Secretary, Department of Planning and Environment v Keelendi Farms Pty Ltd;... [2026] NSWLEC 47 | Case summary

Catchwords

ENVIRONMENTAL OFFENCES: corporate defendants charged with the clearing of native vegetation across two properties – two corporate defendants charged as landholders – third corporate defendant is the entity that managed the farms – individual defendant is the director of the third corporate defendant and was charged with an executive liability offence. ENVIRONMENTAL OFFENCES: alleged clearing of native vegetation by the application of herbicide – subject land concerned two properties with a history of farming and cropping – whether prosecutor permitted to also rely on evidence of the mechanical clearing of native vegetation – elements of the offence – whether defendants cleared vegetation in the charge areas – whether the defendants cleared vegetation in the charge areas during the charge periods – whether defendants cleared native vegetation or whether the defendants cleared weeds or regrowth – whether corporate defendants caused or permitted the commission of an offence – prosecutor not able to prove beyond reasonable doubt that defendants cleared native vegetation in the charge areas during the charge period – defendants found not guilty – charges dismissed. ENVIRONMENTAL OFFENCES: executive liability offence – elements of the offence – whether prosecutor failed to demonstrate beyond reasonable doubt that the individual defendant knew or ought reasonably to have known that the offence was being committed and failed to take all reasonable steps to prevent or stop the commission of the offence – where individual defendant told pilot spraying aerial herbicide not to spray the timbered areas. STATUTORY CONSTRUCTION: meaning of “clearing” of native vegetation in the Local Land Services Act 2013. STATUTORY CONSTRUCTION: meaning of “reasonable steps” in s 13.6 of the Biodiversity Conservation Act 2016.

Judgment

Source: NSW Caselaw (www.caselaw.nsw.gov.au). Converted to plain text by ZoeLaw. Not the official version.

Medium Neutral Citation: Secretary, Department of Planning and Environment v Keelendi Farms Pty Ltd; Secretary, Department of Planning and Environment v Timothy Justin O’Brien; Secretary, Department of Planning and Environment v T J O’Brien Investments Pty Ltd; Secretary, Department of Planning and Environment v Edenmore Farms Pty Ltd (No 3) [2026] NSWLEC 47

Hearing dates: 7, 10, 14, 18, 19, 20, 21, 24, 25, 26, 27 and 28 February 2025 and 24 April 2026 (further written submissions)

Date of orders: 30 April 2026

Decision date: 30 April 2026

Jurisdiction: Class 5

Before: Pepper J

Decision: The orders of the Court are as follows:
(1) the proceedings are stood over for mention to 9.30am on 26 May 2026 before Pepper J; and
(2) liberty to restore on three days’ notice.

Catchwords: ENVIRONMENTAL OFFENCES: corporate defendants charged with the clearing of native vegetation across two properties – two corporate defendants charged as landholders – third corporate defendant is the entity that managed the farms – individual defendant is the director of the third corporate defendant and was charged with an executive liability offence.
ENVIRONMENTAL OFFENCES: alleged clearing of native vegetation by the application of herbicide – subject land concerned two properties with a history of farming and cropping – whether prosecutor permitted to also rely on evidence of the mechanical clearing of native vegetation – elements of the offence – whether defendants cleared vegetation in the charge areas – whether the defendants cleared vegetation in the charge areas during the charge periods – whether defendants cleared native vegetation or whether the defendants cleared weeds or regrowth – whether corporate defendants caused or permitted the commission of an offence – prosecutor not able to prove beyond reasonable doubt that defendants cleared native vegetation in the charge areas during the charge period – defendants found not guilty – charges dismissed.
ENVIRONMENTAL OFFENCES: executive liability offence – elements of the offence – whether prosecutor failed to demonstrate beyond reasonable doubt that the individual defendant knew or ought reasonably to have known that the offence was being committed and failed to take all reasonable steps to prevent or stop the commission of the offence – where individual defendant told pilot spraying aerial herbicide not to spray the timbered areas.
STATUTORY CONSTRUCTION: meaning of “clearing” of native vegetation in the Local Land Services Act 2013.
STATUTORY CONSTRUCTION: meaning of “reasonable steps” in s 13.6 of the Biodiversity Conservation Act 2016.

Legislation Cited: Biodiversity Conservation Act 2016, ss 13.6, 13.29(1)
Criminal Appeal Act 1912, s 5AE
Criminal Procedure Act 1986, s 29(1)(b)
Evidence Act 1995, Pt 3.1
Local Land Services Act 2013, ss 3(e), 60B, 60C, 60D, 60F, 60N
Local Land Services Regulation 2014, cl 106
Native Vegetation Act 2003, ss 7(b), 44
Protection of the Environment Administration Act 1991, s 6(2)
State Environmental Planning Policy (Vegetation in Non-Rural Areas) 2017, cl 4

Cases Cited: Alphapharm Pty Ltd v H Lundbeck A/S [2014] HCA 42; (2014) 254 CLR 247
Director-General, Department of Environment and Climate Change v Walker Corporation Pty Limited (No 2) [2010] NSWLEC 73
Environment Protection Authority v Blacktown Waste Services Pty Ltd [2025] NSWLEC 31
Environment Protection Authority v Charlotte Pass Snow Resort Pty Ltd [2021] NSWLEC 37
Environment Protection Authority v Eastern Creek Operations Pty Ltd (No 2) [2021] NSWLEC 39
Environment Protection Authority v Grafil Pty Ltd; Environment Protection Authority v Mackenzie [2019] NSWCCA 174; (2019) 101 NSWLR 245
Environment Protection Authority v Land and Environment Court of NSW [2004] NSWCA 50; (2004) 144 A Crim R 198
Environment Protection Authority v Maules Creek Coal Pty Ltd (No 3) [2024] NSWLEC 97; (2024) 262 LGERA 350
Environment Protection Authority v O’Brien [2023] NSWLEC 118
Environment Protection Authority v Pullinger (No 2) [2024] NSWLEC 51
Environment Protection Authority v Terrace Earthmoving Pty Ltd & Page [2012] NSWLEC 216
Jago v District Court of New South Wales [1986] HCA 46; (1989) 168 CLR 23
Johnson v Miller [1937] HCA 77; (1937) 59 CLR 467
King v The Queen [1986] HCA 59; (1986) 161 CLR 423
Master Education Services Pty Ltd v Ketchell [2008] HCA 38; (2008) 236 CLR 101
McNamara v The King [2023] HCA 36; (2023) 280 CLR 201
Orr v Cobar Management Pty Ltd [2020] NSWCCA 220
Secretary, Department of Planning, Industry and Environment, v Auen Grain Pty Ltd; Greentree Merrywinebone Pty Ltd; Harris (No 7) [2022] NSWLEC 153
The Queen v Baden-Clay [2016] HCA 35; (2016) 258 CLR 308
The Queen v Hillier [2007] HCA 13; (2007) 228 CLR 618
Turnbull v Chief Executive of the Office of Environment and Heritage [2017] NSWCA 161
Walker Corporation Pty Limited v Director-General, Department of Environment, Climate Change and Water [2012] NSWCCA 210; (2012) 82 NSWLR 12

Category: Principal judgment

Parties: Secretary, Department of Planning and Environment (Prosecutor)
Keelendi Farms Pty Ltd (Defendant)
T J O’Brien Investments Pty Ltd (Defendant)
Edenmore Farms Pty Ltd (Defendant)
Timothy Justin O’Brien (Defendant)

Representation: Counsel:
H El-Hage SC with T Hammond (Prosecutor)
J Ireland KC (Solicitor) and A Connolly (Defendants)
Solicitors:
Department of Planning and Environment (Prosecutor)
McGirr Lawyers (Defendants)

File Number(s): 2022/178680, 2022/178681, 2022/178682, 2022/178683, 2022/178684, 2022/178685, 2022/178686, 2022/178687, 2022/211961, 2022/211962, 2022/211963, 2022/211964

Publication restriction: Nil

JUDGMENT

TABLE OF CONTENTS

TOPIC PARAGRAPH NO

The Defendants are Charged with the Unlawful Clearing of Native Vegetation by Poisoning [1]

Joint Trial [7]

The Charges Summarised [12]

The Offences [14]

Statutory Framework Creating the Offences [14]

Burden of Proof in Circumstantial Cases [27]

Elements of the Native Vegetation Clearing Offences [31]

Elements of the Executive Liability Offence [32]

Evidence Relied Upon by the Parties [33]

The Secretary’s Evidence [34]

Section 60F Certificates [38]

Lay Evidence [41]

Mark Smith [42]

Christine Shrimpton [58]

Scott Beaumont [66]

Spray Tech Documents [69]

O’Brien Interview [83]

ASIC search [87]

Susan Rea [88]

Expert Evidence [94]

Leo Watts [94]

Prof Richard Thackway [130]

Dr Kathryn Hammill [141]

Defendants’ Evidence [162]

Was There “Clearing” of “Native Vegetation” in the Charge Areas During the Charge Periods? [180]

Method of Clearing Particularised by the Prosecutor [180]

The Meaning of the Term “Clearing” [187]

Did “Clearing” of Vegetation Occur in the Charge Areas? [198]

Did the Method of Clearing by the Application of Herbicide Occur During the Charge Periods? [218]

Was There Clearing of “Native Vegetation”? [238]

Was the Clearing in a “Regulated Rural Area”? [248]

Did the Defendants Carry Out or Were Responsible for the Carrying out the Native Vegetation Clearing? [249]

Keelendi Farms and Edenmore Farms [249]

Charges Against T J O’Brien [254]

Executive Liability Charge Against O’Brien [256]

The Defendants Are Not Guilty of the Offences as Charged [257]

Adjournment Application by the Prosecutor to Consider the Making of a Stated Case [268]

Orders [281]

The Defendants are Charged with the Unlawful Clearing of Native Vegetation by Poisoning

Located on the charmingly named Come By Chance Road, Pilliga, are two farming properties known as “Keelendi” and “Edenmore” (“the properties”). These proceedings concern the alleged unlawful clearing of native vegetation by four defendants by the application of herbicide to poison the native vegetation on the properties. All defendants have pleaded not guilty to the charges against them.
By 12 summonses filed on 17 June 2022, the prosecutor, the Secretary, Department of Planning and Environment (“the Secretary”), commenced related criminal proceedings against Edenmore Farms Pty Ltd (“Edenmore Farms”), Keelendi Farms Pty Ltd (“Keelendi Farms”), T J O’Brien Investments Pty Ltd (“T J O’Brien”) and Timothy O’Brien (“O’Brien”) (together, “the defendants”) in Class 5 of the Court’s jurisdiction. To all intents and purposes, O’Brien was, according to the Secretary, responsible for the alleged unlawful clearing as a director of T J O’Brien.
The charges against the corporate defendants are brought under s 60N of the Local Land Services Act 2013 (“LLSA”), whereas the executive liability charge against O’Brien is brought under s 13.6 of the Biodiversity Conservation Act 2016 (“BCA”).
The defendants were prosecuted in several sets of proceedings that were the subject of two separate trials that were heard concurrently in the Court. The 12 summonses filed by the Secretary relate to two charge periods. The charges in the first trial allege offending during the period from 13 to 23 March 2020 inclusive (“the later charge period”) and the charges in the second trial allege offences occurring between 13 February and 13 March 2020 (“the earlier charge period”).
The charges are particularised in the summonses and in the prosecutor’s Amended Statement of Facts. The original Statement of Facts was filed on 10 March 2023. In the summonses, the method of clearing is particularised as “the native vegetation was cleared by means of poisoning”. No other method, such as mechanical clearing, is described.
For the reasons in this judgment, I find that the Secretary has failed to prove beyond reasonable doubt that the defendants cleared native vegetation by the application of herbicide in the charge areas during the charge periods nominated in the summonses. As a consequence, I find all four defendants not guilty of the offences with which they have been charged.

Joint Trial

The proceedings were heard together. As a consequence, on 26 February 2025 the Court ordered that evidence in the first trial be evidence in the second trial. It did so on the basis that, contrary to the initial submission by the defendants, there was nothing inadmissible about a single body of evidence across multiple charges being heard together. As the High Court of Australia stated in McNamara v The King [2023] HCA 36; (2023) 280 CLR 201 (at [67]-[69] per Gageler CJ, Gleeson and Jagot JJ, footnotes omitted):

67 Sections 55(1) and 56(1) combine to produce the result that evidence probative of the existence of one fact in issue in a proceeding is admissible in the totality of the proceeding. The evidence, if admitted, is then available to be used in the totality of the proceeding in the assessment of the probability of the existence of any other fact in issue in the proceeding, including any fact in issue between parties to the proceeding other than the party adducing it, unless and to the extent that its admission is excluded, or its use is limited, by or under another provision of the Evidence Act.

68 The combined operation of ss 55(1) and 56(1) in their application to a multi-party civil proceeding was explained by Austin J in Australian Securities and Investments Commission v Vines. His Honour said:

“It is notable that both ss 55 and 56 address the question whether evidence is admissible in a proceeding. Where a plaintiff seeks to make out separate cases against several defendants in a single proceeding, the question to which the Evidence Act provides an answer is whether evidence is admissible in the proceeding, not whether evidence is admissible to prove the plaintiff’s case against a particular defendant. The answer it gives is that if the evidence is relevant, it is admissible. Once it is admitted, it is evidence in the proceeding, and therefore available to be used for any purpose, unless one of the exclusionary rules of the Act or any surviving general law exclusionary rule applies, or the Court makes use of its statutory discretions to exclude admissible evidence or limit its use.”

69 The combined operation of ss 55(1) and 56(1) can be no different in their application to a multi-party criminal proceeding. For the purposes of s 55(1), the facts in issue in a criminal proceeding are the ultimate facts which establish the elements of the offence or offences charged together with such other facts the existence of which may be probative of the existence of those ultimate facts. Applied to a criminal proceeding constituted by a joint trial on a joint indictment, ss 55(1) and 56(1) produce the result that any evidence adduced on the basis that it is probative of the existence of a fact in issue between the Crown and a co-accused is available to be used by the jury in assessing the probability of the existence of any other fact in issue between the Crown and that co-accused or between the Crown and any other co-accused, unless and to the extent that the admission of the evidence is excluded, or the use of the evidence is limited, by or under some other section. That is so for all evidence adduced in the joint trial, whether by the Crown or any co-accused and whether in examination in chief or in cross-examination.

The admissibility of a single body of evidence across multiple charges heard together has been confirmed on numerous occasions in this Court (Environment Protection Authority v O*’*Brien [2023] NSWLEC 118 at [110] per Robson J; Environment Protection Authority v Pullinger (No 2) [2024] NSWLEC 51 at [38] per Pritchard J and Environment Protection Authority v Blacktown Waste Services Pty Ltd [2025] NSWLEC 31 at [31]-[33]).
It should be recalled in this context, that s 29(1)(b) of the Criminal Procedure Act 1986 provides that proceedings relating to two or more offences alleged to have been committed by the same person may be heard and determined together where the offences arise out of the same factual matrix:

29 When more than one offence may be heard at the same time

(1) A court may hear and determine together proceedings related to 2 or more offences alleged to have been committed by the same accused person in any of the following circumstances—

…

(b) the offences arise out of the same set of circumstances,

The defendants ultimately resiled from their earlier opposition and were content for the Court to treat the evidence in the first set of proceedings to be evidence in the second set of proceedings (T316:37-38).
Although the proceedings were heard jointly, as a matter of administrative convenience the Court must nevertheless consider the case against, and return a verdict in respect of, each defendant separately. Moreover, it is not permissible for the Court to use evidence that was tendered only against one or more defendants to make findings against the remaining defendant or defendants. For example, the record of interview (“ROI”) of O’Brien was tendered against the corporate defendants and not against O’Brien individually.

The Charges Summarised

A summary of the charges is as follows:

| Proceeding an****d defendant and date of alleged clearing | Polygon number, area allegedly cleared and Lot/DP numbers | | 2022/178680 Keelendi Farms Pty Ltd (alleged clearing on Keelendi between 13 February – 13 March 2020) | Polygon 1 (lots 1, 5, 17 of 750249; 35, 36, 37 of 750268) – 813 ha Polygon 14 (lots 2, 6, 13, 14, 32, 33, 34, 36, 37 of 750293) – 43 ha Polygon 15 (lot 2 of 750293) – 27 ha Total area cleared – 883 ha | | 2022/178681 TJ O’Brien Investments Pty Ltd (alleged clearing on Keelendi between 13 February – 13 March 2020) | Polygon 1 (lots 1, 5, 17 of 750249; 35, 36, 37 of 750268) – 813 ha Polygon 14 (lots 2, 6, 13, 14, 32, 33, 34, 36, 37 of 750293) – 43 ha Polygon 15 (lot 2 of 750293) – 27 ha Total area cleared – 883 ha | | 2022/211961 Timothy O’Brien (in connection with primary offence in proceedings 2022/178681) (alleged clearing on Keelendi, between 13 February – 13 March 2020) | Polygon 1 (lots 1, 5, 17 of 750249; 35, 36, 37 of 750268) – 813 ha Polygon 14 (lots 2, 6, 13, 14, 32, 33, 34, 36, 37 of 750293) – 43 ha Polygon 15 (lot 2 of 750293) – 27 ha Total area cleared – 883 ha | | 2022/178682 Edenmore Farms Pty Ltd (alleged clearing on Edenmore between 13 February – 13 March 2020) | Polygon 19 (lots 54, 55 of 750301) Total area cleared – 415 ha | | 2022/178683 TJ O’Brien Investments Pty Ltd (alleged clearing on Edenmore between 13 February – 13 March 2020) | Polygon 19 (lots 54, 55 of 750301) Total area cleared – 415 ha | | 2022/211962 Timothy O’Brien (in connection with primary offence in proceedings 2022/178683) (alleged clearing on Edenmore between 13 February – 13 March 2020) | Polygon 19 (lots 54, 55 of 750301) Total area cleared – 415 ha | | 2022/178684 Keelendi Farms Pty Ltd (alleged clearing on Keelendi between 13 – 23 March 2020) | Polygon 13 (lots 8, 10 and 11 of 750293) Total area cleared – 118 ha | | 2022/178685 TJ O’Brien Investments Pty Ltd (alleged clearing on Keelendi between 13 – 23 March 2020) | Polygon 13 (lots 8, 10 and 11 of 750293) Total area cleared – 118 ha | | 2022/211963 Timothy O’Brien (in connection with primary offence in proceedings 2022/178685) (alleged clearing on Keelendi between 13 – 23 March 2020) | Polygon 13 (lots 8, 10 and 11 of 750293) Total area cleared – 118 ha | | 2022/178686 Edenmore Farms Pty Ltd (alleged clearing on Edenmore between 13 – 23 March 2020) | Polygon 20 (lots 31, 54 and 55 of 750301) – 185 ha Polygon 21 (lots 33, 34 and 35 of 750301) – 138 ha Polygon 24 (lot 1 of 119409 and lot 40 of 750301) – 92 ha Polygon 25 (lot 1 of 119409) – 31 ha Polygon 26 (lots 48, 49 and 60 of 750301) – 106 ha Total area cleared – 552 ha | | 2022/178687 TJ O’Brien Investments Pty Ltd (alleged clearing on Edenmore between 13 – 23 March 2020) | Polygon 20 (lots 31, 54 and 55 of 750301) – 185 ha Polygon 21 (lots 33, 34 and 35 of 750301) – 138 ha Polygon 24 (lot 1 of 119409 and lot 40 of 750301) – 92 ha Polygon 25 (lot 1 of 119409) – 31 ha Polygon 26 (lots 48, 49 and 60 of 750301) – 106 ha Total area cleared – 552 ha | | Timothy O’Brien (in connection with primary offence in proceedings 2022/178687) (alleged clearing on Edenmore between 13 – 23 March 2020) | Polygon 20 (lots 31, 54 and 55 of 750301) – 185 ha Polygon 21 (lots 33, 34 and 35 of 750301) – 138 ha Polygon 24 (lot 1 of 119409 and lot 40 of 750301) – 92 ha Polygon 25 (lot 1 of 119409) – 31 ha Polygon 26 (lots 48, 49 and 60 of 750301) – 106 ha Total area cleared – 552 ha |

The alleged unlawful clearing of native vegetation is said to have occurred over a considerable land area within the properties. An aerial image depicting the boundaries of the two properties (with lots), is below:

The Offences

Statutory Framework Creating the Offences

Section 60N(1) of the LLSA states that:

60N Unauthorised clearing of native vegetation in regulated rural areas—offence

(1) A person who clears native vegetation in a regulated rural area is guilty of an offence unless the person establishes any of the following defences—

(a) that the clearing is for an allowable activity authorised under Division 4 and Schedule 5A,

(b) that the clearing is authorised by a land management (native vegetation) code under Division 5,

(c) that the clearing is authorised by an approval of the Panel under Division 6,

(d) that the clearing is authorised under section 60O (Clearing authorised under other legislation etc),

(e) that the clearing is the carrying out of a forestry operation authorised under Part 5B (Private native forestry). …

An offence against s 60N is a strict liability offence. Any person who clears native vegetation in a “regulated rural area” is guilty of an offence unless they establish a relevant defence (Secretary, Department of Planning, Industry and Environment v Auen Grain Pty Ltd; Greentree Merrywinebone Pty Ltd; Harris (No 7) [2022] NSWLEC 153 at [418]).
A “regulated rural area” is defined in s 60D of the LLSA as, “any area of the State to which this Part [Pt 5A] applies that is category 2-regulated land”. The term “category 2-regulated land” is defined in the same section to mean:

category 2-regulated land means areas of the State to which this Part applies designated as category 2-regulated land on the native vegetation regulatory map (including category 2-vulnerable regulated land that is so designated).

Section 60B relevantly sets out the statutory meaning of the term “native vegetation”:

60B Meaning of “native vegetation”

(1) For the purposes of this Part, native vegetation means any of the following types of plants native to New South Wales—

(a) trees (including any sapling or shrub or any scrub),

(b) understorey plants,

(c) groundcover (being any type of herbaceous vegetation),

(d) plants occurring in a wetland.

(2) A plant is native to New South Wales if it was established in New South Wales before European settlement. The regulations may authorise conclusive presumptions to be made of the species of plants native to New South Wales by adopting any relevant classification in an official database of plants that is publicly accessible.

(3) For the purposes of this Part, native vegetation extends to a plant that is dead or that is not native to New South Wales if—

(a) the plant is situated on land that is shown on the native vegetation regulatory map as category 2-vulnerable regulated land, and

(b) it would be native vegetation for the purposes of this Part if it were native to New South Wales. …

According to s 60C of the LLSA, the action of “clearing” native vegetation means:

60C Meaning of “clearing” native vegetation

For the purposes of this Part, clearing native vegetation means any one or more of the following—

(a) cutting down, felling, uprooting, thinning or otherwise removing native vegetation,

(b) killing, destroying, poisoning, ringbarking or burning native vegetation.

At the time of the alleged commission of the offences, s 60F of the LLSA relevantly made provision for transitional arrangements until native vegetation regulatory maps had been designated:

60F Transitional arrangement until preparation of maps

(1) This section applies to an area of the State to which this Part applies during the period from the commencement of this Part until the area has been designated on a native vegetation regulatory map (the transitional period).

(2) For the purposes of this Part, the area is taken, during the transitional period, to be—

…

(b) category 2-regulated land, if this Part requires the land to be designated as category 2-regulated land on the native vegetation regulatory map (except as provided by paragraph (c)), …

Note–

Generally that section permitted clearing if the vegetation comprised less than 50% of indigenous species of vegetation.

…

(5) The Environment Agency Head may, in connection with any legal proceedings against a landholder, issue a certificate that the land described in the certificate is (for the reasons set out in the certificate) category 1-exempt land or category 2-regulated land (including category 2-vulnerable regulated land). The certificate is, in those legal proceedings, prima facie evidence of the category of the land during the transitional period.

Part 5A of the LLSA, and therefore, the transitional period, commenced on 25 August 2017.
It was not a matter of controversy that certificates were issued by the relevant delegate on 11 November 2021 pursuant to s 60F(5) of the LLSA, certifying that all relevant land the subject of these prosecutions was category 2-regulated land (“the s 60F certificates”).
Clause 106 of the Local Land Services Regulation 2014 (“LLS Regulations”) provided that for the purposes of Pt 5A of the LLSA, “native vegetation” further means:

106 Meaning of “native vegetation”****—species native to New South Wales (s 60B (2))

(1) For the purposes of Part 5A of the Act, a species of plant may be conclusively presumed to be native to New South Wales if it is listed on the official database, unless it is identified on that database as an introduced species only.

(2) In this clause—

the official database means the database of flora known as “New South Wales Flora Online”, maintained by the Royal Botanic Gardens and Domain Trust and published on the website of the Trust.

Under s 13.29(1) of the BCA, a landholder can be held liable for the unlawful acts of third parties:

13.29 Responsibility of landholder for activities carried out on the land

(1) In any criminal or civil proceedings under this Act, the landholder of any land on which an offence or contravention is alleged to have occurred is taken to have carried out the activity constituting the alleged offence or contravention unless it is established that—

(a) the activity was carried out by another person, and

(b) the landholder did not cause or permit the other person to carry out the activity.

This statutorily ascribed criminal liability is separate from any application of common law principles of vicarious liability (Director-General, Department of Environment and Climate Change v Walker Corporation Pty L**imited (No 2) [2010] NSWLEC 73 at [274]). Although in Walker the Court was considering s 44 of the now repealed Native Vegetation Act 2003 (“NVA”), s 44 was not materially dissimilar to s 13.29 of the BCA.
In respect of the charges against O’Brien, s 13.6 of the BCA provides for the executive liability of directors for offences committed by a corporation:

13.6 Liability of directors etc for offences by corporation—offences attracting executive liability

(1) For the purposes of this section, an executive liability offence is an offence against any of the following provisions that is committed by a corporation—

…

(e) section 60N of the Local Land Services Act 2013 (Unauthorised clearing of native vegetation in regulated rural area—offence),

…

(2) A person commits an offence against this section if—

(a) a corporation commits an executive liability offence, and

(b) the person is—

(i) a director of the corporation, or

(ii) an individual who is involved in the management of the corporation and who is in a position to influence the conduct of the corporation in relation to the commission of the executive liability offence, and

(c) the person—

(i) knows or ought reasonably to know that the executive liability offence (or an offence of the same type) would be or is being committed, and

(ii) fails to take all reasonable steps to prevent or stop the commission of that offence.

Maximum penalty—The maximum penalty for the executive liability offence if committed by an individual.

(3) The prosecution bears the legal burden of proving the elements of the offence against this section.

(4) The offence against this section can only be prosecuted by a person who can bring a prosecution for the executive liability offence.

(5) This section does not affect the liability of the corporation for the executive liability offence, and applies whether or not the corporation is prosecuted for, or convicted of, the executive liability offence.

…

(7) In this section—

director has the same meaning as in the Corporations Act 2001 of the Commonwealth.

reasonable steps, in relation to the commission of an executive liability offence, includes, but is not limited to, such action (if any) of the following kinds as is reasonable in all the circumstances—

(a) action towards—

(i) assessing the corporation’s compliance with the provision creating the executive liability offence, and

(ii) ensuring that the corporation arranged regular professional assessments of its compliance with the provision,

(b) action towards ensuring that the corporation’s employees, agents and contractors are provided with information, training, instruction and supervision appropriate to them to enable them to comply with the provision creating the executive liability offence so far as the provision is relevant to them,

(c) action towards ensuring that—

(i) the plant, equipment and other resources, and

(ii) the structures, work systems and other processes, relevant to compliance with the provision creating the executive liability offence are appropriate in all the circumstances,

(d) action towards creating and maintaining a corporate culture that does not direct, encourage, tolerate or lead to non-compliance with the provision creating the executive liability offence.

Relevantly, evidence that an officer of a corporation acting in that capacity had a particular state of mind is evidence that the corporation had that state of mind (s 13.8 of the BCA).

Burden of Proof in Circumstantial Cases

Because these are criminal proceedings, it is axiomatic that the burden rests upon the Secretary to prove beyond reasonable doubt every element and every essential fact that constitutes the offences charged. There is no obligation on the defendants to prove any fact or issue that is in dispute.
Where, as in the present proceedings, the prosecution case rests substantially on circumstantial evidence, the Court cannot return a guilty verdict unless the prosecutor has excluded all reasonable hypotheses (that is, that rests upon something more than mere conjecture) consistent with innocence (The Queen v Baden-Clay [2016] HCA 35; (2016) 258 CLR 308 at [45]-[50]).
All of the circumstances established by the evidence must be considered and weighed in deciding whether there is an inference consistent with innocence reasonably open on the evidence; the evidence must be considered as a whole (The Queen v Hillier [2007] HCA 13; (2007) 228 CLR 618 at [46]).
The prosecutor accepted that proving the facts of “clearing” and “native vegetation” were so fundamental to the process of reasoning as to the guilt of the defendants, that they must be proved beyond reasonable doubt.

Elements of the Native Vegetation Clearing Offences

In relation to the charge against s 60N of the LLSA, the Secretary must prove to the criminal standard the following four elements of the offence:

(a) first, that “clearing” (as defined in s 60C of the LLSA) occurred on the property. The Secretary relied upon spatial imagery and remote sensing to identify changes in vegetation appearance and health and changes to the landscape during the charge periods. The Secretary also relied on the evidence of witnesses who had visited the properties in June and August 2020 and June 2021, who observed browned-off, dead and dying vegetation at various locations on the properties. In addition, the Secretary relied upon the business records of a third party to demonstrate that there had been herbicide spraying on the properties between 10 and 15 March 2020; (b) second, that there was the clearing of “native vegetation” (as defined in s 60B). The Secretary relied upon experts who opined as to the species of vegetation that ought to have been present on lots the subject of the charges; (c) third, that the clearing occurred in a “regulated rural area” (as defined in s 60D). This was established by evidentiary certificates; and (d) fourth, that it was the defendants who carried out or were responsible for the clearing. For the purpose of the deeming provision contained in s 13.29 of the BCA, the ownership of the properties the subject of the clearing Keelendi Farms and Edenmore Farms was proved by registered proprietor searches. This was not in dispute. With respect to T J O’Brien, who was not a landowner, and therefore, against whom the deeming provision did not apply, the Secretary relied upon records produced by the third party who had applied herbicide by aerial spraying, to the effect that it had been engaged by T J O’Brien and that at all times it had acted on instructions from that corporate defendant. O’Brien, as a director of T J O’Brien, made admissions during his ROI that T J O’Brien was the entity responsible for the management of the properties.

Elements of the Executive Liability Offence

In relation to the executive liability offence brought against O’Brien pursuant to s 13.6 of the BCA, the Secretary must prove to the criminal standard that:

(a) first, a corporation has committed an executive liability offence such as an offence against s 60N of the LLSA; (b) second, the person is a director of a corporation or is involved in its management. Company searches established that at all relevant times, O’Brien was the director of T J O’Brien. This evidence was not the subject of challenge; (c) third, the person knows or ought reasonably to have known that the executive liability offence would be or was being committed. The Secretary sought to rely upon various business records from the third party contracted to spray herbicide to demonstrate that at all relevant times O’Brien knew that herbicide was being applied to the properties as charged; and (d) fourth, the person failed to take “all reasonable steps” (as defined in s 13.6(7) of the BCA) to prevent or stop the commission of the offence.

Evidence Relied Upon by the Parties

An understanding of the evidence relied upon by both parties, especially the prosecutor, is fundamental to the resolution of the question of whether the defendants are guilty of the offences charged.

The Secretary’s Evidence

The defendants submitted that much of the evidence adduced by the Secretary, which was the subject of extensive objection, was irrelevant and, as a consequence, should be afforded no weight. This was because the evidence was that of alleged unlawful clearing by the defendants outside the charge areas, outside the charge periods, or by means that were not particularised in the summonses, namely, mechanical.
The evidence complained of was admitted on the basis that it was relevant context which supplemented the evidence led by the Secretary that was specifically directed to the offences charged.
The evidence was not relied upon by the prosecutor as tendency evidence or as evidence of the elements of the offences. Rather, it was relied upon as evidence capable of providing context to the prosecutor’s allegations. It was generally admissible as relevant under Pt 3.1 of the Evidence Act 1995. The evidence was background evidence that placed the defendants’ alleged activities within a series of connected events or course of conduct; it was admissible as circumstantial evidence to prove that the defendants committed the offences as charged. The question was ultimately one of weight.
Despite it being admitted over objection, as is explained later in this judgment, some of this evidence assisted the defendants. This is because the contextual evidence included the following undisputed facts:

(a) the owners of the properties (Keelendi Farms and Edenmore Farms) had, for many years prior to the charge periods, undertaken cropping and grazing over areas extending over thousands of hectares on the properties; (b) the annual cropping cycle involved ploughing, weed suppression by herbicide spraying, planting, and harvesting; (c) broadacre mechanical clearing, not charged against the defendants in these proceedings, had been carried out on the properties between 2017 and 2019. This clearing was being investigated by Mark Smith, a Senior Investigator with the Department of Planning and Environment (“the Department”) and Christine Shrimpton, a Compliance and Regulation Officer in the Biodiversity and Conservation Division of the Department, when they visited the properties in June 2020; and (d) from 2017 to early 2020, when rain came in February 2020, the properties had experienced a period of severe drought.

Section 60F Certificates

The Secretary tendered the s 60F certificates on 11 November 2021. One related to Edenmore and the other to Keelendi. Section 60F(5) of the LLSA states that:

(5) The Environment Agency Head may, in connection with any legal proceedings against a landholder, issue a certificate that the land described in the certificate is (for the reasons set out in the certificate) category 1-exempt land or category 2-regulated land (including category 2-vulnerable regulated land). The certificate is, in those legal proceedings, prima facie evidence of the category of the land during the transitional period.

The s 60F certificates had the effect of categorising parcels of land within the properties as category 2-regulated land. Each certificate identified the lot and DP numbers to which the certificate applied. Importantly, every parcel of land the subject of the charges was covered by the s 60F certificates.
I accept the submission of the Secretary that, in the absence of any cogent evidence giving rise to reasonable doubt about the categorisation of relevant parcels of land, it may be accepted to the requisite standard that the land the subject of the charges was category 2-regulated land and that this element of the clearing offences has been established.

Lay Evidence

In addition to the s 60F certificates, and various other tendered documents, the Secretary relied upon the evidence of four lay witnesses.

Mark Smith

According to the affidavit of Smith sworn on 15 June 2022, Smith made several site inspections to the properties. On 23 and 24 June 2020, in relation to mechanical clearing the subject of an earlier unrelated investigation, he visited the properties and observed areas across the properties with tree canopies that “appeared brown and dead”.
Smith returned to the properties in August 2020 and again saw trees that “appeared dead with brown coloured leaves”. This was in contrast to other areas where the vegetation was green, and according to him, appeared healthy, including the trees.
Smith took a number of photographs of his observations during his site visits, which were in evidence before the Court.
Smith identified, by the use of waypoints, the areas on the properties where his observations were made. This included in land parcels the subject of the charges.
During his site inspections in August 2020 (between 11 and 14 August 2020) he also took soil and vegetation samples (leaves and branchlets) at several waypoints on the properties, that were sent for analysis. None of the soil samples tested positive for herbicides.
The analysis of the vegetation confirmed the presence of the following herbicides in two samples:

(a) glyphosate (more commonly known as “Roundup”); (b) aminomethlyphosphonic acid (or AMPA); (c) 2,4-dichlorophenoxyacetic acid (“2,4-D”); (d) triclopyr; and (e) 2-methyl-4-chlorophenoxyacetic acid (or MCPA) (in one of the two positive samples only).

The purpose of the sampling was to test for the presence of herbicides. The samples were obtained by Smith in the field during his site visits approximately five to six months after the alleged clearing events.
Smith was cross-examined extensively on his observation that trees appeared to be dead or browned-off. He was taken to numerous photographs. He noted the difference between photographs taken to the east, which contained browned-off trees, compared to photographs taken to the west, where he could see “the trees on the other side of Epping Road, they’re nice, and green, and healthy, and there’s - they haven’t got that - that quite obvious brown tinge” (T38:09-11). He was shocked at the contrast between the vegetation on one side of the road compared to the vegetation on the other (T46:26-30).
Smith was asked whether he had encountered apparent poisoning while investigating the alleged clearing. He responded that (T46:26-30):

when I’ve arrived there, I have seen the eucalypt trees and vegetation in the state what it was and I’ve gone, ‘Oh my god’, for want of a better word. I’ve gone, ‘This is not right.’ And you look across one side of the road and it’s all nice and green and lush, and you look over this side of the road, it looks - like, everything is dead.

When asked about his understanding of what constituted clearing, Smith said, “the evidence I saw suggested to me and the - the widespread nature of the evidence and the consistency of it suggested to me that something had occurred that had impacted upon these trees.” He went on to say, “there’s a lot more information in the first investigation that provides clear evidence, in my view, that there was a systemic process in relation to the clearing of the property. And the use of herbicides was one part of that” (T61:22-44).
Smith stated that he had a broad knowledge of native vegetation and knew that eucalyptus trees “stay green” (T41:32). He had obtained an environmental science degree from Charles Sturt University which “included a lot of biology and ecological stuff and one of those subjects was plant identification, which included the keying out of various plants and identification of plants, including native species” (T65:04-07).
Smith also gave evidence of what he believed to be burn piles containing charcoal, clay nodules and remnants of burnt timber on both Keelendi and Edenmore. It was his belief that these burn piles were what remained of the vegetation that had been cleared for the purposes of cultivation.
There were several difficulties with Smith’s evidence. First, even assuming for present purposes that the burn piles constituted evidence consistent with clearing, the circumstances of this clearing are not known. It could, for example, have been the result of mechanical clearing that occurred at an earlier point in time and was not the subject of the charges.
Second, Smith’s evidence was given in the context of an ongoing investigation into unlawful clearing by mechanical means of native vegetation at the properties between 2017 and 2019. Smith’s visits to the properties were part of this investigation. It was as a consequence of his observations during these site visits that he formed the view that herbicide had been used to unlawfully clear native vegetation. However, it was not until the site inspections carried out with Dr Kathryn Hammill and Associate Professor Richard Thackway on 11, 12, 13 and 14 August 2020 that he took soil and vegetation samples, only two of which came back positive for herbicides.
Third, observations made by him in his written and oral affidavit to the effect that the vast majority of the leaves of tree canopies were browned-off and appeared dead (see at T39:25-40:08) were not uniformly consistent with the photographs taken by him and tendered in evidence, some of which showed trees that appeared healthy, or at least not dead.
Fourth, and finally, some of the photographs taken by Smith were from locations outside the charge areas (T372:43-45). According to the Secretary, they were relied upon as part of the context within which the alleged clearing occurred.

Christine Shrimpton

Shrimpton’s duties included investigating alleged breaches of, among other legislation, the LLSA and the BCA. She affirmed an affidavit on 4 February 2025.
Shrimpton attended the properties between 22 and 24 June 2020. She made observations about the vegetation at numerous locations, which included the browning of trees, which she contrasted with green shoots of crops and healthy trees elsewhere on the properties.
On 22 June 2020 she observed that vegetation showed signs of stress and decline on Keelendi. On 23 June 2020 Shrimpton noted cultivated areas, along with other vegetation, within Keelendi that was browning. Vegetation adjacent to Keelendi was, however, green. She also observed felled trees and evidence of burn piles, including charred logs. The same day, she saw further evidence of burnt vegetation at various locations within cultivated areas on Keelendi.
Shrimpton conducted another site inspection at Keelendi on 24 June 2020. She again observed browned-off foliage, felled vegetation and burnt vegetation.
Shrimpton was repeatedly cross-examined to the effect that she was on the properties to investigate clearing by mechanical means. She refuted this saying, “at no time did I make a conclusion, prior to the inspection, that the clearing was undertaken wholly or partly mechanically. The purpose of the inspection was to try and determine what had occurred on site and potentially how.” (T79:42-44).
Nevertheless, she noted that “it’s very hard, without the right maps and the right knowledge, to determine from aerial imagery whether pesticides have impacted any vegetation.” (T79:50-T80:02).
The Court repeats its conclusion that only minimal weight may be afforded to her evidence of burn piles on the properties.
Furthermore, similar to Smith, some of the photographs taken by her were of locations outside the charge areas (T372:42-45). Again, they were relied upon by the prosecutor as part of the factual context surrounding the charge areas and charge periods.

Scott Beaumont

Scott Beaumont affirmed two affidavits, one on 15 June 2022 (“the first Beaumont affidavit”) and one on 8 March 2023 (“the second Beaumont affidavit”). Beaumont is a Senior Investigator with the Department. His duties also included investigating alleged breaches of, among other legislation, the LLSA and the BCA. He attended the properties to conduct site inspections. In addition, he served statutory notices on the defendants. He also served a statutory notice on Spray Tech Aviation Pty Limited (“Spray Tech”), the third party who applied herbicide to the properties. He also undertook a directed interview with O’Brien, as the corporate representative of the three corporate defendants.
Beaumont provided evidence of the lots and DPs of the properties that were investigated giving rise to the charges.
Beaumont attended site inspections of the properties between 22 and 26 June 2020 in the company of expert witnesses. He made several observations of vegetation that appeared brown in colour.

Spray Tech Documents

Beaumont issued a statutory notice to Spray Tech on 5 January 2022. The notice asked for, among other things, the instructions that were given to those undertaking the application of poisons or herbicides, together with all invoices and receipts issued in respect of the aerial spraying of the properties, from 1 January 2020 to 21 June 2021.
On 9 September 2021 a similar request for records had been made to each of the corporate defendants in respect of the same time period.
The records provided in response to the notice indicated that Spray Tech was engaged by T J O’Brien to conduct the aerial application of herbicide and that O’Brien was the person who had authorised the spraying of vegetation on the properties with herbicide.
In response to the notice, Spray Tech answered that the instructions given were “do not spray timbered areas. Areas of heavy timber were left to spray by ground”. Further, it stated that “the spraying undertaken was on fallow country for wheat production. Instructions were given not to spray directly over or near trees. Some timbered areas were left to be sprayed by ground. Other spraying operations took place during this period on cotton.” In addition, Spray Tech commented that “minimal impact was observed on native vegetation”. In terms of the steps to be undertaken by Spray Tech to identify native vegetation and to avoid any impact on it, Spray Tech responded that it “uses the map supplied by the Agronomist who recommends the application. We then transfer this to Google Earth to identify hazards such as powerlines and trees and neighbours.” Moreover, prior to an application, an aerial inspection was made by the pilot in command.
In relation to Edenmore, invoices were produced in respect of herbicide applied on 13 and 14 March 2020, in polygons 20, 21, 22 and 24 the subject of the charge areas. The pilot work order contained a written notation that stated “as per Tim’s request”. It may be readily inferred that this was a reference to O’Brien.
In relation to Keelendi, records showed that herbicide was applied to polygons 1, 13, 14, 15 and 16 on 10 and 13 March 2020. All polygons were the subject of charge areas in respect of that property.
A crude marked up aerial map with some cadastral boundaries for variously named paddocks was produced by Spray Tech (among other records) purporting to show the areas that were sprayed on the properties:
Despite submissions by the Secretary that these areas aligned with at least some of the charge areas for the two properties, this was impossible to tell from the notations on the map and I make no such finding in this regard based on this evidence. Reinforcing this view is the fact that neither the provenance nor the date of the map’s creation was known.
The records produced included a recommendation from Greg Rummery Consulting Pty Ltd, a “Hands On Agronomist”, which was for “Broadacre fallow weed control using an aircraft” on Edenmore. The “problem” species being targeted did not include any native vegetation. The chemicals included, Roundup, 2,4-D (Ester 680), triclopyr (Garlon 600), a “hasten spray adjuvant” and ammonium sulphate. The comments at the bottom of the document were addressed to “Tim” – who I again infer is O’Brien – and included the statement “ . Note that if spray conditions are not favourable then the application should halt until such conditions return” (emphasis added).

O’Brien Interview

Beaumont required the corporate defendants to nominate a representative to answer questions. O’Brien was the nominated officer and was required for a compulsory interview, which took place on 3 June 2022. During the interview, O’Brien made the following admissions against the corporate defendants (not against himself):

(a) T J O’Brien farms the properties by planting and harvesting crops. It also runs sheep and cattle on the properties;

(b) T J O’Brien is responsible for the day-to-day farming activities across both properties;

(c) activities across the properties that have resulted in the removal of native vegetation include clearing fence lines and constructing roads;

(d) on a parcel of land where clearing took place, there was: “routine agricultural stuff” and some regrowth that was “tidied up”; the mechanical removal of vegetation using a bulldozer; and excavators to “pluck out” the regrowth. O’Brien was the person making decisions regarding the works that were carried out;

(e) chemicals were used for weed control. The use of boom sprays were directed by O’Brien. It was the agronomist who made recommendations regarding spraying; Spray Tech was used when it was an emergency or when it was too wet to spray;

(f) T J O’Brien paid Spray Tech for the work undertaken by it;

(g) no contractors or third persons carried out works on the properties without permission from O’Brien;

(h) no person except O’Brien authorised works on the properties; and

(i) at all relevant times, no approvals or consents had been obtained for the clearing of native vegetation.

O’Brien was adamant, however, that only weeds and regrowth were sprayed as part of routine agricultural management activities and that the “big trees” were not to be touched:

| Q | 132. | Yeah. And what do you mean by, “re-growth?” | | A | | (No verbal reply) | | Q | 133. | And as I said I’m not a farmer I’m trying to understand what you mean by regrowth. What do you mean by that? What – explain to me what re-growth is? | | A | | Re-growth is basically just small wispy re-growth that’s come back because just through management practices it hasn’t been, you know, looked after just – it’s just, it’s just re-grown. And as Scott can see like, it gets wet, wet, you know, and we get these dry periods and then these wet periods and then yeah, and it just hasn’t been, it just hasn’t been controlled. | | Q | 134. | Yeah, so that’s the reasoning of the re-growth. What is re-growth? | | A | | (No verbal reply) | | Q | 135. | When I think of re-growth I’m thinking of (indistinct) or a shadow but obviously we’re not talking about beards. | | A | | Yes, yeah, yeah, yeah. | | Q | 136. | Can you explain what you mean by re-growth? | | A | | Re-growth’s just – it’s, it’s – they’re just re-growth trees, you know, just trees that have just – they’re not, they’re not big trees they’re just, they’re just small re-growth in trees yeah. | | … | | | | MR MCGIRR: | Is it weeds? Is it Box-thorns? Is it - - - | | MR O’BRIEN: | It’s, it’s yeah, look, it’s, it’s all those things, you know, it’s yeah, that’s, that’s exactly right. And a lot of the stuff is yeah, just noxious weeds that we tidy up as well like thorns and- - - | | A | | Yeah. | | … | | | | Q | 174. | What can you tell me about the vegetation and clearing works that occurred in that area? | | A | | That’s the same Scott it’s, it’s just re-growth that we, we tidied up there and we’ve just put in some new roads to run through there. | | MR MCGIRR: | So, this is all routine agricultural management? | | MR O’BRIEN: | Yeah. | | … | | | | Q | 179. | What agricultural activities have occurred in that part of the property over the last 5 years for example? | | A | | We’ve run that again, it’s a mixed farm so we’ve run stock there, we’ve run – and we done some, some rotation of the farming there it’s just – yeah, it’s just been run as a mixed, a mixed operation there. | | Q | 180. | Okay. Have there been trees other than re-growth removed from that area that I’ve marked number 2? | | A | | (No verbal reply) | | Q | 181. | Just for the purpose of the recording - - - | | A | | Sorry, no, no there’s – there was only the re-growth trees (indistinct) there Scott. | | … | | | | Q | 185. | And number 6 being within these boundaries here that I’ve put the texta around and number 7 here within that boundary. Across those properties 3, 4, 5, 6 and 7 have any large species of native vegetation been removed for the purpose of your farming activities? | | A | | We only, we only took the re-growth Scott. | | Q | 186. | So my information is that these areas contained large trees that weren’t re-growth in everyone of those areas and including 1 and 2 and 3 that are now in poor health or have been removed. What can you tell me about that? | | A | | We, we didn’t remove the big trees and I – yeah, I, I don’t know with, with the poor health like we haven’t, you know, and I’m not sure what you mean by poor health but yeah. Is that, is that, is that – yeah, we, we - - - | | … | | | | Q | 187. | Do you use poison on the property? | | A | | No. | | Q | 188. | Do you use poison for weed control? | | A | | We use chemicals for weed control yeah - - - | | Q | 189. | Okay. Sorry. | | A | | - - - I don’t call it poison yeah. | | Q | 190. | Okay. Chemicals. What chemicals do you use for weed control? | | A | | We just use pretty general chemicals for weed control just for – just to control the, you know, the weeds for the cropping and just to maintain, to maintain the country, so Roundup and Roundup is really one of the main ones. | | MR MCGIRR: | Is that what you use or do you get the contractors? | | Mr O’BRIEN: | We – yeah, we, we do, we do have our own applicators and we, and we do use contractors as well. | | … | | | | Q | 219. | Okay. In relation to Kealandi [sic] and Edenmore the information that I have Tim is that there were large trees right across both properties and, and in particular in these boxes that I’ve marked on this map in front of you and those trees are no longer there or they’re in poor health. Can you explain that to me? | | A | | I – Scott we, we – as I said we only touch, we only remove the re-growth like I, I didn’t want, you know, look I love, you know, the big trees yeah, but they weren’t, they weren’t to be touched. | | Q | 220. | They weren’t to be touched? | | A | | (No verbal reply) | | Q | 221. | Is that instruction you give to your contractors working on your properties? | | A | | I just give the instructions to only take re-growth. |

The only time that trees were removed was for routine agricultural management, where the trees were stacked and burnt:

| Q | 231. | Has there been any instances where you’ve instructed the removal of larger trees on the property native vegetation - - - | | A | | Ah. | | Q | 232. | - - - that aren’t re-growth? | | A | | Look, the only times Scott if it was, it was in - and you know if we were putting in a, a new road, or a fence line, or something like that that’s the only time yeah, that’s the only time that, that the large trees would, would have been as I understand it routine agricultural management so that’s the only time with the larger trees we’d do it. | | Q | 233. | And how often would that take part? | | A | | It’s not a, not a scheduled thing Scott it’s just as we need to do it type of thing as, you know, as we get time to do it, you know, really it’s not a scheduled thing it’s just, it’s just – look, during the, during the drought we just, you know, we’ll – that’s, that’s when we did a lot of the (indistinct) because there was – it was, it was drought time and we – you know, we, we couldn’t grow crops and we couldn’t run stock so, you know, we just – that was the time that we do fencing the roads and fix things up. | | Q | 234. | So the large trees to remove for routine agricultural management activities. How do you remove a large tree? | | A | | Generally we pluck it out with an excavator, or, or, or, or a bulldozer. | | Q | 235. | So it can be pushed over with a dozer? | | A | | It can be yes, yeah, yeah. | | Q | 236. | Do you use the chain method where you - - - | | A | | No, no. | | Q | 237. | Yeah. No? | | A | | (No verbal reply) | | Q | 238. | Do you – when you say, “an excavator” is it something with a grader on it, or something that you can just knock it over, or? | | A | | It just plucks, plucks it out Scott so it’s just very, you know, it’s just very, accurate I suppose just – it’s just not broad scale it’s just plucking out. | | Q | 239. | So both of those methods would remove the root-ball(?) Is that correct? | | A | | Yes, I suppose they’d dig it out of the ground yes, yeah. | | Q | 240. | And once that tree is out of the ground and moved so you can go about your business. What is done with that tree? | | A | | Well, if it’s – it’ll be racked up and racked up into a pile and then we, and then we – and then just to tidy up we, we normally burn it. |

O’Brien indicated that the drought had affected a number of the trees on the property:

| Q | 244. | - - - causing problems with the health? | | A | | And, and Scott, we had, we had this massive drought too – not droughts – but that – if that’s caused, you know, what you guys are alleging against us a lot of that – we saw a lot of trees die in the drought too and you know, you know, it worries me what you guys are alleging against us and I think, you know, the, the drought, the drought has done a lot of this to the trees you know. It’s mainly – the big trees they need a lot of water and, and during a – you know a massive of the worst drought in history like the, you know, they’re used to getting rain every year so yeah. So, I think yeah, I – look, I’m interested to see what yeah, your experts say but I think - - - | | Q | 245. | Mm. | | A | | - - - you know and I, I think the drought was a big part of it as well you know so - and we’ve seen areas that we haven’t sprayed and – around, around the district that have died because of, because of the drought. | | Q | 246. | Have you had areas on your property where large trees have been affected by the drought that are more susceptible than other areas? | | A | | Yes, yeah. | | Q | 247. | Any why is that? | | A | | I think because of – I think this, this country here at, at Pilliga is, is, you know, is pretty – has a pretty reliable rainfall so. I think there’s trees all the way through the properties Scott that have been affected by the drought like all the way through. Like, and you talk about all your, you know, you think, or, I can, I can say the same thing that the droughts affected all the trees through that property. | | Q | 248. | If trees are affected by the drought what affects does it have on them? | | A | | They, they just die. | | Q | 249. | And when they die what do you do with them? | | A | | I just leave them there and looks it’s – I, I don’t (indistinct) if they’re still there - - - | | Q | 250. | Okay. | | A | | - - - if – I think they’re the trees you’re referring to. | | Q | 251. | So they’re the ones potentially that I am alleging are in poor health and they maybe drought - - - | | A | | But you’re - - - | | Q | 252. | - - - affected trees? | | A | | Yeah, I think so. | | Q | 253. | So you don’t remove any of those trees yourself? | | A | | No, we haven’t, we haven’t removed any of those trees. | | Q | 254. | Do any of them come back? | | A | | Yes. |

ASIC search

Beaumont conducted an ASIC search that established that, at all relevant times, O’Brien had been a director of:

(a) Edenmore Farms since 12 December 2021; (b) Keelendi Farms since 20 April 2016; and (c) T J O’Brien since 3 March 2004.

Susan Rea

Finally, Susan Rea, the Team Leader, Spatial Imagery Services for the Department, affirmed an affidavit on 9 March 2023. Her role involved managing aerial and satellite images within the Department.
Rea confirmed that a number of aerial images depicting the properties were made available to Leo Watts, an aerial imagery interpretation expert witness engaged by the Secretary.
In cross-examination she was asked questions about technical aspects of the images curated by her team. In response to a question about the differences between infrared and colour imagery, she said (T106:29-36):

And what the near-infrared allows you to do is to be able to detect - it’s good for detecting vegetation, and the health of vegetation. So when you see - so if you look at an image - satellite imagery, image using the red, green, blue band, what we call three band, you will see the true colour, so vegetation will appear green. With a four band image, the near-infrared, the image displays red - or, the vegetation will display red, and the health of the vegetation can be depicted by the redness in the image. And different vegetation reflects different values of the red.

Rea’s evidence about the value of infrared to measure the health of vegetation is consistent with that of Watts.
In response to a challenge to her expertise she said (T106:43-48):

So just - the training that I’ve had is probably the 30 years experience that I’ve had using remote - remote sensing data throughout the years. I’ve got a Bachelor of Technology, major in GIS, which included remote sensing techniques. And then working with - and then - and just regular workshops that we had within our own department, and our team, in regards to learning remote sensing skills and techniques.

Rea confirmed that Watts had been provided with daily aerial imagery for specific dates and not merely monthly mosaics.

Expert Evidence

Leo Watts

The Secretary relied upon three aerial imaging and analysis reports prepared by Watts. Watts specialises in the field of aerial surveying and aerial imagery analysis. He has over 35 years of experience in this field. The Secretary engaged him to prepare a report on the changes in vegetation on Edenmore and Keelendi.
Watts was initially engaged by the Department to provide a report in 2021 in respect of the alleged unlawful clearing, which he did on 3 March 2021 (“the 2021 Watts report”). The 2021 Watts report was not tendered in the proceedings and was not relied upon by the Secretary. A dispute arose during the hearing, however, as to the extent to which the content of the 2021 Watts report informed his subsequent reports. Objections were made by the defendants to the use of the subsequent Watts reports tendered in the proceedings on this basis. But a proper analysis of the Watts reports relied upon by the Secretary reveals that his conclusions were not based upon his earlier findings and opinions contained in the 2021 Watts report. Accordingly, they were admitted into evidence.
The first Watts report relied upon by the Secretary in these proceedings was prepared by him on 14 June 2022 (“the first Watts report”). Watts’s instructions were in these terms:

Based on an examination of aerial photographs and satellite images over the Properties, provide an expert opinion report on whether there has been clearing of native vegetation on Category 2 regulated land and more specifically on the following:

(a) Any changes in the vegetation present on the Properties which occurred after 1 January 2020.

(b) Any changes you are able to identify, including the location of such change and the dates between which such change took place.

(c) If you have identified any vegetation as having been cleared, was that vegetation present on the Properties prior to 1 January 1990.

(d) If you have identified any vegetation as having been cleared, the extent of such clearing.

(e) Include within the report a sequence of images (aerial photographs and/or satellite images) which show any vegetation change over time.

In his first report Watts analysed a total of 17 aerial images taken between 23 June 1986 to 8 January 2021, which comprised both colour infrared and natural colour images at differing resolutions. His analysis was confined only to land defined as category 2-regulated land pursuant to the LLSA.
Watts prepared maps and tables assigning polygon numbers to areas that, in his opinion, had been affected by the clearing of ground cover vegetation and trees. Because these properties had been investigated previously for unlawful clearing, he used only “newly identified areas” alleged to have been cleared during the charge periods (as is explained below, these were modified in his second report). Watts’s iterative methodology was set out in the first Watts report, and elaborated upon in his subsequent reports.
The imagery that Watts had regard to in applying this methodology was listed in his first report. Data was requested by him and once received, imported into a geospatial dataset in his global mapper software, whereupon he adopted the coordinate system of the Australian Geospatial System standard. He then checked and confirmed the extent and contents of the cadastral boundary shapefiles to confirm the geographic boundaries of the properties. He visually checked and confirmed the contents of the s 60F certificate category boundary shapefiles to satisfy himself of the lot and DP numbers of the plots within each property.
Watts proceeded to identify the dates of the aerial photography and satellite imagery that he had received. He undertook a process of image rectification (a process where imagery taken from multiple perspectives is merged) and checked the 17 dates of imagery for their spatial, spectral and radiometric suitability. Where appropriate, adjustments were made to the imagery as required.
Using the global mapper software, he extracted the category 2-regulated land boundaries from the s 60F certificate category shapefiles to create a geospatial boundary layer of the areas of interest within the global mapper for later use during the imagery analysis. He checked the spatial accuracy of the datasets to confirm their appropriateness and suitability for use, including the Sentinel-2 imagery.
With respect to his use of infrared imagery (criticised by the defendants), he stated that leaves of healthy, growing vegetation appear red, shades of red, or pink in this type of colour image. Deep red indicates dense, vigorous growth. Thus, as vegetation and tree populations decrease or become sparser, the vegetation appears as lighter shades of red or pink. When the vegetation becomes unhealthy or dies, it appears as various shades of green or tan.
Watts stated that he was able to identify land cover types through the analysis of high resolution image datasets:

Trees were identified by me through a combination of their irregular shaped canopies, canopy tone and green colour, coarse canopy texture and their vertical height profile above ground indicated by the shadows they cast.

I identified areas of ground cover/grass cover through a combination of their green/grey colour, mottled but smoother texture, irregular shaped patterns which were visible in most areas between trees and in open spaces, and their ground profile due to no shadow being cast between the trees. In addition, nearby ground areas I identified as cultivated or cleared paddocks had a completely different appearance.

Areas of cultivated paddocks or cleared areas were identified by me through a combination of their green/grey colour, much finer texture, regular and evenly coloured straight line patterns indicative of a manmade activity and the large contiguous shaped areas they covered.

Areas of crops under cultivation were identified by me through a combination of their deep dark green colour, much finer and more even texture than natural ground cover, regular and fine straight line patterns indicative of a manmade activity and the large contiguous shaped areas they covered.

Areas of ploughed paddocks were identified by me through a combination of their light grey colour, fine texture, repeated and fine straight line patterns indicative of a manmade activity and the large contiguous shaped areas they covered.

According to Watts’s analysis, as at 1 January 2020 there were both significant portions of the properties where vegetation had been removed for possible future cultivation purposes and there were significant undisturbed areas covered with light to medium scattered trees and ground cover:

Areas identified as having undergone mechanical clearing between 31 August 2017 and 1 January 2020.

a) These areas covered the majority of the post 1 January 2020 areas where I observed the suppression or treatment of the vegetation to have occurred.

b) In these areas most to all of the ground cover had previously been cleared prior to 1 January 2020. As such I concluded that any ground cover in these areas was unlikely to have been present prior to 1 January 1990.

c) While significant numbers of trees had previously been removed or thinned from these areas, there were still many remaining trees at varying densities as at 1 January 2020. My comparison with the 1990 imagery concluded that the far majority of these remaining trees were present prior to 1 January 1990. As detailed in the FINDINGS section of this report however, the imagery available for this analysis did not enable me to comment on the relative health of these trees resulting from the observed vegetation suppression within these areas.

Areas where clearing had not previously been identified prior to 1 January 2020.

a) In these areas where the vegetation had previously remained undisturbed since 1990, my comparison with the 1990 imagery concluded that the far majority of the trees in these areas were present prior to 1 January 1990. Again as detailed in the FINDINGS section of this report however, the imagery available for this analysis did not enable me to comment on the relative health of these trees resulting from the observed vegetation suppression within these areas.

b) In regards to the ground cover in these areas, based upon my analysis of the 1986, 1990 and 1 January 2020 images, I observed that these areas had a history of containing ground cover prior to the observed vegetation suppression. While I am of the opinion that in the far majority of cases, these areas contained some form of ground cover in the same areas in 1986, 1990 and 1 January 2020, I cannot comment on whether these areas consisted of the same types and species of ground cover in all dates.

Watts had commenced his analysis by looking at pre-2020 high resolution aerial images to create a baseline. He gave examples of colour images of land types of vegetation that he had observed in the aerial images taken on 24 July 2017, 3 July 2019 and 1 January 2020. Watts then examined what had occurred on the properties during 2020 by reference to several image dates, in particular, images taken between 13 February and 13 March 2020.
Watts made several observations about changes in the vegetation between 13 February 2020 and 13 March 2020 (that is, in one month), which in his view were widespread and noticeable. In relation to Keelendi he noted that:

106 …

b. In regards to Keelendi it appeared that widespread vegetation regrowth that had the appearance of natural ground cover regrowth as opposed to a cultivated crop had occurred over areas both inside the property boundary, and outside and adjacent to it. In addition however, there also appeared to have been some attempt to suppress or remove the vegetation growth over large areas within the Keelendi property. Five polygon areas numbered 1, 2, 14, 15 and 16 are noted in the following image.

And in relation to Edenmore he opined that:

106 …

c. In regards to Edenmore it also appeared that widespread vegetation growth had occurred both inside the property boundary, and outside and adjacent to it. As at this date however there appeared to only be one polygon area numbered 19, highlighted in Leo Watts the following image, where some attempt to suppress or remove the vegetation growth had occurred.

Watts then provided a more detailed analysis as to what he believed had occurred on both farms in the period between 13 February and 13 March 2020. He used reference points identifiable in the imagery relied on to assess ground cover growth. Watts stated that he was able to discern between mechanical and non-mechanical growth restraint by examining the texture, tone and ground patterns in the polygons. He could delineate between the polygons with and without areas of healthy vegetation coverage. The same exercise was repeated for the next alleged clearing date, that is, between 13 March and 23 March 2020, and for each image date until the end of 2020. Thus, for example, the analysis based on the image taken on 5 July 2020, was that areas that had previously been the subject of clearing had not experienced natural ground cover regrowth, but had possibly experienced the growth of other planted vegetation such as a crop.
According to Watts, he was also able to identify vegetation that was ground cover as distinct from cultivated paddocks in the polygons within the charge areas. He was able to identify areas of limited or unhealthy ground cover by the aqua green colour in the infrared images that he had analysed.
Watts’s findings in his first report were summarised by him as follows:

Seventeen epochs of aerial imagery dated between 23 June 1986 and 8 January 2021 were analysed for the purpose of determining changes in vegetation over the Properties.

Land analysed for vegetation change cover only those parts of the Properties designated as Category 2 regulated land as defined by the yellow shading on the section 60F certificates contained within Appendices 36 and 37 (Area of Interest). As per these maps, the areas analysed for vegetation change cover:

• Keelendi Farms 5,128ha

• Edenmore Farms 6,460ha

The image below highlights the location of the Area of Interest on both the Keelendi and Edenmore Properties.

In general, as at 1 January 2020, the lands on both Properties within the Area of Interest have similar appearances. Significant areas had previously had vegetation cleared or thinned. There were also significant areas that appeared undisturbed which can be described as largely open areas covered with light to medium scattered trees in parts and with some form of ground cover. Several watercourses were observed to meander across both Properties and in these areas the tree cover was denser than the general trend of tree cover over other parts.

I observed very little infrastructure on either Property within the Area of Interest apart from some dams, tracks and drainage channels. The most significant piece of infrastructure appeared to be an airstrip on Keelendi Farms.

Within the Area of Interest on both Properties, I observed no obvious cultivation activities being undertaken as at 1 January 2020.

In regards to my vegetation change analysis undertaken within the FINDINGS, I made the following observations:

a) Between February and April 2020 there appeared to have been some attempt to suppress, reduce the health of or remove ground cover vegetation growth over large areas within both Properties. Based upon my observations, the vegetation within these areas had the appearance of either being unhealthy or possibly even dying or dead.

b) While I could not determine the method used to undertake this through the analysis of the imagery it did not have the appearance of having been undertaken by mechanical clearing means, as I believed had been used for the observed vegetation clearing prior to 2020 and as documented in my aerial imagery interpretation report dated 3 March 2021.

c) I observed this suppression or removal of vegetation to have occurred within the following periods:

● Keelendi Farms – between 13 February 2020 and 23 March 2020 over approximately 4,182ha being 80% of the Category 2 land.

● Edenmore Farms – from 13 February 2020 with its full effects observed by 12 April 2020 over approximately 3,493ha being 54% of the Category 2 land.

d) Post 1 January 2020, I observed that parts of the Category 2 land on both Properties had for the first time been worked for the preparation of crops, followed by the planting of crops and then by the start of 2021 their removal and the return of the paddocks to ploughed fields.

e) During the first few months of 2020 I noted that areas within Category 2 land on both Properties, which had and had not previously been cleared of trees and ground cover vegetation, appeared to have experienced widespread natural ground cover regrowth. Similar vegetation growth was also observed over wide areas outside and adjacent to the Properties during this time.

My reasonings and methodology for these findings are detailed within the body of the report.

A diagram showing all areas where Category 2 land within both Properties appeared to have been affected by the removal or treatment of ground cover vegetation is highlighted in green on the image below.

Maps which assign unique polygon numbers to all areas which appeared to have been affected by the removal or treatment of ground cover vegetation can be found in Appendix 39. A table which lists the polygon numbers, their total area in hectares and the Lot and Deposited Plan numbers which fall within each polygon can be found in Appendix 40.

I have previously an aerial imagery interpretation report dated 3 March 2021 which investigated alleged vegetation change between 31 August 2017 and 19 March 2019 over both Properties.

I observed that the areas of modified vegetation within this current report included the far majority of the cleared areas observed within the 3 March 2021 report.

I refer to the areas of modified vegetation identified in this report that were additional to the areas identified in the 3 March 2021 report as the ‘newly identified areas’. I determined the area of the ‘newly identified areas’ to be:

● Keelendi Farms 2,288ha

● Edenmore Farms 1,565ha

TOTAL 3,853ha

Maps of the ‘newly identified areas’ with the same polygon numbers can be found in Appendix 42. A table which lists the polygon numbers, the total ‘newly identified area’ within each polygon in hectares and the Lot and Deposited Plan numbers which fall within each ‘newly identified area’ can be found in Appendix 43.

A diagram showing all ‘newly defined areas’ where Category 2 land within both Properties appeared to have been affected by the removal or treatment of ground cover vegetation is highlighted in green on the following image.

In short, it was his opinion that in the period between the aerial images dated 13 February and 13 March 2020, there were noticeable and widespread changes to the vegetation on both properties. In particular, five polygons on Keelendi (1, 2, 14, 15 and 16) and one polygon on Edenmore (19) had experienced some form of what he believed to be non-mechanical clearing.
Between 13 and 23 March 2020, there was, according to Watts, further widespread vegetation changes over both properties. On Keelendi, polygons 2 to 13 were cleared of most of the vegetation that had been visible 10 days earlier. Similar observations were made about Edenmore in respect of polygons 17, 18 and 20 to 28.
Although outside the charge periods, by 12 April 2020 the extent of the apparent clearing was evident. Further aerial analysis after 12 April 2020, and up to and including 7 September 2020, indicated healthy vegetation growth at the boundaries of the properties and vegetation regrowth within the boundaries of the properties that was consistent with crop cultivation. Areas of crop cultivation appeared to cover most, but not all, of the same areas where suppression or removal of vegetation was observed to have occurred between February and March 2020.
Watts prepared a supplementary report dated 6 February 2025 (“the second Watts report”). The second Watts report was responsive to a report prepared by Dr David Robertson, an ecologist engaged by the defendants, dated 24 January 2025 (“Dr Robertson’s first report”).
As a consequence of Dr Robertson’s first report, Watts made modifications to areas in the first Watts report which he had identified as experiencing supressed vegetation and recategorised them as bare earth, that is, his “Modified Newly Identified Areas”. This had the effect of reducing the total area of land cleared in polygons 1, 11, 17, 23 and 27.
The Executive Summary to the second Watts report stated:

Through my analysis of the high resolution images dated 24 February 2017, 3 July 2019 and 1 January 2020, I have re-assessed the extents of the ‘Newly Identified Areas’ from the Primary Report. I have determined that the areas of what I now refer to as the ‘Modified Newly Identified Areas’ to be:

● Keelendi Farms 2,062 ha

● Edenmore Farms 1,374 ha

TOTAL 3,436 ha

Maps of the ‘Modified Newly Identified Areas’ with the clearing polygon numbers used in the Primary Report can be found in Appendix 1 of this report. This appendix replaces Appendix 42 in the Primary Report. A table which lists the polygon numbers, the total ‘Modified Newly Identified Area’ within each polygon in hectares and the Lot and Deposited Plan numbers which fall within each ‘Newly Identified Area’ can be found in Appendix 2. This appendix replaces Appendix 43 in the Primary Report. Maps which highlight the comparison between the ‘Modified Newly Identified Areas’ and areas removed from the maps in Appendix 42 of the Primary Report can be found in Appendix 3.

A third Watts report was prepared on 17 February 2025 (“the third Watts report”). This report was also prepared in response to Dr Robertson’s first report and the addendum report by Dr Robertson dated 7 February 2025 (“Dr Robertson’s second report”).
Dr Robertson’s first report was highly critical of Watts’s methodology and evidence, in respect of which Watts responded as follows:

(a) first, Dr Robertson criticised Watts’s use of the Sentinel-2 satellite imagery as being inappropriate for identifying vegetation changes. Watts, a very experienced analyst of aerial imagery in this context, maintained its suitability for identifying such changes, including those occurring within short time periods; (b) second, Dr Robertson stated that there was no ground data relied upon by Watts to obtain a baseline measure of the condition of the vegetation on the properties prior to any clearing. According to Watts, however, an analysis of high resolution aerial imagery of the properties between 1996 and 2020 provided a satisfactory method of establishing a baseline of the vegetation that existed, and its location, as at 2020 in the absence of ground data; (c) third, in response to Dr Robertson’s complaint that external reference sites were not used by Watts as a comparison to the internal sites that he analysed on the properties, Watts noted that he had used multiple internal, external and control sites located within the differently dated imagery; (d) fourth, Watts indicated that he saw little utility in “ground truthing” his findings as Dr Robertson suggested, because his observations from all images resulted in “clear conclusions” regarding the state of vegetation in the affected areas. Put simply, it was not necessary to do so; and (e) fifth, Watts observed that if, as Dr Robertson had opined, the causes for the change in vegetation were natural, then his expectation was that the changes would not be confined to fence lines or lot boundaries, that is, along straight lines. Rather, the visual impact would have been more random and would have affected vegetation on immediately adjacent properties.

Watts was cross-examined about the method of clearing. He said (T188:18-24):

There was one very small area in that period between - in - during 2020 where I did observe a very small part of mechanical cleaning and removal, but the far majority was an observe – the change I observed was the change in the appearance and the health of the vegetation. Now, the cause of it, what caused that, I couldn’t determine, except that the impact of what I observed was - was very much confined to within property boundaries or lot boundaries.

Watts was challenged on his use of different resolution imagery and asked whether it impacted his interpretation of the images. Watts’s unequivocal response was (T209:17-37):

…I was able to determine, because of the resolution - the resolution was so - was - was so good - obviously, there are higher - you know, there’s even higher resolution images and satellites of those, but the imagery - the - the resolution of those images was of a resolution where I could identify directly from each individual image whether - what was - what were trees and what were - and what was ground cover. I didn’t - I didn’t need to - you know, I didn’t - I didn’t need to ground truth - it was - it was obvious to me

…

they were good quality, where you could even identify beyond the - you - you could see the tree canopies. You could see the - you could see the shapes of the canopies. You could see the - the different patterns and textures of the - of the tree canopies. You could see the different patterns and textures in the - in the ground. You could see the bare earth that we spoke about before, and I used those same images to refine some of those areas that I’d previously looked at. So I used that as - to establish a baseline…

Watts proceeded to explain how he had used imagery with higher resolution to establish a baseline of vegetation on the properties (T210:05-21):

...once I started my assessment, I was then using other dates of imagery, more of the medium resolution imagery that I refer to from January 2020 onward, and within all of those dates of images I - I could observe locations of vegetation that were external to the site that did not change, trees and grass cover, and these - and I used those as what I call my - my - my control sites or my reference sites.

So it became a process for me to undertake a relative comparison between what had not changed, in my opinion, in places that were mainly external to the properties and areas that had changed internally within the properties, and I used that comparison with both the colour aerial imagery and the - and the colour - the colour infrared imagery. So I did not need to ground truth to establish a calibration point of what was vegetation and what wasn’t vegetation. I could actually do that by - by comparative terms. In addition, the visual changes I observed between the - the changes I observed as - during the days of imagery during 2020 became so - to me, so clear and so obvious. It - but it wasn’t - to me, it wasn’t as if it was a case of marginal differences…

The ground truthing exercise suggested by Dr Robertson was therefore unnecessary according to Watts. He expanded upon this opinion in re-examination, saying (T241:36-242:22):

…So for every date of imagery after my baseline point, 1 January 2020, when I observed - when I analysed the - colour or the - the natural colour imagery, I referred to areas that had changed or hadn’t changed back to those points in regards to their colour, their shapes or - or patterns, tones. Certainly, in the colour imagery - and also, particularly, I used that - I used those reference points in the colour infrared imagery as well because I found that the external calibration sites hadn’t changed at all throughout the whole of 2020. So almost in every instance - well, I think in every instance, when I looked - identified an area that had changed in the natural colour imagery and then I then went to the colour infrared imagery and compared those sites - or the colours within the colour infrared imagery to those test sites, it just confirmed my observations from the natural coloured imagery.

…

Had not changed in their - their appearance of vegetation coverage. So they had been - there had been no - and this is both - both in trees and ground cover, because I wasn’t - I wasn’t seeing changes in health. I wasn’t - I wasn’t seeing changes in land use of those area. There’s be no - there was no clearing of vegetation. There was no clearing of trees. There was no cropping or farming. The only changes that occurred was most likely just the density of vegetation as a course of the seasons. The rain events that occurred in February and March. I could identify that the - the spatial positions of features, vegetation features, particular trees, hadn’t changed. I could identify that the - that the patterns and the texture of ground cover hadn’t changed. That’s - ‘cause how I use it.

…

The lack of change provided - gave me confidence to - to determine that these sites outside the properties had not been affected by any natural cause. Or any man-made cause of clearing. And they provided, for me, excellent comparison points. Where I could compare all these external reference and control sites to areas that I identified as changed vegetation within the properties. And how they changed over time. So I could compare those reference sites between dates of imagery. And then compare the changes of vegetation within a date of imagery with those referenced sites.

Watts maintained that the 10 m resolution satellite imagery that he had used was an appropriate resolution to analyse any changes in vegetation across large areas (T219:29-30). The choice of image resolution depended upon what was being examined. Because he was analysing changes across thousands of hectares, rather than individual shrubs or trees, 10 m resolution imagery was “ideal”. Watts conceded that a 10 m resolution may not be sufficient to discern the health of trees, nevertheless the imagery was, in his view, adequate to assess the health of ground cover.
It was put to Watts that there was a material difference between the vegetation on a roadside reserve and the vegetation on areas that had been annually farmed, and that they could not be meaningfully compared. In response, Watts said that (T227:36-47):

…it’s not just the vegetation of the road reserves that I use as my control and test sites. It’s - there are - there are many other - and I believe the ones - if I’ve got this correct, the road reserves, I think you may be talking to the - to the eastern side of Edenmore. There were some - some areas over there that was undisturbed vegetation. But there were also many other areas surrounding the - the Keelendi and the Edenmore farms where I observed many areas of undisturbed vegetation that weren’t stock roots or - or road reserves that exhibited, at 1 January, identical vegetation, side by side on one side of the fence of Keelendi and on the other side of fence of Keelendi. So as - visually, as - as - as it - as I could determine, the - the appearance of the vegetation, ground cover and trees looked identical.

When asked about his understanding of the difference between vegetation that was suppressed compared to vegetation that was stressed, Watts opined that the difference could be discerned by examining the uniformity of the impact (T233.15-35):

It would depend - to me, it would depend upon the cause of the stress or the suppression...

If it was a - a natural cause of stress or vegetation restraint suppression, I would expect what I see - the impacts of what I see would be more random in nature, as it would be, I would think. For example, if there was - they talk about, I think, insect damage or aphid infestation and those sorts of - those sorts of things. Well, you would see them - I - I would believe - okay, I’m not an expert in those sort of fields.

…

But I would - I would see a natural cause as a more random occurrence within the imagery, as opposed to a - something that was more systematic and deliberate, would be something which I’m seeing in - in this imagery, where the suppression is following paddock boundaries, fence lines and lot boundaries. If it was natural, I would be expecting the - the natural cause to be spilling over into the neighbours’ paddocks, into the - into the road reserves, for example, but that’s not happening.

It is convenient at this juncture to make four observations about Watts’s evidence. First, I do not accept that Watts was able to satisfactorily explain the changes that he had detected in his comparison of the aerial images, especially from 1 January to 23 March 2020. For example, Watts contended that he was able to observe changes in the vegetation across the properties between 13 February and 13 March 2020. However, if, as I accept based on the evidence before me, that the first relevant aerial spraying did not occur until 10 March 2020, this means that the changes that he observed in respect of the properties between 13 February and 13 March 2020 (the first charge period) occurred in just three days. At the landscape scale this seems implausible. If clearing of native vegetation (either ground cover or trees) was occurring during this period, it is therefore likely that it was by means other than by poisoning as alleged.
Second, of concern, and undermining to a significant degree the weight to be attributed to Watts’s ability to distinguish between ground cover and other forms of vegetation such as crops or weeds (which he was adamant he could do), was the area that Watts mistakenly identified as vegetated which, as Dr Robertson demonstrated, was bare earth. This diminishes the credibility of Watts’s methodology. No explanation from Watts was given as to the cause of the error in analysis. While it is commendable that he acknowledged the mistake, that a similar error in his analysis has been made elsewhere cannot be discounted.
Third, there is substance in the defendants’ criticism that Watts placed insufficient emphasis on the fact that the properties were subject to ongoing farming activities, including cultivation and cropping, and clearing prior to the charge periods. Watts was not analysing a landscape that had not previously been modified. This strongly suggests that the areas that Watts concluded had been cleared did not contain native vegetation, but comprised some other form of non-native ground cover, such as weeds, during the charge periods. The complex history of clearing on both properties prior to the charge periods make it impossible to find beyond reasonable doubt that the extent of the changes in native vegetation that Watts observed were due to some form of systematic program of chemical suppression by the defendants during the charge periods.
Fourth, although he remarked that the clearing “did not have the appearance of having been undertaken by mechanical clearing means”, he unequivocally stated that he “could not determine the method used to undertake this [clearing] through the analysis of imagery”.

Prof Richard Thackway

The Secretary relied upon several expert reports of Professor Richard Thackway. The first report was dated 16 March 2023 (“the first Prof Thackway report”). The second report was dated 14 February 2025, and was prepared in response to Dr Robertson’s first and second reports (“the second Prof Thackway report”).
Professor Thackway is an ecologist with particular expertise in land management, vegetation surveying and mapping, including the use and interpretation of satellite imagery.
In his first report, Prof Thackway was asked to opine on the following:

In brief, I was asked by the Department to provide opinion, based on my specialised area of knowledge, addressing the following.

a) Whether there has been any clearing of vegetation by way of poisoning in the Area of Interest, and if so:

i. the extent of such clearing;

ii. whether such clearing has caused or is likely to cause any harm to the environment.

b) If you have identified harm, or likely harm, identify the extent of any such harm including the immediate and ongoing impact on:

vegetation in any identified ecologically endangered communities;

“paddock trees”; and

identified native vegetation.

Provision of a sequence of photographs taken during your site inspection.

c) Provision of a sequence of a table of species identified during your site inspections noting within which polygon each species was identified.

In Prof Thackway’s opinion, herbicide poisoning most likely occurred after mechanical clearing of the properties. It was his view that there were multiple chemical events in the form of repeated applications of herbicide. He eliminated drought, insect damage and groundwater draw-down as causes of his observed changes in vegetation when compared to the roadside reserves.
During a site visit to the properties in June 2021, Prof Thackway tested for the presence of herbicide in samples taken by him. Both biological samples tested positive for herbicide, consistent with the use of herbicide products applied by Spray Tech. Professor Thackway’s sampling results were consistent with those of Smith. That is, the soil samples returned a negative result, but the vegetation samples indicated the presence of Roundup and triclopyr, both of which cause browning and loss of foliage.
The low number of positive results in the soil samples was considered by Prof Thackway to be caused by the delay between spraying and sampling – almost 15 months later – or the chemical leaching down into the soil. This was consistent with Dr Robertson’s evidence that the samples ought to have been collected “soon after the event” (T313:25-26), that is, “days or weeks, or at the most a month or so” (T313:40).
His conclusions were summarised in section 2 of the first Prof Thackway report, including the native vegetation that he had observed on the properties that had been adversely affected by what he concluded to be herbicide poisoning:

The following statements summarise my field assessment of sites:

a) All 77 plots surveyed within Keelendi-Edenmore and the adjacent Roadside reserves sites (24) were surveyed in June 2021. Survey sites were restricted to heavy grey soils (vertisols). Roadside reserves are defined in in Section 4.

b) All sites and plots surveyed in the field in June 2021 within Keelendi and Edenmore, were native vegetation on and prior to 31 October 2014 (World Imagery time series satellite images included the period 15 January 2009 to 2 August 2021, …). Between the imagery dated 23 June 2018 and when I visited the sites in the field in June 2021 (as marked in Figure 6 and Figure 7 and Table 6) Keelendi-Edenmore was variously cleared using a combination of mechanical clearing and herbicide poisoning.

c) Multiple clearing events over time using herbicide poisoning at multiple locations characterise the conversion of native vegetation (ground cover and regrowth trees) in the Keelendi-Edenmore to dryland cropping (e.g. Figure 39 compare a) and b) and compare c) and d)).

d) Mechanical and chemical (herbicide poisoning) land management regimes have transformed or converted candidate native vegetation (Plant Community Type, or PCT 1) to non-native vegetation i.e. cropping production landscapes (Figure 39. This change in land use has resulted in irreversible harm to Category 2 native vegetation in short, medium and longer term).

e) Native trees form part of the overstorey structure of a plant community.

f) Grey cracking clays have been demarcated from red sandy soils (non-cropping soils). Sites surveyed in the field focussed on grey cracking clays. An analysis of those sites within the broader landscape (using World Imagery, ESRI nd) shows the landscape of KeelendiEdenmore has been transformed from native vegetation used for grazing, to a cropped landscape (Figure 43) …(Figure 53, Figure 54, Figure 55, Figure 56, Figure 57) shows this landscape level change i.e. conversion of native vegetation to non-native vegetation. This was validated when I was driving, within the Area of Interest, where I observed that previously informal and irregular farm roads, consistent with farm tracks used for managing grazing animals and accessing farm dams, had been generally replaced with a formal network of rectangular and geometric farm roads consistent with farm tracks used for managing cropped areas. This recent and obvious change in farm-level infrastructure was likely developed to support the recently created cropping landscape (Figure 43). These observations were made as part of upscaling the site-based assessments of native vegetation condition to the landscape level using the Vegetation Assets States and Transitions (VAST) framework (Thackway and Specht 2015; Thackway and Freudenberger 2016). These observations were confirmed after the field survey using high resolution World Imagery (ESRI nd).

Prior to visiting in June 2021 used high resolution World Imagery (ESRI nd) to understand the relationship at a fine scale between transformation of native vegetation to non-native vegetation (cropping). In accordance with the Methodology outlined in this report, I observed isolated paddock trees and patches of remnant woodland showing the crowns dying off i.e. browned-off. Browned-off tree crowns occurred only on grey soils and not on the red soils. Browned-off tree crowns are consistent with broadacre herbicide applied with a crop duster light aircraft.

This report found that no single poisoning event occurred on Keelendi-Edenmore. Rather a progression of multiple poisonings has been used over time in different parts of the Area of Interest.

The Area of Interest and the site surveyed has recently changed land use from one of grazing native vegetation with farm dams to that of broadacre cropping i.e. the annual production of dryland wheat (e.g. Figure 39). This land use change has focused on the heavy grey soils and not the red soils. Seen in this context i.e. the annual production of wheat, Category 2 listed native trees have been, and likely continue to be repeatedly sprayed with herbicide.

…

This report confines my opinions and comments to those sites and plots, and the associated Lots and DPs that were surveyed in the field in June 2021.

2.1 Extent of clearing by way of poisoning

Sites and plots were assessed in the field between Tuesday 22 June to Sunday 27 June 2021.

Metrics were used to assess the clearing impacts of herbicide poisoning on tree health and vitality. Lower scores for tree health and vitality were recorded within Keelendi-Edenmore (treatments) compared to similarly assessed Roadside reserves (i.e. controls), indicating multiple recent herbicide poisoning events. Metrics of tree health and vitality showed stressed trees (yellowing and browning off of tree crowns) and recently dead trees within the cropped area. Conventional herbicide applications (tractor drawn boom sprays) in cropping systems do not result in immediate evidence of widespread tree stress and tree death. Evidence of recent and widespread tree stress and tree death are indicative of aerially applied herbicide.

Laboratory results showed the presence of contact post-emergent non-selective herbicide (used to control woody and non-woody plants (including 2,4-D and Glyphosate) and pre-emergent herbicide (used to control woody plants as well as annual and perennial grasses (Simazine and Diuron).

Metrics used to assess impacts of herbicide poisoning on the resilience and regenerative capacity showed to a lesser extent, poorer scores for the resilience and regenerative capacity within Keelendi-Edenmore, which is attributed to multiple poisonings with a mix of aerially applied herbicides (pre-emergent and non-selective). Figure 37 supports this, noting that defoliation occurs within six months of each spray event (i.e. site 11, plots 37-40). Figure 39 also supports this showing separate spraying events occurred before June 2020 and again before August 2021.

Tests for herbicide residues within soil samples and biological samples (leaves) showed mixed results. No herbicide residues were recovered from the soil cores. However, a mix of four different herbicides were detected in two biological (leaf) samples.

Ongoing annual application of a mix of pre-emergent herbicides and post emergent, non-selective herbicides effectively controls and restricts the capacity of tree, shrub and grass-like growth forms to reproduce and regenerate…

…

The extent of clearing determined by Mr Leo Watts … was accepted as the zonal areas to be used to guide field sampling between Tuesday 22 to Sunday 27 June 2021. …

…

Before the clearing event/s, the Coolibah – Black Box woodland EEC included three woodland types (PCT 37, PCT 39 and PCT 40). The impact of herbicide on these woodlands has resulted in dying and dead native vegetation (trees as well as native groundcover) within what are now cropping paddocks.

Before the clearing event/s, the Myall woodland EEC include only areas of PCT 27 with a dominant overstorey of Acacia pendula (Weeping Myall). Clearing of these woodlands has also resulted in dying and dead native vegetation (trees as well as native groundcover) within what are now cropping paddocks.

…

2.2.2. In relation to “paddock trees”

In my opinion land use change involving initially mechanical clearing, followed by multiple applications of herbicides has irreversibly caused harm to the Category 2 native plant communities.

In some areas (site 2) the native vegetation, comprising varying tree growth stages (age classes) was thinned leaving residual “paddock trees” (…Figure 53). Figure 42 shows an enlargement of this area, where both deliberate and inadvertent herbicide application/s have been used to harm trees. Figure 42 shows severe herbicide poisoning (browned-off tree crowns) within KeelendiEdenmore along with equally impacted tree crowns in the Roadside reserves, outside the Area of Interest. Herbicide poisoning inside the Area of Interest is consistent with aerial spraying associated with cropping production. Inadvertent spray drift from aerial spraying of the cropping paddock onto the adjacent native vegetation in the Roadside reserves are consistent with aerial herbicide spray that drifted from east to west.

The concentration of herbicide that has impacted paddock trees via aerial spraying events is relatively low. The dosage is consistent with herbicide used during wheat production i.e. herbicides are applied to control and restrict weeds in wheat, rather than directly and quickly killing trees. Arguably, from a production perspective, the fact that paddock trees are repeatedly sprayed indicates the harm caused to the trees is secondary and collateral damage.

Older tree growth stages have a remarkable ability to survive and recover from one off herbicide injury. Larger trees sprayed multiple times die slowly over several years rather than dying quickly.

2.2.3. In relation to identified native vegetation

Both deliberate and inadvertent application/s of herbicide have been used to harm native vegetation. This includes within the farming area of the Area of Interest, riparian zones and outside the Area of Interest in the Roadside reserves.

Herbicide poisoning observed in the World Imagery (e.g Figure 39) within the Area of Interest showed dying trees and dead high-quality remnant woodland patches. Effects of clearing using herbicide poisoning on the health and vitality of native trees, shrubs, grasses and herbs was evidenced by yellowing and browning off and recently dead trees, shrubs, grasses and herbs within and adjacent to cropped areas. This was confirmed during the field survey (Figure 31 and Figure 34).

Recent herbicide poisoning of Black-Box Coolibah woodland (e.g. browned-off tree crowns Figure 26 and Figures 31 and 32) including trees, shrubs, grasses and herbs) and associated isolated paddock trees (within recently mechanically cleared and converted to cropping)…Recent herbicide poisoning is shown by trees, shrubs and grasses at the same site that are browned-off. In my opinion clearing of the Black-Box Coolibah woodland EEC (recognised under both state and Federal legislation) has also removed and degraded the habitats of many plant and animal species…

Professor Thackway stated that to spray a tree with herbicide would not necessarily kill it. Rather, covering the tree with herbicide would likely defoliate it and its presentation would be that of no colour on imagery.
During oral evidence, Prof Thackway was taken to several photographs that showed trees that had been purportedly sprayed and those that had not by way of comparison (T148:16-T150:19). He concluded that there was evidence of repeated spraying across multiple paddocks (T151:13-14). He was asked about the timing of the spraying and his response suggested a correlation between the spraying events and the crop produced in 2018, 2019 and 2020, with the trees showing different effects (T153:02-31).
Professor Thackway was unable to state, however, when the herbicide, the effects of which he was opining on, was applied. That is, before, during, or after the charge periods.
This is, in my view, understandable given the delay between the alleged poisoning and the collection of the samples by him for testing, in addition to the fact that Keelendi Farms and Edenmore Farms continued to engage in farming activities, especially cropping, requiring weed suppression by the application of herbicide. For example, Appendix 11 to the first Prof Thackway report contained a table of all of the Spray Tech invoices for work carried out on the properties between 11 March 2020 and 14 June 2021. It showed 86 chemical applications in this period, including for “Fallow Herbicide”. Professor Thackway does not say whether the damage to vegetation that he observed in June 2021, and the samples that he took at that time, were attributable to later spraying activities, rather than the spraying activities occurring during the charge periods in 2021.

Dr Kathryn Hammill

Finally, the Secretary relied upon the opinions of Dr Kathryn Hammill, a consultant ecologist with native vegetation expertise, including in the surveying and mapping of native vegetation. Dr Hammill produced two reports dated 16 June 2022 (“Dr Hammill’s first report”) and 18 February 2025, the latter being a response to Dr Robertson’s first report (“Dr Hammill’s second report”). Dr Hammill has particular expertise in the use of GPS as well as in the analysis and interpretation of spatial datasets such as satellite images and vegetation maps. Dr Hammill’s evidence was primarily relied upon by the prosecutor to prove that what was allegedly cleared was “native vegetation” as defined in the LLSA.
Dr Hammill was asked to address the following questions in her first report:

a) Did native vegetation, within the meaning of the LLS Act, exist in the areas of interest immediately prior to February 2020?

b) If the answer to (a) is yes, please specify the species and communities of vegetation in the areas of interest;

c) Whether there has been any clearing of native vegetation in the areas of interest, and if so:

i. The extent of such clearing;

ii. Whether such clearing has caused or is likely to cause any harm to the environment and if so, the extent of any such harm;

iii. Whether any cleared vegetation in the areas of interest was “regrowth” within the meaning of the LLS Act.

d) Please include within the report a sequence of photograph taken during your site inspection.

e) Please include within the report a table of species identified during your site inspections noting within which polygon each species was identified.

Dr Hammill used a georeferenced shapefile of the polygons identified in the first Watts report, and applied Geographic Information Systems (“GIS”) data; cadastral and topography spatial datasets; the “New South Wales Flora Online” database, also known as the PlantNet database; other satellite imagery; threatened species records from NSW BioNet; published vegetation maps for the region prepared by the Office of Environment and Heritage (“OEH”); and the published NSW Plant Community Types classification system, to assess what species were likely to have been removed during the alleged clearing. She relied upon the PlantNet database to determine whether the species were native or not. For some species, although she was unable to conclude that they were present as individuals or populations at the time of their alleged removal, she was able to infer their presence from historical records of threatened species, which indicated a high likelihood of their presence at the relevant time in the areas affected by the alleged clearing.
Dr Hammill inspected the properties between 22 to 25 June 2021. She went to various sites, including sites within the charge areas, during which she observed the health of the vegetation and identified species and opined as to whether or not they were native. She observed a variety of native trees, shrubs, grasses and plant species that were “native vegetation” as defined, including two endangered ecological communities (“EECs”). Dr Hammill documented her assessment of each of the sites that she visited on the properties.
Dr Hammill was instructed to use areas identified by Watts in making her observations and findings.
Dr Hammill’s conclusions in her first report were that:

18 I observed a total of 85 native species at the Properties in June 2021, comprising: 7 native tree species, 22 native shrub species, 17 native grass and grass-like species, 36 native forb species, 1 native fern species and 2 species of other growth form types. I provide a list of these species in… . Based on my observations of dead and dying plants across the Properties, I consider these species have been affected by clearing, including by poisoning, in the areas I inspected.

19 I observed the following recognised native vegetation communities (Plant Community Types, PCT) at the Properties which I consider have been affected by clearing, including by poisoning.

● PCT 27 Weeping Myall open woodland of the Darling Riverine Plains Bioregion and Brigalow Belt South Bioregion.

● PCT 37 Black Box woodland wetland on NSW central and northern floodplains including the Darling Riverine Plains Bioregion and Brigalow Belt South Bioregion.

● PCT 39 Coolibah - River Coobah - Lignum woodland wetland of frequently flooded floodplains mainly in the Darling Riverine Plains Bioregion.

● PCT 40 Coolibah open woodland wetland with chenopod/grassy ground cover on grey and brown clay floodplains.

● PCT 247 Lignum shrubland wetland on regularly flooded alluvial depressions in the Brigalow Belt South Bioregion and Darling Riverine Plains Bioregion.

Extent of clearing

20 In the areas I inspected in June 2021, I observed extensive areas of dead and dying native trees, shrubs and groundcover amongst ploughed paddocks in various stages of cropping. The dead and dying native vegetation occurred in a matrix of patches of bare ground, paddocks with ploughing furrows rows of stubble and some areas with a newly-emerged green crop, and around pools of water lacking any live wetland vegetation. …

21 The observations I made on site are broadly consistent with the findings of the aerial imagery analysis of Leo Watts of vegetation change and decline in canopy health. I observed trees in an extremely poor state of heath, often dead or dying with very few live canopy leaves. The affected trees included Acacia pendula, Casuarina cristata, Eucalyptus coolabah, Eucalyptus largiflorens and occasionally Eucalyptus camaldulensis. I observed dead groundcover and wetland vegetation with leaves appearing grey or brown and ‘burnt’, in open areas without trees and in patches under trees across the areas of interest. Based on my observation on site, I broadly agree with the extent of clearing identified by Leo Watts in his aerial imagery analysis report and associated maps.

22 Overall, the dead and dying vegetation I observed represents a significant and extensive loss of open native woodlands on the Namoi River floodplain…

Harm to the environment

23 Many of the trees I observed to be dead or dying were large, hollow-bearing trees of Eucalyptus coolibah (Coolibah), Eucalyptus largiflorens (Black Box), Casuarina cristata (Belah) trees, including isolated paddock trees, and also trees which were part of intact woodland ecosystems and riparian corridors. The size of the dead and dying trees indicates they would be of substantial age, at least 300 years old. These trees are almost irreplaceable due to their age and the importance of the habitat they provide…

…

26 I observed dead and dying remnants of native wetlands in the eastern part of Edenmore (refer to observations at site 13, Appendix A). …

27 I observed significant erosion and death of native vegetation, including trees and groundcover, along the immediate the riparian corridor of the Namoi River, Baradine Creek and other tributaries of the Namoi River in this location. …

…

32 As shown in the photographs in Appendix A, the stem diameter of many of the dead and dying trees I observed on the Properties was substantial, and greater than 50 cm diameter. Many of the trees also contained substantial sized hollows of 10 cm diameter or more in large branches. Both the size (stem diameter) of trees and the presence of large hollows are indicators of very old trees.

33 To ascertain the potential age of trees observed on the Properties, I have reviewed a study of dendrochronology of Eucalyptus coolabah in the southern QLD the Diamantina and Warburton River Systems (Gillen 2017)…

34 The Gillen (2017) study was undertaken in broadly a similar semi-arid floodplain landscape and climatic zone as the subject Properties. The Gillen (2017) study assessed the growth rates of Coolibah, which is noted as being a keystone species of the inland river systems. The study used an analysis of carbon isotopes embedded in the woody tissue of stems of Coolibah trees. The analysis used samples from Coolibah within the woody stem tissue collected at distances from the outer trunk. The 14C isotope content in the stem wood was correlated with the known atmospheric 14C at particular times during the last century. This enabled the scientists to measure the amount of growth of the stem (i.e. the stem diameter increase) over a certain time period (number of years). The stem diameter growth rate was determined as being between 1.3mm to 1.8mm per annum for the Coolibah trees examined. At this growth rate a tree with a 30 cm (300mm) diameter trunk would be 166 years old and a tree with a 50 cm (500mm) trunk diameter would be 277 years old (based on the faster growth rate of 1.8 mm per annum). Many of the trees I observed on the site were far larger in diameter than 50 cm and, based on the analysis of (Gillen 2017), I would say that these trees would be at least 300 years old.

35 I also note the general consensus in the scientific community that trees with large hollows are at least 100 years, since it takes at least this length of time for trees to form large hollows. I refer the reader to the information presented in the tree hollow ‘factsheet’ in…which identifies a time period of at least 100 years is required before trees start to develop hollows. In semi-arid landscape where tree growth is relative slow, I think the general understanding of the 100 years required for trees to develop hollows would be a considerable underestimate for the age at which Coolibah and Black Box develop large hollows.

36 I have reviewed the findings of Leo Watts’ aerial imagery analysis which shows many of the same trees and apparent native groundcover vegetation were present in the subject areas prior to 1990…

37 On the basis of the findings outlined in paragraphs 29 to 31 above, I conclude that the trees, where these occurred on the Properties prior to the alleged clearing, and also likely the surrounding native species assemblage and groundcover, was present before 1990 and is therefore not regrowth.

In Dr Hammill’s second report she was asked her views on:

● Maps prepared by Leo Watts of the total area cleared on the Properties, issued to me as pdf aps on 5/02/2025, ahead of the following supplementary report by Leo Watts containing the same maps.

● ‘Supplementary Report on the Interpretation of Aerial Imagery over land located at Keelendi Farms and Edenmore Farms, Pilliga, NSW (the Properties) from 1 January 2020 to 8 January 2021’, a report dated 6 February comprising a revised opinion of Leo Watts, provided to me as a pdf copy on 8/02/2025.

She concluded that none of the sites that she assessed on the properties in June 2021 were located in the areas removed from the polygons in Watts’s supplementary imagery analysis contained in the second Watts report. That is, none of her sites were within the excised areas subsequently identified by Watts as bare ground. As a consequence, Watts’s changes to his polygons did not affect her findings on the plant species listed in her first report.
Dr Hammill also provided a further opinion on species of native vegetation present in the polygons identified by Watts that were not referred to in her first report. Otherwise, her conclusions on the loss of threatened species habitats, the clearing of threatened ecological communities, and the harm to aquatic ecosystems that she expressed in her first report remained unchanged.
Table 2 in Dr Hammill’s second report set out in detail the native vegetation that, in her opinion, was present across the properties and Watts’s “areas of interest”. It demonstrates that the polygons and lots the subject of the charges would have, at some point in time, been populated with “native vegetation” for the purpose of s 60N of the LLSA.
Dr Hammill was challenged on the accuracy of her opinions, especially in respect of sites that she had not visited. Her response was that, “I’ve got no doubt that all the species that I’ve inferred would be somewhere in the polygons” (T284:14-15).
In terms of identifying the vegetation that had existed in the areas allegedly cleared, while she acknowledged that “it makes it more difficult” to test what vegetation had been present prior to clearing, Dr Hammill was of the view that features of the landscape and historical imagery nevertheless “provide the opportunity to infer what was highly likely to be there” (T263:01-05). As she explained (T278:16-26):

…So we - we were discussing before, and it’s also mentioned in different places in my reports that, generally speaking, it’s the same landscape, so it’s the same soil type. It’s got similar - it’s in the floodplain of the Namoi River. The soil types were generally the same across almost the entire properties. There was some red soil in the far south-east, which is not within a polygon where Eucalyptus populneus comes in. It’s a different vegetation community. However, all of the locations I went to and drove through to get to other places were all on the same Namoi Valley black soil plains landscape, which defines - well, it’s a strong influence on the vegetation that’s going to be there. So you’re not going to suddenly get different eucalypt types pop into that landscape.

Dr Hammill explained that the OEH mapping that she relied upon was partly predictive and partly survey based. It was not, however, intended to indicate whether a particular plant exists at a specific location.
Dr Hammill was not cross-examined about the native vegetation that she identified in any particular polygon in the charge areas.
Although not instructed to do so, Dr Hammill proffered the opinion that, consistent with her observations, the plants examined had been affected by herbicide poisoning (T271:40-46) and their condition could not be “attributed to another cause across the extent that I saw” (T272:21-25). She did not see the effects of any other causes, such as insect attack (T272:33-36). Having said this, she did not specifically examine for insect or parasite infestation, but she noted that “she did not go looking for herbicide either. It was very present” (T272:48-49) and that “the appearance of the landscape was just so strikingly impacted” (T285:33-34). In other words, “it’s quite recognisable as an ecologist to see an unnatural, extensive death of plants” (T286:31-32). She compared the photographs of plants affected by herbicide to the photograph attached to Dr Robertson’s report showing what he had described as “dieback” and noted that the photographs were very similar (T289:06-14). Her view was not formed as a result of her conducting a chemical analysis, rather it was based upon her experience as an ecologist.
But as her cross-examination revealed, Dr Hammill’s observations with respect to the application of herbicide to vegetation were not always accurate. For example, there was at least one instance where a tree that she had concluded was “dead” had made a full recovery (T268:31-34).
The views of Dr Hammill concerning the presence of “native vegetation”, as that term is defined, in the areas identified in the summonses may be accepted. That is to say, within the polygons identified by Watts (and as subsequently modified by him) I find beyond reasonable doubt that at some stage “native vegetation” was present on the properties within the meaning of s 60B of the LLSA.
Having said this, however, it must be emphasised that Dr Hammill’s evidence does not prove to the criminal standard when the native vegetation was cleared or by what method or methods. As the evidence makes tolerably clear, both properties have been the subject of clearing prior to the charge periods, the clearing being primarily for agricultural purposes. When and where this clearing occurred, as the Secretary appeared to accept (T381:13-14, 381:19-20 and 388:37-38), is not known. For example, Smith and Shrimpton initially inspected the properties as part of an investigation into prior potentially unlawful clearing of native vegetation by mechanical means and Prof Thackway opined that both mechanical and chemical clearing had occurred on the properties within the charge areas and that the application of poison had occurred on multiple occasions at different points in time.

Defendants’ Evidence

In addition to tendering documents and cross-examining the witnesses referred to above, the defendants relied upon the opinion of Dr Robertson as contained in his first and second reports.
In his first report, Dr Robertson was asked to:

(a) visit the properties and comment on the observations made by Smith, Dr Hammill and Prof Thackway in their principal reports. In particular, he was asked to describe the present state of certain trees that had been identified by the witnesses for the Secretary as either dead or dying; (b) visit the road reserve outside the western boundary of Keelendi and provide his opinion on the extent to which the road reserve was an area that provided a useful ecological comparison with the land that was inside the boundaries of Keelendi and Edenmore as it existed in early 2020; (c) comment on the interpretation of the aerial and satellite imagery by Watts; and (d) provide “an assessment” of the evidence of Smith and the reports of Dr Hammill and Prof Thackway.

The findings arrived at in Dr Robertson’s first report were as follows:

After my critique of the reports prepared for the Prosecution, I concluded there are major challenges to be overcome to verify that poisoning within the charge periods of February 13 to 23 March 2020 was responsible for dieback in the areas of interest. This is for several reasons.

● Ruling out other factors: Within the areas of interest, and in surrounding farmlands, there are many possible factors that could cause dieback and so there is substantial potential for misdiagnosing all or part of the dieback cause. I have endeavoured to discuss some obvious factors, but my list is not exhaustive. I found that none of the reports for the Prosecution considered the full array of causes for the dieback or worked to rule them out.

● Narrow time and late field investigations: The charge periods spanning February 13 to 23 March 2020 is specific and represent a short time – especially within the context of a cropping farm where farm work is ongoing. I believe that there were likely to have been substantial farm works before, during and after the charge period that could have contributed to observed vegetation changes. The Prosecution experts should have recognised this and found out what works occurred in the area. I note that the earliest that investigations took place on the ground after the charge period was by Smith (2022), who visited some of the areas of interest several months after the charge periods. However, the two scientists who visited the areas of interest for more detailed botanical studies did not do so until 15 months after the charge dates. That is, more than a year after the alleged poisoning. This makes such studies extremely difficult because so much could have occurred in the areas of interest during that time. I note that neither Hammill Stone (2022) nor Thackway (2023) focussed on the impacts arising during the time and neither expert has acknowledged the difficulties of doing so.

● Size and variability of areas of interest: The areas of interest are large (spanning thousands of hectares and spread over kilometres), varying in terms of vegetation, soils and other factors. The witnesses who conducted on the ground investigations have not adequately dealt with or sampled to account for these variations across the areas of interest. As discussed earlier in this report, I conclude that the authors of the prosecutors reports have not taken sufficient samples of vegetation, soils and foliage to test adequately for herbicide occurrences. They have also not sampled widely enough across the site to be sure to capture variations in such areas. These flaws mean that the conclusions from the prosecution evidence cannot be relied upon.

● Remote sensing: Due to the size of the areas of interest, it is essential to use remote sensing to access the vegetation changes that have occurred within them during the charge periods. Watts (2022) conducted the remote sensing work but is not a botanist and did not go to the site to validate his analysis, or use reference sites. I also believe that he is not familiar with the farm work that occurs on the subject site. In my experience, remote sensing should always be supported by fieldwork to help clarify what is being detected within aerial imagery and to provide opportunities for calibration. No such work was done and so Watts (2022) did not obtain an accurate understanding of what he was looking at on ground. Neither Hammill Stone (2022) nor Thackway (2023) provided field data to inform Watts work. In my view the lack of coordination of field work and remote sensing means that the results of Watts work are questionable.

● Use of Infrared imagery: Infrared imagery has been used by Watts (2022) to pick out actively growing vegetation and to look for suppressed or removed vegetation. While such false colour imagery is well suited to such analysis, there is potential for misinterpretation of what images show. This is particularly so due a lack of fieldwork to help interpret such imagery. Moreover, analysis of such imagery should take place with due consideration to confounding factors such as soil type, soil moisture and soiI salinity, which was not done by Watts (2022). I have found examples of such flaws in Watts (2022) and believe that they call into question the nature and extent of vegetation suppression or removal observed by Watts and attributed to works during the charge periods.

● Temporal analysis of dieback: My visit to some of the areas of interest in December 2024 illustrated to me that much of the “dead and dying” vegetation reported by Hammill Stone (2022) and impacted trees reported by Thackway (2023) and Smith (2022) remains alive and there is a high proportion of trees and shrubs in good condition within these areas. All of the fieldwork conducted on behalf of the prosecution represented a snapshot at one point in time. This calls into question the impact assessments within such reports.

It was Dr Robertson’s opinion that, first, the aerial images taken in 2018, 2020 and 2023 and interpreted by Prof Thackway, showed that patches of trees and grassland had marginally changed during this time, and in particular, that the image taken in 2023 appeared to show healthy Weeping Myall trees. Dr Robertson found these trees during his site inspection on 18 December 2024, as evidenced by the photographs taken by him.
Second, he stated that:

In my SoE I commented on the work of Watts (2022) where he provided imagery showing what was stated to be false colour imagery of the northwestern polygon in the area of interest that showed blue shading that he described as being moist or containing underlying water. In my SoE I obtained the same imagery and noted that the imagery showed what appeared to be extremely dry, bare land in those areas.

My use of rainfall and Namoi River flow data is consistent with my opinion that the area referred to as moist/wet by Watts (2022) was very dry in 2020. I note that Watts (2020) has prepared other figures with what he refers to as “natural” colour imagery that has what I believe is not natural as it has blue colouration on the same areas of the area of interest. I query the validity of the other mapping too. I believe that it confirms the need to use ground truthing of such sites, and I also believe that other data such as rainfall and river flows would be very informative. Such work would be essential to avoid misidentifying dry areas as moist/wet areas.

Rainfall data presented in Appendix A show that the rainfall in 2019 was extremely low when compared to the average annual rainfall data for the Pilliga Post Office weather station (approximately 5km from Edenmore and Keelendi). The rainfall in October, November and December leading up to the alleged poisoning dates in February/March 2020 was extremely low (eg 3.4 mm in October and no rain in December). In the Pilliga high summer temperatures result in high evaporation and so such low rainfalls would, in my opinion, be lost through evaporation and would not be likely to flood large parts of the areas of interest.

I have also included rainfall data from the Pilliga Post Office station for 2020. I note that much higher rainfall occurred in February and March of that year, but January had very low rainfall.

Similarly, the Namoi Flow report in Appendix B provides a summary diagram in which it states:

February 2019 The Lower Namoi was declared to be in Stage 4 – Critical Drought. The Namoi Valley had received its lowest inflows since 1918.

I believe that the only way that the area could have been wet was via heavy rains and/or flooding by the flood channels associated with the Namoi River. Neither the rainfall nor flow data support this as the area was enduring the worst drought since European settlement occurred.

Dr Robertson’s evidence suffered from a number of deficiencies. First, it was no more than, as he described it, a “critique” of the expert evidence of the prosecution and of Smith’s evidence, premised upon a hypothetical methodology for investigating the “dieback” of plants generally, as devised by him. The term “dieback” was adopted by Dr Robertson from Meat and Livestock Australia and a fact sheet published by CottonInfo dated September 2018 entitled Tree dieback identification and management guide.
Dr Robertson’s evidence provided limited assistance in assessing the expert evidence of the prosecution. He was not instructed to, nor did he undertake, an assessment of the likelihood that dieback was the cause of the browning and unhealthy appearance of the vegetation witnessed by the prosecution’s lay and expert witnesses in the charge areas (T328:42-48).
Furthermore, Dr Robertson was not provided with all of the prosecution evidence. He did not undertake his own analysis and interpretation of the aerial and satellite photography. He did not interrogate data to understand the vegetation that existed on the property and was unable to comment on what state the vegetation was in prior to the first charge period or before 2020.
Having said this, it must be recalled that the defendants bear no part of the burden of having to establish the cause of the moribund appearance or even the absence of vegetation noted by the witnesses. At all times, it was incumbent upon the Secretary to establish that the defendants had “cleared” “native vegetation” (as those terms are defined) for the purposes of s 60N of the LLSA. The criticism of Dr Robertson in this regard is directed to the utility of his evidence and the weight to be placed upon it as a reasonable alternative hypothesis accounting for the impacted vegetation.
Second, Dr Robertson accepted that “elements” of his hypothetical ideal methodology (a method he was not instructed to devise) set out in his first report in order to determine whether the clearing occurred by poisoning, were “not possible to completely get” (T303:10-34). His acknowledgement was apposite. The methodology was premised upon an unrealistic assumption that the person employing this process was wholly unencumbered by resourcing, timing or access constraints. For example, obtaining baseline samples to record vegetation changes. This would have necessitated O’Brien permitting sampling to be carried out on the properties prior to any clearing, lawful or otherwise (T309:31). In making the concession that his preferred methodology was difficult to achieve given the limitations of time, resources and access, Dr Robertson appeared to acknowledge that a more orthodox approach of obtaining baseline data about a particular site was to have regard to other areas as a reference point and to examine published information on each plant community type (T308:11-24). This was precisely the method utilised by Dr Hammill.

Was There “Clearing” of “Native Vegetation” in the Charge Areas During the Charge Periods?

Method of Clearing Particularised by the Prosecutor

Each summons particularised the “method of clearing” as “by means of poisoning”, which was refined further by the Secretary as the application of herbicide or other chemical poisoning. This was consistent with the Prosecutor’s Amended Statement of Facts and original Statement of Facts which likewise stated with respect to both Keelendi and Edenmore that:

The clearing occurred by the following methods:

i. herbicide/chemical poisoning.

The Secretary submitted that it was sufficient if the prosecution demonstrated to the criminal standard that clearing occurred that was either mechanical or by chemical poisoning. There was evidence from both Watts and Prof Thackway in their first reports that the clearing had been both mechanical and herbicidal.
The Secretary accepted in her written submissions that the prosecution case was “a means” by which the landscapes within multiple polygon areas the subject of the charges were cleared of native vegetation during February and March 2020 for the purpose of crop cultivation, and moreover, that the summonses and her alleged facts referred only to poisoning. However, she noted that it was evident from the expert evidence that the defendants had been in receipt of for nearly two years, that other potential forms of clearing had occurred on the land which formed part of the charges.
Particulars serve the function of informing the accused the case they have to meet (Johnson v Miller [1937] HCA 77; (1937) 59 CLR 467 at 489 per Dixon J and King v The Queen [1986] HCA 59; (1986) 161 CLR 423 at 425 per Murphy J).
In the present proceedings, while the use of a tractor or other machinery by the defendants to remove, kill, destroy or fell any native vegetation on the property would amount to “clearing” as that term is defined in s 60C of the LLSA, it is tolerably clear that until the hearing of the charges, the Secretary had consistently maintained the allegation that the “method of clearing” was “by means of poisoning”. It was on this basis that the defendants prepared their defence, including their evidence. In the present case, in light of the particulars provided to the defendants and relied upon by them, it would be misleading if the Secretary was permitted to rely upon mechanical clearing in order to prove the charges (Jago v District Court of N**ew S**outh W**ales [1986] HCA 46; (1989) 168 CLR 23 at 57 per Deane J). Given that particulars are, as the prosecutor accepted, a means of providing natural justice, and that the prosecutor elevated clearing by poisoning as an essential or consequential fact alleged in the summonses, then as a matter of fairness to the defendants, any late change of position by the Secretary ought not be permitted.
Another difficulty facing the Secretary is that even if no unfairness arises out of the prosecutor relying upon mechanical clearing to constitute the offences, at its highest, the evidence is that some “other potential forms of clearing likely also occurred”. In my view, neither the aerial analysis of Watts and Prof Thackway nor the inferential evidence of Smith and Shrimpton of “burn piles”, proves beyond reasonable doubt that mechanical clearing during the charge periods occurred.
I therefore reject any reliance by the Secretary upon mechanical clearing as a basis for the unlawful clearing of the native vegetation charges.

The Meaning of the Term “Clearing”

The defendants submitted that, properly construed, “clearing” had to result in the complete removal or death of the native vegetation, that is, the “loss of the living plant”. Applying poison such as a herbicide could not constitute “clearing” unless it resulted in the death of the native vegetation.
Accordingly, where herbicide is applied to vegetation, deliberately or unintentionally (for example, as a result of drift), the application does not constitute “clearing” if the application injures the vegetation and the vegetation subsequently recovers. It only constitutes “clearing” for the purpose of the LLSA if the application of the herbicide is demonstrated to have killed the native vegetation. It is only at this point that the offence is complete.
The defendants relied upon the following evidence:

(a) first, Watts did not conclude that any vegetation had died, rather the opinions that he expressed were framed as an “attempt to supress”, “remove” or otherwise adversely impact the health of vegetation. He observed that the vegetation in various locations on the properties had the “appearance of” being “unhealthy” or “dying or dead”. He was not, however, able to determine any cause; (b) second, Dr Hammill accepted that the vegetation could look dead but not in fact be dead. This was starkly illustrated by a tree that she asserted was dead but was plainly alive (T268:29-34). She did not know how many of the trees that she had identified in her evidence as “dead” or “apparently dead” were presently alive (T268:01-T269:50). As she acknowledged, vegetation that appeared to be dying or unhealthy can recover (T267:14-34). No samples were taken by her of vegetation to assess its health (T271:22-24). Her evidence was consistent with that of Prof Thackway who stated that healthy trees that have been recently sprayed can look dead but are still alive, and that older tree growth stages have a remarkable ability to survive and recover from herbicide injury; and (c) third, the lay observations of Smith and Shrimpton concerning the vegetation that they saw dying or was dead could not be relied upon to prove this fact.

In Turnbull v Chief Executive of the Office of Environment and Heritage [2017] NSWCA 161, Sackville AJA (with whom Basten and Meagher JJA agreed) stated in respect of a nearly identical definition of “clearing” to that in s 60C(b) of the LLSA (s 7(b) of the NVA), that the definition ought not be narrowly construed (at [72]):

72 The primary Judge’s construction of s 6(1)(c) of the NV Act receives support from s 7, which defines the expression “clearing native vegetation” to include “killing, destroying, poisoning, ringbarking, uprooting or burning native vegetation”. This language is entirely apt to describe the destruction of seedbanks or other subsoil reproductive mechanisms that lie dormant until seasonal conditions or rainfall allow them to flourish. Having regard to the objects of the legislation, there is no reason to give s 7 any narrow construction. Quite the contrary.

To construe the definition of “clearing” as necessitating the death of the vegetation is not warranted having regard to the text, context and purpose of s 60C of the LLSA. Commencing with the text, although the term is exclusively defined (“means”), the definition is, as was observed by Sackville AJA in Turnbull, broad in scope, encompassing both death and damage. For example, the “thinning” or even lopping of vegetation may not kill it. Likewise, “ringbarking” or “burning”. These acts may damage vegetation, but will not necessarily result in its annihilation. There are many species of native vegetation that in fact require “burning” in order to propagate, and will readily regenerate after having been burnt or lopped. The text of the definition of “clearing” therefore admits of harm that is separate to death.
Second, the context within which s 60C is located speaks against the narrow construction afforded to the term “clearing” by the defendants. For example, s 60B(3) extends the definition of “native vegetation” to dead plants in the circumstances prescribed therein. At the risk of being trite, dead plants cannot be rendered more dead by being “cleared” (on the defendants’ definition), but their removal can result in criminal sanction. By way of illustration, hollow bearing dead trees provide important habitat for a variety of fauna, some of which are threatened or critically endangered. Section 60B of the LLSA seeks to extend protection to dead native vegetation if it is situated on certain categories of vulnerable regulated land. Sections 60B, 60C and 60N makes it an offence to remove these trees or otherwise interfere with them, irrespective of the fact that they were dead.
Third, one of the objects of the LLSA is “to ensure the proper management of natural resources in the social, economic, and environmental interests of the State, consistently with the principles of ecologically sustainable development” as described in s 6(2) of the Protection of the Environment Administration Act 1991 (s 3(e) of the LLSA). Those principles are described in that Act as (emphasis added):

(2) For the purposes of subsection (1) (a), ecologically sustainable development requires the effective integration of social, economic and environmental considerations in decision-making processes. Ecologically sustainable development can be achieved through the implementation of the following principles and programs—

(a) the precautionary principle—namely, that if there are threats of serious or irreversible environmental damage, lack of full scientific certainty should not be used as a reason for postponing measures to prevent environmental degradation.

In the application of the precautionary principle, public and private decisions should be guided by—

(i) careful evaluation to avoid, wherever practicable, serious or irreversible damage to the environment, and

(ii) an assessment of the risk-weighted consequences of various options,

(b) inter-generational equity—namely, that the present generation should ensure that the health, diversity and productivity of the environment are maintained or enhanced for the benefit of future generations,

(c) conservation of biological diversity and ecological integrity—namely, that conservation of biological diversity and ecological integrity should be a fundamental consideration,

(d) improved valuation, pricing and incentive mechanisms—namely, that environmental factors should be included in the valuation of assets and services, such as—

(i) polluter pays—that is, those who generate pollution and waste should bear the cost of containment, avoidance or abatement,

(ii) the users of goods and services should pay prices based on the full life cycle of costs of providing goods and services, including the use of natural resources and assets and the ultimate disposal of any waste,

(iii) environmental goals, having been established, should be pursued in the most cost effective way, by establishing incentive structures, including market mechanisms, that enable those best placed to maximise benefits or minimise costs to develop their own solutions and responses to environmental problems.

It would, in my opinion, be antithetical to the purpose articulated in s 3(e) of the LLSA (as expanded by s 6(2) of the Protection of the Environment Administration Act), to confine the meaning of “clearing” in s 60C of the LLSA in the manner contended for by the defendants.
The defendants relied upon the definition of “clear” in cl 4 of the State Environmental Planning Policy (Vegetation in Non-Rural Areas) 2017 (“SEPP”) which defines that term to include:

clear vegetation, includes–

(a) cut down, fell, uproot, kill, poison, ringbark, burn or otherwise destroy the vegetation, or

(b) lop or otherwise remove a substantial part of the vegetation.

This definition does not assist the defendants. First, delegated legislation cannot generally be used to construe principal legislation (Master Education Services Pty Ltd v Ketchell [2008] HCA 38; (2008) 236 CLR 101 at [26] and Alphapharm Pty Ltd v H Lundbeck A/S [2014] HCA 42; (2014) 254 CLR 247 at [39]). Second, the defendants did not explain the application of the SEPP to the present proceedings. Third, the definition of “clear” in cl 4(b) makes it plain, in my view, that an act other than the complete destruction or removal of the vegetation is encompassed within its ambit and will result in its ‘clearing’.
As a consequence, the Court need not make any findings about whether or not the native vegetation was killed or removed as a consequence of the alleged poisoning by herbicide. Put another way, I do not consider that it is necessary for the Secretary to demonstrate to any standard that the defendants killed or removed vegetation in order to be guilty of the offences as charged.

Did “Clearing” of Vegetation Occur in the Charge Areas?

The defendants’ reasons for contending that the boundaries of the charge areas were uncertain, and therefore, that the prosecutor had not demonstrated beyond reasonable doubt that clearing had occurred in the charged areas, were fourfold:

(a) first, the areas of clearing identified by Watts in his first and second reports were derived from the 2021 Watts report, which was not in evidence; (b) second, Watts’s identification of the extent of vegetation changes as at 13 and 23 March 2020, was unreliable because he used low resolution images for vegetation analysis (10 m resolution). He failed to detect that large areas in his polygons were not covered with vegetation but were bare earth. As a result of observations by Dr Robertson, Watts revised his calculation of the cleared areas and reduced the total land alleged to have been cleared by 295 ha; (c) third, the boundaries of the category 2-regulated land used by Watts in preparing his polygons were not proven by the prosecutor. Watts had relied upon the shapefiles provided by Adam Hutchings from the Department on 9 June 2021, and Watts had assumed that that these files were accurate based upon checks and comparisons on other supplied datasets; and (d) fourth, a comparison could not be reliably made by Watts between his baseline aerial image of the properties as at 1 January 2020 and the images dated 13 February and 13 March 2020, given that the former was taken at a resolution of 1.5 m and the latter were taken at a resolution of 10 m. The 1 January 2020 image could not be used as a proxy for the state of the vegetation as at the commencement of either charge periods because Watts claimed not to be able to discern a difference in the state in the landscape as at those dates.

In the result, the defendants submitted that the Court could not safely rely upon Watts’s interpretation of the images that formed the basis of his opinions about vegetation change during the charge periods. Watts’s calculations of the size, location, and even timing of the areas allegedly cleared, were attended by reasonable doubt.
The defendants further submitted that this doubt infected the evidence of the witnesses who inspected the properties. It could not be concluded that their observations were relevant to the charge areas or to the charge periods. Furthermore, the evidence of alleged clearing with respect to one area could not be adopted and applied to other areas on the properties given their lack of geographic proximity.
There are several responses to the defendants’ arguments. Although Watts stated that he had considered his 2021 report in the preparation of the first Watts report, in my view, it does not follow that the first Watts report was derivative of the 2021 Watts report in a manner that taints the former’s contents and renders any reliance upon the first Watts report to be in error.
The defendants contended that there were a number of references in the first Watts report to the 2021 Watts report (for example, at paragraph 36e)). Moreover, in the first Watts report, Watts expressly states that he had considered, along with other documents, the 2021 Watts report in the preparation of his first report.
The “newly identified areas” the subject of the first Watts report are the areas the subject of the charges. They were “newly identified” only insofar as they were additional to the areas of modification identified in the 2021 Watts report, which are not the subject of these proceedings. The previously identified modified areas were not used by Watts to form a conclusion about whether clearing had occurred in the areas particularised in the summonses. Nothing identified by the defendants in their submissions (see, for example, at paragraph 5.4 of their written submissions) warrants the conclusion that Watts “depended” upon the 2021 Watts report to formulate his opinions.
It is therefore incorrect to conclude, as the defendants do, that his first report depends upon the opinions expressed in the 2021 Watts report. That is, read reasonably, it is clear that the opinions that Watts expresses in his first report are not impermissibly reliant upon the findings and conclusions made by him in the 2021 Watts report. To the contrary, the findings made in the first Watts report are exclusively a product of his comparison and his analysis of the images either contained in, or appended to, his first report.
To the extent that Watts makes, at various junctures in his first report, reference to his findings regarding the suppression of vegetation in areas that had, and had not, been previously cleared in the 2021 Watts report, this was, as he describes it, observational evidence only insofar as it did not inform, in the sense that it was not a step in the reasoning process leading to, the findings that he makes in respect of the “newly identified areas” of clearing. The boundaries of these “newly identified areas” were determined by comparing the various aerial images forming part of the first Watts report. It is not the case, and nor was it put to Watts in cross-examination, that the accuracy of the areas that had been cleared the subject of these proceedings was “unverifiable”. They were ‘verified’, as was established by Watts’s recalculation of areas allegedly affected by clearing in the second Watts report.

Q. How were they used as part of your assessment, these areas that you’ve circled?

A. So because I didn’t take any - undertake any ground truthing or I didn’t actually have a - a ground - a field - a field observation calibration point, I - I want - well, I needed - not that I wanted - I needed to establish my own calibration points to assess what I believed had - vegetation had changed and what - and what areas hadn’t changed. So for every date of imagery after my baseline point, 1 January 2020, when I observed - when I analysed the - colour or the - the natural colour imagery, I referred to areas that had changed or hadn’t changed back to those points in regards to their colour, their shapes or - or patterns, tones. Certainly, in the colour imagery - and also, particularly, I used that - I used those reference points in the colour infrared imagery as well because I found that the external calibration sites hadn’t changed at all throughout the whole of 2020. So almost in every instance - well, I think in every instance, when I looked - identified an area that had changed in the natural colour imagery and then I then went to the colour infrared imagery and compared those sites - or the colours within the colour infrared imagery to those test sites, it just confirmed my observations from the natural coloured imagery.

Q. You said that those external sides had not changed. Changed in what respect?

A. Had not changed in their - their appearance of vegetation coverage. So they had been - there had been no - and this is both - both in trees and ground cover, because I wasn’t - I wasn’t seeing changes in health. I wasn’t - I wasn’t seeing changes in land use of those area. There’s be no - there was no clearing of vegetation. There was no clearing of trees. There was no cropping or farming. The only changes that occurred was most likely just the density of vegetation as a course of the seasons. The rain events that occurred in February and March. I could identify that the - the spatial positions of features, vegetation features, particular trees, hadn’t changed. I could identify that the - that the patterns and the texture of ground cover hadn’t changed. That’s - ‘cause how I use it.

Q. And the lack of change was relevant how?

A. The lack of change provided - gave me confidence to - to determine that these sites outside the properties had not been affected by any natural cause. Or any man-made cause of clearing. And they provided, for me, excellent comparison points. Where I could compare all these external reference and control sites to areas that I identified as changed vegetation within the properties. And how they changed over time. So I could compare those reference sites between dates of imagery. And then compare the changes of vegetation within a date of imagery with those referenced sites.

Watts also credibly eschewed the suggestion that he had incorrectly interpreted the infrared imagery that he had employed to opine on the changes to the vegetation during the charge periods. In particular, that he had wrongly interpreted the red and pink colour shown on the images as healthy vegetation, when the properties had been subject to heavy rainfall in February 2020 after a period of severe drought. There was evidence before the Court that the interpretation of the colour red in infrared images can be rendered challenging by rain, because it gives rise to higher levels of moisture retention thereby affecting the colour of the image.
As his evidence indicates, Watts did not use the colour infrared imagery in isolation to express his opinions about the changes in vegetation on the properties. Rather, he compared the colours within the colour infrared imagery to natural colour test sites to confirm his observations from the natural coloured imagery.
Similarly, while Dr Robertson questioned Watts’s analysis of natural colour images, especially those that Watts had interpreted as indicating underlying water by reason of the presence of blue-tinged areas, I do not find that this relevantly undermined Watts’s evidence. The defendants’ submission that the drought made the presence of water or moisture unlikely is inconsistent with the rainfall event in February 2020 that occurred across the properties (the natural colour images were taken on 13 February and 13 March 2020).
In short, I reject the defendants’ challenge to Watts’s evidence on the basis that the boundaries of the charge areas are uncertain.
Having regard to the totality of Watts’s evidence I find beyond reasonable doubt that there was vegetation that was cleared on both Keelendi and Edenmore in the charge areas during the relevant charge periods. However, this does not mean that I accept Watts’s opinion as to the size, time or method of clearing, or whether he was able to discern that the vegetation cleared was something other than crop or weeds.
Given Watts’s inability to distinguish ground cover from bare earth in his first report, I cannot find beyond reasonable doubt that he is able to differentiate between ground cover and cultivated paddocks. I do not accept his evidence in this regard.
Furthermore, by his own admission, Watts’s evidence cannot be used to establish the method of clearing, that is, by the application of herbicide. This is important because a reasonable inference to draw is that some of the clearing that Watts observed as having occurred in the charge areas, including during the charge periods, occurred by mechanical means. This finding is consistent with the ongoing farming activities on the properties, the potential prior clearing of the land by mechanical means, the evidence of burn piles deposed to by Smith and Shrimpton, the limited number of positive results for herbicide in the samples taken by Smith and Prof Thackway (with no positive results having been returned from the soil samples), and even the abundant healthy vegetation present in the road reserves.

Did the Method of Clearing by the Application of Herbicide Occur During the Charge Periods?

The Secretary relied principally upon the evidence of Prof Thackway, Smith and Dr Hammill, to establish that the defendants had cleared the native vegetation by the application of poison or herbicide during the charge periods.
The defendants submitted that the prosecution case suffered from a lack of proof that any herbicide spraying which might have been carried out within the charge dates or within the charge areas caused the adverse effect on particular vegetation in particular parts of the properties observed on the ground by Smith and Shrimpton in June or August 2020, and 15 months after the end of the charge periods in June 2021 by Dr Hammill and Prof Thackway. The submission is compelling.
While Smith took a number of soil and vegetation samples during his site inspection of Keelendi between 11 and 14 August 2020, none of the soil samples returned positive results for herbicides. This was a mere five months after the charge periods. The defendants therefore submitted that the negative results provided a strong factual foundation for finding that none of the areas sampled by Smith had been sprayed with poison during the charge periods. Just as the presence of herbicide would have been a powerful indicator that the cleared areas were sprayed with herbicide during the charge periods, the absence of herbicides in the samples was arguably indicative of a contrary position.
Smith took four vegetation samples collected by hand. The species of the samples were not identified and two of the samples were collected from locations outside the charge areas. It is also not clear whether the remaining two samples were taken within the charge areas because, as the defendants contended, they were not plotted by reference to Watts’s “modified newly identified areas”.
As Dr Robertson noted, there was no explanation as to why so few samples were taken and why the samples were not collected until five months after the charge periods.
As explained above, Prof Thackway used an aerial image of the properties taken in 2018 to express his opinions about what activities had occurred on the land in 2020. He also took soil and vegetation samples during his site inspections in June 2021. However, at the risk of repetition, this was 15 months after the charge periods. Herbicide residue was detected in none of the 38 soil samples tested. No ground cover was tested.
Professor Thackway submitted two leaf samples taken from the south of Keelendi, but it was likewise not possible to determine whether or not the samples were taken within the charge areas because they were not plotted by refence to Watts’s polygons. Both leaf samples returned positive herbicide results, but of the four herbicides found to be present, one of the leaves tested positive to only one herbicide.
Professor Thackway reviewed a list of the aerial spraying events conducted by Spray Tech between February and April 2020 on the properties. He identified that herbicide was sprayed on nine occasions.

Was There Clearing of “Native Vegetation”?

The Secretary relied upon the evidence of Dr Hammill to demonstrate that the vegetation that was cleared from the charge areas was “native vegetation” as statutorily defined. Her evidence is detailed above.
Assuming that the Secretary can establish that vegetation was cleared from the charge areas during the charge periods, the defendants did not accept that it was “native vegetation”. It should be recalled that Watts stated that he could not form an opinion as to whether what he had identified as vegetation was native vegetation.
The defendants submitted that the type of vegetation that grew in the charge areas between 2019 and early 2020 after the previous mechanical clearing and during the drought was similarly unknown. As Dr Hammill stated, weeds were likely to have spread in areas that were previously cleared (T263:11-20).
The defendants further noted that Dr Hammill had inspected very little of the properties, and that her estimates of the plant community types and species in the charge areas and in the non-charge areas were based upon the OEH map for the region which was published in 2015. Dr Hammill described the OEH map as “predictive” and that it could be inconsistent “with actual vegetation types and their extent on the ground”. Based on her inspections of the properties, she rejected some of the plant type communities referred to in the OEH map as being present in the locations that she visited, and therefore, suggested different ones.
Accordingly, the defendants asserted that there was great uncertainty about what plant type communities existed in early 2020 in the areas not visited by Dr Hammill or that had been previously cleared, and as a consequence, the Secretary had not established whether there was native vegetation present in the charge areas that was cleared during the charge periods.
Having regard to the evidence of Dr Hammill I find beyond reasonable doubt that there was native vegetation on the properties that was cleared, either by mechanical or herbicidal means, in the charge areas at some point in time.
But as explained above, this is not the end of the matter. The Secretary must demonstrate that there was native vegetation that was cleared in the charge areas at the relevant time, and not at some earlier or later point in time. The Secretary has, in my opinion, failed to do so.
There are two aspects to this. First, it is entirely possible that, for example, the clearing of native vegetation from the charge areas occurred earlier as part of the broadacre clearing that the Department was investigating in 2019, or was due to the aerial spraying of vegetation, including crops or weeds, again, outside the charge periods.
Second, it is also a reasonable inference that, given the history of farming activity and cropping across the properties, that what was cleared from the charge areas during the charge periods was not native vegetation but weeds. The Spray Tech documents support this proposition. On 13 March 2020 the reason for spraying was identified as “weeds”. This is consistent with the repeated references to the “crop type” being “fallow”. That is, aerial spraying of already cleared fallow land occurred to treat weed growth. To reiterate, no one from Spray Tech was called to give evidence on behalf of the prosecutor.

Was the Clearing in a “Regulated Rural Area”?

Having regard to the s 60F certificates, there is no doubt whatsoever that any relevant clearing occurred in a “regulated rural area”.

Did the Defendants Carry Out or Were Responsible for the Carrying out the Native Vegetation Clearing?

Keelendi Farms and Edenmore Farms

Pursuant to s 13.29 of the BCA, a landholder such as Keelendi Farms or Edenmore Farms is liable for the clearing unless it is established that the activity was carried out by another person that the landholder did not cause or permit to carry out the clearing.
It has been accepted in a number of decisions concerning the antecedent provision contained in s 44 of the NVA, that the defendant has the burden of proving on the balance of probabilities that it did not cause or permit the clearing to be carried out. Those authorities are applicable to s 13.69 of the BCA.
The applicable legal principles were distilled by the Court in Walker, which were endorsed on appeal (Walker Corporation Pty Limited v Director-General, Department of Environment, Climate Change and Water [2012] NSWCCA 210; (2012) 82 NSWLR 12 at [83]-[89]). In Auen Grain, Robson J adopted these principles further stating that (at [250]-[252]):

250 The principles applicable to s 44 of the Native Vegetation Act where liability is being attributed to a corporate defendant were summarised by Pepper J, albeit in relation to different factual circumstances, in Director-General, Department of Environment and Climate Change v Walker Corporation Pty Limited (No 2) [2010] NSWLEC 73. I adopt Pepper J’s summation at [274] where, having considered a number of earlier authorities, including Director-General, Department of Environment and Climate Change v Jack and Bill Issa Pty Ltd (No 5) [2009] NSWLEC 232; (2009) 172 LGERA 225 (‘Issa’) and Department of Environment and Climate Change v Olmwood Pty Limited [2010] NSWLEC 15; (2010) 173 LGERA 366, her Honour stated:

“…

(a) first, s 44(b) provides a statutory basis of landholder liability arising out of the acts of a third party separate from the common law principles of vicarious liability (Olmwood at [354]). However, an offence against s 12 is one of strict liability thereby attracting the principles of vicarious liability (Issa at [80]);

(b) second, the defence provided for in s 44 must be established on the balance of probabilities by the defendant (Olmwood at [333]);

(c) third, a company can be criminally liable through the actions of its officers or employees, as the embodiment of the company (Issa at [79]);

(d) fourth, in determining who is the ‘directing mind and will’ of the company the real question is, on the proper construction of the statute, whose act is intended to count as the act of the company (Issa at [82]). The acts of both a high-level employee or director and the acts of low-level employees may count if that is required by the terms or the offence and the objects of the statute (Issa at [83]);

(e) fifth, in cases concerning a protective regulatory regime such as that contained in s 12 of the Act, the conduct of officers or employees involved in the actus reus of the offence may be attributed to the company at least where such conduct is in furtherance of the company’s interests or not against them (Issa at [97]);

(f) sixth, to “cause” clearing does not require the exercise of particular control over the third party whose actions resulted in the clearing event to the extent that would otherwise be necessary to establish vicarious liability. Where the clearing by a third party arises as a natural consequence of the landholder’s conduct that landholder can be said to have caused the clearing (Olmwood at [355]); and

(g) seventh, to “permit” means to intentionally allow. That is to say, with knowledge or awareness rather than any intentional failure to act. Further, knowledge that something that is a contravention is likely to be done in the future is to “permit” it to be done. However, mere carelessness or negligence in failing to prevent an act giving rise to a contravention is not to “permit” it to occur (Olmwood at [358]-[359]).”

251 Justice Pepper’s analysis, particularly in relation to her Honour’s finding that knowledge in the context of “causing” something to be done is different to knowledge in the context of “permitting” something to be done, was accepted in Walker Corporation Pty Ltd v Director General, Department of Environment, Climate Change and Water (2012) 82 NSWLR 12; [2012] NSWCCA 210 (‘Walker NSWCCA’) at [87].

252 In Director-General of the Department of Environment, Climate Change and Water v Walker Corporation Pty Ltd (No 2) [2011] NSWLEC 229, Preston J at [99]-[102] considered the application of s 44, in a manner I respectfully adopt, as follows:

“[99] One consequence of the evidentiary presumption in s 44 being established is that the landholder, who is taken to have carried out clearing by s 44, may be a person who “carries out” clearing for the purposes of s 12(2) of the NV Act, provided that the clearing is in contravention of s 12(1) of the NV Act. However, s 44 is not itself concerned with whether the clearing is in contravention of s 12(1) of the NV Act. It is only concerned with the attribution of responsibility of the act of carrying out clearing to a person (the landholder of land on which the clearing was carried out) who did not physically carry out the clearing, without regard to the lawfulness or unlawfulness of the clearing.

[100] Once clearing of native vegetation occurs on land, s 44 operates to deem the landholder of the land on which the native vegetation was cleared to have carried out the clearing unless both of the matters in paragraphs (a) and (b) of s 44 are established. Establishment of both of the matters in paragraphs (a) and (b) is necessary to prevent s 44 operating to deem the landholder to be a person who carried out the clearing.

[101] Paragraph (a) requires the establishment of a positive matter: the clearing was carried out by another person other than the landholder. Paragraph (b) requires the establishment of a negative matter: the landholder did not cause or permit the other person (established in paragraph (a)) to carry out the clearing. The negative formulation in paragraph (b) is important: what is to be established is that the landholder did not cause or permit the other person to carry out the clearing, and not that the landholder did cause or permit the other person to carry out the clearing.

[102] The fact that the evidentiary presumption in s 44 operates unless both matters in paragraphs (a) and (b) are established and the fact that the matter in paragraph (b) is formulated in negative terms, has the consequence that the landholder who wishes to rebut the evidentiary presumption, bears the onus of establishing the matters in paragraphs (a) and (b), on the balance of probabilities: Department of Environment and Climate Change v Olmwood Pty Ltd [2010] NSWLEC 15; (2010) 173 LGERA 366 at 444 [333] and Director-General, Department of Environment and Climate Change v Jack and Bill Issa Pty Ltd (No 5) [2009] NSWLEC 232; (2009) 172 LGERA 225 at 247 [96].”

Because there is no unlawful clearing of native vegetation to be attributed to the landholder in the present proceedings for the reasons given above, the Court need not make a determination on whether the landholder caused or permitted the commission of the offences under s 60N of the LLSA.
However, were I required to determine the issue, having regard to the authorities, and assuming that native vegetation was cleared (as those terms are statutorily defined) within the charge areas during the charge periods, I would find that Keelendi Farms and Edenmore Farms both caused and permitted the commission of the offences.

Charges Against T J O’Brien

As the evidence provided by Spray Tech in response to the statutory notice, the answers provided by O’Brien on behalf of T J O’Brien in his ROI, and the business records provided by T J O’Brien demonstrate there is no doubt whatsoever that T J O’Brien was operating both Keelendi and Edenmore and that O’Brien, on behalf of T J O’Brien engaged Spray Tech to carry out the aerial spraying.
However, for the reasons given above in relation to Keelendi Farms and Edenmore Farms, the charges must fail.

Executive Liability Charge Against O’Brien

The Secretary relied upon the evidence in support of the underlying corporate offence charged against T J O’Brien to establish the first element of the executive liability charge brought against O’Brien pursuant to s 13.6 of the BCA.
As stated above, with respect to the second element, the company searches in evidence plainly establish that O’Brien was a director of T J O’Brien at all relevant times (the second element).
In relation to the third element, the Secretary relied upon the ROI of O’Brien and the Spray Tech records to submit that O’Brien knew or ought reasonably to have known that the executive liability offence would be or was being committed by T J O’Brien.
Finally, fourth, the Secretary argued that there was “no evidence” (T365:36-37) that O’Brien had failed to take all reasonable steps to prevent or stop the commission of the offence by T J O’Brien.
There was considerable debate as to whether the fourth element had been proven by the Secretary to the requisite standard. The Spray Tech records and the O’Brien ROI suggested that O’Brien had engaged an agronomist to advise him and that he had acted on those instructions. In particular, he had specifically instructed Spray Tech not to spray directly over or near the timbered areas.
The prosecutor submitted that, especially when regard was had to the expansive scope of the definition of “reasonable steps” in s 13.6(7) of the BCA, this was insufficient. The obligation was on the person to take “all reasonable steps”, which required more than the direction that O’Brien gave Spray Tech. It was not known, for example, what instructions were given to each of the pilots who undertook the spraying or when they were given. In short, “no evidence has been called about that by the defendant” (T392:17-18).
With respect to the Secretary’s submission, at all times the onus rests on him to establish the elements of the executive liability offence (s 13.6(3) of the BCA), including that O’Brien failed to take all reasonable steps to prevent or stop the commission of the offence.
If it is accepted, as it must be for the reasons explained above, that the prosecution has failed to demonstrate that the vegetation that it alleges was removed in the charge areas during the charge periods was native vegetation, then in my opinion, the Secretary has not discharged its burden. This is because the evidence indicates that O’Brien expressly told Spray Tech not to spray the trees and it cannot be demonstrated that what was sprayed during the charge periods was native vegetation and not, for example, weeds on fallow land that had already been the subject of clearing.
An entirely reasonable inference to be drawn, for example, is that the cause of the brown and dead-looking trees observed and photographed by various witnesses was due to drift when the properties were sprayed by Spray Tech, or because Spray Tech ignored, or at the very least were cavalier, with the instructions provided by O’Brien. Again, no one was called on behalf of Spray Tech by the prosecution.

The Defendants Are Not Guilty of the Offences as Charged

For the reasons given above I find that the Secretary has not proven beyond reasonable doubt all of the elements of the offences. It follows that I find the defendants not guilty of the offences as charged and all summonses ought to be dismissed.

Adjournment Application by the Prosecutor to Consider the Making of a Stated Case

By email communication to chambers on 24 April 2026, the Secretary requested that the Court not to make final orders to enable her to consider the Court’s reasoning pursuant to a potential application under s 5AE of the Criminal Appeal Act 1912 (“CAA”) for a stated case on a point of law. The application was opposed by the defendants principally on the grounds of finality of litigation given that the proceedings have been ongoing for five years (they were commenced on 8 March 2021).
Section 5AE(1) of the CAA provides as follows (emphasis added):

5AE Point of law stated during summary proceedings

(1) At any time before the completion of proceedings before the Supreme Court in its summary jurisdiction, the Land and Environment Court in its summary jurisdiction, the District Court in its summary jurisdiction or a Court of Coal Mines Regulation in its summary jurisdiction, the judge hearing the proceedings may, or if requested by the Crown must, submit any question of law arising at or in reference to the proceedings to the Court of Criminal Appeal for determination.

On any construction of s 5AE(1) of the CAA, if the request for a stated case is made by the Crown, the Court does not have a discretion not to submit the question of law.
In further written and oral submissions, the defendants submitted, first, that s 5AE of the CAA was not engaged. Rather, the provision was only enlivened when a specific question of law had been identified and stated and that this had not been done by the prosecutor. Mere speculation that a question might arise was insufficient to justify withholding final orders. The defendants further argued that the prosecutor has had more than adequate time to consider its position and deferring orders at this stage would be inconsistent with the overriding purpose of the Criminal Procedure Act 1986. Second, the prosecutor had not provided any evidence explaining why the application had been made so late. In support of its contentions the defendants relied upon the decision in Orr v Cobar Management Pty Ltd [2020] NSWCCA 220; (2020) 103 NSWLR 36.
The purpose of s 5AE of the CAA was described in Orr as (at [103]):

103 The purpose of s 5AE of the Criminal Appeal Act was correctly identified by Simpson J (as her Honour then was) and with whom Bell and Buddin JJ agreed in Environment Protection Authority of NSW v Goulburn Wool Scour Pty Ltd (2004) 137 LGERA 289 at 299; [2004] NSWCCA 439 at [36] (Goulburn Wool). There, her Honour observed that:

“…s 5AE is designed to facilitate the giving of advice by this Court to a first instance judge in order to enable the proper resolution of the issues in the proceedings. It arises where proceedings are still on foot and enables the proper and just determination of those proceedings, according to law, and in the correct application of the law. Section 5AE is expressed to apply to proceedings that have not come to an end, and which may, therefore, themselves be resolved by the determination of the questions of law...”

There is some force in the defendants’ submissions, as was recognised by Johnson J in Orr (at [129]-[133]):

129 The present proceedings provide another telling illustration of the difficulties and complications which can arise with use of the stated case procedure in the Criminal Appeal Act 1912 (NSW). In Collins v State Rail Authority of New South Wales (1986) 5 NSWLR 209, Street CJ (Slattery CJ at CL and Yeldham J agreeing) said at 211:

“It should be recognised at the outset that a stated case is well-known as a cumbersome and often unsatisfactory means of bringing a matter up for consideration on appeal. There are occasionally issues of law which can conveniently be dealt with through this appellate procedure. In general, however, it is a procedure which is fraught with difficulties and the present case is no exception from that generality.”

130 The stated case procedure originates from an era before the ready availability of transcripts and court judgments.

131 Until 1998, appeals by way of stated case could be brought from the Local Court to the Supreme Court in summary criminal proceedings under s101 Justices Act 1902 (NSW). The stated case procedure was abolished by the Justices Legislation Amendment (Appeals) Act 1998 (NSW). In the second reading speech concerning that Act, the Attorney General said (Legislative Council, New South Wales, Parliamentary Debates (Hansard), 17 September 1998) that “[p]rocedures for appeals by way of stated case are generally regarded as being cumbersome and unwieldy” with the procedure having “few supporters among the judiciary or practitioners”: at 7597.

132 Part 5 of the Crimes (Appeal and Review) Act 2001 (NSW) now provides for appeals by both the prosecutor and the defendant from the Local Court to the Supreme Court in criminal proceedings. Those provisions have functioned without particular difficulty in the years which have passed since the demise of the stated case procedure.

133 In 2014, the New South Wales Law Reform Commission, in Report 140 entitled Criminal Appeals, recommended the abolition of the stated case procedure under s 5B Criminal Appeal Act (pars 5.14–5.150). A further recommendation was made for the repeal of provisions allowing a trial Judge to submit a question of law arising during or after proceedings to the Court of Criminal Appeal as contained in s 5A, s 5AE, s 5B, s 5BA and s 5BB of the Criminal Appeal Act (pars 10.83–10.89). A recommendation was made for an expanded right of appeal against acquittal.

As the Court of Criminal Appeal emphasised (at [106]):

106 In terms, the purpose of s 5AE is for this Court to determine questions of law which have been submitted to it for determination. That language presupposes that those questions of law have not first been determined by the judge who submits the questions. This is consistent with the section not operating as a surrogate right of appeal. As Hunt J (with whom Grove and Allen JJ agreed) observed in Budget Nursery Pty Limited v Commissioner of Taxation (1989) 42 A Crim R 81 at 82:

“The purpose of the stated case procedure is not to provide an appeal after the judge has made his decision; the purpose is to provide him with assistance in coming to his decision.”

This passage was quoted with approval by Kirby P in Cassell at 166, in turn cited by Tobias JA in the EPA Case at [24].

But, as the joint judgment of Bathurst CJ and Bell P (as his Honour then was) held, the procedure has been used in this Court repeatedly and correctly (at [6]) in cases such as Environment Protection Authority v Land and Environment Court of NSW [2004] NSWCA 50; (2004) 144 A Crim R 198 and Environment Protection Authority v Grafil Pty Ltd; Environment Protection Authority v Mackenzie [2019] NSWCCA 174; (2019) 101 NSWLR 245.
Section 5AE of the CAA has also been used in this Court to state a case in Environment Protection Authority v Maules Creek Coal Pty Ltd (No 3) [2024] NSWLEC 97; (2024) 262 LGERA 350 (at [18]); Auen Grain (at [488]); Environment Protection Authority v Charlotte Pass Snow Resort Pty Ltd [2021] NSWLEC 37 (at [153]); Environment Protection Authority v Terrace Earthmoving Pty Ltd & Page [2012] NSWLEC 216 (at [245]); and Environment Protection Authority v Eastern Creek Operations Pty Ltd (No 2) [2021] NSWLEC 39 (at [47]).
Far from novel, its routine deployment is no doubt explicable on the basis that, unlike a person convicted of an offence (s 5AB of the CAA), the Crown does not enjoy a right of appeal to the Court of Criminal Appeal in respect of proceedings before this Court in its summary jurisdiction.
While the Court appreciates the frustration of the defendants, it does not agree that the prosecutor has had ample time or resources to formulate any question of law for resolution by the Court of Criminal Appeal. This cannot be ascertained in the absence of the reasons of the Court (Land and Environment Court of NSW at [54]-[58]).
As for any delay by the prosecutor in bringing the application, this must be rejected. As soon as the parties were notified that judgment was to be delivered, the prosecutor wrote to the Court making the request. Contrary to the submissions of the defendants, in the present circumstances, the request can only be made upon receipt of the reasons of the Court.
I therefore accede to the Secretary’s request to adjourn the proceedings for it to consider its position.

Orders

Accordingly, the orders of the Court are as follows:
the proceedings are stood over for mention to 9.30am on 26 May 2026 before Pepper J; and
liberty to restore on three days’ notice.

**********

Read the judgment on NSW Caselaw

(2019) 101 NSWLR 245

At a glance

Court

Decision date

Before

Catchwords

Source

Email me a plain-English summary of this case

Catchwords

Judgment

JUDGMENT

The Defendants are Charged with the Unlawful Clearing of Native Vegetation by Poisoning

Joint Trial

The Charges Summarised

The Offences

Statutory Framework Creating the Offences

Burden of Proof in Circumstantial Cases

Elements of the Native Vegetation Clearing Offences

Elements of the Executive Liability Offence

Evidence Relied Upon by the Parties

The Secretary’s Evidence

Section 60F Certificates

Lay Evidence

Mark Smith

Christine Shrimpton

Scott Beaumont

Spray Tech Documents

O’Brien Interview

ASIC search

Susan Rea

Expert Evidence

Leo Watts

Prof Richard Thackway

Dr Kathryn Hammill

Defendants’ Evidence

Was There “Clearing” of “Native Vegetation” in the Charge Areas During the Charge Periods?

Method of Clearing Particularised by the Prosecutor

The Meaning of the Term “Clearing”

Did “Clearing” of Vegetation Occur in the Charge Areas?

Did the Method of Clearing by the Application of Herbicide Occur During the Charge Periods?

Was There Clearing of “Native Vegetation”?

Was the Clearing in a “Regulated Rural Area”?

Did the Defendants Carry Out or Were Responsible for the Carrying out the Native Vegetation Clearing?

Keelendi Farms and Edenmore Farms

Charges Against T J O’Brien

Executive Liability Charge Against O’Brien

The Defendants Are Not Guilty of the Offences as Charged

Adjournment Application by the Prosecutor to Consider the Making of a Stated Case

Orders

Legislation Cited (17)

Cases Cited (42)