden of the evidence. The burden of the evidence from the defendants' witnesses was that within the four-day forecast, the ignition could safely be completed on the second day, with the result that the heavy smoke from the core ignition ought have minimal impact on the plaintiffs' vineyards. At the same time, the overall threat of wildfire would be minimised by starting the edging on the first of the four days, and having the whole burn completed and mopped-up before the west coast trough arrived, with the risk of gusty winds from the north and north-west.

426 A typical example of the exchanges in cross-examination may be seen in the evidence of Mr Walker. Before quoting the exchange, I would note the following. In 2006, the plaintiffs applied for (and were refused) an interlocutory injunction to restrain the burning of an area by another prescribed burn, DP60, until after their grape harvest. The area of DP60 was two to four kilometres to the south-west of the plaintiffs' land. Mr Walker gave evidence in that interlocutory application in which he referred to the burn of DPHB8 in 2004. He was cross-examined in the trial of this action on that evidence. In his cross-examination, Mr Walker conceded that an aspect of his evidence in the earlier interlocutory application was inaccurate. Notwithstanding that error, I found him to be a generally reliable and impressive witness, with extensive experience in fire management and prescribed burning.

Was that not the case with DPHB8 that the burn could have been commenced on east to north-east winds as Mr Marshall proposed?---It's my view that that would have presented an unnecessary risk or an unreasonable risk that the winds would have swung to the north during the conduct of the burn that would have represented a significant risk to containment of the burn and the values to the south of the burn were such that would have been a unreasonable and ill-advised commencement.

Would you look at ... the quotation from Mr Marshall where he suggested the burn be carried out with winds from the east to north-east - that this will keep the smoke off the grape vines ... He does not make any such qualification as you have just given in evidence, does he?---What we can make of that is that the preference would have been for winds to be blowing from the east to the north-east. That's indeed what was done in respect of the core lighting and the subsequent days of burning where the bulk of the smoke was generated. So the actual conduct of the burn was in line with the strategy that Mr Marshall was advocating.

And with the core ignition, would that mean that the bulk of the fire was also caused from which you say the bulk of the smoke arose?---Yes, the ignition.

That apparently, to your officers, didn't pose any concern about safety of forest areas to the south to start the core ignition on north-easterly winds, did it?---The - - -

That's right, isn't it?---The ignition on east to north-east winds as part of that cycle would have been before the wind swung to the north presenting the unstable and less predictable circumstances of northerly winds.

Starting, as your officers did, on 1 April - sorry, not starting it. Doing the core ignition, as your officers did, on 1 April on north-easterly winds, obviously was not a matter of safety concern to them with regard to forests to the south?---Generally the wind - -

Can you answer that yes or no? That's the case, isn't it?---The forecast for a north-easterly wind at eight kilometres an hour would not have represented a significant risk of containment of that burn. By contrast, had the wind swung to the north-west and the wind strength had been greater, that would have represented a significant risk.

The four-day forecast was for three days of easterly and north-easterly winds as the prevailing winds during those three days?---Yes.

That did not present any risk?---With the wind velocity that was forecast and the stability of the weather pattern, it represented a relatively low risk.

And indeed if you look at what you said on page 44, immediately after the quotation of Mr Marshall's note, you were asked this question, 'Is that right - winds of east to north-east?' and your answer, 'For the portion of that burn that lay immediately south of the vineyard, that would have been the most favourable prevailing wind to conduct the burn. That is correct, isn't it?---Yes, that's consistent with the discussion we just had.

...

Is not the upshot of all of that, Mr Walker, that the late Mr Denis Marshall was correct in amending the prescription for this burn to burn on east to north-east winds to keep the smoke off the grapevines?---That was a favourable wind and that was the wind direction that was chosen for the core lighting on subsequent days.

It was in fact not the prescription that was followed by the officers who conducted the burn because they commenced the burn on south-south-easterly winds, did they not?---The edging commenced on a day where there were south-south-easterly winds forecast as part of the winds for that day, but knowing that the winds were going to swing to the east and the north-east for the core lighting, that was the chief concern in choosing the first day to commence the edging.

428 As I have noted (see [373] above), all the experts, apart from Dr Tolhurst, agreed, 'it is correct to commence the prescribed burn on south-easterly winds, at the commencement of the wind cycle' (emphasis added). North-easterly winds, particularly of gentle velocity, do not present the danger associated with north-westerly winds. Also, easterly and (albeit to a lesser extent) north-easterly winds, generally occur in the earlier, and safer, part of the weather cycle.

The extent of the Department's knowledge of smoke taint in relation to the conduct of prescribed burning

429 In the south-west of Western Australia for over 40 years, there had been only one instance brought to the Department's attention where a concern regarding smoke taint had arisen, and that was in relation to the plaintiffs' harvest following the prescribed burn in 2001.

430 The Department, through Dr McCaw, initially sought information from the Australian wine industry and other sources concerning any scientific data or literature in relation to the possibility of smoke tainting wine grapes, and the circumstances under which that might occur, and the levels of smoke which created any risk in that regard. None was then available.

431 In March 2003, the Department carried out a burn on land adjacent to the Gloucester Ridge Winery. There was no complaint of smoke taint following that burn.

432 In late 2003, the AWRI published a paper which indicated that intensive smoke from bushfires would impact upon the quality of grapes used in wine production. The report referred to the 'possibility' of prescribed burning also affecting wine grapes. In January 2004, the author of that report met with representatives of the Department. The author of the report said that he was unable to say whether prescribed burning was likely to have any significant effect on wine grapes.

433 In hindsight, having regard to the expert evidence in this case, the risk of smoke taint in wine grapes is at its highest in the period seven days after veraison and up to the time of harvest, but it is not insignificant in the pre-veraison period prior to then.

434 There was no exploration in the course of the trial as to why the plaintiffs' crops had been affected, when there had otherwise been no complaints in the previous 40 years, and when a burn of land adjacent to the Gloucester Ridge Winery had not provoked any complaints of smoke taint to the grapes in that vineyard. The expert evidence adverted to the potential for environmental and physiological restraints to affect the uptake of smoke compounds in the berries and other plant material. There was no examination in this case concerning any specific environmental or physiological factors with respect to the plaintiffs' vines.

The plaintiffs' case, and preliminary observations on it

435 The plaintiffs did not plead that the defendants owed them a duty of care, although their case was in negligence, so it depended upon, inter alia, the existence of a duty of care.

436 In opening written submissions, the plaintiffs articulated a duty in these terms:

[The defendants] owe a duty of care to the plaintiffs to avoid the reasonably foreseeable risk of injury to the grapes and vines in the Vineyards.

437 In closing written submissions, the plaintiffs articulated the alleged duty as follows:

[The defendants] ... had a duty of care in managing the forest not to cause harm to the plaintiffs' vineyards by the lighting of fires in the forest adjacent to the vineyards.

438 The plaintiffs, in their opening submissions, described the duty of care as a 'non-delegable duty of care'. They said:

The special relationship of proximity between the forest under [the Department's] control and the Vineyards attracts the rule in Rylands v Fletcher and the Burnie Port Authority principles, so that the Defendants are subject to a non-delegable duty of care, which is heightened according to the dangerousness of the product or substance, or the activity involved (Torette House Pty Ltd v Berkman (1939) 39 SR (NSW) 156 per Jordan CJ at 164-165 and Burnie Port Authority v General Jones Pty Ltd [1994] HCA 13; (1994) 179 CLR 520 per Mason CJ, Deane, Dawson, Toohey and Gaudron JJ at 550-4).

439 Burnie Port Authority v General Jones Pty Ltd [1994] HCA 13; (1994) 179 CLR 520 was a case in which the owner of a building was liable to a licensee of the building for the destruction of the licensee's property when work was carried out on the building by an independent contractor (ie, not a servant or agent of the owner) in a negligent manner causing a fire to break out. The majority said (560):

In [the] circumstances, the overall work which the independent contractor was engaged to carry out on the premises was a dangerous activity in that it involved a real and foreseeable risk of a serious conflagration unless special precautions were taken to avoid the risk of serious fire. It was obvious that, in the event of any serious fire on the premises, [the plaintiff's] frozen vegetables would almost certainly be damaged or destroyed. In these circumstances, the [defendant], as occupier of those parts of the premises into which it required and allowed [the hazardous materials] to be introduced and the welding work to be carried out, owed to [the plaintiff] a duty of care which was non-delegable in the sense we have explained, that is to say, which extended to ensuring that its independent contractor took reasonable care to prevent the [hazardous materials] being set alight as a result of the welding activities. It is now common ground that [the independent contractor] did not take such reasonable care.

440 The nature of a non-delegable duty of care was discussed in New South Wales v Lepore [2003] HCA 4; (2003) 212 CLR 511, see particularly [30] (Gleeson CJ), [105] (Gaudron J), [144] (McHugh J), [254] - [257] (Gummow and Hayne JJ), [290] (Kirby J) and [339] (Callinan J).

441 In my view, this case does not raise for determination the existence or otherwise of a non-delegable duty of care as explained and applied in Burnie Port Authority v General Jones Pty Ltd or as discussed in New South Wales v Lepore. This is not a case raising for consideration the question of whether the defendants would be liable for acts done by independent contractors. Nor is it a case where there was a risk of conflagration if appropriate precautions were not taken. Here, fire was not a risk of the conduct, but the very nature and purpose of the conduct. Also, unlike the owner of premises in Burnie Port Authority v General Jones Pty Ltd, here the Department had statutory functions which included an obligation to cause and use fire; ie, by the carrying out of prescribed burning. Here, the outbreak of fire was not the product of negligence. Rather, fire was a deliberate strategy, legislatively sanctioned, to achieve certain important social goals.

442 Further, in my view, neither formulation of the alleged duty is tenable in the circumstances. Each formulation specifies the content of the duty without reference to the exercise of reasonable care, and both, although particularly the latter, tend to suggest an obligation simpliciter to prevent harm. In these respects, the formulations are, in my view, inapt: Vairy v Wyong Shire Council [2005] HCA 62; (2005) 223 CLR 422 [25], [118]; Roads and Traffic Authority of NSW v Dederer [2007] HCA 42; (2007) 234 CLR 330 [18], [43], [51], [54]; Rosenberg v Percival [2001] HCA 18; (2001) 205 CLR 434 [68]; Maloney v Commissioner for Railways (NSW) (1998) 52 ALJR 292, 292.

443 Also, the suggested formulations pay insufficient attention to the particular nature of the harm suffered, viz, smoke taint, and the relevant risk, viz, the risk of smoke tainting the grapes: see Modbury Triangle Shopping Centre Pty Ltd v Anzil [2000] HCA 61; (2000) 205 CLR 254 [14], [104] - [105]; Cole v South Tweed Heads Rugby League Football Club Ltd [2004] HCA 29; (2004) 217 CLR 469 [1], [82]; Roads and Traffic Authority (NSW) v Dederer [18], [59] - [61].

444 Further, the plaintiffs' vineyard is not the only vineyard in the Donnelly Pemberton area. There are approximately 60 vineyards which are adjacent or in close proximity to the areas managed by the Department, and which are also susceptible to smoke taint. The plaintiffs are members of the class of persons owning or operating vineyards in the area. Any duty would be owed to the class.

445 Accordingly, it seems to me that, in substance, any arguable duty would be along the following lines - that in planning and carrying out prescribed burns on land which is adjacent to or in close proximity to vineyards, the defendants owe a duty to exercise reasonable care to avoid the risk of smoke taint or damage to the fruit or vines on the vineyards.

446 A duty in those terms would not, in my view, be expressed with such particularity to, in effect, circumvent the requirement of reasonableness at the breach stage, nor would it be too abstract so as to provide no adequate legal mean by which to determine the issues in the case: cf Amaca Pty Ltd (Formerly James Hardie & Co Pty Ltd) v Hannell [2007] WASCA 158; (2007) 34 WAR 109 [350].

447 In relation to the matters said to give rise to a duty, the plaintiffs alleged that a duty was established by:

(a) the proximity of the forest to the adjoining vineyards;

(b) the fact that the Department's officers knew grapes were ripening and were due to be harvested within the proposed burning period;

(c) the fact that the Department's officers knew that produce from the vineyards had been damaged from the earlier adjacent prescribed burn in 2001;

(d) the fact that the Department's officers knew of the possibility that smoke would taint wine grapes and render them unfit for winemaking; and

(e) the risk of harm from the smoke from the burn tainting the plaintiffs' grapes was a real risk and was not insubstantial.

448 In substance, the above matters constitute an allegation that the duty is established by reason of the fact that injury or damage to the plaintiffs' grapes was reasonably foreseeable.

449 As noted below, however, reasonable foreseeability of harm is a necessary, but not sufficient, condition for the existence of a duty of care in these circumstances.

450 The plaintiffs, in par 11 of their statement of claim, allege that the second defendant (the State of Western Australia) and its servants or agents in knowing or having ought to have known of the existence of the plaintiffs' vineyard and the fact that the grapes were ripening and ready, or nearly ready, for harvest, were negligent in that they:

(a) lit the fires when they could have delayed doing so until after the grapes in the vineyard had been harvested;

(b) lit the fires when they were aware, or should have been aware, of the possibility of smoke from the fire spreading across to the plaintiffs' land and falling upon the vineyard;

(c) lit the fires when the weather conditions were unsuitable for avoiding smoke and ash from the fires entering the plaintiffs' land and falling upon the vineyard;

(d) did not control, or attempt to adequately control, the fires so as to prevent smoke and ash from the fires entering the plaintiffs' land and falling upon the vineyard;

(e) by commencing, continuing and increasing the fires along the southern boundary of the plaintiffs' land, the first defendant allowed the volume of smoke and ash to increase; and

(f) did not take any, or any reasonable, precautions to ensure that smoke and ash would not escape from the fires on Crown land onto and into the plaintiffs' land on which the vineyard was situated.

451 The allegation in subpar (a) was developed in written closing submissions by alleging, in effect, that the burn at DPHB8 could have been 'rolled over' into the next spring 2004 burn season.

452 In closing submissions, the allegations of breach in subpars (b) - (e) reduced, in substance, to the proposition that the second defendant was negligent in commencing the burn on forecast south, south-east winds. This proposition is the converse of the propositions contended for in subpar (f), discussed below.

453 With respect to the allegation in subpar (f), the plaintiffs contended that:

(i) the burn at DPHB8 was conducted contrary to the burn prescription by failing to commence the burn on winds in accordance with Mr Marshall's note of 1 February 2002 or Mr Gillard's note of 3 February 2004; and

(ii) the burn could have been commenced on 1 April 2004, rather than on 31 March 2004, and thereby conducted on forecast north-east and east winds on 1 and 2 April 2004, with the intention of blowing the edging and core ignition smoke away from the vineyards.

454 The allegation in subpar (i) is relevant to the application of s 5W(d) of the Civil Liability Act 2002 (WA), referred to below. Although not expressly mentioned in submissions, the allegation in subpar (i) above was also perhaps intended to relate to the proposition that a failure to observe a statutory obligation (assuming for the moment that the burn prescription created a statutory obligation pursuant to the Forest Management Plan) may be used in an evidentiary sense as indicative of negligence at common law: Crimmins v Stevedoring Industry Finance Committee [1999] HCA 59; (1999) 200 CLR 1 [161], [164].

455 In addition, in the plaintiffs' submissions, there was an unpleaded allegation in negligence to the effect that the defendants failed to consult meaningfully with grape growers on land adjacent to the area of the burn. This allegation led back to the substance of the earlier allegation regarding deferral of the burn. It was said that the defendants would, in effect, have carried out the burn at a different time, or in different circumstances, to accommodate the risks to the plaintiffs' vineyards, had meaningful consultation occurred.

456 The plaintiffs said that the matters they had identified as evidence of breach of duty also went to proving that s 5B of the Civil Liability Act did not operate to prevent the defendants being liable to the plaintiffs. Section 5B provides:

5B. General principles

(1) A person is not liable for harm caused by that person's fault in failing to take precautions against a risk of harm unless -

(a) the risk was foreseeable (that is, it is a risk of which the person knew or ought to have known);

(b) the risk was not insignificant; and

(c) in the circumstances, a reasonable person in the person's position would have taken those precautions.

(2) In determining whether a reasonable person would have taken precautions against a risk of harm, the court is to consider the following (amongst other relevant things) -

(a) the probability that the harm would occur if care were not taken;

(b) the likely seriousness of the harm;

(c) the burden of taking precautions to avoid the risk of harm;

(d) the social utility of the activity that creates the risk of harm.

457 Questions of whether the Civil Liability Act supplants the common law relating to duty of care, standard of care and causation, or modifies it, or reflects it, and if so to what extent, were considered recently by the Court of Appeal: Department of Housing and Works v Smith [No 2] [2010] WASCA 25. Pullin JA, (Newnes JA agreeing) said that he gained the impression that the High Court in Adeels Palace Pty Ltd v Moubarak [2009] HCA 48; (2009) 260 ALR 628 regarded the relevant provisions of the Civil Liability Act 2002 (NSW) therein considered as supplanting the common law. His Honour appears, however, to have left open these questions for further determination [11] - [20]. Buss JA held at [77] that s 5B of the Civil Liability Act relates to a breach of a duty of care and does not modify or supplant the common law principles which determine whether a duty of care exists or not.

458 I respectfully find Buss JA's reasoning persuasive on the operation of s 5B. In any event, in these proceedings the parties argued the case on the basis of the existence or otherwise of a common law duty of care. For reasons set out later, in my view no duty under the common law exists in this case.

459 Section 5B sets out the requirements that must be satisfied before a person can be found liable for failing to take precautions against a risk of harm: Penrith Rugby League Club Ltd (t/as Cardiff Panthers) v Elliott [2009] NSWCA 247 [22] (a decision with reference to the analogous, but not identical, provision in the New South Wales legislation).

460 Both the issues of duty and breach also fall to be considered having regard to s 5W of the Civil Liability Act.

461 Section 5W provides:

5W. Principles concerning resources, responsibilities etc. of a public body or officer

The following principles apply in determining whether a public body or officer has a duty of care or has breached a duty of care in proceedings in relation to a claim to which this Part applies -

(a) the functions required to be exercised by the public body or officer are limited by the financial and other resources that are reasonably available to the public body or officer for the purpose of exercising those functions;

(b) the general allocation of those resources by the public body or officer is not open to challenge;

(c) the functions required to be exercised by the public body or officer are to be determined by reference to the broad range of its activities (and not merely by reference to the matter to which the proceedings relate);

(d) the public body or officer may rely on evidence of its compliance with the general procedures and applicable standards for the exercise of its functions as evidence of the proper exercise of its functions in the matter to which the proceedings relate.

462 Each of the Crown in right of the State of Western Australia, the Department, and the Executive Director as an incorporated body, is a 'public body or officer': s 5U(a), (b) and (g) of the Civil Liability Act; s 3 Crown Suits Act 1947 (WA).

463 The plaintiffs submitted that s 5W(a), (b) and (c) have no bearing because their case does not involve an allegation of a failure to allocate resources, and the conduct of the burn of DPHB8 was not, in any event, limited by the extent of the resources available to the Department. In relation to s 5W(d), they rely on the allegation that the burn was not conducted in accordance with the burn prescription.

464 The plaintiffs also contended that s 5X of the Civil Liability Act had no application. Section 5X provides:

5X. Policy defence

In a claim for damages for harm caused by the fault of a public body or officer arising out of fault in the performance or non-performance of a public function, a policy decision cannot be used to support a finding that the defendant was at fault unless the decision was so unreasonable that no reasonable public body or officer in the defendant's position could have made it.

465 Section 5U of the Civil Liability Act defines 'policy decision' as 'a decision based substantially on financial, economic, political or social factors or constraints'.

466 Although some reference had been made to the prospect of burning DPHB8 prior to 31 March 2004 in the plaintiffs' written opening submissions, the plaintiffs' counsel, in closing, indicated that the plaintiffs did not allege that the defendants were negligent in failing to conduct the burn at some time in history prior to 31 March 2004 (ts 1110 - 1111).

Principles in relation to, and findings on, duty of care

467 In Drexel London (a firm) v Gove (Blackman) [2009] WASCA 181 [262] - [268] McLure JA (as her Honour then was) (Le Miere J agreeing) set out the following propositions relevant to the existence or otherwise of a duty of care of a public authority under the general law. Reasonable foreseeability of harm of the kind suffered is a necessary, although insufficient, condition for the existence of a duty of care. The imposition on a public authority of a duty of care directly connected with the existence of a public power or duty raises discrete considerations, for example, there must be no incompatibility between the public authority's statutory powers and duties, and its common law duties. The mere existence of a statutory duty or power does not create a common law duty, even if harm of the kind suffered is reasonably foreseeable. Something additional is required. What that something is depends on all the circumstances. Ordinarily, a public authority will have no liability for a pure omission - a term intended to exclude an omission occurring in the course of positive conduct. That is a consequence of the need for appropriate accommodation of public and private law principles and remedies. The existence of a statutory power or performance of a statutory duty may (not must) import a common law duty to take care. There will be no common law duty of care if a statute expressly or impliedly excludes a parallel common law duty. Although the mere existence of a statutory duty or power does not create a common law duty, the scope and purpose of a statutory duty or power can inform and affect the existence and content of any common law duty.

468 Her Honour's reference to a public authority 'ordinarily' not being liable for a pure omission would, as I understand it, accommodate the particular position of highway authorities, and allow for the possibility of liability for nonfeasance in other authorities where, in all the circumstances, there exists sufficient control over the source of risk of harm to persons injured: Brodie v Singleton Shire Council [2001] HCA 29; (2001) 206 CLR 512 [103].

469 In relation to her Honour's reference to incompatibility between an asserted common law duty and the statutory scheme, the High Court in Sullivan v Moody [2001] HCA 59; (2001) 207 CLR 562 [50], [55], had said:

Different classes of case give rise to different problems in determining the existence and nature or scope, of a duty of care. Sometimes the problems may be bound up with the harm suffered by the plaintiff ... Sometimes they may concern the need to preserve the coherence of other legal principles, or of a statutory scheme which governs certain conduct or relationships. The relevant problem will then become the focus of attention in a judicial evaluation of the factors which tend for or against a conclusion, to be arrived at as a matter of principle.

...

More fundamentally, however, these cases present a question about coherence of the law. Considering whether the persons who reported their suspicions about each appellant owed that appellant a duty of care must begin from the recognition that those who made the report had other responsibilities. A duty of the kind alleged should not be found if that duty would not be compatible with other duties which the respondents owed.

470 In relation to the question of coherence between the law of tort and a statutory scheme, Spigelman CJ (albeit in dissent in the result), in Hunter Area Heath Service v Presland [2005] NSWCA 33; (2005) 63 NSWLR 22 [21] said:

The imposition of a duty of care may be inconsistent with some aspect of the scheme or, if not directly inconsistent, may be otherwise inappropriate by reason of the scope and purpose of the legislation:

• liability in tort may 'distort [the] focus' of the statutory decision-making process;
• the decision may be made in a 'detrimentally defensive frame of mind';
• a common law duty should not be imposed if it 'would ... have a tendency to discourage the due performance of ... statutory duties';
• the imposition of a duty of care may 'undermine the effectiveness of the duties imposed by the statute';
• 'a common law duty could distort the performance of the functions of the statutory body'. (citations omitted)

471 Reasonableness is another important criterion for determining the existence or otherwise of a duty of care: Tame v The State of New South Wales [2002] HCA 35; (2002) 211 CLR 317 [9], [12], [14], [35] (Gleeson CJ), [109] (McHugh J), [185] (Gummow & Kirby JJ), [272] (Hayne J) and [331] (Callinan J).

472 In considering the existence and scope of any duty of care in this case, the provisions of s 5W of the Civil Liability Act are also applicable (see [461] above).

473 Whilst the Department had a statutory duty to manage the Warren National Park, including the area in DPHB8, this is not a case which raises for consideration the nature and scope of the duty arising from the management and control of the land owed to entrants upon the land: cf Romeo v Conservation Commission of the Northern Territory [1998] HCA 5; (1998) 192 CLR 431; Vairy v Wyong [117].

474 Nor is it the sort of case which might have arisen if DPHB8 had not been incorporated into the Master Plan and burned by prescription, or if the burn had been deferred, and the area subsequently became engulfed by a wildfire (eg, from lightning or anthropogenic causes) which then escaped to the plaintiffs' adjoining land and destroyed their vineyards and buildings or injured their workers, or caused loss of life: cf Pyrenees Shire Council v Day [1998] HCA 3; (1998) 192 CLR 330.

475 Nor is it a case where, in the positive exercise of its functions in the carrying out of the prescribed burn, the Department failed to exercise reasonable care to avoid hop-overs escaping and destroying the plaintiffs' vineyards and other property: cf Burnie Port Authority v General Jones Pty Ltd 520; Caledonian Collieries Ltd v Speirs [1957] HCA 14; (1957) 97 CLR 202, 220.

476 As indicated earlier, in my view, the context of any duty must be considered with respect to the injury, and the risk of injury, to the plaintiffs' grapes by smoke taint.

477 In this context, it is to be noted that the Department has no control over the way smoke will drift once a fire is started. Smoke moves from the fire through the atmosphere, blown by wind. As Mr Sneeuwjagt said (ts 911), both during the course of the day, and over the course of a period of a few days in which a burn is carried out, the wind can and does fluctuate, affecting the direction in which smoke is blown. There was no suggestion in this case that large fans or other mechanical devices could be erected or utilised to divert the natural movement of smoke to a particular area, or in a particular direction. Once the burn commenced, the smoke would travel in whatever direction the wind, in combination with the relevant atmospherics, would take it.

478 This fact focuses attention, then, on the decision to undertake the burn. If there is no ability to control the drift of smoke once the burn commences, the only measure open to the Department to affect potentially the movement of smoke from a burn is to choose the timing of the burn, so that it occurs in conditions which are most likely (but without any guarantee) to be conducive to the smoke being blown in a particular direction.

479 In the context of an alleged duty to exercise reasonable care to avoid smoke taint to grapes, it is important to recall that:

(a) grapes are contaminated by smoke coming in contact with the vines, and the berries directly absorbing smoke compounds, or other parts of the vine absorbing the smoke compounds and the translocation of the smoke compounds to the berries via the phloem of the grapevine;

(b) it is likely that any smoke exposure, even of a low density or for a short period, has the potential to contribute to the uptake of smoke compounds in the berries and other parts of the vines;

(c) the risk of uptake of smoke compounds into the grapes is likely to be at its highest for seven days post-veraison to harvest, but the risk is not insignificant in the preceding period between when the berries are pea-size to three days post-veraison, and there remains a risk, albeit a low risk, in the vine-dormancy and early shoot-emergence period; and

(d) whilst it is likely that any event of smoke exposure has the potential to contribute to the uptake of smoke compounds, the level of contribution will be dependent on, inter alia, undefined environmental or physiological restraints.

480 It is also relevant to recall (see [142] above) that the actions proposed by the Forest Management Plan, including relevantly prescribed burning, are to be undertaken with the objectives of:

• optimising the maintenance of forest ecosystem health and vitality;
• promoting the conservation of biodiversity;
• controlling adverse impacts of wildfire on the social, cultural and economic values of land managed by the Department and adjoining land; and
• minimising the risk of smoke emanating from prescribed burns impacting on population centres, and on other sensitive areas.

481 At least on one view of it, the plaintiffs' case involves the proposition that the Department should not have performed its statutory obligation to manage the land by the carrying out of a prescribed burn in accordance with its fire management plan pursuant to the Forest Management Plan (relevantly, the Master Burn Plan under which, for some years, DPHB8 had been a part and which fitted the Department's criteria for burning on 31 March 2004), in order to avoid the risk of smoke taint to the plaintiffs' grapes. Insofar as that is the nature of the plaintiffs' case, it is, in my view, inconsistent with the statutory functions of the Department and no duty of care arises under the general law.

482 If the plaintiffs' case is characterised as a claim that the Department had a common law duty, ie, a legal obligation the breach of which would render it liable in damages, to exercise care to avoid the risk of smoke taint to grapes, to which its statutory duty to carry out prescribed burning so as to serve the overall objectives prescribed by the Forest Management Plan is subservient, the common law duty would also, in my view, be inconsistent with the statutory scheme.

483 If the plaintiffs' case is characterised as involving a duty by the Department to take into account relevant considerations and give proper consideration to the risk of smoke taint to the plaintiffs' grapes in planning and implementing prescribed burning so as to minimise the effect of smoke on 'sensitive areas' within the meaning of cl 17 of the Forest Management Plan, the breach of a duty so expressed is more consonant with the potential availability of public law remedies, rather than common law damages. (In any event, the Department did have regard to the plaintiffs' grapes by amending the burn prescription and undertaking the core ignition on winds which were thought likely to minimise the impact of smoke on the plaintiffs' vineyards).

484 Further, the existence of such a common law duty of care would tend, at least, to distort, by skewing, the focus of the statutory decision-making process, and distort the performance of the Department's statutory functions.

485 For these reasons, I would not regard the defendants as coming under a duty of care to the plaintiffs to avoid smoke taint to grapes, when the Department was undertaking the planning and implementation of prescribed burning.

486 Further, in the circumstances, the imposition of a duty of care would not, in my view, be properly regarded as reasonable.

487 Looking at the matter in a broad and practical way, it is important to note that:

(a) karri forest, particularly with K1 and K2 understorey, cannot realistically be safely or effectively burned in the period May to about mid-December, which leaves the potential for burning to be undertaken to the period between late December and April, and even within that period, there are limited opportunities when fuel moisture levels and weather conditions combine to allow the burn to be appropriately undertaken;

(b) the greatest risk of smoke taint is between seven days post-veraison to harvest, which falls generally within January to May, with the next greatest risk in the preceding period, from when the berries are pea-size through to three days post-veraison, which, I infer, would be measured in a period of at least weeks, rather than days (there is also some small risk from early-shoot development).

488 Accordingly, practically speaking, the time when it is safe and effective to carry out prescribed burning in karri forest with K1 and K2 understorey is the time when the grapes are at stages of development where smoke taint poses a not insignificant risk. A duty to exercise reasonable care to avoid the risk of smoke taint to the plaintiffs' grapes in planning and implementing a prescribed burn would involve the imposition of a legal obligation which was, in practical terms, incapable of being discharged.

489 Also, given the risk of smoke taint to grapes, the distance over which smoke may be carried on the wind, and the existence of 60 vineyards in the area in question, the existence of a common law duty is likely to involve competing duties to vignerons within the class. A wind forecast to take smoke away from one vineyard may well take it on to another. The point is illustrated by the vineyards around the prescribed burn 'Big Brook' (exhibit 33B), which has vineyards to the north and north-west, to the west, to the south, and to the east and north-east. As Mr Sneeuwjagt said, any burn of that area is likely to affect some, if not all, of the surrounding vineyards.

490 For these reasons also, in my view, it would not be reasonable for such a duty of care to be imposed.

491 However, I will proceed to consider the question of breach of duty in case I am wrong in finding that a duty of care was not owed to the plaintiffs.

492 The question of breach of duty must be considered by reference to the relevant provisions of the Civil Liability Act: Adeels Palace Pty Ltd v Moubarak [27].

493 In relation to s 5B of the Civil Liability Act, the 'risk of harm' to which the section refers is, relevantly, the risk to grapes of smoke taint. The first question, under s 5B(1)(a), is whether it was a risk of which the Department knew or ought to have known. As I understood the defendants' submissions, it was not really disputed that the risk of smoke taint from the burning of DPHB8 was foreseeable.

494 The second question under s 5B(1)(b) is whether the risk of smoke taint to grapes was 'not insignificant'. In adopting the language of 'not insignificant', the legislature seems to have eschewed the language of the common law in Wyong Shire Council v Shirt [1980] HCA 12; (1980) 146 CLR 40, 46 - 47, concerning the existence of a risk which is 'real' in the sense that it is 'not far-fetched or fanciful'. (This point is helpfully discussed by Dominic Villa, Annotated Civil Liability Act 2002 (NSW) (2004) [1A.5B.070].) Whilst recognising that aspect of the use of the language in s 5B(1)(b), the word 'insignificant' is an ordinary English word and, it seems to me, no utility is gained by searching for synonyms to unlock its meaning. An evaluative judgment needs to be made as to whether, on the evidence, the relevant risk is 'not insignificant'.

495 In relation to s 5B(1)(b), in my view, the following evidential considerations are relevant. First, there was good evidence, both anecdotally from the reports of vignerons affected by the fires in south-eastern Australia in early 2003, and from the AWRI Report, that the risk of smoke taint to grapes from fires with the intensity and smoke level of bushfires was at least not insignificant. Secondly, notwithstanding Mr Godden's statement, in effect, that he was unaware of the level of risk from prescribed burning, the evidence was that a core ignition would involve a substantial intensity of fire and associated high levels of smoke. I infer that a core ignition burn of over 500 ha would exhibit characteristics of fire and smoke intensity which would not be dissimilar to those associated with a bushfire. Thirdly, there is the fact that, as Mr Sneeuwjagt's evidence indicated, within a day, and in the course of a burn over several days, there is a significant chance of the occurrence of swings and changes in wind direction. Accordingly, even with the core ignition scheduled for a day with favourable winds in the north-east quadrant, and with forecast easterly winds on 2 April, there was a not insignificant risk that the winds would move around from time to time in the course of the edging and core ignitions. Also, whilst the core ignition of DPHB8 would be completed on 1 April 2004, it was expected that two further days would be needed to mop-up. During this time also the smoke could be expected to continue to drift. Fourthly, there was the evidence that, regardless of the wind direction during the day, there was a potential for the smoke to return and linger during the evening as a result of being trapped by an overnight inversion, the occurrence of which is not uncommon in the region. Fifthly, the previous prescribed burn adjacent to the eastern boundary of the plaintiffs' vineyards, the smoke from which covered the vineyards, had created a smoke taint problem. Sixthly, the plaintiffs' grapes were within one of the stages of development in which susceptibility to smoke uptake was not insignificant.

496 For these reasons, I would conclude that the burn of DPHB8 presented a not insignificant risk of smoke taint to the plaintiffs' grapes.

497 The third question is whether a reasonable person in the position of the Department would have taken the precautions that the plaintiffs allege should have been taken: s 5B(1)(c) of the Civil Liability Act; Adeels Palace Pty Ltd v Moubarak [28]. That question is to be determined prospectively, without the use of hindsight: Adeels Palace Pty Ltd v Moubarak [40].

498 When considering the application of s 5B(1)(c), s 5B(2) requires the court to consider, but not confine its consideration to, the matters in s 5B(2)(a), (b), (c) and (d). Under this provision the court would be entitled, in my view, to take into account matters relevant to breach of duty under the common law, except to the extent that their potential application is inconsistent with s 5B or any other relevant provisions of the Civil Liability Act, properly construed.

499 In relation to the question of whether a reasonable person in the Department's position would have taken the precaution of deferring the burn of DPHB8 until after harvest, or otherwise rolling it over to another season, in my view, the matters referred to in [406] - [413] above are relevant. The Department had, in the words of Mason J in Wyong Shire Council v Shirt (47) 'other conflicting responsibilities', as a result of which a reasonable person in the Department's position would not have deferred the burn as alleged.

500 In terms of the particular factors referred to in s 5B(2), my findings are as follows. As to s 5B(2)(a), the risk of smoke taint to grapes, whilst not insignificant, would not, in my view, be regarded as high, in light of the absence of scientific evidence connecting prescribed burning to smoke taint, and in light of the Department's previous experience of prescribed burning in the winegrowing areas of south-western Australia in the previous 40 years. As to s 5B(2)(b), the nature of the potential harm from the core ignition, were such harm to occur, was, in my view, serious in that smoke taint could well destroy the economic value of the plaintiffs' harvest for that year. With respect to s 5B(2)(c) and (d), those factors bring to account the matters in [406] - [413] to which I have referred in [499] above. In my view, they indicate that the burden of deferring the burn (including the wasted opportunity when opportunities are limited), and the important social utility of carrying it out, outweighed the risk of harm to the plaintiffs' grape crop for a year.

501 I find that having regard to the above matters, a reasonable person in the Department's position would not have 'rolled-over' or otherwise deferred the burning of DPHB8.

502 In relation to the alleged failure to take the precaution of deferring the commencement of the burn from 31 March to 1 April, the matters to which I have referred in [414] - [428] are relevant and are to be considered in the context of the matters in [406] - [413]. The probability that harm would occur from the edging burn was low. Given that the fire intensity and associated smoke level from an edging burn is of quite a different order from that of a core ignition, an edging burn on south to south-easterly winds would not carry the kind of dense smoke associated with a bushfire, or core ignition. Thus, in itself, the edging burn does not carry a high or even significant probability of harm. Moreover, the deferral of the edging burn for a day would not alter the fact that a core ignition, with its high density of smoke, would still be required. In light of the evidence of the uptake of smoke in grapes, and the possibilities of changes in wind direction, and of inversion, the deferral of the edging burn for a day would not, in the context of the burn in its entirety, have materially altered the probability of the risk of harm to the plaintiff's grapes occurring. If harm were to occur, however, it would be serious as it could damage the whole grape crop for the year. The burden of deferral for a day was to run the risk of not completing the burn before the arrival of north and north-west winds, and thereby run the risk of losing the opportunity to do the burn or, if the burn went ahead, run the risk of it not being completed safely. The carrying out of a safe and effective burn was an activity of high social utility. In my view, no reasonable person in the Department's position would have deferred the burn for a day as alleged.

503 In relation to the alleged failure to consult meaningfully with the plaintiff, that allegation in my view fails for two reasons. First, I find that Mr Marshall and Mr Gillard both consulted the plaintiffs concerning the proposed burn of DPHB8, and that the consultations were meaningful in that they resulted in amendments to the planning of the prescribed burn, each of which was designed to minimise smoke over the vineyards. The second, and final amendment to the burn prescription, also resulted in the procurement of additional resources to manage the extra risks involved in undertaking the burn on easterly or north easterly winds in order to minimise smoke over the plaintiffs' vineyards. Secondly, even if I am wrong in finding that there was meaningful consultation, in my view a reasonable person in the Department's circumstances would not have engaged in meaningful consultation with a view to delaying the burn, having regard to the matters referred to in [499] - [502] above.

504 It was suggested by Dr Tolhurst, in effect, that the burn of DPHB8 could have been managed differently, and better, from the plaintiffs' perspective, by breaking up the area of DPHB8 into smaller parcels and burning each separately. Multi-stage burning is routinely undertaken by the Department in the south-west. It has, however, its disadvantages. The experts in their conferral (including Dr Tolhurst) agreed: 'Multi-stage burning requires a greater level of resources to complete than single stage burns and carries additional risks'. The defendants' witnesses (Mr Keppel ts 557, 588; Dr McCaw ts 875; Mr Sneeuwjagt ts 925 - 926 and Mr Cheney ts 1004) did not accept the proposition as a reasonable method of burning in this case. Their evidence, which I accept, indicated that larger burns maximise burn efficiency and fuel reduction, and that a series of smaller burns requires finding more windows of opportunity for burning, when opportunities are in any event very limited. Multiple burns also increase the risks of escape, and involve inefficient use of limited resources. Apart from Dr Tolhurst's suggestion that a multi-stage approach would have been a 'viable option' for the burn of DPHB8, there was, in my view, no cogent evidence to suggest that the probability and likely seriousness of harm to the plaintiffs' grapes would, in any event, be lessened were a multi-stage approach adopted. The evidence of the 'wine' experts was that the effect of smoke exposures on grapes is additive and even small exposures may contribute to the final taint observed in wine (see [262] above). Furthermore, consideration of the burdens involved in adopting such a method, referred to above, and the social utility of burning in a safe and effective way, confirms, in my mind, the evidence of the defendants' witnesses that a multi-stage burn was not a reasonable method of burning in the circumstances. Accordingly, I find that a reasonable person in the Department's position would not have adopted such a method.

505 The circumstances to which I have referred in the preceding paragraphs are consistent with Mr Muller's evidence, which I have accepted, that, in relation to the burn of DPHB8, reasonable steps were taken to minimise smoke over the vineyards and that a prudent fire manager would have taken advantage of the window of opportunity on 31 March 2004 to burn DPHB8. If there were a duty of care owed to the plaintiffs, in my view, the Department exercised reasonable care in the circumstances.

506 I also find, as a fact, that the burn was carried out in accordance with the burn prescription, and in accordance with the standards and procedures applied by the Department in connexion with its Master Burn Plan. These included the avoidance of burns on north to north-west winds, particularly in respect of large areas of land. With respect to the plaintiffs' allegation that the burn prescription required the burn to be commenced on east-south-east to easterly winds, I have found that this allegation mischaracterises Mr Gillard's amendment, which contemplated the burn to be completed (not commenced) on winds of that kind. I should also add here that, as I understand it, the plaintiffs do not complain that the burn was undertaken with a view to it being completed on forecast easterly winds with a northerly component, rather than on easterly winds with a southerly component as Mr Gillard's amendment had specified. Even if this complaint had been made, the burn prescription, construed purposively, would not in my view have precluded the core ignition from being undertaken on easterly winds with a northerly component on the second day of the weather cycle and, in any event, the burn of DPHB8 was approved by Fire Management Services. In construing the burn prescription, it should be read as conveying the meaning which a reasonable person in the Department involved in the burn would understand it to mean. In this regard, the experts' agreement, in the following terms, is of relevance:

Prescribed burning prescriptions can be applied with some flexibility and discretion provided that approval to vary the prescription is received at the appropriate level. Professional judgment should be used to assess relevant risk mitigation or other operational factors ...

507 I also accept Mr Sneeuwjagt's evidence as to the object and purpose of the burn prescription as amended so far as it concerned the plaintiffs' position:

... The fact of the matter is, Mr Sneeuwjagt, that you authorised the burn to be carried out contrary to both the note affixed to the prescription by Mr Marshall and the note affixed to the prescription by Mr Gillard. That's the fact, isn't it?---No, it's not the fact. I think you are trying to put words in my mouth here. As I said to you, I agreed with the edging lighting to be undertaken under those south-south-east winds, knowing full-well, based on the forecast that more favourable weather was ahead to allow the bulk of the smoke to go away, thereby minimising the impact. Which was always our objective.

508 Moreover, winds from the east without a southerly component would likely have benefitted the plaintiffs rather than winds from the east with a southerly component.

509 The above matters indicate that there was a proper exercise of the Department's functions, and provide grounds for confirming that if, contrary to my conclusions, a duty of care exists, it was not breached: s 5W(d) of the Civil Liability Act.

510 Furthermore, if my conclusions above are erroneous, the decision to burn DPHB8, when it was burnt, was, in essence, a decision based substantially on political or social factors or constraints, and was accordingly a policy decision within the meaning of s 5X of the Civil Liability Act. On either basis upon which the plaintiffs put their case, namely, the deferral of the burn until after harvest or some other unspecified time, or a deferral from 31 March to 1 April, the gravamen of the complaint is that an appreciation of the risk of harm to the plaintiffs' grapes should have led to the burn not being carried out when it was. The question of whether the risk of smoke taint to the grapes, and thus to the economic interests of winegrowers in the region, should be determinative in deciding whether to proceed in accordance with the prescribed burning programme for 2004, was the very matter which the Minister was asked to consider. The plaintiffs' particular request for deferral was also before the Minister. The decision to burn in accordance with the prescribed burning programme irrespective of the risk of damage to grape crops in early 2004, was, in my view, clearly a decision based substantially on political and social factors. This is so having regard both to the nature of the decision, and the circumstances in which it was made. It was made against the broad background of the political history of prescribed burning in this State, and in the context of the particular political sensitivities prevalent at that time, as revealed in the Minister's statements to the House. The operational decision to burn on 31 March 2004 gave effect to the Minister's decision.

511 Having regard to the matters in [406] - [428], it was not, in my view, a decision so unreasonable that no reasonable public body or officer could have made it. In these circumstances, any fault by the Department in deciding to carry out the burn on 31 March 2004, or in failing to defer the burn, was, in substance, a policy decision in the performance of a public function. Such a policy decision cannot be used to support a finding of fault, by reason of s 5X of the Civil Liability Act. Accordingly, in my view, to make a finding of liability on the basis that the defendants failed to defer the burn, would be contrary to s 5X.

512 The potential operation of s 132 of the CALM Act draws attention to the precise basis or bases upon which the plaintiffs contend that each defendant is liable. As noted earlier, the particularised plea of negligence in the statement of claim refers to negligence by the 'second defendant [ie, the State] and its servants or agents'. For the purposes of this plea, I understand that the plaintiffs' contention is that the Executive Officer, as a corporate body, is an agent of the Crown in right of the State of Western Australia. The plea, insofar as it refers to 'servants or agents' might be thought to suggest the State is alleged to be vicariously liable for the conduct of the employees in the Department ('servants') and the conduct of the Executive Director (as an 'agent'). As noted below, however, the plaintiffs appear to disavow any reliance on vicarious liability.

513 The plaintiffs, in their submissions, generally used the word 'defendants' as a compendious term, without discrimination between the two defendants, in advancing their case in negligence in terms of both breach and duty. Reference to each defendant was, however, separately made in relation to submissions on s 132. The plaintiffs' submissions (page 17 closing submissions) in that regard said:

Section 132 of the CALM Act does not avail either of the Defendants as a defence to their breach of their common law duty to the Plaintiffs. S 132 only extends to the 'personal liability' of an individual; not to the liability of the State, qua the State, nor to the common law duty of care owed by the Department, that is, the Executive Director in his corporate capacity, not to cause harm to the Plaintiffs in the exercise of its statutory duty or powers; that is to say, the common law duty which the Executive Director owes to the Plaintiffs by virtue of the principles stated and applied, for example, in Crimmins, Pyrenees, Burnie and see Leichhardt Municipal Council v Montgomery (1997) 230 CLR 22; [2007] HCA 6 as to non-delegable duty and note per Gleeson CJ at [18] regarding the non-delegable duty that arises from the conduct of extra-hazardous activities, which is the case with respect to prescribed burning. The Executive Director is not sued personally but in his statutory corporate capacity as it was at the time of the DPHB8 Burn; see NSW v West [2008] ACTCA 14; (2008) 165 ACTR 47, (2008) Aust Tort Reports, 81-967; [2008] ACTCA 14; and the State is sued because the Department through the corporate capacity of the Executive Director, is an agency of the State which is responsible, and liable, for any breach of duty by its Agency. This is not a case of vicarious liability, either by the Department or the State; cf Bell v State of Western Australia [2004] WASCA 205.

Notwithstanding the last sentence in the above quote, the penultimate sentence might be thought to suggest the concept of vicarious liability.

514 The plaintiffs' senior counsel, in closing oral submissions, also said:

We are not relying on a case of vicarious liability ... this is not a case where the plaintiff relies on the personal liability of an individual; rather, it is relying upon the liability of the department for a breach of its duty to carry out its function with proper care.

515 The plaintiffs also referred, in closing submissions, to the existence of a non-delegable duty of care owed by the Department in accordance with authorities such as Burnie Port Authority v General Jones Pty Ltd.

516 For the reasons given earlier at [441], in my view, this case does not involve the existence or otherwise of a non-delegable duty of care as explained and applied in cases such as Burnie Port Authority v General Jones Pty Ltd. Further, as the Executive Director, as a corporate body, is an agent of the Crown in right of the State of Western Australia, in my view, the State could only be liable for any tort committed by the Executive Director on the basis of vicarious liability. Also, the Executive Director as a corporate body would not, in the circumstances, in my view, be liable for the acts of the Department's employees. Rather, the State of Western Australia, and not the Executive Director, would be vicariously liable for the tortious conduct of employees within the Department: see Hogg P & Monahan PJ, Liability of the Crown, (3rd ed) 119, 341 - 342; and State of Western Australia v Bond Corporation Holding Ltd (1991) 5 WAR 40, 66 - 67, in relation to government employees.

517 It was not clear whether, and if so how, the plaintiffs were putting a case based on vicarious liability and, as I have found, a non-delegable duty of care as postulated by the plaintiffs has not been established.

518 Having regard to my earlier findings on duty and breach, the proper construction of s 132 of the CALM Act is not ultimately determinative of this action, and the matters referred to above have meant that the debate on s 132 has lacked some clarity and focus. It is, accordingly, unnecessary to come to a concluded view on the proper construction of s 132. It is not necessary to consider whether, in particular, having regard to the relevant principles of construction, an express reservation of vicarious liability is required or whether an implied reservation is sufficient, and whether the use of the word 'personally' in s 132 signifies an implied reservation of vicarious liability in the State of Western Australia. See generally Bell v The State of Western Australia [2004] WASCA 205 [4], [34] and [35]; Nada v Knight (1990) Aust Tort Rep 81-032 at 67,918; Hogg & Monahan, Liability of the Crown, 120, Puntoriero v Water Administration Ministerial Corporation [1999] HCA 45; (1999) 199 CLR 575.

519 In light of the above reasons, I find that a duty of care, as alleged by the plaintiffs, or otherwise, did not exist. If such a duty did exist, I find that it was not breached. Accordingly, the plaintiffs' action must be dismissed.