Weed Society of Victoria Inc.

PROCEEDINGS THIRD VICTORIAN WEED CONFERENCE

Earth Wind Fire Water and Weeds

SPONSORED BY Blank page PROCEEDINGS THIRD VICTORIAN WEED CONFERENCE Earth Wind Fire Water and Weeds

3–4 October 2007

All Seasons International Hotel, Bendigo

Weed Society of Victoria Inc. PO Box 987, Frankston, Victoria 3199 Tel/fax 03 9576 2949 Email [email protected] Website www.wsvic.org.au This spring the Defeating the Weed Menace (DWM) Programme, the Australian Government’s $44.4 million commitment to national action on Australia’s most threatening weeds, will launch a new website - www.weeds.gov.au - in mid September to help people identify which plants can be a problem in their local area and to provide advice on how to manage these plants.

A print advertising and editorial campaign targeting gardeners and hobby farmers in peri-urban areas will run from September 2007 to March 2008. Copies of the advertisements and other resource materials will be available on the website.

While the majority of people targeted by the campaign will be directed to the website www.weeds.gov.au for further information, enquiries are also likely to be received by state, regional and local governments and nursery retailers for information about weeds issues specific to their local area. PROGRAM

DAY 1 WEDNESDAY 3 OCTOBER 2007

07.15 – 08.50 REGISTRATION

08.50 – 09.00 WELCOME President, Weed Society of Victoria

SESSION 1 KEYNOTE SPEAKERS Chair: Ken Young 09.00 – 10.00 Future weed management in a changing environment – socially, politically and physically Rick Roush, The University of Melbourne, Parkville 10.00 – 10.30 Climate change impacts on weeds in southern Australia Darren Kriticos, Ensis, Canberra 1

10.30 – 11.00 MORNING TEA

SESSION 2 EARTH WIND FIRE WATER AND WEEDS Chair: Daniel Joubert 11.00 – 11.30 EARTH The effects of land management on weed seed longevity Ken Young, The University of Melbourne, Dookie 6 11.30 – 12.00 WIND How far do weed seeds actually travel? Roger Cousens, School of Resource Management, The University of Melbourne, Burnley 9 12.00 – 12.30 FIRE Post-fi re weed management: lessons from the Victorian Alps Charlie Pascoe, Parks Victoria, Bright 10 12.30 – 13.00 WATER Identifying appropriate aims and management approaches for aquatic weeds Nigel Ainsworth, Department of Sustainability and Environment, Melbourne

13.00 – 14.00 LUNCH AND POSTER SESSION

SESSION 3 EARTH WIND FIRE WATER AND WEEDS Chair: Ian Faithfull 14.00 – 14.20 How can willow sawfl y aid our willow management? Fiona Ede, Department of Primary Industries, Frankston 19 14.20 – 14.40 Response of Brazilian milfoil (Myriophyllum aquaticum) to salinity and water-level fl uctuations # and its potential to invade wetlands of the Gippsland Lakes, south-east Victoria Mark Toomey, Institute for Sustainability and Innovation, Victoria University 22 14.40 – 15.00 Aquatic weeds and their impact in irrigation systems experiencing drought within Victoria # Tim Nitschke, Goulburn-Murray Water Aquatic Plant Services 28 15.00 – 15.20 Predicting the potential distribution of orange hawkweed on the Bogong High Plains using a dispersal constrained habit suitability model Nick Williams, School of Resource Management, The University of Melbourne 30

15.20 – 15.50 AFTERNOON TEA

SESSION 4 TACKLING WEEDS ON PRIVATE LAND Chair: Lisa Minchin 15.50 – 16.20 Measurement of changes in stakeholder weed management attitudes and actions over the life of the Tackling Weeds on Private Land initiative # Catriona King, Department of Primary Industries, Seymour 31 16.20 – 16.50 Tackling weeds on private land Helen Anderson, Department of Primary Industries, Hamilton 37 16.50 – 17.20 Weed management practice change within organisations Mark Farrer, Department of Primary Industries, Stawell

19.00 CONFERENCE DINNER

Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 v DAY 2 THURSDAY 4 OCTOBER 2007

08.00 – 08.30 REGISTRATION

SESSION 5 WEED MANAGEMENT Chair: Andrew Clark 08.30 – 09.00 Innovations in spraying techniques, will they enable less herbicide to be used? # Harry Combellack, Spray Smart Enterprises 42 09.00 – 09.30 Prioritising roadsides and reserves for weed control in the Shire of Yarra Ranges # David Blair, Shire of Yarra Ranges 48 09.30 – 10.00 Grassy weeds as fi ery competitors or phantom companions? – Lessons learnt from conservation and land management Colin Hocking, Institute for Sustainability and Innovation, Victoria University 55

SESSION 6 WEED ALERT Chair: Daniel Joubert 10.00 – 10.30 The Victoria Weed Alert Program Stuart Lardner, Department of Primary Industries, Horsham

10.30 – 11.00 MORNING TEA

11.00 – 11.30 Horsetail (Equisetum spp.) and knotweeds (Fallopia spp.) – progress on eradication of two of Victoria’s State Prohibited Weeds Michael Hansford, Department of Primary Industries, Box Hill 58

CONCURRENT SESSIONS

SESSION 7 Chair: Sarah Partington SESSION 8 Chair: Chris Knight 11.35 – 11.55 Early detection for water weeds in 11.35 – 11.55 Response of Chilean needle grass (Nassella Australia neesiana) and phalaris (Phalaris aquatica) Fiona McPherson, NSW Department of seedlings and mature plants to changes Primary Industries, Grafton Agricultural in soil phosphorus, nitrogen and soil pH Research and Advisory Station 61 Charles Grech, DPI, Attwood 77

11.55 – 12.15 Weed proofi ng Australia: a way forward 11.55 – 12.15 Arrowhead in Victoria: current control on invasive garden plants methods and potential for biological control # Nicola Thomson, WWF-Australia Jean-Louis Sagliocco, Department of [for Andreas Glanznig] 64 Primary Industries, Frankston 78

12.15 – 12.35 The role of the National Herbarium of 12.15 – 12.35 Weeding out Australia’s worst willows Victoria (MEL) in the documentation Jackie Steel, Department of Primary of new weeds. Val Stajsic and Alison Industries, Frankston 80 Vaughan, National Herbarium of Victoria 70

12.35 – 12.55 Improving provincial Victoria’s bio- 12.35 – 12.55 Community based quality assurance security by minimising the risk of new scheme weed introductions Brian Feldtmann, Dookie Land Stephen Young, Department of Primary Management Group, Dookie 90 Industries, Ballarat 75

13.00 – 14.00 LUNCH AND POSTER SESSION

vi Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 CONCURRENT SESSIONS

SESSION 9 Chair: Michael Hansford SESSION 10 Chair: Kate Blood 14.00 – 14.20 An investigation of application techniques 14.00 – 14.20 Containing orange hawkweed (Hieracium for the control of various noxious and aurantiacum) infestations in the Ballarat environmental weeds with picloram area gel based herbicides # Simon Martin, Department of Primary Geoffrey Keech, Macspred Australia 92 Industries, Ballarat 109

14.20 – 14.40 Empowering gardeners to make 14.20 – 14.40 Alert and action on Nassella charruana informed choices when buying garden and a range of other National Alert List plants # species in Victoria Nicola Thomson, WWF-Australia 98 Sarah Partington, Department of Primary Industries, Bacchus Marsh 112

14.40 – 15.00 The use of radiation wavelengths to 14.40 – 15.00 The Victorian Weed Spotter network manage weeds # Catherine McInerney, Department of Ken Young, The University of Primary Industries, Geelong 116 Melbourne, Dookie 104

15.00 – 15.20 The impact of weeds on our roadside 15.00 – 15.20 Water hyacinth illegal trade in Victoria: secrets the trace back and forward of a State Karen Jones, Rural City of Wangaratta 107 Prohibited Weed Ryan Melville, Department of Primary Industries, Geelong 119

15.20 – 15.40 Seed bank spatial dynamics and plant invasions in Victoria’s western basalt plains grasslands Aaron Dodd, Department of Primary Industries, Bacchus Marsh 121

15.45 CLOSING REMARKS

POSTERS Practice change by Victorian linear reserve managers Claire Norris, Department of Primary Industries, Geelong 122 Practice change within the Victorian fodder industry Claire Norris, Department of Primary Industries, Geelong 122

LIST OF ATTENDEES 123

REFEREED PAPERS The papers in this proceedings that went through the refereeing process are indicated by # at the end of their title in the program above. Referees: Ken Young and Kelly Raymond

Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 vii Blank page combined effects of climate (temperature, precipitation) and other global change Climate change impacts on weeds in southern components. The focus of this review is on temperature, soil moisture and the atmos- Australia pheric concentration of carbon dioxide

(CO2). This paper discusses the projected Darren J. Kriticos, Ensis, PO Box E4008, Kingston, Australian Capital changes in climate and the mechanisms Territory 2604, Australia. by which global climate change is likely to affect weeds and their management in southern Australia. The likely impacts of these mechanisms are demonstrated in Summary Projected increases in atmos- future socio-economic development paths the context of southern Australia using pheric carbon dioxide concentrations and consequent greenhouse gas emis- the CLIMEX modelling package (Sutherst and concomitant increases in global tem- sions. These scenarios covered a very large et al. 2007). peratures are likely to result in signifi cant range in emissions increases, and when changes in weed distributions. Tropical and used as the basis for forcing global climate Direct impacts of increasing sub-tropical weeds are likely to advance models resulted in a wide range of global atmospheric carbon dioxide further south. Increases in carbon dioxide warming projections. The model based Carbon fertilisation effect could increase growth rates of plants due ‘best estimates’ range from 1.8°C for the In situations where CO2 is a limiting to the carbon fertilisation effect (CFE), and conservative B1 scenario, to 4.0°C for the growth factor, increased CO2 can have a increase their water use effi ciency (WUE). A1FI scenario (IPCC 2007). Subsequent to fertilising effect. This phenomenon has Whether potential gains in crop yield are the release of the IPCC Fourth Assessment been termed the carbon fertilisation effect realised will depend upon the net effects Report, Rahmstorf et al. (2007) have pub- (CFE). Under the CFE, the relative growth of climate change on the competitiveness lished evidence that indicates that whilst rates of plants growing at elevated CO2 of weeds. Climate change is also threat- global emissions of carbon dioxide are fol- can be enhanced compared with similar ening the effectiveness of herbicides. The lowing projections closely, the resulting plants growing under current (pre-climate greatest uncertainties surround the effects increases in global temperature and sea change) CO2 levels. Where other factors of climate change on the effectiveness of level are each tracking on the high end of are limiting (toward the upper and lower weed biological control systems, which projections. They conclude that the pre- regions of the curve in Figure 1), the car- may either be enhanced, or diminished by vious projections by the IPCC may have bon fertilisation effect is reduced. Weeds climate change. underestimated the likely changes in tem- with the C3 photosynthetic pathway are Keywords: Climate change, CLIMEX, perature, underlining concerns about glo- generally expected to enjoy an increased species ranges. bal climate change. competitive advantage through enhanced

Global climate change, caused by in- growth rates compared to C4 crop species Introduction creased emissions of greenhouse gases, is (Ziska 2000, 2001, 2003). However, as Wil-

The Intergovernmental Panel on Climate likely to affect weed management in many liams et al. (2007) observed, sometimes C4 Change recently reaffi rmed and reinforced ways, but the outcome, for instance, as a plants could gain a greater competitive ad- its earlier messages about the human caus- shift in weed distributions, depends on the vantage from elevated CO2, and so specifi c es of global warming, concluding that there is now a ‘…very high confi dence that the global net effect of human activities since 1750 has been one of warming’ and fur- 1.4 ther, that ‘Warming of the climate is un- equivocal, as is now evident from observa- tions of increases in global average air and 1.2 ocean temperatures, widespread melting Current climate of snow and ice, and rising global average Carbon Enhanced CO2 Fertilisation seal level’ (IPCC 2007:3-5). The primary 1 driver of this warming has been identifi ed as anthropogenic emissions of so-called long-lived greenhouse gases (carbon diox- 0.8 ide, methane, nitrous oxide and an assort- ment of halocarbons). The 100-year linear trend in global surface temperatures (1906 0.6 to 2005) has been updated to 0.74°C. It is already clear that these observed changes are already having an infl uence on the Relative growth rate distribution and abundance patterns of a 0.4 range of short generation, mobile species such as Lepidoptera and plant pathogens (Parmesan et al. 1999, Parmesan and Yohe 0.2 2003, Woods et al. 2005, Parmesan 2006). The observed changes in climate and the concomitant species responses are pro- 0 jected to continue to increase throughout 0 0.2 0.4 0.6 0.8 1 1.2 the 21st Century. The IPCC has developed a set of climate change scenarios (IPCC Environmental factor 2000). These scenarios take the form of a set of storylines that describe plausible Figure 1. Schematic representation of the carbon fertilisation effect.

Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 1 situations should be verifi ed. Because of 1.2 their association with nitrogen-fi xing bac- teria, legumes may be expected to benefi t from the CFE compared with non-legumes Current climate in non-agricultural settings where nitro- 1 Enhanced CO2 gen is often limiting.

Water use effi ciency

Increased CO2 has the effect of increas- 0.8 ing plant water use effi ciency (Morison and Gifford 1984a,b). This is because for a given level of stomatal CO , a plant can 2 0.6 Increased water operate with a more restricted stomatal use efficiency opening, resulting in less loss of leaf mois- ture. The dashed line in Figure 2 indicates the effect of enhanced CO2 on the species 0.4 soil moisture growth response curve. Un- growthrate Relative der elevated CO2, plants may be able to grow better at moderate (sub-optimal) soil moisture levels. The minimum soil mois- 0.2 ture level at which plant growth can occur (permanent wilting point) is unlikely to be altered by changes in CO2. Increasing CO2 may therefore allow species to persist in 0 drier regions than they can under reduced 00.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 CO as they are better able to make use of 2 Soil moisture index lower levels of soil moisture.

Emergent effects of climate change Figure 2. Schematic representation of the effect of elevated CO on plant on weeds 2 Length of growing season growth rates and water use effi ciency. In cool to cold climates (e.g. temperate, Mediterranean and alpine), increasing temperatures could increase the length of the growing season for some weeds. This may become apparent through an earlier onset of spring growth as is projected to 1 occur for bridal creeper at Hobart (Figure 3). The pattern of changes in climate suit- 0.9 ability however is complicated by the in- Current climate terplay of soil moisture and temperature. 0.8 Had 3 2070s As is apparent in Figure 3 during the early part of the year, future conditions could 0.7 become drier during the late summer and 0.6 early autumn period. This period of sub- optimal growth could partly offset gains 0.5 in winter growth. At the trailing edge of the range shifts, where conditions get too 0.4 hot or dry for a species in the future, the Weekly growth index opposite change could be observed, where 0.3 the growing season contracts (Figure 4). 0.2 Shifting species ranges A common factor limiting the poleward 0.1 spread of plants appears to be a minimum threshold of growing degree days neces- 0 sary to complete a generation in the case 0 50 100 150 200 250 300 350 400 of annual and ephemeral species, or, in Day of year the case of perennial species to complete the reproductive cycle and produce vi- Figure 3. CLIMEX weekly growth index for bridal creeper (Asparagus able mature seeds. It is common to fi nd asparagoides) at Hobart (43°00’S, 147°15’E) modelled using the reference species growing beyond their ecoclimatic climate (1961–1990 average) and the 30 year average around 2075 as limits for long-term persistence (e.g., Aca- simulated using the Hadley mark 2 model with high sensitivity and high cia nilotica in Sydney botanic gardens, and emissions scenario. Under the future scenario, growing conditions for Grevillia robusta in Canberra). It is diffi cult to imagine however, that a plant growing bridal creeper are expected to improve slightly at this site. The climate outside of this range boundary could pose change scenarios were supplied by CSIRO Marine and Atmospheric an immediate weed threat problem. The Research, and adapted using the techniques described in Stephens et al. presence of lethal low temperatures can (2007).

2 Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 also limit a plant species in terms of lati- 0.45 tude and/or altitude. Evidence for these mechanisms comes from both mechanis- 0.4 tic studies (Woodward 1988), and infer- ential modelling approaches (Beerling et 0.35 Current climate al. 1995, Sykes et al. 1996, Holt and Boose 2070s Hadley High 2000, Kriticos and Randall 2001, Kriticos 0.3 et al. 2003a,b, Walden et al. 2004, Kriticos et al. 2004, Kriticos et al. 2005a, Kriticos et 0.25 al. 2005b, Kriticos et al. 2006, Kriticos and Potter 2006) 0.2 As temperatures increase, we should expect that many species would be able to 0.15 overcome their current poleward limits.

For southern Australia, this means that Weekly growth index 0.1 weeds whose current distribution lim- its are further north such as rubber vine 0.05 (Cryptostegia grandifl ora), and prickly aca- cia (Acacia nilotica) will probably invade 0 further south (Figure 5). 0 50 100 150 200 250 300 350 400 For weeds with a more temperate or -0.05 Mediterranean climatic preference such as Scotch broom (Cytisus scoparius), butterfl y Day of year bush (Buddleja davidii), and bridal creeper (Asparagus asparagoides) (Figure 6) climate Figure 4. Weekly growth index for Scotch broom (Cytisus scoparius) at change is likely to reduce their potential Bendigo (36°46’S, 144°18’E) modelled using CLIMEX. Under the reference range. The mechanisms by which this oc- climate (1961–1990) and the 30 year average around 2075 as simulated curs are likely to include direct heat stress, using the Hadley mark 2 model with high sensitivity and high emissions as well as competition from plants that are better adapted to the warmer conditions scenario. Under the warming scenario, growing conditions for Scotch expected in the future. As most weeds broom are expected to deteriorate at this site. The climate change scenarios have effi cient dispersal mechanisms, this were supplied by CSIRO Marine and Atmospheric Research and adapted displacement of cool-adapted weeds may using the techniques described in Stephens et al. (2007). well come as a result of invasion by other weeds.

Range changes in the context of an active invasion It is worth bearing in mind that few weeds AB have reached their Ecoclimatic limits in Australia. Therefore, whilst they are still invading, the effect of climate change will be to shift the geographic limits of that invasion (Figure 7). Therefore in order to understand the effects of climate change on the distributions of weeds it is neces- sary to consider ecoclimatic models such as CLIMEX (Sutherst et al. 2007). Such models provide a means of understanding the synoptic view of the invasion under current climate, as well as exploring the likely effect of climate change scenarios on that invasion.

Decreased herbicide effi cacy Bailey (2004) has shown that increasing temperatures and CO2 can result in re- duced herbicide effi cacy through more rapid chemical breakdown, reducing the period during which effective concentra- Figure 5. The current and potential range of prickly acacia (Acacia nilotica) tions of the herbicide are maintained. Zis- in Australia under (A) current climate, and (B) the climatic conditions ka et al. (1999) have shown that elevated projected for the 2070s using the Hadley mark 2 GCM. Water use effi ciency CO2 could lead to increased tolerance of and growth rates due to the carbon fertilisation effect are assumed to glyphosate by weeds. The mechanism by remain unchanged. The large unsuitable area in the centre of the 2070s which this occurs could have to do with decreased stomatal conductance. Archam- scenario is due to decreased rainfall and increased temperatures reducing bault et al. (2001) found highly variable soil moisture availability. The climate change scenarios were supplied response of herbicide effi cacy to elevated by CSIRO Marine and Atmospheric Research, and adapted using the

CO2 and temperature. This variability techniques described in Stephens et al. (2007).

Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 3 spanned plants with C3 and C4 photosyn- AB thetic pathways and three herbicide chem- icals tested.

Changing carbon to nitrogen ratios

Increasing CO2 levels are likely to increase C:N ratios in many plant tissues. This could result in either more or less herbiv- ore damage to plants. Nitrogen is frequent- ly a limiting resource for herbivores. If the herbivores can effi ciently and effectively discard excess carbon, as they extract the nitrogen that they need, then they may be able to maintain similar population dy- namics to current CO2 conditions. Under this scenario they may increase the total amount of damage they infl ict on a plant population because they increase their per capita consumption of tissues in order to maintain a given level of nitrogen intake. If, on the other hand, they are not able to extract the necessary levels of nitrogen, Figure 6. The current and potential range of bridal creeper (Asparagus their population dynamics may be nega- tively affected, leading to less plant dam- asparagoides) in Australia under (A) current climate, and (B) the climatic age. The implications of these mechanisms conditions projected for the 2070s using the Hadley mark 2 GCM. Water are that biological control efforts against use effi ciency and growth rates due to the carbon fertilisation effect are weeds may be either enhanced or retarded assumed to remain unchanged. The climate change scenarios were supplied under future climate scenarios. by CSIRO Marine and Atmospheric Research and adapted using the techniques described in Stephens et al. (2007). Conclusions Climate change is likely to pose a series of threats to weed management in south- ern Australia. Northern weeds will invade further south, and many existing weeds may grow more rapidly, and become more fecund. Many weeds will be able to invade further into arid regions, and grow better under reduced rainfall. Whilst the weeds themselves are growing more vigorously, B A C some herbicides at least will become less effective. Even the positive implication that some temperate weeds will be forced further south is a mixed blessing. They are likely to be forced southward by oth- er weeds, and the temperate crops upon which the valuable southern Australian agriculture depends are likely to suffer similar consequences. Similarly, native fl ora and fauna are likely to suffer similar southward range shift pressures. Given the significance of the likely changes in weed management, and the areas of uncertainty, it seems clear that we need to emphasise our ability to adapt rapidly our weed management systems in Figure 7. Diagrammatic representation of the relationship between (A) response to emerging threats and challeng- current distribution, (B) potential distribution under current climate, and es. Modelling and experimental work to (C) potential distribution under climate change. date has indicated the likely general ways in which these systems may change, but this work is not suffi ciently sophisticated or reliable to be able to predict when these of the telltale harbingers of the identifi ed as the trailing edge of their range becomes changes might come about, or even what (and unidentifi ed) climate change threats, unsuitable, it is quite likely that weeds specifi c outcomes might be expected for and to be able to rapidly assemble and ap- will invade following a disturbance such given combinations of weeds and valued ply the resources needed to develop adap- as fi re, fl ood or wind throw. species (crops and conservation targets). tations to the emerging challenges. Thus, there will be a need for adaptive or The projected impacts of climate change Acknowledgement nimble governance of land management on weeds will be paralleled by similar Thanks are due to Dr Agathe Leriche for enterprises and activities. This approach changes for native and production ecosys- help with some of the graphics. to land management will need to be aware tems. When native plants become stressed

4 Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 References Kriticos, D.J., Sutherst, R.W., Brown, J.R., Hennecke, Q. Paynter and C. Wilson, Archambault, D.J., Li, X., Robinson, D., Adkins, S.A. and Maywald, G.F. (2003b). pp. 11-21. (CSIRO Entomology, Can- O’Donovan, J.T. and Klein, K.K. (2001). Climate change and the potential distri- berra).

The effects of elevated CO2 and temper- bution of an invasive alien plant: Acacia Williams, A.L., Wills, K.E., Janes, J.K., ature on herbicide effi cacy and weed/ nilotica ssp. indica in Australia. Journal of Vander Schoor, J.K., Newton, P.C.D. crop competition. Prairie Adaptation Applied Ecology 40, 111-24. and Hovenden, M.J. (2007). Warming

Research Collaborative. Kriticos, D.J., Yonow, T. and McFadyen, and free-air CO2 enrichment alter de- Bailey, S.W. (2004). Climate change and R.E. (2005b). A revised estimate of the mographics in four co-occurring grass- decreasing herbicide persistence. Pest potential distribution of Chromolaena land species. New Phytologist. Management Science 60, 158-62. odorata (Siam weed) in relation to cli- Woods, A., Coates, K.D. and Hamann, A. Beerling, D.J., Huntley, B. and Bailey, J.P. mate. Weed Research 45, 246-54. (2005). Is an unprecedented dothistro- (1995) Climate and the distribution of Morison, J.I.L. and Gifford, R.M. (1984a). ma needle blight epidemic related to Fallopia japonica: use of an introduced Plant growth and water use with lim- climate change? Bioscience 55, 761-9.

species to test the predictive capacity of ited water supply in high CO2 concen- Woodward, F.I. (1988). Temperature and response surfaces. Journal of Vegetation trations. I. Leaf area, water use and the distribution of plant species. Plants Science 6, 269-82. transpiration. Australian Journal of Plant and Temperature, Symposium of the Holt, J.S. and Boose, A.B. (2000). Poten- Physiology 11, 361-74. Society for Experimental Biology, eds tial for spread of Abutilon theophrasti in Morison, J.I.L. and Gifford, R.M. (1984b). S.P. Long and F.I. Woodward. (Com- California. Weed Science 48, 43-52. Plant growth and water use with lim- pany of Biologists, Cambridge).

IPCC (2000). Special report on emissions ited water supply in high CO2 concen- Ziska, L.H. (2000.) The impact of elevated

scenarios: a special report of Working trations. II. Plant dry weight, partition- CO 2 on yield loss from a C3 and C4 weed Group III of the Intergovernmental ing and water use effi ciency. Australian in fi eld-grown soybean. Global Change Panel on Climate Change. Cambridge Journal of Plant Physiology 11, 375-84. Biology 6, 899-905. University Press, Cambridge, United Parmesan, C. (2006) Ecological and evo- Ziska, L.H. (2001) Changes in competi- Kingdom. lutionary responses to recent climate tive ability between a C-4 crop and a IPCC (2007). Climate change 2007: the change. Annual Review of Ecology Evolu- C-3 weed with elevated carbon dioxide. physical science basis. Contribution of tion and Systematics, 37, 637-669. Weed Science 49, 622-7. Working Group I to the Fourth Assess- Parmesan, C., Ryrholm, N., Stefanescu, C., Ziska, L.H. (2003) Evaluation of yield loss ment Report of the Intergovernmental Hill, J.K., Thomas, C.D., Descimon, H., in fi eld sorghum from a C-3 and C-4

Panel on Climate Change. Cambridge, Huntley, B., Kaila, L., Kullberg, J., Tam- weed with increasing CO2. Weed Science United Kingdom and New York, NY, maru, T., Tennent, W.J., Thomas, J.A. 51, 914-18. USA, Cambridge University Press. and Warren, M. (1999). Poleward shifts Ziska, L.H., Teasdale, J.R. and Bunce, J.A. Kriticos, D.J., Alexander, N.S. and Kolom- in geographical ranges of butterfl y spe- (1999) Future atmospheric carbon diox- eitz, S.M. (2006). Predicting the poten- cies associated with regional warming. ide may increase tolerance to glypho- tial geographic distribution of weeds Nature 399, 579-83. sate. Weed Science 47, 608-15. in 2080. Proceedings of the Fifteenth Parmesan, C. and Yohe, G. (2003). A glo- Australian Weeds Conference, eds C. bally coherent fi ngerprint of climate Preston, J.H. Watts and N.D. Crossman, change impacts across natural systems. pp. 27-34. (Weed Management Society Nature 421, 37-42. of South Australia, Adelaide). Rahmstorf, S., Cazenave, A., Church, J.A., Kriticos, D.J., Lamoureaux, S., Bourdôt, Hansen, J.E., Keeling, R.F., Parker, D.E. G.W. and Pettit, W. (2004). Nassella and Somerville, R.C.J. (2007). Recent cli- tussock: current and potential distri- mate observations compared to projec- butions in New Zealand. New Zealand tions. Science 316, 709. Plant Protection 57, 81-8. Stephens, A.E.A., Kriticos, D.J. and Kriticos, D.J., Potter, K.J. and Alexander, Leriche, A. (2007). The current and fu- N. (2005a). The potential distribution ture potential geographic distribution of Buddleja davidii in Australia and of the Oriental fruit fl y, Bactrocera dor- New Zealand. Rotorua, New Zealand, salis, (Diptera: Tephritidae). Bulletin of Ensis. Entomological Research. Kriticos, D.J. and Potter, K.P. (2006). Some- Sutherst, R.W., Maywald, G.F. and Kriti- thing’s bugging butterfl y bush in Tas- cos, D.J. (2007). CLIMEX Version 3: Us- mania. Proceedings of the Fifteenth er’s Guide. (Hearne Scientifi c Software Australian Weeds Conference, eds C. Pty Ltd, www.Hearne.com.au). Preston, J.H. Watts and N.D. Crossman, Sykes, M.T., Prentice, C. and Cramer, W. pp. 731-4. (Weed Management Society (1996). A bioclimatic model for the po- of South Australia, Adelaide). tential distributions of north European Kriticos, D.J. and Randall, R.P. (2001). A tree species under present and future comparison of systems to analyse po- climates. Journal of Biogeography 23, 203- tential weed distributions. In ‘Weed 33. Risk Assessment’, eds R.H. Groves, Walden, D., van Dam, R., Finlayson, M., F.D. Panetta, and J.G. Virtue, pp. 61-79. Storrs, M., Lowry, J. and Kriticos, D. (CSIRO Publishing, Melbourne, Aus- (2004). A risk assessment of the tropical tralia). wetland weed Mimosa pigra in north- Kriticos, D.J., Sutherst, R.W., Brown, J.R., ern Australia. Research and Manage- Adkins, S.A. and Maywald, G.F. (2003a). ment of Mimosa pigra: papers presented Climate change and biotic invasions: a at the 3rd international symposium on case history of a tropical woody vine. the management of Mimosa pigra, eds Biological Invasions 5, 145-65. M. Julien, G. Flanagan, T. Heard, B.

Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 5 (Cousens and Mortimer 1995), 88% for gi- ant ragweed (Ambrosia trifada) (Westerman The effects of land management on weed seed et al. 2006) and modelling predicting up to 65% predation in a range of cropping rota- longevity tions in the USA (Westerman et al. 2006). Under Australian conditions these losses K.R. Young, The University of Melbourne, Dookie College, Victoria 3647, appear to be less , or at least more highly Australia. variable (0–100%) with more predation on the paddock edges (Spafford Jacob et al. 2006). Predators include ants, voles, earth- worms, rodents, birds and carabid beetles (Cousens and Mortimer 1995, Spafford Seed longevity Before tackling what affects has man- Jacob et al. 2006, Westerman et al. 2006). agement practices have on seed longevity, Once seed is buried (e.g. by tillage, fall- ‘One years seeding, seven years weeding’ it is important to understand what are the ing into soil cracks) predation by animals This old adage bears great witness to causes of loss of seeds from the seedbank. is dramatically reduced (Hartzler et al. weed seeds ability to persist, as most of 2006, Spafford Jacob et al. 2006, Westerman us who farm, garden or are involved in Seed losses et al. 2006) but predation can occur from weed management know! So what causes The dynamics of the seedbank is due to soil microbes. The effect of soil microbes seed longevity and how do we manage the infl ux and effl ux of seeds. The infl ux has been poorly studied with most infor- this in the plants that we wish to target. within a paddock has also been called mation coming from treating seeds with In arable agricultural systems, this will seed rain. While seed shedding from the fungicides and comparing to untreated be all of the seedbank as not only do we mother plant is the main infl ux, seed im- seeds with losses estimated to up to 20% want low weed seed numbers but also we portation via wind, crop seeds, machinery (Cousens and Mortimer 1995). do not want any carry over of crop seeds and animals can also be major infl uxes or The loss of seed from the seedbank due from one year to the next as this can cause more particularly the major reasons for to germination depends on the seeds dor- disease and pest issues and quality issues new infestations. The effl uxes are seed mancy status and its position within the for the following crop (adventitious pres- losses due to predation, germination soil profi le. After ripening of seeds (either ence), a growing issues with the advent and loss of viability (Cousens and Mor- artifi cially or naturally) can cause a large of greater product diversity and manage- timer 1995). Unfortunately there is little germination pulse after soil water reaches ment of GM crops. In pastoral and natural understanding of the relative importance adequate levels. Seed germination of wild systems, not all the seed bank is targeted, of these three effl uxes as most seedbank oats increased linearly from 0% to 90% as we need to maintain a viable seed bank studies have not recorded differences be- over 18 month of after ripening at 25°C of our pasture and native species to main- tween the seedbank losses. Many studies (Adkins and Ross 1981). Different species tain a reservoir of genetic material for re- have just looked at before and after seed- have different germination rates and puls- placement recruitment. bank numbers (Bekker et al. 2000, Felix es as was demonstrated by Young (2001) in There are inherent differences between and Owen 2001, Thorne et al. 2007). Others comparing wild oats, annual ryegrass (Lo- species in their longevity. Seed size has have recorded, initial seedbank numbers, lium rigidum) and wild radish (Raphanus been put forward as an indicator of seed seedling emergence and fi nal seedbank raphanistrum). Both wild oats and annual longevity, with small seeds having longer numbers but can not distinguish between ryegrass had germination losses from the persistence than larger seed (Thompson fatal germination, decay and predation for seedbank primarily in the fi rst year, where et al. 1993). The theory behind this is that the differences between pre management wild radish predominately had its germi- smaller seeds have a greater ability to be and fi nal seedbank numbers (Young 2001, nation losses in its second year especially incorporated into the soil, where larger Davis et al. 2005, Brainard et al. 2007). Also following cultivation (Table 1). seeds are more likely to be targeted by the studies tend to investigate only one or Seed decay or loss of viability has been predator species, and moved to the soil two of these effl uxes at a time. For exam- the least studied phenomena of the three surface through soil movement (Spokas et ple, many predation studies are conducted effl uxes with estimates of up to 20% loss al. 2007). This theory has been challenged only on seeds on the soil surface. Most per- due to decay (Cousens and Mortimer by Leishman and Westoby (1998) ,who sistence studies involve burial of seeds at 1995) but is more likely less than 10% in demonstrate that with Australian native different depths, utilising mesh bags to most situations (Young 2001). species there is no correlation between contain seeds. Longevity is recorded by Weed management targeting the seed- seed size and shape and seed persistence. recovery of and germination of seeds from bank should be aimed at either increasing They summarise the difference between within the mesh bag. predation or increasing germination as the papers are due to the natural history However, large seed losses from pre- these two effl uxes have the greatest po- characteristics between British species dation can occur, with losses of 85% be- tential to reduce the seedbank. and Australian species. The later having ing recorded for wild oats (Avena fatua) many species that are more adapted to fi re ecology hence having heavier seeds than the herbaceous fruits of British species. Table 1. The effect of cultivation on seed longevity and seedling emergence However, in our pastoral and arable farm- ing ecosystems, most of our weed species from the seedbank (Young 2001). have been derived from overseas and par- Emergence (%) ticularly from Europe, so the Thompson Year 1 Year 2 et al. (1993) theory should hold true for Without With Dormancy (%) the majority of our arable weeds. But for Species cultivation cultivation after 21 months invaders of Australian native ecosystems, Wild oats 57.6 5.7 8.0 0.5 0.3 seed size and shape may not be a good indicator of seed persistence. Annual ryegrass 42.8 24.0 6.1 4.4 0.5 Wild radish 11.3 28.3 49.1 32.7 11.3

6 Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 Management effects on seed 100% longevity Seed burial Seed burial has the effect of increasing 80% seed longevity the deeper the burial. In many studies, seed on the soil surface may have a slightly longer longevity initially 60% than seed in shallow depths (1–5 cm) due to germination losses at shallow burial depth. Over time seeds on the surface and 40% at shallow depths (<2 cm) have similar low

levels of persistence, but once the depth bank Proportion seed of increases the longevity of seeds increases 20% as depth increase (Figure 1) (Young 2001). So for weed management, it is an option to deep bury seeds to avoid germination 0% of new propagules. However this man- agement can only occur once a decade or surface 1 cm 2 cm 4 cm 8 cm longer (based on decay rates) as if these Emerged Viable seed buried seeds are brought back to the sur- face they will again germinate. Non-viable seed Decayed or predated In arable agricultural systems, the effect of cultivation on weed seed movement has Figure 1. The effect on seed burial of wild radish on seedling emergence, been studied (Mohler et al. 2006) and is viable seed, non viable seed and decayed or predated seed percentages after incorporated into models used to predict 18 months seed burial (Young 2001). weed seedling emergence (Sester et al. 2007, Spokas et al. 2007). Cultivation equip- ment such as mouldboard plough invert the soil, so initially a large difference in Australian arable dryland ecosystems, the In arable situations, heat has been the seed bank can be obtained. Over time legume and oilseed crops in the rotation utilised to destroy many weed seeds. however, continual inversion will result are seen to be the most vulnerable to weed Soil solarisation capturing heat through in a fairly uniform seedbank throughout outbreaks, through their lower competi- black plastic covers, or heat accumulating the plough depth (Cousens and Mortimer tive ability (slow growth rate, low sowing through organic matter decomposition 1995, Sester et al. 2007, Spokas et al. 2007). densities) and limited herbicide options. in composts has been used to kill weed Seeds can also work themselves from low- In the oilseed canola this has seen the de- seeds. Use of steam to control weeds has er soil profi les into the upper layers over velopment of herbicide tolerant varieties. been developed but has had very limited time, negating some of the effects of burial An increase in the intensity of cropping effect on killing seeds unless implemented (Mohler et al. 2006). Tine implements tend has seen an increase in weed seedbanks in a pressurised device. As soil is a good the mix the seeds in a horizontal action (Cousens and Mortimer 1995). This in- insulator, seeds even at relative shallow with little displacement of seeds at differ- crease in weed seedbanks has been record- depths are well protected from heat as a ent depths, as do rototillers. ed in pastures with increasing disturbance control management technique. The change in tillage systems from (animal hooves, slashing or cutting hay Fire has been used to manage weeds multiple tillage operations to a single pass etc.) (Renne and Tracy 2007, Wellstien et al. after germination in many organic crops, tillage operation (zero tillage) has also 2007). These increases in weed seedbanks and the effect on seeds has been investi- brought with it a change in weed spectra. from more intensifi ed agriculture are of gated in managing weeds being returned Reduced tillage systems tend to increase magnitudes of 60–100%. to the fi eld after crop harvest. Walsh and the level of perennial species relatively Crop rotations can have a large impact Newman (2007) found that temperatures to annual species (Carter and Ivany 2006, on seedbank diversity and seed longev- in excess of 400°C and 500°C for 10 sec- Conn 2006). Under reduced tillage there ity within a species. In a study of Powells onds were required to kill seeds of annual is less burial of weed seeds causing an in- amaranth, (Brainard et al. 2007), found ryegrass (Lolium rigidum) and wild rad- crease in the weed seedbank in the top 10 that seed dormancy was selected for un- ish (Raphanus raphanistrum) respectively. cm of the soil (Conn 2006), which in turn der continuous lucerne (dairy pasture) Utilisation of conventional and narrow increases the turnover of the seedbank. compared to shorter dormancy under cul- windrows achieved greater than 96% con- Though, Carter and Ivany (2006) got great- tivated vegetable cropping. trol of both of these species (Walsh and er seedbank build up in the subsoil (>20 Newman 2007). cm) than the topsoil (<10 cm), which they Heat explained was due to the soil type being Heat can be applied to the soil via various Soil quality a fi ne sandy loam. A general observation means such as fi re, solar radiation, steam, Weed suppressive soils based on inputs from the literature suggests that annual decomposition, and microwaves (Brodie of carbon through green manure crops, grasses respond more to cultivation (less and Young 2007). Heat can be used to ei- animal manure application, have led to persistence) than either perennial grasses ther destroy the seeds or promote germi- the formation of colonies of deleterious or broadleaves (Cousens and Mortimer nation. In natural ecosystems there may rhizosphere bacteria which in tune has 1995). be a selectivity between native species led to a decreased growth of weed seed- and introduced species in their tolerance lings (Kremer and Li 2003). The increased Rotations to heat, with many native species being bioactivity within the soil under these Weed seedbanks are altered by crop rota- more ecologically adapted to heat (fi re) ‘biological’ systems may also decrease the tions through different crops having differ- with fi re stimulating germination, where longevity of weed seed banks, though lit- ent sowing times, competitive abilities and introduced species may be more vulner- tle work has yet been done on this. But types of herbicides used. In south-eastern able. biological systems show promise as Davis

Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 7 et al. (2005) showed a negative correlation Brodie, G. and Young, K. (2007). The use Thompson, K., Band, S.R. and Hodgson, between weed seed persistence and hy- of radiation wavelengths to manage J.G. (1993). Seed size and seed shape drothermal time, indicating that soils with weeds. Proceedings of the 3rd Victorian predict persistence in soil. Functional higher moisture retaining properties will Weeds Conference, pp. 104-6. (Weed So- Ecology 7, 236-41. decrease the persistence of a seed. This ciety of Victoria, Frankston). Thorne, M.E., Young, F.L. and Yenish, J.P. is most probably due to greater activity Carter, M.R. and Ivany, J.A. (2006). Weed (2007). Cropping systems alter weed by fungal and bacterial decay at high soil seed bank composition under three seed banks in Pacifi c Northwest semi- water levels. Though high C:N ratios can long-term tillage regimes on a fi ne loam arid region. Crop Protection 26, 1121-34. impede soil microbe decay of weed seeds in Atlantic Canada. Soil and Tillage Re- Walsh, M. and Newman, P. (2007). Burn- (Shem-Tov et al. 2005). Higher organic search 90, 29-38. ing narrow windrows for weed seed soils, with greater trash conservation also Conn, J.S. (2006). Weed seed bank affected destruction. Field Crops Research. increases the protection of seed predators, by tillage intensity for barley in Alaska. Wellstien, C., Otte, A. and Waldhardt, R. increasing seed predation of the seedbank Soil and Tillage Research 90, 156-61. (2007). Seed bank diversity in mesic (Spafford Jacob et al. 2006; Westerman et Cousens, R. and Mortimer, M. (1995). ‘Dy- grasslands in relation to vegetation al. 2006). namics of weed populations’. (Cam- type, management and site condition. bridge University Press, Cambridge). Journal of Vegetation Science 18, 153-62. Conclusion Davis, A.S., Cardina, J., Forcella, F., John- Westerman, P.R., Liebman, M., Heggen- Incorporating a fully integrated weed son, G.A., Kegode, G., Lindquist, J.L., staller, A.H. and Forcella, F. (2006). In- management techniques into land man- Luschei, E.C., Renner, K.A., Sprague, tegrating measurements of seed avail- agement by targeting not only the emerged C.L. and Williams M.M. (2005). Envi- ability and removal to estimate weed weeds but also the seedbank will lead to ronmental factors affecting seed per- seed losses due to predation. Weed Sci- better long term management. The two ar- sistence of annual weeds across the US ence 54, 566-74. eas to target in reducing weed seedbanks corn belt. Weed Science 53, 860-8. Young, K.R. (2001). Germination and are increasing predation and germination Felix, J. and Owen, M.D.K. (2001). Weed emergence of wild radish (Raphanus losses. The former fi ts well into the present seedbank dynamics in post conserva- raphanistrum L.). The University of Mel- paradigm of biological farming and with tion reserve program land. Weed Science bourne, 205. the advent of new molecular tools, it will 49 780-7. be easier to determine which methodolo- Hartzler, B., Liebman, M. and Westerman, gies/technologies are most appropriate P. (2006). Weed seed predation in agri- for different ecosystems. In pastoral and cultural fi elds, Iowa State University. natural ecosystems where the seedbank 2007. not only consists of weed seeds but also Kremer, R.J. and Li, J. (2003). Developing desirable seeds, it is still to be determined weed-suppressive soils through im- whether there is suffi cient effi cacy win- proved soil quality management. Soil dow for these techniques. and Tillage Research 72, 193-202. Targeting the germination losses from Leishman, M. R. and Westoby, M. (1998). the seedbank, then requires a follow up Seed size and shape are not related to to manage the emerged population. This persistence in soil in Australia in the has to then be incorporated within other same way as in Britain. Functional Ecol- management tasks such as sowing crops ogy 12, 480-5. or fertilising pastures, which may cause Mohler, C.L., Frisch, J.C. and McCulloch, time delays and hence yield losses. If her- C.E. (2006). Vertical movement of weed bicides are relied on too heavily, there will seed surrogates by tillage implements be weed resistant issue to be managed. and natural processes. Soil and Tillage The future to managing weed seed- Research 86, 110-22. banks is in the use of computer simula- Renne, I.J. and Tracy, B.F. (2007). Distur- tion models in order for land managers bance persistence in managed grass- to predict possible outcomes under dif- lands: shifts in aboveground commu- ferent scenarios that they are faced with nity structure and the weed seed bank. annually. These management aids need Pant Ecology 190, 71-80. to incorporate all different management Sester, M., Durr, C., Darmency, H. and Col- options as well as economic outcomes. A bach, N. (2007). Modelling the effects good example of such a model is Ryegrass of cropping systems on the seed bank Integrated Management (RIM). dynamics and the emergence of weed beet. Ecological Modelling 204, 47-58. References Shem-Tov, S., Klose, S., Ajwa, A. and Fen- Adkins, S.W. and Ross, J.D. (1981). Studies nimore, S.A. (2005). Effect of carbon:ni- in wild oat seed dormancy. Plant Physi- trogen ration and organic amendments ology 67, 358-62. on seedbank longevity. Weed Science So- Bekker, R.M., Verweij, G.L., Bakker J.P. and ciety of America Abstracts 45, 97. Fresco L.F. (2000). Soil seed bank dy- Spafford Jacob, H., Minkey, D.M., Gallagh- namics in hayfi eld succession. Journal er, R.S. and Porger, C. (2006). Variation of Ecology 88, 594-607. in postdispersal weed seed predation in Brainard, D.C., DiTommaso, A. and Moh- a crop fi eld. Weed Science 54 148-55. ler, C.L. (2007). Intraspecifi c variation in Spokas, K., Forcella, F., Archer D. and seed characteristics of Powell amaranth Reicosky, D. (2007). SeedChaser: ver- (Amaranthus powellii) from habitats with tical tillage distribution model. Com- contrasting crop rotations histories. puters and Electronics in Agriculture 57, Weed Science 55 218-26. 62-73.

8 Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 How far do weed seeds actually travel?

Roger Cousens, School of Resource Management, The University of Melbourne, Burnley Campus, 500 Yarra Boulevard, Richmond, Victoria 3121, Australia.

Abstract Ecologists invariably focus on number of seeds deposited per unit area long-distance dispersal, since this has a at different distances or the frequency dis- large infl uence over the rate of spread of tribution of dispersal distance (these two invasive species through a landscape. We relationships are very different shapes), marvel at the distances that (some) seeds the vast majority of seeds are deposited travel and the astonishing array of mor- within a few metres or centimetres. Thus, phological adaptations by which this is we need to measure short-distance disper- achieved. Theoretical ecologists can now sal, not just long-distance dispersal. predict the distances dispersed by wind Examples are given of studies that have using highly sophisticated models, al- measured short distance dispersal by air, though our ability to predict distances dis- water, fi re and earth. It is shown that the persed by other vectors, such as animals, general characteristics of primary disper- is crude by comparison. Once they have sal-density curves are consistent across reached the ground, secondary dispersal these vectors: scale of dispersal distance can re-distribute seeds by considerable varies, but the general shape does not. distances. However, short-distance disper- However, secondary dispersal can change sal has a major contribution to patchiness these relationships considerably. It is also within fi elds, the level of competition that shown how different vectors acting se- plants experience, the microsites in which quentially and/or in parallel can combine seedlings emerge and the maintenance of to produce distinctive dispersal-density species diversity. Whether we look at the curves.

Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 9 soil and by breaking seed dormancy and instigating mass germination of seed that survives the fi re. The reduction or loss of Post-fi re weed management: lessons from the ground stratum, understorey and canopy Victorian Alps vegetation (depending on fi re intensity) can also reduce competition for light, Charlie Pascoe, Parks Victoria, PO Box 20, Bright, Victoria 3741, Australia. moisture, nutrients and physical space for plants that either survive a fi re, whether unburnt or only partially burnt above ground or as underground organs (roots, bulbs, etc.), or that are able to germinate Summary Wildfires can create both On 7–8 January 2003 and again on 1 quickly on the burnt ground. The creation threats and opportunities for weed man- December 2006 extensive dry lightning of vast areas of bare ground can provide agement. An understanding of these can storms started widespread fires in the ideal germination opportunities that are help land managers maximise benefi ts and Victorian Alps. 1.08 million hectares was otherwise rarely available to many colo- minimise risks from post-fi re weed control burnt in the ‘Alpine Fires’ from January nising species. The availability of light and programs. The primary aims of post-fi re to March 2003, including 489,000 ha of the fl ush of nutrients in the ash bed after weed control should be to minimise the national parks and conservation reserves. a fi re can promote rapid growth of seed- impacts of weeds on important natural The 2003 Alpine Fires burnt most of the lings. and other assets, including neighbouring Alps east of the Buffalo and Wonnangatta By their inherent nature, and despite freehold land, and to take advantage of valleys as well as similarly extensive areas the long period of adaptation of much of new or enhanced weed control opportuni- in the high country of New South Wales Australia’s indigenous fl ora to fi res, many ties resulting from the fi re. and the Australian Capital Territory. weeds of native vegetation can take ad- The management of each weed species The ‘Great Divide Fires’ burnt 1.05 vantage of post-fire establishment and following fi re should depend on how that million hectares between December 2006 growth opportunities as well as, or better weed is likely to respond to the fi re, how and February 2007, including most of the than, many native species. Some weeds, fi re control works may have affected the western and southern parts of the Victo- such as Cirsium vulgare (Savi) Ten. (spear weed, and how both the fi re and fi re control rian Alps. Three other major fi res during thistle), may only proliferate and prosper may have affected existing weed control this period: Tawonga, near Mt Beauty, for a short period post-fi re and will be out- programs. With most of the Victorian Alps (33,600 ha); Tatong, north of Mansfi eld, competed once native vegetation re-estab- burnt in wildfi res in 2003 and 2006/07, sig- (31,800 ha); and Hermit Mountain, south lishes, while others, such as introduced nifi cant resources have, and are being, de- of Corryong, (2800 ha) burnt further sub- Rubus spp. (blackberries) can maintain voted to post-fi re weed control on public stantial areas within the Victorian Alps. their place in the regenerated ecosystem, land. Lessons learnt following these fi res The combined burnt area included 362,000 competing very ably with native species. have been formulated into a set of princi- ha of national parks and reserves. Ap- While not a direct effect of fi re, there is ples for post-fi re weed management that proximately 130,000 hectares was burnt in increasing evidence of a strong association are broadly applicable on public land. A both 2003 and 2006/07, including most of between fi res and fl oods (including fl ash number of species case studies illustrate the Tawonga and Hermit Mountain fi re fl oods) in the Victorian Alps (Lynch 2007, the application of the principles to post- areas. PV and DSE 2007). Steep catchments de- fi re weed control programs. Weed management on public land in nuded of vegetation by fi re are vulnerable the Victorian Alps is a high priority and to accelerated erosion, particularly during Introduction receives substantial investment, particu- heavy rain events post-fi re. Floods are well The Victorian Alps covers the extensively larly on Parks Victoria’s park and reserve known as both disturbance events depos- forested mountainous region straddling estate. Most of Parks Victoria’s weed con- iting sediment for seed germination and the Great Dividing Range in eastern Vic- trol programs are designed to meet at least vectors for dispersal of weed propagules toria. For the purposes of this paper, it is one of three broad objectives: (e.g. Howell and Benson 2000, and refer- loosely defi ned as the area from Mansfi eld 1. Protect important natural, cultural ences therein). In post-fi re fl oods there is in the west to Bonang, north of Orbost, and other values within parks and re- potential for long-distance transport of in the east. It encompasses the upper and serves, weed propagules (particularly seeds and mid-catchments of many of eastern Vic- 2. Reduce the risks of weeds expanding easily-broken branches of species such toria’s major rivers and includes a wide their distributions to new areas within as Salix spp. (willows) and Vinca major L. range of geologies, landforms, altitudes, parks and reserves, and (blue periwinkle), although this does not climates and vegetation types, from alpine 3. Prevent the spread of weeds onto appear to have been studied in Australia. peaks above the tree line at over 1800 m neighbouring public and private land. Signifi cant rain events caused substan- to deeply-incised, forested valleys at less tial erosion in parts of the Buckland and than 300 m. Much of the Victorian Alps Fire and weeds Mitta Mitta catchments following the 2003 remains as forested public land, although Wildfi res and fi re control efforts can have Alpine Fires, and massive erosion in upper there are many areas of cleared and for- signifi cant effects on weeds and weed con- parts of the Gippsland Lakes catchments ested freehold land, particularly in fertile trol programs, an understanding of which after the 2006/07 Great Divide Fires. Sedi- valleys and on fl atter uplands, on both is important for post-fi re weed manage- ment deposited along river banks follow- sides of the Great Divide. ment. ing such events can be many tens of cen- Within the Victorian Alps, Parks Victo- timetres deep and could provide a nutri- ria manages over 800,000 ha of parks, con- Impacts of fi re on weeds ent-rich substrate for the germination and servation reserves and historic areas, in- Fire can directly affect both weeds and na- growth of water-dispersed weeds. Again, cluding Alpine National Park (ANP) and tive plant species in many ways. Fire can there appear to have been no studies of Mount Buffalo National Park (MBNP). kill plants, both through combustion and this to date and anecdotal observations Most of the remaining public land is State radiant heat. Fire can also deplete seed following the 2003 fl ood events were in- forest managed by the Department of Sus- banks, both by killing viable seed stored adequate to distinguish between weeds tainability and Environment (DSE). on plants and in the upper layers of the in situ pre-fi re growing up through the

10 Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 deposited sediment and new incursions positive outcomes can include: north-westerly winds predominate dur- taking advantage of the new substrate. • Killing of mature plants, ing the main fi re season (summer to early Following the extensive post-fi re fl ood- • Depletion of weed seed banks, autumn). Wind-dispersed weeds that oc- ing in Gippsland in mid-2007, there will • Easier chemical or physical control of cur upwind of recently burnt areas could be a heightened need to monitor for, and even-aged post-fi re regeneration, and pose a high risk of invasion into post-fi re respond to, the establishment of weeds in • Improved visibility of, and access to, ash beds. riparian areas. weed infestations. Investigations of wind dispersal of These opportunities are discussed further weeds and inter-relationships with fi re Impacts of fi re control on weeds under ‘Principles for post-fi re weed con- would be a valuable area of research. An The Code of Practice for Fire Management trol on public land’, below. investigation of this nature, involving Sa- on Public Land in Victoria (DSE 2006) rec- The most immediate effect of fi res on lix cinerea L. (grey sallow) on and around ognises the potential for fi re control ac- weed control is the disruption of weed the Bogong High Plains, parts of which tivities to facilitate weed invasion. Min- management programs underway, or pro- have been twice burnt in four years, is eral earth control lines built to contain fi res grammed to commence, during a major commencing in 2007/08 (J. Moore person- create extensive areas of, often, severely wildfi re event. While managers can at- al communication). disturbed, bare ground. In addition, the tempt to keep priority weed programs Native and introduced animals, which machinery, vehicles and even personnel going, many agency staff are likely to be can carry seeds in their fur or gastro-intes- used to control fi res can carry weed seeds diverted to assist with the fi re control ef- tinal tracts (Carr et al. 1992, Muyt 2001, Ey- into and around a fi re area. Weeds could fort. A program manager’s duty of care to les 2002), may be forced to forage widely potentially be transported vast distances weed control contractors and fi eld-based in search of adequate food post-fi re, and in this way, depending on how far away staff, and inability to oversee their works could help disperse weeds into previously fi re fi ghting equipment and personnel are and safety, make it almost inevitable that uninfested areas. Weeds can also be intro- sourced from. Weeds can also be trans- on-ground works will be disrupted, even duced into new areas in fodder brought ported relatively short distances, from in areas not directly threatened by fi re. onto fi re-affected properties. With losses weedy to previously ‘clean’ sections of a This disruption can occur at critical times of fencing during fi res and therefore in- fi re ground, either as seed or vegetative for some weed control programs, particu- creased levels of stock wandering, as well propagules. The complete absence of es- larly for weeds whose periods of fl owering as licensed grazing in State forests and tablished vegetation and the scarcity of and seed production coincide with peak other crown land, stock could spread both native plant propagules on mineral earth fi re season (e.g. many exotic Asteraceae existing and new weeds into burnt areas. control lines means that any weed prop- (daisies and thistles) at higher elevations). agules deposited during fi re control face Other than attempting to keep vital con- Fire recovery planning little competition from other plants. trol programs rolling along with minimal The ‘Guidelines and Procedures for During the 2003 Alpine Fires, over 7500 supervision, there is little weed managers Managing the Environmental Impacts of km of fi re control lines were constructed, can do in these situations. In many cases, Weeds on Public Land in Victoria’ (DSE including over 6000 km on public land it is possible to recommence control works 2007) summarise the comprehensive State (DSE 2005). Many of these were actively later in the season (e.g. blackberry control and Federal legislative and policy frame- rehabilitated in the fi rst three months post- programs are often effective during au- work around weed control on public land. fi re by respreading windrowed vegetation tumn); otherwise works need to be held With the exception of the Victorian ‘Code and topsoil back across them. The respread- over until the following year. With the po- of practice for fi re management on pub- ing of topsoil, in particular, may help to tential for increased fi re frequency due to lic land’ (DSE 2006), which requires that constrain weed establishment, to some global warming weed managers should weed management issues be addressed in degree, by facilitating the re-establish- consider building contingency plans for fi re-recovery plans, none of these instru- ment of competing native species whose fi re events into weed control programs. ments address weed control in relation to propagules are contained in the soil. The fi re. However, all remain as applicable in effectiveness of control line rehabilitation, Other relationships between weed the post-fi re environment as at any other including the distribution and abundance dispersal and fi re time. of weeds in relation to disturbance and There appears to have been little detailed Following each of the major wildfi re rehabilitation history, and landscape con- investigation into the specifi c dispersal events in Victoria in recent years, the State text, are presently being investigated (S. vectors and patterns of key weeds in the government has announced fi re recovery Cutler personal communication). To date, Victorian Alps, let alone the effects that programs, which have included signifi cant 27 weed species have been recorded on fi res might have on these. The availabil- pest plant and animal control components 2003 fi re control lines in ANP and MBNP ity of post-fi re ash beds, however, is likely (e.g. Victorian Government 2007). Parks (S. Cutler, personal communication), in- to enhance germination success of weed Victoria and DSE have worked collabo- cluding many species not normally found seeds that are able to disperse into burnt ratively in developing integrated public in undisturbed forest environments. These areas. land fi re recovery plans (e.g. PV and DSE weeds are predominantly pasture grasses Wind can disperse the seeds of a wide 2003, 2007), in accordance with the require- and herbaceous weeds commonly found range of weeds (e.g. Asteraceae, Salix ments of the ‘Code of practice for fi re man- in paddocks, roadsides and other dis- spp.) over both short and long distances agement on public land’ (DSE 2006). turbed areas. When completed, this study and many species rely on wind as a key The fi re recovery plan for the 2006/07 should provide public land managers dispersal vector (Carr et al. 1992). The fi res (PV and DSE 2007) recognises that: with improved knowledge for managing strong winds and powerful updrafts often ‘The post-fi re environment provides both weeds on control lines constructed during associated with wildfi res may contribute opportunities and threats [for weed con- future fi res. to the dispersal of certain weeds that set trol]. For example, prompt action can seed during the fi re season. An under- minimise regrowth of weeds, but failure Impacts of fi re on weed control standing of wind patterns both during to act can result in even more dense infes- Fires can provide signifi cant opportuni- and following fi res could help weed man- tations of fi re-responsive species, which ties for improved weed management agers predict potential dispersal patterns will choke regeneration of native species outcomes, as well as negative impacts of certain wind-dispersed weeds. In the and potentially spill over onto neigh- on weed control programs. Some of the Victorian Alps, strong and gusty north to bouring private land.’

Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 11 The ‘2007 Report from the Ministerial assessment can help to identify priority health and safety of fi re fi ghters, public Taskforce on Bushfi re Recovery’ (Victorian fi re recovery tasks, it needs to be augment- safety, and protection of public and pri- Government 2007) identifi es two broad ed by other knowledge and information. vate assets, will also be addressed in fi re priorities for post-fi re weed control: pub- A summary of the broad weed manage- control plans. Even during the height of a lic land close to private land boundaries ment priorities for public land following fi re, land managers have an opportunity to and sites of vulnerable native vegetation the 2006/07 fi res is contained in Table 1. seek to incorporate into fi re control plans or fauna. The report states that activities measures to prevent or minimise the risks are to focus on invasive and fi re-sensitive Principles for post-fi re weed control of spread of priority weeds. Such measures weeds (both new and emerging) to limit on public land could include information for crew brief- regeneration. A number of principles can be identifi ed ings, vehicle and machinery washdown Specifi c post-fi re priority weed man- that can assist land managers to minimise requirements, and ‘no go’ areas). Ideally, agement actions need to be identified the impacts of weeds following wildfi res hygiene plans for the prevention of spread through a combination of systematic risk and to maximise the benefi ts that can be of critical weeds would be developed in assessment, collation of other information achieved from post-fi re weed manage- advance and incorporated into pre-fi re- (including local knowledge) and consid- ment. season planning, ready for immediate roll- eration of basic principles of weed man- out should a fi re occur in, or threaten, any agement (e.g. Robinson 1996, DSE 2007), Prevention priority weed containment areas. including principles of post-fi re weed con- Principle 1: Minimise the availability of trol (see below). weed propagules post-fi re by maintaining Principle 3: Facilitate the stabilisation and The risk-management approach in- a high level of pre-fi re weed control. For revegetation of bare ground as quickly as volves identifying key values and threats many weeds, there are likely to be fewer possible post-fi re. Bare ground provides to those values within the fi re area. How- opportunities to establish post-fi re if their enhanced opportunities for weed estab- ever, risk assessment depends to a high propagules are not present, or reduced, lishment. Once bare areas have become degree on the availability of knowledge pre-fi re. Unfortunately, this requires a high revegetated, any weed propagules present on both values and threats. Values vary in level of continuous weed control, which is or newly arriving need to compete with their distribution and signifi cance, while often well beyond the resource capacities established plants to establish themselves. threats vary in distribution and sever- of most public land managers. However, Stabilisation and prevention of soil erosion ity. Just as importantly, different values managers of fi re-prone areas should, as a assists native vegetation establishment (as- are affected to varying extents by differ- priority, try to maintain a high level of con- suming suitable propagules are present), ent threats. Values that may be affected trol over weeds which have a high poten- thereby helping to limit opportunities for by weeds include: threatened species and tial to spread during or following fi re. weed germination. Conversely, the reten- communities, waterways and wetlands, tion of bare ground can maintain receptive visitor access and amenity, and neigh- Principle 2: Incorporate weed hygiene seedbeds for weed invasion and height- bouring land. While information on val- principles and practices into fi re control ened risks of soil instability and erosion ues is often mapped and stored system- plans. While the primary focus of any (and thus further bare ground for weed atically, information on threats, usually, fi re control effort will always be on fi ght- establishment). is not. Therefore, while a systematic risk ing the fi re, many other matters, such as

Table 1. Broad public land weed control priorities following the 2006/07 Great Divide Fires (from PV and DSE 2007). Key Value Broad project description

Native vegetation across the fi re Eradicate new and emerging weeds wherever they occur (e.g. Himalayan honeysuckle, affected area black knapweed, Paterson’s curse and African lovegrass). (Includes supporting surveillance activities for these and other new and emerging species.)

Prevent establishment of new invasive species along fi re control lines, water fi lling points, access tracks, etc.

Native vegetation across the fi re- Reduce the threat of established, high-risk, fi re-responsive weeds, focusing on the affected area with a focus on the highest value areas at risk (e.g. Cape and English broom, willows, gorse and hawkweeds, highest priority asset areas particularly where threatening alpine bogs, alpine heathlands and Poa grasslands).

Reduce the threat of established weeds with a focus on the highest value areas at risk (e.g. blackberry, willow, blue periwinkle and other riparian weeds in priority river reaches and other priority sites – e.g. those containing threatened species or communities).

Alpine bogs Remove willows and other pest weeds.

Private land adjoining parks and On public land adjoining private land, decrease the cover of fi re responsive weeds, such as reserves and forests English and Cape broom.

Decrease the impact of key weeds along the interface, such as blackberry.

Decrease the impact of gorse by reducing its potential to move into new areas.

Undertake surveillance and treat new and emerging weeds on public land along interface areas, including ragwort, Paterson’s curse, and other weeds.

12 Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 Principle 4: Undertake surveillance Principle 6: Remove remaining indi- riparian weed infestations can also be dif- for, and respond to, new and emerging vidual mature plants from infestations. fi cult to access, without the construction of weeds. Following a major disturbance Where fi re has killed the majority of ma- expensive and environmentally damaging event, such as a wildfi re, there is an in- ture plants in a weed infestation, the win- access tracks. The economic and environ- creased likelihood that new weed infesta- dow of opportunity to achieve effective mental costs of access construction can tions will appear. In some cases, these may control of the weed can be signifi cantly outweigh potential weed control benefi ts. only be small range extensions of existing improved if any surviving mature plants However, with the removal of vegeta- infestations (e.g. a weed establishing in the or unburnt patches are also killed. This tion during a fi re there can be signifi cant next catchment over from where it was needs to occur before these survivors have opportunities to gain relatively easy access previously known). However, new weeds a chance to set seed and replenish post- to previously inaccessible weed infesta- from much further away could also ap- fi re seed stores. For weeds that take two tions. Post-fi re, surviving pines and ripar- pear. Weed managers need to implement or more years to become reproductively ian weeds can be much more visible and programs to enable them to be both vigi- mature, preventing the replenishment of accessible to ground crews. Even where lant for, and able to rapidly respond to, seed stores from surviving individuals can new access tracks need to be constructed, any new incursions as soon as possible. It signifi cantly increase the window of op- this may be accomplishable with less cost is important that new infestations be de- portunity to accomplish their control or and impact than if the site were unburnt. tected before they have an opportunity to eradication. Sometimes fire control lines may even set seed (recognising this time frame can have been constructed in locations that co- vary greatly between species). Principle 7: Capitalise on post-fi re de- incide with weed access needs. However, Weed managers should use their knowl- pletion of weed seed stores. Weed seed as rehabilitation of new fi re control lines edge of the weed assemblages in and stores can be signifi cantly depleted post- often occurs quite early post-fi re, weed around burnt areas, and local knowledge fi re, particularly where fi re intensity has managers need to identify potentially use- of weed responses following previous been high. Like many indigenous Austral- ful control lines for access to weed infes- disturbance events in similar locations, ian plants, which have adapted to fi re by tations as soon as possible and negotiate to forecast possible new weed incursions storing seed, either on the plant or in the to keep them open while weed control is that may occur. Particular consideration soil, for germination en masse post-fi re, undertaken. should be given to possible ‘sleeper weeds’ many weeds have a similar capability (e.g. that are known to be present but are con- Cytisus scoparius (L.) Link, Genista monspes- Risk management sidered to be relatively benign, but which sulana (L.) L.A.S.Johnson, Leycesteria for- Principle 9: Protect priority values at risk could rapidly invade burnt areas post-fi re. mosa Wall. and Ulex europaeus L.). Post-fi re from weeds post-fi re. All environmental As a precaution, such sleeper weeds, as mass germination is an ecological strategy weed management programs should, as a well as any known, isolated occurrences that allows seedlings to take advantage priority, aim to protect the highest value of invasive species in or near burnt areas, of the nutrient-rich post-fi re ash bed and natural and other assets in an area that are should be eliminated as soon as possible reduced competition from adult plants considered to be at the greatest risk from post-fi re. during establishment and growth. Heat the impacts of weeds. This is termed an While a general alertness for new and during a fi re can also kill weed seed. The asset-based risk management approach emerging weeds may suffice in many greater the fi re intensity, the deeper the (DSE 2007). DSE (2007) outlines a frame- burnt areas, some situations may demand potential heat penetration into the soil and work and detailed planning process for a more proactive level of surveillance. Fol- the greater the potential mortality or ger- identifying priority values at risk from lowing the 2003 Alpine Fires, and in recog- mination of the seedbank. weeds. nition of the potentially altered weed risk Effective control of post-fi re weed seed- Due to the large scales of the 2003 and post-fi re, Parks Victoria commissioned a lings, following mass germination, can 2006/07 fi res and the need to produce survey and assessment of the weeds and prevent re-establishment of infestations in draft recovery plans promptly post-fi re, cultivated plants at Clover Arboretum in which mature plants have already been this process was not able to be followed in ANP near Mt Beauty (Ecology Australia eliminated (either through fi re-mortality its entirety. However, the key steps of iden- 2006). This assessment recommended the or post-fi re control, or a combination of tifying and prioritising values, identifying control or elimination of 25 cultivated the two). However, post-fi re weed seed- threats (weeds being only one of many), and/or naturalised taxa. Based on local ling density can sometimes be so high that assessing the risks of particular weeds to knowledge of the invasiveness of some it is diffi cult to kill all individuals. There specifi c values, and selecting priority sites of the cultivated and naturalised exotic can also be subsequent germination from for weed control were undertaken (PV and plants in Falls Creek village, Parks Victo- the remnant seedbank following initial DSE 2003, 2007). ria also commissioned a cross-tenure post- seedling control. Follow-up treatment of Table 1 identifies, in general terms, fi re survey for invasive weeds in the Falls surviving or newly germinating seedlings some of the highest priority values threat- Creek area (Carr et al. 2004). This survey may therefore be required in the same ened by weeds identifi ed in the 2006/07 also recommended a number of priority season and/or subsequent years. A good Great Divide Fires fi re recovery plan (PV actions for land managers. understanding of a weed’s ecology and and DSE 2007). Relatively weed-free ex- regular monitoring of its response to both amples of vegetation that has an inherent Opportunism fi re and treatment will help determine high susceptibility to weed invasion (e.g. Principle 5: Take advantage of post-fi re the necessary ongoing treatment require- moist environments, fertile riparian veg- reductions in weed biomass. Many large ments. etation communities) are considered to and dense weed infestations can be costly be of high value and therefore are a high and diffi cult to control. The sometimes Principle 8: Take advantage of improved priority for attempting to maintain in a massive reductions in weed biomass fol- post-fi re visibility and access to weed low-weediness or weed-free state. lowing fi res can provide opportunities to infestations. Many weed infestations control infestations with substantially less in the Victorian Alps are in steep and/ Principle 10: Expect the unexpected. effort and resources than would otherwise or thickly vegetated terrain. Weeds such There are often likely to be unknowns be required. as Pinus spp. (pine wildings) growing when post-fi re weed programs are being amongst a dense understorey can be very developed. For example, there could be diffi cult to both detect and reach. Many weeds whose response to fi re is not fully

Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 13 known. While the literature may provide priority, attempt to at least maintain, if not requires a holistic, cross-tenure approach, indications of likely responses (e.g. time build upon, their pre-fi re weed control with each land manager doing their bit to to reach reproductive maturity, based on programs. achieve a common goal. studied or cultivated populations), what In recent years, around the Alps, Parks actually occurs locally, at the range of el- Principle 13: Address key post-fi re com- Victoria has developed a wide range of evations and landscape positions within munity and political expectations. Fol- weed management partnerships, both for- the fi re area and with potential interacting lowing a wildfi re local communities, par- mal and informal, with other agencies (in effects of other factors, such as drought, ticularly neighbours of public land, can be particular catchment management author- may differ. It is also possible that new in- highly concerned and emotional about the ities, alpine resorts, Department of Prima- festations of weeds that were previously causes, impacts and control of the fi re, pre- ry Industries (DPI) and DSE), researchers present, but undetected, could be discov- fi re land management, and post-fi re recov- and neighbours. These cooperative pro- ered and may require a signifi cant level of ery (Collins 2006). Politicians can some- grams are achieving better and more sus- response to contain. Conversely, some ex- times refl ect community angst by being tainable results than previously, and are isting occurrences of fi re-sensitive weeds critical of public land management and sometimes able to attract additional fund- could be eliminated by fi re. fi re management practices (Collins 2006). ing due to their cooperative nature. The It is important that land managers It is important that public land managers development of a cross-tenure post-fi re maintain suffi cient fl exibility in their post- recognise and accept these reactions and weed control plan, or broader fi re recov- fi re programming to respond to any new implement post-fi re management actions ery plan (e.g. PV and DSE 2003, 2007) is a or emerging threats or opportunities. In that demonstrate sound land manage- valuable fi rst step in cooperative manage- the fi rst summer following the 2003 Alpine ment. ment, but it needs to be followed through Fires, alpine bogs on the Bogong High Following a fi re, effective control of into implementation and evaluation, so Plains were quite unexpectedly invaded weeds on the private land interface, par- that cooperative management becomes by S. cinerea seedlings (see case study, be- ticularly where neighbours have reason- the norm, rather than the exception. low). Due to the signifi cance of the threat able levels of weed control on their own to these valuable wetlands, substantial properties, is an important demonstration Resourcing additional funding was obtained from of sound land management. The identifi - Principle 15: Constrain the ambition of the North East Catchment Management cation and control of any new or emerging post-fi re weed control programs to re- Authority and a signifi cant response pro- weeds following fi re needs to be a very alistic objectives. Following a fi re, and gram initiated. high priority. There also needs to be effec- possible associated signifi cant reductions One of the most important unknowns tive communication with neighbours to in weed biomass, some widespread, in- following the 2006/07 Great Divide Fires listen to any concerns they may have and transigent weed problems may suddenly is what the cumulative effects of being to allow them to have input into plans for seem to be containable. While this may be burnt twice in four years will be for many weed control near the boundaries of their so for certain species in some locations, for weeds. Again, the literature will assist, to land. Through discussion, opportunities many weeds it will not be the case. Some some degree, but on-ground monitoring for mutually benefi cial cooperative weed weeds in the Victorian Alps, (e.g. intro- will be vital for an accurate understanding control works may be identifi ed. duced Rubus spp. and Hypericum perfora- of weed responses. More broadly, post-fi re control of weeds tum L. – St. Johns wort) are so widespread in areas of high visitor activity, such as pic- that it would be logistically impossible to Principle 11: Consider the likely conse- nic and camping areas, fi shing spots and control all, or even most, infestations, even quences of not controlling fi re-respon- along major access routes, can help miti- with a very large post-fi re weed control sive weeds post-fi re. Many weeds are gate ongoing community concerns about budget. well adapted to fi re and can respond vig- public land management. Icon sites that The objectives of post-fi re weed control orously if left unchecked. Land manag- attract high visitor numbers, such as the programs therefore need to be constrained ers must consider the risks posed if these Wonnangatta Valley and Howqua River by the realities of both short-term fi re re- weeds are not actively controlled post-fi re. fl ats in the Victorian Alps, may require covery budgets and realistic expectations Whilst effective weed control can be re- a higher level of weed control than sites of likely future budgets to maintain post- source-demanding in the initial post-fi re with lower visitation. Recognition of, and fi re gains. There is little value in under- period, the potential consequences of not appropriate measured response to, pub- taking massive weed control programs for implementing effective weed control are lic expectations can assist land managers two to three years post-fi re, only to see the increased risks of vastly greater and more to keep focused on implementing prior- weeds recover in subsequent years, due to intractable weed management problems ity post-fi re works, rather than reacting to a lack of resources that should have been in the future. Small or predominantly ju- community discontent. foreseeable at the outset. venile infestations are likely to be much In relatively undisturbed sites, many easier and more cost-effective to control Principle 14: Enhance partnerships post-fi re weed control programs should than large, established, mature infesta- and integration of programs across generally be self-maintaining once native tions. boundaries. The signifi cant impacts of vegetation cover reaches suffi cient height fi res on weeds and existing weed control and density to out-compete invading Continuity activities provide an opportunity for land weeds. In disturbed areas, however, there Principle 12: Protect and capitalise on pre- managers to review the effectiveness of the is likely to be a greater requirement for on- existing weed control investments. One management and delivery of their existing going weed management, for other weeds of the priorities for post-fi re weed man- weed control programs. Except for small are as likely as native species to replace agement should be the protection of any and isolated infestations, effective strate- weeds that are killed. pre-fire weed control investments. Re- gic weed control usually requires a tenure- sources for weed control on public land blind, cooperative, collaborative and con- Principle 16: Allocate suffi cient resources are often tight and land managers cannot sistent approach (DSE 2007). Individual for post-fi re weed management. Fires can afford to waste any investment in strate- land managers can achieve short-term and provide a unique opportunity to achieve gic and effective weed control. If within localised control of certain weeds by work- benefi cial weed control outcomes, but can the potential scope of post-fi re weed man- ing in isolation from their neighbours, also initiate the development or expansion agement budgets, managers should, as a but broadscale and/or long-term control of substantial post-fi re weed infestations.

14 Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 Weed managers need to allocate suffi cient Broad L. formosa control areas were de- Introduced Rubus spp. (blackberries) resources for increased weed vigilance lineated, based on known pre-fi re distri- Introduced Rubus spp. are extremely wide- and survey for at least two to three years bution and likelihood of occurrence in ar- spread in the Victorian Alps, particularly post-fi re. Increased resources are also like- eas not previously searched in detail. The in gullies and riparian areas, but also in ly be needed for at least the same period, highest priority post-fi re was to visit, map disturbed areas and moist environments or potentially much longer, for intensive and control all known infestations. In the away from waterways. Rubus spp. are so on-ground weed control. Due to the spe- fi rst post-fi re season, the commencement ubiquitous and readily dispersed by native cifi c impacts of fi res and fi re control ac- of the control program was delayed until and introduced birds and mammals that tivities on weeds, local staff are unlikely to early February, much later than in pre-fi re in most areas eradication is not possible. be able to meet legislative obligations (see years, to allow more time for young plants Broadscale control over wide areas, while DSE 2007), community expectations or to grow suffi ciently to be visible above the technically feasible, would be too costly, environmental needs without additional rapidly regenerating understorey. Any even for a well-funded fi re-recovery pro- resources. surviving mature plants were killed. The gram. Control programs therefore need to second priority was to search all gullies be strategic and targeted at protection or Monitoring and evaluation and other moist areas considered to have maintenance of specifi c values, e.g.: Principle 17: Monitor infestations to reas- potential to harbour L. formosa, within the • threatened species (and their habitats) sess control requirements and evaluate broadly delineated areas, to ascertain its or communities; effectiveness. All known weed infesta- presence. All such areas were searched in • areas of predominantly blackberry-free tions should be assessed following fi re to the three years following the 2003 fi res. vegetation; determine both the immediate impacts By 2–3 years post-fire, the vigorous • access along roads and tracks for rec- of the fi re and the post-fi re responses of regeneration of native understorey veg- reational users and along rivers and the weeds. Repeated monitoring may be etation had significantly increased the streams for anglers; required, as different weeds will respond diffi culty of accessing areas and locating • the natural visual landscapes that visi- to fi re at different rates and the responses L. formosa seedlings. As rangers had ob- tors enjoy, particularly around look- of some weeds may change over time. served that it takes 2-3 years for L. formosa outs and picnic and camping areas; Monitoring is also essential following to produce viable seed (D. Smithyman and commencement of control programs to al- personal communication), a strategy was • the boundaries of freehold land, where low evaluation of treatment effectiveness. adopted of leaving some catchments for the adjacent land is either blackberry- If treatment results are not satisfactory or two years between searches for new plants. free, or has an effective control pro- off-target impacts signifi cant, the control It was felt that this would allow seedlings gram. program may need to be adjusted. An to grow suffi ciently to increase their vis- Rubus spp. recover very well following fi re effective monitoring and evaluation pro- ibility amongst the heavy undergrowth, and are capable of redeveloping a dense gram allows such adjustments to be made without allowing the production of seed. canopy from roots protected from exces- as early as possible. However, there remained a risk that any sive heat in the soil, within two to three plants missed would produce seed prior years. In addition, except in very intense Case Studies to any subsequent visit. At the time of the fi res, at least some above-ground parts of The following species case studies help to 2006/07 fi re, many infestations consisted some Rubus plants often survive the pas- illustrate application of the above princi- of only small scattered individual plants sage of fi re, due to their preference for ples for post-fi re weed control. of 1–2 years of age, hidden under the for- growing in moist habitats, thereby further est understorey, requiring much careful enhancing the recovery of Rubus post-fi re. Leycesteria formosa (Himalayan searching to discover. Very few larger, The re-establishment of Introduced Rubus honeysuckle) mature plants remained. spp. is often quicker than the recovery of Leycesteria formosa, a native of the Himala- Due to the presence of mature plants in many native plants, particularly woody yas and western China that was imported gullies on crown and private land outside species. into Australia as a garden plant, can rapid- MBNP, park staff are working with neigh- Parks Victoria’s approach to post-fi re ly colonise undisturbed habitat (Hosking bouring land holders and land managers control of introduced Rubus spp. in the et al. 2006). It is spread by mammals, birds to encourage and assist them to control L. Alps has been to build on pre-existing and water and is capable of out-competing formosa infestations on their land, to help strategic control programs and generally native plants to the point of completely prevent reinfestation of the park. only to tackle new sites where improved altering the fl oristics of a forest understo- The combined impacts of the 2003 fi re access or other changed circumstances rey if left uncontrolled. Prior to the 2003 and the intensive post-fi re control program provide an opportunity, that was not pre- Alpine Fires, there were established infes- have had a signifi cant impact on L. formosa viously apparent or available, to achieve a tations across a considerable area of the in MBNP, with many previously heavily level of control that is likely to be able to be foothill forests of MBNP and in the east infested areas having only low infestations maintained in the medium- to long-term. Kiewa Valley of ANP, south of Mt Beauty. when burnt again in 2006/07. L. formosa The temptation to undertake large-scale Pre-fi re, efforts to control the species had is capable of withstanding repeated fuel Rubus control programs, for which follow- been carried out for many years in MBNP reduction burning (Robinson 1996), but its up funding is unlikely to be available, is and to a lesser extent in ANP. response to two wildfi res in four years is generally avoided. Following the 2003 Alpine Fires, the uncertain. However, it is likely that, given majority of existing L. formosa infesta- its ability to produce viable seed within 2– Cytisus scoparius (English broom) tions in both parks were reduced to ju- 3 years, seedbanks are likely to have been Cytisus scoparius has probably been present venile plants, germinated from the soil at least partially replenished by any ma- in the Victorian Alps for over 100 years, seed bank post-fi re. The post-fi re control ture or regenerating plants missed during with infestations now present in many program was focused on MBNP, to build the 2003–2006 control programs. However, areas, including the Mitta Mitta, Kiewa, on the long-term pre-fi re control work un- the 2006/07 fi re provides another oppor- Ovens, Buckland, King, Howqua and De- dertaken there and because L. formosa was tunity to resurvey and retreat all known latite Valleys and Bogong High Plains. It considered to pose a direct threat to sev- infestations again, without the hindrance is a highly invasive woody weed that can eral threatened plant and animal species of a dense native understorey. form both a dense canopy and deep lit- on the slopes of Mount Buffalo. ter that is able to out compete most native

Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 15 species. The high density and fl ammabil- be applied at a limited number of sites. a disturbed area on the edge of the Falls ity of mature C. scoparius stands makes Thus, a level of control suffi cient to pre- Creek alpine resort. Isolated infestations surrounding vegetation more susceptible vent post-fi re fl owering and seeding was of it, too, have since been found in a south to the effects of fi re, due to the intensity only achieved in a few areas. easterly direction from Falls Creek, on the with which C. scoparius burns (Parsons Other priority actions following the Bogong High Plains. and Cuthbertson 1992). 2003 Alpine Fires included a comprehen- The 2003 Alpine Fires burnt substantial Cytisus scoparius can flower and set sive post-fi re broom mapping program areas on the Bogong High Plains and some seed in the second season post-fi re (Al- (Wells 2004), re-establishment and further areas within the Falls Creek Alpine Resort, lan et al. 2004), although at high elevations expansion of the biological control pro- although very few of the sites where H. reproductive maturity may take four years gram by DPI, establishment of an adap- aurantiacum had previously been recorded (Robertson et al. 1999). Soil-stored seed vi- tive experimental management program were burnt (refer maps in Williams et al. ability remains high for many years, up to trial the effectiveness and impacts of 2006). Nevertheless, following the 2003 Al- to 80 years if stored dry (DNRE 1998). different post-fi re chemical control tech- pine Fires, Parks Victoria was concerned Broom stands can therefore form massive niques and timing, including off-target that H. aurantiacum could disperse into soil seed banks of up to 65,000 seeds per impacts on native species (Allan et al. 2004, post-fi re ash beds. A targeted survey for square metre (DNRE 1998). Seed is read- 2005, 2006), and establishment or enhance- H. aurantiacum was therefore commis- ily spread by water and in mud attached ment of partnerships with a wide range sioned in late 2003, although this was sub- to vehicles, machinery, animals and foot- of agencies and organisations to improve sequently expanded to incorporate other wear. It can spread rapidly down water- cross-boundary collaboration and effec- invasive weeds known to occur in the Falls ways and along roads and tracks (DNRE tive landscape-scale management of C. Creek area, including H. praealtum (Carr 1998). In some areas of the ANP, there is scoparius. Following the failure to prevent et al. 2004). Fortunately, despite extensive a very strong association between the oc- post-fire seedbank replenishment over searching, only one new occurrence of H. currence of C. scoparius and the movement wide areas, the post-fi re strategy for C. aurantiacum and two of H. praealtum were pathways of cattle up into summer graz- scoparius has also been further updated to located (Carr et al. 2004). However, further ing areas (Robertson et al. 1999). guide ongoing control. occurrences of both species have since been Effective management of C. scoparius re- Following the 2006/07 fi res, PV and recorded, although there is no clear link be- lies on depletion of soil seed banks (Rob- DSE will be targeting the mostly iso- tween these new sites and fi re history. ertson et al. 1999, McArthur 2003). Fire of lated C. scoparius infestations within the To better refi ne search effort for H. au- suffi cient intensity can contribute substan- fire area, particularly those previously rantiacum, Williams et al. (2006) developed tially to this outcome by heat-killing seed treated pre-fi re. Surviving plants will be a constrained habitat suitability model in the upper part of the soil profi le and sprayed to prevent seed set in spring 2007. for the species (including fi re history as a by stimulating mass germination of much Spraying of seedlings will commence at disturbance factor), which they combined of the seed bank that remains (Robertson the same time and, based on monitoring with a dispersal model, based on wind et al. 1999, Muyt 2001). Seed bank deple- of effectiveness and ongoing recruitment, direction and frequency during the fl ow- tion of up to 90% can be achieved post-fi re continue as frequently as required each ering season. This model was to be used (DNRE 1998). year to control survivors and new germi- to guide survey effort in the Falls Creek/ The 2003 and 2006/07 fi res burnt the nants. Considerable effort will be made to Bogong High Plains area during sum- majority of C. scoparius infestations in the prevent any further seed set in controlled mer and autumn 2007, but survey plans Victorian Alps, killing most mature plants. infestations. were thwarted by the Tawonga Gap fi re in The 2003 Alpine Fires, however also de- A trial of the strategic use of prescribed December 2006, which again burnt parts stroyed all of the C. scoparius biocontrol re- fi re to re-deplete seedbanks that have re- of the Bogong High Plains, although not search and nursery sites, which had been charged following the 2003 Alpine Fires is any known Hawkweed sites. An intensive established and nurtured over many years planned for 2007/08. Again, chemical con- Hawkweed survey effort is now planned (Allan et al. 2005). Following the 2003 fi res, trol of any surviving plants and all regen- for summer 2007/08, to attempt to locate Parks Victoria and DPI developed a post- erating seedlings will be undertaken re- any unidentifi ed infestations. fi re plan for the management of C. sco- peatedly, as required, following the burn. At Mt Buller, H. aurantiacum was dis- parius, which proposed a multi-pronged If successful, and subject to assessment of covered in 2002/03, growing on a little- strategy to achieve progressive control cumulative impacts on native vegetation, used track (Williams and Holland 2007). of the weed (McArthur 2003). The over- further strategic use of prescribed fi re may Despite control efforts, it subsequently all aim was to take advantage of post-fi re be employed in the control of C. scoparius expanded onto the adjacent batter and seedbank depletion and prevent replen- infestations in the Victorian Alps. downslope into adjacent dense sub-alpine ishment by killing any surviving plants vegetation, beneath which it was very dif- in spring 2003, then controlling post-fi re Hieracium aurantiacum L. (orange fi cult to detect. Again, intensive surveys regeneration from autumn 2004 onwards. hawkweed) and H. praealtum Vill. ex planned for 2006/07 were frustrated by Unfortunately, the size and complexity Gochnat (king devil hawkweed) fi res. The Great Divide Fires burnt all of of the overall post-fi re recovery program All Hieracium species are State Prohibited the vegetation on both sides of the track, meant that resources were not allocated weeds. Hieracium aurantiacum is thought to but left the track itself unscathed (L. Per- soon enough and chemical control, even have been present in Falls Creek village for rin personal communication). Surveys for of most of the surviving patches of mature over 20 years, although it was only record- potentially undetected infestations will plants, did not commence until autumn ed at naturalised there in 1999 (Williams now be easier over the 2007/08 summer, and spring 2004. By that time, but particu- and Holland 2007). From Falls Creek, it while native vegetation remains relatively larly the second spring post-fi re (2004), has invaded nearby areas of the Bogong sparse in burnt areas. the density and height of seedling regen- High Plains in ANP, apparently dispersing During the 2006/07 fi res at both the eration in many areas restricted chemical on north-westerly winds (Carr et al. 2004, Bogong High Plains and Mt Buller, pro- penetration, thus allowing many plants Williams et al. 2006), although human and cedures were incorporated into daily fi re to survive and produce fl owers and seed animal vectors may also have contribut- control plans to minimise the risks of (Allan et al. 2005, 2006). The application ed to its expanding range (Williams and spread of Hawkweeds. Fire crews work- of multiple chemical treatments prior to Holland 2007). Hieracium praealtum was ing in areas in which Hawkweeds may fi rst seeding in late spring 2004 could only only discovered in late 2003, growing in have been present were briefed on the

16 Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 species and risks by local staff and vehi- source populations of mature S. cinerea (Weed Management Society of South cles and machinery were excluded from were not present to invade bogs post-fi re. Australia, Adelaide). areas of known infestations. As a precau- With hindsight, a more proactive approach Allan, C., Wright, J. and Raymond, K. tion, a washdown point was established to willow control prior to 2003 may have (2004). An adaptive experimental man- in a suitable location in Falls Creek village prevented, or at least restricted, the sever- agement program for English broom to clean down potential soil and seed car- ity of S. cinerea invasion of bogs post-fi re. Cytisus scoparius (L.) in Victoria. Pro- rying areas of all vehicles and machinery However, it is possible that at least some ceedings of the 14th Australian Weeds that operated off formed vehicle tracks on of the S. cinerea source populations were Conference, eds B.M. Sindel and S.B. the Bogong High Plains. located well beyond Parks Victoria’s area Johnson, pp. 191-4. (Weed Society of of management within ANP. Importantly, New South Wales, Sydney). Salix cinerea L. (grey sallow) due to the slow recovery time of alpine Carr, G.W., Roberts, N.R., Wearne, L.J. and Salix cinerea is a highly invasive willow bogs after fi re, many bogs can retain ex- McMahon, J.B. (2004). Alpine National of waterways and wetlands. It seeds pro- posed, moist, bare peat for several years Park post-fi re mapping of orange hawk- lifi cally and seed can be wind dispersed post-fi re. These areas, if moist at the time weed and other pest plants. Ecology over tens of kilometres (National Willows of seed dispersal, provide ideal seedbeds Australia Pty Ltd for Parks Victoria. Taskforce 2006). Many alpine bogs burnt for invading S. cinerea. Fire-affected bogs Carr, G.W., Yugovic, J.V. and Robinson, intensively during the 2003 Alpine Fires. may therefore need inspection and follow- K.E. (1992). ‘Environmental weed in- These bogs are a threatened vegetation up willow control annually, until bare ar- vasions in Victoria’. (Department of community listed under the Victorian Flo- eas are fully revegetated. Conservation and Environment, Mel- ra and Fauna Guarantee Act (1988) and Parks Victoria will be actively monitor- bourne). provide habitat for a range of threatened ing bogs across the 2006/07 fi re areas over Collins, P, (2006). ‘Burn: the epic story of fl ora and fauna species. With the burning coming years to assess, and respond to, bushfi re in Australia’. (Allen and Un- of bog vegetation, extensive areas of ex- post-fi re weed invasion. It is hoped that win, Crows Nest, NSW) posed bare peat remained. the extensive control of both mature and DNRE (1998). Landcare notes – English In the summer of 2004, a year after the juvenile S. cinerea between 2004 and 2006 broom. Department of Natural Resourc- fi res, park rangers and alpine botanists on and around the Bogong High Plains es and Environment, Keith Turnbull Re- alike were surprised to discover S. cinerea and Mount Hotham (and, to a lesser search Institute, Frankston Victoria. seedlings appearing in vast numbers in degree, Mount Buffalo) will have lessened DSE (2005). ‘The recovery story: the 2003 alpine bogs on the Bogong High Plains. the seed source for further invasion there. Alpine fi res’. Department of Sustain- Such a signifi cant incursion was not ex- However, given the potential for long dis- ability and Environment, Melbourne. pected, as willows had not invaded alpine tance wind dispersal of S. cinerea, it is like- DSE (2006). ‘Code of practice for fi re man- bogs following previous major fi res in the ly that many substantial potential source agement on public land’. Revision 1. Alps, such as Black Friday in 1939 and populations have not yet been treated. Department of Sustainability and Envi- the Caledonia fi re on the Gippsland high ronment, Melbourne. plains in 1998. However, searches in gul- Acknowledgements DSE (2007). ‘Guidelines and procedures lies and other moist areas on the Bogong Much of the material for the case studies for managing the environmental im- High Plains, nearby Mount Hotham, and in this paper came from discussions with, pacts of weeds on public land in Victo- the surrounding valleys showed that ma- and unpublished project reports writ- ria 2007’. Department of Sustainability ture S. cinerea were widespread and very ten by, current and former Parks Victoria and Environment, Melbourne. much alive, despite the recent fi res. Smaller rangers and weed project offi cers. I thank Ecology Australia (2006). ‘An assessment S. cinerea seedling infestations in Mount them for so freely sharing their observa- of cultivated plants as environmental Buffalo bogs were also recorded. These tions and project results and for allowing weeds at Clover Arboretum, Alpine searches, and subsequent more systematic me to utilise their collective knowledge. National Park’. Ecology Australia Pty aerial surveys during autumn, benefi ted Fiona Ede kindly provided comments and Ltd for Parks Victoria. from the sparseness of native vegetation references on fl ooding and weeds. Sera Eyles, D. (2002). ‘Sambar Deer (Cervus post-fi re. Cutler provided preliminary results from unicolor) as a potential seed vector for Following the discovery of S. cinerea in- her study of fi re control line rehabilitation the spread of the environmental weed vasion into bogs Parks Victoria developed success. Louise Perrin provided informa- Himalayan honeysuckle (Leycesteria for- a two-pronged strategy for control: hand tion on H. aurantiacum at Mt Buller. Stefan mosa) at Mount Buffalo National Park’. pulling of seedlings in bogs and other Kaiser provided valuable comments on a B.Sc. (Hons) thesis, University of Mel- areas as they appeared; and survey and draft of this paper. bourne, Parkville, Victoria. control (mainly cutting and painting) of Hosking, J., Sainty, G. and Jacobs, S. (2006). potential source populations. A systematic References ‘Alps invaders – weeds of the Austral- approach was adopted for the survey of Allan, C., Raymond, K. and Wearne, L. ian high country’, 2nd edition. Austral- source populations, with waterways and (2005). The role of monitoring in weed ian Alps Liaison Committee. other moist areas closest to, and situated management: a case study from the Vic- Howell, J. and Benson, D. (2000). Predict- to the north of, the Bogong High Plains torian Alps. Proceedings of the Weed ing potential impacts of environmental and Mount Hotham being targeted fi rst. society of Victoria 2nd Victorian Weed fl ows on weedy riparian vegetation of Considerable funding was provided by Conference, pp. 63-6. (Weed Society of the Hawkesbury–Nepean River, south- the North East Catchment Management Victoria, Frankston). eastern Australia. Austral Ecology 25, Authority to assist with this work and Allan, C., Wearne, L., Price, J., Keatley, 463-75. Mount Hotham and Falls Creek Alpine M. and Tumino, P. (2006). Best-prac- Lynch, A. (2007). The climate of the Alpine Resorts also enthusiastically cooperated. tice chemical control of English broom Shire – dealing with extremes. Proceed- Large numbers of volunteers, as well as (Cytisus scoparius) evaluated in Alpine ings of the Australian Alps National staff and contractors, pulled out many National Park, Victoria, through adap- Parks, Climate Change Management tens of thousands of S. cinerea seedlings tive experimental management pro- Implications, Science Management between 2004 and 2007. gram. Proceedings of the 15th Austral- Workshop, ed. G. Anderson, pp. 133-9. It seems likely that across the Victorian ian Weeds Conference, eds C. Preston, Australian Alps Cooperative Manage- Alps in 1939, and at Caledonia in 1998, J.H. Watts, N.D. Crossman, pp. 243-6. ment Program.

Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 17 McArthur, K. (2003). English broom post fi re plan. Department of Primary In- dustries, Frankston, Victoria. Muyt, A. (2001). ‘Bush invaders of south- east Australia’. (R.G. and F.J. Richard- son, Melbourne). National Willows Taskforce (2006). Wil- lows in the spotlight – Salix cinerea. The Willow Strip: Newsletter of the Nation- al Willows Taskforce 1, 5. Parsons, W.T. and Cuthbertson, E.G. (1992). ‘Noxious weeds of Australia’. (Inkata Press, Melbourne.) PV and DSE (2003) ‘Statewide public land ecological and cultural fi re recovery plan’. Parks Victoria and Department of Sustainability and Environment, Melbourne. PV and DSE (2007) ‘Great divide fi re re- covery plan’ (Draft submitted for ap- proval, June 2007). Parks Victoria and Department of Sustainability and Envi- ronment, Melbourne. Robertson, D.C., Morgan, J.W. and White, M. (1999). Use of prescribed fi re to en- hance control of English broom (Cyti- sus scoparius) invading a subalpine snowgum woodland in Victoria. Plant Protection Quarterly 14, 51-6. Robinson, K. (1996). ‘Weed management manual for the Australian Alps Na- tional Parks.’ Australian Alps Liaison Committee. Victorian Government (2007). ‘2007 Re- port from the ministerial taskforce on bushfi re recovery’. Department of Inno- vation, Industry and Regional Develop- ment, Melbourne. Wells, N., (2004). ‘English broom (Cyti- sus scoparius) mapping post 2003 bush fi res’. Internal report, Parks Victoria, Omeo. Williams, N., Hahs, A., Morgan, J. and Hol- land, K. (2006). A dispersal constrained habitat suitability model for orange hawkweed (Hieracium aurantiacum) on the Bogong High Plains, Victoria. Parks Victoria Technical Report. Parks Victo- ria, Melbourne. Williams, N.S.G. and Holland, K.D. (2007). The ecology and invasion history of hawkweeds (Hieracium species) in Aus- tralia. Plant Protection Quarterly, 22, 76- 80.

18 Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 defoliated in late November, and although several fl ushes of new leaves were pro- How can willow sawfl y aid our willow management? duced through the season, high willow sawfly numbers meant that the trees Fiona Ede, Department of Primary Industries, PO Box 48, Frankston, Victoria were moderately to severely defoliated for the remainder of the season (Ede et al. 3199, Australia. 2007). Tree deaths as a result of willow sawfl y activity have been reported from South America (Dapoto and Giganti 1994, Al- derete and Fidalgo 2004), southern Africa Summary Willows are serious weeds of Northern Hemisphere include both males (Urban and Eardley 1997) and New Zea- waterways, impacting on physical and and females whereas populations in the land (Charles et al. 2004). The experience biotic processes. The discovery of willow Southern Hemisphere appear to con- of catchment managers in the North Island sawfl y (Nematus oligospilus) in Australia po- sist entirely of females (Koch and Smith of New Zealand suggests that tree deaths tentially has implications for willow man- 2000). can occur in the second season of severe agement activities. Willow sawfl y larvae The adults lay eggs on fully expanded willow sawfl y infestations (I. McIvor per- eat willow leaves and when populations willow leaves, generally on the upper leaf sonal communication). There have yet to increase to high levels, entire willow trees surface. The eggs hatch after about a week be any deaths of willow trees in Australia can be defoliated. Experience with willow and larvae initially feed on the leaf material that can be categorically blamed on severe sawfl y incursions in other Southern Hemi- adjacent to the eggs. There are fi ve to sev- defoliation as a result of willow sawfl y sphere countries has shown that ongoing, en larval instars before pupation (Charles activity, but it is possible that some trees severe defoliation of trees can result in tree and Allan 2000). Pupation occurs in a co- badly affected in the last two or three sea- deaths. It is possible that sustained high coon attached to an intact leaf, branch or sons will fail to produce new foliage in the populations of willow sawfl y in Victoria stem, or other available surface (Ede et al. coming spring. will result in a change of investment focus 2007). Adults emerge after four to fourteen However, it has been reported that in in willow management programs and may days, depending on conditions, and com- some Southern Hemisphere populations, aid the transition from willow dominated mence oviposition immediately (Ovruski willow sawfly does not maintain high riparian communities to those dominated 1994). The duration of the life cycle from population levels in the long term. For ex- by healthy native riparian vegetation. oviposition to adult emergence can be as ample, in several regions in the North Is- short as 22 days under ideal conditions land of New Zealand, willow sawfl y activ- Introduction (Charles and Allan 2000). Six generations ity initially resulted in tree defoliation and In Victoria and other parts of south eastern per season have been reported from sites death, but sustained activity over more Australia, willows (Salix spp.) are serious in New Zealand (Charles et al. 2004), but than two or three seasons was not evident weeds of waterways and wetlands, due in Chile, only two generations per season (Cowie 2006). The only region where wil- to their highly invasive nature and det- were observed (Gonzalez et al. 1986). low sawfl y populations have maintained rimental impacts on river health. These As willows are deciduous, willow saw- high numbers and caused ongoing defoli- impacts include modifi cation of stream fl y overwinter in a dormant state. In order ation and tree death is Hawkes Bay on the channels and hence river fl ows, exacer- to do this, the larvae which develop late east coast of the lower North Island (Ede bation of riverbank erosion, exclusion of in the season spin dark brown cocoons in 2006). It is not known whether physical or native species from riparian vegetation the soil or leaf litter under willow trees biotic factors such as predation, parasit- communities, alteration of in-stream food (Ede et al. 2007). The larvae remain in ism or disease result in declining willow webs and reduction of access for recrea- these cocoons in the larval form through- sawfl y populations. tional pursuits. For these reasons, several out winter and pupate into adults at the Salix species are declared noxious in Vic- beginning of spring. The emergence of Willow sawfl y in Australia toria and are listed as Weeds of National adults commences in October in New Zea- The fi rst reported observations of willow Signifi cance (WoNS). Signifi cant resources land (Charles and Allan 2000) and Chile sawfl y in Australia were made in March are currently invested in willow manage- (Gonzalez et al. 1986), which also appears 2004, when infestations were observed ment, with control costs in Victoria alone to be the case in Victoria (Ede et al. 2007). defoliating crack and weeping willows totalling about $10m per annum. Emergence is asynchronous resulting in a in the Canberra area, with further infesta- The recent arrival of willow sawfly range of age cohorts in the population at tions of willow sawfl y recorded 150 km (Nematus oligospilus) in Australia has the any time through the season. In Victoria south of Canberra in 2004 (Bruzzese and potential to affect some willow manage- in 2007, willow sawfl y larvae entered the McFadyen 2006). ment programs. High populations of this overwintering phase in early May (Ede et In the 2004/05 season, willow sawfl y insect can result in tree defoliation and al. 2007). was confi rmed at additional sites in the eventually tree death, and willow sawfl y ACT and southern NSW, in the Adelaide has caused significant damage to wil- Impact of willow sawfl y on willows Hills and in a suburban garden in north low populations in other countries in the Willow sawfl y larvae eat willow leaves. west Melbourne (Bruzzese and McFadyen Southern Hemisphere, particularly New They consume about two leaves in total 2006). The known distribution of the insect Zealand (Disbury et al. 2004). during their development, and can eat an increased in the 2005/06 season, with wil- entire leaf in their last day as a larva prior low sawfl y reported at several sites across Willow sawfl y biology to pupating (Ede 2006). New leaves de- south eastern Australia (Finlay and Adair Willow sawfl y, Nematus oligospilus (Förster velop to replace those eaten by the larvae, 2006). Tree defoliation occurred at several (Hymenoptera: Tenthredinidae)), is native but if larval numbers are high, these new sites in north east Victoria and in central to the Northern Hemisphere. It is wide- leaves are rapidly consumed. NSW, in addition to those sites previously spread in the Holarctic (northern) Region Defoliation of entire trees occurs when affected (Ede 2006). of Europe and is also found in Eurasia willow sawfl y population levels are high. It is not known how willow sawfly and North America (Koch and Smith In the 2006/07 season, willow trees at a arrived in Australia. It is possible that 2000). Populations of willow sawfl y in the site in north east Victoria were initially cocoons may have been inadvertently

Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 19 imported in packaging material or contain- up to damaging levels for a few years only, willow sawfl y activity will result in an ers, or that adults blew across the Tasman or at a limited number of sites, and over increase in understorey light levels. This Sea from New Zealand in easterly wind the long term, willow sawfl y has very lit- change in light environment may facilitate systems (Bruzzese and McFadyen 2006), tle impact on willow populations, and on invasion by other riparian weeds or pro- as has occurred for other insects. their management. The second scenario mote the establishment of native species, involves wide scale and longer term im- depending on the surrounding vegetation Willow sawfl y in Victoria pacts, with high willow sawfl y popula- and land use. At sites with limited connec- Willow sawfl y was fi rst observed in Victo- tions continuing to cause tree defoliation tions to native vegetation communities, it ria in April 2005 in north west Melbourne and death for several years, and/or in may be desirable to undertake planting (Bruzzese and McFadyen 2006), with a many locations across the state. This out- or seed bank augmentation with native number of further sightings occurring in come would have a signifi cant impact on species in the early stages of a willow the past two seasons (Finlay and Adair willow management. sawfl y infestation. By minimising distur- 2006, Ede et al. 2007). The most extensive It is not yet possible to predict which bance under the existing willow canopy, outbreak is in the Kiewa Valley in north of these scenarios is the most likely. How- and through appropriate species selection, east Victoria, where defoliation of S. × ru- ever, the results from the 2006/07 season this process would favour the rapid estab- bens trees was fi rst noticed in December indicate that high willow sawfl y popula- lishment of a native riparian community 2005. Willow sawfl y populations in this tions have developed simultaneously at which is able to fl ourish once the willows area have caused severe defoliation of large disparate locations (north east Victoria, are no longer live. This management tech- numbers of willow trees along at least 15 central Victoria and southern Victoria) nique would be best suited to sites where km of the valley for the past two seasons. causing tree defoliation in willows of sev- willows can be left in situ once dead, and Defoliation of willow trees has now eral different taxa. is less likely to result in severe infestations been recorded at several other sites in Vic- There are a number of implications for of other riparian weeds. toria, including the Murray River in the willow management under the second Further research is required to under- north east; the Broken River and tributar- scenario. At sites frequented by the public, stand these interactions between willow ies, Goulburn River, and Campaspe River or where downstream infrastructure such sawfl y impacts and willow management in central Victoria; the Tarago River, Pheas- as bridges is a concern, death of whole and restoration of native riparian com- ant Creek, and Fosters Creek in southern trees or of large branches may result in munities. As well, the impact of willow Victoria; and at Boneo on the Mornington the need to remove dangerous material, sawfl y on reproductive potential of seed- Peninsula (Ede et al. 2007). which will require resource investment, ing willows is a topic of interest to willow and may shift investment priorities in the managers. If willow sawfl y decreases tree Willow taxa affected by willow short term. Tree deaths along rivers where fi tness, does this in turn decrease seed pro- sawfl y willows currently form part of the river duction to such an extent that seeding wil- Field experience in New Zealand indicates bank stabilisation infrastructure will also lows are no longer a signifi cant problem that willow sawfl y is more likely to de- require additional investment to ensure in the environment? The relationship be- foliate tree willows than shrub willows, that the loss of willows in those areas does tween tree defoliation (both extent and fre- although the shrub willow S. purpurea not result in signifi cant bank erosion. quency) and declines in tree root biomass (purple osier) is susceptible to willow However, in many cases the death of is another potential avenue for research. sawfl y (Cowie 2006, Ede 2006). Charles et willow trees will be viewed as a benefi cial Preliminary fi ndings from New Zealand al. (1998) found in laboratory trials that outcome by catchment managers. If wil- suggest that root biomass declines dra- adults selected shoots of tree willows or low sawfl y is effective at killing willows of matically with defoliation (Ede 2006) and purple osier for oviposition, but not shoots particular taxa, then it may be possible to a better understanding of this relationship from other shrub willows. shift resources from these species to those would aid willow management. For ex- In Victoria to date, willow sawfl y has less affected by willow sawfl y. Currently, it ample, at sites where high willow sawfl y defoliated crack willow (S. fragilis), golden appears that willow sawfl y is more likely populations initially cause tree defoliation upright willow (S. alba var. vitellina), crack to adversely affect tree willows such as but do not persist for a suffi cient length × golden willow (S. × rubens), black wil- crack willow, golden willow, black willow of time to kill trees, it may be desirable to low (S. nigra), Chilean pencil willow (S. and various hybrids, with shrub willows kill affected trees using chemical or physi- humboldtiana), tortured willow (S. matsu- such as grey sallow less affected. If wil- cal methods while they are in a weakened dana ‘Tortuosa’), weeping willow (S. ba- low sawfl y is able to signifi cantly impact state. bylonica), and golden weeping willow (S. on populations of tree willows, then re- Although it is highly unlikely that wil- sepulcralis var. chrysocoma). sources could be reallocated to addressing low sawfl y will adversely affect all wil- In addition, willow sawfl y has been the threats posed by grey sallow and other lows across all of Victoria, it is possible found on New Zealand hybrid willow (S. taxa. that this insect will have suffi cient impact matsudana hybrids), purple osier (S. pur- In situations where willow sawfl y caus- to allow a redirection of current invest- purea) and grey sallow (S. cinerea) (Ede et es occasional tree defoliation but does not ment in willow control and to maximise al. 2007). lead to tree death, it is likely that current opportunities for the replacement of wil- willow control techniques will be affected low dominated riparian zones with those Implications for willow to some level. Most obvious is foliar ap- dominated by healthy native vegetation management in Victoria plication of herbicides, which requires fo- communities. At sites in Victoria where willow sawfl y liage, but the effi cacy of frill and fi ll and populations establish high densities, it is stem injection techniques would also be Acknowledegments likely that defoliation of susceptible wil- affected by partial or total defoliation as Funding for the willow sawfl y project: low taxa will occur, and tree deaths are a consequence of diminished sap fl ows ‘Adapting willow management strategies possible. However, it is unclear whether through the trees. to deal with spread of the willow sawfl y’ large willow sawfl y populations will per- Many willow management activities is provided by the Environmental Water sist over time and whether they will be currently seek not only to remove willows Reserves and River Health Division, and widespread through the state. from riparian areas, but to restore native the Pest Management Unit, both of the De- Two scenarios are possible. In the fi rst vegetation communities in these areas. partment of Sustainability and the Envi- scenario, willow sawfl y populations build Defoliation of willow trees as a result of ronment (DSE); and by Melbourne Water.

20 Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 Information relating to the New Zea- y Neuquen (Argentina). Bosque 15(1), land experience of willow sawfly has 27-32. been generously provided by The Hor- Disbury, M., Cane, R.P., Haw, J.M., Gilmer, ticulture and Food Research Institute of S.E. and Garner, S.J. (2004). Field evalu- New Zealand (HortResearch), particularly ation of Spinosad for control of the Dr Brent Clothier, Mr Ian McIvor and Mr willow sawfl y (Nematus oligospilus) in John Charles; and by Mr Garth Eyles, Mr Hawke’s Bay. New Zealand Plant Protec- Graeme Hansen and Mr Darren Gorst of tion 57, 244-7. the Hawke’s Bay Regional Council, New Ede, F.J. (2006). Willow sawfl y (Nematus Zealand. oligospilus) in Victoria: Status report, Several people provided assistance with July 2006. Published by Department of fi eld work during the 2006/07 season, par- Primary Industries, Frankston, Victoria, ticularly Valerie Caron, Daniel Clements, Australia, 30 p. http://www.weeds. Leah Fensham, Gamini Nugapitiya and org.au/WoNS/willows/docs/Willow_ staff from the North East Catchment Man- Sawfl y_Status_Report_July_2006.pdf agement Authority. Dr Robin Adair provid- Ede, F.J., Caron V. and Clements, D. (2007). ed helpful comments on the manuscript. Willow sawfl y activity in Victoria: The 2006/07 season. Published by Depart- References ment of Primary Industries, Frankston, Alderete, M. and Fidalgo, P. (2004). The Victoria, Australia, 64 p. http://www. parasitoids of Nematus oligospilus Fo- weeds.org.au/WoNS/willows/ erster (Hym.: Tenthredinidae) in Cali- Finlay, K.J. and Adair, R.J. (2006). Distribu- fornia (USA) for a programme of bio- tion and host range of the recently in- logical control in Argentina. Abstract of troduced willow sawfl y, Nematus oligos- paper presented at 22nd Session of the pilus Förster, on willows (Salix spp.) in FAO International Poplar Commission southeast Australia. Proceedings of the (IPC2004), Santiago, Chile, November 15th Australian Weeds Conference, eds 2004. http://www.fao.org/forestry/ C. Preston, J.H. Watts and N.D. Cross- foris/webview/common/media.jsp? man, pp. 791-4. (Weed Management So- mediaId=7544&langId=1 ciety of South Australia, Adelaide). Bruzzese, E. and McFadyen, R. (2006). Gonzalez, R.H., Barria, G. and Guerrero, Arrival of leaf-feeding willow sawfl y M.A. (1986). Nematus desantisi Smith, Nematus oligospilus Förster in Australia an important new species in Chile (Hy- – pest or benefi cial? Plant Protection menoptera: Tenthredinidae). Revista Quarterly 21, 43-4. Chilena de Entomologia 14, 13-15. Charles, J., Chhagan, A., Allan, D., Fung, Koch, F. and Smith, D.R. (2000). Nematus L., Hurst, S. and McIvor, I. (2004). The oligospilus Förster (Hymenoptera: Ten- willow sawfl y, Nematus oligospilus, in thredinidae), an introduced willow New Zealand: 1997 – 2004. Abstract of sawfl y in the Southern Hemisphere. paper presented at 22nd Session of the Proceedings of the Entomological Society FAO International Poplar Commission of Washington 102(2), 292-300. (IPC2004), Santiago, Chile, November Ovruski, S.M. (1994). Oviposition be- 2004. http://www.fao.org/forestry/ haviour of Nematus desantisi Smith foris/webview/common/media.jsp? (Hymenoptera: Tenthredinidae), a seri- mediaId=7544&langId=1 ous defoliator of willow and poplars. Charles, J.G. and Allan, D.J. (2000). Devel- Revista de Investigacion - Centro de Investi- opment of the willow sawfl y, Nematus gaciones para de Regulacion de Poblaciones oligospilus, at different temperatures, de Organismos Nocivos 9(1-4), 7-13. and an estimation of voltinism through- Urban, A.J. and Eardley, C.D. (1997). Wil- out New Zealand. New Zealand Journal low sawfl y: a contentious issue. Plant of Zoology 27, 197-200. Protection News 47, 20-4. Charles, J.G, Allan, D.J. and Fung, L.E. (1998). Susceptibility of willows to ovi- position by the willow sawfl y, Nematus oligospilus. Proceedings of 51st New Zealand Plant Protection Conference 1998, pp. 230-4. Cowie, B. (2006). Overcoming the threat posed by willow sawfl y Nematus oli- gospilus. A review of research needs and possible options. Prepared for the Foundation of Research, Science and Technology by Environment Manage- ment Services Ltd., Christchurch, New Zealand, 24 p. Dapoto, G. and Giganti, H. (1994). Bio- ecologia de Nematus desantisi Smith (Hymenoptera: Tenthredinidae: Nema- tinae) en las provincias de Rio Negro

Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 21 westerly wetland in the Gippsland Lakes Ramsar site and is probably the freshest: Response of Brazilian milfoil (Myriophyllum wetlands to the east become progressively more saline as they approach Lake Wel- aquaticum) to salinity and water-level fl uctuations lington and the entrance to the Southern and its potential to invade wetlands of the Gippsland Ocean at Lakes Entrance (e.g., see Roache et al. 2006: Raulings et al. 2007). Myriophyl- Lakes, south-east Victoria lum aquaticum was mentioned specifi cally as one of fi ve pest plant species in the Mark Toomey and Paul I. Boon, Institute for Sustainability and Innovation Strategic Management Plan for the Gipps- and School of Biomedical Science, Victoria University (St Albans Campus), land Lakes Ramsar Site (Department of PO Box 14428, Melbourne City Mail Centre, Victoria 8001, Australia. Sustainability and Environment 2003) and as the highest priority pest plant, before even blackberries and willows, in the Lake Wellington Wetlands Draft Management Summary Brazilian milfoil (Myriophyllum slow-moving streams in south-east Plan (Parks Victoria 1997). The basis of aquaticum: Haloragaceae) is an introduced Queensland, near-coastal areas of New this concern is that excessive growths of freshwater aquatic macrophyte that has South Wales, much of Victoria, Tasmania M. aquaticum have manifold impacts on been found in some freshwater wetlands and parts of south-western Western Aus- aquatic ecosystems. Dense stands restrict of the Gippsland Lakes in south-eastern tralia (Parsons and Cuthbertson 2001). stream fl ow, increase the deposition of Victoria and has the potential to spread to Myriophyllum aquaticum is a rhizoma- sediment and organic matter (Evans et al. other fresh and brackish-water wetlands in tous perennial which is distinguished 2003), deplete oxygen concentrations in the region. This paper reports on the effects from other species in the genus by its the water column (Sytsma and Anderson on M. aquaticum of salinity and water-level feathery, glaucous green emergent stems 1989) and increase rates of water loss from fl uctuations, in order to improve under- with pinnate leaves arranged in whorls water bodies via transpiration (Cilliers standing of the invasive potential of this (Aston 1973). By taking root at the margins 1999). Dense accumulations of shoots and species and to determine the effectiveness of shallow lakes and slow fl owing rivers, leaves in the water column provide habi- of water-level and salinity manipulations M. aquaticum extends into wetlands from tat for mosquito larvae, since predation by as potential control strategies. Laboratory the water’s edge up to a depth of about fi sh is much decreased among the tangled experiments showed that plant biomass 1.5 m as a dense mass of tangled rhizomes plant material (Orr and Resh 1991). The was reduced by exposure to salinities as with many emergent shoots (Sytsma and ability of M. aquaticum rhizomes to com- low as 2 g L−1 (= approx. 3300 µS cm−1 at Anderson 1989). It can persist also on pletely colonise moist or fully submerged 25°C) for instantaneous exposures and 4 moist mudfl ats and on gravel (Aston 1973; sediments also affects colonisation by na- g L−1 (= approx. 6600 µS cm−1 at 25°C) for Orchard 1979). Growth of M. aquaticum tive taxa (Washington State Department of gradually increasing salinities. A complete can be problematic under certain condi- Ecology 2003). draw down of water levels in Sale Com- tions and, although it is a perennial, M. Chemical control of isolated outbreaks mon, a wetland in the Gippsland Lakes aquaticum commonly exhibits the growth and manipulations of water regimes for complex, over three months in summer characteristics of an annual; when water larger infestations were proposed by 2006–2007 caused signifi cant reductions in temperature rises in spring shoots rapidly Parks Victoria (1997) as the two most vi- the biomass and surface extent of M. aquat- grow from rhizomes (Sutton 1985). able control options for M. aquaticum in icum. Laboratory incubations with dried All M. aquaticum plants in Australia wetlands of the Gippsland Lakes. Certain- sediments, however, demonstrated that are female, and Sainty and Jacobs (1981) ly, a number of studies have reported on plants could recover quickly from dried warned specifi cally of the danger of acci- water-level draw down as a weed control sediments, so partially desiccated plants dentally introducing male plants through strategy for problematic aquatic species, may re-establish when the wetland fl oods the aquarium trade. As only female plants including Cambomba caroliniana (Goldsby again. Because if its sensitivity to salinity, are present in Australia, reproduction is and Saunders 1977), Hydrilla verticillata we conclude that M. aquaticum has limited entirely vegetative, by way of stem and (Poovey and Kay 1998) and Egeria densa ability to permanently invade brackish- rhizome fragmentation resulting from (Thomaz et al. 2006). Where infrastructure water or salinised wetlands of the Gipps- the action of wind and waves or animal is present to allow hydrological manipula- land Lakes: invasions may occur during disturbance. Where conditions and sub- tions, a draw down of water levels can be periods of inundation with fresh water but strata are suitable, fragments of M. aquati- a cheap and effective weed management plants are likely to be killed or inhibited by cum can readily establish via adventitious technique that has long-lasting effects on subsequent periods of higher salinity. roots, which form at the node and form problematic aquatic vegetation. Previous new individuals (Jacot-Guillarmod 1979). research has reported on the accumulation Introduction The clonal growth habit facilitates long of M. aquaticum biomass under favoura- Myriophyllum aquaticum (Vell.) Verdc.; distance dispersal where there is fl owing ble environmental conditions (Monteiro syn Myriophyllum brasiliense Cambess. is water to distribute vegetative components and Moreira 1990, Sytsma and Anderson an introduced freshwater macrophyte in to new areas (Evans et al. 2003). 1993a), but we are unaware of any report the Family Haloragaceae, native to South Myriophyllum aquaticum was fi rst re- on the response of M. aquaticum to draw America and commonly sold as an aquar- ported in Sale Common, a freshwater downs of water level over the critical sum- ium plant because of its aesthetic appeal wetland in the Lake Wellington wetlands mer period, when desiccation effects are and ease of cultivation (Sutton 1985). As complex of the Gippsland Lakes Ramsar likely to be most severe under Austral- with many exotic plants, M. aquaticum site, in 1992 (Parks Victoria), although ian conditions. Accordingly, the effect of has a propensity to escape cultivation Aston (1973) reported that it was already water-level changes on the growth of M. and it is now naturalised in many wet- present on farm dams around the town- aquaticum was the fi rst objective of the re- lands and slow-moving aquatic systems ship of Lakes Entrance in 1969. Sale Com- search reported in this paper. outside of its native range. It was fi rst re- mon is a 300 ha, shallow freshwater marsh To date there has been little or no corded in Australia in 1908 and has since located 3 km south of Sale in south-eastern consideration given to the role played spread throughout wetlands, ponds and Victoria (38°30’S, 147°51’E). It is the most by salinity in the potential for spread of

22 Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 M. aquaticum in Australia, nor as a possible time dead plants were discarded and re- of the experiment at Week 9. At Week 9 control mechanism that could be imple- placed with healthy specimens to ensure a all plants were washed, sorted, dried and mented by wetland managers under suit- healthy cohort at T0. Plants were exposed weighed as described previously. able conditions. Salinity may play a major to one of fi ve saline solutions: 1, 2, 4, 8 and role in limiting the spread of M. aquaticum 16 g L−1 and were grown with a 12 h light Statistical analysis Plant biomass data in the more brackish-water wetlands of to 12 h dark cycle with light provided by were not distributed normally and, since the Gippsland Lakes, because the Lakes an array of hydroponic fl uorescent lamps attempts to transform them to conform to complex has become increasingly sa- (Gro-lux®, Sylvania Lighting Co., Germa- the assumption of normality were unsuc- linised as a result of the artifi cial opening ny). The temperature was maintained at cessful, the non-parametric Kruskal-Wal- of the lakes to the sea at Lakes Entrance a constant 20°C for the duration of both lis Single Factor Analysis of Variance by in 1889 (Bird 1962, 1966). It is noteworthy experiments. Ranks was used to analyse the effects of that M. aquaticum has been reported in Sale salinity on plant performance. Compari- Common, the freshest of the wetlands in Experiment 1: Sudden salinity increas- sons of biomass to control groups within the Lake Wellington complex and there is es Ten plants each were randomly as- each treatment were made with Nemenyi little or no information on its spread to signed to one of fi ve treatments and an- non-parametric Multiple Comparisons. other, more salty, wetlands to the east. The other set of ten plants set aside as con- Mann-Whitney U tests were used to com- limited amount of research undertaken trols. Plants were exposed to one of the pare responses following the sudden expo- on the species in the past has suggested fi ve saline solutions and any water lost via sure to salt with those of a gradual salinity that M. aquaticum can colonise saline evapotranspiration was replaced with de- increase, always comparing responses at habitats: Haller et al. (1974) reported that, ionised water up to the 1.8 L mark every equivalent fi nal salinities. Data were ana- over a four week period, M. aquaticum in second day; this method maintained sta- lysed using SPSS v. 15.0., with the excep- the USA grew in salinities up to 10 g L−1. ble saline solutions for the duration of the tion of Nemenyi non-parametric multiple This result, however, is in stark contrast to experiment. The experimental exposure comparisons, which were completed long- what is known of the salinity tolerance of ran for nine weeks, after which plants hand as per Zar (1999). most species of freshwater macrophytes were removed and necrotic plant tissue in Australia. There are a range of thresh- was discarded. Roots were discarded Responses to water-level fl uctuations olds below which Australian freshwater also because epiphytic algae sometimes Observations at Sale Common Obser- macrophytes can tolerate salinity and still grew so profusely that they confounded vations of M. aquaticum in Sale Common remain viable, but most do not tolerate sa- an accurate assessment of root biomass. were made on four occasions during the linities greater than about 1–2 g L−1 (Hart The remaining living rhizomes and leaves summer of 2006–2007: 9 November 2006, et al. 1991) and 4 g L−1 appears to be an up- were washed and oven dried at 90°C for 14 December 2006, 15 January 2007 and 12 per limit of nominally freshwater species a minimum of 24 h until constant weight February 2007. Water levels were drawn (Brock 1981). was attained. down by natural evaporation over the Salinity impacts on plants are not hot dry summer of 2006–2007; levels had simple, and there is some evidence that Experiment 2: Gradual salinity increas- fallen markedly after the November 2006 aquatic plants are more tolerant of gradual es Fifty individual plants were chosen sampling and, upon the visit in December increases in salinity (‘presses’) than to sud- for exposure to gradually increasing sa- 2006, the wetland was completely dry. den increases or ‘pulses’ (e.g., see McKee linities. The experimental design aimed Fifteen permanent transects were laid and Mendelssohn 1989). Thus a second to generate, after fi ve weeks, a suite of out randomly, in groups of three at fi ve aim of our research was to determine the plants that had been exposed to gradually locations in Sale Common that had sub- effects of increased salinity on M. aquati- increasing salinities ranging from 1 to 16 stantial accumulations of M. aquaticum. cum and whether the adverse effects of g L−1, with ten replicate plants at each of Each transect started 3 m to the landward salinity were reduced by gradual, rather the fi ve salinities. Plants would then be side of the shoreline or the nearest vegeta- than sudden, increases in salt concentra- kept at their designated salinities for a fur- tion and all transects extended 15 m into tions. ther four weeks, until the experiment was the wetland, with start/end points perma- stopped. nently marked by GPS and stakes. Biomass Method At the beginning of the fi rst week, all was sampled using a 20 × 20 cm quadrat A suite of laboratory incubations were 50 plants were exposed to the 1 g L−1 treat- placed at the mid point of the largest patch used to determine the effects of salinity ment and then were conceptually divided of M. aquaticum along each transect. This and re-fl ooding on M. aquaticum, whereas into fi ve groups (each of ten plants) to be approach was used to determine the maxi- the effects of water-level draw down were allocated to the various fi nal salinities of mum biomass of M. aquaticum present at inferred from a set of fi eld-based observa- 1, 2, 4, 8 or 16 g L−1. Plants were incubated each transect. In November 2006, biomass tions over the summer of 2006–2007. as described above and, after one week, samples were collected by cutting the 40 of the initial 50 plants moved to the 2 fl oating rhizomes and emergent shoots Responses to increases in salinity g L−1 treatment; the remaining 10 plants with shears around the inside perimeter Two laboratory experiments were under- were kept at a salinity of 1 g L−1. These of the quadrat from the top to bottom of taken: 1) a trial in which plants were ex- 10 plants remained at this salinity for the the water column. In subsequent months, posed to a sudden increase in salinity; and duration of the experiment. After a sec- after the water-level draw down, surface 2) a trial in which salinities were increased ond week, 30 of the 40 plants in the 2 g vegetation was collected by hand from gradually up to a maximum of 16 g L−1. L−1 treatment were moved to the 4 g L−1 within a quadrat placed on the ground. Mats of M. aquaticum were separated into treatment and the remaining 10 plant left In December 2006 the quadrat sampling single rhizomes of similar size and condi- in the 2 g L−1 treatment for the duration of position was offset 20 cm to the right by tion and placed in individual 2 L contain- the experiment. Similarly, 20 of these 30 of the November 2006 sampling position ers fi lled with tap water. Each rhizome con- plants were moved at Week 4 to the 8 g sampling position, so as to avoid the pre- sisted of emergent shoot and leaf material, L−1 treatment and 10 left at 4 g L−1; a fi nal vious sampling area. In January 2007 the submerged rhizome, aquatic adventitious 10 plants were moved to the 16 g L−1 treat- quadrat was offset 20 cm to the left and in roots and some submerged leaves. Plants ment at Week 5. All plants then remained February 2007 it was offset again a further were acclimated for two days before the exposed to their designated salinities for 20 cm to the left. Cleaned samples were experiment commenced, during which a further four weeks until the conclusion oven dried at 90°C until a constant weight

Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 23 was attained then dry-weight biomass to salinity of 16 g L−1 having become com- biomass remaining at the end of the sud- was recorded. pletely necrotic by the end of the experi- den exposure trial. The lateral spread of M. aquaticum was ment. quantifi ed each month by measuring the There were also signifi cant adverse ef- Responses to water-level fl uctuations start and end points of vegetation that fects on living biomass in the exposure trial Myriophyllum aquaticum underwent an came into contact with the transect line. where salinities were increased gradually initial growth phase between Novem- Where M. aquaticum surface coverage was (P = 0.021). The harmful effects of salinity, ber 2006 and December 2006, with both patchy or disjointed, it was necessary to however, were less severe when salinities biomass and surface coverage reaching a record several vegetation start and end were increased gradually in a ‘press’ dis- peak in December 2006. The mean maxi- points. The sum of the differences of the turbance than when increased suddenly mum dry-weight biomass of M. aquaticum vegetation start and end points was used in a ‘pulse’ disturbance. At the end of the was 269 g m−2 and 701 g m−2 in Novem- to calculate total lateral spread along each experiment, plants that had been exposed ber and December, respectively. Follow- transect. to gradual increases in salinity had more ing the draw down of water levels in Sale viable rhizome biomass than plants ex- Common over the summer of 2006–2007, Effects of experimental re-inundation of posed suddenly to an increase in salinity both plant biomass and surface extent of dry sediment Upon the fi nal visit to Sale to the same fi nal salt concentration (Figure M. aquaticum fell signifi cantly each month Common in February 2007, 16 soil sods 1). Although stem necrosis in the sudden (Figure 2 and 3) and the mean maximum (15 cm × 10 cm) containing M. aquaticum exposure trial was substantially greater, biomass across all 14 transects had fallen rhizomes were dug from the dry wetland. there was no signifi cant difference in mean to 221 g m−2 in February 2007. A sub-sample was taken from the soil sods biomass between the two treatments with Patterns in surface cover and lateral to measure moisture content and the re- the exception of the 16 g L−1 treatment lev- extent mimicked that of the biomass: mainder of the samples inundated with el (U = 25, P = 0.013), which had no living there was an initial expansion during the fresh water to a depth of 10 cm for 29 days. The experiment took place outdoors and water lost via evaporation was constantly 0.9 replaced. After 29 days the number of new 0.8 Sudden shoots that emerged from M. aquaticum Gradual rhizomes within and above the surface of 0.7 the sediment samples was recorded. 0.6 Statistical analysis Biomass and surface 0.5 coverage data were analysed by a Repeat- ed Measures ANOVA. The assumption of 0.4 normality was verifi ed (Kolmogorov-Smir- Biomass (g DW) 0.3 nov test) but Mauchly’s test for sphericity violated the assumption of equality of var- 0.2 iances. Therefore, the Repeated Measures ANOVA F statistic we used was the more 0.1 conservative Lower-Bound test statistic. 0.0 Bonferroni post-hoc analysis was used to 0124816 compare differences across months. Data −1 from one transect were not analysed due Salinity treatment (g L ) to substantial insect grazing. Figure 1. Changes in mean dry-weight biomass of M. aquaticum under Results sudden and gradual increases in salinity. Means and standard errors are Responses to increases in salinity shown, n= 10. Plant biomass was lost from shoots that developed necrotic regions along the length of the rhizome and following death 900 of the emergent tips of plants. In both the sudden and the gradual experimental tri- 800 als, there was a trend of declining biomass ) 700 of live rhizomes in response to increasing −2 salinity (Figure 1). In the sudden expo- 600 sure trial, salinity signifi cantly affected live rhizome biomass after nine weeks of 500 treatment (P <0.005). The development of 400 necrotic regions among plants exposed suddenly to a salinity of as low as 2 g L−1 300 caused a signifi cant reduction of viable, Biomass m (g DW 200 live rhizome biomass compared with the control group kept under freshwater con- 100 ditions (Q statistic = 5.65, q critical = 2.64). Although there was a steady loss of bio- 0 mass with increasing salinity at all treat- Nov 2006 Dec 2006 Jan 2007 Feb 2007 ment levels, the harmful effects of salinity were most pronounced at the upper range Figure 2. Temporal changes in dry-weight biomass of M. aquaticum from of the salinity treatments (i.e. in excess of transects at Sale Common from November 2006 to February 2007. Means ± 8 g L−1), with the stems of plants exposed standard errors are shown, n=14.

24 Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 early summer and a subsequent reduction As noted earlier, the salt concentration re- in the Lake Wellington complex is Dowds in cover and extent as a consequence of ported by Haller et al. (1974) to be toxic to Morass State Game Reserve, and measure- the water-level draw down in December. M. aquaticum was between 10 and 13 g L−1. ments of salinity in that wetland indicate During the initial growth phase, the mean a salinity regime of from 0.5 to 11 g L−1 extent of M. aquaticum cover expanded In the present study it was not possible between 1992 and 2003 (Boon et al. 2007). by 2.4 m along the transects, refl ecting a to quantify mortality in the way used by Future experiments may, however, ad- signifi cant increase from a mean extent of Haller et al. (1974), who measured growth dress the salinity tolerance of M. aquaticum 5.1 m in November 2006 to one of 7.5 m rates. The extent of stem necrosis, how- obtained from sites with a higher back- in December 2006 (P = 0.001). In response ever, did provide a good indication of the ground salinity than Sale Common, such to water-level draw down, however, M. harmful effects of salt; we found the loss as the Heart Morass to the north of Dowd aquaticum beds began to contract after of biomass was most pronounced at high Morass, and in these cases we might fi nd December 2006. The reduction in surface salinities (i.e., 8–16 g L−1). The harmful ef- more salt-tolerant strains. cover was statistically signifi cant between fects of salt on non-halophytes occur as Our results indicate that salt effects December 2006 and January 2007 (P = much from duration of exposure as from may be delayed or decreased when salini- 0.036) and again between January 2007 salt concentration (James and Hart 1993, ties are increased gradually in a ‘press’ dis- and February 2007 (P = 0.004). By Febru- Howard and Mendelssohn 1999) and it turbance rather than suddenly in a ‘pulse’ ary 2007 the extent of M. aquaticum cover was apparent that the duration of sud- disturbance. Viable biomass was reduced was very disjointed and patchy, and it ex- den exposure to salinity increases was signifi cantly by a sudden exposure to salt tended merely an average of 2.7 m along suffi cient to cause a severe reduction in at only 2 g L−1, whereas the fi rst sign of the 14 transects. M. aquaticum vegetative vigour. Although an impact following gradual increases in The experiment in which dried wetland the results obtained by Haller et al. (1974), salinity occurred 4 g L−1. Moreover, plants sediment was experimentally re-fl ooded for plants in North America, indicate an in the gradual exposure group had higher indicated that re-shooting from desiccated ability of M. aquaticum to invade environ- fi nal live biomass than did those at the rhizomes occurred within one week of in- ments with salinities of up to about 8 g equivalent salinity when salinity was in- undation. By the end of the experiment L−1, our results suggest that M. aquaticum creased suddenly. Under natural condi- at 29 days, the mean number of emergent in south-eastern Australia would be un- tions, both gradual and sudden increases shoots per sediment clod was 14.1, equiva- likely to permanently colonise wetlands in salinity may occur in wetlands: the lent to a plant density of 970 shoots per m2. with a salinity of greater than about 4 to former during, for example, evaporative The mean soil moisture from the sediment 8 g L−1. Although we found signifi cant concentration and the latter during sud- samples taken when the wetland was dry adverse effects of salinities of as low as 2 den, wind-and tidal-mediated intrusions in February 2007 was 49% on a gravimetric g L−1 when salinities were increased sud- of seawater from the adjacent Gippsland basis. denly, it is likely that M. aquaticum could Lakes. invade wetlands during periods of very Discussion low salinity (e.g., after overbank fl ooding Responses to water-level fl uctuations Effect of salinity of wetlands from adjacent rivers, such as The dramatic increase in M. aquaticum bio- The few published reports of the salinity the Latrobe or Avon Rivers in the Gipps- mass at Sale Common between November tolerance of Myriophyllum species suggest land Lakes complex) but then retreat again 2007 and December 2007 was a function of they are more tolerant of saline water than as salinities increased in the wetlands as the synergistic effects of wetland inunda- many other submerged nominally fresh- water levels fell due to evaporation, or fol- tion and late spring climatic conditions, in- water macrophytes. For example, the le- lowing intrusions of saline water from the cluding high temperatures and a long pho- thal salinity for Myriophyllum spicatum in adjacent Lake Wellington. In view of these toperiod, which favoured rapid growth. the USA was reported to be greater than fi ndings, we consider that the more saline This explanation for seasonal growth pat- 13 g L−1 (Haller et al. 1974) and 52% of an wetlands of the Lake Wellington complex terns is consistent with what is known of Australian native milfoil, Myriophyllum are probably not at high risk of permanent the life cycle of M. aquaticum; in temperate crispatum, withstood 72 days of exposure to invasion by M. aquaticum dispersed from climates, warm spring conditions aug- salinities of 7 g L−1 (James and Hart 1993). Sale Common. The nearest large wetland ment rapid growth and expansion (Sutton 1985, Parsons and Cuthbertson 2001) un- til maximum biomass is achieved, which 9 then remains constant throughout sum- 8 mer (Monteiro and Moreira 1990, Sytsma and Anderson 1993b). The maximum 7 mean biomass recorded in December 2007 in Sale Common (>700 g m−2) was 6 lower than that recorded by Sytsma and 5 Anderson (1993b) for a Californian Lake during summer (>1000 g m−2). It is plausi- 4 ble, however, that biomass accumulation and surface coverage had not peaked prior 3 Linear extent (m) to the drought-related draw down of Sale 2 Common in early December 2007 and, had the summer been more mild, biomass may 1 have continued to increase until autumn 0 to reach values similar to those reported for the Northern Hemisphere. Nov 2006 Dec 2006 Jan 2007 Feb 2007 When rooted in shallow water, M. aquaticum typically spreads from lake mar- Figure 3. Temporal changes in extent of M. aquaticum along transects at Sale gins into deeper water as a fl oating mat of Common from November 2006 to February 2007. Means ± standard errors emergent and submerged foliage (Orchard are shown, n=14. 1979, Sytsma and Anderson 1989). The

Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 25 lateral spread of M. aquaticum observed in Summer draw downs lasting longer inventory and management plan for December (mean increase 2.4 m) occurred than three months would probably be re- the Hanford Reach National Monu- at both the ends of several transect lines quired to suffi ciently dry vegetation and ment, p. 141. – at the lakeward and landward limits of soils to control M. aquaticum. Whilst our ex- Goldsby, T.L. and Sanders, D.R. (1977). the vegetation. Wetland inundation in No- perimental results indicated that M. aquati- Effects of consecutive water fluc- vember 2007 was beyond the innermost cum recovered rapidly when recently dried tuations on submersed vegetation margin of the M. aquaticum beds, which al- sediments were re-fl ooded, colonisation of of Black Lake, Louisiana. Journal of lowed the plant to spread laterally in two exposed sediments at Sale Common by Aquatic Plant Management 15, 23-8. directions along some transects (i.e. into tall emergent terrestrial species (e.g. Per- Haller, W.T., Miller, J.L. and Garrard, the wetland and simultaneously toward sicaria decipens) during draw downs have L.A. (1976). Seasonal production and the shore), thereby facilitating surface the potential to create strong competition germination of Hydrilla vegetative spread and biomass increase. Had water for light, space and nutrients when re- propagules. Journal of Aquatic Plant remained at this level throughout sum- fl ooding does occur. In any case, the tim- Management 14, 26-9. mer, it would be possible for M. aquaticum ing of re-inundation is important because Haller, W.T., Sutton, D.L. and Barlowe, to permanently establish a larger surface re-fl ooding may augment nutrient release W.C. (1974). Effects of salinity on coverage by spreading up the shoreline from decaying vegetation and re-fl ooded growth of several aquatic macro- gradient as well as out into the wetland. soil, thereby facilitating rapid plant recov- phytes, Ecology 55, 891-4. The signifi cant reduction of both bio- ery. If re-inundation occurs in late summer, Hart, B.T., Bailey, P., Edwards, R., Hor- mass and surface cover from December it is possible that any benefi ts gained from tle, K., James, K. and McMahon, A. 2006 through to February 2007 is a clear in- the draw down may be quickly lost as M. (1991). A review of the salt sensitivity dication of the deleterious effects that pro- aquaticum grows back rapidly under the of Australian freshwater biota. Hyd- longed summer water-level draw down favourable, warm and nutrient-rich condi- robiologia 210, 105-44. has on M. aquaticum. Owing to the chronic tions. Further fi eld-based research is need- Howard, R.J. and Mendelssohn, I.A. drought experienced in south-eastern ed to determine whether draw downs of (1999). Salinity as a constraint on Australia over the study period, water lev- water level over summer can be used as a growth of oligohaline marsh macro- els were completely drawn down in Sale control measure for M. aquaticum. phytes II. Salt pulses and recovery Common over the summer of 2006– 2007. potential. American Journal of Botany. Despite these severe conditions, M. aquati- References 86, 795-806. cum recovered strongly when dry soil clods Aston, H.I. (1973). ‘Aquatic plants of Aus- Jacot-Guillarmod, A. (1979). Water were inundated with fresh water, indicat- tralia’, p. 82. (Melbourne University weeds in southern Africa. Aquatic ing that this species is tolerant of at least Press, Melbourne, Australia). Botany 6, 377-91. short-term desiccation. The persistence of Bird, E.C.F. (1962). The swamp paperbark. Julien, M.H. (1995). Alternanthera phi- subterranean rhizomes of M. aquaticum in The Victorian Naturalist 79, 72-81. loxeroides (Mart.) Griseb. In ‘The biol- otherwise dry sediments may be an effec- Bird, E.C.F. (1966). The impact of man ogy of Australian weeds’, eds R.H. tive mechanism for surviving water-level on the Gippsland Lakes, Australia. In Groves, R.C.H. Shepherd, and R.G. draw down and sediment desiccation, as ‘Geography as human ecology. Meth- Richardson, p. 1. (R.G and F.J Rich- it has been shown with other species of odology by example’, eds S.R. Eyre and ardson, Melbourne). aquatic plant that moisture loss through G.R.J Jones, pp 55-73. (Edward Arnold, James, K.R. and Hart, B.T. (1993). Effects plant surface material severely reduces London). of salinity on four freshwater macro- plant survival rates (e.g. Alternanthera phi- Australian Government Bureau of Mete- phytes. Australian Journal of Marine loxeroides: Julien 1995, Kruger 2004). orology. (no date). Daily weather obser- and Freshwater Research 44, 769-77. Water-level draw down can success- vations for Victoria. http://www.bom. Kruger, A.L. (2005). Seasonal variation fully limit the growth and colonisation of gov.au/climate/dwo/ICDJDW0300. in the effect of desiccation and dam- aquatic plants providing that it is aimed shtml (Accessed 20 April 2007). age on the propagules of survival of at vulnerable aspects of the target species’ Boon, P.I., Hatton, M., Morris, K., Raulings, alligator weed. Masters thesis, Uni- life cycle. Notoriously problematic aquatic E., Roache, M., Robinson, R. and Salter, versity of New England, Australia. plant species, such as Hydrilla verticillata, J. (2007). Ecology and management of McKee, K.L. and Mendelssohn, I.A. possess underground storage organs (tu- the Lake Wellington wetlands, Gipps- (1989). Response of a freshwater bers) that retain nutrients within the plant land Lakes. Report on the R&D project, marsh plant community to increased (Sculthorpe 1967) and facilitate rapid re- 2003-–2006. Unpublished report by Vic- salinity and increased water level. growth at the onset of more favourable toria and Monash Universities to Land Aquatic Botany 34, 301-16. conditions (Van and Steward 1990). In the and Water Australia (Canberra). Monteiro, A. and Moreira, I. (1990). case of H. verticillata, a late summer draw Brock, M.A. (1981). The ecology of halo- Chemical control of parrotfeather down is most effective as it kills summer phytes in the south-east of South Aus- (Myriophyllum aquaticum). Proceed- re-growth and prevents the formation tralia. Hydrobiologia 81, 23-32. ings EWRS 8th Symposium on of new tubers that could be activated in Cilliers, C. (1999). Lysanthia n.sp. (Cole- Aquatic Weeds. p. 163-4. subsequent years (Haller et al. 1976). As optera: Chrisomelidae), a host specifi c Orchard, A.E. (1979). Myriophyllum M. aquaticum lacks comparable storage beetle for the control of the aquatic (Haloragaceae) in Australasia. I. organs, its ability to withstand sediment weed Myriophyllum aquaticum (Halora- New Zealand: a revision of the genus desiccation is dependent upon the per- gaceae) in South Africa. Hydrobiologia and synopsis of the family. Brunonia ennial rhizomes that facilitate re-growth 415, 271-6. 2, 247-87. once favourable conditions return. This Department of Sustainability and Environ- Orr, B.K. and Resh, V.H. (1991). Infl u- aspect of M. aquaticum biology indicates ment (2003). ‘Gippsland Lakes Ramsar ence of Myriophyllum aquaticum cover that the exact timing of a draw down may Site strategic management plan’. De- on Anopheles mosquito abundance, not be critical other than in terms of the partment of Sustainability and Envi- ovi position and larval microhabitat. extent of sediment drying. The desiccation ronment, Melbourne. Oecologia 90, 474-82. tolerance exhibited by rhizomes, however, Evans, J.R., Nugent, J.J. and Meisel, J.K. Parks Victoria (1997). Lake Wellington suggests that the duration of draw down (2003). Parrot feather: Myriophyl- Wetlands draft management plan. is likely to be critical. lum aquaticum. Invasive plant species Parks Victoria, Kew.

26 Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 Parsons, W.T. and Cutherbertson, E.G. (2001). ‘Noxious weeds of Australia’, p 485. (CSIRO Publishing, Melbourne). Poovey, A.G. and Kay, S.H. (1998). The potential of summer drawdown to manage monoecious Hydrilla. Journal of Aquatic Plant Management 36, 127-30. Raulings, E., Boon, P.I., Bailey, P.C., Mor- ris, K., Roache, M.C. and Robinson, R.R. (2007). Rehabilitation of swamp paper- bark (Melaleuca ericifolia) wetlands in south-eastern Australia: effects of hy- drology, microtopography, plant age and planting technique on the success of community-based revegetation tri- als. Wetland Ecology and Management 15, 175-88. Roache, M.C., Bailey, P.C. and Boon, P.I. (2006). Effects of salinity on the decay of aquatic macrophytes. Aquatic Botany 84, 45-52. Sainty, G.R. and Jacobs, S. (1981). ‘Wa- terplants of New South Wales’, p. 219. (Water Resources Commission New South Wales, Sydney). Sculthorpe, C.D. (1967). ‘The biology of aquatic vascular plants’, p. 340. (Ed- ward Arnold, London). Sutton, D.L. (1985). Biology and ecology of Myriophyllum aquaticum. Paper pre- sented to 1st International Symposium on Water Milfoil (Myriophyllum spica- tum), Vancouver, BC, Canada, 23-24 July 1985. Sytsma, M.D. and Anderson, L.W. (1989). Parrotfeather: impact and management. Paper presented to Proceedings, 41st California Weed Conference, Ontario, CA, 16-18 January 1989. Sytsma, M.D. and Anderson, L.W. (1993a). Transpiration by an emergent macro- phyte: source of water and implications for nutrient supply. Hydrobiologia 271, 97-108. Sytsma, M.D. and Anderson, L.W. (1993b). Biomass, nitrogen, and phosphorus al- location in parrotfeather (Myriophyllum aquaticum). Journal of Aquatic Plant Man- agement 31, 244-8. Thomaz, S.M., Pagioro, T.A., Bini, L.M. and Murphy, K.J. (2006). Effects of reservoir drawdown on biomass of three species of aquatic macrophytes in a large sub- tropical reservoir (Itaipu, Brazil). Hyd- robiologia 570, 53-9. Van, T. and Steward, K.K. (1990). Longev- ity of monoecious Hydrilla propagules. Journal of Aquatic Plant Management 28, 74-6. Washington State Department of Ecol- ogy (2003). Technical information about parrot feather. http://www. ecy.wa.gov/programs/wa/plants/ weeds/aqua003.html (Accessed 9 Oc- tober 2006). Zar, J.H. (1999). ‘Biostatistical analysis’, p. 223. (Prentice-Hall, Inc, New Jersey).

Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 27 drying out a waterway where practical. It may also include aspects of biological con- Aquatic weeds and their impact in irrigation systems trol, if available. With a sound knowledge of other possible approaches to aquatic experiencing drought within Victoria weed management, a good IWM program may be able to be implemented. Currently A B Tim Nitschke and Roger Baker the principal method of control is through A Goulburn-Murray Water Aquatic Plant Services, PO Box 264, Kerang, the use of herbicides. Victoria 3579, Australia. B Goulburn-Murray Water Aquatic Plant Services, 41 High Street, Rochester, Herbicides Victoria 3561, Australia. A number of herbicides are currently used for aquatic weed management. Given the restrictions of the environments such as natural carriers and Ramsar Wetlands in which infestations may grow, options are Summary Aquatic weeds greatly reduce plant between Echuca and Torrumbarry very regulated and restricted. However in the effectiveness of water distribution Weir. There are several other aquatic plant irrigation channels and drains herbicides systems. Aquatic weeds have a negative species that have the potential to create are likely to continue having an ongoing impact on native fl ora and fauna species similar impacts as sagittaria, both within role in aquatic weed management. Regula- and on the integrity of natural waterways. irrigation systems and in natural water- tory bodies on irrigation systems continue Drought is a key factor in the prevention of ways. to expand over time and continually con- further spread of most aquatic weed spe- strain operations. cies. The results of lower channel levels The problem and reduced fl ows due to drought, how- Annual expenditure on the control of Physical/environmental control ever can lead to an increase in abundance aquatic weeds by Goulburn-Murray Wa- The management of aquatic weed infes- of aquatic weed infestations within the ir- ter (G-MW) alone is estimated at over tations by removal of silt by excavation rigation system. $3,000,000 depending on seasonal vari- continues to be undertaken. Problematic Keywords: Aquatic, irrigation, infesta- ables that govern the growth of aquatic issues in channels due to this method are tions, drought, biodiversity. weeds. Aquatic weeds are managed by time to implement, access, cost, and loss G-MW because they impact on service de- of clay lining on channel beds. Mechanical Introduction livery to customers. They block channels control can be very effective through the Aquatic weeds have been a ‘thorn in the and drains, causing increased water levels removal of silt and hence the absence of backside’ to irrigation managers since the that lead to ineffi ciencies in water delivery a suitable environment for aquatic weed early inception of irrigation systems in Vic- and damage to infrastructure. They may growth, however stem fragments through toria. Aquatic weeds are managed because also cause fl ooding where water fl ows in mechanical control pose a threat for fur- they block channels and drains, causing drains are retarded during rain and peri- ther spread. Therefore this method is not increased water levels that lead to inef- ods of high drain fl ow (Gunasekera and suitable for many species. fi ciencies in water delivery and damage Krake 2001). They also have a negative Cutting of Typha spp. is an accepted to infrastructure. One such aquatic weed impact on native species and on the in- method of control, particularly when the example, sagittaria (Sagittaria graminea tegrity of natural waterways. They may plant is cut below water level, allowing Michx.), was not initially treated as a ma- also contribute to the reduction in volume the plant to ‘drown’ (Apfelbaum 2001). jor threat, until the early 1980s, when the of available water within a system due to This process may not be effective against distribution of the plant increased rapidly. uptake and loss from the system through all species, as some species can respond The reasons for this dramatic increase are evapo-transpiration. to cutting by actively putting on new unclear, though one theory may be that growth. the number of propagules produced con- Methodology review Shading is another method of mechani- stantly since the 1960s by smaller popula- History has shown that aquatic weeds can- cal control that is gaining in popularity. tions reached a critical level by the 1980s not be effectively controlled through one Anecdotal evidence that aquatic weed that allowed the plant to spread beyond method alone, but rely on several aspects growth in small, on-farm channels is re- established populations. This is in accord- of control which can determine the success duced by the presence of large shade trees, ance with established principles of aquatic or failure of a program. Early detection and is backed up by the more intense shade weed infestation (Arthington and Mitchell prevention has proven to be the most cost- provided by the use of plastic sheeting 1986) that invasion by aquatic species is effective method of aquatic weed control, (Carter et al. 1994). This sort of control has followed by a period of establishment be- where small infestations can be managed its disadvantages, however, being pro- fore dispersal. effi ciently and effectively. Where aquatic hibitively expensive for large areas, such Sagittaria now infests drains and chan- weeds have established and continue to as Goulburn-Murray Water’s 7000 km of nels across four of the six of Goulburn- disperse an Integrated Weed Management open channels, and being more appropri- Murray Water’s Irrigation Areas in Victo- (IWM) approach needs to be implement- ate for submersed vegetation, over which ria, the Murray and Murrumbidgee Irriga- ed. IWM can be defi ned as the integration the sheeting can sit. As well as this, re-col- tion Areas in southern New South Wales, of effective, environmentally safe and so- onisation is rapid after removal of matting and many natural systems in Northern cially acceptable control tactics that reduce (Eichler et al. 1995), and matting can get Victoria and Southern New South Wales. weed interference below the economic covered with sediment in a dynamic sys- These include the Edwards River, Goul- injury level (Elmore, 1996). In practical tem, providing a fresh substrate for weeds burn River, Broken Creek and associated terms, this means the development of a to colonise. Draw-down and drying of the Nine-Mile and Boosey Creeks, the Ovens management plan that includes aspects of irrigation system allowing over-wintering River, particularly at its confl uence with the target species’ biology, along with tar- to naturally control aquatic weeds is an the River Murray and the River Murray geted or specifi c herbicide use and other operational practice employed by G-MW. itself. By the end of 2005 it was the most management techniques, such as mini- In periods of drought where all water is widespread introduced emergent aquatic mising the spread of weed propagules or precious, outfalling irrigation water at the

28 Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 end of an irrigation season into the natu- Control based on plant biology and ecol- Cother, E.J. (1999). Host range studies of ral carriers has been signifi cantly reduced. ogy morphology, seed dormancy and ger- the mycoherbistat fungus Rhynchospo- Channels have been left at their supply mination, physiology of growth, competi- rium alismatis. Australasian Plant Pathol- level for stock and domestic purposes and tive ability and reproductive biology are ogy 28, 149-55. this has resulted in increased aquatic weed all aspects that need to be used for man- Crump, N.S., Cother, E.J. and Ash, G.J. growth within the irrigation system. agement of aquatic weeds. Information (1999). Clarifying the nomenclature in on seed banks, root reserves, dormancy microbial weed control. Biocontrol Sci- Biological control and longevity of propagules may be used ence and Technology 9, 89-97. This method can be defi ned as ‘the use of to better predict infestations. Weed seed Eichler, L.W., Bombard, R.T., Sutherland, living organisms to suppress a pest popu- bank densities and root reserves can be J.W. and Boylen, C.W. (1995). Recolo- lation, making it less abundant and thus greatly reduced by eliminating seed pro- nization of the littoral zone by macro- less damaging than it would otherwise duction for a few years (Buhler et al. 1997) phytes following the removal of benthic be’ (Crump et al. 1999). It can be broad- or through interference with dormancy barrier material. Journal of Aquatic Plant ly divided into two categories, classical or germination requirements (Bhowmik Management 33, 51-4. biological control, where an organism is 1997), or can increase rapidly if plants are Elmore, C.L. (1996). A reintroduction to released into the environment to repro- allowed to produce seed. Being able to ac- integrated weed management. Weed duce and proliferate and, from there, to curately identify establishment thresholds Science 44, 409-12. infect, compete with or consume the tar- for aquatic weed species and implement- Flower, G.E. (2003). The biology and ecol- get organism, and inundative biological ing a method of control prior to experi- ogy of arrowhead (Sagittaria montevi- control, where the controlling organism is encing dispersal, is a key management densis Cham. et Schlecht), a weed in rice cultured and applied directly to the pest tool. in NSW. PhD thesis, Charles Sturt Uni- organism. versity, Wagga Wagga. Classical biological control is hampered Conclusion Gunasekera, L. and Krake, K. (2001). Ar- by the large amount of money required to Drought has a pronounced effect in alter- rowhead – a serious aquatic weed in implement it (Chokder 1967) and some- ing aquatic plant abundance and diver- northern Victoria. Victorian Landcare and times variable success rates. An example sity. It has a positive effect in reducing the Catchment Management 19, 7. of the successful implementation of classi- spread of weed species, as most species Szczepanski, A.J. (1977). Allelopathy as cal biological control is the introduction of rely on spread by water as the main vec- a means of biological control of water the Cactoblastis moth into Australia to con- tor for the spread of propagules and seed. weeds. Aquatic Botany 3, 193-7. trol prickly pear. Such unbridled success Drought has a negative effect in increas- stories are somewhat rare, however. ing the abundance of certain weed spe- As well as examples, like prickly pear, cies due to changes in management prac- of biological control using an organism tices. that eats or infects the pest species, in- troduced species may compete with the Acknowledgements pest plant for resources (allelospoly), or The author would like to acknowledge interfere with the pest species by releasing Roger Baker for his ‘on-ground’ knowl- compounds into the environment that act edge. upon the pest species, a process known as allelopathy (Szczepanski 1977). Literature References on allelopathy in aquatic plants is very Apfelbaum, S.I. (2001). Cattail (Typha spp.) limited, however, and effects are often management. www.appliedeco.com/ mistakenly attributed to this process. Projects/CattailManage.pdf, accessed The most successful method of inunda- 15th October, 2002. tive biological control for alismataceous Arthington, A.H. and Mitchell, D.S. (1986). species has been the mycoherbicide ap- Aquatic invading species. In ‘Ecology of proach, where a mycoherbicide is defi ned biological invasions’, eds R.H. Groves as ‘a fungal pathogen which, when applied and J.J. Burdon, pp. 34-53. (Cambridge inundatively, kills plants by causing a dis- University Press, London). ease’ (Crump et al. 1999). In the Australian Bhowmik, P.C. (1997). Weed biology: im- rice industry, most work has been done portance to weed management. Weed using the fungus, Rhynchosporium alismatis Science 45, 349-356. (Cother 1999), but other pathogens have Buhler, D.D., Hartzler, R.G. and Forcella, F. been investigated overseas (Chung et al. (1997). Implications of weed seedbank 1998). dynamics to weed management. Weed Science 45, 329-36. Issues Carter, D.R., Carter, S. and Allen, J.L. Drought has led to higher water values (1994). Submerged macrophyte control and lower returns for irrigation water sup- using plastic blankets. Water Science pliers due to the lack of sales water. Pro- Technology 29, 119-26. tection of irrigation assets, including the Chokder, A.H. (1967). Biological control of water reserves themselves, is a vital factor aquatic vegetation. Agriculture Pakistan in effective management of an irrigation 18, 225-9. system. Knowing and recognising which Chung, Y.R., Koo, S.J., Kim, H.T. and Cho, aquatic plant species pose the most risk K.Y. (1998). Potential of an indigenous to our waterways, early identifi cation and fungus, Plectosporium tabacinum, as a treatment, and constant surveillance of our mycoherbicide for control of arrow- waterways is required to prevent further head (Sagittaria trifolia). Plant Disease infestations of new aquatic weed species. 82, 657-60.

Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 29 Predicting the potential distribution of orange hawkweed on the Bogong High Plains using a dispersal constrained habit suitability model

N.S.G. WilliamsA, A.K. HahsB and J.W. MorganC A School of Resource Management, Faculty of Land and Food Resources, The University of Melbourne, 500 Yarra Boulevard, Richmond, Victoria 3121, Australia. B Australian Research Centre for Urban Ecology, Royal Botanic Gardens Melbourne c/- School of Botany, The University of Melbourne, Victoria 3010, Australia. C Department of Botany, La Trobe University, Bundoora, Victoria 3083, Australia.

Abstract Developing tools to predict the location of Native tussock grasslands and disturbed new biological invasions is essential if ex- areas had high probabilities of H. aurantia- otic species are to be controlled before they cum establishment. Extensive fi eld search- become widespread. Currently, alpine ar- es failed to detect new populations how- eas in Australia are largely free of exotic ever time-step validation, using 1998–2000 plant species but face increasing pressure locations, accurately predicted the occur- from invasive species due to global warm- rence of all post-2003 populations in areas ing and intensifi ed human use. To predict identifi ed as having a high probability of the potential spread of highly invasive or- occurrence. No H. aurantiacum was found ange hawkweed (Hieracium aurantiacum) in areas outside the predicted dispersal from existing founder populations on the plume. This suggests our model has good Bogong High Plains in southern Austral- predictive power and will improve the ia, we have developed a spatially-explicit ability to detect populations and prioritise dispersal-constrained habitat suitability areas for on-going surveillance. model. The model combines a habitat suit- This research (Williams et al. in press) ability index, developed from disturbance, has been accepted for publication in Eco- site wetness and vegetation community logical Applications and should be pub- parameters, with a phenomenological dis- lished in late 2007 or early 2008. persal kernel that uses wind direction and observed dispersal distances. After gen- Reference erating risk maps that defi ned the prob- Williams, N.S.G., Hahs, A.K. and Morgan, ability of invasion across the study area, J.W. (in press). A dispersal-constrained we intensively searched several locations habitat suitability model for predicting to validate the model. The highest prob- invasion of alpine vegetation by Hier- ability of H. aurantiacum establishment acium aurantiacum. Ecological Applica- was southeast from the initial infestations. tions.

30 Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 the stakeholders’ on-ground weed control programs (in the case of Municipal Coun- Measurement of changes in stakeholder weed cils, CMAs and Linear Managers). DPI had not previously worked with the Garden management attitudes and actions over the life of the and Fodder industries in great depth in Tackling Weeds on Private Land initiative relation to weed management and so this project worked at building relationships, Catriona M. King, Department of Primary Industries, PO Box 879, Seymour, raising awareness about the problem of spread of both declared weeds and non Victoria 3660, Australia. declared, invasive plants through their business activities, and on the co-develop- ment of strategies to reduce this risk. The aim of this research was to determine the Abstract The Tackling Weeds on Private railway authorities), the Garden industry effectiveness of the project by measuring Land initiative focused on developing ef- and the Fodder industry. It was felt that changes in stakeholder awareness and ac- fective partnerships with, and building it was a more effective approach to work tions in terms of their weed management the capacity of key stakeholders whose with these groups and encourage them to responsibilities over the life of the project. actions directly impact on or infl uence the infl uence the communities or businesses management of weeds on private land. An within their sphere of operations, rather Methods adoption model was used by the project than spread the project’s resources thinly Adoption model both to develop engagement strategies by trying to work at the community lev- The project team developed a model of and to measure the impact and effective- el. Some project activities, such as media adoption which recognised that adoption ness of these strategies over the life of the campaigns and compliance operations, of new or different practices is a staged project. The model is based on stakehold- were however directed at the community process. The model used by the initiative ers progressing along a continuum of level. (Figure 1), describes a progression along ‘awareness’ and ‘acceptance’ to ‘action’. The project delivered a variety of strat- a cycle of ‘awareness’, and ‘acceptance’ to Both quantitative and qualitative methods egies aimed at raising awareness of the ‘action’. The model also maps some of the were used to analyse shifts over time for stakeholders’ weed management respon- elements of a Bennett’s hierarchy (Bennett fi ve different stakeholder groups (CMAs, sibilities and encouraging them to take 1977); Knowledge, Attitudes, Skills, Aspi- Municipal Council, Linear Reserve man- action. These strategies ranged from ex- rations and Practice Change, in relation to agers, the Garden industry and the Fod- tension, through to incentives and the de- the Aware, Accept and Act segments of the der industry). Data was collected and ana- velopment of partnerships between gov- adoption cycle. lysed at the beginning (to provide a base- ernment and stakeholders, and amongst line), mid and end of project. Municipal the stakeholders themselves. These part- Domains of change – what does success Councils and the Garden industry showed nerships jointly developed innovative look like? the greatest level of progression along the approaches to weed management which In order to measure the effectiveness of continuum from the baseline to the end of were focused on sustainability beyond the project, it was necessary to define the three year project. the life of the three year initiative. Project what success looked like, or what changes activities were designed to be strategic we wanted the stakeholders to make, in Introduction with a long term focus to complement order to act on their weed management Tackling Weeds on Private Land (TWoPL) was a three year, $9 million state-wide initiative, delivered by the Department of Primary Industries (DPI) on behalf of External Context the Department of Sustainability and En- vironment (DSE), Victoria. It was aimed at ensuring that government and the community share responsibility for weed ATTITUDE management and private land managers, agencies and industries are capable and KNOWLEDGE empowered to act to decrease the impact AWARE of weeds on the State’s social, economic •Know •Seen ACCEPT and environmental values. The focus of •Heard •Think the initiative over these three years has •Recall •Feel been on developing and maintaining ef- •Believe •Say

fective partnerships with key stakehold- ACT SKILLS ers whose actions and activities impact CHANGE •Plan on or infl uence the management of weeds •Invest on private land. The initiative recognised •Do that if we are to be successful in tackling •Tell weeds, a coordinated approach is required PRACTICE to engage the many land custodians in- ASPIRE cluding State Government, Municipal Councils, private landowners, industry and the community. The project worked with five key stakeholder groups: Municipal Coun- cils, Catchment Management Authorities (CMAs), Linear Managers (VicRoads and Figure 1. Aware, Accept, Act Adoption model.

Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 31 responsibilities. Research questions were operational representative (e.g. Environ- Additional quantitative and qualitative then formulated to measure the degree ment/NRM Offi cer) of each organisation. survey questions were designed to pro- of change for each stakeholder in each Samples for the end of project survey are vide information about the stakeholders’ domain, over the life of the project. The shown in Table 2. attitudes, motivators and barriers towards domains of change were defi ned to be as their weed management responsibilities generic as possible for each stakeholder Data analysis and the information from these responses group, but some differences were neces- A research methodology was designed was used in conjunction with the dichoto- sary for the agency groups and industry which would enable objective and quan- mous ones to provide a richer picture. The groups (Table 1). tifi able measurement of change over time, focus of this paper is on the quantifi cation of the stakeholders’ awareness, acceptance of the stakeholders’ changes in the Aware, Data collection and actions in relation to their weed man- Accept and Act segments of the adoption Data was collected at the beginning of the agement responsibilities. model. project to provide a baseline from which to A range of dichotomous (Yes/No) sur- measure shifts in each of the domains; at vey questions were designed and grouped Results the mid point and at the end of the project. according to the Domain of Change and The following series of tables show the re- Surveys were designed to collect both whether they belonged to the Aware, Ac- sults for each stakeholder at the beginning quantitative and qualitative information cept or Act segment of the adoption model. of the project (baseline) and at the end of and were administered by telephone. The This is described in Table 3 for the agency the project. For simplicity, the results col- same survey instrument was used for each group of stakeholders. lected at the mid point of the project are of the agencies (CMAs, Municipal Coun- For each of the cells in Table 3, the re- not shown. These mid project results, to- cils and Linear Managers) and modifi ed sults were expressed as the proportion of gether with the qualitative information ones were designed for the Garden and respondents answering ‘Yes’ out of the to- were used to inform the delivery of the Fodder industry representatives. Survey tal number of respondents who answered remainder of the project activities and samples were kept as uniform as possible those questions. Statistical signifi cance of modify strategies accordingly. for the three surveys, however there was the changes between the baseline and end some variation due to staff turnover. In of project results was tested using Chi2 Catchment Management Authorities the case of the agencies, surveys were con- tests and P values of <0.05 or lower were Overall, CMAs showed an increase in ducted with a senior management and an considered signifi cant. awareness for the fi ve domains for change combined, but no signifi cant increases in the Accept and Act segments of the adop- Table 1. Domains of change. tion model. The improvements in aware- Agencies - CMAs, Councils, Linear Industries – Garden and Fodder ness were mainly attributable to improve- Managers ments in the Managing priority weeds, Prevention of spread and Working with 1. Manage priority weeds on land they are 1. Not selling or promoting declared weed other land managers domains. Results for responsible for species, or products containing weeds each individual domain which contribute 2. Prevent the spread of weeds resulting 2. Prevent the spread of weeds resulting to this fi nding are as follows: from their business operations from their business operations Manage priority weeds on their land 3. Report new weed species 3. Report new weed species CMAs showed a signifi cant increase (P <0.01) in awareness of their responsibili- 4. Work with other land managers 4. Work with other industry members ties of managing weeds, the priority spe- in a coordinated approach to weed in a coordinated approach to weed cies and their locations within their catch- management management ments. There was a slight decrease in the 5. Show leadership in weed managementA Accept segment which relates to the or-

A ganisation setting goals for weed manage- CMAs and Councils only. ment improvement, and there was no sig- nifi cant change in the Act segment which

Table 2. Survey samples. Table 3. Survey design. Stakeholder group Number of Domain of change Aware Accept Act respondents 1. Manage priority weeds on land 3 questions 2 questions 4 questions CMAs 16 they are responsible for (agencies).

Municipal Councils 38 2. Prevent the spread of weeds 2 questions 2 questions 3 questions resulting from their business A Linear Managers 4 operations

Garden industry 250 3. Report new weed species 2 questions 2 questions 1 question

Fodder industry 150 4. Work with other land managers 1 questions 3 questions 1 question in a coordinated approach to weed A Eight Linear managers were interviewed management in the baseline survey but numbers were not able to be sustained at the end 5. Show leadership in weed No questions No questions 2 questions of project survey. As the sample size is management so small, results for this group are not included in this paper. Total all Domains 8 questions 9 questions 11 questions

32 Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 incorporates issues such as allocating spe- locations within their shires. There was Report new weed species There were sig- cifi c staff and funds to weed management, however, a signifi cant decrease (P <0.01) nifi cant increases in awareness (P <0.01), having scheduled works and staff training in the Accept segment, and no change in acceptance (P <0.001) and actions (P <0.001) programs in place. the Act segment. Indicators for these seg- of Councils in terms of understanding the ments were the same as previously de- risk of new weed species, implementing Prevent the spread of weeds There was a scribed for CMAs. procedures to minimise this risk, and re- signifi cant increase (P <0.01) in awareness porting incidences of these species. of the ways in which CMAs could spread Prevent the spread of weeds There weeds during their works programmes were signifi cant increases in awareness (P Work with other land managers Sig- and how they could reduce this risk (for <0.01) of the ways in which Councils could nifi cantly more Councils were aware (P example through vehicle hygiene). There spread weeds during their works pro- <0.001) of other land managers and agen- was also a signifi cant increase (P <0.05) grammes and how they could reduce this cies and their responsibilities for weed in the CMA respondents whose organisa- risk. There were no signifi cant increases management and had also initiated col- tions had Codes of Practice in place with in the Accept and Act indicators (as previ- laborations (Act, P <0.05). procedures and systems to support them. ously described for CMAs). These were indicators used for the Accept segment. In terms of the Act segment, there was no signifi cant change in the pro- portion of CMAs which inserted weed hy- Table 4. Catchment Management Authorities. giene clauses in their contracts (e.g. with % of responses works contractors) and provided facilities Aware Accept Act and equipment, and time for staff to im- plement vehicle wash-down procedures. Baseline End Baseline End Baseline End Manage priority 67 90 61 56 40 42 Report new weed species There were no weeds on their land signifi cant changes in awareness of new P value <0.01 n.s n.s. high risk weed species, risk areas and Prevent the spread of 61 94 19 53 19 23 potential pathways of introduction and weeds spread; in acceptance as measured by the P value <0.01 <0.01 n.s. number of CMAs actively distributing in- Report new weed formation to staff about these new species, 75 84 31 50 61 75 and also implementing a reporting pro- species cedure; and in actions in terms of actual P value n.s. n.s n.s reports of new weed species. Work with other 44 100 87 85 83 69 land managers Work with other land managers Signifi - P value <0.001 n.s n.s cantly more CMAs were aware (P <0.001) Show leadership in ––––7881 of other land managers and agencies and weed management their responsibilities for weed manage- P value – – n.s ment. However, there was a decrease in those who were actively working with Total – All Domains 65 91 54 64 47 49 others and with resources committed to P value <0.001 n.s n.s these collaborative activities (Accept in- dicators). Similarly, fewer CMAs had ac- tually initiated collaborations since the baseline survey (Act indictor). Table 5. Municipal Councils. % of Responses Leadership in weed management There Aware Accept Act was no signifi cant increase in the propor- Baseline End Baseline End Baseline End tion of CMAs saying that their organisa- Manage priority tion was the lead agency for a collabora- 79 81 54 34 58 55 tive project or an active advocate for weed weeds on their land management (Table 4). P value n.s <0.01 n.s Prevent the spread 90 100 41 53 43 54 Municipal Councils of weeds Overall, Councils showed signifi cant in- P value <0.01 n.s n.s creases in awareness, acceptance and Report new weed actions for the fi ve domains of change 56 79 24 62 40 87 species combined. These were mainly attribut- able to improvements in the Prevention of P value <0.01 <0.001 <0.001 Work with other spread, Reporting new weeds and Work- 38 100 70 76 55 79 ing with other land managers domains, as land managers described below. P value <0.001 n.s <0.05 Show leadership in ––––5367 Manage priority weeds on their land weed management There was no signifi cant increase in the P value –––– n.s proportions of Councils demonstrating awareness of their responsibilities of man- Total – All Domains 71 88 50 58 51 62 aging weeds, the priority species and their P value <0.001 <0.05 <0.01

Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 33 Leadership in weed management There which might be a source of weeds. and the Acting indicators were whether was no signifi cant increase in the propor- There was a significant increase in the respondents were currently participat- tion of Councils saying that their organisa- awareness (P <0.05) of the risks from in- ing in such collaboration and contributing tion was the lead agency for a collabora- vasive garden plants, garden supplies to industry discussions about current and tive project or an active advocate for weed and landscaping works. Acceptance was future problem species (Table 6). management (Table 5). measured by indicators about educating customers about invasive plants, imple- Fodder industry Garden industry menting procedures to ensure that sup- Fodder industry respondents reported a Garden industry respondents showed plies were weed free, and implementing signifi cant increase in awareness, but not significant increases in the Awareness, weed hygiene procedures for landscap- acceptance or action for all domains com- Acceptance and Action segments of the ers. This indicator had not increased sig- bined. The signifi cant awareness increase adoption model for all four domains of nifi cantly since the baseline. There was a is attributable only to the Preventing the change combined. Signifi cant improve- signifi cant increase in the Act segment (P spread of weeds domain. ments in awareness were attributable <0.01) which was measured by indicators to all domains except for Working with about suggesting alternative, non invasive Not selling or promoting weeds There other industry members; signifi cant im- plants to customers, requiring weed free was a slight but not signifi cant increase in provements in acceptance were seen only guarantees on garden supplies and insert- the indicators contributing to the Aware in the Reporting new weeds domain and ing machinery hygiene clauses into con- rating; knowledge of the provision of the signifi cant improvements in actions were tracts for landscapers. CaLP Act 1994 that it is illegal to sell de- seen only in the Prevention of spread of clared weeds, or offer them for sale as part weeds domain. Report new weeds There was a signifi - of another product (i.e. weed contaminat- cant increase (P <0.001) in awareness of ed fodder), understanding the difference Not selling or promoting weeds There the species of invasive garden plants of between declared and non declared weeds was a signifi cant increase (P <0.001) in concern, the ways in which they might be- and having access to a declared weed list. garden industry respondents being aware come naturalised, and appropriate report- There was a slight decrease in both the of the declared noxious weed list, having ing procedures. There was also a signifi - Accept and Act ratings. Accept indicators access to the list and in their knowledge of cant increase (P <0.05) in acceptance which were related to use of Codes of Practice which species are on the list (Aware indi- was measured by respondents’ perceived which specifi cally address the presence of cators). There was no signifi cant change responsibility for identifying and report- weeds in the fodder, and the Act indicators in the Accept indicator which was that ing these new weed species and in ensur- which concerned respondents who had ex- businesses provided training to their staff ing that their staff are also trained to be perience with selling fodder which they on declared weed identifi cation. Indica- able to do this. There was no change in the subsequently found out to be contaminated tors of Action for this domain were that proportion of respondents who had actu- with declared weeds and if they had taken the businesses had experience with selling ally reported these weeds (non declared steps to prevent this happening again. (or using, for landscapers) plants, which weeds); the Act indicator. they subsequently found out to be weeds Prevent the spread of weeds There were and if they had taken steps to prevent this Work with other industry mem- signifi cant increases both in awareness (P happening again. There was no signifi cant bers There were no significant differ- <0.001) and in acceptance (P <0.05) but increase in this indicator. ences in awareness, acceptance or actions no signifi cant increase in acting for this for this domain. The Aware indicator was domain. Awareness was measured by Prevent the spread of weeds For this do- knowledge of other members of the gar- the respondents’ rating of their knowl- main, indicators were developed which den industry that the respondents could edge of the ways in which weeds could described the spread of weeds in terms work with to minimise the sale or spread be spread during fodder production and of selling or using invasive plants (other of weeds. Accept was measured by the /or transport, and acceptance related to than declared weeds), and using garden respondents’ beliefs that it was both benefi - using a Code of Practice which addressed supplies (such as soils, mulches, rocks etc.) cial and practical to collaborate in this way, machinery and vehicle hygiene. The Act indicators were whether hygiene systems and processes were actually implemented Table 6. Garden industry. and complied with by respondents’ staff % of Responses and contractors. Aware Accept Act Report new weed species There were no Baseline End Baseline End Baseline End signifi cant changes in this domain. Aware- Not selling or 74 83 42 35 69 73 ness was measured by knowledge of new promoting weeds high risk weeds not yet known in Victoria; P value <0.001 n.s n.s acceptance measured by respondents’ rat- Prevent the spread ing of their responsibility for identifying 91 94 59 64 52 60 of weeds and reporting these new weeds, and act- P value <0.05 n.s <0.01 ing was whether they had actually report- ed any of these species. Report new weed 57 68 60 67 8 7 species Work with other industry members As P value <0.001 <0.05 n.s for the garden industry, this domain con- Work with other cerned collaborative activities to minimise 49 56 84 89 21 26 industry members the spread of weeds. The Accept rating was P value n.s n.s n.s signifi cantly higher (P <0.05) due to an in- crease in respondents believing that it was Total – All Domains 70 77 64 68 40 46 both benefi cial and feasible to collaborate P value <0.001 <0.05 <0.001 with other industry members, however

34 Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 this was not carried through into action, The Aware, Accept, Act adoption model solely on the quantitative data presented with a signifi cant decrease (P <0.05) in re- proved to be an effective basis for design- in this paper is that it does not take into spondents who were currently participat- ing a quantifi able method of measuring consideration contextual factors operating ing in collaborative activities (Table 7). behavioural and attitudinal shifts over during the period of the research. These time. Caution however, must be exercised factors are critical in order to determine Discussion in relying solely on these measures to de- whether changes (improvements or other- The following table (Table 8) compares termine the effectiveness of an initiative wise) to the Aware, Accept and Act ratings significant improvements in different such as Tackling Weeds on Private Land. can be attributed to the project interven- domains of change for each stakeholder. For the stakeholders with smaller sample tions alone or whether external factors Municipal Councils have progressed to sizes, such as CMAs and Municipal Coun- are having an impact. For example, the the Act stage of the adoption model in the cils, the qualitative fi ndings (not presented serious state-wide drought and bushfi res domains of Reporting new weed species in this paper) created a far richer picture of in some regions negatively impacted on and Working with other changes and impacts and provided greater most stakeholders’ abilities to implement Agencies and the Garden industry are insights into causalities than the quanti- as much weed management activity as at the Act stage for preventing the spread tative results alone. For the Garden and they might otherwise have done. of weeds. The least amount of progression Fodder industries, where the sample sizes When the qualitative information, as across all stakeholder groups is evident were larger, the quantitative results were well as additional evaluation data col- for the fi rst domain (Managing priority more reliable, but the qualitative informa- lected throughout the life of the initiative weeds/ not selling or promoting weeds). tion was also invaluable in providing in- is brought to bear on interpreting these The Fodder industry shows the least de- sights into causalities and reasons behind results, greater insights can be gained for gree of progression. the numerical results. A risk of relying each stakeholder and are summarised be- low:

Catchment Management Authorities Table 7. Fodder industry. While signifi cant gains have been seen in % of Responses several areas of weed management, the Aware Accept Act CMAs’ statutory position of leadership for Baseline End Baseline End Baseline End natural resource management within Vic- toria might have been expected to lead to Not selling or 77 81 51 48 91 86 even greater progression along the adop- promoting weeds tion cycle. This result however, might be P value n.s n.s n.s explained by the feedback that CMAs did Prevent the spread not perceive themselves to be land man- 38 61 48 70 70 73 of weeds agers with direct weed management re- P value <0.001 <0.05 n.s sponsibilities per se, but rather to hold a Report new weed more strategic, policy setting role. Engage- 52 55 91 88 9 11 species ment strategies delivered by the project were designed mainly towards building P value n.s n.s n.s awareness, particularly of CMAs’ roles Work with other 50 67 89 91 46 37 and responsibilities with regard to weed industry members management. There were also some en- P value n.s <0.05 <0.05 gagement strategies designed at gaining Total – All Domains 61 69 78 79 56 53 acceptance and action, particularly in re- P value <0.001 n.s n.s lation to minimising weed spread during CMAs works programmes (such as vehi- cle hygiene practices), however the project had limited success engaging CMAs on Table 8. Comparison of signifi cant improvements in domains of change by these issues. stakeholder groups. Municipal Councils Municipal Garden Fodder Signifi cant improvements were seen in CMAs Councils industry industry nearly all areas measured for Municipal Councils. A large portion of the TWoPL Manage priority weeds on Aware – Aware – initiative’s resources was allocated to this their land / Not selling or stakeholder group, with an incentive pro- promoting weeds gram, educational resources and a series Prevent the spread of weeds Aware Aware Aware Aware of information and networking forums Accept Accept amongst the key engagement strategies. Act Additional data was collected which Report new weed species – Aware Aware – showed the degree of impact of each of the Accept Accept strategies on Municipal Councils’ attitudes Act and actions towards their weed manage- ment responsibilities. This data supported Work with other agencies Aware Aware – a causal link between the TWoPL engage- /industry members Accept ment and the signifi cant improvements in Act the Aware, Accept and Act ratings over All domains combined Aware Aware Aware Aware the life of the project. In terms of under- Accept Accept standing the external factors, the quali- Act Act tative information showed a consistent

Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 35 concern about an issue of confusion over industry peak body expressed strong industry representatives recognise the the legal responsibility for managing concern about the possibility of negative competitive advantage of ensuring weed weeds on council managed roadsides. publicity for the industry if any exhibi- free products and implementing hygiene Municipal Councils expressed high levels tors were seen to be prosecuted at such practices to reduce weed spread during of frustration over the length of time that a showcase event. This led to the compli- their operations. However, they also pro- it was taking State government to resolve ance activity being undertaken in a very vide strong evidence that implementing the issue and many perceived the likely discreet manner, with offi cers conducting stringent weed hygiene practices is fre- outcome to be another example of cost inspections out of opening hours and in quently not cost effective or practical, par- shifting from State to Local Government. It plain clothes. By the third year, the rela- ticularly for the smaller industry players. can only be hypothesised that even greater tionship between this peak body and the In addition, these fi ndings clearly confi rm gains may have been made in the absence TWoPL team had improved to the extent that market forces dictate a far higher tol- of this cause of tension between the project that the peak body requested the compli- erance amongst consumers for weed con- sponsors and this stakeholder group. ance offi cers to attend during public open taminated fodder during times of product hours, and in uniform. In this way, the in- scarcity, such as the recent drought. Garden industry dustry was seen to be actively supporting As for Municipal Councils, the signifi cant the message to the public about the threat Limitations improvements seen in the Garden indus- of introductions of invasive plants. The results presented in this paper rep- try can also be attributed, at least in part resent the stakeholders’ self reporting. In to the TWoPL initiative through analysis Fodder industry the full evaluation of the TWoPL initia- of respondents’ ratings of the impacts of Active engagement with the Fodder in- tive, other metrics of project success such specifi c engagement strategies. In addi- dustry by the TWoPL project only began as numbers and types of projects initiated, tion, however, understanding the social approximately 12 months prior to the fi - partnerships developed etc., have been context is important for interpretation of nal survey. In addition, engagement was used to ‘ground truth’ this self reporting these fi ndings. Qualitative information mainly with two key industry peak bod- data. showed that Garden industry members ies so additional time is required for the It was diffi cult to design survey ques- have very high ratings for their industry’s messages to reach the 13,000 Fodder in- tions to represent Accept indicators. As a responsibility for not contributing to weed dustry players in Victoria. Understanding result, there was often a non linear rela- spread through their business activities this context helps to interpret the Aware, tionship from Aware to Accept to Act, as and a high personal motivation for caring Accept, Act ratings and the fact that there one might expect from the adoption mod- for the environment. The change required was limited progression along the adop- el. Greater importance is therefore placed of this industry, if managed carefully, need tion cycle for this stakeholder during on the Aware and Act indicators not have an impact on their profi tability. the life of this project. An innovation or For example, a key project with this stake- practice change in the agricultural sec- Acknowledgements holder was the development of a brochure tor is more likely to be adopted when it Most of the data collection for this paper entitled ‘Grow me Instead’ which provid- has a high ‘relative advantage’ over the was conducted by Down to Earth Research ed safe alternative species to correspond- product or practice it supersedes (Pan- Pty Ltd, 10 Caladenia Circuit, Frankston ing invasive garden plants. This substitu- nell et al. 2006). In addition, a research South, Victoria 3199. The Tackling Weeds tion, given suffi cient lead time was per- study indicated that time lags exceeding on Private Land initiative was funded by ceived not to negatively affect profi tability. one year existed between the time when the Department of Sustainability and En- Indeed, competitive advantage could be the fi rst farmer discovered an innovation vironment, Victoria. conferred on businesses portraying the and the time when 50% of farmers in the ‘sustainability’ message. Research at the study area had discovered it (Gibbs et al. References beginning of the project indicated that a 1987). This is only the time it takes to be Bennett, C.F. (1977). Analysing impacts of barrier for practice change adoption with aware of the innovation, not to adopt it. extension, programs. Extension Service, this industry was a lack of trust of State A further study indicated that the adop- US Department of Agriculture, Wash- government and a concern that signifi cant tion time from delivery of an extension ington, D.C. numbers of new species would be declared package to reaching the adoption ceiling Gibbs, M., Lindner, R.K. and Fischer, A.J. under the CaLP Act 1994 very quickly with (when the rate of adoption plateaus) was (1987). The discovery of innovations no consultation or lead time to allow busi- seven to eight years (Marsh et al. 2000). by farmers. The Journal of the Austral- nesses to prepare. Some representatives These studies relate to the adoption of a ian Institute of Agricultural Science 53 (4), of the Garden industry also expressed new farming production system, such as a 254-61. concern that Government and other or- new crop variety, with an improvement in Marsh, S.P., Pannell, D.J. and Lindner, ganisations appeared to focus mainly on production and/or profi tability being the R.K. (2000). The impact of extension on the introduction of weeds through their main driver for adoption. In the case of adoption and diffusion of lupins as a industry compared with other vectors of adoption of natural resource management new crop in Western Australia. Austral- weed introduction and spread (e.g. for practice change, it is also known that the ian Journal of Experimental Agriculture agricultural weeds). Understanding this rate of adoption is slow and non-adop- 40, 571-83. context allowed the TWoPL project to tion is relatively high (Pannell et al. 2006). National Land and Water Resources Audit design engagement strategies targeted at Environmental attitudes generally show a (2002). Australian farmers – relating to building relationships with the industry limited relationship with natural resource natural resource management in Aus- peak bodies to increase trust, and then to management practice change, with fi nan- tralians and Natural Resource Manage- jointly develop strategies to address the cial risk and management skills being far ment 2002. weed spread risk. One anecdote alone more powerful infl uencers of adoption Pannell, D.J., Marshall, G.R., Barr, N., indicates the success of this approach. (National Land and Water Resources Au- Curtis, A., Vanclay, F. and Wilkinson, R. Compliance staff conducted audits for dit (2002)). It is not surprising therefore, (2006). Understanding and promoting declared and state prohibited weeds at that we have not yet seen greater adoption adoption of conservation practices by three annual Melbourne International of weed spread minimisation practices by rural landholders. Australian Journal of Flower and Garden shows. Before the fi rst the Fodder industry. The qualitative fi nd- Experimental Agriculture 46 (11), 1407- year’s compliance activity, the Garden ings of this study do indicate that some 24.

36 Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 Materials and methods The fi ve key stakeholders comprised Lo- Tackling weeds on private land cal Government (62 Municipal Councils), Catchment Management Authorities (ten Helen AndersonA, Mark FarrerB, Mark WattC, Ugo MantelliD, Catriona KingE CMAs), Linear Reserve Managers (VicRo- and Tony LovickF ads and four rail companies), the Garden A Department of Primary Industries, Private Bag 105, Hamilton, Victoria 3380, Industry, and the Fodder Industry. Stake- holder analysis was undertaken for each Australia. of the fi ve stakeholders to gain an under- B Department of Primary Industries, 23 Patrick Street, Stawell, Victoria 3380, standing of organisation/industry struc- Australia. ture; operational systems and networks; C Department of Primary Industries, cnr Mair and Doveton Streets, Ballarat, decision-making processes and key per- Victoria 3350, Australia. sonnel; inter and intra-organisational de- D Department of Primary Industries, PO Box 4166, Parkville, Victoria 3052, pendencies; size of organisation/industry Australia. and major organisations/industry infl u- E Department of Primary Industries, PO Box 879, Seymour, Victoria 3660, ences (Mantelli 2005). Australia. Eight project objectives linked to the F project goal of ‘key stakeholders accept- Department of Primary Industries, 89 Sydney Road, Benalla, Victoria 3672, ing and acting on their weed management Australia. responsibilities in a collaborative manner’. Implementation of the project objectives was achieved through four subprojects; Partnerships; Prevention and early inter- Summary The Victorian Government ini- vention; Enhanced enforcement; and En- tiative ‘Tackling Weeds on Private Land’ Sixty percent of Victoria’s land is pri- gagement. The Partnerships subproject was a three-year initiative that sought to vately owned and all of it requires coor- supported key stakeholders to: gain a encourage landowners to work collabora- dinated weed management. There are better understanding of their roles and tively to manage weeds. many examples of successful community responsibilities; actively integrate weed The $9 million initiative accelerated partnerships tackling weeds and success- management into their programs; and in- implementation of the Government’s Vic- ful key stakeholder policies and programs crease collaborative action on weeds. The torian Pest Management Framework – A for weed management. The Government Prevention and early intervention sub- Framework for Action. recognises however, that more can be done project established a process to prioritise The project worked with fi ve key stake- to coordinate the range of stakeholders new high-risk weed species in Victoria for holders whose activities signifi cantly in- involved in weed management. Coordi- improved surveillance, and aimed to im- fl uence how weeds impact upon and are nated approaches to weed control will prove stakeholder adoption of prevention managed on private land; Local Govern- encourage all land managers and organi- of spread principles (with particular em- ment, Catchment Management Authori- sations to play their part in reducing the phasis on pathways) and hygiene systems ties, Linear Reserve Managers, the Garden impact of weeds on social, environmental and practices. The Enhanced enforcement Industry and the Fodder Industry. and economic assets. subproject provided increased support TWoPL built on the momentum of The Tackling Weeds on Private Land to stakeholders and community priority a range of existing programs by build- (TWoPL) initiative was launched in Oc- weed control programs through targeted ing the profi le of weed management as a tober 2004 as a policy initiative under the compliance programs. The Engagement core component of business operations. Victorian Government’s Victorian Pest subproject supported delivery of the other Through partnerships with key stakehold- Management – A Framework for Action subprojects through expertise in commu- ers, a range of new initiatives promote (VPMF). The $9 million, three-year ini- nication, social research and evaluation. weed management responsibilities, share tiative sought to expand and promote in- knowledge, build capacity and ensure novative and collaborative approaches to Changes in stakeholder awareness, integrated and cooperative approaches reducing the impacts of weeds on private acceptance and action to preventing new weeds and managing land. Victoria’s Department of Primary In- Adoption of new or different practices is existing problems. Significant practice dustries (DPI) delivered key components a staged process. The model used by the change in weed management responsibili- of the initiative on behalf of the Depart- initiative, is a progression along a con- ties was achieved by all fi ve stakeholders ment of Sustainability and Environment tinuum of ‘awareness’ to ‘acceptance’ to over the life of the project. (DSE). ‘action’ (AAA continuum). Baseline po- Between July 2004 and June 2007 TWo- sitioning of key stakeholders along the Introduction PL worked with five key stakeholders AAA continuum was established through The cost of weeds to Victorian agriculture whose activities significantly influence initial attitudinal research in 2005. Shifts is estimated at more than $900 million per how weeds impact upon and are man- along the continuum were assessed mid year (Sinden et al. 2004). This is a conserva- aged on private land; Local Government 2006 and at end of project, mid 2007 (King tive estimate as it does not include costs (MC), Catchment Management Authori- 2007a,b,c). Comparison of results with and losses associated with weed manage- ties (CMAs), Linear Reserve Managers, baseline data and with qualitative data ment in natural environments or with loss the Garden Industry and the Fodder In- was used to determine the effectiveness of amenity. Weed invasion is ranked sec- dustry. ‘Weeds are everyone’s property’ of the initiative in bringing about practice ond only to habitat loss as the major cause was the slogan for TWoPL. It recognises change. Evaluation took into account the of biodiversity decline (Sattler et al. 2002). that all land managers and Victorians have potential differences in positioning on Changing environmental conditions, in- a role or responsibility in weed manage- the AAA continuum between individuals creased knowledge and a rising number of ment, not only for the management of our within stakeholder organisations and the new weed incursions, demand the devel- existing weed problem, but also for the organisations themselves. opment of new and improved approaches prevention of new and emerging weeds. Four key areas of weed management to managing weeds. responsibility were defi ned. These were

Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 37 used to design relevant engagement strat- state or region need to be prioritised, so provide support to regional staff involved egies and to evaluate their effectiveness resources can be directed appropriately to- in supporting community groups and in increasing stakeholders’ awareness and ward prevention strategies. A risk assess- building the capacity of DPI to carry out action in each area. These were: ment framework was developed to estab- compliance under the Catchment and Land 1. Managing priority weeds on land the lish a quantifi able risk assessment of weed Protection Act 1994. .stakeholder is responsible for (agen- spread pathways in Victoria to inform cur- cies – CMAS, MCs, Linear Reserve rent weed management projects and en- Use of social research, evaluation and Managers) OR Not selling, promot- able objective prioritisation and strategic coordinated communication ing or transporting declared weeds or development of future weed management Social research was conducted to inform products containing declared weeds projects and initiatives. the development of the project’s engage- (garden and fodder industries). The risk assessment framework identi- ment strategies and communication mes- 2. Preventing the spread of weeds result- fi ed 10 potential activities or vectors im- sages and to evaluate their effectiveness. ing from the stakeholders’ business op- plicated in potential weed introduction Research was conducted in two stages. At erations. and spread in Victoria (Australia). A ma- the commencement of the project, studies 3. Reporting new potential weed species. trix was developed referencing the activi- were conducted to gain a better under- 4. Partnerships and collaborative weed ties with 28 industries and organisations standing of the size and complexity of the management activities with other agen- which may introduce and spread weeds stakeholders to assist in project scoping cies or industry members. by one or more of these activities. A set and the development of a broad engage- Social research, stakeholder analysis and of 15 criteria, each with fi ve associated ment and communication strategy. More partnership discussions were used to intensity ratings was developed to assess detailed research was then conducted to identify appropriate approaches. Engage- 98 organisation/activity combinations for collect both qualitative and quantitative ment activities included development of their risk of introducing and distribut- information on knowledge, attitudes, cur- printed and website extension material, ing 86 high priority weeds into Victoria. rent practices, motivators and barriers to media releases, regional forums, presenta- Confi dence scores, indicating the level of change relating to weed management. tions, interviews with senior executives, certainty and availability of data were also This information built on the scoping stud- partnerships project grants, technical provided. A decision support system, An- ies and together formed a detailed picture advice and stakeholder participation in alytic Hierarchy Process (AHP), was then of each stakeholder which was used to shaping policy. used to assist decision making on the rela- identify and target specifi c engagement TWoPL tailored engagement tools and tive weightings of each of the criteria. strategies. This research was particularly materials to the needs of the individual or- The risk assessment framework can valuable for understanding stakeholder ganisations to ensure that the project built be readily adapted to other regions and groups with whom DPI had not previ- on previous work undertaken. Oppor- pests. ously worked in depth (such as the gar- tunities for networking and information den and fodder industries), and enabled sharing have led to collaborative action Support for community led action relevant and targeted engagement and resulting in a more coordinated approach Demand from the community for action on communication strategies to be designed to weed management. weeds is high. The community weed mod- to achieve the necessary change for each el employed by groups in Victoria such as stakeholder group. Prioritising high risk weeds to the Victorian Serrated Tussock Working Key messages were developed and Victoria Party and Gorse Task Force, has allowed delivered through targeted communica- There are some 1300 weed species already key stakeholders to pool resources and tion materials as well as the mass media. in Victoria, with an estimated 40,000 po- knowledge, and work cooperatively and Effective project communications were tential weeds that could still enter Austral- strategically toward controlling weeds. designed with distinctive branding, iden- ia. A preliminary investigation was under- This collaborative approach has proved tifying the different stakeholders. All taken to identify the most likely candidate very effective in encouraging the major- communication strategies used consistent species posing a weed risk to Victoria. A ity of landowners to participate in weed project quality assurance and approval rapid assessment methodology was de- management. Stakeholders are aware processes. veloped (Blood 2005) to enable the can- that all land owners must participate for Evaluation was conducted throughout didate weeds to be prioritised to identify projects to be effective, as land managers the project, enabling continuous improve- the highest risk species to Victoria. Criteria that refuse to control weeds threaten the ment to the delivery of the project’s activi- included weed history, current recorded efforts and investment of the community ties as well as at the end of the project to distribution and potential threat. and impact on the natural or agricultural measure achievement of the project objec- environment. This generates signifi cant tives. Improvements in surveillance demand for government to provide com- Preventing the introduction and estab- pliance support to ensure the small per- Results lishment of serious new weeds to Victo- centage of land owners that refuse to meet The TWoPL initiative was a major con- ria is dependant on early detection and their obligations, does not jeopardise the tributor to implementation of the Weed eradication. Weed Spotters are a network success of these projects. Management Strategy of the Victorian Pest of people who keep an eye open for and The Enhanced enforcement component Management Framework – A Framework report potential new and emerging weeds. used a targeted mobile compliance ap- for Action. The initiative contributed to 16 Resource materials and training programs proach as an effi cient means to support of the 18 objectives of the Weed Manage- were developed to build the capacity of stakeholders tackling weed problems. ment Strategy, and 31 of the 50 strategic Weed Spotters to detect, identify and re- The approach, utilising a team of special- actions identifi ed to meet these objectives port Victorian Alert Weeds. ist compliance staff (approximately 4.5 full (Victorian Government 2002). Surveillance for Victorian Alert Weeds time equivalents), has been piloted across The three project annual reports (An- can be streamlined by focusing on the Victoria over the last three years to pro- derson et al. 2005, 2006, 2007) provide most likely ‘pathways’ these weeds might vide compliance support to community detailed information on TWoPL achieve- travel to enter Victoria. The methods by groups that are actively working toward ments including stakeholder advocacy of which weeds enter or spread (or pathways control of established weeds. The project benefi ts of weed management partner- of distribution) throughout the country, also utilised specialist compliance skills to ships in the media, industry and weed

38 Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 management forums. A brief summary Table 1. Practice change in awareness, acceptance and action on weed of TWoPL achievements is provided management by stakeholders by project end. below. CMAs Municipal Linear Garden Fodder Changes in stakeholder awareness Councils Managers acceptance and action Aware at end project 91% 88% 84% 77% 69% Increases in awareness and acceptance, % change 26% *** 17% *** 4% 7%*** 8% *** of the four weed management respon- sibilities noted above, were seen for all Accept at end project 64% 58% 50% 68% 79% fi ve stakeholders (Table 1), with statisti- % change 10% 8% * 6% 4% * 1% cally signifi cant increases in awareness achieved by CMAs (26%), Municipal Act at end project 49% 62% 57% 46% 53% Councils (17%), the Fodder Industry (8%) % change 2% 11% ** 5% 6% *** (3%) and Garden Industry (7%) and in accept- Statistically signifi cant change from beginning to end of project * P <0.05, **P <0.01, ***P <0.001 ance by Municipal Councils (8%) and the Sample sizes for the Linear reserve managers were too small for signifi cance testing. Garden Industry (4%). Sample sizes for Linear Reserve Managers were too small for statistical analysis. Increases in action on weed manage- participation in planning and running Prioritising high risk weeds to Victoria ment responsibilities were achieved by forums. The forums were designed to Some 1268 candidate species were iden- CMAs, Municipal Councils, Linear Re- showcase achievements, share informa- tified through initial assessment as a serve Managers, and the Garden Indus- tion, enhance networking and partnership potential risk to Victoria. Application of try. Municipal Councils and the Garden development, distribute extension materi- rapid assessment criteria to the candidate Industry showed the greatest level of im- al and allow the sharing of weed manage- weeds (Blood 2005) identifi ed 410 weeds provement with statistically signifi cant in- ment stories. Ten forums were held during with high potential threat to Victoria. creases of 11% and 6% respectively. the project with approximately 600 people These species are known as Victorian Alert The Fodder Industry was the only from nearly 270 organisations attending Weeds and included; stakeholder where a decline in action on these events. − 212 taxa not currently known in Victo- responsibilities occurred although this was Through engagement activities and ria (priority for prevention of introduc- not statistically signifi cant. This may have collaborative development, 20 signifi cant tion), been a consequence of the signifi cant pres- new weed management partnerships have − 141 taxa with less than 10 known distri- sure on the fodder industry imposed by been established. These partnerships take bution records (priority for eradication) the drought in combination with the fact the form of joint projects where resources, and that active engagement with this group risks and decision-making are shared. Oth- − 57 taxa with 11–20 known/unknown did not occur until the fi nal 12 months of er activities include; agreed participation distribution records. the project. in decision making committees, strategic The Victorian Weed Alert Plan describes planning, sharing information, business Victoria’s approach to prevention and Partnerships partnerships and commitment to regular eradication of new weed incursions. The To aid partnership development, innova- ongoing dialogue and collaboration. Weed Alert Plan has been updated to in- tion and capacity building in weed man- clude approaches for surveillance and re- agement a three year grants program was Communication sponse to Victorian Alert Weeds. established to provide opportunities for A scoping report on development of weed the key stakeholders to manage weeds management extension material (Snell Improvements in surveillance into the future. Fifty six organisations, and Norris 2005) identifi ed the needs of Thirty of the 410 Victorian Alert Weeds representing municipal councils, linear key stakeholders and priorities for fur- were selected to provide focus to the reserve managers, fodder and nursery in- ther development. Materials developed Victorian Weed Spotters Network. These dustry associations, implemented a total included 47 fact sheets incorporating key weeds were promoted to the Weed Spot- of 107 projects with $1.9 million funding messages, industry specifi c subjects and ters to encourage reporting of these new assistance and co-investment of $3.7 mil- technical information. Over the life of the and emerging species to improve surveil- lion. Projects implemented incorporated project, 25,344 fact sheet kits each contain- lance measures. A Weed Spotters website one or more of the following areas; com- ing up to seven fact sheets were sent to or- has been developed (www.dpi.vic.gov. munity education and awareness, weed ganisation/industry executives and staff au/weedspotters) as a one-stop informa- spread prevention training, weed map- members. tion and resources site for current Weed ping and marker systems, codes of prac- Media was utilised to raise awareness Spotters and those inquiring about the net- tice development, implementation of and communicate TWoPL achievements. work. The website includes a photo library weed hygiene protocols, local law devel- Sixty seven media articles were released by of the 30 priority taxa to assist identifi ca- opment, planning guideline creation and the project over the three years, resulting tion and enable improved surveillance. project evaluation. A case studies booklet in 237 published media articles. Analysis Weed Spotters are provided with a Ref- (Roche et al. 2007) highlights progress in of the extent to which the seven key mes- erence Manual and have the opportunity weed management achieved through 24 sages were reported in the media found to attend training workshops. Six train- projects implemented by Municipal Coun- that 67% reproduced all or more than half ing modules were developed including; cils, Linear Reserve Managers, the Garden of the related media release content. To- an introduction to Weed Spotters, plant and Fodder Industries. wards the end of the project there was an biology, recognising weed alert species, Engagement was also achieved through; increase in the number of articles on weeds hygiene health and safety, reporting weed high level meetings with Municipal Coun- produced by stakeholders themselves and alert species, and collecting and submit- cils to gain executive support for TWoPL in the number of unsolicited media en- ting plant specimens. objectives, feedback to key stakeholders quiries. The overall tone achieved in me- The results of the weed spread path- on the performance of their weed man- dia coverage was 89% positive, 8% neutral ways risk assessment are shown in Table agement programs and key stakeholder and 3% negative. (King 2007d). 2. The analysis produced a list of potential

Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 39 weed spread pathways ranked by relative Table 2. Top 10 Weed spread risk industries and pathways. risk score. A number of industries, (seed, Industry/Organisation Pathway Risk Rank aquarium, landscaping and nursery) and Rating specifi c pathways (deliberate introduction via business, deliberate introduction via Seed Deliberate introduction via business 0.894 1 community and contaminated equipment, Aquarium/pet shop Deliberate introduction via business 0.893 2 produce and vehicles) were amongst the Landscaping Deliberate introduction via business 0.827 3 highest risk for introducing these high pri- Public Deliberate introduction via community 0.819 4 ority weeds into Victoria (Thomas et al. Nursery Deliberate introduction via business 0.808 5 2007). Landscaping Contaminated vehicles 0.801 6 Support for community led action Aquarium/pet shop Contaminated goods/produce 0.788 7 Seven compliance projects were carried Earth moving Contaminated equipment 0.787 8 out over three years on a range of estab- Earth moving Contaminated vehicles 0.770 9 lished weeds in areas including the south Forestry Contaminated vehicles 0.766 10 west, north east, north central, central and eastern areas of Victoria. Species targeted Source: Thomas et al. 2007. were gorse, blackberry, serrated tussock and ragwort. The project also provided compliance support to regional projects TWoPL are outlined in the following dis- underestimated the complexity of devel- targeting other species on a smaller scale. cussion. oping an Exit Plan. During the project de- The seven projects targeted areas where The analysis underpinning TWoPL velopment phase, suffi cient time should a relatively high percentage of land own- practice change programs has been criti- be allocated in the project work break- ers had failed to meet their obligations cal to success of project delivery. Analysis down structure if it is intended to include for weed control in the past. Inspections of industry and organisation structure, a project Exit Plan. were carried out on 370 priority proper- networks, size and infl uence as well as Stakeholder groups were selected based ties covering 19,797 hectares. Notices un- attitudinal information on barriers and on the best available knowledge at the der the Catchment and Land Protection drivers of change has been critical for un- time, of each group’s relative risk for weed (CaLP) Act were issued to 314 land owners derstanding the current context and the spread and their ability to minimise the requiring them to take action on weeds. type and scale of change required through impact of weeds through successful part- The fi rst Priority Area Notice under the stakeholder engagement. Suffi cient time nerships and increased weed management CaLP Act, covering approximately 100 allocated to this type of analysis can pro- activities. It was recognised that there are properties, was also implemented un- vide a greater level of confi dence both to a number of other key organisations and der the project following Ministerial ap- the project investor and project manager industries, which are likely to have signifi - proval. Of the 27 land owners prosecuted that successful practice change activities cant potential for weed spread and there- in the Magistrate’s Court, 18 were issued can be delivered within the project man- fore might be important stakeholders for infringement notices (fi nes) and a further agement framework. future weed management projects. 14 land owners were issued offi cial warn- TWoPL experience in engaging with The weed spread pathways risk assess- ing letters for lesser breaches of the CaLP industry bodies and government agen- ment process was developed to provide a Act. cies has found that initial meetings and quantifi able assessment of the relative risk The compliance projects generated dialogue at the executive level of organi- of weed spread through the activities of signifi cant media interest for the groups sations and industry to elicit support for different industries and organisations. The involved, allowing promotion of their ef- intended engagement with their organi- results of the risk assessment have been forts and key messages. Positive feedback sation is critical to the success of practice used to inform the focus of subsequent was received from all groups supported, change projects. Victorian Government initiatives includ- along with requests to be involved in any The TWoPL partnership approach has ing the $6.2 million ‘Improving Provincial similar projects in the future. confi rmed that collaboration and partner- Victoria’s Biosecurity’ and $30.1 million ships between government, community ‘Weeds and Pests’ Initiative announced in Discussion and industry are essential to the success May 2007. The fi ve key stakeholders involved in TWo- of weed management programs. The ex- The identification of new, high risk PL have a signifi cant infl uence on success- perience of the TWoPL project has shown weed species to Victoria (Victorian Alert ful weed management in Victoria. They it is possible to develop partnerships and Weeds) required consideration of how have all contributed to weed management collaborative activities with organisations. to respond to the introduction of these to varying degrees in the past and believe Critical to the success of TWoPL partner- weeds. There is no legislative requirement weeds to be a major problem that requires ships has been establishing trust, respect to control Victorian Alert Weeds; therefore, collaboration and additional resourc- and dialogue with partners as well as any management action is entirely volun- ing. While key stakeholders were already agreeing on mutual benefi t and allowing tary. This situation requires the develop- aware, accepting or acting on their weed partners to participate in the shaping of ment of a ‘Voluntary Removal Concept’ management responsibilities at the start of policy. for responding to Victorian Alert Weed the project, participation in TWoPL has en- For initiative projects with a focus on introductions and incursions. The devel- hanced and built their capacity further. stakeholder engagement where produc- opment of this concept has been included Project learnings identifi ed throughout tive partnerships have been developed as a project deliverable in a new DPI initia- the project are described in detail in the and there is mutual benefi t to maintain tive called Improving Provincial Victoria’s three project annual reports (Anderson et those partnerships, there is merit in con- Biosecurity. al. 2005, 2006, 2007). Learnings were used sidering planning to strategically exit from Setting of targets for stakeholders’ pro- to inform continuous improvement of the such projects. Relationships between DPI gression along the ‘Triple A’ continuum project throughout its life with recom- and TWoPL key stakeholders are benefi t- of Aware – Accept – Act was based on the mendations incorporated into the follow- ing from the production and implementa- assumption that the stakeholder surveys ing year’s program. Key learnings from tion of the project Exit Plan. Project staff conducted at the beginning, mid point and

40 Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 end of the project would provide objective, Watt, Shane Bettess, Abby Stratton, Julie in Australia. CRC for Australian Weed quantifi able data with which to inform Graham, Jeff Wilkie, Paul Major, Andrew Management Technical Series No. 8. target setting (Baseline survey results) Staley, Brendan Roughead, Tony Lovick, (CRC for Australian Weed Manage- and to measure progress against these Donna Smithyman, Kylie Woods, Rob ment, Adelaide). targets (mid and end of project surveys). Walker, Julie McGeary, Ian Dreher, Kelly Snell, K. and Norris, C. (2005). Develop- The project team set targets based on their Snell, Simon Bonwick, Ben Matthews, ment of extension material for mini- estimate of the increase in the Aware, Ac- Melanie Martin, Nigel Thomas. Landscape mizing weed spread scooping report, cept and Act ratings as measured at the Protection Portfolio, Catchment and Agri- Tackling Weeds on Private Land Initia- Baseline survey which might be expected culture Services, Department of Primary tive (Department of Primary Industries, after delivery of the project activities. It Industries, Victoria. Melbourne). is likely that these targets were set too DSE Investors: John Burley, Anne Den- Thomas, N., Steel, J., King, C., Hunt, T. and high. nis, Andrew Hodges, Beth Jones. Pest Weiss, J. (2007). Weed spread pathway The Enhanced enforcement component Management Branch, Department of Sus- risk assessment Stage 2 – June 2007 (De- of the TWoPL project, which effectively tainability and Environment, Victoria. partment of Primary Industries, Victo- funded approximately 4.5 full time staff, ria). was responsible for producing almost References Victorian Government (2002). Victorian 23% of all DPI compliance actions on Anderson, H., Farrer, M., McGeary, J., Pest Management Framework: a frame- pests (briefs of evidence/ warning letters/ Smithyman, D. and Watt, M. (2005). An- work for action – Weed Management prosecutions/ infringements) over the nual Report 2004 – 2005, Tackling Weeds Strategy (Department of Natural Re- three year period. More importantly, the on Private Land Initiative (Department sources and Environment, Melbourne). subproject provided signifi cant support to of Primary Industries, Melbourne). active community groups showing lead- Anderson, H., Farrer, M., Watt, M., Joubert, ership in managing weed issues and as- D., King, C. and Mantelli, U. (2006). An- sisted in protecting the investment of gov- nual Report 2005 – 2006, Tackling Weeds ernment and the much larger number of on Private Land Initiative (Department landowners meeting their responsibilities of Primary Industries, Melbourne). for pest control. The targeted mobile com- Anderson, H., Farrer, M., Watt, M., Joubert, pliance approach has been shown to be an D., King, C. and Mantelli, U. (2007). An- effective tool for supporting stakeholders nual Report 2006 – 2007, Tackling Weeds that are actively working toward control on Private Land Initiative (Department of established weeds and also assisted in of Primary Industries, Melbourne). building the capacity of DPI to carry out Blood, K. (2005). Criteria for rapid assess- compliance under the CaLP Act. ment of the Top 30 Alert Weeds in Vic- TWoPL communications activity played toria, Tackling Weeds on Private Land a role in raising the Department’s reputa- Initiative (Department of Primary In- tion with industry groups, local govern- dustries, Melbourne). ments and, to some extent, with regional King, C. (2007a). End of Project Stakehold- communities. Communication materials er Attitudes Report; CMAs, Municipal and activities were seen as professional, Councils, Linear Managers Report high quality and well organised. Land- – June 2007 (Department of Primary In- scape Protection senior management and dustries, Victoria). the DSE investor were generally very sat- King, C. (2007b). End of Project Stakehold- isfi ed with the way in which they were er Attitudes Report: Garden Industry kept informed about TWoPL and found it June 2007 (Department of Primary In- useful for advocating the project to exter- dustries, Victoria). nal stakeholders. King, C. (2007c). End of Project Stakehold- Overall, TWoPL has progressed weed er Attitudes Report: Fodder Industry management in Victoria by building the June 2007. (Department of Primary In- capacity of key stakeholders to manage dustries, Victoria). weeds and encouraging and fostering King, C. (2007d). Communications strat- partnerships and collaboration that will egy evaluation June 2007 (Department last well beyond the life of the initiative. of Primary Industries, Melbourne). Mantelli, U. (2005). Stakeholder analysis Acknowledgements Stage 1 Report – June 2005, Tackling Project Partners: Nina Rogers, Municipal Weeds on Private Land Initiative (De- Association of Victoria; Robert Chin, Nurs- partment of Primary Industries, Mel- ery and Garden Industry Victoria; Mary bourne). Trigger, Sustainable Gardening Australia; Roche, K., Dodd, A., Farrer, M., King, C., Wayne Potter, Australian Rail Track Cor- Mantelli, U., McCarthy, M., Norris, C. poration; Matt Mooney VicRoads; Greg and Young, S. (2007). Tackling weeds Payne, Powercor; Len Smythe, Australian success stories: case studies from organ- Agricultural Contractors Association; Co- isations and industry 2004–2007. lin Peace and Duncan Hanley, Australian Sattler, P., Creighton, C., Lawson, L. and Fodder Industry Association. Tait, J. (2002). Australian terrestrial The DPI project team: Mark Farrer, Ste- biodiversity assessment 2002. (Natural ve Young, Claire Norris, Megan McCarthy, Heritage Trust Canberra). Kristy Roche, Aaron Dodd, Ugo Mantelli, Sinden, J., Jones, R., Hester, S., Odom, Catriona King, Neil Smith, Daniel Joubert, D., Kalisch, C., James, R. and Oscar, C. Kate Blood, Catherine McInerney, Mark (2004). The economic impact of weeds

Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 41 structure when the same nozzle and spray is used. These fi gures would change if dif- Innovations in spraying techniques, will they enable ferent nozzles were used. If droplet drift less herbicide to be used? had to minimised then a coarser higher velocity spray would be an option. This can be achieved by either using a narrower A B Harry Combellack and Jerzy A Zabkiewicz angle fl at fan nozzle, 65° to 80° rather than A Spray Smart Enterprises, 7 Michelle Drive, Maiden Gully, Victoria 3551, 110°, or higher fl ow rate nozzle 03 (12.5 m Australia. sec−1) rather than an 01 (5.5 m sec−1). The B Plant Protection Chemistry NZ, 49 Sala Street, Rotorua, New Zealand. narrower spray angle increases droplet size as well as droplet velocity as do high- er fl ow rate nozzles (see Table 4). Further droplet velocity can be increased by using selected adjuvants (Butler-Ellis et al. 1997, Abstract This review considers innova- when selecting nozzles and spraying sys- Holloway et al. 2000, Combellack 2004). tions in the delivery, retention and up- tems for specifi c applications. This has led Typically oil based adjuvants increase take of herbicide sprays. The complexity to an awareness that these have to be con- droplet size and velocity while non ionic of each, current knowledge limitations as sidered in relation to the spray formula- surfactants reduce droplet size and veloc- well as future requirements needed to in- tion, as droplet size can vary signifi cantly ity (Butler-Ellis et al. 1997, Miller personal crease overall effi cacy are considered. The with formulation when used with differ- communication, Table 4). A further option lack of understanding of plant structure ent nozzles or operating conditions (Butler is to use a twin fl uid or venturi nozzle as and how it affects form of deposit is poorly Ellis et al. 2001, Combellack 2004). How- both can generate low velocity coarse spray. recognised but is a key determinant of out- ever, although some generalisations can be This is illustrated by the mean velocity for come. Overall, satisfactory progress will made, we are still a long way from being a twin fl uid nozzle which was around 1.9 not be made until appropriate models are able to predict appropriate combinations m sec−1 (Miller et al. 1990) compared to developed for each process and integrated accurately. In particular, the infl uence of the 14.3 m sec −1 for coarse spray using a into a comprehensive agrichemical man- plant surface characteristics on retention 11008 fl at fan (Miller personal communica- agement system. It is surmised that im- and uptake of herbicides is still poorly un- tion, Table 4). The addition of a non-ionic plementation of such models will enable derstood and may still be the dominant surfactant will increase the droplet size of less herbicide to be used and a lowering of variable in some of the spray application/ twin fl uid nozzles but lower their veloc- adverse environmental contamination. formulation/effi cacy processes. ity due to the induction of bubbles of air Keywords: Spray application; formu- An illustration of possible overall her- into the droplets (Miller et al. 1990). Slower lation; effi cacy; deposition; retention; up- bicide effi ciency for early and late spray- moving droplets from such nozzles, par- take; translocation. ing of a crop is given in Table 1. These ticularly if generated with additional non estimates confi rm what was emphasised ionic surfactant, are better retained on Introduction previously (Combellack 1981) that the target surfaces covered in crystalline wax The challenges facing users of herbicides proportion of herbicide misplaced within (Miller et al. 1990). These examples show today have increased in complexity over the sprayed area is far higher than that that manipulation of droplet velocity and recent years. On the one hand consum- lost outside. The estimates also show that the physical characteristics of the spray ers require the highest quality of produce spraying later results in reduced misplace- can signifi cantly change the percentage of whilst regulators insist on minimal prod- ment within the sprayed area. This is an the spray retained. However, to date, little uct residues and risks to the operator and important consideration if contamination or no consideration has been given to this environment. Herbicide effi cacy can be of ground water is important, but as crop aspect and its infl uence on the percentage defi ned as the interactions of: delivery: in- yields benefi t from the early removal of of spray misplaced. terception: retention: uptake: translocation weeds, optimal timing is contentious. Such Studies on the effi ciency of the spray- and active ingredient activity (Zabkiewicz estimates are also crucial in demonstrating ing process have evolved from empiri- 2003). Risk can be reduced by not using where spraying effi ciencies can be made. cal tests and fi eld trials to more detailed specifi c herbicides or reducing the amount For example in the estimates presented, fundamental studies. Most attention has used (Combellack 1989, Haas 1989). Such the amount of spray intercepted to that re- been focused on herbicides as they ac- options can be implemented as demon- tained, varies with crop age due to chang- count for over half of all pesticides used strated by the use of integrated weed con- es in leaf area, phenology and surface (Underwood et al. 2001). The behaviour trol (Combellack 1989); the use of less per- sistent herbicides and lower frequency of herbicide use in cropping systems (Haas Table 1. Estimates of herbicide destination at two growth stages of wheat if 1989) and through the introduction of ge- −1 netically modifi ed crops. However these sprayed at 50 L ha . are neither without cost nor controversy. % of applied Regulatory requirements for herbicide Parameter GS 13 GS 39 use have increased but have been mainly Intercepted [% which hits the target] 10 40 aimed at reducing off-target effects. The requirement to increase biological effi cacy Retained [% intercepted less % refl ected plus % retained by non 525 and reduce detrimental environmental ef- target plants] fects within the sprayed area can only be Taken up [%which passes through cuticle of target] 2 10 met by improved spray application and Effective [% of total applied that affects effi cacy] 0.5 3 formulation technologies. They must be considered together because together they Misplaced within sprayed area 80 50 vary the proportion of product intercepted Lost outside sprayed area [mostly as droplets or vapour drift but 10 10 by and subsequently retained on the tar- can also be on soil or in water as run off or through soil profi le] get plants. There are recognised challenges

42 Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 and performance of spray formulations moving at 20 km h−1 (Stafford and Miller soil profi les into underground water and can be greatly infl uenced by the formu- 1993). Additionally decontamination of or contamination of bystanders. It has lants and or adjuvants, either by affecting the injection system can be diffi cult. Such been pointed out that physical application fl ow rate, droplet size, dynamic surface systems do have advantages in that con- and formulation factors need to be tailored tension, spreading and wetting of foliage, centrates can be returned to the container to specifi c applications (Miller et al. 2001). or the resultant form of deposit, infl uenc- if not used; if the weather during spraying In the case of ground applications, droplet ing uptake and translocation. becomes adverse the tank will not contain drift effects are typically in the order of a unused spray mixture; decontamination few metres, compared with the hundreds Delivery of the tank, the main pump and lines may of metres possible with aerial applications. There have been a number of changes not be required; However, this may in turn generate a in sprayer design to accommodate user The complex interaction of boom higher pesticide deposit per unit area, and demands. In Australia, average sprayer width, tank size, sprayer speed and ap- if close to water bodies, exceed permitted speed is now around 20 km h−1 and with a plication volume on the time it takes to levels or be a risk to soil and aquatic organ- few spraying at 40 km h−1; average boom empty a tank are given in Table 2. A user isms or hazard to bystanders (Mathers et width is now over 25 m and with many at should aim for a minimum of one hours al. 2006). One proposal is that spraying at 40 m; tank volume is mostly over 5000 L spraying time as often fi lling/ferrying can fi eld edges should be through nozzles that with a few at 10,000 L; and many sprayers be equally as long. A simple program is produce coarser droplets. This could again are now fi tted with twin booms to accom- required showing these interactions so as be semi-automated or under the control of modate a wider range of sprayer speed. to enable a person purchasing a sprayer the operator. Though this would improve These demands have meant that droplet to select appropriate tank size/sprayer deposition on-site, it may not do so on- drift is exacerbated because as sprayer speed/ boom width combinations on a target, as it is well known that there is less speeds increases detrainment of droplets, more assured basis. retention of larger drops by (hard to wet) hence drift, increases (Nilars 2002, Nuyt- Considerable attention has been given plant foliage. It is possible that a deliberate tens et al. 2006) in part because 110° noz- to the physical process of spray delivery change in formulation during application, zles remain the most popular when it has and its off-site component, ‘drift’. The ba- such as the addition of adjuvants, could been clearly shown that 80° nozzles gener- sic determinants are droplet size, droplet be a solution. ate much less droplet drift (Nilars 2002). velocity, release height and meteorological One further aspect of drift mitigation Wider booms mean that boom move- effects. Since fall distance is also greater that has been investigated experimentally ment, particularly in the horizontal plane, for wider booms, typically 600 to 700 mm, and theoretically, is the use of windbreaks can be worsened with increasing speed droplet evaporation can be an issue when or shelterbelts. These are common in or- leading to less uniform deposition due to spraying in summer, leading to potential- chard situations, but less so in other agri- the ends of the boom virtually stopping, ly more drift. Hence a simple selection of cultural sectors. However, their value as a thus increasing deposition, and then ac- nozzle type and droplet size may be inad- spray drift mitigating option is becoming celerating, hence lessening deposition. equate. In this case the addition of anti- recognised as it is found that spray drift While there are twin boom sprayers evaporant or anti-drift agent (both intend- can be reduced to less than a percent of that can have either or both spray lines ed to maintain or increase droplet size) has in-swathe deposition (Ucar and Hall 2001, delivery the solution there are few nozzle been exploited. A real-time meteorological Brown et al. 2004). However, although it combinations that can be used to maintain system could additionally prevent appli- is recognised that interception effi ciency the same spray quality when switching cation when the wind direction and or varies with shelterbelt material or dimen- from one line to the other or when using thermal air movement are inappropriate. sions, the factors contributing to this with both at the same time. In most instances Easy to say but diffi cult to implement or live shelterbelts is still poorly understood a doubling in fl ow is required before any justify economically. Perhaps the ultimate change can be effected and such a pres- solution is real-time interactive spray for- Adhesion, retention, distribution sure change varies the spray quality with mulation production, based on operating There have been a few studies of spray re- most hydraulic nozzles, but the user has conditions and location of the applicator tention by crops throughout their growing no knowledge of this. within the spray zone. This may sound season aimed at quantifying soil deposi- There have been no tests to determine esoteric, but smart systems such as these tion rather than spray effi cacy effects (van mixing effi ciency in large tanks, but it is have been suggested and are under con- de Zande et al. 2003). There have also been considered unlikely that dispersions of dry sideration for ground based application a number of other studies to measure spray fl owables are within the ±15% current EN- systems (Ganzelmeier 2005). retention by crops. For example the effect TAM [EN 12761-2:1.1.1.5 Mixing] require- Ground based application systems of changing droplet trajectory on reten- ment (Ucar et al. 2000). Recently direct have focused on nozzle selection and ap- tion by wheat and barley (Combellack and injection equipment has been introduced plication criteria, through a more prescrip- Richardson 1985); the effect of hydraulic in part to overcome the challenge of even tive approach, than through optimisation nozzle size on retention and effi cacy on mixing, but advice about its limitation is of application models. However, this may weeds and wheat (Dempsey et al. 1985, often not proffered. For example, users be changing, as concerns develop over Moerkerk and Combellack 1992); nozzle are not advised that not all formulation eco-system loading, herbicide persistence, type and spray volume on retention by types are suitable. For example dry fl owa- run-off from arable systems into irrigation weeds and crop (Ayres et al. 1985); effect bles tend to settle out when mixed, thus ditches or waterways, movement through of adjuvants on the location of herbicide unless a mixing device is included in the concentrate container hence the delivered concentration may vary with time; with Table 2. Effect of tank size and application volume rate on time (Hrs) to −1 suspension concentrates viscosity increas- empty tank using 30 m boom at 25 km h . es with decreasing temperatures making 25 L ha−1 50 L ha−1 75 L ha−1 100 L ha−1 it diffi cult to transfer them from the drum Tank size Ha Hrs Ha Hrs Ha Hrs Ha Hrs to the concentrate tank; or that the time 2500 L 100 1.07 50 0.33 33.3 0.22 25 0.17 taken to reach the nozzle from the point of 5000 L 200 2.13 100 1.07 66.7 0.44 50 0.33 entry can be as much as 20 seconds. The sprayer will have travelled over 100 m if 7500 L 300 3.20 150 2.13 100 1.07 75 1.07

Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 43 deposits on weeds (Wynen and Combel- Table 3a. Summary surface features monocotyledons (Adapted from Harr lack 1992); retention on peas, beans and et al. 1991). barley (Holloway et al. 2000). Though such Hairs mm2 Stomata mm2 Contact angle° studies give guidance for improved reten- tion for some arable crops, they are not Species Wax Upper Lower Upper Lower H2O NIS applicable for all crops in all situations. Agropyron repens xlline 1.8 5.1 4.6 0.7 R 84 Hence, correct formulation properties are Alopecurus myosuroides xlline 1.2 0.2 3.1 2.0 R 119 essential to provide biological effi cacy. It is well known that small droplets Apera spica-venti xlline 1.1 0 3.3 1.0 R R (>50 µm diameter) will adhere to plant Avena fatua xlline 0.7 0.1 1.6 1.2 R 84 surfaces but larger droplets (250 to 500 µm and upwards) can rebound or shat- Brachiaria plantaginea xlline 0.1 0.5 7.4 4.3 R 79 ter (Hartley and Brunskill 1958). Since Bromus secalinus xlline 0.4 1.8 2.7 1.3 R 84 small droplets are prone to drift, we have Cynodon dactylon xlline 0.5 0.3 2.5 0.6 R 77 the classic conundrum of how to achieve maximum adhesion/retention, with mini- Cyperus rotundus smooth 0 0 0 3.8 <20 <20 mum drift. The situation is complicated Digitaria sanguinalis xlline 0.7 0.8 1.2 1.0 R 93 by different crops and weeds, as well as Echinochloa crus-galli the herbicide, having different optimal re- xlline 0.1 0 3.6 2.9 R 82 quirements because of differences in sur- Panicum dichotomifl orum xlline 0.1 0 4.3 3.1 R 85 face structure (Harr et al. 1991). The ‘wetta- Poa annua xlline 0.8 0 3.9 0.9 R R bility’, as measured by the droplet contact angle of upper vs lower surfaces, varies Setaria faberi xlline 0.2 0 4.2 2.5 R 84 between species (Harr et al. 1991, Forster Sorghum halepense xlline 0 0 2.3 2.6 R 74 and Zabkiewicz 2001). This implies that Legend: R = refl ected; NIS = 0.1% Sandovit (surface tension 30 mN m−1). not only adhesion/retention will vary, but also droplet spread after adhesion will vary. Data shows that refl ective surfaces Table 3b. Summary surface features dicotyledons (Adapted from Harr et al. 1991). are typically dominated by crystalline wax 2 2 platelets and those that are less refl ective Trichomes mm Stomata mm Contact angle° by smooth wax (Harr et al. 1991). Species Wax Upper Lower Upper Lower H2O NIS Initial retention is dependent on physi- Abutilon theophrasti smooth 0.4 1.4 2.1 1.8 66 51 cal processes comprising three competi- tive forces: inertial, surface tension and Amaranthus retrofl exus smooth 0 0 1.8 2.5 71 54 visco-elastic. These are best described by Capsella bursa-pastoris smooth 1.1 0.7 2.5 3.2 99 55 three characterisation parameters: Rey- Cassia obtusifolia xlline 0 0 1.4 3.9 R 79 nolds No. = ratio inertial/visco-elastic forces; Weber Number = ratio inertial/sur- Chenopodium album xlline 0.6 1.3 0.4 0.3 R 76 face tension forces and Capillary number Datura stramonium smooth 0.1 0.4 0.3 0.3 87 56 = visco-elastic/surface tension forces. A model for primary adhesion that omits the Gallium aparine smooth 0.4 0 0 0.4 90 58 physical forces has been suggested (For- Ipomoea purpurea smooth + 0.1 0.1 0.2 0.4 var var ster et al. 2005). It considers the infl uences ribbed of leaf surface character, droplet velocity Polygonum lapathifolium smooth + 0 0 0.4 0.4 99 57 and spray solution dynamic surface ten- fi lamentous sion. Similar plant and spray solution pa- rameters are involved in retention models Portulaca oleracea smooth 0 0 3.9 1.8 78 58 (Grayson et al. 1993, Forster et al. 2004a) Senecio vulgaris smooth 0 0.1 0.9 2.8 81 62 but such models are for specifi c crop types Sida spinosa and plant phenology (which still requires smooth 5.3 4.1 1.8 5.5 <20 <20 better characterisation), and do not em- Sinapis arvensis smooth 0.2 0.4 7.5 0.7 88 55 brace the three competitive forces, so a Solanum nigrum smooth 0.2 0.4 0.3 0.6 78 34 universal model does not yet exist. These retention models are also limited Stellaria media smooth 0 0 0.5 0.4 100 47 because they do not include droplet ve- Xanthium orientale smooth 2.2 2.1 2.0 3.9 <20 <20 locities which vary with nozzle type, size Legend: R = refl ected, var = variable, NIS = 0.1% Sandovit (surface tension 30 mN m−1). and formulation nor the three competitive forces (inertial, surface tension and visco- elastic) on impact. In the case of fi eld crop Table 4. Effect of nozzle size and formulation on droplet velocity (m sec−1). sprayers or air-assisted systems, droplet velocities may be higher than terminal, so Nozzle primary retention interactions will be dif- Spray 11001 11003 11008 ferent. Practical solutions which have met with Water 5.5 12.1 14.3 some success, are the addition of ‘stickers’, Water +Agral 4.0 11.6 14.0 which are generally high molecular weight additives, providing greater viscosity or Water + LI700 8.2 14.3 16.1 visco-elasticity to reduce droplet rebound. Water + Axion 6.0 11.9 13.8 Although providing better adhesion, such

44 Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 additives may not be very useful where cuticular wax by plasticising the amor- (related to the initial dose) is dominant good foliar coverage is essential, as this phous wax fraction. Bucholz (2006) con- in typical plant systems using accelerator would again require small droplet produc- sidered the role of the lipophilic pathway adjuvants and hydrophilic to moderately tion and potentially increase drift. for diffusing non-electrolytes across plant lipophilic herbicides. However, it is un- An alternative approach may be the cuticles and concluded, as others have likely to be predominant in all situations opposite strategy, which is the use of ad- (Schonherr and Baur 1996, Kirkwood and all plant species. Therefore a much ditives to make larger droplets shatter on 1999, Schonherr 2006), that there are three better understanding of plant leaf cuticu- impact with the leaf, and create smaller steps involved: sorption into the cuticular lar structures, as well as structure activity droplets in situ within the canopy. This lipids, diffusion across the cuticular mem- relationships for adjuvants, is still required would be an elegant solution to spray brane and fi nally desorption into the apo- to progress to a successful quantitative drift but may only be applicable to certain plast of the epidermal cells. model of uptake. canopy types. These adhesion enhancing However, there is a much simpler ef- Translocation has received the least adjuvants may give not only good adhe- fect on the leaf surface that needs to be attention, though long distance trans- sion/retention, but by their nature also considered. It has been known that in the port has been well studied and reviewed cause droplet spread. A further possibil- case of herbicides such as glyphosate, (Price 1976, Price 1982, Coupland 1988). ity is to fi ll the droplet with air bubbles higher concentrations, or lower carrier In the case of foliar applied herbicides, it using twin fl uid or air induction nozzles volumes (essentially higher concentra- is known that lipophilicity is important, and appropriate adjuvants (Combellack tions also) can improve uptake. If a spray as these compounds have to cross hydro- and Miller 2001) thus reducing impact formulation contains adjuvants that cause phobic membranes or structures other velocities and reducing rebound forces to droplet spread on a leaf surface, this will than the cuticle proper (Schreiber 2006). improve adhesion. in effect lower the mass of active per unit Hydrophilic molecules are readily trans- Though adhesion and retention prop- area, without any change in solution con- ported in either the phloem or the xylem, erties can be modifi ed substantially by centration until the spray droplet begins though their initial movements through adjuvants, their infl uence on subsequent to evaporate. Ultimately there will be a the cuticle, epidermal cells and into the performance must also be taken into ac- ‘solution residue’, where the concentra- mesophyll are not well understood (De- count, in particular their effect on uptake tion of the active bears no resemblance to vine and Hall 1990). The presence of sepa- and translocation. the starting spray solution concentration rate ‘hydrophilic’ and ‘lipophilic’ path- (Zabkiewicz 2003). It has been found (For- ways as part of the uptake process, may in Uptake and translocation ster et al. 2004b) that this solution residue turn determine the effi ciency of the subse- It was postulated several decades ago that or ‘initial dose’ (ID) can be related to the quent translocation pathway, but it is also there could be mass fl ow of spray solution mass uptake of pesticides with different diffi cult to defi ne when uptake becomes into leaves through their stomata (Schon- lipophilicities, in combination with differ- translocation. Adjuvants are known to fa- herr and Bukovac 1972, Green and Buko- ent surfactants and on several plant spe- cilitate cuticular ‘transport’ (foliar uptake) vac 1974). The latter authors found that cies. This relationship has been validated but are not thought to play any signifi - there was evidence of stomatal infi ltration with a wide range of formulations and cant part in further short or long-distance after just two minutes of exposure albeit plants, representing typical fi eld rates and translocation processes. However, in the- in only 0.5 to 4.4% of stomata. Stomatal formulations (Forster et al. 2006). ory, if adjuvants could reach the cellular penetration was achieved in practice in the Transport through the cuticle is consid- plasmalemma, then they could affect the 1980s (Field and Bishop 1988, Stevens and ered to be a three stage diffusion mecha- initial stage of the sub-cuticular transport Zabkiewicz 1988) through the introduc- nism. Namely absorption into the cuticle, process. The advantage of the mass up- tion of organosilicone surfactants. How- diffusion through the cuticle and fi nally take relationship developed recently is ever, this mechanism is not effective ge- desorption from the cuticle into the inter- that it can provide information akin to that nerically for the reason that not all plants nal leaf cells (Kirkwood 1999, Buchholz used for drug delivery dose prescriptions. have stomata on upper leaf surfaces or 2006, Zabkiewicz 2006). The principal fac- Knowing the mass uptake, an estimation the size of the stomatal pore is too small tors affecting uptake rates of non-electro- can be made of the mass translocated in or upraised. Furthermore, under stress, lytes are: specifi c systems; with subsequent studies stomata close, which also occurs at night Solute mobility: which is affected by on the infl uence of physiological and en- making spraying at this time less reliable temperature, solute molar volumes, and vironmental infl uences, appropriate ‘dos- with some products. . cuticular wax composition. ages’ could be applied at specifi c growth Various aspects of cuticular structure in Limiting skin: or the limiting skin tor- stages or conditions. It can also be used to relation to the fundamental mechanisms tuosity which is the length of the diffusion estimate the change in overall effi cacy of of uptake have been considered (Riederer path through the ‘limiting skin’. This is spray formulations, as to date no generic and Markstadter 1996). More recently re- only a proportion of, and not the entire quantitative relationship had been identi- views of the mechanisms involved have cuticle, and is infl uenced by the size and fi ed. been published. Kerstiens (2006) reviewed orientation of the cuticular wax crystals. water transport in plant cuticles and con- Driving force: which is affected by the Conclusions cluded that cuticles act as a solution-dif- starting and continuing concentration of The future focus in the use of pesticides fusion membrane for water. Schonherr a.i. in the ‘solution’ on the cuticle surface, should be on delivery of the active to the (2006) concluded that ionic polar com- in the cuticular layers, and in the epider- target site, which is not the total planted pounds rely on dipole-dipole, valency and mal cell wall. Overall, in simple terms: area, but the plant(s) within it, by using ap- electrostatic interactions to enable passage propriate formulations and droplet char- of herbicides across cuticular membranes uptake = solute mobility × cuticle tortu- acteristics to control spray distribution, through structures called aqueous pores. osity × driving force. uptake and translocation (if required). It In his review of sorption and diffusion of would seem that due to the complexity lipophilic molecules in cuticular waxes Adjuvants have been classifi ed into ‘ac- of the system, it is inevitable that this can Schreiber (2006) noted that accelerators, celerator’ and ‘passive’ categories, and it is only be completed through the develop- selected alcohol ethoxylates and n-alkyl probable that they will affect each of these ment of computer based decision support esters, did not solubilise crystalline wax terms. The results of Forster et al. (2004b) systems. This in turn requires the develop- but enhanced pesticide transfer across would indicate that the driving force ment of models for each of the processes.

Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 45 Deposition into target areas can be mod- effects on spray drift deposition in a reduces the critical rainfall period. Pes- elled, though this approach is not used simulated wetland. Pest Management ticide Science 24, 55-. much by most of the primary sectors. The Science 60, 1085-90. Forster, W.A., Kimberley, M. and Zabk- requirement to place deposits onto specifi c Buchholz, A. (2006). Characterization of iewicz, J.A. (2005). A universal spray parts of a plant, or all over a plant, needs the diffusion of non-electrolytes across droplet adhesion model. Pest Manag. to be addressed through a much better ap- plant cuticles: properties of the li- Sci., in press. proach to modelling retention (instead of pophilic pathway. Journal of Experimen- Forster, W.A., Steele, K.D., Gaskin, R.E. empirical data sets) which in turn requires tal Botany 57, 2501-13. and Zabkiewicz, J.A. (2004a). Spray a better description of plant development Butler Ellis, M.C., Tuck, C.R. and Miller, retention models for vegetable crops: and (foliar) surface characteristics. The P.C.H. (1997). The effect of some adju- preliminary investigation. New Zea- principal factors controlling foliar uptake vants on sprays produced by agricul- land Plant Protection, Hamilton, New appear to be solute mobility in the cuticle, tural fl at fan nozzles. Crop Protection 16, Zealand, ed S.M. Zydenbos, Vol 57, pp. cuticle tortuosity and solute driving force. 41-50. 125-8. Apart from the last factor, the other two are Butler Ellis, M.C., Tuck, C.R. and Miller, Forster, W.A. and Zabkiewicz, J.A. (2001). poorly understood and not easy to meas- P.C.H. (2001). How surface tension of Improved method for leaf surface ure. Whole plant translocation models surfactant solutions infl uences the char- roughness characterisation. Proceed- exist, but it is the short-term or short-dis- acteristics of sprays produces by hy- ings of the sixth International Sympo- tance controls that need to be elucidated draulic nozzles for pesticide applicant. sium on Adjuvants for Agrochemicals, and related to the initial mass uptake. Colloids and Surfaces A: Physiochemical ed H. de Ruiter, pp. 113-18. More complex models of uptake or to- and Engineering Aspects 180, 267-76. Forster, W.A., Zabkiewicz, J.A. and Rieder- tal pesticide effi cacy are being developed Combellack, J.H. (1981). An assessment er, M. (2004b). Mechanisms of cuticular and show promise (Satchivi et al. 2000a,b, of the problems of effi ciently spraying uptake of xenobiotics into living plants: Lamb et al. 2001), but their universal ap- herbicides onto weeds in cropped ar- 1. Infl uence of xenobiotic dose on the plicability and acceptance will not occur eas. Proceedings of the Sixth Australian uptake of three model compounds, until the input parameters have been sub- Weeds Conference Vol. 1, pp. 93-8. applied in the absence and presence stantially simplifi ed or can be measured Combellack, J.H. (1989). The importance of surfactants into Chenopodium album, easily. At present, as with all models, their of weeds and advantages and disad- Hedera helix and Stephanotis fl oribunda greatest value may lie in identifying the vantages of herbicide use. Plant Protec- leaves. Pest Management Science 60, gaps in our knowledge, so that we focus tion Quarterly 4, 14-32. 1105-13. on the most important elements Combellack, J.H. (2004). Interaction of noz- Forster, W.A., Zabkiewicz, J.A. and Ried- The integration of spray deposit pro- zles with spray fl uids on droplet drift. erer, M. (2006). Mechanisms of cuticular fi les within geographic information sys- Plant Protection Quarterly 19, 119-24. uptake of xenobiotics into living plants: tems can provide information that can Combellack, J.H. and Miller, P.C.H. (2001) 2. Infl uence of xenobiotic dose on the be subsequently related to pest or weed Effect of adjuvants on spray patterna- uptake of bentazone, epoxiconazole competition levels, crop productivity and tion and the volume of air inducted by and pyraclostrobin, applied in the pres- cumulative residue profi les. The develop- selected nozzles. Proceedings of the 6th ence of surfactants into Chenopodium al- ment of a comprehensive decision support International Symposium on Adjuvants bum, Sinapis alba and Triticum aestivum system (DSS) has the further benefi ts of for Agrochemicals, ed. H. de Ruiter, pp. leaves. Pest Management Science 62, 664- providing tailored sub-sets of information 557-62, Amsterdam, The Netherlands. 72. that can be used by individual operators, Combellack, J.H. and Richardson, R.G. Ganzelmeier, H. (2005). GIS-based appli- for operational optimisation, such as selec- (1985). Effect of changing droplet tra- cation of plant protection products. An- tion of appropriate pesticides or applica- jectory on collection effi ciency BCPC nual Review of Agricultural Engineering tion technology and maximising on-target Monogram No 28. Symposium on Ap- 4, 245-56 deposits. This DSS can in turn become part plication and Biology, ed. E.S.E. South- Grayson, B.T., Pack, S.E., Edwards, D. and of an agrichemical management system, combe, pp. 227-33. Webb, J.D. (1993). Assessment of a math- which can be applied on a local, regional Coupland, D. (1988). Factors affecting the ematical model to predict spray depo- or national scale and related to long term phloem translocation of foliage ap- sition under laboratory track sprayer pesticide eco-system interactions. It is sug- plied herbicides. British Plant Growth conditions. II: Examination with further gested that implementation of such a sys- Regulation Group Monograph 18, plant species and diluted formulations. tem will enable less herbicide to be used eds R.K. Aitken and D.R. Clifford, pp. Pesticide Science 37, 133-40. and a lowering of adverse environmental 85-112. Green, D.W. and Bukovac, M.J. (1974). Sto- effects. Dempsey, C.R., Combellack, J.H. and Rich- matal penetration: effect of surfactants ardson, R.G. (1985). The effect of nozzle and role in foliar penetration. American Acknowledgements type and 2,4-D concentration on spray Journal of Botany 61, 100-6. This work was supported by the New Zea- collection by wheat and weeds. BCPC Hartley, G.S. and Brunskill, R.T. (1958). land Foundation for Research Science, and Monogram No 28. Symposium on Ap- Refl ection of water droplets from sur- Technology. plication and Biology, ed. E.S.E. South- faces. In Surface Phenomena in Chemis- combe, pp. 235-40. try and Biology, eds J.F. Danielli, K.G.A. References Devine, M.D. and Hall, L.M. (1990). Im- Parkhurst, and A.C. Riddiford, pp. 214- Ayres, P., Taylor, W.A. and Cotterill, E.G. plications of sucrose transport mecha- 3. (Pergamon Press, New York). (1985). The infl uence of cereal canopy nisms for the translocation of herbi- Haas, H. (1989). Danish experience in ini- and application method on spray depo- cides. Weed Science 38, 299-304. tiating and implementing a policy to sition and biological activity of a herbi- Eichert, T. and Burkhardt, J. (2001). Quan- reduce herbicide use. Plant Protection cide for broad-leaved weeds. In BCPC tifi cation of stomatal uptake of ionic sol- Quarterly 4, 38-44. Monogram No 28. Symposium on Ap- utes using a new model system. Journal Harr, J., Guggenheim, R., Schulke, G. and plication and Biology, ed. E.S.E. South- of Experimental Botany 52, 771-81. Falk, R.H. (1991). The leaf surface of combe, pp. 241-50. Field, R.J. and Bishop, N.G. (1988). Promo- major weeds. Sandoz Agro, Agrobiol. Brown, R.B., Carter, M.H. and Stephenson, tion of stomatal infi ltration of glypho- Research Station, CH-4108 Witterswil, G.R. (2004). Buffer zone and windbreak sate by an organosilicone surfactant Switzerland.

46 Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 Holloway, P.J., Butler-Ellis, M.C., Webb, Alvin and C.E. Price, pp. 237-52, (Aca- 21st century. Proceedings of the 6th In- D.A., Western, N.M., Tuck, C.R., Hayes, demic Press, London, UK). ternational Symposium on Adjuvants A. L. and Miller, P.C.H. (2000). Effects of Riederer, M. and Markstädter, C. (1996). for Agrochemicals, ed. H. de Ruiter, pp. some agricultural tank-mix adjuvants Cuticular waxes: a critical assessment 608-620. (ISAA 2001 Foundation, Am- on the deposition effi ciency of aque- of current knowledge. In Plant cuticles sterdam). ous sprays on foliage. Crop Protection – an integrated functional approach, ed. Van de Zande, J.C., Parkin, C.S. and Gil- 19, 27-37. G. Kerstiens, pp. 189-200. (Bios Scien- bert, A.J. (2003). Application technolo- Kerstiens, G. (2006). Water transport in tifi c Publishers, Oxford, UK). gies. In Optimising Pesticide use, ed plant cuticles: an update. Journal of Ex- Satchivi, N.M., Stoller, E.W., Wax, L.M. M. Wilson, pp. 23-44. (Wiley and Sons, perimental Botany 57, 2493-9. and Briskin, D.P. (2000a). A nonlinear New York). Kirkwood, R.C. (1999). Recent develop- dynamic simulation model for xeno- Wynen, P. and Combellack J.H. (1992). The ments in our understanding of the plant biotic transport and whole plant al- effect of adjuvants on location of herbi- cuticle as a barrier to foliar uptake of location following foliar application I. cide deposits on wild oats. Proceedings pesticides. Pesticide Science 55, 69-77. Conceptual foundation for model de- Adjuvants for Agrichemicals, ed. C.L. Lamb, S., Bean, M. and Chapman, P. (2001). velopment. Pesticide Biochemistry and Foy, pp. 319-28. Pub CRC. Foliar uptake of pesticides by math- Physiology 68, 67-84. Zabkiewicz, J.A. (2003). Foliar interactions ematical modelling: a tool for infl uenc- Satchivi, N.M., Stoller, E.W., Wax, L.M. and uptake of agrichemical formula- ing adjuvant selection. Proceedings of and Briskin, D.P. (2000b). A nonlinear tions – present limits and future poten- the 6th International Symposium on dynamic simulation model for xeno- tial. In Chemistry of Crop Protection, Adjuvants for Agrochemicals, ed, H. de biotic transport and whole plant allo- eds G. Voss, and G. Ramos, pp. 237-51. Ruiter, pp. 106-12. (ISAA 2001 Founda- cation following foliar application II. (Wiley VCH, Weinheim). tion, The Netherlands). Model validation. Pesticide Biochemistry Zabkiewicz, J.A. (2006). Spray formulation Mathers, J.J., Glass, C.R., Harrington, P. and Physiology 68, 85-95. effi cacy-holistic and futuristic perspec- and Smith, S. (2006). Techniques for the Schönherr, J. (2006). Characterization of tives. Crop Protection 26, 312-19. estimation of bystander contamination. aqueous pores in plant cuticles and In Aspects of Applied Biology 77, 2006 permeation of ionic solutes. Journal of International Advances in Pesticide Ap- Experimental Botany 57, 2471-91. plication. Vol. 1, pp. 253-8. Schönherr, J. and Bukovac, M.J. (1972). Miller, P.C.H., Merritt, C.R. and Kempson, Penetration of stomata by liquids. De- A. (1990). A twin-fl uid nozzle spraying pendence on surface tension, wetta- system: A review of research concerned bility and stomatal morphology. Plant with spray characteristics, retention Physiology 49, 813-19 and drift. Proceedings Crop Protection Schönherr, J. and Baur, P. (1996). Effects in Northern Britain 1990. (BCPC, Croy- of temperature, surfactants and other don, England). adjuvants on rates of uptake of organic Miller, P.C.H., Lane, A. and Wheeler, H. compounds. In Plant Cuticles – an in- (2001). Matching spray applications to tegrated functional approach, ed. G. canopy characteristics in cereal crops. Kerstiens, pp. 134-54. (Bios Scientifi c Pesticide Outlook 100-102. Publishers, Oxford, UK). Moerkerk, M.R. and Combellack, J.H. Schreiber, L. (2006). Review of sorption (1992). The relationship between spray and diffusion of lipophilic molecules in retention and herbicidal effi cacy of di- cuticular waxes and the effects of ac- clofop methyl. Proceedings – Adjuvants celerators on solute mobility. Journal of for Agrichemicals, ed. C.L. Foy, pp. 311- Experimental Botany 57, 2515-23. 18. Pub. CRC. Stafford, J.V. and Miller, P. C.H. (1993). Nilars, M.S. (2002). Some nozzle perform- Spatially selective application of herbi- ance considerations when using wide cide to cereal crops. Computers and Elec- booms at higher spraying speeds. As- tronics in Agriculture 9, 217-29. pects of Applied Biology 66, 2002 In- Stevens, P.J.G. and Zabkiewicz, J.A. (1988). ternational Advances in Pesticide Ap- Effect of surfactants on foliar uptake: plication. pp. 95-106. interactions with species, chemicals Nuyttens, D., De Schampheleire, M., and concentrations. Proceedings of Steurbaut, W., Baetens, K., Verboven, the EWRS Symposium: Factors affect- P., Nicolai, B., Ramon, H. and Sonck, ing herbicidal activity and selectivity, B. (2006). Experimental study of factors Wageningen, pp. 145-50. infl uencing the risk of drift from fi eld Ucar, T., Ozkan, H.E., Fox, R.D., Brazee, sprayers, Part 2: Spray application tech- R.D. and Derkson, R.C. (2000). Experi- nique. Aspects of Applied Biology 77, mental study of jet agitation effects on 2006 International Advances in Pesti- agrochemical mixing in sprayer tanks. cide Application. Vol. 1, pp. 331-40 Journal of Agricultural Engineering Re- Price, C.E. (1976). Penetration and trans- search 2, 195-207. location of herbicides and fungicides in Ucar, T. and Hall, F.R. (2001). Windbreaks plants. In Herbicides and Fungicides- as a pesticide drift mitigation strategy: Factors Affecting their Activity, ed. N.R. a review. Pest Management Science 57, McFarlane, The Chemical Society Spe- 663-75. cial Publication 29, London, pp. 42-66. Underwood, A., Roberts, S. and Yopp, F. Price, C.E. (1982). Penetration and translo- (2001). An overview of the commer- cation of herbicides and fungicides. In cial agrochemical and adjuvants mar- The Plant Cuticle, eds D.F. Cutler, K.L. kets and trends impacting each for the

Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 47 data. In the case of implementing this at the Shire, the desktop information was Prioritising roadsides and reserves for weed control in gained from the Shire’s GIS Intramaps sys- tem, but for other users this may include the Shire of Yarra Ranges publicly available internet based layers (e.g. DSE MapShare, Victorian Resources David Blair, Shire of Yarra Ranges, PO Box 105, Anderson Street, Lilydale, Online, Viridans, Google Earth etc) that Victoria 3140, Australia. help describe the sites in question. The Our Common Ground Pty Ltd, 17 Kalamunda Terrace, Healesville, Victoria fi eld assessment only needs a fi eld da- 3777, Australia. tasheet, but does require the knowledge of how to do a Habitat Hectare assessment and the ability to identify weeds. The information collected from Parts Summary The Shire of Yarra Ranges is The problem of how to allocate a weed A and B were (initially) collated onto an responsible for managing over 3000 sites budget across such a large and varied mu- Excel spreadsheet, but other database sys- (reserves and roadsides) spread across a nicipality as the Shire of Yarra Ranges was tems could also be used. large and diverse landscape that contains what led to the development of the Threat a wide variety of ecological and social Matrix. The Shire’s Weed Management Project development and methods values. The Shire’s Weed Management Strategy (2005) noted that reserves and How the Threat Matrix was developed Strategy (2005) recognised many limita- roadsides were valued for a wide range of In developing a system to prioritise and tions in adopting a species based approach reasons, but that the protection of biodi- rank sites, an evaluation was carried out to weed management for its bushland ar- versity was the most signifi cant reason for of what information was available that eas, particularly given the vast majority of weed control works. could usefully describe the sites as well as weed species encountered were not listed In such a large municipality, and with what could reasonably be expected to be as Noxious. thousands of sites (reserves and roadsides) collected. Each type of information to be A site based approach rather than spe- the Shire is responsible for managing, it is used was given a weighting in terms of its cies based relies on the ability to prioritise a very diffi cult task to adequately manage importance in the overall value of the site a large number of highly variable sites. such a diversity of sites without a database and following fi eld testing of the scoring, With a lack of existing mechanisms to do describing – and putting values to – each some adjustments were made. (It should this prioritising, the Weed Management of the sites concerned and their own indi- be noted that the system can be modifi ed Strategy developed the Threat Matrix, de- vidual values. to suit whatever information is available signed specifi cally for this task. The Threat During the process of writing the Weed and the values given to different descrip- Matrix gives value to a range of descrip- Management Strategy, an innovative tors is subjective so could be changed to tors for each site, some of which are held method of prioritising all these sites was give more weighting, for example, to so- within existing GIS layers, others collected developed, the Threat Matrix. The Threat cial values instead of ecological values). in the fi eld. Matrix gave weighting to a range of ‘val- Care was taken to balance values that The methodology and collection tech- ues’ each site possessed, from ecological applied to one type of site but not others niques are not diffi cult and do not involve to social, that could be measured either (e.g. roadsides but not reserves), to avoid specialist or expensive equipment. A large in the offi ce or in the fi eld, and by assign- unfair weighting to sites with additional number of sites can be surveyed in a rela- ing value scores to these, has resulted in a descriptors (e.g. one descriptor – Roadside tively short period of time, providing a unique way of providing a ranking score Conservation Signifi cance, only applies to very useful dataset of information that for each site. roads). gives the ability to ‘see’ the whole Shire in By combining the new Threat Matrix To avoid the necessity of fi eld visiting a much more holistic way. Only by having scoring with the current State-wide sys- all 3000+ sites, but at the same time not the data describing and comparing all the tem of ‘Habitat Hectares’ (describing veg- wanting to leave any out of the analysis, sites in one place, can well informed deci- etation quality), a site valuation methodol- the Threat Matrix was divided into two sions be made. Additionally, the resulting ogy has been developed that can be used sections, Part A and Part B (as described ‘snapshot’ can be used as a monitoring to assist the allocation of resources across above). Initially it was hoped Threat Ma- function in the future to critically appraise the Shire’s sites and to recognise the value trix Part A (TMA) would be a coarse fi lter, weed management progress. in weed control. using values that could be auto-generated The fi eld assessment has also assisted from existing GIS layers. Threat Matrix Introduction with other projects with weed relevance Part B (TMB) would then provide addi- The Shire of Yarra Ranges hosts some of (schools programs, community engage- tional fi ltering with additional informa- the most important biodiversity assets in ment, education etc) and when combined tion collected from site visits to only the Victoria. It is home to all three land based with statistically targeted mapping done best sites identifi ed in TMA. However, the State emblems (common heath, Leadbeat- with the GPS, allows monitoring mapping Part A assessment alone proved far less er’s possum and helmeted honeyeater) of weed populations which assists with predictive with some sites with excellent and contains critical habitat for many en- evaluation of on ground management vegetation quality getting poor scores dangered and endemic species. The Shire practices into the future. (due to being a common Ecological Vege- covers 2472 km2 which takes in the eastern tation Class – EVC) while some poor qual- urban fringe of Melbourne (Lilydale to Bel- Materials ity sites gained high scores due to having grave), the Dandenong Ranges, signifi cant The Threat Matrix site prioritisation system an Endangered EVC – even if it was now agricultural lands including the Yarra Val- requires very little in the way of physical a mown football ground). ley winery region and large areas of state materials beyond a computer and datash- As a result, the process of putting sites forest, National Parks and Melbourne’s eets, it is more a matter of the information through TMA became a manual one as water catchments. The Shire covers the that it is based on that is of importance. did the method of reducing the overall upper 51% of the Yarra catchment and 10% Part A of the Threat Matrix uses ‘desk- number of sites to something manageable. of the Dandenong catchment which fl ow top’ information accessed via computer This was done through manually check- into Port Phillip Bay. while Part B is based on fi eld assessment ing aerial photographs of all 765 reserves

48 Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 and removing those that did not appear to Threat Matrix A + B scores (to give site way’ in weed management and initia- have enough vegetation to warrant Shire ranking) tives contractors carrying out weed control. Additional fi eld notes Enhanced environment and increased The list of 2234 roadsides was cut using o ‘Nutshell’ description of site, main cost effectiveness due to more strategic the Roadside Conservation Signifi cance features locating of control works GIS layer, so all ‘High’ and ‘Very High’ o Management comments, possible Benchmark data for Habitat Hectare conservation roads were retained on the priorities, time needed scoring collected list while ‘Medium’ and ‘Low’ were re- o Additional site location information Greater ability to receive grants and moved. Some additional sites also made and access diffi culties funding due to proven strategic ap- the list that were currently being managed As the database has developed, additional proach for other reasons or under separate pro- information has been added including: Greater ability to work with State agen- grams. ‘Friends of’/community groups work- cies in weed control, data sharing capa- Threat Matrix Part B (TMB) assess- ing on site bility ments made in the fi eld became a fairly Budgets for weed work The data gathered provides a suite of quick process with an average 10–25 sites Contractors work carried out information that can be used for a wide a day able to be visited and assessed. The Mobile phone coverage (for OH&S, to range of purposes. The primary purpose Habitat Hectare sheet established by the see where coverage extends) of developing this spreadsheet has been DSE was streamlined to a faster version the need for the Shire to make strategic to speed up this assessment, while main- Results and discussion decisions around the prioritisation of sites taining its comparative value to the state The ‘weed mapping project’ (as it was for contractor action and budget alloca- system. Field assessments provided veg- known, which was more accurately a site tion. However the spreadsheet is capable etation quality (HabHa score), an overall prioritisation project) began on 6 February, of much wider analysis - it can be ‘sorted’ weed cover score (from HabHa score sheet) 2006. The project was delivered by David to manipulate the data in a number of and a basic weed list. The weed list was Blair from ‘Our Common Ground’ as a interesting ways that allows a better un- a simple presence/ absence list of what consultant, working full time for the fi rst derstanding through creating various ‘big was on site. This information was not re- six months and part time for the remain- pictures’ of the Shire and its assets. quired for the TM ranking, but has proved der until funding ceased on 30 June, 2007 For example, see Table 1. valuable for coarse scale weed mapping when the project ended. and as a reminder of what is on each Ranking sites – triple bottom line site. Main achievements All 600 sites are now ranked with a com- 3000 sites of Shire managed land listed bined TMA/TMB score, ranging from 1 The fi nal database (2234 roads, 765 reserves) (least valuable site) to 44 (most valuable An Excel spreadsheet database was set up 600 sites fi eld assessed for: – note, the TM is an open ended scale, so with the information collected including: o Threat Matrix score (site values and if additional data fi elds are used, scores Site location details budget allocation tool) for these same sites could be higher). This o Name of reserve, address, suburb/ o Vegetation quality gives a clear ranking that, given the range town o Weed cover and weed species of environmental and social values con- o Melway ref present sidered important, each site can now be o Shire of Yarra Ranges (SYR) GIS site ‘Top 20’ list of most common weeds judged against all others across the Shire identifi cation code across the Shire produced to determine the relative value of each Area (ha for reserves) or length (km for Integration of weed distribution and site in an environmental and social way roads) of site threat knowledge with other programs (through TM) as well as an economic way Area of vegetation on site or the area as Raising of SYR profi le as ‘leading the (site value – see below). This has assisted described by the assessment (e.g. site may have sports ovals, car park, build- ings etc. that are not included in this Table 1. total) Spreadsheet sorted by: Analysis possible: Threat Matrix A information o Botanical or Zoological Signifi cance Threat Matrix score (A+B) Prioritise all sites for their ‘value’ Overlay (from Planning Scheme) Weed species Coarse scale maps of weed locations showing: o Roadside Conservation Signifi cance o Identifi cation of ‘new and emerging’ species (RCS) score o Distribution analysis, how widespread different species o EVCs and threat level (e.g. Endan- are gered, Rare, Least Concern etc) Assist with management (know what spp are on site) o Land adjoining the site (agricultural, urban, forest, river, National Park) Assist community group enquiries re spp. present o Importance as a corridor (habitat ‘Top 20’ most common weed species list connection) Track spread of weed species over time o Edge score (ratio of edge length to Weed cover Proportion of sites under different weed loads site area) Locality (town/suburb or Which areas have working community groups Threat Matrix B information from site electoral regions, Melway visit Distribution of resources ref) o Habitat Hectare score (vegetation Weed loads (% cover × area of land) quality) Weed species in certain area o Overall % weed cover (from Habitat Habitat Hectares Total # Habitat Hectares, therefore monitoring of weed Hectare score – see discussion be- cover provides possible evaluation of management strategy low) success through time o Weed species present on site Budget allocations Distribution of resources, how this relates to TM score (1=present, 0=absent/not observed)

Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 49 with directing where resources for weed The site value divides the total weed for the current weed loads and therefore management should go. management budget by the number of how much work was required to control Habitat Hectares being managed across what was there. For some sites the budget Vegetation accounting – NVMF the Shire. In this way large sites or sites allocated would allow only for coarse con- It was found that across the 600 sites (which with excellent vegetation quality gain trol of the very worst mature weeds (i.e. cover 1332 ha of forested land) there were more funding than smaller/poorer qual- sites with very high weed loads) while a total of 609 Habitat Hectares. This was ity sites (Equation 1). others will be ‘clean’ before the budget is the fi rst time this had been measured for The site value can then be further ad- spent (i.e. sites with very few weeds). the Shire and therefore set a benchmark to justed using the Threat Matrix scores, The allocations suggest what each site see if the Shire is achieving a ‘Net Gain’ as thereby giving extra weighting to our was worth from our Habitat Hectare/ described under the State Government’s ‘best’ sites (‘best’ due to all the various Threat Matrix point of view. In cases where ‘Native Vegetation Management Frame- values weighed up in the TM calculations the site was in good condition and little work’. This method of vegetation account- – ecology, social, safety etc) so the concen- needed to be spent (but a healthy budget ing will increasingly be used across the tration of value is on the better sites, with was available), the extra budget ‘cascaded’ State and to have this baseline information less on the poorer sites (Equation 2). down to the fi rst sites under the ‘cut off’. will be valuable for many future planning Given large budgets are found at the top and vegetation management aspects of Cascading budget The original intention of the list (where the best sites are) and the Shire’s work or for the setting up of a of the Threat Matrix ranking of sites was sites at the bottom of the list have smaller ‘Bush Broker’ system. that in the fi rst year, the best sites (what- suggested budgets, excess from good sites ever proportion of the overall number of would most likely be spread across several Bush Broker – Vegetation offset trading sites that ends up being) would have weed sites, doing a little on each (coarse weed scheme Under the ‘Native Vegetation control carried out on them. The following control). In this way, over time, as better Management Framework’, development year, given some of the best sites are now budgets become available for these lower in Victoria requires an overall ‘Net Gain’ to under control, the budget can spread fur- sites, the coarse scale containment of the be achieved in vegetation quality or quan- ther down the list, picking up the next best worst species would have been done, and tity. An improvement may not always be sites that were missed in the previous year. a fi ner scale of control works (which takes possible on the site where the develop- The intention was that over time, the best greater time/budget) could be implement- ment is to occur, therefore some landhold- sites took less and less to maintain so the ed. But budgets are always kept in propor- ers may choose to trade vegetation offsets. budget could cover an increasing number tion to the values of the site. The Bush Broker system was developed of sites. In this way the budget savings The primary problem with this method to allow the trading between those who from having better vegetation quality on was that it required the fi eld contractor to have made ‘habitat gains’ and those who the top sites, ‘cascades’ down the list. make the decision when to stop and move need to offset what ‘habitat losses’ their If it is found the budget was not reach- on to the next reserve and to keep a record development has resulted in. ing down far enough, or a greater propor- of how much budget would then cascade tion of the overall list needed to be covered, down if it had not all been spent. To over- Site values and predictive budgets the TMA/B ‘average’ can be adjusted (to come this problem, a ‘predictive budget’ Placing dollar values on environmental a higher number) to allow the budget to was produced which attempted to use the sites is notoriously diffi cult – but this is reach further down the list. site values combined with the weed cover effectively what is demanded of the Veg- Rather than trying to estimate what to predict what might be spent on each etation Management Offi cer who must de- each site: reserve (Equation 3). cide how to divide an overall budget for a) deserves (how signifi cant it is from an Weed cover is scored as <5% weed cov- bushland maintenance over hundreds of ecological and social perspective) and er = 0.05, 5–25% weed cover = 0.25, 25– sites. The modelling produced by this TM then trying to work out therefore what 50% weed cover = 0.5, >50% weed cover analysis does not attempt to put overall it = 1.0. The Budget balancing coeffi cient is values on these land assets, however with b) requires (dependant on weed loads and the number required to make the total for the information now gathered, the overall area to be weeded) all sites (or however many you wish to weed budget can be divided amongst the This system would provide a budget that spread the budget over) equal the overall sites using a combination of what the site fi tted the importance of the site and what it budget. It is likely to be around 5, depend- has (size and vegetation quality) and how had on it to protect. It made no allowance ing on how weedy the sites are. important it is (through the TM scoring). The budget/site data can be manipulat- ed in a wide variety of ways. An example Overall budget of this is given by the following: Equation 1. Site value (HabHa) = • Site value Total Habitat Hectares • Predictive budget Overall budget × TMA + TMB score for site Site value This should not be confused Equation 2. Site value (HH and TM) = with the amount that needs to be spent on TMA + TMB ‘average’ for all sites* each site as it only looks at what a site is *Note that the fi gure used for the TMA + TMB ‘average’ for all sites is really just a coef- worth as a proportion of the overall weed fi cient that makes the books balance. This value can be varied to give greater or lower budget, not what it needs to control its value to all sites which could allow for more or fewer sites to be covered by the available weeds, nor what the site is ‘worth’ eco- budget, for example if these fi gures are used as predictive budgets. If the ‘average’ is logically or socially. This value can assist increased arbitrarily (e.g. to 22), the budget will be spread over more sites, conversely, if with budget allocation, but there are both the fi gure is lowered (e.g. to 15), the same budget would cover fewer sites (or a greater external variables and inaccuracies within budget would be needed to cover all sites). the model that require actual budgets to be set for each site taking these additional variables into account, the most obvious Equation 3. being weed cover – see example below. Predictive budget = Site value × Weed cover % × Budget balancing coefficient

50 Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 Weed cover scoring It is worth noting to maintain (and the ecological benefi ts are ‘new and emerging’ weeds or highly local- that the weed cover scores were all ob- much greater also). ised species such as bridal creeper, aspara- tained from the Habitat Hectare scores gus fern or wild tobacco tree which are all which assume any non-indigenous veg- Assumptions and inaccuracies The in- spreading quickly through different parts etation to be a ‘weed’. Obviously many formation gathered and the resulting data- of the Shire. As managers of these weeds, exotic plants will not be treated by SYR base used to produce the predictive budg- the Shire can now target outlying popula- contractors – they may not be serious et relied on a number of assumptions and tions for eradication and make genuine enough to warrant treatment or they may had certain known inaccuracies. Despite attempts at containing the core popula- be part of landscaping, gardens or playing this, the overall integrity of the database tions. It also allows highly targeted com- fi elds/lawns. A better method (see future and the conceptual ideas held within it are munity education campaigns in the areas directions, below) would be to record a still robust enough to be worthwhile. affected. second weed cover score, just for those Assumptions and inaccuracies includ- Knowing the weed species on a site also species contractors are likely to treat. This ed: helps contractors know what herbicides/ will give a much more useful value to this Most roadsides were assumed to have tools to take as well as improving timing predictive budgeting. an average 10 m wide ‘footprint’ for the (seasonal and daily weather variation) road (6 m of road surface and 4 m of such as doing dig/pulling work on wet Minimum site budget For sites on the poor or no vegetation). Therefore 1 km days, spraying on dry days. lower end of the scoring system, their pre- of road with the standard 20 m reserve dictive budget allocation may be as little width was assumed to = 1 ha of vegeta- Weed cover across Shire sites as a few dollars. Obviously contractors tion. Wider road reserves were calcu- Analysing the categories of weed cover could not even drive to a site let alone lated individually, particularly where (<5%, 5–25%, 25–50% and >50% cover) carry out control work on these sites for road width also varies gave us an overall snapshot of how weedy that amount, so the setting of a minimum Vegetation quality (HabHa score) was the Shire was at the time of surveying budget (e.g. $200?) is recommended, so all consistent along the lengths described (2006/07). sites with predictive budgets lower than Area of native vegetation could be ac- As can be seen in Figure 1, there was this be allocated this amount of money curately measured from aerial photos a high proportion (26%) of sites that had (Table 2). Treeless vegetation (seen on aerial photos) was assumed to be exotic (e.g. Separate budget items – other works, new grass) and emerging weeds and requests Giv- Information for where the community 26% 23% en there are numerous pressures on the carry out weeding work is not com- weed control budget, some of which arise plete, particularly where weeding is throughout the year (new found or season- done by individuals, not recognised al outbreaks, responses to residents, un- groups forseen events, control of ‘new and emerg- Larger reserves were assessed looking ing species’ etc), as well as other works at the more accessible parts and assum- such as revegetation, fencing or signage, ing the rest of the reserve was of similar it may be prudent to keep a proportion of quality 17% the budget aside for general works not re- RCS data mostly 10+ years old lated to the core weed management. This Limitations on Habitat Hectare scor- could be a set % of the overall budget, it ing method; not suited to all situations 34% could be budget that has ‘cascaded’ down where assessments were made from the original list, or is separated in some other way. Coarse scale weed species mapping <5% w eed cover This separate budget would allow a pre- With the weed species listed for each site determined amount to go to contingencies (presence/absence only) and subsequent- 5-25% w eed cover so the primary control work is not com- ly added to the GIS database as a separate promised from requests, but similarly to layer for each weed species, a coarse scale 25-50% w eed cover ensure that these requests are not ignored. map can now be produced to show the lo- cation of all the main weed species found >50% w eed cover Example of budgeting (Table 2) The two on Council land. reserves in Table 2 have the same TM val- This information is useful to show the ex- Figure 1. Proportion of SYR sites ue. The table shows that one reserve (Ev- tent of different weed species, particularly with varying percentage weed cover. erard Park) is valuable because it has an Endangered EVC and is popular for rec- reation, despite having high weed loads and poor vegetation quality. The other re- Table 2. Example of budgeting. serve (Wards Road), is highly valuable due Wards Road Reserve (Monbulk) Everard Park (Yarra River, Healesville) to its excellent vegetation quality. How- 26 ha 12 ha ever despite Wards Road Reserve having a higher site value (and could therefore jus- Threat Matrix score 25 Threat Matrix score 25 tify more being spent on it), because it has Vegetation quality 0.73 (excellent) Vegetation quality 0.35 (fair) very low weed cover, it requires only $154 Vegetation type (EVC) – Least Concern Vegetation type (EVC) – Endangered ha−1 to manage, where Everard Park with (common) weeds, revegetation and fencing require- ments needs $1250 ha−1. This shows the Weed cover <5% Weed cover 25–50% value of getting (and keeping) reserves in ‘Site value’ $21,752 ‘Site value’ $ 4712 good condition. Once the weed loads are Budget 2006/07 $4000 Budget 2006/07 $15,000 under control, it takes comparatively little

Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 51 virtually no weed cover (<5%). These sites Occurrence of weed species used to inform DSE of which weed species could be tackled confi dently, knowing that With the major weed species listed for the SYR felt should be listed as Noxious. with minimal effort, weeds on these sites each site visited, a calculation was made Should the key environmental weeds of would be under control (or even eradicat- showing what proportion of sites had this Shire be listed, in the long term this ed). These sites should subsequently only each particular weed species. The result- would signifi cantly assist our efforts to require monitoring and occasional light ing graph (Figure 2) shows all species that control weeds as any listed species are weeding. The largest proportion of sites were found on at least 10% of sites. no longer able to be bought and sold and (34%) had light to moderate weed infesta- It should be noted some species are landholders have an obligation to remove tions (5–25% cover). The aim for these sites common because they are highly invasive them from their properties – one of the would most likely be initial containment and have spread widely (e.g. Cotoneas- major limitations and weaknesses in with the aim of eradication (of certain spe- ter, Blackberry), while others are common the Shire’s ability to effectively manage cies) over a 2–10 year period. primarily due to their popularity in being weeds. The remaining 40% of sites (17% with planted in gardens (e.g. Agapanthus). 25–50% cover, 23% with >50% cover) For a full list of species, see Appendix Additional programs and benefi ts arising would require substantial control efforts. A. from weed mapping. Initially this would simply be aiming at Field observations One of the benefi ts containment of the worst/ most invasive Victoria’s worst? The case for expanding from weed mapping most of the reserves species with gradual improvement over the Noxious weed listing and roadsides in the Shire was incidental time until full containment could occur. As can be seen from the list in Appendix observations. These ranged from damage This may take 20–40+ years (depending A, the vast majority of weeds threaten- to infrastructure, illegal activities (rubbish on how this is funded) and in many cases ing the Shire’s native forests are not listed and weed dumping, fi rewood collection, would require the cooperation of adjoin- under the Catchment and Land Protection motorbike riding, hunting etc), rediscov- ing landholders. Act 1994 (CaLP) as Noxious (as marked ery of otherwise forgotten sites and the ob- It should be remembered that the most with *). Indeed the only species that DPI servance of new weed populations. While weedy 40% of sites represents 250 sites or actively targeted during the project period most of these activities were too late for 460 ha of highly weed infested land, which was ragwort (and some minor blackberry action, some, like dumped weeds proved gives some scale to the weed problem control), which due to its shade intoler- highly cost effective through stopping within this Shire. It is diffi cult to estimate ant nature, is of little threat to most forest new weed infestations becoming estab- what it would take to bring these reserves ecosystems. lished. and roadsides to <5% weed cover, but is This list demonstrates the need for ur- likely it would be in the order of $5–50 mil- gent revision of the CaLP listings to extend Letters to residents requesting weed re- lion over fi ve years. However to maintain beyond ‘agricultural’ weeds, to include moval Where Shire/public land was in these sites at <5% (if we got them to this) more ‘environmental’ weeds, a review good condition, but under threat of infec- for fi ve years would probably be in the which had begun as the project was near- tion by weeds from neighbouring private order of $1 million ($200,000 p.a.), show- ing completion. land, letters were being sent to residents re- ing the signifi cant cost savings of getting The comprehensive dataset obtained questing they control their weeds. Letters reserves to a manageable level. through the weed mapping project was were highly targeted using observations

100

90

80

70

60

50

40

% of sites present on 30

20

10

0

Ivy

Trad Holly

broom *

Montbretia Blackberry* Nightshade Agapanthus Pittosporum Cotoneaster Radiata pine Spear thistle*

Spanish heath Blue periwinkle English

Sycamore maple eeping buttercup Jap. honeysuckle Wild tobaccotree

Montpellier broom* Cr Cootamundra wattle

Species targeted by DPI under Noxious weed program: ragwort (2% – and none of those sites have >5% cover of this weed) * denotes a Noxious weed

Figure 2. Occurrence of different weed species on SYR sites (% of sites present on).

52 Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 made by the Weed Offi cer to the specifi c database up to date and allow progressive SYR to urgently address this weed. The species they had on their land being in- monitoring to occur. strategy involves several aspects: cluded in the letter. • Urgent control of isolated populations Weed cover – species controlled by con- with the hope of eradication from those Greater understanding of weed ecolo- tractors As mentioned above, to gain areas (Healesville, Coldstream, Narre gy Through extensive fi eld observations more useful predictive budgets, the weed Warren East, Montrose) across the Shire and discussions with con- cover that is recorded would be more use- • Containment of larger populations tractors and the public, a much greater ful if it were just of species likely to be (Yarra Glen, Chirnside Park) and ‘core’ knowledge of weed ecology for this Shire controlled by SYR bushland contractors. area, working strategically to ‘pull in was achieved and therefore the ability to When the full list of sites is resurveyed, the edges’ of these populations and put realistic management into action was this additional information should be col- prevent further spread with an aim of greater. Knowledge included how prob- lected. eventual eradication over many years lematic different species were and in what • Community consultation – feedback on conditions. Integration with community pro- new Bridal Creeper locations, encour- grams Weeds on private land continue agement and advice relating to control Trimble GPS mapping Following the to be one of the most serious issues ham- on their properties initial site visit and ranking of all 600 sites, pering long term control of weeds in the • Coordination of weed control efforts a selection of sites had weed locations ac- Shire. Current laws, obligations and com- between other agencies within the curately mapped using a Trimble GPS pliance are all insuffi cient to prevent new SYR, and between SYR and adjoining unit. This involved a second site visit and weeds being planted and existing weeds Shires. walking over the whole site (or sub section from being controlled. There are a number Bridal creeper highlights the ‘broad pic- of the site) mapping the coverage of the of programs that have been suggested in ture’ that this mapping project produces. major weeds (those likely to be managed the Weed Management Strategy and sev- Without this overall snapshot of the Shire, by contractors). eral that have begun implementation. It is recognition of ‘new and emerging’ weeds is A total of 50 sites were GPS mapped. recommended that a revision be done of very diffi cult and an effective overall man- These were selected to provide a statisti- the WMS programs taking into account agement solution is nearly impossible. cally valid sample that could be compared the additional information now collected All locations of bridal creeper known to weed cover in the future to measure the through the weed mapping project. by the Weed Strategy Offi cer have been success of management practices. The mapped with the Trimble GPS, thus move- proposed statistical modelling was Gen- Case study – bridal creeper ment of this species should be able to be eralised Linear Mixed Model (LIMS) with During the mapping of the 600 sites across closely monitored. the design variables being Road, Reserve the Shire, it became clear that certain weeds and TM score. Co-variable was whether were limited in their extent, localised to Glossary the site was a Blackspot (specifi c weed specifi c areas. For most species, given the CaLP Catchment and Land Protec- control program) or not and the response lack of weed spread history, it is diffi cult tion Act (1994) variables were the weeds on site. While to determine if it has limited extent due EVC Ecological Vegetation Class the sampling was essentially random to a lack of weedy potential, or if they are HabHa Habitat Hectare within these design variables, sites were genuinely ‘new and emerging’ and should Intramaps Shire’s GIS mapping system checked to ensure adequate spread across be controlled. One weed that was found GIS Geographical Information the geographical region of the Shire was to be quite limited in spread, but known System achieved as well. to be highly invasive across Victoria was GPS Global Positioning System Mapping of sites with the GPS took bridal creeper. RCS Roadside Conservation Sig- considerably longer than the initial site Bridal creeper or smilax (Asparagus as- nifi cance visits, with an average of 2–4 sites per day paragoides) is listed as a Weed of National NVMF Native Vegetation Manage- being mapped. The sites were mapped us- Signifi cance (WONS), and as such is recog- ment Framework ing both point and polygon data with a nised as being one of the worst 20 weeds TM Threat Matrix range of description fi elds for each point in Australia. Bridal Creeper is wide spread TMA Threat Matrix Part A collected based around the ‘11 Core At- across Victoria and has just (in 2006), TMB Threat Matrix Part B tributes’ described in National and State been listed as Noxious (Restricted) un- SYR Shire of Yarra Ranges Weed Strategies. der the Catchment and Land Protection Act WMS Weed Management Strategy 1994. Future directions Through the preliminary mapping References Future surveying, expanding the list The done, and through casual observations Blair, D.P. and Blair, S. (2005). Shire of Yar- 600 sites assessed were selected from over made during this process, it was noticed ra Ranges Weed Management Strategy. 3000 sites, the remainder of which have that Bridal Creeper had a core popula- Shire of Yarra Ranges, Lilydale, Victo- not been assessed other than brief visual tion in the Seville/Wandin North/Gru- ria. checking of aerial photographs. In cutting yere area, but only isolated occurrences Commonwealth Government (1999). The the list, there were many sites that nearly of limited extent outside this core area. National Weeds Strategy: a strategic ap- made the cut, and would be good to do Some of the outlying populations includ- proach to weed problems of national after this fi rst year of assessment. ed Yarra Glen, Healesville, Narre Warren signifi cance. (Revised Edition). www. East, Montrose, Chirnside Park, Yellingbo weeds.org.au. Future surveying, corrections to the da- and Coldstream. While this appears wide Commonwealth Government/Weeds taset and monitoring Over time, weed spread, in total there are less than 30 sites Australia (2007). Weeds of National Sig- cover on these sites will change. It is es- beyond the core area with many of these nifi cance. www.weeds.org.au/natsig. timated all 600 sites could be re-surveyed infestations covering less than 20 m2. For a Victorian Government, Department of in 40–60 days of surveying plus time for shire as large as Yarra Ranges, this is a very Natural Resources and Environment data input (note that Trimble GPS map- limited spread. (2002). Victorian Pest Management: a ping takes longer – closer to 1–4 sites per This information, gathered through the framework for action. State Govern- day on average). This would keep the weed mapping program, now allows the ment of Victoria. www.dse.vic.gov.au.

Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 53 Appendix A. Occurrence of different weed species on SYR sites. Weed Species % sites on Noxious? Blackberry (Rubus fruticosus spp. agg.) 80% Y Ivy (Hedera helix and Delairea odorata) 59% Sweet pittosporum (Pittosporum undulatum) 54% Cotoneaster (Cotoneaster spp.) 40% Agapanthus (Agapanthus praecox) 39% Radiata pine (Pinus radiata) 37% Japanese honeysuckle (Lonicera japonica) 31% Montbretia (Crocosmia × crocosmiifl ora) 24% Holly (Ilex aquifolium) 22% Wandering trad (Tradescantia fl uminensis) 20% Black nightshade (Solanum nigrum) 16% Creeping buttercup (Ranunculus repens) 16% Spanish heath (Erica lusitania) 16% Wild tobacco tree (Solanum mauritianum) 15% Blue periwinkle (Vinca major) 14% Sycamore maple (Acer pseudoplatanus) 14% English broom (Cytisus scoparius) 12% Y Spear thistle (Cirsium vulgare)11%Y Cootamundra wattle (Acacia baileyana) 10% Montpellier or Cape broom (Genista monspessulana) 10% Y Sallow wattle (Acacia longifolia)9% Willow (Salix spp.) 9% Y Bulbil watsonia (Watsonia meriana var. bulbillifera)9%Y Hawthorn (Crataegus monogyna)9%Y Pampas grass (Cortaderia selloana)8% Arum lily (Zantedeschia aethiopica)8% Angled onion (Allium triquetrum)9%Y Cedar wattle (Acacia elata)7% Tutsan (Hypericum androsaemum)7%Y Mirrorbush, taupata (Coprosma repens)6% Ginger lily (Hedychium gardnerianum)6% Red cestrum (Cestrum elegans)6% Boneseed (Chrysanthemoides monilifera)5%Y Butterfl y bush (Buddleja davidii)5% Laurel (various) (Prunus spp.) 5% Bridal creeper (Asparagus asparagoides)5%Y Asparagus fern (Asparagus scandens)5% Privet (Ligustrum vulgare)4% Desert ash (Fraxinus angustifolia ssp. angustifolia)4% Bluebell creeper (Sollya heterophylla)4% Acanthus, bear’s breeches (Acanthus mollis)4% Banana passionfruit (Passifl ora sp. Aff. Mollissima)3% Forget-me-not (Myosotis sylvatica)3% Loquat (Eriobotya japonica)3% Himalayan honeysuckle (Leycesteria formosa)3% Willow hakea (Hakea salicifolia)3% Bamboo (various) (Phyllostachys spp.) 3% Strawberry tree (Arbutus unedo)2% Karamu (Coprosma robusta)2% Gorse, furze (Ulex europaeus)2%Y Ragwort (Senecio jacobaea)2%Y Fuchsia (Fuchsia magellanica)2% Cape Leeuwin or false wattle (Paraserianthis lopantha)1% Winter heliotrope (Petasites fragrans)1%

54 Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 within Queensland, will be a test of our capacity to control weeds overall – for cur- Grassy weeds as fi ery competitors or phantom rent and potential spread of CNG in Aus- tralia, see (WONS 2000). companions? – Lessons learnt from conservation and A great deal of effort has already gone land management into understanding the biology and ecol- ogy of CNG. As far as fi re and water is Colin Hocking, Institute for Sustainability and Innovation, Victoria concerned, it appears to be well adapted to our climate and ecological conditions. University, PO Box 14428, MCMC Melbourne, Victoria 8001, Australia. Burning can reduce CNG seed forma- tion for a year (data not yet published), but burning kills few plants, and seed set Recently, while googling, I came across a animals (including domestic stock), and resumes in only one season after burning. map of the distribution of Chilean Needle the effectiveness of mowing at seed-set on This is not surprising as CNG and its many Grass (CNG) in Victoria (DPI 2003) which fl inging the seed up roadsides and out into troublesome Nassella-like relatives now struck me for the story it told. CNG (Nas- paddocks. in Australia, derive from the temperate sella neesiana) is one of Australia’s twenty CNG is perhaps as good as any weed to grasslands of South America, with similar worst ‘Weeds of Significance’ (WONS use for examining our relationships with climate, fi re, soils and grazing pressures 2000). I have been fi ve years on the Chil- noxious and other highly harmful weeds, to our own (WONS 2000). It may even be ean Needle Grass National Task, a group because many of the aspects of CNG biolo- that CNG, like serrated tussock and other attempting to take a national level ap- gy, ecology and infestation are reasonably Nassella species, is even ahead of the game, proach to limiting, controlling and, hope- well known. It is a good case study because being adapted to a climate which is only fully in some areas, eradicating CNG from it is relatively new as a major infestation in now developing in Australia, as we feel the landscape. It is estimated that CNG eastern Australia. CNG is also one of nine the local and regional consequences of has the potential to cause up to 50% in lost closely related grassy weeds from South global climate change. Many of our newly production on farmland with high levels America that have recently become promi- emerging weeds may have already re- of infestation, due to poor palatability of nent in eastern Australia, which include sponding, via selection, to the possibility fodder, puncture of carcasses by needle- serrated tussock (Nassella trichotoma) and that we may no longer be in an extend- sharp seeds, spoilage of wool due to seed cane needle grass (Nassella hyalina). So a ed drought, but that the weather pattern contamination, and other effects. Control case study of CNG addresses the question: has changed permanently and what we of CNG on farmland has been estimated ‘What do we need to do to control, limit are now experiencing is more the aver- to be around $25 ha−1. CNG is also at and eradicate major weeds undergoing ac- age than the exception (Young 2007). We least as damaging to biodiversity as it is tive invasion?’ perhaps more clearly than also know that competitive replacement to agricultural production, being an ac- by looking at weeds fully out of control of CNG by more desirable plant species, tive invader of native grassy ecosystems across diverse landscapes, with numerous rather than simple removal and hope for which have numerous rare and threat- issues for control, which tend to muddy the best, is important for effective control ened species (WONS 2000). What struck underlying trends. CNG is still only across of CNG, partly because it has such a large me about the Victorian distribution map a small percentage of its potential range. It and long-lived seed bank (WONS 2000). of CNG was the strong association of CNG has only recently been detected (and prob- CNG is highly successful at being a occurrence with towns and cities, within ably only recently introduced) to the states phantom companion, as indicated in the a web of highways that stretch out from of Tasmania, Queensland and South Aus- distribution referred to at the start of this Melbourne. tralia, and is yet to be detected in West- paper, because our current social and envi- The spread of CNG is related to its ern Australia, where according to climate ronmental practices promote its distribu- heavy animal-borne seed. It is now rea- modelling, it has the potential for major tion and establishment. sonably well established that highways spread. Whether CNG can be prevented Any solution that controls CNG will are a major pathway for spread due to the from entering WA, eradicated in Tasmania need to address each of these factors, ways it can attach to humans as well as and SA, and either removed or contained and their interactions. In addition, any

Figure 1. Map of Chilean needle grass in Victoria in Figure 2. Potential spread of CNG in Victoria. Source 2003. Source (DPI 2003). (DPI 2003).

Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 55 effective strategy will need to take into ac- know that information alone does not lead begins at least equally from an understand- count our emerging understanding of how to effective learning and change. ing of the institutional and social settings people best learn and change to achieve An alternative approach to CNG weed of the weed, and investigates integrated more sustainable outcomes. We know, management that takes account, not only approaches to weed management that can for example, that information alone, and of what is needed from scientifi c research, fi t with institutional and wider commu- the distribution of information, to those but also from institutional understanding nity interests and expectations. spreading the weeds and those responsi- and action for change, has been outlined in Weeds may be defi ned as ‘Plants that are ble for controlling spread, will not solve Williams and Hocking (2005). This is based unwanted in the place in which they grow’ the problem. What leads to effective learn- on extensive experience of working with but this defi nition skips over the reality ing and change is taken up further below. land managers, to progressively manage that it depends on who doesn’t want them, A good example of the challenges facing CNG down, in an integrated way with the for what reasons, how much they don’t us in the control of weeds, and the need for resources, timescales and procedural con- want them, and what resources that can be integrated approaches across ecology, con- straints of the land managers. For exam- brought to bear on the problem. Learning trol science, institutional context and ef- ple, it is easier for most roadside managers and behaviour change for sustainability fective behaviour change, is the challenge to build a small allocation for CNG into tells us there is something more than this of how to prevent CNG from spreading their yearly budgets, and spread this over also for weed management. Action for along roadsides. several years, than for a large scale, one control and eradication of weeds, just as Current mowing along major roadsides off allocation, especially when the weed for other environmental issues, does not is organised to meet specifi c institutional is not of high priority. There are also insti- fl ow automatically from raising awareness objectives. These are primarily for safety, tutional expectations that the amount of that a weed is a problem or providing a with short mowing at regular intervals money and resources assigned to control range of methods for control. Information through the year on the verge of the road would decrease over time, or be allocated alone is an important but insuffi cient part (e.g. within 5 m), and aesthetics, with more to new areas of control, as the man- of action for change. People do not change less frequent mowing in other roadside aging-down strategies take effect. their behaviour or take action to protect grassy areas. There are huge areas to be The approach reported in Williams and the environment just because they are managed, so the process is streamlined Hocking (2005) was based around man- told, and become aware of, the need for ac- and strategically planned to provide for agement of CNG in high quality conserva- tion and the impacts of no action. This has safety for workers and drivers, as well as tion areas. In a subsequent trial study, we been highlighted in a major project I have identifi cation of special services and natu- have worked with VicRoads to investigate recently been involved in, to implement ral areas requiring specifi c management. control methods along roadsides in Vic- professional development for facilitators To schedule specifi c weed management toria. A joint analysis, which included in- of sustainability (Hocking, Ray & Day into this demanding and complex strate- stitutional constraints, revealed that CNG 2006). What research tells us, validated by gic approach is potentially diffi cult, but if was sitting up higher than most other veg- the Guide Beside action research program methods can be found to integrate weed etation in road verges, and this provided which captures the practical experience management with existing systems and the opportunity to integrate weed man- of numerous facilitators of sustainability, primary needs of the institutions, then the agement with existing mowing regimes, is that information and awareness is only strategic systems should be able to carry by replacing the mowing at key times of one strand in a multiplicity of factors that these out. The challenge is how to fi nd the year with a wick wiping methodology, lead to learning and change. Table 1 out- ways to best manage CNG using our lat- which brushed herbicide on the projecting lines some key elements. est understandings of both its biology and CNG leaves and fl owering tillers. Previ- Notice that high on the scale of impor- ecology, and how our current institutional ous wick wiping trials on serrated tussock tance in actually triggering change (al- practices can be put to best effect to wind it using similar principles had not been fully though all of the above interact together) back, rather than promoting its spread. effective (Miller and Wilsher 1999). How- is personal support and encouragement Removal by spraying out, the cur- ever, we pursued this approach because – the ‘trusted other’. Each of the above ele- rent main method for CNG control along we envisaged that it would be possible to ments interact to enhance the likelihood of roadsides, may not work effectively. This integrate wick-wiping into existing mow- change – for example, access to informa- is partly because of the high cost in any ing regimes over a number of years, and tion and development of skills may be- one year of a major spray-out program, that this would progressively manage the come much more effective if the people partly because CNG is not easy to identify densities of CNG along roadsides down, and context of information provision and from other more useful grasses, and part- rather than the current mowing practices skills development has a personally con- ly because large areas need to be treated. which inadvertently manage CNG den- nected element – that is, a sense of trust An additional complicating factor is that sities up. What we have found (Hocking and engagement that results in more effec- herbicide can remove not just CNG, but 2007), to be reported soon via research pa- tive learning, and that allows individuals other useful competitive plants, leaving pers, is that the fi rst run of wick wiping to come at the problem in individual ways, CNG to re-establish quickly from its high has been more successful than predicted, taking account of their particular circum- seedbank. Slashing practices along road- and that one or two additional years of stances, and to fi nd individually appropri- sides continue to promote CNG because treatment, or follow-up selective spray out ate solutions. ways to control CNG have not been ef- of individual tussocks, may remove CNG According to the literature and experi- fectively built into yearly work cycle of altogether from these areas. This method ence, there is an alternative to the learn- mowing and weed control. This is because is partly successful because it leaves more ing and change approach outlined above, CNG is not seen as a priority weed, and preferable vegetation in place as competi- but it is not a very attractive one. This to see it so is challenging because most of tive repressors of CNG regeneration from approach is perhaps best summed up in the current control methods do not fi t in the seedbank. two of the outcomes of a major Victorian with standard institutional practices and What the CNG case study highlights statewide survey on attitudes to the envi- budgetary conventions. Also, approaches is that we need to approach research on ronment, which was representative of Vic- to education of road management institu- weed management not just from a scien- torians (DSE 2006) – note that the survey tions and workers has focused primarily tifi c perspective, although understanding was conducted in 2004, prior to the current on distributing information and expecting weed biology and ecology is an important water crisis in Victoria (and Australia). To change – as is explained below, we now contribution. We also need research that the question:

56 Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 Table 1. Some key elements of learning and change for sustainability (adapted from Robinson 2005). Change Factor Estimated Contribution to Change Pre-disposing/motivation factors 70–90% of voluntary change is motivated ‘I want to’ by the individual’s frustration, dissatisfaction or guilt with their current behaviours. Enabling resources/factors 75–80% of change is dependent on external ‘I know I should’ services/products Triggering situations and factors 64–75% of voluntary change is triggered ‘OK … I’ll give it a go!’ by interactions with family, friends or workmates Satisfying factors 100% of sustained change is accompanied ‘That was OK’ by sustained personal satisfaction

A. Do you think climate change is already References having an impact on our natural envi- DPI (2003). Current distribution of Chilean ronment? – 76% of Victorians said YES needle grass Nassella neesiana in Victo- B. What is needed for us to take action ria, Department of Primary Industries, on environmental issues? – 72% of Victoria, Melbourne, http://www.dpi. Victorians believe that environmental vic.gov.au/dpi/vro/map_images.nsf/ concerns need to become a crisis before Images/weeds_potential_distribution_ they are responded to. chilean_needlegrass/$File/pot_chil- A crisis is one way of eliciting a response ean_needlegrass.jpg. to weeds. We have already seen the effec- DSE (2006). Draft learning to live sustain- tiveness of a crisis in mobilising people ably strategy. Victoria’s approach to for learning and change with a number learning based change for environmen- of Australia’s worst weeds. Serrated tus- tal sustainability. Department of Sus- sock (Nassella trichotoma), a close relative tainability and Environment, Victoria, of CNG is a good example of a weed that Melbourne. began as a minor problem in the 1990s and Hocking, C. (2007). Chilean needle grass rapidly became a major crisis, with per- along roadsides: wick wiping as a new haps climate change being an underlying option for control. VicRoads, Mel- contributing factor. CNG has all the hall- bourne. marks of a similar building crisis. Hocking, C., Ray, S. and Day, T. (2006). The The question for those of us who work Guide Beside approach to professional on weeds is, do we need to reach a crisis learning for facilitators of environmen- before we have effective action? Is it possi- tal sustainability. A summary of the out- ble to intervene through our understand- comes of Stages 1 and 2 of the Profes- ing of weed biology and ecology, as well sional Development for Sustainability as human institutions, and what it takes Educators Project., Victorian Association for effective learning and action to take for Environmental Education supported place, to avert the major impact of a weed by the Department of Sustainability and spreading so that it becomes fully out of Environment, Victoria (Learning to Live control? Sustainably Strategy), Melbourne. If we count ourselves as amongst the Miller, L. and Wilsher, M. (1999). Serrated community of researchers, extension of- tussock control. Department of Natural fi cers, educators and concerned citizens Resources and Environment AG0391. who want to make a difference in our fi ery Robinson, L. (2005). Enabling change. Tools engagement with weeds, then we must for designing and delivering commu- address all of the factors needed for ef- nity change programs, Social Change fective action, unless we want to commit Media Workshop Manual, Sydney. ourselves to the pessimistic inevitability Williams, L. and Hocking, C. (2005). Man- of action following a crisis. agement of Chilean needle grass in low- Is CNG a fi ery competitor or a phantom land native grasslands. Plant Protection companion? I suspect it is being successful Quarterly 20, 11-7. at spreading because it is able to be both. WONS (2000). Chilean needle grass (Nas- To ‘defeat the menace’ we will need to take sella neesiana) strategic plan. Weeds of effective action to counter both its compet- National Signifi cance (WONS). Com- itive and companionable attributes – the monwealth of Australia and the Na- scientifi c and social in an integrated way tional Weeds Strategy Executive Com- – which is in line with the best of learn- mittee, Canberra, ACT. ing and action for sustainability as we are Young, R. (2007). Post-winter update for beginning to understand it. water storages and fl ows. Paper pre- sented to ‘Is Australia still in drought?’, Melbourne, 4th September.

Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 57 according to prescriptions developed for Horsetail (Equisetum spp.) and knotweeds (Fallopia the management of these species. spp.) – progress on eradication of two of Victoria’s Site monitoring and data management DPI’s Integrated Pest Management Sys- State Prohibited Weeds tem (IPMS) is the database used to col- lect the data for infestations, assessments, Michael Hansford, Catchment and Agriculture Services, Department of and treatments of declared weeds in Primary Industries, Locked Bag 3000, Box Hill, Victoria 3128, Australia. Victoria.

Biology and threat of horsetail and knotweeds Abstract Horsetail (Equisetum species) Materials and methods Horsetail (Equisetum species) and knotweeds (Fallopia species) are State Framework Equisetum is the single genus in the family Prohibited Weeds, a category of noxious In Victoria, State Prohibited Weeds are Equisetaceae, a primitive group of spore- weed considered to have the highest biose- managed under the Weed Alert program. bearing plants, allied to ferns. There are curity risk to Victoria. Overseas, especially Weed Alert is a State Government program about 30 world species of Equisetum, al- in North America and Europe, Horsetail developed to prevent the introduction though the genus has a complicated tax- and Knotweeds are considered very se- of serious new weeds to Victoria, and to onomy, which includes naturally occur- rious weeds and are much more widely eradicate the most serious incursions that ring hybrids, subspecies and recognised distributed than they are in Australia. The have established. The program, delivered varieties that are often diffi cult to iden- main means of reproduction and dispersal by the Department of Primary Industries tify (Burrill and Parker 1994, Rook 2002, of both weeds is via root fragments fol- (DPI) focuses on State Prohibited Weeds Faithfull 2003). The genus is almost cos- lowing soil disturbance, and in the case of and Victorian Alert Weeds by overseeing mopolitan, with only Australia and New knotweed, both root and stem fragments. the surveillance, collection, identifi cation, Zealand lacking native species. Although This means that excavation of infestations assessment of and response to these spe- Equisetum species produce spores, the as a means of eradication is risky and, cies. ‘Weed Spotters’ are enlisted to assist main means of reproduction and disper- without proper guidance and adherence with the surveillance part of the process. sal in many environments is vegetatively to strict hygiene and disposal protocols, The ‘Weed Alert Plan Victoria 2006/2007’ via root fragments following soil distur- is actually likely to spread propagules to (Blood in publication) describes how the bance, and some species are able to root other areas. Chemical control is usually surveillance and response to potential, from stem fragments (K. Blood personal the preferred option, but there are also new and emerging weeds is managed in communication 2006). Equisetum species challenges in that established horsetail in- Victoria. often favour wet places. The long rhi- festations can be diffi cult to kill with herbi- zomatous roots divide frequently, spread cides. In Victoria, horsetail and knotweeds Surveillance underground many metres horizontally have been banned from sale as garden DPI has utilised a variety of means of de- and may descend to depths well over one plants since 2003, and the known infesta- tecting infestations of Equisetum and Fal- metre. Some species also produce tubers tion sites are largely confi ned to private lopia. For example, the Weed Spotter net- on the rhizomes. From these rhizomes and public gardens, mostly resulting from work is always on the look out for any new arise jointed, hollow stems. Some species historical plantings. The Department of incursions of these species. Other means of die back for the winter and new stems Primary Industries’ Weed Alert team runs detection of these species has resulted from re-emerge each spring, fertile stems with the detection and eradication program searching botanical databases; receiving spore cones emerge fi rst, followed by new for these species in Victoria. This paper reports from herbarium staff, other botani- stems and leaves. Other species have aerial will provide background to the biology of cal professionals or the general public; as a stems which persist all year round. Spore- horsetail and knotweeds, the threat they result of direct DPI inspections of gardens producing cones are borne at the ends of pose as weeds, their known distribution in and nurseries by DPI; and investigations the main stems of these species (Burrill Victoria, and the progress on their eradica- ‘tracing-forward’ and ‘tracing-back’ from and Parker 1994, Faithfull 2003). tion in Victoria. known infestation sites. Keywords Horsetail, Equisetum, knot- Weed status and impacts In parts of North weeds, Fallopia, Japanese knotweed, Fal- Control of State Prohibited Weeds America E. arvense is a pest in pastures, lopia japonica, giant knotweed, Fallopia In Victoria, the public is encouraged to hayfi elds, grain fi elds, orchards, nurseries, sachalinensis, State Prohibited Weed, eradi- report occurrences of Equisetum, Fallopia and small fruit crops, especially on poorly cation. and any other State Prohibited Weeds to drained soils (Cody and Wagner 1980). DPI (through the Weed Alert Program), In high densities, horsetails reduce crop Introduction but landowners with these weeds growing yields by producing inhibitory substances In May 2003, all species of horsetail (Equi- on their properties are not encouraged to that depress the growth of neighbouring setum spp.) and three species of knotweed try to treat this class of weed themselves. plants (Cooperative Research Centre for (Fallopia japonica, F. sachalinensis and F. Without adherence to best-practice man- Australian Weed Management 2003). They × bohemica) were declared State Prohibited agement of Fallopia or Equisetum, a treat- can be very diffi cult to control because of Weeds under the Catchment and Land Protec- ment operation may fail, or worse still, their extensive, inaccessible, underground tion Act 1994. These species were declared end up spreading propagules to other ar- organs (Faithfull 2003). Equisetum arvense in this category to help prevent their estab- eas. Since these species are under a State is often considered the most troublesome lishment in Victoria, as they were assessed Government eradication campaign, the weed in the genus. This species is a weed as highly invasive, known to be serious key message is for the public to report oc- in Europe, North America, South America, weeds overseas, and known infestations currences of these weeds to DPI but then to Africa, Asia and New Zealand (Weiss and in Victoria were considered very small and leave it to DPI to manage the infestations. Iaconis 2002). In Australia, all Equisetum largely confi ned to gardens and nurseries. To this end, DPI may use chemical, physi- species are declared noxious weeds in all Plants declared as State Prohibited Weeds cal or a combination of chemical and phys- states and territories of Australia (Weeds are to be eradicated from the State. ical means to treat State Prohibited Weeds Australia database 2007). At the national

58 Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 level, live Equisetum plant material is a Weed status and impacts Knotweeds can grow into new plants. Knotweeds do prohibited import to Australia (Australian were introduced to the UK in the early produce seeds, but it seems that seedling Quarantine and Inspection Service 2007), 1800s and to the USA in the late 1800s as reproduction is rarely encountered, at least and the genus is a listing on the National ornamentals. They are now serious weeds in North America and Europe (Ainsworth Environmental Alert List (Department of in the UK, Europe, USA, Canada and New and Weiss 2002). Fragments of rhizomes or Environment and Water Resources 2007). Zealand. Knotweeds rapidly invade river- green stems can be spread to other sites by banks and sites subject to disturbance, digging, road-making equipment, grading Reproduction, growth and spread Re- displacing all other vegetation by shad- and cartage of contaminated soils, in mov- production occurs almost entirely by veg- ing and root competition. They die back ing water, or in dumped garden waste. It etative means (mainly from rhizomes). over winter, leaving bare soils open to seems Fallopia species have historically Small pieces of tuber or rhizome broken erosion. The rhizomes are able to regrow been far less commonly offered for sale in from the parent plant can grow into new from small cuttings, making mature in- Victoria than Equisetum species, and there plants (Royer and Dickinson 1999), and festations extremely diffi cult to remove, have been no reports of illegal trade since stem sections can form roots in moist en- and they can emerge from deep down to Fallopia was declared in Victoria in 2003 vironments (K. Blood personal commu- penetrate asphalt (Gillespie and Faithfull (Author personal observations). nication 2006). Horsetails also produce 2004). Knotweeds are also a pest in urban tiny spores that may be spread by wind gardens and degrade urban infrastruc- Discussion and water, but these spores die of water ture by sending up shoots from their deep Infested area and infestation site stress unless they land on a site with pro- underground rhizomes which penetrate numbers longed wet conditions, such as in wetland asphalt, paved areas, and building mate- Groves and Panetta (2002) provide some habitats, and once released the spores are rials. In Australia, Fallopia species are not general principles for feasibility of eradi- only viable for about 48 hours (Royer and declared in any State or Territory except in cation of incursions. These include an in- Dickinson 1999). With these limitations on Victoria (Weeds Australia database 2007) fested area limit of less than 100 ha and reproduction by spores, vegetative repro- and their importation to Australia is not the number of infestation sites being three duction is far more common. Fragments prohibited (Australian Quarantine and In- or less as being indicative as feasible for of rhizomes and tubers can be spread to spection Service 2007). eradication. Table 1 shows that the infest- other sites by cultivation, road-making ed areas of Equisetum (0.39 ha) and Fal- equipment, grading and cartage of con- Reproduction, growth and spread Re- lopia (0.11 ha) are actually very small, and taminated soils, in dumped garden waste production occurs most commonly by are well within the feasibility of eradica- and water (Blood 2001). The sale/trade of vegetative means. Fragments of rhizome tion criteria. Table 2 and Table 3 show the potted plants of Equisetum by nurseries, or green stem broken from the parent plant number of current infestation sites for both markets, or enthusiasts is a major vector for horsetail dispersal. Such trade was more frequent prior to the declaration of Table 1. Equisetum and Fallopia number of infestation sites detected in horsetail in Victoria in 2003, but horsetails Victoria as at August 2007. are still occasionally being discovered as Botanical name Common name Total number of known Total infested traded or in gardens as historical plantings infestation sites area (ha) at resulting from past trade (Author personal at August 2007 August 2007 observations). Equisetum species Horsetail 43 0.39 Knotweeds (Fallopia species) Fallopia species Knotweeds 11 0.11 Fallopia is a genus in the family Polygo- naceae. Fallopia are fast-growing perenni- als that form dense leafy thickets, fl ower- Table 2. Equisetum species detected as garden infestations in Victoria as at ing in summer with most of the foliage August 2007. dying down over autumn and re-emerg- Species Number of known Eradication ing in the spring. They hybridise readily. infestation sites statusA Only three taxa are declared in Victoria: F. detected in Victoria japonica (Japanese knotweed), native to Ja- Equisetum hyemale 40 Treatment/monitoring/eradicated (6) pan, Korea, Taiwan and China; F. sachalin- Equisetum arvense 1 Treatment ensis (giant knotweed), native to northern Equisetum sylvaticum 1 Treatment Japan and the Sakhalin Peninsula (Rus- sia); and F. × bohemica (Japanese knotweed Equisetum palustre 1 Treatment hybrid – a hybrid between F. japonica and TOTAL 43 sites, 37 active sites F. sachalinensis). Only F. japonica and F. A Eradication status, after Panetta (2007). The fi rst status is the ‘treatment’ phase, the sachalinensis have been recorded in Vic- second is ‘monitoring’, and ‘eradicated’ is the fi nal status of a site. toria. F. japonica tends to grow to heights of 1–2 m, whilst F. sachalinensis tends to grow 2–4 m high. The stems of these spe- Table 3. Fallopia species detected as garden infestations in Victoria as at cies are hollow and bamboo-like, arising August 2007. from rhizomes. The leaves are often heart- Species Number of known infestation Eradication shaped, although variable, with leaves of sites detected in Victoria statusA F. sachalinensis the largest, measuring 15– Fallopia japonica 8 Treatment/monitoring 40 cm long by 10–25 cm wide. The roots Fallopia sachalinensis 3 Treatment/monitoring consist of spreading rhizomes that are able Fallopia × bohemica 0n/a to penetrate to depths of 6 m and to lat- erally spread 20 m from a growing plant TOTAL 11 (Gillespie and Faithfull 2004). A Eradication status, after Panetta (2007). The fi rst status is the ‘treatment’ phase, the second is ‘monitoring’, and ‘eradicated’ is the fi nal status of a site.

Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 59 Equisetum (37) and Fallopia (11). It can be see these sites eradicated in the near fu- biodiversity/invasive/weeds/alert- seen these are over the three site limit for ture. It is expected that there will be new list.html (Accessed: August 2007). feasible eradications. However, it needs to detections of as yet unknown historical Faithfull, I. (2003). Horsetails: State Pro- be realised that these principles were sup- plantings in gardens of Equisetum hyemale, hibited Weed. Landcare Note. LC0375. posed to be general, and may not apply to and possibly of other Equisetum species. It Department of Primary Industries. every eradication campaign. It also needs is expected that few new Fallopia sites will Gillespie, P. and Faithfull, I. (2004). Knot- to be noted that many of the sites in the be found in Victoria. It is eventually ex- weed: State Prohibited Weed. Landcare ‘treatment’ phase will be progressed into pected that the number of new detections Note. LC0377. Department of Primary the ‘monitoring’ phase within a year or of historical plantings in gardens of Equi- Industries. two, and then some of these sites should setum and Fallopia will fall to a very low Groves, R. H. and Panetta, F. D. (2002). be able to be confi rmed as eradicated with- level, as it is assumed that there are now Some general principles for weed in a few years. This will bring down the few, if any, Equisetum and Fallopia still be- eradication programs. Proceedings of total site numbers. Of course, this doesn’t ing traded in Victoria to provide a source the 13th Australian Weeds Conference, take into account the possibility of further of new garden infestations. eds H. Spafford Jacob, J. Dodd and J.H. detection of new sites, which is considered Moore, pp. 307-310. (Plant Protection highly likely for Equisetum hyemale, with Acknowledgements Society of Western Australia, Perth). as yet undetected historical plantings in I thank the Department of Sustainability Panetta, F.D. (2007). Evaluation of weed gardens still likely to be detected in the and Environment for providing funds to eradication programs: containment and future. Given the current low site numbers support the work of the Weed Alert pro- extirpation. Diversity and distributions of Fallopia and the fact that Fallopia was not gram; DPI for support at all levels, in par- 13, 33-41. traded as much as Equisetum in the past, ticular the support of other Weed Alert Rook, E.J.S. (2002). Equisetum species. it is expected that only a small number of team members; and Val Stajsic (National The horsetails and scouring rushes. undiscovered historical plantings of Fal- Herbarium of Victoria) who has provided Available: http://www.rook.org/earl/ lopia will be detected in the future. authoritative identifi cations of the Fallopia bwca/nature/ferns/equisetum-WIP. and Equisetum species detected in Victo- html (Accessed: 17 May 2007). Lack of a seedbank ria. Royer, F. and Dickinson R. (1999). ‘Weeds Since Victorian infestations of Fallopia and of Canada and the Northern United Equisetum appear to be largely clonal infes- References States’. (University of Alberta Press tations and have no effective ‘seedbank’, Ainsworth, A. and Weiss, J. (2002). Fallopia and Lone Pine Publishing, Edmonton, after treatment, the ‘monitoring’ phase japonica (Houtt.) Ronse Decr. (Japanese Alberta, Canada). need only extend as long as no regrowth knotweed) – an underrated threat to Weeds Australia database. Noxious Weeds from rhizomes has been noted. Of course riparian zones in Australia. Proceed- List. Available: http://www.weeds. this assumes that rhizomes will not have a ings of the 13th Australian Weeds Con- org.au/noxious.htm) (Accessed: 4 July dormant state and act as a ‘budbank’. With ference, eds H. Spafford Jacob, J. Dodd 2007). the monitoring of Fallopia and Equisetum and J.H. Moore, pp. 130-3. (Plant Pro- Weiss, J. and Iaconis, L. (2002). Horsetails infestations in Victoria, it has been pre- tection Society of Western Australia, – all Equisetum species. Weed risk as- sumed that if there has been no emergence Perth). sessment data, Pest Plant Assessment of any shoots from a monitored infesta- Australian Quarantine and Inspection Project, Keith Turnbull Research Insti- tion site for three consecutive years after Service (2007). ICON database. Avail- tute (unpublished). the last treatment of any shoots that have able: http://www.aqis.gov.au/icon emerged, eradication of the infestation is (Accessed 4 July 2007). very likely to have been achieved, and it Blood, K. (in publication). Weed Alert Plan has been noted as such on the database. Victoria 2006 2007, Department of Pri- mary Industries, Bacchus Marsh. Targeting individual species of Blood, K. (2001). ‘Environmental weeds Equisetum and Fallopia for eradication. – a fi eld guide for SE Australia’. (C.H. Table 2 and Table 3 indicate that three spe- Jerram Science Publishers, Mt Waver- cies of Equisetum and one species of Fallo- ley). pia have very low infestation site numbers Burrill, L.C. and Parker, R. (1994). Field for the State, just one site each in the case horsetail and related species. A Pacifi c of the Equisetum species. Notwithstanding Northwest Extension Publication. Pub- that there may still be further detections of lication PNW 105. Published by Oregon these species made in the future, it should State University, Washington State Uni- be very likely for DPI to be able to eradi- versity, and University of Idaho. Avail- cate the currently known infestations of able: http://extension.oregonstate. these species for Victoria within just a few edu/catalog/html/pnw/pnw105/ years. (Accessed: 15 May 2007). Cody, W.J. and Wagner, V. (1980). The biol- Conclusion ogy of Canadian weeds. 49. Equisetum Provided there is continued adherence to arvense L. Canadian Journal of Plant Sci- best-practice treatment and monitoring of ence 61, 123-33. known infestation sites, it is likely that the Cooperative Research Centre for Austral- known infestations of Fallopia sachalinen- ian Weed Management (2003). Weed sis, F. japonica, Equisetum arvense, E. sylvati- management guide. Horsetails (Equise- cum and E. palustre will be eradicated from tum species). Victoria in the very near future. Continued Department of the Environment and Wa- vigilance with the treatment and monitor- ter Resources (2007). Weeds on the Na- ing of the more numerous known infes- tional Environmental Alert List. Availa- tation sites of Equisetum hyemale should ble: http://www.environment.gov.au/

60 Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 Early Detection Project and the existing weed detection organisation, Weed Spot- Early detection for water weeds in Australia ters, has been vital for developing and implementing early detection procedures Fiona McPherson, NSW Department of Primary Industries, Grafton for water weeds in Victoria. This collabo- Agricultural Research and Advisory Station, PMB 2, Grafton, New South ration ensures that new infestations are reported to and followed up by the rel- Wales 2460, Australia. evant agencies. The Early Detection Project has provid- ed Weed Spotter members with training and education materials for identifying and detecting water weeds in their local Summary A new project initiated by the authorities and waterway managers can area. Furthermore, over 60 project partici- National Aquatic Weeds Management play a vital role in detecting aquatic weeds pants have registered as Weed Spotters. Group and NSW Department of Primary early, assisting in the subsequent manage- The Weed Spotters network and Weed Industries aims to increase the adoption ment of our waterways. Alert Contact Offi cers (WACOs) have pro- of early detection procedures for water vided the participants of the project with weeds in Australia. The project training in the reporting protocols that Formal early detection procedures have The Aquatic Weed Early Detection Project have been previously established for re- been developed for weed control authori- is an initiative of the National Aquatic porting Victorian State Prohibited, Alert ties and waterway managers to enable Weed Management Group and its initial and new and emerging weeds. Weed Spot- a proactive approach to early detection. development began in 2005 when com- ters also provide participants with access These procedures build on the current munity networks and control authorities to further training, resources and links to practices of weed control authorities and were approached and training resources the large network of volunteers. increase their capacity to identify water were developed. weeds, assess the water weed risk in their The project began in November 2006 Target audience of the Early area and to conduct surveys at high risk funded by the Australian Government Detection Project sites. Defeating the Weeds Menace program • Community groups, Informal early detection procedures and NSW Department of Primary Indus- • Water quality monitoring networks have been developed for community tries. The main objective of the two year (Waterwatch), groups and regular users of the waterways project is to increase the adoption of early • Weed Spotters, with the aim of increasing the likelihood detection procedures for water Weeds of • Regular waterway users, of opportunistic water weed detections. National Signifi cance (WONS) and other • Local and state government, They involve increasing understanding high priority water weeds in three key ar- • Catchment Management Authorities of the water weed threat and abilities to eas (southeast Queensland, northern New (CMA’s), and recognise and report incursions to the rel- South Wales and north central regions of • Field workers and weed management evant authority. Victoria). contractors. These procedures improve our ability The formal early detection procedures to detect water weeds early. This increases have been developed for weed control au- Target Weeds of the Early Detection the likelihood of successful containment thorities and waterway managers to take Project in north central Victoria or eradication and reduces control costs a proactive approach to early detection. State government, regional CMAs and lo- by treating incursions before they become These procedures build on the current cal councils were consulted to determine well established. practices of weed control authorities and what water weeds pose the greatest threat This paper will discuss the processes for increase their capacity to identify water in the north central regions of Victoria. implementing an early detection program weeds, assess the water weed risk in their These include the WoNS salvinia (Salvinia and the benefi ts this approach will bring to area and to conduct surveys at high risk molesta), alligator weed (Alternanthera phi- waterway management in the north cen- sites. This is being achieved by providing loxeroides) and cabomba (Cabomba carolini- tral regions of Victoria and nationally. training, technical advice and on ground ana); the National Alert species Senegal tea assistance with the initial assessment and (Gymnocoronis spilanthoides), lagarosiphon Introduction survey of the management area. (Lagarosiphon major) and horsetails (Equise- Water weeds are fast growing invasive The informal early detection proce- tum sp.); the state prohibited weed water plants that can rapidly infest a waterway. dures have been developed for commu- hyacinth (Eichhornia crassipes), and region- They can form large fl oating mats and nity groups, regular users of the water- ally signifi cant such as sagittaria (Sagittaria dense stands under the water or along ways, and those outside of the weeds platyphylla) and elodea (Elodea canadensis). banks where they interfere with the nor- fi eld. The aim is to increase the likelihood mal functioning of the waterbody, de- of opportunistic water weed detections. It Training stroying environmental, economic and involves developing their understanding Workshops are presented by the Early De- social values. of the water weed threat and their ability tection Project in conjunction with Weed Managing water weed infestations to recognise and report incursions to the Spotters who provide training in the estab- is challenging due to their rapid growth relevant authority. This is being achieved lished reporting protocols for Victoria. The rates, their ability to spread by seed and/ by providing training, education materi- workshops are based on the Recognising or plant fragments, and the diffi culties and als and technical advice to interested com- Weeds in our Water Training Course (VET costs associated with controlling weeds munity members and water quality moni- resource RTC2016A) which includes the in an aquatic situation. The most strate- toring organisations such as Waterwatch following subjects: gic and cost effective form of water weed Victoria. • The impacts of water weeds, management is to detect and treat the in- • How they are introduced and spread, festation early. Collaboration with existing weed • Where do they grow, Community groups, regular users of detection organisations • Recognition techniques, waterways, fi eld workers, weed control The successful collaboration between the • Disposal of unwanted plant material,

Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 61 • Identifying characteristics of noxious - water bodies in urban areas i.e. Table 1. Vectors capable of water weeds and similar looking plants, stormwater retention ponds. spreading water weeds. and • Identifying and recording the sites that • Boats and recreational craft • How to conduct a water weeds risk as- provide an ideal habitat for water weed sessment and survey in you local area. growth. Water weeds require predomi- • Boat trailers nantly slow moving, permanent and • Eel trapping equipment Resources to the project shallow freshwater to grow. They are • Fishing nets A number of resources are utilised by the more likely to grow well in degraded Early Detection Project, including a re- waterways with elevated nutrient lev- • Dumping of aquarium or fi shpond viewed version of the Recognising Weeds els, direct sunlight and warmer water plants in our Waters presentation, the new Early temperatures. They can be found grow- • Irrigation channels Detection procedures presentation and ing in the following types of water- live water weed specimens. Each partici- ways: • Irrigation equipment pant in the workshop receives the follow- - streams, • Floods ing training materials: - lakes, • Hobbyists • Recognising Weeds in our Waters - wetlands, Workbook, - farm dams, • Earth moving equipment, trench • The Early Detection Survey Manual, - water storage facilities, diggers • An water weed WEEDeck, - stormwater retention ponds, • Wildlife (birds, mammals etc.) • Brochures and weed alerts from State/ - irrigation channels, • Floods Territory agencies, - canal estates, • A Early Detection Project fl yer, - urban drains, and • Contaminated plants • A Workshop Evaluation form, and - irrigated crops (rice and turf). • Mulching • A Weed Spotter reporting sheet. • Prioritising the recorded sites by using the matrix in Table 2 by determining if • Incorrect disposal of plants The Early Detection Survey – a the site is a high, medium or low risk. • Fire fi ghting equipment formal approach As a general rule at least 80% of your (i.e. Heletankers) The Early Detection Survey is a formal sites should fi t into the high risk cat- • Cropping (turf production and and proactive approach to detecting water egory. Sites with low or medium risk distribution) weeds, aimed at weed control authorities are useful to survey if there are signifi - but also appropriate for community organi- cant environmental, economic or social • Cattle and horses sations ready to carry out more formal pro- assets downstream. • Slashers cedures. It allows us to determine which • Determining how many sites can be • Deliberate plantings by aquarium sites within a management area could be surveyed and how often. Generally plant traders at risk of water weed introduction and/or 10–15 sites per day can be inspected by provide ideal habitat and conduct routine a two person crew and at least two in- • Mistakenly grown as a vegetable surveys at these sites. It contains parts: (1) spections of each site per year is recom- (Alligator weed) the local area assessment; (2) the survey mended. and (3) data management/ reporting. Table 2. A matrix for prioritising aquatic weed survey sites. Part 1: Local area assessment The local area assessment is an in-offi ce Risk of introduction evaluation of the management area that (human activity) determines which sites could be at risk of ————————————❯ water weed introduction and/or provide the ideal conditions for growth by: Moderate human High human • Learning to identify water weeds and Limited human access (bridge access (bridge determine which ones threaten your lo- access to stream crossings, parks, crossings, parks cal waterways by attending the Recog- boat ramps) boat ramps) nising Weeds in our Waters workshops or utilise identifi cation recourses. • Seeking expert advice and collating Fast fl owing. Few if maps, photos and information about any ponded areas. Low Risk Low Risk Medium Risk previous water weed infestations and Riparian cover intact. actions in your area. • Identifying and recording the sites where weeds could be introduced to your local waterways. Water weeds Slow moving and are predominantly introduced and ponded. Moderately Low Risk Medium Risk High Risk spread by human activities (see Table degraded. Partial

1). Therefore, weeds are often found in Conditions for growth riparian cover. accessible areas of the waterways at: - bridge crossings, - major roads, Slow moving and - public parks beside water bodies, ponded. Highly - boat ramps, degraded. Limited Medium Risk High Risk High Risk - streams or wet areas where earth riparian cover. High moving activities have recently oc- nutrient loads. curred, and

62 Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 • Determining when to survey the sites. • Water weeds present, fi nalised to be applied nationally, to en- The survey should be conducted when • Habitat area, sure that stakeholders have the capacity the conditions are safe for the partici- • Coverage, to continue early detection beyond the end pants and plants are easiest to detect • Growth stage, of the project. e.g. when fl owering. • Treatment, and • Photos available. Conclusions Part 2: Survey The Early Detection Project has already Once the high risk sites have been identi- Informal early detection procedures raised the awareness of water weeds and fi ed its time to get out of the offi ce and con- Community members, fi eld staff and those the importance of early detection in the duct a survey. At each site walk along the outside the weed industry can play a vi- north central regions of Victoria by engag- bank and closely observe the water sur- tal role in detecting and reporting water ing with the community and relevant or- face, banks and submerged environments weeds early by: ganisations, providing training, resources for floating, submerged and emergent • Learning how to recognise and report and linkages with established weed detec- weeds. Structures such as bridge crossings water weeds that threaten local water- tion networks. Community groups, fi eld will allow you to inspect the middle of the ways at the Recognising Weeds in our workers, weed control authorities and wa- water body. Even if the site appears to be Waters Workshops or utilise identifi ca- terway managers can play an important weed free check around logs, fences or tion recourses. role in detecting water weed infestations other snags where water weeds may have • Keeping an eye out for water weeds early. This will enable weed managers to been trapped during previous fl ooding. while conducting monitoring and other have a better understanding of the distri- Record information about the site in your activities in the waterways. Being par- bution of water weeds and prioritise the database or complete a Site Survey Data ticularly vigilant at high risk sites where subsequent management actions to pre- Sheet. If you fi nd a water weed: water weeds are likely to be introduced vent their spread. • Double check the plant’s identifi cation and/or provide the ideal conditions for Due to the keen interest shown by the using the identifi cation resources. growth. See table 2. participants, the project appears well on • Determine the upper most limit of the • If you fi nd a suspected water weed target to meet the objective of increasing infestation. Work your way upstream contact the relevant authority (Report the adoption of early detection procedures until upper limits are found (some- all State Prohibited, Alert and new and for water weeds in Australia. times a boat may be necessary useful emerging weeds to Weed Spotters). for this). The upper limit may be an iso- Acknowledgements lated swamp or farm dam which may Evaluation Thanks to Rod Ensbey, Andrew Petro- only join the stream in times of high The Early Detection Project has already eschevsky and Elissa van Oosterhout fl ow. trained over 180 people in the high priority for providing comments on this paper. I • Photograph the weed close- up and in areas of Southeast Queensland, Northern would also like to thank Weed Spotters, its habitat. New South Wales and the north central Waterwatch and all workshop participants • Report the weed to the relevant author- regions of Victoria. As a result of the joint for your willing participation, feedback ity (Report all State Prohibited, Alert workshops conducted with Weed Spotters and enthusiasm. and new and emerging weeds to Weed in Victoria, over 80 participants registered Spotters). Weed Spotters has protocols as Weed Spotter members. References established to ensure that the plant’s All workshop participants are given the Thackway, R., McNaught, I. and Cunning- identifi cation is confi rmed, that the rel- opportunity to rate the effectiveness of the ham, D. (2004) A national set of core evant authorities are contacted and that workshop, the methodology and resources attributes for surveying, mapping and the Weed Spotter is kept informed by and provide comments and suggestions in monitoring Weeds of National Signifi - the Weed Alert Contact Offi cer. the Workshop Evaluation Form. The re- cance. Proceedings of the 14th Austral- sponse to the workshops has been very ian Weeds Conference, eds B.M. Sindel Part 3: Data management and reporting. positive with all participants reporting and S.B. Johnson, pp. 690-3. (Weed So- It is important to record the site informa- an increase in their knowledge of water ciety of NSW, Sydney). tion, survey results and follow-up survey weeds and rating the workshops as use- dates in your database or complete a Site ful or very useful. This feedback has also Survey Data Sheet and report this to the been used to review and update training Early Detection Project for mapping. All techniques, course content, resources and the site details need to be recorded even survey methodology. when no water weeds are present. Please Many enthusiastic participants are keen note that because survey sites are not ran- to apply the early detection techniques domly selected, data cannot be used to in- learnt at the workshops to their local area. fer characteristics of areas not surveyed. Several organisations have already con- The following nationally agreed at- ducted an assessment of their local area tributes for surveying and mapping Weeds and carried out the initial surveys with a of National Signifi cance are required to en- small number of new infestations detected sure that consistent and reliable informa- and reported. tion is collected (Thackway et al. 2004): • Site name, The future • Date, The Early Detection Project will continue • Latitude and longitude of site (point to provide training, support and a range format), of activities in the high priority regions, • Length of water body surveyed, probably expanding to other areas of Aus- • Waterway description, tralia in 2008. • Site facilities/structures/modifi cation, New water weed detections will be • Depth, collated and mapped. The early detec- • Substrate, tion methodology will be reviewed and

Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 63 Many garden plant species currently being traded, however, are or have the Weed proofi ng Australia: a way forward on invasive potential to cause serious harm to farm- 1 ers, the agricultural industry, areas of high garden plants conservation value and the broader envi- ronment. In short, these invasive plants Andreas Glanznig, WWF-Australia, GPO Box 528, New South Wales 2001, present a clear strategic risk to Australia’s Australia. agricultural profi tability and our natural assets, and as such warrants concerted action by governments, industry and the community to mitigate this risk and fa- Introduction cilitate a transition toward a prosperous costly escaped invasive garden plants. garden plant market based on low-risk The control of invasive plants represents This paper outlines a strategic response plants. an immense new challenge for Australia to this challenge – one that can enable both It is important to also recognise that the in the opening decades of the 21st cen- the garden industry and Australian com- existence of these high risk plants in Aus- tury. For the sake of our future economy munities to play a signifi cant and positive tralia’s garden plant market is a legacy of and our environment, we must tackle role in weed proofi ng Australia. It includes Australia’s quarantine regime that existed the problem more effectively. opportunities to create new markets for before the mid-1990s that focused on con- Dr Rachel McFadyen, Killing Us low risk plants, and to enable the garden taminants and a relatively small number Softly – Australia’s Green Stalkers industry to move decisively onto the front of serious agricultural weeds. These plant (2003, p. 3) foot in responding to the invasive plant species would not be able to be legally im- problem and future civil liability risk. ported into Australia today. In 2004, an infestation of Mexican feath- A policy package is needed that strives ergrass (Nassella tenuissima) – the fi rst in Issues to achieve a 2015 vision of a prosperous Australia – was found escaping from a Some of the issues that need to be resolved garden industry built on the sale of low- Tamworth garden. It is an attractive orna- through the proposed policy package are: risk garden plants, encouraged by an em- mental tussock grass imported legally into • Uncertainty about the weed status of powered and enabled community that Australia in about 1996, propagated and garden plant species in the garden in- wants to reduce their weed spread risk then widely sold. It is a known grazing dustry, footprint. and environmental weed that economists • Lack of comprehensive and detailed The policy package needs to deliver the estimated would cost graziers $39m if it knowledge about the nature and ex- following outcomes by 2010: escaped, and has the potential to invade 14 tent of the garden industry pathway for million hectares (Groves et al. 2005). Meat propagation and movement of invasive Environmental and Livestock Australia are on the record plant species, • Only new low risk plant species are le- saying that: ‘it’s the last thing we want. • A suite of poorly harmonised weed gally permitted into Australia by 2006, It would be an absolute disaster for Aus- lists, • No high risk garden plants are traded, tralia (Williams 2004). • No convenient information measures focusing on those yet to naturalise or The example of this high-risk invasive in place to enable consumers to easily become widespread, garden plant highlights a bigger problem. identify and properly care for medium • Increase in detection and eradication of Seven out of 10 of Australia’s environmen- risk garden plant in response to grow- new high risk garden plant incursions. tal and agricultural weeds are escaped in- ing social concern about weed impacts, vasive garden plants, and the bulk of the and Social potential invasive plants already here are • Potential for industry leaders to be • Australians are empowered and ena- introduced garden plants. Those that have commercially undermined by those bled to reduce their weed spread risk, already escaped cost farmers $100m’s each in the industry not adopting positive • In major cities and towns, individu- year – just three escaped invasive garden measures. als able to join community-expert net- plants cost farmers nearly $100m y−1: rub- works to detect and eradicate new high ber vine costs $27m y−1 in control lost pro- A way forward: a 10 point plan risk garden plant incursions, duction costs, Paterson’s curse costs $30m Australia has the challenge of working out • Garden industry supports, and is ena- y−1, and lippia costs $38m y−1 – and con- an effective, least cost solution to tackling bled to play a signifi cant role in reduc- tribute to the $1 in every $7 of income that the causes and sources of the invasive ing Australia’s weed spread risk, farmers’ lose to weeds each year (Sinden garden plant problem. We have a superb • Garden plant species that present a 2004). Just one escaped invasive garden opportunity to work out this solution in high risk of demanding signifi cant Aus- plant, lantana, degrades 4 million hectares a strategic and systematic way so that all tralian community group effort to re- of Australia, and many others pose direct stakeholders go into a change process with store bush and land are removed from threats to threatened species. their eyes open and are able to participate sale. As a consequence, a key focus of Aus- in an ordered transition. tralia’s weed defence system needs to be The starting point is the need to recog- Economic tackling the invasive garden plants issue nise that most garden plants in trade have • Garden plant species that present a head on. We need to get it right so farmers, no weed history and can be considered high risk (and potential and/or actual national park managers, bushcare, land- environmentally safe. This creates an op- cost) to Australia’s agricultural indus- care and other community groups aren’t portunity to shift the market towards the tries are removed from sale, subjected to continued and new waves of majority of low risk plant species. • Garden plant species that present a high risk (and potential and/or actual cost) to government agencies respon- Footnote sible for managing national parks and 1 This paper was fi rst published in the document Weeds in the Media (2006) CRC for other crown lands are removed from Australian Weed Management. Adelaide. sale,

64 Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 • Garden businesses unduly fi nancially cultivars and varieties of garden plants, Menace’ Research and Development Plan, impacted by the policy receive one-off fruits and vegetables already on the list. and the opportunity exists to make this a transition reimbursement where ap- There has been some misinformation that research stream in the proposed new Inva- propriate (i.e. signifi cant loss in mar- new varieties of common garden plants or sive Plants CRC. ket value of existing Plant Breeder vegetables already on the permitted list, Rights), such as roses, lettuce and tomatoes, would Research to produce genuine sterile • New market demand created for low- be banned. This is simply not true. cultivars risk garden plants. There is a market opportunity to produce This can be achieved through a 10 point 2. Give garden industry and genuinely sterile cultivars of profi table policy package: communities certainty about the invasive garden plants. To ensure that 1. Close Australia’s front door to new weed status of garden plants governments recognise these 100% sterile weeds. National list of invasive plant species cultivars, they would need to be vetted 2. Give garden industry and communi- The weed issue is characterised by a mul- through a national sterile cultivar accredi- ties certainty about the weed status of titude of weed lists, combined with vary- tation scheme (see Plan Point 7), and to garden plants. ing degrees of uncertainty and confusion enable effi cient compliance they would 3. Better understand the extent and risk about which garden plant species are in- need to be able to be easily told apart from from continued trade in invasive gar- vasive. To fi x this problem, both the Nurs- invasive varieties by having unique fea- den plants. ery and Garden Industry Australia (NGIA tures, such as unique coloured fl owers or

4. Build knowledge about sterile garden 2004) and WWF (Glanznig et al. 2004) are stems for example. plants and the dynamics of invasive- calling for governments to develop one The garden industry has the oppor- ness. national ‘master’ list of invasive plants un- tunity to explore this avenue further by 5. Build garden industry understanding der which State, regional and local lists are becoming a partner of the new Invasive about the risks and costs associated nested. This should build on the existing Plants CRC bid, and including a new re- with invasive garden plants, and ca- national list of naturalised plant species search stream on sterile cultivars. pacity for positive action. (Groves et al. 2003). 6. Mobilise garden industry to respond The National List should be divided 5. Build garden industry positively to the invasive species chal- into various threat or risk based catego- understanding about the risks lenge. ries, which clearly delineates between and costs associated with invasive 7. Protect garden industry leaders and re- high-risk and medium risk plant species. garden plants, and capacity for duce transaction and compliance costs positive action by establishing a streamlined national 3. Better understand the extent Garden industry invasive plants capacity regulatory framework. and risk from continued trade in building program 8. Phase out supply and trade of high risk invasive garden plants The garden industry needs to strategi- invasive plants nationally. The landmark 2005 CSIRO ‘Jumping the cally reposition itself to play a major and 9. Encourage gardeners to increase Garden Fence’ report highlighted the im- positive role in the solution to the inva- product demand for low risk garden pacts of invasive garden plants on the en- sive plant problem. Governments and key plants. vironment and agriculture, the signifi cant scientifi c bodies, such as the Weeds CRC, 10. Mobilise communities to search and number of high-risk invasive garden plant need to support these efforts. destroy new infestations of escaped in- species still in trade, as well as poorly har- Opportunities include training, incor- vasive garden plants. monised State and Territory noxious weed poration of invasive plants in accredita- The diagram in Figure 1 shows how this lists. However, the last reasonably compre- tion standards, positioning garden centres 10 point plan targets different stages and hensive audit of known weeds in trade in as knowledge providers on invasive plant audiences along the pathway from propa- the garden industry was done for 1998/99 solutions to consumers, and becoming gation to escape of high-risk invasive gar- by the WA Department of Agriculture. leaders in community involvement pro- den plants. There is a pressing need for a national grams to replace high risk with low risk audit of garden plant species in recent plants (e.g. Garden future fi tting schemes 1. Close Australia’s front door to trade, to identify a comprehensive shad- (see Plan Point 9)). new weeds ow list of high risk plant species that need Comprehensive permitted list/weed risk to be removed from trade to reduce the National weed information system/portal assessment system by 2006 risk of them causing signifi cant harm to Currently, information to identify and In early 2005, the Australian Government agriculture and/or the environment, par- manage weed issues is scattered among committed to close a quarantine law loop- ticularly those yet to naturalise or become a wide range of sources. This makes it hole that allowed nearly half of all plant widespread. diffi cult for the community and garden species on Earth to be imported into Aus- This national shadow list should be a industry alike to keep abreast of new in- tralia with no weed risk assessment – in- focus for full weed assessments by gov- formation, as well as the changing legal cluding over 3335 known weeds not yet ernment agencies (see Plan Point 8), and status of plant species. What is needed is a found in Australia (Glanznig 2005). These self regulation by nursery growers (see one-stop-shop national weeds information known weeds became prohibited imports Plan Point 6). portal that becomes the premier gateway in June 2005 (Macdonald 2005). In 2006, for information about weeds in Australia. when the loophole was fully closed all pro- 4. Build knowledge about sterile Both WWF and the Australian Institute of posed imports of new plant species not on garden plants and the dynamics of Horticulture support the development of the national permitted list became subject invasiveness a one-stop-shop national weeds informa- to a weed risk assessment, with only those Research into why and what plants tion portal. that present a low risk to Australian agri- become invasive The Australian Government, acting on culture and the environment able to be le- Research into invasion biology and ecol- advice from the National Weeds Advisory gally imported. This strengthened quaran- ogy is a growing fi eld of scientifi c endeav- Group (NWAG), is now developing a na- tine regime will keep new invasive weeds our. The Australian Government recently tional weeds information portal under its out of Australia, but still allow plant spe- committed to provide some additional ‘Defeating the Weeds Menace’ program. cies on the permitted list, including new funding through the ‘Defeating the Weed

Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 65 Weed Proofing Initiative The Garden Plant Invasion Pathway

Comprehensive National Import of new overseas Border Permitted Plant List plant species

National List of Invasive Existing introduced and native Plant Species plant species under cultivation

Nursery National Invasive Plant Plants Handler Capacity Biosecurity Nursery Growers Building Alert Program System

National Post-border Permitted Plant List (permitted for propagation and sale) Nurseries Supermarkets/ Landscapers National Noxious Weed List hardware chains (prohibited from propagation and sale)

National Weed Awareness and Education Campaign Consumers Developers National Mandatory Invasive Plant Labelling Scheme

National Voluntary Grow Me Instead Labelling Scheme Gardens Future Fitting Garden Schemes

National Weed Alert Pro-active: Emerging invasive plants in areas Early Warning and Rapid Weed Alert Network surrounding urban gardens Response System Reactive: Bushcare/ Widespread invasive plants in areas Coastcare restoration surrounding urban gardens

Serious weeds of the Serious weeds of agriculture environment

Figure 1. How this 10 point plan targets different stages and audiences along the pathway from propagation to escape of high-risk invasive garden plants.

6. Mobilise garden industry to Industry Australia (NGIA) policy was re- risk assessment process, a national weed respond positively to the invasive leased in late May 2006. An opportunity information portal, and a national manda- species challenge exists in these policies to spell out the in- tory invasive plant labelling system. Garden industry policies dustry’s duty of care to mitigate the risk These industry policies also have an im- Peak garden industry groups have a posed by invasive garden plants (particu- portant visioning role, exemplifi ed by the strong leadership role to play in the in- larly high-risk invasive plants), as well as AIH commitment to support all programs dustry’s response to the invasive garden committing to develop appropriate codes that will ensure no new invasive plants plant challenge. The Australian Institute of conduct. become naturalised in Australia from 2010 of Horticulture (AIH) recently released its It is instructive to note that WWF and onwards (AIH 2006). updated weed and invasive plant policy the AIH share many policy positions, such (AIH 2006) and the Nursery and Garden as the need for a national standard weed

66 Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 Beyond compliance: institutionalising National post-border permitted plant list part of a broader education program, that corporate social responsibility (CSR) The most cost effective way for govern- gives consumers ‘care’ information – like Australia is in a period of policy fl ux as ments to prevent new weed problems is medicines advising on number and size governments move to strengthen measures to implement permitted list/weed risk of doses – rather than ‘don’t’ information for weed prevention and control; an exam- assessment systems – essentially making – like cigarette labelling. Labelling will be ple is the number of States and Territories any proposed new plant species a prohib- particularly useful for the growing market that are currently reviewing their noxious ited import until proven low-risk by a rig- of garden plants sold through hardware weed lists (e.g. Vic, NSW, NT). Rigorous orous weed risk assessment process. and supermarket chains, which do not of- weed risk assessments of candidate weeds Western Australia already has a com- fer advice to consumers. take time and as a result many States have prehensive permitted list/risk assessment WWF believes that the label should be a large backlog of plant species to be as- system in place, while the Northern Ter- designed in a way to show regional differ- sessed. The consequence is that there are ritory has a partial permitted list/risk as- ences in invasiveness – perhaps a map of signifi cant lead times between State agen- sessment system. Queensland and Tasma- Australia showing suitable growing area cies identifying a candidate weed and its nia are examining the feasibility of imple- where the plant is not known to be inva- fi nal declaration as a noxious weed. menting a permitted list system. It makes sive in one colour, with areas where it is An alternative to this situation, is for a lot of sense to bring all these efforts to- invasive highlighted in another colour. It garden plant growers to move beyond gether and fuse them into a national post- also needs to include the proper botanical compliance by using resources such as the border permitted plant list to complement name to help avoid confusion. Global Compendium of Weeds and emerging the Australian Government’s national bor- A mandatory labelling scheme would studies to identify high risk candidate in- der permitted plant list. This double per- also enable the garden industry to move vasive garden plants, and in consultation mitted list approach is the most effective to contain future civil liability risk. A study with weed scientists, put them in a hold- and effi cient policy approach to prevent by the University of New England’s Aus- ing pattern until they can be properly risk new weed problems, and also remove the tralian Centre for Agriculture and Law assessed. incentive to smuggle new invasive plant identifi ed that the practice of knowingly seeds into Australia from overseas and selling invasive plants with no consumer National plant handler biosecurity alert then exploit the weaknesses in narrow disclosure presents a civil liability risk for system State and Territory prohibited lists. the garden industry, and that a signifi cant The garden industry is a national market. benefi t of a national labelling scheme is to The movement of plants and soils around National noxious weed list reduce the probability of the success of fu- Australia could provide a vector for the It doesn’t make sense, and is a waste of ture claim/s against garden plant growers movement of disease or invertebrate pests scarce funds, to have one government try- and sellers to pay for the ‘clean up’ costs of (such as Phytophthora or fi re ants) to new ing to control a plant species while it is their invasive plants. This risk mitigation regions. This is an important risk that being sold in another jurisdiction. CSIRO is analogous to fast food chains introduc- needs to be managed. Part of the solution found that 40% of the naturalised inva- ing low fat product lines and strengthen- is a national rapid alert system to let plant sive garden plants declared noxious in ing consumer disclosure of product ingre- handlers quickly know about new biose- one jurisdiction were available for sale in dients to help contain the liability risk of curity risks and what they need to do to another (Groves et al. 2005). This also ap- claims that fast food caused their obesity manage any risk. plies between the Australian Government problems (Martin et al. 2005). and State Governments, where one AQIS Experience shows that to be work prop- 7. Protect garden industry leaders quarantine target weed and six national erly, the labelling scheme needs to be man- and reduce transaction and Alert weeds were still available for sale in datory (Wilkenfeld 2003). compliance costs by establishing 2004. (Glanznig et al. in press). Both the Australian Institute of Horti- a streamlined national regulatory There are four leading countries work- culture (AIH 2006) and WWF support a framework ing on weeds: South Africa, New Zealand, national mandatory labelling scheme (Ta- The garden industry is a national market, the United States, and Australia. ble 1). though domestically weed control is cov- All of them, except Australia, have na- ered by eight separate jurisdictions each tional noxious/pest plant lists. It is time National sterile cultivar accreditation with different laws and lists. The result for Australia to also implement one na- scheme is ineffi ciency and inconsistency. Govern- tional list that refl ects the national nature There is currently some controversy about ments need to take a stride forward and of the garden plant market and how it is whether new varieties of lantana and black develop a stream-lined and coherent na- promoted, and the national spread path- agapanthus are truly sterile. Some lantana tional framework of laws that build on a ways of invasive plants. The focus of varieties promoted as sterile, for example, national standard weed risk assessment the proposed national noxious weed list were subsequently found to be able to protocol, clear national weed priorities and should be on those high-risk plant species cross breed with weedy forms of lantana. nation-wide control of high risk invasive that are yet to naturalise, or are not yet (Neal and Playford nd). To give consum- plants through a national noxious weed widespread where restricting supply will ers certainty and peace of mind, and also list, as well as a national labelling scheme. result in a reduction of establishment and ensure that claims of ‘non-invasiveness’ Experience in the United States (Caton invasive success. on labels are not misleading, sterile plant 2005), New Zealand and Australia (Moss, varieties could be vetted through an inde- and Walmsley 2005) shows that to protect National mandatory invasive plant pendent and scientifi cally robust national those industry leaders that do the right labelling scheme sterile cultivar accreditation scheme. Vari- thing and remove high-risk species from Consumers have a right to know if they eties that were scientifi cally proved to be trade, laws are needed to create a level are buying a potentially invasive plant. sterile could be accredited and branded playing fi eld so all growers and sellers are They also need to know how to reduce as ‘environmentally safe’. To provide a required to remove high-risk plant spe- their weed spread risk by knowing if the market incentive, these varieties would cies, so leaders are not commercially dis- plant is suitable for the proposed location, need to be exempted from sale bans, but advantaged. Voluntary approaches alone how to maintain it, and how to dispose of to ensure easy compliance people would do not work. green waste responsibly. This can be fa- need to be able to easily tell the difference cilitated at Point of Sale through a label, as between invasive varieties and the sterile

Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 67 Table 1. Relationship of proposed key elements in an effective national regulatory framework.

National List of High Risk Plant National Noxious Weed List (prohibited for Quarantine List Invasive Plant Species sale (accredited sterile cultivars/varieties Species excepted)) Alert List Type 1: Species/ Taxa Action Plan Type 2: No specifi c plan needed

Control List Type 1: Species/ Taxa Action Plan Type 2: No specifi c plan needed

Medium Risk National Permitted Plant List (certain species Plant Species in current trade permitted for sale subject to mandatory labelling. For new proposed plant species for trade, only those that are low risk plant would be permitted). National Mandatory Invasive Species Labelling Scheme

Low Risk Plant National Permitted Plant List (permitted for Species sale) Voluntary ‘Grow Me Instead’ Labelling Scheme

Defi nitions High Risk Invasive Plant Species refers to those introduced, and native plant species under cultivation that are known or have a reasonable probability of becoming invasive and harmful to agriculture and/or the environment in Australia. This would include plant species that are ‘transformers’, and/or have a direct impact on rare and threatened native species. A specifi c Weed Risk Assessment Score range may be used to assist determine high risk plant species.

Medium Risk Invasive Plant Species refers to those introduced, and native plant species under cultivation, that are known or have a reasonable probability of naturalising, and are or have a reasonable probability of being a minor to signifi cant problem in Australia. A specifi c Weed Risk Assessment Score range may be used to assist determine medium risk plant species.

Low Risk Plant Species refers to those introduced and native plants plant species that have been assessed as low risk to the environment and agriculture at present. A specifi c Weed Risk Assessment Score range may be used to assist determine low risk plant species.

variety, by breeding in a unique coloured The experience and process used to mobilise community support and involve- fl ower or stem into the sterile varieties for determine the Weeds of National Signifi - ment. The increase in awareness about the example (see Plan Point 4 for R&D com- cance, as well as the New Zealand experi- large impacts of weeds and where they are ponent). ence and process to develop the statutory coming from over the next several years ‘National Pest Plant Accord’ offer useful will lead to strong community expecta- 8. Phase out supply and trade of models that could assist develop the as- tions for governments and the garden in- high risk invasive plants nationally pect of the proposed National Noxious dustry to implement effective solutions to National invasive garden plant accord Weed List that pertains to high-risk inva- the invasive plant problem. There is a pressing need to phase out the sive garden plants. supply and trade of high risk invasive A national invasive garden plant ac- National voluntary Grow Me Instead plants nationally, particularly those that cord, or something analogous, is needed labelling scheme are not yet naturalised or widespread. for Australia, that follows the basic proc- A possible fl ip-side to the national manda- An example is the Australian Quarantine ess outlined below. tory invasive plant labelling scheme men- and Inspection Service (AQIS) quarantine tioned in Plan Point 7 is a national volun- weed, Ceylon hill cherry (Rhodomyrtus to- 9. Encourage gardeners to increase tary labelling scheme that leverages off the mentosa) that failed a weed risk assessment product demand for low risk garden ‘Grow Me Instead’ brand. It could promote and is now a prohibited import into Aus- plants low risk garden plants as an alternative tralia, is a serious weed in Florida and Ha- National weed awareness and education to higher risk garden plants. The scheme waii, but according to Aussie Plant Finder campaign could complement a set of educational 2004 is still advertised for sale in NSW and Currently there is low awareness about materials (e.g. brochures, posters, guides) Qld (Hibbert 2004). weed issues by urban Australians. How- developed for major urban centres. To achieve this aim requires a national ever, market research shows they have process that is scientifi cally robust, ena- strong latent demand to do adopt behav- Future-fi tting garden scheme bles all stakeholders to have a say, and iours that reduce weed spread risk, if it is Many gardens contain high risk invasive fi nancial implications to be fully consid- easy and convenient. garden plants, including those yet to es- ered. If for example, Plant Breeder Rights A national weed awareness and edu- cape widely into the environment, such as (PBR) are unduly impacted, industry has cation campaign is being planned by the Mexican feathergrass (Nassella tenuissima). strong grounds to negotiate a structural Australian Government under its ‘De- Encouraging home owners to participate adjustment package. feating the Weeds Menace’ program to in schemes that audit their gardens and

68 Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 help them replace high risk plants with reduce the weed spread risk and future the Australian Centre for Agriculture those that are environmentally safe needs costs posed by invasive garden plants. and Law, University of New England. to be part of the weed proofi ng solution. Australia deserves nothing less. WWF-Australia, Sydney, pp. 19-20. A council led model is the Greenweb pro- Moss, W. and Walmsley, R. (2005). Con- gram being implemented by local govern- References trolling the sale of invasive garden ments in Sydney, though apparently there Australian Institute of Horticulture(AIH) plants: why voluntary measures alone is also a garden centre led model – Wa- (2006). Weed Busting. Weed and Inva- fail, WWF-Australia Discussion Paper, terwise gardens – being trialled in West- sive Plant Species Policy 2006. AIH, WWF-Australia, Sydney. ern Australia (Rebecca Dawson, personal Bendigo. Neal, J. and Playford, J. (nd). Investigat- communication). Caton, B.P. (2005). Availability in Florida ing sterility in ornamental varieties of Getting in early even before these inva- nurseries of invasive plants on a volun- south east Queensland. sive plants have the chance to escape and tary “do not sell” list. US Department of Nursery and Gardening Australia (NGIA) become a problem is one of the most cost- Agriculture, Raleigh, USA. (2004). Submission to the Federal Sen- effective actions that we can take. This ‘fu- Edelman, R. (2003). The fourth Edelman ate Environment, Communications, ture-fi tting’ is analogous to homes install- survey on trust and credibility. Pre- Information Technology and the Arts ing more effi cient lights or showerheads to sented at the World Economics Forum, (ECITA) References Committee Inquiry save energy and water respectively. Davos, Switzerland, 23 January 2003. on Invasive Species. NGIA, Sydney. If widely promoted, this scheme has Glanznig, A. (2005). Closing Australia’s Sinden, J., Jones, R., Hester, S., Odom, D., the potential to create a new market for quarantine law loophole to new weeds. Kalisch, C., James, R. and Cacho, O. low risk plants. Now imagine if the NGIA WWF Issues Paper, WWF-Australia, (2004). The economic impact of weeds and/or AIH partnered with organisations Sydney. in Australia. Technical Series Paper No. like WWF and/or major farmer bodies to Glanznig, A., McLachlan, K. and Kessal, 8. CRC for Australian Weed Manage- promote this positive scheme. The WWF O. (2004). Garden plants that are inva- ment, Adelaide. panda is one of the 10 most trusted brands sive plants of national importance: an Sullivan, J., Williams, P., Cameron, E.K. in the developed world (Edelman 2003), overview of their legal status, commer- and Timmins, S. (2004). People and time and we have over 80 000 supporters in cial availability and risk status. WWF- explain the distribution of naturalized Australia. We also reach 100 000s of Aus- Australia, Sydney. plants in New Zealand. Weed Technol- tralians through our public campaigns, Glanznig, A., McLachlan, K., Kessal, O. ogy 18, 1330-3. such as ‘The Future is Man Made’ that and Casali, S. (In press). Garden plants Wilkenfeld G. and Associates Pty Ltd et includes invasive species as a campaign that are invasive plants of national sig- al. (2003). A mandatory water effi cien- issue. nifi cance: an overview of their legal cy labelling scheme for Australia. Final status, commercial availability and risk report prepared for Environment Aus- 10. Mobilise communities to search status. 2nd edition. WWF-Australia, tralia. and destroy new infestations of Sydney. Williams, B. (2004). Ban sought to block escaped invasive garden plants Groves, R., Boden, R., and Lonsdale, M. super weed invasion. The Courier Mail, National Weed Alert early warning and (2005). Jumping the garden fence: in- Thursday, 4 November 2004, p.11. rapid response system vasive garden plants in Australia and Early detection of new invaders is es- their environmental and agricultural sential for cost-effective intervention. impacts. CSIRO report for WWF-Aus- Evidence from Australia shows that most tralia. p. 31. (WWF-Australia, Sydney). new plant invaders are escaped invasive Groves, R.H., Hosking, J.R., Batianoff, garden plants that appear around popula- G.N., Cooke, D.A., Cowie, I.D., John- tion centres (Hosking et al. 2004). In New son, R.W., Keighery, G.J., Lepschi, B.J., Zealand, which has studied this pattern Mitchell, A.A., Moerkerk, M., Randall, in more detail, research found that of the R.P., Rozefi elds, A.C., Walsh, N.G. and fi rst collection of naturalised plant species Waterhouse, B.M. (2003). Weed catego- between 1985–2000, 91.5% were found ries for natural and agricultural ecosys- within 1 km of the nearest building and tem management. Bureau of Rural Sci- 67% were found within 2 km of a town ences, Canberra. (Sullivan et al. 2004). Hibbert, M. (2004). ‘Aussie plant fi nder Victoria already has a world leading 2004’, p. 279. (Florilegium, Sydney). ‘Weed Alert and Rapid Response System’ Hosking, J.R., Waterhouse, B.M. and Wil- (WARRS) in place. This needs to be rolled liams, P.A. (2004). Are we doing enough out nationally to encourage communities about early detection of weed species to take part in surveillance efforts. It has naturalizing in Australia? Proceedings the potential to build on urban Bushcare of the 14th Australian Weeds Confer- programs, since people already active in ence, eds B.M. Sindel and S.B. Johnson. restoring bushland may also be interested (Weed Society of NSW, Sydney). in monthly surveillance efforts to fi nd new Macdonald, I. (2005). More than 3000 plant invaders that may harm the bush potential weeds prohibited. Media re- they hold dear. Garden centres could also lease by the Minister for Fisheries, For- promote local surveillance efforts. estry and Conservation, 1 June 2005. DAFF05/102M. http://www.mffc.gov. Conclusion au/releases/2005/05102m.html. The 10 point plan proposed in this paper Martin, P., Verbeek, M., Thomson, S. and has the potential to strategically reposition Martin, J. (2005). The costs and bene- governments, the garden industry, NGOs fi ts of a proposed mandatory invasive and communities, so that they are work- species labelling scheme, A Discussion ing together collaboratively to markedly Paper prepared for WWF-Australia by

Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 69 The judicious purchase of important private herbaria (such as those of Otto The role of the National Herbarium of Victoria (MEL) Sonder, James Drummond and Joachim 1 Steetz) contributed invaluable foreign ma- in the documentation of new weeds terial, as well as some of the earliest Aus- tralian collections, to Mueller’s botanic Val Stajsic and Alison C. Vaughan, National Herbarium of Victoria, Royal museum (Cohn 2003). These acquisitions Botanic Gardens Melbourne, Birdwood Avenue, South Yarra, Victoria 3141, provided valuable comparative mate- Australia. rial that allowed the Australian fl ora to be accurately described. As well as being a valued resource for botanists around the world, the foreign collections are a vital Abstract Any attempt to document new of Ferdinand Mueller as the fi rst Govern- tool in the identifi cation of newly natural- weed species is reliant upon the accurate ment Botanist of the colony of Victoria. As ised taxa in Victoria. identifi cation of the plants concerned. This Government Botanist, Mueller’s directive The collection continued to grow after can only be achieved with access to taxo- was to investigate Victoria’s fl ora – a brief Mueller’s time, though not at the same nomic literature, well-curated reference which remains a fi tting description of the phenomenal rate. Significant additions specimens and expert botanical knowl- work of the National Herbarium today. include the purchase of herbaria from Fe- edge. The National Herbarium of Victoria Born in Germany in 1825, Mueller lix Reader, Max Koch and Raleigh Black (MEL) is the central authority on the plant trained as a pharmacist at the University (Short 1990), while continuing efforts from biodiversity of Victoria and the chief re- of Kiel and became a skilled botanist (Mor- herbarium staff; exchange of specimens pository of plant specimens. A major role ris 1974). Upon arriving in South Australia with other herbaria; and, notably, contri- of the Herbarium is to collect and docu- in 1847, Mueller began enthusiastically in- butions from Cliff Beauglehole, Jim Willis, ment the State’s fl ora, including both na- vestigating the Australian fl ora. His botan- and the Victorian Field Naturalists, also tive and naturalised species. The Herbar- ical pursuits took precedence over other enhanced the collection (Cohn 2005). ium houses a collection of approximately endeavours, and he amassed a private Today the National Herbarium contains 1.2 million pressed and dried plants, and collection of over 4000 specimens. These around 1.2 million specimens – approxi- is particularly rich in historic and foreign- specimens provided the foundation for mately 800,000 Australian specimens, and collected specimens. These specimens pro- the Melbourne collection, which, in 1894, 400,000 collected overseas – and includes vide an invaluable permanent record of was estimated to number over one million representatives of most of Australia’s vas- the occurrence of a plant species at a par- specimens (Cohn 2003). cular fl ora, as well as algae, bryophyte, ticular locality and time. The collections Mueller was an exceptional botanist lichen and fungi collections of interna- held at the National Herbarium, together and a prolifi c collector, and became the fo- tional signifi cance. It is one of the largest with its comprehensive botanical library cus of botanical endeavours in Australia. herbarium collections in Australia, and and wealth of botanical expertise, make it Upon being appointed Government Bota- the richest in historic and type specimens an essential resource for research on Vic- nist, he was quick to undertake botanical (Willis 1990). Approximately 37% of the toria’s fl ora. explorations, and collected around 1500 Australian collection is Victorian, and pro- previously unrecorded species for Victo- vides a permanent and verifi able record Introduction ria in his fi rst 15 months in offi ce (Willis of the State’s fl ora. Records of all species, The National Herbarium of Victoria (MEL) 1990). including new and emerging naturalised is the central authority on the plant biodi- In addition to his own efforts, Mueller’s species, are substantiated by herbarium versity of Victoria. Situated at the Royal collections were supplemented by materi- voucher specimens, making the National Botanic Gardens in South Yarra, the her- al obtained from a wide variety of sources. Herbarium a valuable resource for infor- barium houses the State Botanical Collec- Specimens collected on both offi cial and mation on weedy taxa. tion – a collection of approximately 1.2 private exploring expeditions (includ- million pressed and dried plant speci- ing those of Ludwig Leichhardt, Ernest The documentation of Victoria’s mens; the most comprehensive library Giles, John McDouall Stuart and William weed fl ora at MEL of botanical literature in Australia, and a Tietkens) were sent to Mueller for identifi - Plant invasions, along with land-clearing signifi cant collection of botanical artwork. cation and incorporated into the growing and global warming, are a major contrib- The State Botanical Collection is an es- collection. Mueller also received speci- uting factor to the loss of global biodiver- sential resource for research on Victoria’s mens from prominent botanists in other sity. Naturalised taxa comprise approxi- fl ora. colonies; through the exchange of material mately 20% of the total Australian fl ora The National Herbarium is the oldest with overseas collectors, and from enthu- (Hosking personal communication 2005) scientifi c institution in the State. It was siastic individuals throughout the colonies and 30% of the Victorian fl ora (Walsh and founded in 1853 with the appointment who he engaged as plant collectors. Stajsic 2007). More than 65% of Victoria is invaded by wholly or predominantly exotic vegetation (Stuwe 1986, Carr et al. 1992). The area occupied by naturalised Footnote species increases annually, and records of 1 The term ‘weeds’ in this paper includes both naturalised and incipiently naturalised newly established species continually ac- plants. As defi ned by Walsh and Stajsic (2007), naturalised plants are those alien plants cumulate (Carr 1993). Hosking (personal that sustain self-replacing populations without direct intervention by people or in spite communication 2004) estimated that 13 of human intervention, by recruitment from seeds or vegetative propagules (e.g. the new weeds establish in Australia each bulbils of many exotic Oxalis species) or by vegetative spread (e.g. the extensive rhizome year. The National Herbarium of Victoria system of *Spartina × townsendii). Incipiently naturalised taxa are those where the taxon is regards the issue of plant invasions and known to be not indigenous in Victoria and is represented by one or more populations, the corresponding threats that are posed but the extent of naturalisation is uncertain and there is doubt whether it has become to native ecosystems and biodiversity as a truly naturalised (as defi ned above) yet. Taxa in this category demonstrate the potential critical conservation issue. to become truly naturalised.

70 Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 Documenting the State’s fl ora 6000 Naturalis ed The documentation of new weeds is de- pendent on a thorough understanding of 5000 Nativ e both the native and the naturalised fl ora that occur in the State; we can’t know if a taxon is new to the State if we don’t know 4000 what taxa are already present. One of the most important resources 3000 produced by the National Herbarium is ‘A Census of the Vascular Plants of Victoria’2.

Number of taxa 2000 The Census, now in its eighth edition, lists the scientifi c names of all vascular plant taxa known to occur in the State. All taxa 1000 listed in the Census are substantiated by herbarium voucher specimens, most of 0 which are housed at MEL, although some 1984 1990 1993 1996 2000 2003 2007 are housed in other recognised herbaria (Walsh and Stajsic 2007). Documenting the Year of publication occurrence of new weeds in publications such as the Census plays a pivotal role in Figure 1. Native and naturalised vascular plants in Victoria (from ‘A Census early intervention strategies and weed of the Vascular Plants of Victoria’, editions 1–8). Data from the 1988 Census alert procedures. was omitted due to an error in the fi gures. An examination of past editions of the Census reveals valuable data about weed incursions in Victoria. As illustrated in Figure 1, the proportion of naturalised to native fl ora recorded in the Census has 80 increased steadily in successive editions, 65 with the greatest proportional increase in naturalised taxa occurring between the 60 seventh (2003) and eighth (2007) editions. 47 It is interesting to note that not all of the 132 additional naturalised taxa listed 40 in the 2007 Census were newly established in the period since the publication of the 18 previous edition. Many of these taxa be- taxa of Number 20 came naturalised many years prior to 2003 (Figure 2), but had previously escaped de- 2 tection or classifi cation as ‘naturalised’. 0 The systematic databasing of the Aus- pre – 1975 1975 – 1995 1996 – 2003 2004 – 2007 tralian Collection at MEL – initially un- der the J.T. Reid project (2000-2001) and Year of collection later by the Australia’s Virtual Herbarium (AVH) project (2001-2006) – has greatly fa- Figure 2. Collecting date of naturalised taxa in Victoria fi rst recorded in ‘A cilitated the detection of weedy taxa in the Census of the Vascular Plants of Victoria’(2007). herbarium collection that had previously escaped attention, and thus been omitted from early editions of the Census. It is important to note that the natural- or eradication procedures can be initiated Another infl uential development has ised status accorded to the taxa in Walsh prior to an authoritative determination be- been the burgeoning interest in the en- and Stajsic (2007) is based strictly on the ing provided by a herbarium taxonomist. croachment of weeds in Victoria as an im- evidence available from specimens held This is one of the pivotal roles that her- portant conservation issue at MEL. The at MEL. Although some taxa that are listed baria play in early detection and notifi ca- greater consideration of our weed fl ora as incipiently naturalised may in some in- tion protocols. prompted a refi nement of the terminology stances be fully naturalised, MEL does not In order to identify plant specimens cor- used in the Census, and the introduction hold voucher specimens that verify that rectly, it essential to possess a knowledge of the ‘incipiently naturalised’ category in status. It is vital that voucher specimens of taxonomic methods, extensive knowl- Ross (2000). This term is used to describe of naturalised taxa are lodged in herbaria edge of the taxonomic literature, and taxa that are known to be introduced in in order to accurately and verifi ably docu- experience in the identifi cation of plants Victoria, but are not yet known to be fully ment weed establishment and expansion. (Jones and Luchsinger 1979). Knowledge naturalised. In previous editions of the of the horticultural fl ora is particularly im- Census, taxa were generally listed only if Identifying new weeds portant given that almost 80% of recently they were fully naturalised. Consequently, Plant identifi cation (also called determina- recorded naturalisations in Victoria are of many taxa which are now included as in- tion) and the correct application of a name garden origin (Walsh and Stajsic 2007). Re- cipiently naturalised were absent in earlier (nomenclature) are integral components liable determination can only be achieved editions. of all taxonomic work. It is important to by having access to comprehensive taxo- appreciate that, in formal processes of nomic literature, and a comprehensive early detection and notifi cation of new herbarium collection that includes native, Footnote naturalisations, such as the Weed Alert naturalised and horticultural taxa. 2 Hereafter referred to as the Census. Rapid Response (WARR) plan, no control

Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 71 Determination Herbaria (CHAH) agreed to produce a new applied to specimens in the State Botani- When confronted with a new naturalised cooperative census of Australian vascular cal Collection are up-to-date, the National plant, an experienced botanist can usually plants. This project is actively underway, Herbarium also provides nomenclatural recognise the plant to either genus or fami- and it is hoped that the Australian Plant advice, and can help guide external cli- ly level (if not species level). Following the Census (APC) will standardise the appli- ents through some of the more complex initial recognition, the unidentifi ed speci- cation of botanical names in Australia (for nomenclatural issues. men needs to be keyed out using botanical more information, see: http://www.anbg. literature such as fl oras, monographs and gov.au/chah/apc/introduction.html). The importance of getting the revisions, where available. Specimens are Botanical name changes are often a identifi cation right then compared with species descriptions source of confusion; however, they are an The Herbarium’s plant collection and li- and any available illustrations. important component of plant taxonomy, brary form a priceless storehouse of scien- Once we obtain a name for the uniden- which, like any other science, is dynamic tifi c information that is an indispensable tifi ed specimen, it is then compared with and constantly evolving. Plant names are aid for identifying new weeds. The advan- a voucher specimen of the taxon, if one usually changed for one of the following tage of using herbarium records is that they is held at the Herbarium. This is where two reasons: provide a permanent and verifi able record MEL’s extensive foreign collection, unique 1. Changed views on species delimitation: that substantiates the presence of a plant at among Australian herbaria, provides an Two species, previously considered dis- a particular place and time. Specimens can invaluable source of comparative material tinct, may, with new evidence, turn out be re-examined if there is dispute over the that greatly assists the identifi cation of ex- to be the one taxon. Only one name can identity of the record. When document- otic species. be applied to a single taxon, so the ear- ing the fl ora of a given area one should be Due to the lack of comprehensive global liest published name will take priority, mindful of the fact that sight records not revisions, specimens from some taxonomi- and the other name should no longer substantiated by voucher specimens are cally diffi cult groups sometimes need to be used. On the other hand, what was not verifi able, and may present problems be sent to botanists at overseas herbaria considered a single species might turn to future researchers (Albrecht 1993). for determination. A Callitriche specimen out to include more than one taxon. If The National Herbarium encourages received at MEL, apparently of a taxon so, new names must be applied to the those with an interest in naturalised plants new to Victoria, provides a recent exam- newly recognised taxa, in accordance to submit good quality, well preserved ple. The specimen was originally believed with the rules set out in the Interna- specimens with adequate fi eld notes. This to be *C. marginata Torr., and was sent to tional Code of Botanical Nomenclature maximises the likelihood of obtaining a an authority on the genus in England for (ICBN). One of the greatest benefi ts of confident determination and provides confirmation of the identification. The collecting voucher specimens is that us with the option of incorporating the specimen was redetermined to *C. defl exa they enable previous identifi cations to specimens into the State Botanical Collec- A.Br. Such misdeterminations can occur be verifi ed or amended when species’ tion if desired (Albrecht 1993). Incomplete where there is poor literature and the lack delimitations change, whereas it is of- or poorly preserved specimens with in- of keys that include all the known taxa. In ten impossible to establish the current adequate accompanying information are another instance, material of an invasive name for identifi cations based solely on often impossible to determine with con- alien, Ludwigia, that was collected from the sight records. fi dence. Guidelines for the preparation outer eastern suburbs was determined at 2. Changes in classifi cation: Changes in and submission of voucher specimens can MEL as *Ludwigia arcuata × repens; the de- plant groupings may result in a taxon be found on the Royal Botanic Gardens termination was confi rmed by an author- being transferred from one genus to an- website: http://www.rbg.vic.gov.au/re- ity on Ludwigia in Taiwan. other. For example, based on the latest search_and_conservation/plant_identifi - evidence, *Achnatherum caudatum (Trin.) cations/voucher_specimens. Nomenclature S.W.L.Jacobs & J.Everett has now been Many people do not submit specimens During plant determination work, one transferred to a new genus, Amelichloa, to a herbarium for identifi cation because may notice that the taxonomic literature and is now referred to as *Amelichloa they assume that they know the identity might cite one or more synonyms for caudata (Trin) Arriaga & Barkworth. of the plant they’re dealing with (Hosk- certain species which are treated in that ing et al. 1996). Failing to correctly identify account. Assuming that a taxon hasn’t It is important to note that, in Australia, a specimen can result in the inadvertent been divided into several other taxa, most many weed names have been misapplied. dispersal and establishment of weedy spe- synonyms are easily correlated with the A misapplication of a name occurs when a cies, and delay their eradication or control. currently accepted name. Nomenclatural plant has been initially incorrectly identi- The potential implications of misidentify- checks are usually less labour intensive fi ed, and the wrongly-applied name is sub- ing plants are highlighted by the following than the determination process, but there sequently published in botanical literature. examples. Several years ago, the weedy al- are exceptions to this, and the nomencla- For example, the name *Physalis viscosa L. ien *Lampranthus tegens (F.Muell.) N.E.Br. ture of some taxa can be very complex and has been misapplied in Australia, and re- was planted at Moonee Ponds Creek in ur- convoluted. determined as several different taxa, with ban Melbourne, in the belief that it was the The purpose of nomenclature is to facil- the Victorian populations now considered native pigface, Disphyma crassifolium (L.) itate unambiguous communication about to be *Physalis hederifolia A.Gray (J. Sullivan L.Bolus. In a similar case, MEL received a a taxon, i.e. to ensure that we all refer to personal communication 2007). Other ex- specimen of *Amelichloa caudata (Trin.) Ar- a particular taxon using the same name. amples of misapplied weed names in Aus- riaga and Barkworth (syn. *Achnatherum This usually works well, but there can be tralia include: *Carthamus glaucus M.Bieb., caudatum (Trin.) S.W.L.Jacobs and J.Everett) differences of opinion about nomenclature misapplied to *Carthamus leucocaulos Sm; about two or three years ago, which had and taxonomic concepts between taxono- *Leucanthemum maximum (Ramond) DC., been planted along a stream in Wangaratta mists. Consequently, different names may misapplied to *Leucanthemum × superbum in the belief that it was the native tussock be applied to the same taxon in different (Bergmans ex J.W.Ingram) D.H.Kent; and grass, Poa labillardierei Steud. states (or different countries). *Lantana camara L. var. camara, misapplied A number of native plant nurseries in This lack of uniform nomenclature can to *Lantana × strigocamara R.W.Sanders. Melbourne have propagated and used sometimes cause confusion. That is why, in As well as keeping track of nomenclat- the alien *Carpobrotus aequilaterus (Haw.) 2004, the Council of Heads of Australasian ural changes and ensuring that the names N.E.Br. in revegetation schemes in the

72 Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 belief that they were using the native Car- 24% were primarily fortuitous fi nds, de- discolor Andrews at Pomonal (near Gram- pobrotus rossii (Haw.) Schwantes (Stajsic tected mainly by members of the public. pians National Park), which yielded sev- personal observations). Similarly, one cli- One of the chief means by which bota- eral other new weed records for Victoria, ent submitted a specimen of a Carex ob- nists at MEL detect new weeds is by tar- including Erica glandulosa Thunb. subsp. tained from a native plant nursery in Mel- geting areas that have proven likely to glandulosa and Hakea prostrata R.Br. bourne, which was distributing the plant contain previously unrecorded natural- Although the above high risk areas are as Carex incomitata K.R.Thiele (an uncom- ised taxa. Hosking et al. (2004) argue that targeted by MEL botanists and others with mon native in the Melbourne area); the an increased detection effort should ini- a good knowledge of, and concern for, the plant turned out to be the invasive alien tially target areas around population cen- native and naturalised fl ora, there are cur- *Carex divulsa Stokes subsp. divulsa. tres because most plants fi rst naturalise rently no systematic and regular surveys More recently, MEL received specimens in urban bushland and areas with a high of these areas in Victoria. Unless we col- of Acacia cyclops A.Cunn. ex G.Don from concentration of gardens. Many new in- lect, document and positively identify Barwon Heads near Geelong, for deter- cursions originate from discarded garden new naturalisations, those responsible for mination. This non-Victorian species had refuse, or from propagules dispersed from managing the impact of weeds will be un- been propagated and used in revegetation gardens to native bushland by birds. In aware of potential new problems in their schemes in the belief that it was the native Victoria, two such areas that have yielded jurisdictions (Hosking et al. 2004). Acacia uncifolia (J.M.Black) O’Leary (Reid many new weeds over the last ten years and Murphy in press 2008). This was the are the Dandenong Ranges, and the Mt Accessing the resources of the fi rst report of this invasive species for Vic- Macedon area. National Herbarium of Victoria toria. Revegetation areas, particularly road- The National Herbarium of Victoria is an side plantations and soil conservation invaluable and irreplaceable source of How are new weeds detected? schemes, are another rich source of new verifi able information on the State’s fl ora. Seventy-nine percent of the additional al- weeds. Many of the taxa that naturalise It is a dynamic collection that relies on the ien taxa listed in Walsh and Stajsic (2007) at these sites are Australian natives. Carr submission of good quality voucher speci- since 2003 are regarded as being incipi- et al. (1992) reported that at Anglesea in mens to maintain an up-to-date reference ently naturalised. This strongly supports Victoria 36 of a total of 45 species planted collection representative of Victoria’s na- the view that it is essential to detect new in two revegetation areas were present as tive and naturalised fl ora. weed incursions early, while ‘effective ac- escapes in adjacent heathland and heathy When Mueller founded the National tion is still possible and before the cost of woodland. Many of the Australian na- Herbarium, he envisioned that it would be control escalates and the weed infestation tive naturalised taxa listed in both Ross ‘at all times accessible to the public’ (Willis has compromised natural values’ (Tim- and Walsh (2003) and in Walsh and Stajsic 1990). Unfortunately, a lack of resources mins and Braithwaite 2002). Early detec- (2007), including Acacia, Eucalyptus, Hakea has meant that this has rarely been the tion maximises the likelihood of success- and Melaleuca species, were collected case throughout the Herbarium’s history; fully eradicating the weed, and reduces from roadside plantations. At the White however, access to the collections has im- the negative effects on the natural envi- Elephant Reserve in the Parwan Valley proved dramatically in recent years, with ronment. (near Bacchus Marsh), several Western the databasing of the Australian plant col- Since 2002, MEL has collaborated with Australian Eucalyptus species (including lections as part of the Australia’s Virtual the Department of Primary Industries Eucalyptus astringens (Maiden) Maiden, Herbarium (AVH) project. This collabo- (DPI) in the WARR plan. The chief aims E. kondininensis Maiden and Blakely, and rative project has allowed unprecedented of the plan are to detect new incursions of E. occidentalis Endl.) have become natu- online access to the species’ distribution new weeds before they become natural- ralised. Other Australian natives natural- data held in Australia’s major govern- ised, and to instigate rapid notifi cation of ised at this site include Callitris endlicheri ment-funded herbaria. The AVH website DPI staff in order to enable early remedial (Parl.) F.M.Bailey, C. columellaris F.Muell. can be accessed via the Royal Botanic Gar- or eradication strategies. One of the poten- and Casuarina glauca Sieber ex Spreng. dens portal, at http://www.rbg.vic.gov. tial benefi ts of this partnership for MEL is Several new weed taxa recently re- au/avh/. the receipt of herbarium voucher speci- corded for the fi rst time in Australia were Requests for more detailed informa- mens to substantiate new weed records. found at alpine ski resorts in Victoria. tion from MEL’s collection are serviced at MEL botanists also provide expert advice These include Hieracium praealtum Vill. ex the discretion of the Collections Manager, and warning of potential new weeds by Gochn. subsp. bauhinii (Besser) Petunn. at and charges may be incurred. Enquiries helping compile ‘target lists’ of taxa that Falls Creek, and Juncus ensifolius Wikstr., should be directed to [email protected]. are either known to be serious environ- which is abundant around Baw Baw Vil- gov.au. Access to the Library and Collec- mental and economic weeds, but have not lage and has spread along roadsides, ski tions is available to accredited researchers yet been recorded in Australia, or that have runs and along the East Tanjil River (Stajsic by appointment. been recorded as weeds in other States, and Hosking personal observation 2005). The Identifications and Information but, in Victoria, are currently known only Juncus ensifolius Wikstr. has also been dis- Service provides plant identifications, from limited populations. Target lists are covered at other ski resorts in Victoria. nomenclatural advice and information used to promote awareness of potential Regional areas of long established ag- on plant distributions, and is the contact threats, and to focus survey efforts (Wa- riculture should also be targeted, which is point for people wishing to submit vouch- terhouse 2003). something that we are conscious of at MEL, er specimens to the Herbarium. The serv- Hosking et al. (2004) point out that the and aim to do. These sites are subject to ice is open from 10:00 AM – 1.00 PM each majority of new plant naturalisations are signifi cant movement of stock, stock feed weekday. Please direct any correspond- detected or fi rst recognised by those with and agricultural machinery from around ence to: a good knowledge of both the native and the state and throughout Australia, which Identifi cations and Information Serv- exotic fl ora. This observation is supported may act as vectors for new weeds. ice, Royal Botanic Gardens Melbourne the Census data, where 76% of newly list- MEL botanists also work with DPI staff Private Bag 2000, South Yarra Victoria ed naturalised taxa in Walsh and Stajsic in conducting targeted searches for poten- 3141. Ph: (03) 9252 2315. (2007) were fi rst detected by herbarium tially invasive species. A recent example Copies of ‘A Census of the Vascular Plants staff or other botanists, or by knowledge- of this cooperation was an inspection of of Victoria 8th edition’ (2007) can be ob- able members of the public. The remaining a newly discovered population of Erica tained by contacting the Identifi cations

Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 73 and Information Service at the above ad- Ross, J.H. (2000). ‘A census of the vascular dress. plants of Victoria 6th edition’. (Royal Botanic Gardens, South Yarra). References Ross, J.H. and Walsh, N.G. (2003). ‘A cen- Albrecht, D. (1993). ‘Collecting and pre- sus of the vascular plants of Victoria 7th serving herbarium specimens.’ (Na- edition’. (Royal Botanic Gardens Mel- tional Herbarium of Victoria, Royal bourne, South Yarra). Botanic Gardens, South Yarra). Short, P.S. (1990). Politics and the purchase Carr, G.W., Yugovic, J.V. and Robinson, of private herbaria by the National K.E. (1992). ‘Environmental weed inva- Herbarium of Victoria. In ‘History of sions in Victoria: conservation and man- systematic botany in Australasia: pro- agement implications.’ (Department of ceedings of a symposium held at the Conservation and Environment Ecolog- University of Melbourne, 25–27 May ical Horticulture, East Melbourne). 1988’, ed. P.S. Short, pp. 5-12. (Aus- Carr, G.W. (1993). Exotic fl ora of Victoria tralian Systematic Botany Society Inc., and its impact on indigenous biota. In South Yarra). ‘Flora of Victoria’, Volume 1: Introduc- Stuwe, J. (1986). An assessment of the con- tion, eds D.B. Foreman and N.G. Walsh, servation status of native grasslands pp. 256-97. (Inkata Press, Melbourne). on the western plains of Victoria, and Cohn, H.M. (2003). 150 years: the National sites of botanical signifi cance. Environ- Herbarium of Victoria, 1853-2003. Muel- mental Studies Program No. 412. De- leria 17, 3-14. partment of Conservation, Forests and Cohn, H.M. (2005). The close union be- Lands, Heidelberg. tween the Herbarium and the Natural- Timmins, S.M. and Braithwaite, H. (2002). ists. Victorian Naturalist 122, 282-9. Early detection of weeds on islands. Forbes, S.J., Gullan, P.K., Kilgour, R.A. and In ‘Turning the tide: the eradication of Powell, M.A. (1984). ‘A census of the invasive species’, eds C.R. Veitch and vascular plants of Victoria.’ (National M.N. Clout, pp. 311-18. (IUCN SSC In- Herbarium of Victoria, Melbourne). vasive Species Specialist Group. IUCN, Forbes, S.J. and Ross, J.H. (1988). ‘A census Cambridge). of the vascular plants of Victoria 2nd Walsh, N.G. and Stajsic, V. (2007). ‘A cen- edition.’ (National Herbarium of Victo- sus of the vascular plants of Victoria’ ria, Melbourne). 8th edition. (Royal Botanic Gardens Hosking, J.R., Sainty, G.R. and Jacobs, Melbourne, South Yarra). S.W.L. (1996). Certainty and uncertain- Waterhouse, B.M. (2003). Know your ty in plant identifi cation. Proceedings enemy: recent records of potentially of the 11th Australian Weeds Confer- serious weeds in Northern Australia, ence, ed. R.C.H. Shepherd, pp. 464-7 Papua New Guinea and Papua (Indo- (Weed Science Society of Victoria, Mel- nesia). Telopea 10 (1), 477-85. bourne). Willis, J.H. (1990). Melbourne: a focal point Hosking, J.R., Waterhouse, B.M. and Wil- for early botanical activity. In ‘History liams, P.A. (2004). Are we doing enough of systematic botany in Australasia: about early detection of weed species proceedings of a symposium held at naturalising in Australia? Proceedings the University of Melbourne, 25-27 May of the 14th Australian Weeds Confer- 1988’, ed. P.S. Short, pp. 1-3. (Australian ence, eds B.M. Sindel and S.B. Johnson, Systematic Botany Society Inc., South pp. 68-72. (Weed Society of NSW, Syd- Yarra). ney). Jones, S. Jr. and Luchsinger, A.E. (1979). ‘Plant systematics.’ (McGraw-Hill Book Company, New York). Morris, D. (1974). Mueller, Sir Ferdinand Jakob Heinrich von [Baron von Muel- ler] (1825 – 1896). In ‘Australian Diction- ary of Biography, Volume 5’, pp. 306-8. (Melbourne University Press, Carlton). Reid, J.C. and Murphy, D.J. (in press 2008) Some case studies of Acacia as weeds and implications for herbaria. Muelleria 26. Ross, J.H. (1990). ‘A census of the vascular plants of Victoria 3rd edition’. (National Herbarium of Victoria, South Yarra). Ross, J.H. (1993). ‘A census of the vascular plants of Victoria 4th edition’. (Royal Botanic Gardens, South Yarra). Ross, J.H. (1996). ‘A census of the vascular plants of Victoria 5th edition’. (Royal Botanic Gardens, South Yarra).

74 Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 for business purposes and deliberate or in- advertent introduction of plants, seeds or Improving provincial Victoria’s biosecurity by plant parts by community members. Dif- ferent industries are involved in activities minimising the risk of new weed introductions that can potentially introduce and spread new weeds. As gardens are the most sig- nifi cant source of new weed introductions, A B Mark Farrer and Stephen Young the IPVB project will focus its engagement A Department of Primary Industries, State Government Offi ces, 23 Patrick activities on the management of new weed Street, Stawell, Victoria 3380, Australia. introduction pathways associated with the B Department of Primary Industries, Ballarat, Victoria 3350, Australia. garden industry and community garden interest groups in lifestyle regions. The IPVB project aims to understand the risk, location and introduction path- ways of Victorian Alert Weeds. Educat- Summary Recent research into the spread community activities associated with these ing and working strategically with the of weeds has found ordinary gardens are growth areas that are linked to pathways community, industries and agencies can a major source of new weed introductions of introduction of new weeds are the focus reduce the risk of damaging new weeds (Groves 2005). As part of the Victorian of this project. being introduced resulting in signifi cant Government’s ‘Moving Forward: Mak- This paper presents a summary of a cost savings to agriculture, business and ing Provincial Victoria the Best Place to working project ‘Improving Provincial the environment. Live, Work and Invest’ initiative, the De- Victoria’s Biosecurity – High Priority Vic- partment of Primary Industries (DPI) is torian Alert Weeds in High Risk Provincial Approach delivering a four year, $3.6 million project Areas’ that is working collaboratively with One of the most signifi cant transformations entitled ‘Improving Victoria’s Provincial these industries and communities. taking place in Victoria is the expansion of Biosecurity’ (IPVB) which aims to reduce the ‘new’ urban lifestyle (amenity) regions the risks surrounding the introduction Assessing Victorian Alert Weeds through a growing demand for rural living of high priority Victorian Alert Weeds in To work most effectively to protect Victo- (Barr 2005). The lifestyle (amenity) region identifi ed provincial urban and lifestyle ria from potential pest plant biosecurity provides a key focus for preventing the regions in collaboration with DPI’s Weed threats, the IPVB project is working with introduction and spread of new pests to Alert Program. Social research has identi- a suite of weeds referred to as Victorian Victoria. New species introduced through fi ed that lifestyle regions are expanding Alert Weeds. These are new weeds to Vic- garden and landscape plantings can es- (Barr 2005), increasing the threat of new toria that have the potential to threaten the cape across the interface and threaten bio- weed introductions in provincial Victoria. State’s environment and agriculture. Some diversity, waterways, agricultural produc- Strategic engagement with industries of these are thought to be naturalised in tion and community assets. Engagement and communities whose activities are small numbers but still eradicable, others of agencies, industries, community groups linked to high risk weed introduction are yet to reach Victoria but present a sig- and Weed Spotters involved in high risk pathways in provincial Victoria provides nifi cant risk if they were to arrive and nat- introduction pathways, provides an op- an opportunity for proactive surveillance uralise (Victorian Weed Alert Plan 2007). portunity for pro-active surveillance and and adoption of preventative actions to As these weeds are not declared under the adoption of preventative actions that will reduce weed introduction. Catchment and Land Protection Act 1994, reduce the risk of introduction. Keywords: Provincial Victoria, lifestyle there is no legislative barrier preventing The project comprises two major com- regions, introduction pathways, Victorian their promotion, trade and movement ponents - the industry and agency com- Alert Weeds, Improving Provincial Victo- around the State. These plants are being ponent of the project that will have a fo- ria’s Biosecurity project (IPVB). assessed using the Victorian Pest Plant cus across the central Victorian ‘Lifestyle Prioritisation Process, to identify the most Amenity Region’ and a more spatially dis- Introduction serious threats. The distribution of each crete community component focusing on Pest plants are key biosecurity threats to species is validated through surveillance smaller priority regions (Figure 1). provincial Victoria’s environmental, agri- to inform appropriate management strate- The ‘IPVB – High Priority Victorian cultural and community assets. In recent gies. The most serious will be recommend- Alert Weeds in High Risk Provincial Are- years, action to improve Victoria’s sur- ed for declaration as noxious weeds. as used a Priority Assessment Framework veillance and response to invading weeds to develop a ranked list of priority areas. through the Victorian Pest Management Recognising invasion pathways This comprised; potential/current distri- – A Framework for Action (VPMF) Weed Invasion pathways are any means that bution of previously assessed Victorian Management Strategy, has seen the devel- lead to entry or spread of pests. The pri- Alert Weeds, proximity to state borders opment and delivery of capacity building mary objective of a pathways analysis is to and the major Hume Highway corridor, programs with key stakeholders on both identify the species, assess the probability the number of households and potential public and private land to support a na- of its entry, the pathways by which it may gardens as an indicator of potential intro- tional biosecurity approach. enter Victoria or by which it may spread, duction points and level of interface with Parts of regional Victoria are enjoying and subsequently the consequences of its social, environmental and agricultural as- a period of rapid growth in population, introduction. Pathway management is the sets. infrastructure, industry and investment. most effective way to address unintention- Recent research into pest plant origins al introductions (King 2006). Industry and community has identifi ed gardens as major sources Weed spread pathways are many and engagement of pest plant introduction (Groves 2005). varied, examples include: naturally oc- DPI works proactively with a range of Urban environments and their many gar- curring (waterways), movement of inad- Victoria’s key pest management stake- dens that interface with lifestyle regions vertently contaminated plant, equipment, holders and has established strong part- are an area of ongoing weed introduction goods, livestock or produce, deliberate in- nerships with local government, industry, risk in provincial Victoria. Industry and troduction of plants, seeds and plant parts and the community. Understanding the

Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 75 Figure 1. Focus areas for industry (shaded) and community groups (striped) across 17 municipalities for tackling the introduction of high risk Victorian Alert Weeds.

needs and drivers for each stakeholder and industry participants implementing Acknowledgements group has been essential to developing risk mitigation strategies for introduction Thank you to the Victorian Government this relationship. The business activities of Victorian Alert Weeds. for providing funding to the project and of many organisations have the potential The selection of engagement activities the following people for their input and to introduce new weeds into Victoria. For and products used to infl uence change supply of information: Catriona King, example when a pet shop sources aquatic will be tailored to meet the specifi c needs Helen Anderson, Neil Smith, Beth Jones plants from interstate. This level of risk of each stakeholder as identifi ed in the and Tony Lovick. has signifi cantly increased in recent times, stakeholder analysis. particularly where provincial urban areas References interface with rural lifestyle areas. It is Discussion Barr, N. (2005). The changing social land- important to understand and work with The IPVB initiative will invest in tackling scape of Rural Victoria – 2005. Depart- these key segments of the community and the introduction pathways of Victorian ment of Primary Industries, Tatura. industry to reduce the risks of introducing Alert Weeds as opposed to the more tradi- Groves, R.H., Boden, R. and Lonsdale, new weeds into Victoria. tional management focus of tackling estab- W.M. (2005). Jumping the garden fence: The IPVB project aims to infl uence short lished weeds in rural segments. The IPVB invasive garden plants in Australia and term and long term behaviour change re- initiative will be delivered in an environ- their environmental and agricultural lating to weed management, in order to ment of signifi cant change. Globalisation, impacts. CSIRO report prepared for reduce the risk of introduction and incur- land use change and environmental events WWF-Australia. WWF-Australia, Syd- sion of high risk Victorian Alert Weeds such as drought, are beyond the infl uence ney. in identifi ed lifestyle regions. Industry of the project. It is within this context that King, C. (2006). Weed spread pathways mapping and stakeholder analysis has the IPVB project seeks to infl uence change risk assessment (Tackling Weeds on identifi ed the stakeholders that are in the in project stakeholder behaviour over the Private Land initiative). Victorian Weed best position to assist with making these long term. To provide certainty of project Alert Plan (2007). changes. They include environmental and impacts in such a dynamic environment non-environmental groups, industries and a project management framework has government agencies that can improve been developed. A key component of this surveillance capacity to spot and report framework is a project evaluation plan Victorian Alert Weeds through the Weed which will measure project progress and Spotter network. They also include plant its impact upon completion. The project interest groups, garden industry segments framework will also take into account the and government agencies who can assist need for fl exibility to respond to infl uences with the development and implementa- beyond the control of the project. tion of mitigation strategies in high risk With these project management proc- introduction pathways. esses in place over the next three years, The IPVB project has undertaken stake- the IPVB project will build upon previous holder analysis to inform the development initiatives and work in collaboration with of engagement strategies to achieve the de- the DPI Weed Alert program to reduce the sired changes sought by the project. Exam- risk of introduction and incursion of high ples of these desired changes include, in- risk Victorian Alert Weeds in identifi ed ru- creased surveillance of high risk Victorian ral lifestyle areas, thereby increasing weed Alert Weeds in identifi ed lifestyle regions management support to the community.

76 Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 Response of Chilean needle grass (Nassella neesiana) and phalaris (Phalaris aquatica) seedlings and mature plants to changes in soil phosphorus, nitrogen and soil pH

Charles Grech, Department of Primary Industries, 475 Mickleham Road, Attwood, Victoria 3049, Australia.

Abstract Chilean needle grass (Nassella neesiana and pH was relatively larger than that of (Trin. & Rupr.) Barkworth, CNG) is com- CNG plants. Overall, plant productivity monly found growing in pasture swards was reduced across both species in acid or with more desirable and more palatable alkali soils without additional phospho- pasture species such as Phalaris aquatica. rus. CNG is known to replace more desirable Plants that were treated with nitrogen species and compete for soil resources in fertiliser were excluded from the analysis pasture and grazing situations. This pa- as they had symptoms of nitrogen burn- per describes an experiment which evalu- ing and their growth was detrimentally ated the relative responses of CNG and affected. phalaris to different soil pH levels and the Under glasshouse conditions, CNG was addition of phosphorus and nitrogen in able grow more rapidly than phalaris, pro- a glasshouse environment. The applica- ducing larger and taller plants, although tion of these treatments was to establish its general trends were similar to phalaris. what soil conditions favoured the growth Although phalaris may be more respon- of each of the species. sive during early growth to certain soil The addition of phosphorus generally management techniques, such as soil pH, increased the growth of both species. Pha- the overall response of phalaris during laris plants were taller than CNG plants later growth was less than that of CNG. during the late seedling stages although These results suggest that CNG is likely to CNG was taller by plant maturity across have a competitive advantage over phala- all soil pH treatments. Soil with neutral pH ris in the fi eld and that alteration of pH or favoured both species. The plant height soil phosphorus level is unlikely to alter response of phalaris to soil phosphorus that.

Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 77 A risk assessment has been carried out by scientists from the Department of Pri- Arrowhead in Victoria: current control methods and mary Industries as part of the Victorian noxious weeds review, in consultation potential for biological control with Catchment Management Authorities, and the weed was ranked as relatively in- Jean-Louis SaglioccoA, Nigel AinsworthB and Robin AdairA vasive with a score of 0.58. A process to as- A Department of Primary Industries, Primary Industries Research Victoria, sess Sagittaria spp. for possible declaration PO Box 48, Frankston, Victoria 3199, Australia. as noxious weeds is underway. B Department of Sustainability and Environment, PO Box 500, East Melbourne, Chemical control Victoria 3002, Australia. Several herbicides are registered for con- trol of Sagittaria montevidensis in rice crops. No herbicides are registered specifi cally for control of Sagittaria platyphylla. A small number of minor use permits have been is- Summary The exotic emergent aquatic identifi ed as Sagittaria graminea Michx. var. sued by the Australian Pesticides and Vet- weed arrowhead (Sagittaria platyphylla), weatherbiana (Fernald) Bogin (Aston 1967). erinary Medicines Authority (APVMA) originating from north and central Amer- Specimens collected in irrigation channels to authorise work by Goulburn Murray ica, is particularly invasive in northern between Nathalia and Yarrawonga were Water Authority, Murray Irrigation Ltd Victoria and southern New South Wales identifi ed as Sagittaria graminea var. platy- and the Central Murray County Council, including the Murray River. Chemical phylla Engelmann (Conn 1994). Due to the NSW and are available online at http:// control is currently used to manage the diffi culty in identifying arrowhead spe- www.apvma.gov.au/permits/permits. species, but there would be signifi cant ad- cies and the small number of herbarium shtml. Under permit, 2,4-D is used to vantages in developing biological control specimens kept at the Melbourne herbar- control S. platyphylla in irrigation chan- to provide a cost-effective means for con- ium, surveys were conducted in 2005–06 nels and specifi ed natural waterways, a trolling existing populations and limiting throughout invaded areas in the irrigation commercial product containing imazapyr further spread. districts of northern Victoria to collect suit- and glyphosate is used in irrigation chan- Keywords: Arrowhead, aquatic weed, able material for identifi cation (Sellwood et nels and drains and glyphosate is used in chemical control, biological control. al. 2006). Forty-one arrowhead infestations a range of circumstances. However, very were surveyed and at each site, specimens high glyphosate rates (up to 40 L per hec- Introduction were collected and preserved. Specimens tare) are required. Water resources in temperate Australia from all surveyed sites were sent to Dr Bri- A problem with herbicide application is are under threat not only due to drought, an Keener, a north-American specialist of that while emergent plants can be killed, over-exploitation and pollution, but also the genus Sagittaria, at University of Ala- submerged rosettes survive and require due to biotic invasions. A number of in- bama, Tuscaloosa, and all were identifi ed re-treatment when they become emergent. troduced aquatic plants are serious threats as Sagittaria platyphylla (Engelmann) J.G. The use of herbicides has to be carried out to Victorian waterways, wetlands, lakes, Smith. This taxon, previously known as with great care to ensure that maximum pondages and farm dams. The majority of Sagittaria graminea var. platyphylla Engel- concentrations set to protect the environ- these were introduced for the ornamental mann, was raised to the species level by ment or (in some cases) drinking water and aquarium trades, while others were Kortright (Kortright 1998). S. platyphylla quality are not exceeded and that irriga- introduced accidentally, for example, as is native to north and central America. Its tion water is safe to use. A particular prob- contaminants. Indiscriminate dumping distribution includes predominantly the lem arises in billabongs where the shallow of unwanted aquarium and ornamental southern states of the USA from Florida to water and lack of fl ow make it diffi cult to material has seen the naturalisation and Texas and extends to Mexico and Panama. apply the required herbicide rates without rapid spread of a number of aquatic weed In the USA, S. platyphylla grow in streams exceeding the allowable concentrations in species, among them arrowhead. and lakes from sea level up to an altitude water. When large areas of Sagittaria re- of 900 m ( Haynes and Barre Hellquist quire treatment, it is sometimes necessary Taxonomy 2000). to spray sections progressively to avoid Arrowhead is a generic name that refers the water quality effects that would oc- to exotic species from the Alismataceae Status cur if a large weed mass was all killed at family naturalised in Australia (Hnatiuk Sagittaria spp. are not currently declared once. 1990); two species are considered as inva- weeds in Victoria (under the Catchment and sive. The fi rst species, Sagittaria montevi- Land Protection Act 1994), New South Wales Potential for biological control densis, has two sub-species, both of which or Queensland. Sagittaria platyphylla and Due to the many diffi culties associated with are present in Australia (Aston 1973): S. Sagittaria montevidensis are declared N1 chemical control and the rapid spread of montevidensis calycina originates from weeds in South Australia (the presence of S. platyphylla infestations, the Department north America while S. montevidensis mon- the weed is notifi able throughout the State of Primary Industries was contracted by tevidensis originates from south America. (in some parts of the State only for S. mon- the Goulburn-Broken Catchment Manage- S. montevidensis is invasive in NSW in rice tevidensis) and plants must be destroyed ment Authority to investigate the poten- fi elds and irrigation channels (Sainty and throughout the State (in some parts of the tial for biological control of arrowhead. An Jacobs 1981). Some confusion existed on State only for S. montevidensis)). In Western extensive literature review was conducted the taxonomic identity of the second spe- Australia, they are declared as P1/2 weeds to identify the fl ora and fauna associated cies of Sagittaria, which is invasive in wa- (P1 prevents the trade, sale or movement with the weed in its native range and in terways in northern Victoria and southern of the weed, P2 enforces the eradication Australia (Sagliocco and Bruzzese 2005). NSW, as the names Sagittaria graminea or S. of a weed not yet widely established in References to pathogens associated with platyphylla are found in the literature. Ma- WA). In Tasmania, the weed is a declared Alismataceae overseas were rare and no terial originally collected by Aston from species for which restrictions measures are record of fungal pathogens on S. platy- northern Victoria, near Wunghnu, was specifi ed in weed management plans. phylla was found. A number of weevils

78 Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 (Coleoptera: Curculionidae) are known to Biological control of aquatic weeds has O’Brien, C.W. (1981). The larger (4.5 mm) attack Sagittaria spp. in the USA (Blatchley been successful, even spectacular in some Listronotus of America north of Mexico and Leng 1916, Center et al. 1999, O’Brien cases, and aquatic weeds count among the (Cylindrorhininae, Curculionidae, Co- 1981, 1977). Thirteen species of Listrono- most successful cases in the history of bio- leoptera). Transactions of the American tus are described as being associated with logical control. While water supplies, usage Entomological Society 107, 69-123. Sagittaria spp. (S. sagittifolia, S. engelman- and conservation come under increased O’Brien, C.W. (1977). The semiaquatic niana, S. lancifolia, S. latifolia, S. graminea, scrutiny, it is desirable that co-operative weevil genus Listronotus in Mexico and S. longiloba, S. cuneata or Sagittaria sp.) arrangements are developed between the central America (Coleoptera: Curculio- Thirteen additional Listronotus species are different organisations involved in water nidae: Cylindrorhininae). Annals of the also described, but their host-plant(s) are delivery and conservation, such as water Entomological Society of America 70 (5), still unknown. However, they are likely users, the different levels of government 804-14. to be within the genus Sagittaria due to and research scientists, to further investi- Sagliocco J-L. and Bruzzese, E. (2005). the very high specialisation within the gate biological control of arrowhead. Biological control of arrowhead – fea- Curculionidae family. Listronotus spp. sibility study, Department of Primary have been found to be the most common Acknowledgements Industries, unpublished report, 18 pp. species feeding and breeding on Sagit- The feasibility study on biological control Sainty, G.R. and Jacobs, S.W.L. (1981). ‘Wa- taria spp. and were observed feeding on of S. platyphylla was funded by the Goul- terplants of New South Wales.’ (Water root collars, leaves, stems, stalks, fl owers, burn-Broken Catchment Management Au- Resources Commission, NSW). fruiting heads (Center et al. 1999, O’Brien thority. Sellwood, M., Ainsworth, N. and Sagl- 1981) or to cause stem galls (O’Brien 1981). iocco, J-L. (2006). A survey of Sagittaria Also in the Curculionidae family, Brachy- References spp. (Monocotyledons: Alismataceae) bamus electus Germar, Anchodemus angus- Aston, H.I. (1967). Aquatic Angiosperms: populations throughout the irrigation tus LeConte and Barinus bivittatus LeConte records of four introduced species new districts of northern Victoria. Proceed- have been recorded on Sagittaria spp., but to Victoria. Muelleria 1, 169-74. ings of the 15th Australian Weeds Con- their plant association is not mentioned Aston, H.I. (1973). ‘Aquatic plants of Aus- ference, eds C. Preston, J.H. Watts and (Blatchley and Leng 1916). Outside cur- tralia, a guide to the identifi cation of N.D. Crossman, pp. 798-800. (Weed culionids, Plateumaris (Coleoptera: Chry- the aquatic ferns and fl owering plants Management Society of South Austral- somelidae) are also reported associated of Australia, both native and natural- ia, Adelaide). with Sagittaria roots (Jolivet and Hawkes- ized.’ (Melbourne University Press, wood 1995). Finally, the diptera leaf- Melbourne). miner Hydrellia deceptor Deonier (Diptera: Blatchley, W.S. and Leng, C.W. (1916). Ephydridae) is reported associated with ‘Rhynchophora or weevils of North Sagittaria sp. in the USA (Deonier 1971, Eastern America.’ (The Nature Publish- 1998). ing Company, Indianapolis). Center, T.D., Dray, F.A., Jubinsky, G.P. and Discussion Grodowitz, M.J. (1999). Insects and oth- Although mapping of S. platyphylla infes- er arthropods that feed on aquatic and tations is incomplete, the rate and extent of wetland plants. United States Depart- invasion has increased dramatically from ment of Agriculture, Agricultural Re- creeks to irrigation channels and wetlands, search Service, Technical Bulletin 1870, with the greatest threat being the invasion p. 200. (USDA-ARS). of the whole Murray River system. The Conn, B.J. (1994). Alismataceae. In ‘Flora lack of registered herbicides combined of Victoria, Volume 2’, eds N.G. Walsh with the diffi culties of application and the and T.J. Entwisle. (Inkata Press, Mel- need to protect water resources from con- bourne). tamination make the chemical control of Deonier, D.L. (1971). A systematic and eco- arrowhead extremely diffi cult. The genus logical study of Neartic Hydrellia (Dip- Sagittaria is not native to Australia and tera, Ephydridae). Smithsonian Contri- does not contain economically important butions to Zoology 68, 1-147. species. Thus biological control represents Deonier, D.L. (1998). ‘A manual of the an attractive option to achieve long term common North American species of control of existing populations and limit the aquatic leafmining genus Hydrel- further spread. lia (Diptera: Ephydridae).’ (Associated As a priority, it is necessary to conduct Publishers, Gainesville, Florida, USA). genetic studies on populations of the weed Haynes, R.R. and Barre Hellquist, C. in Australia to identify precisely its par- (2000). Alismataceae. In ‘Flora of North ent populations in North America. While America north of Mexico, vol. 22. (Ox- the initial literature review has shown that ford University Press, New York). a number of specialist herbivores exist in Hnatiuk, R.J. (1990). Census of Australian the genus Listronotus, further surveys in vascular plants. the southern USA and central America Jolivet, P. and Hawkeswood, T.J. (1995). are necessary to complete this study and ‘Host-plants of Chrysomelidae of the to conduct detailed observations on the world.’ (Backhuys Publishers, Leiden, ecology of S. platyphylla and its natural en- Netherlands). emies. In North America, the wide range Kortright, R.M. (1998). The aquatic plant of S. platyphylla would allow for selection genus Sagittaria (Alismataceae): phyl- of appropriate species and biotypes of bio- ogeny based on morphology. M.Sc. the- logical control agents adapted to climatic sis. University of Alabama. situations similar to Australia.

Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 79 the answers to the 10 impacts questions for each willow taxon assessed. In all cases, Weeding out Australia’s worst willows the ‘worst-case scenario’ was used. If there was evidence that a willow was capable of having a large impact in a particular Jackie Steel and Amy Scott, Department of Primary Industries – Frankston, environment, it scored highly, even if its PO Box 48, Frankston, Victoria 3199, Australia. impact was lower in other types of envi- ronments. A range of literature was used to per- form the assessments, from journals, books and internet sites, to expert opinion. Summary We compared two risk assess- can be considered safe enough to leave in A confi dence score was attributed to each ment methods that can be used to rank situ, or even continue to be planted. question answered to give an indication and categorise willow taxa based on their of the quality of the data used to assess potential to have serious impacts on Aus- Methods each taxon, according to the descriptions tralia’s social, economic and environmen- Two risk assessment methodologies were in Appendix 2. tal values. The fi rst was a generic weed used to assess willows. The methods sec- The descriptive scores of high (H), me- risk assessment and the second, a risk as- tion is in two parts: an analysis of the dium high (MH), medium low (ML) and sessment process tailored to discriminate Victorian Pest Plant Prioritisation process low (L) were converted into the following between willows. We then used the rank- used initially to assess willows, and a de- numerical scores: H = 1, MH = 0.67, ML = ings from the tailored risk assessment to scription of the process of developing a 0.33 and L = 0. determine which willows pose the biggest specifi c willows risk assessment tool. If there was insuffi cient evidence avail- threat to natural resources and hence rec- able to answer a question for a particular ommend them as a high priority for weed Analysing the Victorian Pest Plant taxon, a score of medium (M) was chosen, management. The tailored assessment Prioritisation Process with a value of 0.5; likely to cause the least identifi ed 13 taxa that can confi dently be The Victorian Pest Plant Prioritisation amount of error, as it could only be inac- considered to have the most serious im- Process (PPPP) method used to assess the curate by ±0.5. In such a case as this, a con- pacts. It also identifi ed several low impact willows (Weiss et al. 2004) accords with fi dence score of L was chosen. taxa, but the quality of the data available the National Post border Weed Risk Man- Once all the questions had been an- to assess these taxa was of a much lower agement Protocol (SAI 2006). As such, the swered for a willow taxon, the descriptive calibre. A complete weed risk assessment complete assessment process considered: scores were converted to numerical scores. of willows should also include consid- 1. Invasiveness (or biological traits), Each numerical score was divided by 10, eration of invasiveness and distribution. 2. Potential for spread (by comparing cur- so that when a taxon’s scores for each These are yet to be undertaken, and will rent and potential distributions), and question were added together, the score be crucial to determining the full risk po- 3. Impacts on land use and ecosystems (or would fall between the values of 0 and tential of the members of this genus. Being ecology). 1. The higher the score, the greater was able to differentiate willows in terms of degree of impact that the willow could their impacts takes us one step closer to This paper is primarily concerned with the have. The willows were then ranked from being able to set strategic priorities for wil- impacts component of the assessment. The the highest score to the lowest. Similarly, low management across Australia. criteria that were used for this assessment confi dence scores were converted to a nu- are listed in Appendix 3. We examined the merical value from 0 to 1. The full assess- Introduction spread of scores that resulted from this as- ments, including the evidence used for Willows (Salix spp.), is a taxonomically sessment to determine how many ques- each taxon are available at www.weeds. complex genus comprising more than 400 tions were able to discriminate between org.au/WoNS/willows. A summary of the taxa worldwide (van Kraayenoord et al. the willows. A subset of 14 willow taxa scores is in Appendix 4. 1995). More than 100 willow taxa have was selected for this analysis. been introduced to Australia. Eleven spe- Which willows to assess? cies are naturalised (ARMCANZ 2001, Developing the criteria for assessing the The 16 willow taxa assessed using the APC 2007, APNI 2007), and there are also impacts of willows Victorian PPPP were chosen because they at least 25 described subspecifi c taxa and In February 2007 criteria were developed are declared noxious in Victoria, and are many hybrid combinations. to assess the relative impacts of a range naturalised in Australia. However, to de- Willows were nominated, assessed and of willow taxa. Criteria were developed termine which taxa that should be high listed at the generic level for classifi cation that would be used to assess the degree priorities for management, we assessed as Weeds of National Signifi cance (WoNS) of impact that each willow taxon might a larger list that included non-declared (Thorpe and Lynch 2000); however three have on social, economic and environmen- taxa, potential low risk candidates, and Salix taxa were not included in the WoNS tal values. These criteria needed to differ- also tried to identify willows that might as they appeared to be ‘non-weedy’, and a entiate between willow taxa, and provide become weedy in Australia in the future. further six were exempt from declaration evidence for their impact on social, envi- We limited the assessment list to taxa that in Victoria. ronmental or economic values. have been introduced to Australia, includ- Willow infestations are targets for either The ranking criteria are presented in ing taxa that have naturalised in Australia; eradication or containment in Australia Appendix 1. Each criterion was based or have the potential to naturalise in Aus- (ARMCANZ 1999). Determining the fea- around a question, and scored either as tralia, either due to a history of natural- sibility control methods is complicated by ‘high impact’, ‘moderately high impact’, ising overseas or because they exhibited the large number of taxa involved and the ‘moderately low impact’ or ‘low impact’, invasive traits; or that appeared unlikely likelihood that some of these pose a larger according to the descriptors that were de- to become serious weeds in Australia. weed threat than others. Furthermore, to veloped for each intensity rating. We also assessed individually the three ensure that willow management strategies Once the assessment criteria were major groups of willows, the subgenera are cost effi cient, it is important to decide determined, a literature search was per- Salix, Vetrix and Chamaetia (Skvortsov which willows require control, and which formed, and expert opinion sought, to fi nd 1999). Species within each Salix subgenus

80 Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 often share many biological and ecologi- other three questions relating to abiotic was no information in the literature to in- cal traits. impacts did not differentiate the willow dicate that any of the willow taxa have: A recent modelling exercise highlighted taxa assessed. All willows assessed were affected agricultural quality (20) or land the weed risk associated with some exempt considered to have a high impact on wa- value (21), caused a change in land use taxa (Stokes and Cunningham 2006), so ter quality (6); to be capable of increasing (22), or provided a host to pests or diseases willows that were exempted from noxious vegetative biomass (8); and the potential of agriculture (24), hence all taxa attained weed legislation in any Australian state to greatly reduce the frequency and inten- low scores for the criteria addressing these were also assessed. sity of fi re risk (9) in the habitats in which impacts. Impact on agricultural yield (19) they occurred. varied from low to medium low and there Analysing the willows impacts was varying ability of willows to increase assessment Community/habitat impacts All as- in harvest costs (23). We examined the spread of scores that re- sessed taxa were considered to have a sulted from this assessment to determine high or moderately high impact on the Results for the willows risk assessment how well these questions were able to dis- structure of vegetation communities (11). Compared to the willow impact scores criminate between the willows. A subset of There was some variation in the impacts from the Victorian PPPP (in Figure 1), 14 willow taxa was selected for this analy- on high value Ecological Vegetation Class- the willows risk assessment resulted in a sis; those that were also assessed using the es (EVCs) (10a), however, willows were greater spread of scores across all the cri- Victorian PPPP process. not considered to impact on any low or teria (Figure 2). medium value EVCs (10b and c). The scores and ranks of the willows as- Results Impacts on threatened fl ora (12) and sessed appear in Table 1. The annotations Results for the Victorian Pest Plant fauna (13) were considered high for one relate to the reason(s) that each taxon was Prioritisation Process species, medium high for most, and me- chosen for assessment. Many of the criteria used to assess the dium (which is the score for ‘unknown’) Of the 20 taxa that had an impact score weediness of willows were not able to dis- for a couple of the willow taxa. at or above the mean (0.60), 17 are natural- criminate between different taxa. In 13 of ised in Australia This group includes two the 26 questions, all willows achieved the Impacts on fauna All willows were subgenera, Salix, ranked at two; and Ve- same score. Furthermore, in question 11 documented as having a moderately high trix, ranked at 10. Salix daphnoides, ranked all the willows rated either H or MH, and impact on native fauna (14). The ability at 17, is not naturalised in Australia, but in question 19, either L or ML (Figure 1). of willows to provide some assistance in For more than half of the impacts assess- shelter to desirable species was considered ment there was very little separation in the high for one taxon, medium high for most, scores attained by each willow taxon. and medium (which is the score for ‘un- Footnote known’) for three of the willow taxa. No 1 A numbered list of the criteria used in Social impacts Only two social impacts willow taxa were found to possess proper- the Victorian PPPP is in Appendix 3, and criteria provided any separation value ties injurious to fauna (16). these numbers appear in brackets in the between the willow taxa: restricting ac- text to indicate which criterion is being ad- cess (1)1 and reducing tourism (2). All taxa Pest animal impacts There was some dressed. The criteria were broadly catego- attained a low score for ability to injure variation in the willows’ potential to pro- rised according to the types of values that to people (3) and high for ability to cause vide a food source to minor pest species may be impacted by weeds: social, abiotic, major damage to cultural sites or infra- (17) and capacity to provide harbour to community/habitat, fauna, pest animals structure (4). pest animals (18). and agricultural values. Full assessments for all of the plants that have been as- Abiotic impacts Whilst some willows Agricultural impacts The Victorian PPPP sessed using Victoria’s PPPP can be found scored higher than others for impacts on of willows indicated that willows general- at www.dpi.vic.gov.au/dpi/vro/vrosite. water fl ow (5) and soil erosion (7), the ly have few impacts on agriculture. There nsf/pages/lwm_pest_plants.

100 90 80 H 70 60 MH

50 M

Percent 40 ML 30 20 L 10 0 1 2 3 4 5 6 7 8 910a10b10c1112131415161718192021222324

Question number

Figure 1. Spread of scores from the Victorian PPPP for the impacts of willows. Proportion of willows that scored H, MH, M, ML or L for each question.

Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 81 100 90

80 H 70 60 MH

50 M

Percent 40 30 ML 20 L 10 0 12345678910 Question number Figure 2. Spread of scores from the willow impacts risk assessment. Proportion of willows that scored H, MH, M, ML or L for each question.

Table 1. Scores and rank, from highest impact (1) to lowest (35) of willows assessed with the new impacts assessment. Taxon Common name Score Rank Conf. *S. × rubens white crack willow 0.97 1 0.67 *S. subgenus Salix tree willows, true willows 0.93 2 0.67 *S. fragilis crack willow 0.93 2 0.67 *S. nigra black willow 0.93 2 0.67 *S. alba white willow 0.90 5 0.67 *†S. babylonica weeping willow 0.88 6 0.60 *S. × sepulcralis golden weeping willow 0.87 7 0.65 *S. cinerea grey sallow 0.83 8 0.67 *S. triandra almond willow 0.80 9 0.67 *S. subgenus Vetrix shrub willows, sallows, osiers 0.77 10 0.67 *S. purpurea purple osier 0.77 10 0.58 *S. viminalis common osier 0.77 12 0.62 %S. exigua sandbar willow 0.75 13 0.60 *‡S. alba × matsudana NZ hybrid willow 0.75 13 0.42 *S. × pendulina Wisconsin weeping willow 0.73 15 0.37 %S. glauca Arctic grey willow 0.65 16 0.45 #S. daphnoides violet willow 0.62 17 0.60 *S. × mollissima 0.60 18 0.35 *†S. × reichardtii pussy willow 0.60 18 0.43 *S. matsudana (‡‘Aurea’ and ‘Tortuosa’) tortured willow 0.60 20 0.63 *†S. × calodendron pussy willow 0.58 21 0.30 #S. eriocephala 0.57 22 0.67 #S. pentandra bay willow 0.57 23 0.30 %S. gracilistyla 0.55 24 0.63 *S. caprea (‡‘Pendula’) goat willow (great sallow) 0.52 25 0.58 *S. × sericans pussy willow 0.52 26 0.18 *S. glaucophylloides broadleaf willow 0.45 27 0.32 *S. aegyptiaca Egyptian willow 0.42 28 0.43 #S. elaeagnos hoary willow (bitter willow) 0.35 29 0.58 ‡S. myrsinifolia dark-leaved willow 0.30 30 0.28 *S. humboldtiana ‘Pyramidalis’ Chilean pencil willow 0.22 31 0.48 S. subgenus Chamaetia alpine, arctic willows 0.12 32 0.48 ‡S. alba var. caerulea cricket bat willow 0.10 33 0.52 ^S. integra ‘Hakuro-nishiki’ Nishiki willow 0.10 33 0.57 ^S. × ‘Boydii’ 0.00 35 0.53 Mean 0.60 – 0.53 *Naturalised in Australia. #Naturalised overseas. †Exempted from the WoNS. ‡Exempt from declaration in Victoria. %Exhibits invasive traits. ^Appears to be low-risk.

82 Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 has become invasive in New Zealand (van No reference to willows being directly climatic match for the species. Climatic Kraayenoord et al. 1995). The top 20 also injurious to people was found, therefore modelling showed that some taxa are not includes two taxa that were exempted all taxa attained a low score for this cri- likely to occur as invasive plants in Vic- from the WoNS, and three taxa that were terion. Damage to buildings may occur toria, giving these taxa a low score both exempt from declaration in Victoria. The when stream channel routes are altered for this criterion, and also for impact on remaining two species in this highly as a result of blockage by willow roots, threatened fl ora. The vegetation commu- ranked group, S. exigua and S. glauca do (Sarah Holland-Clift personal communi- nities of all water bodies in Victoria were not have a history of becoming invasive, cation 2006) and during fl ooding episodes considered to comprise high value EVCs, but were chosen for assessment because the build up of woody material from wil- therefore willows were not considered they exhibited traits that suggest they may lows in rivers can destroy bridges (ARM- to impact on any low or medium value become invasive. CANZ 2001). Thus, willows generically EVCs. The taxa with impact scores below were considered to have the potential to Although willows are well documented the mean included many that have not cause major damage to cultural sites or as having a signifi cant impact on vegeta- naturalised beyond their native range, infrastructure and all taxa scored high for tion communities, little information was however, six of them have naturalised in the criterion addressing this impact. found in regard to their impact on threat- Australia: S. × calodendron, S. caprea, S. × ened fl ora. Only S. cinerea was specifi cally sericans, S. glaucophylloides, S. aegyptiaca Abiotic impacts Willows were consid- identifi ed, being described as the most se- and S. humboldtiana ‘Pyramidalis’. Three ered to have a high impact on water fl ow, rious willow preventing the recruitment of the taxa have naturalised overseas: S. if they had the capacity to grow within of Eucalyptus camphora, a dominant com- eriocephala, S. pentandra and S. eleaegnos. S. streambeds, therefore interrupting the ponent of the rare sedge-rich E. camphora gracilistyla, a non-naturalised willow that fl ow of water (Purtle et al. 2001), or if their Swamp community, listed under the Flora exhibits invasive traits, achieved a an im- roots intruded extensively into stream and Fauna Guarantee Act 1988 (Ladson et pact score below the mean. beds (Ladson et al. 1997). Taxa confi ned to al. 1997). Salix × boydii, not known to be natu- riverbanks were considered less likely to Climatic suitability was the main fac- ralised anywhere in the world, attained impact on fl ow. All willows assessed were tor separating willows in the two ques- the lowest possible impact score of 0. This considered to have a high impact on wa- tions that showed much variability in the indicates that it was rated as low for every ter quality because, as deciduous plants community/habitat section of the impact impact assessment criterion. Also ranked (Carr 1996), mass leaf fall in autumn can assessments. in the bottom six were four further non- decrease dissolved oxygen levels, and naturalised taxa. Two of these are exempt willows can also cause intense shading as Impacts on fauna Willows are docu- from noxious weed declaration in Victoria. their canopies tend to be denser than na- mented as having a signifi cant impact on The only naturalised willow in the bottom tive species (Ladson et al. 1997). Willows native fauna, with the ability to decrease six was S. humboldtiana ‘Pyramidalis’. that encroach into the centre of streams available habitat and reduce population The confi dence score is an indication of interrupt water fl ow, which is then di- numbers. For example, shading from wil- the quality of data that was used to assess rected into banks causing erosion. In se- lows decreases primary production and each willow. The average confi dence score vere cases, extreme blockages can occur, impacts on aquatic invertebrates and fi sh was 0.53. Lower confi dence scores were causing streams to change course (Purtle (Ladson et al. 1997). Willows suppress and found to be associated with taxa having et al. 2001). Therefore, taxa that encroach kill indigenous vegetation that would oth- low impacts. The top 13 ranked willows into streams have a high probability of erwise provide valuable habitat and food all had confi dence scores above the mean. increasing soil erosion. As woody shrubs for insects, birds and other vertebrates, The very lowest confi dence scores, below or trees with the capacity to form dense and bare banks beneath willows provide 0.33 (or ML in the scoring system) were thickets, (Carr 1996) all willows assessed little protection for fauna such as frogs, attained by some Australian naturalised were capable of increasing the vegetative water rats, snakes and lizards. Willows do taxa: S. calodendron, S. × sericans, and S. biomass where they replaced lower or less not provide nectar for birds, have few hol- glaucophylloides; by S. pentandra , which has dense vegetation, a regular occurrence on lows (Purtle et al. 2001) and provide less naturalised in the USA (Argus 1986); and disturbed sites (Cremer 1999). Willows large woody debris in stream than native by S. myrsinifolia. All of these taxa scored have low combustibility and fl ammability tree species, important habitat for aquatic below the mean on the impacts ranking, (Carcallet et al. 2001), and all taxa had the fauna (Ladson et al. 1997). but we cannot be sure that their scores are potential to greatly reduce the frequency Again, no information was found spe- accurate, due to a lack of data. and intensity of fi re risk in the habitats in cifi cally documenting the impact of par- which they occurred. Whilst some willows ticular willow taxa on threatened fauna, Discussion scored higher than others for impacts on except for S. cinerea, which is described The Victorian Pest Plant Prioritisation water fl ow and soil erosion, the other three as having the potential to destroy impor- Process questions relating to abiotic impacts did tant habitat of the endangered Leadbeat- Social impacts Certain willows form not differentiate the willow taxa assessed. ers possum and helmeted honeyeater, as thickets as well as grow in-stream, which well as the rare broad toothed rat (Ladson gives them the capacity to signifi cantly re- Community/habitat impacts The for- et al. 1997). strict human access to watercourses, and mation of dense thickets, intense canopy Possums are known to graze and defo- well as restricting access by boats. These shade, and mat-forming roots of willows, liate willows (ARMCANZ 2001). Thicket characteristics, combined with their ability can suppress and exclude indigenous un- forming willows provide cover for wildlife to reduce channel capacity (Cremer 1995, derstorey (Cremer 1999, Purtle et al. 2001), in the USA (Anderson 2006); therefore wil- Purtle et al. 2001) can lead to a reduction in with the result that all assessed taxa were lows are likely to provide some assistance tourism, and activities potentially affected considered to have a high or moderately in shelter to desirable species (15). Salix include: boating, fi shing, swimming, ca- high impact on the structure of vegetation glaucophylloides is unpalatable (Webb et al. noeing and bird watching. Single-trunked communities. They would consequently 1998) and as having a single trunk (Haines species and those valued aesthetically have a similarly high impact on invaded 2003), so it was considered to provide very such as Salix matsudana and S. glaucophyl- Ecological Vegetation Classes (EVCs), little benefi t to fauna. No reference was loides (Ladson et al. 1997) were not consid- however, the degree of impact was con- found in the literature to suggest that any ered to reduce tourism. sidered dependant on the suitability of the willow taxa possess properties injurious to

Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 83 fauna, and consequently all taxa attained a different species, which the Victorian PPPP ised in New Zealand (van Kraayenoord et low score for the criterion associated with was developed to achieve. However, it al. 1995), but only recorded in horticulture this impact. was not designed to detect the differences in Australia. This species should be desig- between similar and closely related taxa, nated as a target for eradication. Salix glau- Pest animal impacts Only S. purpurea was such as the willows. Aquatic plants also ca and S. exigua scored above average in the documented as providing a food source to tended to attain similar scores using this impacts assessment but, despite being in- rabbits, a serious pest (Dickerson 2002). PPPP, so an aquatic weed risk assessment troduced to several new countries for their Salix alba and S. exigua were described as was also developed for Victoria (Weiss aesthetic value (Newsholme 1992), have being palatable (Anderson 2006), and ro- 2007), New Zealand has a weed risk as- not naturalised beyond their native range dents are documented as eating the buds sessment tailored to aquatic plants as well anywhere in the world. A full risk assess- of S. nigra, therefore, these willows have (Champion and Clayton 2001). ment process that includes the invasive- potential to provide a food source to minor Defi ned groups of weeds affecting par- ness and potential distribution for these pest species. ticular values or a specifi c habitat type can taxa will have a provide better assessment Willow stands are described as pro- be subject to specialised weed risk assess- of which of these taxa are likely to be the viding excellent cover for wildlife in the ment tools. New weed risk assessment highest priority for management. USA (Anderson 2006), therefore, thicket tools should be tailored to pick up specifi c Salix caprea, S. aegyptiaca and S. hum- forming taxa, such as S. exigua, S. cinerea, negative characteristics and impacts of boldtiana ‘Pyramidalis,’ are naturalised S. fragilis, S. purpurea, S. × rubens and the target group. Willows contain many in Australia, but had low impact scores S. viminalis (Anderson 2006, Cremer 1995, potentially serious weedy taxa, therefore indicating these species should be lower 1999, 2001, Webb et al. 1988) are likely to developing an effective weed risk assess- priority for management. There are very have the capacity to provide harbour and ment tool identifi es which are likely to be few records of these taxa naturalising and permanent warrens to rabbits and foxes. the most serious weeds. assessments of their invasiveness and dis- Non-thicket forming willows are unlikely The impacts that willows have are tributions may add to this preliminary to provide harbour. largely related to several main character- evidence that they don’t pose a large risk istics, such as, the ability to form dense to Australia’s natural resources. The low Agricultural impacts The Victorian PPPP thickets, grow within streams, undertake scores of several taxa (S. × calodendron, S. of willows indicated that willows general- mass autumn leaf drop and develop large, × sericans and S. glaucophylloides) were also ly have few impacts on agriculture. There invasive root systems. Impact assessment associated with low confi dence scores, in- was no information in the literature to criteria for willows therefore need to util- dicating that little high quality data could indicate that any of the willow taxa have ise these characteristics. However, the im- be used for assessing these species. affected agricultural quality or land value, pact criteria must discriminate between An ability to naturalise elsewhere in caused a change in land use, or provided different willow taxa. The aim of the wil- the world is cited as the best predictor a host to pests or diseases of agriculture, low risk assessment is to determine which of a plant’s ability to naturalise in Aus- hence all taxa attained low scores for the are the worst willows in Australia; which tralia (Panetta 1993). Whilst S. eriocephala, criteria addressing these impacts. Several should be the focus of our management S. pentandra and S. eleaegnos have not yet species were described as agricultural efforts. naturalised in Australia, their invasive weeds in USA and New Zealand (Holm histories in UK, USA. and New Zealand et al. 1979), but as they are not recorded as Willows impacts assessment (Stace et al. 2006, Argus 1986, van Kraaye- invaders of pasture or crops, their impacts The willows risk assessment process re- noord et al. 1995) respectively, makes them are likely to be associated with water- vealed that the two subgenera Salix and high risk plants in Australia. Although, S. ways. Their ability to form dense thickets Vetrix have above-average potential for gracilistyla was given a score below the restricting access for irrigation is likely to impact, and the subgenus Chamaetia was mean and is not known to be naturalised, be the only impact they have on agricul- very low. The assessment of the three sub- it does exhibit invasive traits, such as the ture, and willow root mats are described genera aimed to give an overview of the ability to spread to form dense thickets as reducing access to fl owing water (Sa- weed risk posed by each subgeneric group (van Kraayenoord et al. 1995). Pre-emp- rah Holland-Clift personal communica- of willows. As Chamaetia are not natural- tive removal from horticulture would be a tion 2006). This may have a minor impact ised anywhere in the world, despite a prudent approach for these species. on agricultural yield, as well as cause a history of translocation dating back to at The subgenus Chamaetia and horticul- minor increase in harvest costs due to the least 1906 (Tennant 2004) it appears likely turally-valued taxa, such as S. × boydii, S. requirement to maintain waterways for ir- that many of the willows in this subgenus myrsinifolia, S. integra ‘Hakuro-nishiki;’ rigation purposes. present a low weed risk. and the cricket bat willow S. alba var. caer- Willows are clearly able to seriously Of the 20 willow taxa with high impact ulea had amongst the lowest impact scores. impact many of our social and environ- scores, the 17 naturalised taxa are of most None of these taxa have a history of inva- mental values, but their impacts tend to concern. Since they have proven capable siveness, despite having been cultivated be confi ned to habitat niches in riparian of establishing in the Australian landscape for many years. Introductions of these taxa and wetland areas (Richardson et al. 2006), they are likely to become the highest prior- occurred in 1900 (for S. alba var. caerulea), largely affecting abiotic components of ities for management. Two willow species whilst S. × boydii has been in Australia for aquatic systems as well as biodiversity. (S. babylonica and S. × reichardtii) from this 14 years, S. myrsinifolia for 19 years and Many similarities exist between different high impact group that were exempted members of the subgenus Chamaetia began willows because they are a closely related from WoNS are of particular concern, as arriving more than 20 years ago, although group of taxa that grow in similar habitats are the taxa that were exempted in Victo- some as recently as 2002 (Robert Ingram and the minor variation was not adequate- ria (S. alba × matsudana, and S. matsudana personal communication 2007). This sug- ly captured by the Victorian PPPP. Anoth- vars. ‘Aurea’ and ‘Tortuosa). Whilst they gests that many of these willows may be er factor that reduced the separation value are valued for their aesthetic qualities (van suitable for cultivation, although invasive- of the Victorian PPPP of willows was that Kraayenoord et al. 1995), their potential to ness and distribution assessments are re- many of the criteria are not particularly affect social, economic and environmental quired. The decision to exempt S. alba var. applicable to willows. entities suggests that they will also require caerulea and S. myrsinifolia from noxious A generic weed risk assessment needs management intervention. weed legislation in Victoria the appear to to be suitable to assess a large number of Also of concern is S. daphnoides, natural- be supported by this assessment.

84 Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 The apparent low risk of the dwarf/ Acknowledgements Cremer, K.W. (1995). Willow identifi ca- alpine willows remain equivocal. Salix Thanks go to the participants of the im- tion for river management in Australia. glauca belongs to the subgenus Chamaetia pacts criteria workshop for their time and CSIRO Division of Forestry, Orange, and it did achieve an above-average im- effort in determining and the criteria, and NSW. pact score. However this species differs to John Weiss for running the workshop; Cremer, K. (1999). Willow management for from most Chamaetia in its ability to form to Geoff Carr for his insights into the biol- Australian rivers. CSIRO Forestry and thickets, and extensive shrublands (Welsh ogy and ecology of willows; Rob Ingram Forest Products, Kingston, ACT. 1974, Skvortsov 1999). for his work on plant introductions into Cremer, K. (2001). Exterminate wild pussy The above-average confi dence scores Australia and Sarah Holland-Clift (WoNS willows and its most invasive relatives. of the top 13 taxa highlights these as the national willow coordinator) for initiating CSIRO Forestry and Forest Products, highest priority for management. The very and supporting this work. We are grateful Kingston ACT. low confi dence scores for some natural- to Sarah Keel and Robin Adair for their Department of Primary Industries (DPI) ised taxa suggest further risk assessment suggestions for improvements to drafts of (2006). Review of noxious weeds in effort is required particularly for S. × cal- this paper. Victoria. Department of Primary Indus- odendron, S. × sericans, S. glaucophylloides tries, Frankston, Victoria. and S. pentandra. References Dickerson, J. (2002). Purple osier willow. The willows that aren’t naturalised are Anderson, M. (2006). Salix exigua. In Fire Natural Resources Conservation Serv- the ones that we know the least about. The effects information system. (Online). ice, United States Department of Ag- association of low confi dence scores with US Department of Agriculture, Forest riculture. lower impacts may be due to a tendency Service, Rocky Mountain Research Sta- Haines, A. (2004). Salix myrcoides (bayber- of scientifi c interest towards species that tion, Fire Sciences Laboratory (Produc- ry willow) Conservation and Research have been observed impacting on social, er), viewed: 27th August, 2007, http:// Plan for New England. New England agricultural or natural resources. www.fs.fed.us/database/feis/plants/ Flower Society, Framingham, Massa- Of the non-naturalised taxa, we can be shrub/salexi/all.html. chusetts, USA. more sure that S. integra and S. gracilistyla Argus, G.W. (1986). The genus Salix Sali- Holm, L., Pancho, J.V., Herberger, J.P. and are low impact taxa, and that S. exigua may caceae in the Southeastern United States. Plucknett, D.L. (1979) ‘A geographi- have high impacts, than of S. myrsinifolia Systematic Botany Monographs 9. cal atlas of world weeds.’ (Wiley, New being low impact. Further research to in- Agriculture and Resource Management York). crease the quality of the data used for as- Council of Australia and New Zealand Ladson, A., Gerrish, G., Carr, G. and Thex- sessing impacts is recommended for most (ARMCANZ) (1999). The National ton, E. (1997). Willows along Victorian of the lowest-ranked taxa. Of particular weeds strategy: a strategic approach to waterways. Waterways Unit, Depart- concern is S. myrsinifolia, which had an weed problems of national signifi cance. ment of Natural Resources and Envi- extremely low confi dence score, and S. Agriculture and Resource Management ronment, Victoria, Australia. humboldtiana ‘Pyramidalis,’ already natu- Council of Australia and New Zealand, Newsholme, C. (1992). ‘Willows: the ge- ralised here. Without further research we ACT. nus Salix’. (Timber Press, Portland and run the risk of assuming that taxa, such as Agriculture and Resource Management Oregon). this last one, are low risk, because our as- Council of Australia and New Zealand Panetta, D. (1993). A system of assessing sessments have been based on little data. (ARMCANZ) (2001). Weeds of National proposed plant introductions for weedy Some criteria used to assess the impacts Signifi cance Willow (Salix taxa, exclud- potential. Plant Protection Quarterly 8(1), of willows may be considered more im- ing S. babylonica, S. × calodendron and S. 10-14. portant in determining which taxa should × reichardtii) Strategic Plan’. National Purtle, C., Stelling, F., Martin, D., Gross- be a prioritised for management. The rela- Weeds Strategy Executive Committee, man, T., Frankenberg, J., Campbell, S. tive importance of each criterion can be Launceston. and Dwyer, C. (2001). Willows along manipulated using a pair wise compari- Australian Plant Census (APC) (2007). IBIS watercourses: an introduction. Depart- son tool in an Analytical Hierarchy Proc- database. Centre for Plant Biodiversity ment of Land and Water Conservation, ess, as was used for the Victorian PPPP Research, Council of Heads of Austral- NSW. (Weiss et al. 2004). New scores and ranks ian Herbaria, viewed 12th July 2007, Richardson, F.J., Richardson, R.G. and can be calculated using weighted factors www.chah.gov.au/apc/index.html Shepherd R.C.H. (2006). ‘Weeds of the which could change the relative rankings Australian Plant Name Index (APNI) south-east – an identifi cation guide for of willow taxa. (2007). IBIS database, Centre for Plant Australia’. (R.G. and F.J Richardson, The post-border weed risk management Biodiversity Research, Australian Gov- Melbourne). protocol (SAI 2006) describes how weed ernment, Canberra, viewed 12th July Skvortsov, A.K. (1999). ‘Willows for Rus- risk can be determined by comparing the 2007, www.cpbr.gov.au/cgi-bin/apni sia and adjacent countries: taxonomical likelihood that a plant might naturalise in Carcallet, C., Bergeron, Y., Richard, P.J.H., and geographical revision’, (translated an area, with the consequences of that natu- Frechette, B., Gauthier, S. and Prairie, from Skvortsov, A.K. (1968). ‘Willows of ralisation. Likelihood will be assessed by Y.T. (2001). Change of fi re frequency the USSR. Taxonomic and geographic analysing the invasiveness, or biological in the eastern Canadian boreal forests revision’, Nauka, Moscow), Joensuu traits of each taxon, such as ability to es- during the Holocene: does vegetation University, Joensuu. www.salicicola. tablish, reproduce and disperse. The con- composition or climate trigger the fi re com/announcement/skv/skvortsov. sequences of a taxon’s ability to naturalise regime? Journal of Ecology 89, 930-46. pdf can be determined by the product of its Carr, G.W. (1996) Salix. In ‘Flora of Vic- Stace, C., van der Meijden, R. (ed.) and de impact and its potential distribution. The toria’ Vol. 3, eds N.G. Walsh, and T.J. Kort, I. (ed.) (2006). Salix in Interactive potential distribution of each willow will Entwisle, (Inkata Press, Melbourne). fl ora of north western Europe. ETI BioIn- be extrapolated from current distribution Champion, P. and Clayton, J.S. (2001). A formatics, viewed: 2006, http://ip30. data. When this process is complete, our weed risk assessment model for aquatic eti.uva.nl/bis/fl ora.php?selected=besc willow rankings can be utilised to fully weeds in New Zealand. In ‘Weed risk hrijving&menuentry=zoeken&zoekna inform management requirements for wil- assessment’, eds R.H. Groves, F.D. Pan- am=salix# lows in Australia. etta and J.G. Virtue (CSIRO Publica- Stokes, K.E. and Cunningham, S.A. (2006). tions, Melbourne). Predictors of recruitment for willows

Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 85 invading riparian environments in van Kraayenoord, C.W.S., Slui, B. and Sustainability and Environment, Mel- south-east Australia: implications for Knowles, F.B. (1995). ‘Introduced for- bourne, Australia. weed management. Journal of Applied est trees in New Zealand recognition, Weiss, J., Edgar, R., Hunt, T. and Morfe, Ecology, 43(5), 909-21. role and seed source. No. 15. The wil- T. (2004). Victoria’s noxious weeds re- Standards Australia International (SAI) lows Salix spp.’ (New Zealand Forest view: roll out not fall out. Proceedings (2006). ‘Handbook 294:2006 National Research Institute). of the 14th Australian Weeds Confer- post-border weed risk management Victorian Government Gazette (2005). ‘G43 ence, eds B.M Sindel and S.B.M. John- protocol’. Standards Australia, Sydney. – Declaration of certain plants to be re- son, pp. 707-10. (Weed Society of New Tennant, D.J. (2004). A re-assessment of stricted weeds, 27 Oct 2005’, p. 2389. South Wales, Sydney). montane willow (Salix L., Salicaceae) Victorian Government, http://www. Welsh, S.L. (1974). Anderson’s flora of hybrids in Scotland. Watsonia, 25, 65- gazette.vic.gov.au/Gazettes2005/ Alaska and adjacent parts of Canada. 82. GG2005G043.pdf. Brigham Young University Press, Provo, Thorp, J.R. and Lynch, R. (2000), ‘The Webb, C.J., Sykes, W.R. and Garnock-Jones, UT. http://infor.dec.state.ak.us/ci- determination of Weeds of National P.J. (1988). ‘Flora of New Zealand Vol. imms/newdocs.htm Significance.’ National Weeds Strat- 4’. (Manaaki Whenua Press, NZ). egy Executive Committee, Launceston, Weiss, J.E.R. (2007). Victoria’s aquatic weed Australia. risk assessment process. Department of

Appendix 1. Criteria used in the willows impacts assessment Criteria Intensity rating Score Socio-economic 1. How much damage could Visual effect; little to negligible structural damage. L be caused to human-built Able to be remedied as a normal part of everyday maintenance e.g. pruning ML infrastructure? Maintenance requires specialised equipment, such as for clearing drains MH Major damage to bridges, culverts, weirs, dams etc. requiring repair. H 2. How much horticultural/ Attractive/useful foliage/catkins/stems; OR valued as a fodder, shade, wood, L agricultural value does the willow windbreak or avenue tree; AND require knowledge to propagate, AND are single-sex have? clones Attractive/useful foliage/catkins/stems; OR valued as a fodder, shade, wood, ML windbreak or avenue tree; AND require knowledge to propagate. Some horticultural/agricultural value, but easy to propagate, bisexual or male and MH female, naturalised in Australia No aesthetic value, easy to propagate, bisexual or male and female, naturalised in H Australia 3. To what extent could the willow Little to no impact on public safety. Willow is low-growing and/or has branches that L impact on the health and safety of are too thin to cause serious health damage. waterway/riparian users? n/a ML Moderately likely to cause serious injury or death of waterway/riparian users. E.g. MH tree willow/tall shrub with large, but fl exible branches. Most likely to cause serious injury or death of waterway/riparian users. E.g. tree H willow with brittle branches. Risk of death to water skiers. 4. To what extent could the Little to no impact on activities. Weeds not obvious to average visitors. L taxon impact on recreation in/ <4 activities affected. Minor effects to aesthetics or uses. E.g. willows able to form ML on waterways? E.g. swimming, monocultures that reduce bird life and impede river views for passive enjoyment, but boating (including canoeing, skiing, access for swimming, boating and fi shing is still possible. rafting), fi shing, bird watching, passive enjoyment e.g. picnics 4+ activities impeded. E.g. stream deep enough to boat/swim, but access impeded by MH willows on the bank. Major impact. 4+ activities prohibited. E.g. willows encroach into stream, making it H too shallow to swim/boat. continued/…

86 Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 …/Appendix 1 continued. Criteria Intensity rating Score Stream health 5. To what extent could the willow Little or negligible impact on fl ow capacity or water availability. Willow grows L impact on the fl ow of water in offstream with no root or stem growth in stream. A streams and on water availability ? Minor impact of fl ow by roots or foliage. Roots sometimes grow instream. Capable of ML removing more water than vegetation lacking instream root systems. Major impact on fl ow by roots and foliage including major root structure. Roots and MH stems often grow instream. Capable of using large quantities of water. Always extensive roots and stems growing in stream, causing streams to become H shallower and wider. Also capable of using the most water. 6. To what degree could the willow Low probability of large scale soil movement. Does not grow in riparian areas. L cause bed and bank erosion? Moderate probability of large scale soil movement. Terrestrial species that suppresses ML the understorey and lacks extensive root system, allowing erosion of the banks by overland runoff. High probability of large scale soil movement, but effects remain in stream. Willow MH roots and stems encroach instream to create a wider, shallower stream. High probability of large scale soil movement and major off site implications and H bank failure. Willows are confi ned to the banks, but under fl ood conditions the stream is diverted behind the willows, scouring out large areas of land. E.g. S. exigua 7. To what extent could the Plant is low growing and unlikely to affect the shading of waterways or to drop many L willow affect water quality (and leaves into the stream AND/OR doesn’t grow near waterways consequently, instream native n/a ML biodiversity) i.e. nutrient pulsing, light availability and temperature Grows along the bank to 4+ m tall such that large amounts of leaf litter will fall into MH change. the stream and/or has the ability to cause unseasonal opening in the canopy by signifi cantly out competing native vegetation. Plant overhangs stream, or encroaches into stream such that most of its foliage will H fall into the water. Weeping tree or prostrate form. Biodiversity 8. To what extent could this willow Minor or negligible effect on <20% of the fl oral strata/layers present; usually only L affect riparian/wetland habitat affecting one of the strata OR not known as a weed anywhere in the world. structure/layers? i.e. ground layer Minor effect on 20–60% of the fl oral strata. Does not form large thickets. ML (forbs, grasses, herbs) shrub layer, Minor effect on >60% of the layers or major effect on <60% of the fl oral strata. Large MH tree layer thickets interspersed with other vegetation Major effect on all layers. Able to form monocultures; virtually no other intact strata/ H layers present. 9. To what extent could riparian Coexists with other vegetation in any of the riparian niches and is not dominant OR L areas (in-stream, margins, banks, does not grow in riparian environments. fl oodplain, wetlands) be impacted Occurs as the dominant species in any one of the riparian niches. ML by this willow? Occurs as the dominant species in any two of the riparian niches. MH Occurs as the dominant species in any three of the riparian niches, and/or is capable H of invading wetlands. 10. To what extent could this Suppresses (e.g. Glyceria). No associations formed with other invasive species L willow affect other invasive species May occur in association with minor pests, such as blackbirds or non-declared weeds ML (fl ora and fauna)? n/a MH May occur in association with serious (declared) pests, such as rabbits, foxes or H blackberry

Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 87 Appendix 2. Criteria used to determine confi dence scores for the literature used to answer each question. Document type or information source Rating Peer-reviewed scientifi c papers H High quality science or plant specifi c books (e.g. fl oras), non-peer reviewed scientifi c papers (e.g. conference MH proceedings), personal communications from expert on species being assessed, unpublished reports from highly reliable source (e.g. commercial reports or honours theses, etc.), internet information from herbaria or internet information that cites sources. Personal communications from people with experience with the species under assessment, information from M general plant books, (e.g. Encyclopaedia Botanica, Gardening Flora, etc.), unpublished reports from uncertain sources, internet information from government or university websites that does not cite sources. Anecdotal data from non-experts, internet information from uncertain/uncited sources, or horticultural, nursery ML notes or general web pages. No data or reference material available. L

Appendix 3. Victorian Pest Plant Prioritisation Process impacts criteria. Social values 1. How could the weed restrict human access? 2. How could this weed reduce the tourism /recreational use of the land? 3. Is the plant injurious, toxic, or spines affect people? 4. How much damage is done to indigenous or European cultural sites? Natural resources 5. How could this weed impact on water quality 6. How could this weed impact on quantity? 7. How could the weed increase soil erosion? 8. How could this weed reduce the biomass of the community? 9. How could the weed change the frequency or intensity of fi res? Fauna and Flora/Vegetation 10. How could this weed impact on the vegetation composition on the following: a. High value vegetation? b. Medium value vegetation? c. Low value vegetation? 11. How could this weed effect the structure of a vegetation community? 12. How could the weed have on threatened fl ora spp.? Threatened Flora and Fauna 13. How could the weed impact threatened fauna spp. 14. How could the weed impact non-threatened fauna spp. 15. How could this weed provide benefi ts or facilitate the establishment of indigenous fauna? 16. How is the plant toxic, its burrs or spines affect indigenous fauna? Pest animal 17. How could this weed provide a food source to assist in success of pest animals? 18. How could this weed provide habitat to assist in success of pest animals? Agriculture 19. How could this weed impact on the quantity of agricultural produce? 20. How could this weed impact on the quality of agricultural produce? 21. How could this weed affect land value? 22. How could this weed cause a change in priority of land use? 23. How does the weed increase the cost of harvest? 24. Does this weed act as an alternative disease host or vector?

88 Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 Appendix 4. Scores from the willows impacts risk assessment.

Taxon name Q1 Q2 Q3 Q4 Q5 Q6 Q7 Q8 Q9 Q10

S. subgenus Chamaetia MMLLLLLLLLML S. subgenus Salix HMHHHHMHHHHH S. subgenus Vetrix MHMHL H HMHHMHHH S. aegyptiaca MHMHLMLLMLMHMM M S. alba HMHHHMHMHHHHH S. alba var. caerulea LLLLLLLLLH S. babylonica HMHMHHMHHHHH S. caprea MHMHLMLLMLMHHH M S. cinerea MHHLHHMHHHHH S. daphnoides ML M L MH L H MH H H H S. elaeagnos MHMLLMLMHMMHMLLL S. eriocephala MH MH L MH ML ML MH MH MH H S. exigua MHM L H HMHHMHHH S. fragilis HHHHHMHHHMHH S. glauca M M LMHM HMHMHHH S. glaucophylloides L H L MLM MMHMM M S. gracilistyla MH M L MH ML H ML MH ML H S . humboldtiana ‘Pyramidalis’ MH MH L L ML LLLLM S. integra ‘Hakuro-nishiki’ L M LLLLLLLM S. matsudana H MH H ML ML ML H ML L H S. myrsinifolia LMLMLMLMMM S. nigra HHHHMHMHHHHH S. pentandra MMMM L MMHMHH S. purpurea MHMHL H HMHHMHHH S. triandra MHMHLHHMHHHHH S. viminalis MHH L H MHMHMHH H H S. alba × matsudana HMHH MMHMMHH M H S. × ‘Boydii’ LLLLLLLLLL S. × calodendron. MH H L M M M MH M H M S. × mollissima MH H L M M M MH MH M H S. × pendulina HMHHMHM H H M M M S. × reichardtii MH M L M H MH MH M M H S. × rubens HHHHHMHHHHH S. × sepulcralis HMHHHHMHHHMHH S. × sericans MH L MMMMHMMM

Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 89 Why are growers using the scheme? There are two main reasons why grow- Community based quality assurance scheme ers are using the scheme. The fi rst is to get acknowledgement that their property Brian FeldtmannA and Ken YoungB is managing weeds well. This gives the A Dookie Land Management Group, Dookie, Victoria 3646, Australia. land manager a status within the district B University of Melbourne, Dookie College, Victoria 3647, Australia. and puts pressure on neighbouring prop- erties to improve weed management on their properties. This recognised status also aids land managers in discussions on weed management with Government Introduction Recognised high level weed authorities, whether this be on weed man- Managing weeds on private and public managed property agement issues on adjacent land or on pol- land is problematic; with some land man- In order for a property to be included as icy matters to do with land management. agers being very diligent and others no so. a Recognised High Level Weed Managed Some of the topics within this policy area Where weeds are not being managed can Property, the landholder or their agent have included a rate incentive scheme for cause weed incursion on to adjacent land. must: properties recognised under the DLMG Legislation is targeting those who do not • Nominate a defi ned area boundary to ‘Recognised High Level Weed Managed manage weeds in accordance to the regu- be inspected. Property’ scheme. lations set down by Government and their • Give authorisation of DLMG repre- The second reason why growers have representative bodies. This can tie up a lot sentatives to inspect the land referred taken up the scheme and stayed with the of resources with little positive outcomes to above and to erect or remove the scheme is in marketing their own pro- in the fi eld. The Dookie Land Management Recognised High Level Weed Managed duce. When the scheme was fi rst initiated, Group (DLMG) decided to turn the table Property Sign in accordance to the in- several of the properties that joined the on this process and recognise those in the spections. scheme were growing crops for seed pro- district that do a good job of managing • Give approval for Inspectors appointed duction, others had hay production as a their weeds as part of their Weed Manage- by DLMG to enter this portion of land major component of their farm enterprise. ment Implementation Strategy. at least three times of year, and The scheme gave them a quality assurance In 2002, the Dookie Land Management • Have less than fi ve plants of any one programme for their customers. Group started the Recognised High Level species and no more than 20 plants of Weed Managed Property Quality Assur- combined species that are nominated Effectiveness and uptake of the ance Scheme. The Scheme is supported by within the Goulburn Broken Catch- scheme the State Department of Primary Indus- ment Authorities declared weed list, Since its inception in 2002, the DLMG tries, the City of Greater Shepparton and observed by the inspectors. ‘Recognised High Level Weed Managed the University of Melbourne. This support Property’ scheme has had fi ve to six prop- is both through agreement of the principle How the scheme operates erties listed. In its inaugural year, there aims of the scheme and allocating time for The property is inspected at least twice a where seven properties that requested to staff to be involved in the inspection re- year, after the autumn break and again in join the schemes but only fi ve achieved the gime set up by the scheme. spring. level required by the scheme. While there The inspectors travel over the whole is presently only six properties recognised DLMG Weed Management property and look for any of the weeds under the scheme, the level of weed infes- Implementation Strategy listed within the Goulburn Broken Catch- tations both on these properties (as noted Aims of the strategy are to: ment Authorities declared weed list. A for- on the inspection reports with little fol- • Implement actions in accordance with mal report is written (see form at end of low up actions required) and on adjacent the Goulburn Broken River Catchment paper) and any recommended actions are land has been reduced since the start of Weed Action Plans, also noted within the report. The report is the scheme. The scheme has had no for- • Prevent the incursion of new weeds signed off by a member of the DLMG and mal publicity other than the signage on the within the DMLG area, and a DPI offi cer. Weeds that are not on the list respective properties and word of mouth. • Reduce the weed levels within the are in great numbers are also be noted on Several of the surrounding councils are DMLG area. the inspection report. keeping a watching brief on the scheme to If the property has some recommended see if it fi ts within their weed management The initial objects set out in 2002, where: actions listed during the inspection a fol- implementation plans. • In 2002–2003, have weeds managed low up inspection is conducted within along roadsides in the area bounded by six weeks of the inspection. If the recom- South Boundary Rd, Dookie – Nalinga mended actions are have not been done Rd, Cashels Rd, and the property losses it status. • Have landholders involved in a co-or- If the property meets the criteria of hav- dinated roadside weed management ing less than fi ve plants of any one spe- programme by 2006. cies and a total of lees than 20 combined • Management of all weeds listed in the species the property can have access to book ‘Weeds of the Goulburn Broken’ the signage ‘Recognised High Level Weed within fi ve years. Managed Property’. The cost to the grower • Recognise property owners who have is an annual fee of $50.00 as well as main- successfully implemented a high- taining their high level weed managed level weed management programme properties. through an inspectorial accreditation scheme.

90 Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 Recognised high level Weed Managed Property Inspection Report Date: ...... Property: ...... Owner/ Manager: ...... Phone No.: ...... Property Boundary References (either Grid Ref form CFA map or GS readings) ...... I authorise DLMG representatives to inspect the land referred to above and to erect or remove the Recognised High Level Weed Managed Property Sign in accordance to the inspections.

Name: ...... Signature: ...... –––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––––– Weeds Observed Weed <5 >5 <5 >5 <5 >5 plants plants plants plants plants plants

Alligator Weed Wild Garlic Great Mullein Black Knapweed Kyllinga Hardheads Ivy-leaf Sida Parramatta Grass Hoary Cress Mesquite Arrowhead Horehound Water Hyacinth Carpet grass Noogoora Burr Chilean Needle Grass Amsinckia Parrot’s Feather Ragwort Artichoke Thistle Parthenium Weed Serrated Tussock Bathurst Burr Saffron Thistle Sliverleaf Nightshade Boxthorn Salvinia African Lovegrass Caltrop Slender/Shore Thistle Blackberry Camelthorn Spear Thistle English Broom Cape Broom Spiny Burr Grass Furze Cape Tulip Star Thistle Paterson’s Curse Devils Claw Stemless Thistle Prairie Ground Cherry Dodder Thorn Apple St John’s Wort Golden Thistle Variegated Thistle Sweet Briar Viper’s Bugloss Inspectors Name: ...... Signature: ...... Name: ...... Signature: ...... Name: ...... Signature: ...... Name: ...... Signature: ...... Property Approved ...... Yes ...... No Recommended Actions: ......

President of the Dookie Land Management Group DPI Representative

Signatures: ......

Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 91 for this specifi c purpose. Vigilant® herbi- cide gel containing 43 g kg−1 picloram as An investigation of application techniques for the the potassium salt has provided a solution to this problem and offers a low-impact control of various noxious and environmental weeds way to control woody and rhizomatous with picloram gel based herbicides plants. Vigilant is registered for application by Geoffrey G. Keech, Macspred Australia, 714 Duri-Dungowan Road, cut stump, blazing (stem injection), drill- Timbumburi, New South Wales 2340, Australia. ing and foliar application (on wandering Jew with a paint roller). There remains however some species and growth stages which are not suited to these techniques. To overcome this anomaly and to de- Abstract Over 2500 species of introduced defoliation. The Senna trial was established velop a more effi cient means of control, plants are now established in the wild in in October 06 and needs to be monitored trials in conjunction with several Local Australia, and many threaten the integrity until Spring 07 for meaningful results. Government Weed Control Authorities of some of our most valued places. For Foliar and stem injection shows a great were set up to investigate both species many of these environmental weeds, there deal of promise on the Opuntia species and susceptibility and application techniques is no registered herbicide. Are we winning the results may lead to an inclusion on the for the most effi cient control of: African the battle, do we have the tools to win and label. Basal bark treatments on thin and olive (Olea europaea ssp. africana), arum is research keeping up with the spread? green barked species have proven to be lily (Zantedeschia aethiopica), broad-leafed In an effort to facilitate better control effective to date and offer a signifi cant sav- pepper tree (Schinus terebinthifolia), cat’s methods, trials using a picloram based gel ing in time. claw creeper (Macfadyena unguis-cati), For- herbicide, were set up in conjunction with mosan lily (Lilium formosanum), glory lily several Local Government Weed Control Introduction (Gloriosa superba), Madeira vine (Anredera Authorities. The aim was to investigate In an ideal world we could have grown cordifolia), prickly pear or common pest the application techniques for the most what we liked in our gardens, and re- pear (Opuntia stricta), smooth tree pear effi cient control of the following species: tained unspoiled forests and beaches to (Opuntia vulgaris), senna (Senna pendula African olive (Olea europaea ssp. africana be enjoyed on holidays. According to Mc- var. glabrata). (P.Mill.) P.Green), arum lily (Zantedeschia Fadyen (2003), ‘The arrival of Europeans aethiopica (L.) Sprengel), broad-leafed in Australia has been followed in the last Techniques and methodology pepper tree (Schinus terebinthifolia Raddi), 200 years by over 28,000 foreign plants. Those methods chosen for investigation cat’s claw creeper (Macfadyena unguis-cati Some introductions were accidental, but were standard label claims and some com- (L.) A.Gentry), Formosan lily (Lilium for- most were imported for pasture, horti- monly used methods in bushland regener- mosanum Wallace), glory lily (Gloriosa su- culture or for ornamental reasons. Of the ation which included cut stump, stem in- perba (L.), Madeira vine (Anredera cordifolia nearly 300 plants known to have estab- jection by cutting blazes and wedges and (Ten.) Steenis), prickly pear or common lished themselves as weeds in the wild by drilling and fi lling (Ward and Henzell pest pear (Opuntia stricta var. stricta (Haw.) between 1971 and 1995, for example, two- 2002, 2004). Other methods considered Haw.), smooth tree pear (Opuntia vulgaris thirds were introduced as ornamentals. were foliar, basal bark and scrape and Miller) and senna (Senna pendula var. gla- Of the 460 pasture and legume species paint methods. The treatment methods for brata (Vogel)). trialled in northern Australia 1947–85, 60 each species are shown in Table 1. Application techniques investigated in- became weeds and 13 of these are now se- cluded: cut stump, scrape and paint, stem rious crop weeds. Only four proved useful Cut stump injection, basal bark and foliar smearing, without also causing weed problems. One Stems treated by this method were cut not all techniques were trialled on each plant introduced for pasture became a ma- horizontally (Figure 1) so as to leave a species. jor weed within a decade’. ‘Over 2500 spe- fl at non-sloping surface to receive the gel. All species treated according to label cies of introduced plants are now estab- Stems were cut at 20 mm to 50 mm above directions for cut stump application were lished in the wild in Australia, and many the ground (except for one multi-stemmed satisfactorily controlled. The exceptions threaten the integrity of some of our most plant each of African olive and golden were for African olive and golden rain tree valued places’. Introduced plants have rain tree). A further trial was conducted where one multi stemmed plant for each found their way into nooks and crannies on senna, cutting the stems of dense juve- species was cut and treated at a height where some have become so fi rmly estab- nile stands at about knee to waist height. of 30 to 40 cm above ground level. These lished that it’s been close to impossible to This was done in an attempt to achieve plants re-shot and may survive. Long term remove them. They are now posing a sig- easier application for elderly volunteers. observation of the re-shot plants indicate nifi cant threat to our native species and The stems at each site were cut with ei- that there is still herbicide activity occur- our biodiversity. We need to ask ourselves ther secateurs, loppers or a pruning saw, ring. As picloram may take up to two sea- are we winning the battle, do we have the depending on stem diameter. Vigilant gel sons to fully control some species, further tools to win and is research keeping up was applied to the cut surface in a layer observation will take place. It should be with the spread? 3–5 mm thick within fi ve minutes of cut- noted that all plants cut at 2 to 5 cm were In recent times there has been an up- ting. satisfactorily controlled indicating the surge in the concern over environmental need to follow label directions. weeds, in particular those weeds which Foliar/stem smear Scrape and paint and basal bark tech- have escaped from garden situations. For Opuntia spp. Vigilant gel was applied niques were employed for Madeira vine, Many of these species which threaten as a thin smear covering 75–80% of the cat’s claw creeper and senna. Madeira native vegetation areas are of the woody leaf surface. One side of the stem/leaf was vine was successfully controlled and also weed variety and are well known to bush- treated using a stiff bristled brush, ensur- a very high percentage of the aerial tubers. land regenerators. Until recently, the con- ing treatment to below or on the lowest The cat’s claw creeper trial was compro- trol of these weeds has been hampered leaf. For the lily group, treatment of the mised, however early indications showed due to the lack of a registered herbicide stems and/or leaves or midrib of the leaf

92 Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 was achieved with the Vigilant brush bot- Table 1. Treatment methods and species trialled. tle (see species trial results below). Species Cut Stem Leaf and Basal Scrape and stump injection stem smear bark paint Scrape and paint African olive X Stems were scraped on 2 sides (Figure 2) Arum lily X X to expose the outer cambium layer for a combined area of approximately one third Broad-leafed pepper tree X X to half the circumference of the stem from Cat’s claw creeper X X X X a height beginning at approximately 20–50 Formosan lily X mm and for a distance of 300 mm up the Glory lily X stem. Vigilant gel was applied as a smear Madeira vine X X over the stem using a stiff bristled brush, Prickly pear X X sold in supermarkets as a dishwashing Smooth tree pear X X brush, dispensing the gel from the 1.8 kg Senna X X X tub. For single plant treatments, the Vigi- lant brush bottle was used (not recom- mended because of possible damage to the soft bristles on the brush).

Stem injection The technique for stem injection varied ac- cording to the species trialled. For soft tis- sue plants, a hole was cut in the base of the plant stem with a narrow bladed knife or a specially developed lance. Vigilant gel was placed into the hole with a squeeze bottle fi tted with a long nozzle. Woody stemmed species were treated by cutting blazes with a tomahawk or wedges with a pruning Figure 1. Cut stump. Figure 3. Stem injection saw (Figure 3). Vigilant was applied from either the brush bottle, a squeeze bottle or with a stiff bristled brush from the 1.8 kg tub. Further work is anticipated with both drilling and frilling techniques.

Basal bark Vigilant gel was applied by direct applica- tion to the stem of the target plant (Figure 4) using a stiff bristled brush. The stems were treated from 20 mm above ground level for a distance of approximately 300 mm up the stem.

Results African olive (Olea europaea ssp. africana) (Cessnock City Council) Figure 2. Scrape and paint. Figure 4. Basal bark. Established: 8/3/05 Treatment method: cut stump Stems treated: 11, Control (untreated):3 Results Assessments made on 5/7/05 re-shooting on the multi-stemmed plant Stems treated: 10 each method and 1/9/05 showed the untreated controls had died however there was still one shoot Results Assessed 22/2/06 and December were already re-shooting and all treated with fresh green leaves at the very base 06 showing death and desiccation of all stems appeared to be controlled. One of of the plant. These leaves still exhibited treatments. The site was inspected by Ed- the treated stems was a multi stemmed symptoms of herbicide activity. This is in die Lanting, Weeds Offi cer with Gosford plant and was cut at about 300 mm from agreement with the known data that in- City Council on 12/12/06 and all treat- ground level. Initially it appeared that this dicates activity for up to two seasons in ments show 100% control. plant along with those cut at 20–50 mm, plant tissue. A fi nal assessment on 3/4/07 was controlled. On 2/11/05 this plant had showed no green shoots and it would ap- Cats claw creeper (Macfadyena unguis- shoots approx 300 mm long growing from pear that control has been achieved. cati) (NSW DPI) the base and a smaller deformed shoot The trial was compromised by an over- Established: 18/5/05 (in conjunction with from further up the stem. All other stems spray with Grazon® DS in June 05. The DPI as part of a larger trial) were still controlled and showed signs of untreated controls recovered from this ap- Treatment method: cut stump, stem injec- decay. On 16/2/06 the multi stemmed plication and are all actively growing. tion, scrape and paint, basal bark. plant showed evidence of some dead Stems treated: 3–5 each treatment. shoots but also had several healthy shoots Arum lily (Zantedeschia aethiopica) Results Initial assessments on 23/6/05 while the plants cut close to the ground (Gosford City Council) indicate excellent control with cut stump were still controlled and showed more Established: 2/11/05 applications. Basal bark applications on evidence of decay. Further investigation Treatment method: cut stump, leaf/stem larger stems had defoliated the canopy on 21/10/06 showed that the majority of smear. but it was too early to ascertain the level

Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 93 of control. Small diameter (<10 mm) were The trial was duplicated on a smaller The second assessment for Yamba and controlled above ground level. Stem injec- scale and Vigilant gel applied with a paint fi rst for Pt. Macquarie 9–10/2/06 showed tion treatments on large diameter stems brush wiped across the top of the plot. severe damage to the smaller plants and had defoliated the canopy, assessment at Some phytotoxicity resulted and some signifi cant damage to large plants. Except a later date indicated that the control was plants were controlled but insufficient one large plant which had been treated effective. All uncut stems including the product was applied to be totally effective. as a basal bark only, this plant has not stem injection plants were compromised This trial may need to be re-established been fully controlled due to insuffi cient by the ‘Vine Weeds Work Team’, cutting using a paint roller technique as for wan- coverage; only the treated leaves have and treating the trial plots with a glypho- dering Jew. Vigilant has proved effective died. sate solution, at approximately nine weeks as a treatment for isolated patches and in- The assessment of 21–23/3/06 showed after the original Vigilant treatments had dividual plants shown by in fi eld use, but severe damage and death of smaller plants been applied. The only valid data avail- would prove to be uneconomic for large and further evidence of increased dam- able from the site was for the cut stump infestations. For best results use only a age to large plants. Coverage of at least method which yielded 100% control af- thin smear on the stem and leaves, foliar 75–80% of the surface of one side of all of ter 12 months. Small diameter stems (<5 application with a paint roller may be an the main leaves and stem is needed for mm), which were treated by the basal bark option but at this time is unproven. optimum control as plants only partially method but were lost due to cattle move- treated have survived. All plants treated ment, also appeared to be satisfactorily Madeira vine (Anredera cordifolia) (Coffs by leaf smear have been controlled and it controlled. The major problem in achiev- Harbour City and Port Macquarie- is the preferred option. Stem injection is ing control of cat’s claw creeper is the Hastings Council) slower and all main stems of large plants seedling regeneration and ground cover A trial was established by Tony Cook, DPI need to be treated. Complete control is creepers re infesting the trees. Inspection NSW at Ellenborough, using several dif- attainable with follow up treatments on at two years shows all material below ferent herbicides. At the initial assessment leaves which drop off or are broken off due ground level for the cut stump treatments all Vigilant treatments were deemed to be to the rapid collapse of the main stems. had rotted away. controlling the plants. Unfortunately the No root production from the majority site was fl ooded and most of the data was (>95%) of collapsing plants has been ob- Formosan lily (Lilium formosanum) lost and no conclusions were drawn. served. The fallen leaves/stems in most (Port Macquarie-Hastings Council) Madeira vine was treated at Coffs Har- cases have eventually desiccated and are Established: 22/2/05 bour on 28/9/04 as part of the demonstra- no longer viable. Treatment method: cut stump. tion site on Coffs Creek. Only two plants Stems treated: 10 were treated by scrape and paint method. Smooth tree pear (Opuntia vulgaris) Results Assessments in March 05 and Both were successfully controlled. Aerial (Greater Taree City Council) April 05 indicated no regrowth from the tubers were collected on 25/11/04 and ex- Established: 23/11/05 bulbs. Treated and untreated plants were amined internally. There was signifi cant Treatment method: Stem injection, leaf dug up on 25/5/05 for comparison. The evidence that the tubers were controlled. smear treated plants showed fragmentation and Other plants have been treated as a basal Plants treated: 10 each method separation of the bulb from the stem/root bark treatment and anecdotal evidence in- Results Assessments were made 22/ system while the untreated controls were dicates good control of plants and tubers. 12/05, 22/2/06 and 21/3/06. The initial still intact. Several samples of both treated A large vine was treated by DEC staff and assessment was made by George Wise- and untreated bulbs were collected and total control was achieved. The tubers on mantle, Weeds Offi cer at Greater Taree planted in a garden for observation. The this plant also appeared to be controlled. City Council. Significant phytotoxicity untreated bulbs sprouted and grew the Basal bark treatments will always work was observed in both treatments with the next season while there was no growth better when applied to green bark as pen- stem injection treatments burning the in- exhibited by the treated samples. etration of mature bark on older plants is jection site so severely that the plants had shown to be variable. snapped of at this point. Some concern Glory lily (Gloriosa superba) (Bellingen Plants were also treated at Hornsby on was held that translocation of the product Shire Council and Bellingen Landcare) 14/9/04 with similar results. would be reduced because of this effect. Established: 25/11/04 Subsequent assessments have shown that Treatment method: Foliar and stem Prickly pear (Opuntia stricta) (Port this is not the case, continued monitoring smear. Macquarie-Hastings and Clarence Valley has resulted in the observation of complete Stems/area treated: Plot size 2 m × 2 m. Councils) tissue desiccation and satisfactory control #1–310 plants, #2–265 plants, approx 50% Established: 8/11/05 – Yamba, 23/11/05 has occurred. treated. #3–untreated 195 plants. – Port Macquarie. The foliar treatment has shown excel- Results Plot 1. The top 100 mm every Treatment method: (a) Stem injection by lent control as long as all of the main stems stem was treated by holding in a gloved cutting a plug out of the base of the stem and leaves receive treatment on at least hand and smearing with Vigilant gel. Se- or placing slits in leaves, and (b) Foliar one side of the leaf. Where a large leaf was vere burning of all plants was evident at smear. not treated on one plant, there seems to seven days with the exception of some un- Plants treated: Yamba – 15 stem injection, have been little translocation of the prod- treated plants, an indication of overdosing 10 leaf smear. Port Macquarie – 40+ leaf uct from the stem to the leaf. The reason which may have reduced translocation of smear, 3 stem injection. for this is may be that there was insuffi - the herbicide. Reassessed 22/2/05 and Results Yamba 7/12/05. All treatments cient product applied to the stem below counted 20 plants with severe phytotoxic- showed low to moderate phytotoxicity in the junction with the leaf. ity, 51 slight to nil effect, some of these had the main leaves at fi rst assessment. How- Picloram moves both in the xylem and obviously not been treated. 77% reduction. ever the fl owers and small fruit and some phloem tissues but does not translocate 124 plants present 9/12/05, mostly fresh of the outer leaves showed severe phyto- between these tissues. In other words, germinations. Plot 2. Only the treated toxicity and burning, severe site burning it moves up and it moves down but not plants were burnt at seven days, 152 plants was noted at stem injection sites which down from higher applications and back present 22/2/05, 438 plants 9/12/05. Plot may have reduced translocation. Port up the stem beneath this application. 3 had increased to approx 640 plants. Macquarie was not assessed in December.

94 Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 Senna (Senna pendula var. glabrata) re-shooting while untreated plants have Broad-leaved pepper tree Stem injection (Port Macquarie-Hastings Council and vigorous regrowth. by blazing carried out on a large tree in Dept. Conservation and Climate Change Taree has proved to be ineffective to date NSW–Parks and Wildlife Division) 3. Established: 14/12/06 and a smaller tree treated similarly was Several trials on this species were estab- Treatment method: basal bark not completely controlled initially but lished as listed below. Plants treated: 53 plants up to 15–20 mm eventually died, however cut stump appli- 1. Established 9/02/06 A small scale tri- in diameter were treated on one side of the cations on other plants of the same species al was established on 9/2/06 using cut stem from 20 mm above ground level for a have proved to be effective. Observation stump both high and low cuts and basal length of 300 mm. of the stem injection technique indicates bark applications. Insuffi cient numbers of Results The site was inspected on 21/2/ that cuts should be made at distances of plants were treated to derive meaningful 07 and 10/5/07; 52 plants were dry and no greater than 50 mm apart and should data although all treated plants appeared brittle and easily removed from the soil. be located so that a cut is made directly be- to have been controlled. A second site Adjacent live plants were not able to be low each branch. This species may require was established on senna with stem di- removed by hand pulling. One double further refi ning of the stem injection tech- ameters <10 mm diameter in conjunction stemmed plant which had been treated nique with a complete frill being preferred with Dept. of Conservation and Climate above the fork showed half the plant still as the next option. Change NSW, National Parks division, alive on 21/2/07. This is typical of the cutting at above knee height to trial an eas- mode of action of picloram. All stems of Conclusions ier method for elderly volunteers, assess- multi stemmed plants need to be treated The gel formulation of picloram (Vigilant) ment in June 06 showed re-shooting of the or treatment should be below the lowest has shown to be effective on a range of majority of these stems however all shoots branch. invasive plant species. The most effective were showing typical Vigilant damage. An application technique for all woody spe- assessment on 14/3/07 showed that two Small scale trials and cies with stem diameter larger than 10 mm plants >10 mm diameter were controlled demonstrations is the cut stump method. For smaller stem and eight plants which were <10 mm in Other small scale trials (look and see type diameters, scrape and paint application or diameter had regrowth. This indicates that demonstrations) have also been conduct- for thin or green bark species, basal bark it is diffi cult to apply suffi cient product to ed during 04/05. Some of the results are application has proved effective. Basal small diameter stems to achieve control by surprising and reveal the need to under- bark treatment offers the added benefi t the cut stump method. stand not only the botanical structure of of time saving along with reduced bend- the plant but also the activity and mode ing down time for elderly volunteers. Cut 2. Established: 27/9/06 of action of the active ingredient. The Af- stump application is best suited to manage Treatment method: Scrape and paint, basal rican olive and smooth tree pear experi- larger trees and shrubs and those plants bark ences have been described above. Other with thick bark, application should be Plants treated: 80+ plants from 5 mm to 20 instances have also been discovered where within 100 mm of the ground for a faster mm diameter at two sites were treated as application technique needs to be fine and more effective kill. The experience scrape and paint and two plants (multi- tuned and adapted to suit the species. for senna, African olive and golden rain stemmed) to 20 mm diameter were treated tree, where stems were cut at a height of by basal bark application. Golden rain tree One multi-stemmed 300–400 mm and higher; which is outside Results Rainfall was experienced shortly plant which was cut at about 300 mm re- label claims, indicates that plants treated after application however early results ob- sprouted from the base, while others cut in such a manner may re-shoot and sur- served on 24/10/06 were encouraging with low, as in the Cessnock/African olive ex- vive or take longer for complete control all plants showing signs of defoliation. On perience were effectively controlled. Fol- to be effected. This problem is likely to be 29/11/06 all small diameter plants treat- low label instructions, cut at about 20 species related as according to Ward et al. ed by the scrape and paint method were mm above ground and certainly no more (2002) there was little difference in effi cacy completely defoliated and the stems above than 100 mm. against grey willow (Salix cinerea) between ground were dry and brittle. The larger stems cut at 300 mm and those cut at 0.5 to plants had also completely defoliated and Camphor laurel Cutting blazes in the 1 metre. Long term observation of the re- the smaller branches were dry and brittle. trunk of the tree and applying product to shot olive and senna plants indicate that The larger multi-stemmed plant treated the fl at part of the cut (bottom) does not there is still herbicide activity occurring with the basal bark method had defoli- allow product to move up the plant and with stems greater than 10 mm for Senna ated except for one stem which may not control the top, whereas saw cuts with now being controlled. This indicates that have been suffi ciently covered. The fresh product contacting both top and bottom insuffi cient product can be applied to the leaves on this stem were showing signs of of the cut appeared to give better control. smaller surface area for control to be effec- herbicide activity. There were green stems A better method for stem injection of cam- tive. As picloram may take up to two sea- at the canopy of the plant showing severe phor laurel may prove to be the frilling sons to fully control some species, further tip burning, indicating ongoing herbicide technique with axe cuts close together and observation will take place. It should be activity. The smaller plant treated in this penetrating the cambium layer so that the noted that all plants cut at 20–50 mm were manner had defoliated and there was no product is in contact with both upper and controlled. The olive at Cessnock is now sign of any green stems. lower portions of the cut. One plant each controlled as at April 07. An inspection of both sites on 21/2/07 at Gosford 2/11/05 and Port Macquarie Foliar smear and stem injection shows a and 10/5/07 revealed all treated plants 27/9/06 were treated by scrape and paint great deal of promise on the Opuntia spe- (both methods) were dry and brittle, indi- method. The stem diameters were approx- cies and the results may lead to an inclu- cating a control. The plants were all loose imately 20 to 30 mm. The plant at Gosford sion on the label. The ease of control of in the soil and were easily puled up, in- remains controlled as at 12/12/06 and the small pear plants by foliar application will dicating that the root systems were also Port Macquarie plant had completely de- enhance the activities of all land manage- dead. There was no regrowth from any foliated on 29/11/06 and is dry and brittle ment groups. There is no longer a need for treated plant from either method. One at 14/3/07. Cut stump application was ef- these people to physically remove all plant area has been mulched and replanted; fective on this species. material to prevent dispersal by fragmen- the treated plants show no evidence of tation. Optunia species are best treated

Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 95 using the leaf smear method while the Ward B.G. and Henzell R.F. (2002). Con- scrape and paint and basal bark tech- trolling key environmental weed spe- niques for Madeira vine also successfully cies using non-spray techniques: guide- controlled a very high percentage of the lines for land managers and weed con- aerial tubers. Anecdotal evidence from trol contractors. The Horticulture and New Zealand indicates that good control Food Research Institute of NZ Ltd. of wild tobacco can also be achieved with ISBN 0478068328. basal bark applications as long as the bark Ward, B.G. and Henzell, R.F. (2004). Use of is green at application. A trial of this meth- herbicidal gels on woody weeds, DOC od on wild tobacco has been established at Science Internal Series 162, Department Pacifi c Palms on 8/3/07 on 50+ plants and of Conservation, Wellington, New Zea- will be monitored by Great Lakes Shire land. Council. The use of these techniques on Ward, B.G., Henzell R F. and Pritchard G H. any species should be trialled on small (2002). Eradication of grey willow (Salix areas before attempting a large scale op- cinerea) using herbicide gel technology. eration and should initially be restricted In Proceedings of the Thirteenth Aus- to green or thin barked species. Operators tralian Weeds Conference, eds H. Spaf- will also need to be more alert to weather ford Jacob, J. Dodd and J.HG. Moore, conditions, as these methods leave a larger pp. 488-91. Plant Protection Society of surface area of gel exposed and rainfall Western Australia, Perth. soon after application may reduce the ef- fectiveness of the product and/or increase * Vigilant is the Registered Trademark of the risk of off target damage. The Horticulture and Food Research Insti- For a list of species known to be sus- tute of New Zealand. ceptible to Vigilant, see Appendix A. As with all pesticides, extensive research has been carried out, the results of which im- pact directly on the wording of the label. In all cases therefore, read and follow the instructions on the label and you will en- sure quality results.

Acknowledgements The author would like to thank the fol- lowing people for their assistance in pro- viding trial sites and or monitoring of the results: • Barry Powells, Coffs Harbour City Council. • Barry Shepherd and Maria Edmonds, Cessnock City Council. • Colin Matthews, Bellinger Landcare Inc. • Eddie Lanting, Gosford City Council. • George Wisemantle and Rod Spicer, Greater Taree City Council. • Ian Turnbull, Bellingen Shire Council. • Mike Dodkin, Department of Conserva- tion and Climate Change NSW– Parks and Wildlife Division. • Paul O’Connor, Grant Taylor and Graeme Guy, Port Macquarie-Hastings Council. • Reece Luxton and Darrin Heron, Clarence Valley Council. • Terry Inkson, Great Lakes Shire Coun- cil. • Tony Cook, NSW Department of Pri- mary Industries – Weed Management Unit.

References McFadyen, R., (2003). In ‘Killing us soft- ly-Australia’s green stalkers’ A call to action on invasive plants, and a way forward. ed P. Martin. CRC for Austral- ian Weed Management. Available on http://www.weeds.crc.org.au/docu- ments/kus_part_one.pdf

96 Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 Appendix A. Species known to be susceptible to Vigilant Herbicide Gel at 1/5/07. Known susceptible species Application method Cut Basal Scrape Blaze Common name Botanical name Foliar stump bark paint inject African olive Olea africana X cut close to ground Agapanthus Agapanthus orientalis X drill tubers Arum lily Zantedeschia aethiopica XX Balloon vine Cardiospermum grandifl orum XX Bitou/boneseed Chrysanthemoides monilifera X Blackberry Rubus fruticosus X Broadleafed pepper Schinus terebinthifolia XX Camphor laurel Cinnamomum camphora XXX Canna lily Canna indica X Cassia Senna pendula var. glabrata X Castor oil plant Ricinus communis X Cat’s claw creeper Macfadyena unguis-cati X Chinee apple Ziziphus mauritiana X Chinese celtis Celtis sinensis X Coastal morning glory Ipomea cairica XX Cockspur Maclura cochinchinensis X Coral tree (smooth and rough bark) Erythrina spp. X X Cotoneaster Cotoneaster spp. X Formosan lily Lilium formosanum X Glory lily Gloriosa superba isolated plants only X Golden rain tree Koelreuteria paniculata X cut close to ground Gorse Ulex europaeus X Grafted apricot Unknown X Green cestrum Cestrum parqui X Groundsel bush Baccharis halimifolia X Hiptage Hiptage benghalensis X Honey locust Gleditsia triacanthos X Hybrid pine Pinus elliottii × carribea X Japanese honeysuckle Lonicera japonica X Kahili ginger Hedychium gardnerianum XX Lantana Lantana camara X Lemon tree Citrus limon X Madeira vine Anredera cordifolia XX Mesquite Prosopis spp. X Monsterio Monstera deliciosa X Moth vine Araujia sericifera X Mysore thorn Caesalpinia decapetala X Night scented cestrum Cestrum nocturnum X Ochna Ochna serrulata X Ornamental fruit tree Unknown X Paddy’s lucerne Sida spp. X Prickly pear Opuntia stricta XX Privet (broad and narrow leaf) Ligustrum spp. X Rubber vine Cryptostegia grandifl ora X Senna Senna fl oribunda XXX Siam weed Chromolaena odorata X Smooth tree pear Opuntia vulgaris XX Sweet briar Rosa rubignosa X Umbrella tree Scheffl era actinophylla X Wild tobacco tree Solanum XX Willow Salix babylonica X Yellow bells Tecoma stans X Yellow oleander Nerium oleander ssp. X

Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 97 territory legislation under a given control category. Empowering gardeners to make informed choices Each state/territory law has different control categories which indicate differ- when buying garden plants ent management requirements depending on the level of risk that the species poses Nicola Thomson and Andreas Glanznig, WWF-Australia, PO Box 528, of developing into a weed of primary in- Sydney, New South Wales 2007, Australia. dustry or the environment and the level of opportunity that exists to get it under control. These categories typically range between plants that are banned outright Summary This paper highlights the that can be legally sold in Australia many in the entire state to prevent them from risk posed by the garden plant invasion of which have yet to be screened by gov- establishing, to those that are already well- pathway, particularly the on-going sale of ernments to determine whether they are established in the state but are damaging around 40 per cent of naturalised garden likely to develop into weeds of agriculture enough to warrant legislative control to plants that have been declared noxious in or the environment. prevent its further spread (Australian one or more Australian states or territories. The source of this problem is a com- Weeds Committee 2006). This risk is compounded by an existing bination of the inadequacies of existing In some states/territories, the legisla- backlog of around 4,065 non-naturalised regulatory control measures and public tion includes weed categories that declare referenced weeds identifi ed by the Weeds information systems and the scale of the a species in a given region or local man- CRC that have already been imported into Australian garden plant retail industry. agement area of the state (Victoria, West- Australia for cultivation as garden plants The factors that pose a challenge to ad- ern Australia, South Australia, New South many of which have yet to be screened by dressing this problem will be considered Wales), while other State’s weed declara- governments to determine whether they including a heterogenous and dispersed tion categories apply throughout the state should declared as noxious weeds. The industry which is itself poorly informed regardless of the regional or local manage- causes of this problem include the limi- about risk posed by invasive species ment areas (Tasmania, ACT, Queensland, tation of existing government regulatory and the potential legal liability that this Northern Territory). The categories also systems and public information systems infers. vary in terms of management responsi- as well as the scale and increasingly com- It proposes a way forward for industry bility, the higher-risk categories typically plex structure of the Australian garden and government that, in addition to bet- place responsibility on the state/territory plant retail industry. ter-resourcing existing systems to remove departments to enforce, the medium-risk WWF-Australia proposes a two-tiered backlogs before they become an uncon- categories place responsibility on local policy response to reduce the weed spread trollable problem and centres around miti- or regional governments, some catego- risk associated with invasive garden gating the risk posed by the garden plant ries place responsibility on land-owners plants. High-risk invasive garden plants invasion pathway at the point of sale. to manage the weed on their land, other need to be identifi ed as matter of urgency categories place responsibility on traders and banned from sale before they natural- Background to remove them from trade. ise, while medium-risk invasive garden Weeds cost Australian agriculture $4 bil- Some 430 weeds are declared under the plants should be carefully managed – re- lion a year (Sinden et al. 2004), and degrade different pieces of state/territory legisla- tailers and consumers need to be alerted to tens of millions of hectares of Australia’s tion. However, due to the different catego- their invasive potential and how they can environment (CRC for Australian Weed ries, a declaration does not automatically responsibly care for them to reduce weed Management 2003). While there are many ban a plant from being traded, and places spread risk. This paper focuses on the role weeds, there are relatively few invasion no obligation on other states/territories to of a labelling standard and scheme in re- pathways; the garden plant invasion path- do so in their jurisdictions. As more and ducing the weed spread risk associated way has been identifi ed by the CSIRO, the more potentially invasive garden plants with medium-risk invasive plants. Weeds CRC, Meat and Livestock Aus- emerge, the ongoing spread of declared A national labelling scheme is proposed tralia, and WWF-Australia as Australia’s weeds between jurisdictions via the gar- for Australia that is regionally fl exible (a dominant weed spread pathway (Groves den plant trade will become an increas- plant may be high-risk in one State and et al. 2005, Glanznig 2005, Barker et al. ingly signifi cant invasion pathway. low-risk in another); mandatory (to create 2006, Randall unpublished). About 70 per This situation has arisen due to the gen- a level playing fi eld in a fragmented and cent of Australia’s agricultural and envi- erally narrow and reactive nature of State dispersed industry); and provides a range ronmental weeds are escaped invasive and Territory laws to prevent and control of benefi ts for the nursery industry at garden plants (Virtue et al. 2004). Studies weeds which exposes these jurisdictions minimal cost and disturbance. The scheme suggest that introduced garden plants will to the risk of continued and sustained im- would incur minimal costs to industry and comprise an even greater portion of the to- port of many new weeds (Glanznig 2005). protect it a potential future costly civil li- tal of future naturalised species (Groves et It is important to note at the outset of this ability risk. al. 2005). Figure 2 provides an illustration paper that the garden industry has gen- of the garden plant invasion pathway. erally operated within the bounds of the Introduction Agricultural and environmental weeds law. The importation of hundreds of in- This paper examines the issue of the gar- that are present in Australia are generally vasive garden plant species was entirely den plant invasion pathway in Australia identifi ed and regulated on the State/Ter- legal – as such, responsibility for the his- as a major source of agricultural and en- ritory level (with the exception of the 20 torical weeds problem squarely rests with vironmental weeds in Australia. In par- nationally-agreed Weeds of National Sig- governments. This policy failure is being ticular, it recommends policy measures to nifi cance and regional/local government progressively addressed – the introduc- tackle the risk posed by the on-going sale lists). Each State/Territory has its own list tion of ‘guilty until proven innocent’ im- of garden plants that have been declared of declared weeds which it has assessed port protocols at the borders of both Aus- by one or more state/territory as a noxious to determine its ‘weediness’ (a process tralia and Western Australia are two of the weed. This risk is compounded by a back- often referred to ‘weed risk assessment’) outstanding policy innovations that have log of potentially invasive garden plants and then listed under the relevant state/ occurred in the late 1990s.

98 Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 Issues Table 1. Naturalised invasive and potentially invasive garden plants and Many weedy garden plants are still their noxious status and availability for sale in 2002, both nationally and by advertised for sale jurisdiction (source: Groves et al. 2005). Due to the poor harmonisation between Declared Noxious in State and Territory declared weed lists, Naturalised jurisdiction and available for noxious weeds declared in one jurisdic- invasive garden sale in at least one other state tion can still be sold in another. Table 1 Jurisdiction plantsA Declared NoxiousB or territory jurisdiction presents the results of analysis conducted No. No. % No. % by Groves et al. (2005) that compared a list Australia 720 178 24.7 72 40.4 of invasive garden plants declared under State and Territory legislation with a pub- NSW 205 99 13.8 36 36.7 lished catalogue of garden plants avail- QLD 158 57 7.9 20 35.1 able for sale from nurseries in Australia SA 161 66 9.7 19 28.8 (Hibbert 2002). The catalogue lists about TAS 152 51 7.1 16 31.4 30,000 garden plant taxa (including varie- VIC 409 60 8.3 18 30.0 ties) derived from the catalogues of 280 WA 314 171C 23.8 69 40.4 individual nurseries and 31 seed suppliers spread across all states and mainland ter- ACT 104 23 3.2 6 26.1 ritories. It is important to note that Hibbert NT 63 42 5.8 13 31.0 (2002) does not include many of the other Notes retail sources for garden plants illustrated A Number naturalised includes all listed taxa (genera, species, sub-species) that are in Figure 1, including e-commerce and recorded as naturalised in jurisdiction (Randall and Kessal 2004). non-specialised retail outlets such as dis- B Declared noxious refers to taxa that are declared noxious under relevant State/Territory count department stores. It also important government legislation in respective jurisdiction. The Australian total refers to the number to note that this audit was based on 2002 of taxa that are declared noxious in at least one State or Territory jurisdiction. Percentage data on plant availability. There is clearly is portion of total naturalised invasive plants that are declared noxious in respective an urgent need to conduct a national audit jurisdiction (AWC January 2004). The NSW fi gure includes regional declarations, and of garden plants currently in trade. these listed species may be available for sale in non-control regions. Nationally over 40% of the invasive gar- C This includes those species that are both declared noxious or unassigned. Unassigned den plants that are declared noxious in one species are subject to a weed risk assessment if importation into the State is sought. or more jurisdictions were still available Sources: Randall and Kessal (2004), Australian Weeds Committee (2004), Hibbert (2002) for sale in 2002 (Table 1), with all states and cited in Groves, Boden and Lonsdale (2005). territories signifi cantly exposed to the sale of declared plants in other jurisdictions. The most exposed states were Western Australia with 40 per cent of the plants de- will become future weeds, the probability identifi ed by the report were still avail- clared noxious in WA being available for that they will is signifi cantly higher than able for sale in 2004 (Coutts-Smith et al. sale in at least one other Australian state or for species with no overseas weed history. 2006). territory, NSW with 37% and Queensland This list of species indicates that there is a These fi gures give an indication of sheer with 35%. large pool of plants already in Australia scale of the threat posed by the garden This situation creates a signifi cant risk imported as garden plants that have not plant trade invasion pathway and high- of ‘leakage’ from one jurisdiction to the yet established themselves and that pose a light the need for better control measures State where the weed is declared. This signifi cant risk of becoming weeds of agri- as these plants inevitability spread from leakage could take the form of movement culture and/or the environment. state to state/territory. As governments of a pot plant when people move inter- Two reports recently released by the risk assess more of these plants the num- State or purchase seeds over the internet. CRC for Weed Management have further bers of declared weeds should theoreti- highlighted the risk posed by invasive gar- cally swell from the hundreds to the thou- Backlog of unscreened potentially den plant species that are still advertised sands. invasive garden plants for sale but have yet to naturalise: Data complied by the Weeds CRC indi- • A report commissioned by Meat and A fragmented and dispersed industry cates that the lists of noxious weeds de- Livestock Australia identifi ed 281 intro- The risk posed by potentially invasive gar- clared in each state/territory should grow duced garden plants and 800 lower pri- den plants is worsened by the dispersed substantially in the future. The CRC has ority species that are likely to develop and fragmented nature of the plant retail estimated that there are about 4065 plant into signifi cant grazing weeds should industry – plants can be bought from a species that have already been imported they naturalise in Australia. Of the 281 wide range of outlets including over the into Australia for cultivation as garden potential high risk grazing weeds, over counter at hardware stores, supermarkets plants which are referenced weeds yet to two thirds (70%) were still available for and large discount store chains, as well naturalise in the Australian environment1 sale from Australian nurseries in 2004 as via mail order and over the Internet. (Randall unpublished). While it is impor- (Barker et al. 2006). The ‘Australian Garden Market Monitor tant to note that not all of these species • A report commissioned by the NSW 2005’ shows that for retail channels, retail Department of Environment and Con- nurseries make up 26.3% of the market servation found that weeds threaten share, while the combined market share Footnote 419 NSW plant and animal species of hardware, supermarkets and discount 1 A referenced weed is a plant species that listed as threatened under the State’s department stores is 40.3% (Figure 1). has been identifi ed as a weed of any type legislation. The report estimated that The peak body, Nursery and Garden overseas which is listed in the Australian around 93% of these threatened species Industry Australia (NGIA), covers rough- Exotic Species Database which is main- were directly threatened by weeds that ly one third of the production businesses tained by the West Australia Department are escaped garden plants. The report recorded by the Australian Bureau of Sta- of Agriculture. found that 44% of the 127 weed species tistics in 2001 (Martin et al. 2005). This is

Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 99 an important consideration when looking Retail nursery at tackling the issue of invasive garden Garden supplies plants, as any effort to infl uence grow- Hardware ing and purchasing practices that is run through the peak body will only target a Discount dept. store portion of the sector. Supermarkets Potential civil liability for plant traders Wholesale direct In recent years, many industries have en- Mail order/ E-comm countered civil liabilities that had long been dormant. Suppliers of goods and Markets services that have or are facing civil ac- tions include: the tobacco industry; ‘fast- Figure 1. Summary of the Garden Retail Market Share, year ending June food’ outlets; liquor companies; asbestos 2006 (Freshlogic 2006). suppliers; and local governments. The common feature in all such claims is the argument that the industry was aware of the harm-causing potential of its product, a reliable source for plant identifi cation that ‘purchase transactions’ are an impor- and chose not to take the responsibility for (Savage and Rondeau 1992). tant intervention point in the garden plant either warning of the risks and required invasion pathway (refer Figure 1) because controls, or to directly control that risk Poor information systems this intervention point can be used to: (Martin et al. 2005). The Weeds CRC attempted to measure • Encourage consumer purchasing pref- A study by the University of New Eng- public awareness about weeds through a erences toward plants…that are benign land’s Australian Centre for Agriculture national survey conducted in 2003. The or positive in their impacts, and Law identifi ed that the practice of survey revealed that the urban public is • Stimulate action to reduce the risk of knowingly selling invasive plants with no still largely uninformed and unaware of future invasions, such as retro-fi tting consumer disclosure presents a civil liabil- invasive plants as an issue (Martin 2006). gardens…, and ity risk for the garden industry (Martin Current information sources include • Provide information to enable purchas- et al. 2005). The report demonstrated the state lists of declared weeds which plant ers to implement actions that prevent grounds for future claims against garden retailers within the state are legally obliged the spread of any invasive species they plant growers and sellers to pay for the to comply with. Nurseries may also have have purchased (Martin et al. 2005, ‘clean up’ costs of invasive plants that they to comply with weed lists generated by p.5). have sold. Under Section 52 of the Trade their local government which may or may A rare example of this approach in Austral- Practices Act (1974), a failure to disclose a not have a legal status. In the case of inva- ia is the pioneering efforts of a large Aus- relevant fact of which a supplier is aware sive plants that have not been listed, vari- tralian national mail-order garden plant could be interpreted as a misleading of- ous government and non-governmental and seed retailer: ‘Diggers’. ‘Diggers’ has fence. The report notes that, from a legal bodies have produced communication voluntarily introduced regionally-specifi c perspective, adopting a preventative ap- materials such as websites, booklets and invasiveness warnings in their catalogues. proach may be in the best interests of the posters aimed at consumers to encourage Along with the standard plant characteris- suppliers of potentially invasive plant spe- them to avoid selecting plants that can be tics and care information, each catalogue cies. legally traded but are invasive or poten- entry indicates if the plant has been listed tially invasive in the their areas, notable as a weed in State legislation and fl ags Dependence on industry knowledge examples include the Nursery and Garden that it is unavailable for sale these juris- As most invasive garden plants tend not Industry Association’s ‘Grow me Instead’ dictions. ‘Diggers’ also goes one step fur- to be banned outright, so responsible use campaign which presents safe alternatives ther by warning consumers about plants of them depends on effective communi- to around 17 invasive garden plants that that have not been listed as a weed by any cations. The current system relies on at- are in trade in NSW and ACT. states/territories but are potentially inva- tending staff at plant retail outlets to be The low level of public awareness com- sive in certain areas. These plants are indi- informed and conscientious in warning bined with the widespread availability of cated in their catalogues with an icon that customers about plants that are declared declared invasive plants through the gar- denotes plants that can be invasive and as invasive in other jurisdictions or plants den trade illustrates the ineffectiveness of the climate zones where this risk exists. that are not prohibited but are known to current consumer awareness-raising ef- ‘Diggers’ independently identifi es these be invasive in certain local environments. forts. Current systems rely on nurseries weedy species through their own research Qualifi cations are not required for a per- taking the initiative to supply the infor- including international weeds databases. son to work in the Australian nursery in- mation materials in their outlets and refer In the words of a ‘Diggers’ representative dustry (Adler et al. 2000). to them at the point of sale, or they rely ‘This information is not required by law, A casual survey by the lead author of on consumers taking an active interest in but we feel that by providing these notes four Sydney garden centres in January this the issue, seeking out the information and we can help gardeners to make informed year revealed that even well-informed, applying it when choosing what plant to decisions about what to plant (or what not conscientious garden centre staff only pro- buy. to plant)’ (Sansom 2005). vided a warning in relation to a maximum A major fault of the existing system is of three plants from a list of over 40 inva- that the information is not well coupled Consumer demand for warnings at point sive plant species that were requested and, to the activity it is trying to infl uence, i.e. of sale in some cases purchased. This indication consumer purchasing behaviour. For con- The increase in awareness about the large of nursery staff ability to identify plants sumers to be able to make wise choices impacts of weeds and where they are com- and warn consumers of their properties about which plants to choose and where to ing from over the next several years will was refl ected in a more comprehensive plant them they need to be provided with lead to strong community expectations American study that conclusively found information about the plants properties at for governments and the garden indus- that American plant nurseries were not the point of sale. Martin et al. (2005) note try to implement effective solutions to the

100 Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 invasive plant problem. Urban Australian in the country that pose a risk of devel- • An awareness-raising campaign that have low awareness about weed issues. oping into weeds. It is also clear that the targets garden plant consumers and However, market research shows they declaration and control of weeds requires garden industry operators to encour- have strong latent demand to adopt be- better national coordination, particularly age both parties to proactively prefer- haviours that reduce weed spread risk if it due to the national nature of the garden ence low-risk weeds. is easy and convenient (Martin 2006). plant industry and the lack of existing These on-going ‘soft’ measures need to be This latent demand was affi rmed by a coordination between states/territories accompanied by a focused effort to pro- recent national galaxy poll commissioned on weed identifi cation and control. Non- vide the key information at the point-of- by WWF Australia that surveying 1100 naturalised plants that are determined to sale so that consumers start to factor the Australians by telephone. The poll dem- be high-risk through weed risk assessment invasive plant threat into their purchasing onstrated that a very low level of Austral- should be nationally banned as a matter decisions. Hence WWF-Australia also rec- ians have knowingly brought invasive of urgency. ommends that: plants for their gardens (5%). It also dem- • The Nursery and Garden Industry As- onstrated very strong consumer demand Infl uence demand for medium-risk sociation (NGIA) in conjunction with for warnings on plant labels: plants the Australian Weeds Committee (un- The focus of this paper is on fi nding a so- der the Natural Resource Management ‘Almost all Australians (96% of re- lution to those plants that are identifi ed Ministerial Committee) introduce a na- spondents) believe that plants which can as ‘medium risk’. For the purposes of this tional mandatory plant labelling stand- become harmful agricultural or environ- paper, medium-risk plants are those that ard and scheme. Such a scheme would mental weeds should be clearly labelled have been listed in state/territory legisla- enable garden plant retailers and con- with a warning. This belief is consist- tion as noxious weeds but are not banned sumers to identify at the point of sale ently held by all types of Australians, from trade2. species that have been determined as regardless of factors such as gender, age, Rather than being banned outright, medium-risk on a national level and in or [where they live], these species need to be contained to areas what areas these species pose a risk of and uses that will prevent them from be- becoming invasive. …Australians believe the warning label coming a problem. This containment will should contain plant care information to require a range of measures that together A labelling standard and scheme reduce the risk of weed spread (93%), the target each stage of the garden plant inva- While the NGIA is the appropriate peak parts of Australia where the plant can sion pathway (Figure 2). As this paper has body to lead the development of a label- become a weed (89%) and how to get demonstrated, a critical intervention point ling scheme, responsible plant labelling more information about which plants are in the garden plant invasion pathway is at should not be the sole responsibility of the invasive (88%)’ (Galaxy Research 2007, the point of sale. Consumers have a right NGIA, particularly as its membership only p. 4). to know if they are buying a potentially covers a small proportion of the plant re- invasive plant. They also need to know tail industry. The plant labelling standard Virtually all garden plants are sold with how to reduce their weed spread risk by would need to apply to the whole indus- a label which varies from basic and often knowing if the plant is suitable for the pro- try; hence, once it had been developed it inadequate or incorrect descriptive infor- posed location, how to maintain it, and would need to be established as a national mation to elaborate marketing devices. how to dispose of green waste responsi- mandatory code possibly under the Trade Plant labels are unusual in the commercial bly. Furthermore, plant sale staff require Practices Act 1974. world in not addressing the range of com- better clarity about which plants to warn Models of best practice in labelling have munity expectations that apply in most consumers about and what to advise - par- begun to emerge in Australia and overseas. other consumer product categories such as ticularly as more invasive plants are iden- The example of industry-leadership by food, clothing, and most household prod- tifi ed and as the garden plant retail sector ‘Diggers’, an Australian national mail-or- ucts. As a Senior Horticultural Botanist at becomes more diverse. der retailer, was profi led in this paper. The the Melbourne Royal Botanic Gardens re- WWF recommends a number of meas- United Kingdom’s peak horticultural in- cently asserted; ures to bring industry practice up to speed, dustry body adopted a national plant retail these are: labelling code and a specifi c labelling code ‘consumers expect to be provided with • An industry code of practice that sets for hazardous plants in 1994 (HTA 2000). accurate labelling…we have every right a standard of responsible conduct to Based on these existing models and on the to expect that a plant label faithfully in- ensure compliance with legislation experiences of other labelling schemes in dicates what is in the pot…Accuracy and and to prevent the spread of invasive Australia, WWF has identifi ed a number of truthfulness in labelling may not always garden plants into the environment. In- key characteristics that the scheme would suit the marketing objectives of plant ternational models of best practice are need to adopt. These include: producers, but consumers have the right available including the UK’s horticul- to know if the plant they are purchas- tural code titled ‘Helping to prevent the Nationally structured but regionally ing is an environmental weed’ (Spencer spread of invasive non-native species’ flexible The labelling scheme would 2007, p. 1). (UK DEFRA 2005). need to be nationally coordinated so as to overcome the lack of inter-jurisdictional A way forward – policy solutions coordination that has contributed to the Stop the supply of high-risk plants Footnote weed problem that exists today. However, It is clear that there is an urgent need for 2 By this defi nition, in any given state or the label should be designed in a way to states and territories to tackle the reservoir territory, a medium-risk plant would be a show regional differences in invasiveness, of potential future Australian weeds that plant that has been declared as a noxious for example one option would be a map have been identifi ed by the Weeds CRC weed in the jurisdiction under a legal cat- of Australia showing suitable growing before they ‘jump the garden fence’ and egory that does not ban trade, and/or, a area where the plant is not known to be move beyond affordable control. Better re- plant that has been declared as noxious invasive in one colour, with areas where it sourcing of weed risk assessment systems in another state/territory (hence can be is invasive highlighted in another colour. will enable governments to process the traded but poses a risk of ‘leaking’ into the This would enable consumers and retail- backlog of non-naturalised garden plants area where it is banned). ers to quickly identify which medium-risk

Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 101 plants are appropriate for which areas. It to homes installing more effi cient lights References would also enable a simple system where or showerheads to save energy and wa- Adler, M. (2000). An analysis of hazardous the same symbol and warning would ap- ter respectively. This second phase would plants in the Australian nursery indus- ply to a given plant regardless of where it present an obvious opportunity for the try. Horticultural Research and devel- is being sold in Australia. Beyond specify- nursery industry to develop a new market opment Corporation, NSW. ing the presentation of the warning, the for low-risk plants. Australian Institute of Horticulture (2006). standard would not need to place any re- In terms of costs to industry, implemen- Weed busting. Weed and invasive plant strictions on creativity or innovation in the tation of the scheme would be timed to species policy 2006. Australian Institute label design. work with the normal replacement cycle of Horticulture, Bendigo. p. 13. for labels allowing for existing stocks to Australian Weeds Committee (2004). Nox- Mandatory There is strong evidence to be used and then replaced, with an appro- ious weed list for Australian states and suggest that a labelling scheme that seeks priate sunset clause. This approach would territories, Version 10.00. Web: www. to enable consumers to compare between keep transition costs to a minimum. A weeds.org.au/docs/weednet6.pdf. products needs to be mandatory. In the mandatory labelling scheme would also Australian Weeds Committee (2006). Sum- case of the retail grocery industry’s code of enable the garden industry to move to mary of State and Territory noxious conduct and Australia’s water effi ciency contain future civil liability risk which weeds legislation, version 8.0. http:// labelling code, both were scheduled to presents the risk of massive costs to the www.weeds.gov.au. be converted from voluntary to manda- industry in future years. Barker, J., Randall, R. and Grice, T. (CRC tory due to the failure of an initial volun- for Australian Weed Management) tary approach (Buck and Associates 2003, Conclusion (2006). Weeds of the future? Threats to Wilkenfi eld and Associates 2003, Camp- This paper demonstrates the risk to Aus- Australia’s grazing industries by gar- bell 2006). A University of New England tralia’s environment and agricultural in- den plants. Meat and Livestock Aus- report, ‘Costs and Benefi ts of a Proposed dustry posed by invasive plants available tralia, Sydney. Mandatory Invasive Species Labelling through the Australian garden plant trade. CRC for Australian Weed Management Scheme’ (Martin et al. 2005), looked at the The source of the problem is partly govern- (2003). Killing us softly – Australia’s issues associated with a national label- ment weed regulation systems which have green stalkers: a call to action on in- ling scheme for invasive garden plants. It been slow to catch up with the evolving vasive plant, and a way forward. CRC concluded that unless such a scheme was scale and complexity of the garden plant for Australian Weed Management, Ad- mandatory it was unlikely to be univer- invasion pathway, and partly the industry elaide. sally adopted which would result in an itself which has also not been proactive in Freshlogic (2006). Australian garden mar- unfair fi nancial load being borne by the tackling the threat posed by its products. ket monitor: for year ending 30 June most responsible members of the industry The nursery industry now faces a poten- 2006. Report commissioned by the and ‘free-riding’ by others. tial civil liability for failing to disclose a Nursery and Garden Industry Associa- plant’s invasive properties. tion Australia. Supported by awareness-raising As rec- The problem is made more complicated George Wilkenfeld and Associates Pty Ltd ommended by Martin et. al. (2005) the la- by the fragmented and dispersed nature et al. (2003). A mandatory water effi - belling code would need to be promoted of Australia’s garden industry which in- ciency labelling scheme for Australia. to retailers and the public and supported cludes internet sales, supermarkets and Final report prepared for Environment by education materials such as: a user- discount department stores. Garden plant Australia. friendly booklet and posters and/or a retailers are not required to have any Glanznig, A. (2006). Weed proofi ng Aus- website. The code could also be factored plant-related qualifi cations or knowledge. tralia: a way forward on invasive gar- into existing awareness-raising and edu- A recent national poll demonstrated strong den plants. Proceedings of Weeds in cational efforts such as horticultural train- consumer demand for plant labelling to the Media. CRC for Australian Weed ing programmes and nursery industry ac- enable consumers to choose plants that are Management, Adelaide. creditation schemes. not invasive. Glanznig, A. (2005). Making State weed This paper recommends that govern- laws work, WWF-Australia Issues Pa- Complementary to industry interests ments prioritise processing the backlog of per. WWF-Australia, Sydney. The labelling scheme would reinforce thousands of potentially invasive garden Groves, R., Boden, R. and Lonsdale, M. the direction the nursery industry has al- plants that can be legally sold in Australia (2005). Jumping the garden fence: in- ready taken in with its ‘Grow me Instead’ and before these plants develop into Aus- vasive garden plants in Australia and campaign which encourages consumers tralia’s next generation of agricultural their environmental and agricultural to preference low-risk plants. A labelling and environmental weeds. Plants identi- impacts. CSIRO report for WWF-Aus- scheme could incorporate and support the fi ed as high-risk need to be immediately tralia. WWF Australia, Sydney. ‘Grow me Instead’ campaign and brand, banned from sale. Plant determined to Hibbert, M (2002). Aussie plant finder for example by indicating on labels if a be medium-risk (hence allowable for sale 2002/2003. (Florilegium, Glebe, NSW). plant has been identifi ed as a recommend- in certain parts of Australia) need to be HTA (2000). United Kingdom Horticul- ed non-invasive alternative plant. brought under more thorough control by tural Trade Association code of prac- A likely result of a national labelling government and industry so that they are tice for recommended retail practice scheme would be increasing consumer confi ned to areas and uses in which they relating to the labelling and display of recognition of invasive plants that have will not develop into weeds. A key tool in potentially harmful plants, 2000 review already made it into their gardens. En- this effort will be a national mandatory la- and update. UK Horticultural Trade couraging home owners to participate in belling standard supported by an industry Association. schemes that audit their gardens and help code of practice. The code would result in Martin, P. (2006). Measuring the level of them replace high-risk plants with those consumers being consistently informed of public weeds awareness. Proceedings that are environmentally safe is the logi- a plant’s properties in a way that enables 15th Australian Weeds Conference, eds cal next step in tackling the reservoir of them to choose the right plant for the right C. Preston, J. Watts and N. Crossman medium-risk invasive plants that have yet purpose. pp. 111-13. (Weed Management Society to naturalise in Australia (refer Figure 2). of South Australia Inc). This ‘retro-fi tting’ of gardens is analogous

102 Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 Pathway Intervention Measures The Garden Plant Invasion Pathway

Comprehensive National Import of new overseas BorderPermitted Plant List plant species

Existing introduced and native National Noxious Weed List (High-risk plants prohibited for plant species under cultivation sale)

Nursery Growers

National Mandatory National Invasive Weed Plant Awareness Nurseries Supermarkets/ Labelling Campaign Landscapers Scheme Hardware chains

Consumers Developers

Retrofitting Garden Schemes Gardens

Early Warning and Rapid Emerging invasive plants in areas surrounding urban Response System gardens

Weed Bushcare/ Widespread invasive plants in areas surrounding urban gardens containment Coastcare and control restoration by agricultural industry and Serious weeds of Serious weeds of the governments agriculture environment

Figure 2. Key stages of the garden plant invasion pathway and intervention measures (Glanznig 2006).

Randall, R. and Kessal, O. (2004). National about what’s in a pot? Global Gar- list of invasive and potentially invasive den: Internet Gardening Magazine. garden plants. WWF Australia, Syd- www.global-garden.com.au/burnley/ ney. june00dte.htm. Sansom, T. (2005). The weed debate: gar- Virtue, J., Bennett, S. and Randall, R. deners rights. Diggers Information (2004). Plant introductions in Australia: Note. (http://www.diggers.com.au/). How can we resolve ‘weedy’ confl icts Sinden, J., Randall, J., Hester, S., Odom, D., of interest? Proceedings of 14th Austral- Kalisch, C., Rosemary, J. and Cacho, O. ian Weeds Conference, ed B.M. Sindel (2004). The economic impact of weeds and S.B. Johnson, pp. 42-8. (Weed Man- in Australia. CRC for Australian Weed agement Society of South Australia, Management, Adelaide. Adelaide). Spenser, R. (2007). Plant labelling: how well do plant labels convey information

Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 103 have indicated utilising light fi lters can al- ter weed emergence. Light fi lters are used The use of radiation wavelengths to manage weeds extensively in the horticultural industry to promote or delay fl owering (Pearson and Graham Brodie and Ken Young, The University of Melbourne, Dookie, Khattak 2006). Victoria 3647, Australia. Heat generating photons (microwaves) Over the past 40 years, microwave heat- ing has frequently been proposed as an al- Introduction ternative method of controlling soil-borne Light exists as quantum states of electro- This paper summarise recent research in pests such as weed seeds, insects, nema- magnetic fi elds (Dirac 1927), simultane- to the effect of photons that either regulate todes and pathogens (Nelson 1996). In ously exhibiting both particle and wave plant development (i.e. light wavelengths) particular, treatment of soils as a method behaviour. These quantum states, or pho- or provide heating of organic tissues (i.e. of weed control (i.e. killing the weed seed tons, transport energy and information microwaves and solarisation). bank) has been proposed for some time over vast distances. Information is modu- (Nelson 2003). lated onto light by variations in intensity Plant regulating photons Barker and Craker (1991) demonstrated (number of photons) and energy. The en- Photons that are visible to humans (wave- that treatment of soil, containing ‘Ogle’ ergy associated with a photon is frequency lengths between 400 nm and 900 nm) affect Oats (Avena sativa) and an undefined dependent, such that E = hυ; where h is plants in various ways. The main photons number of naturalised weed seeds, in a Planck’s constant (6.626 × 10−34 J s) (Ein- that affect plants are the red and blue parts microwave oven prevented seed germi- stein 1951). It is these differences in photon of the spectrum. Plants have three types nation when the soil temperature rose energy that differentiate the behaviour of of receptors that respond to blue light: a) above 80°C. Other experiments (Brodie photons. crytpochromes which regulate hypocotyl et al. 2007a) have demonstrated that mi- Some photons provide energy that expansion; b) phototropin which regulates crowave treatment of soil signifi cantly re- stimulates biological activity, while others phototropism and c) zeaxanthin which re- duced wheat seed germination when the in different parts of the spectrum can se- gualtes stomatal opening (Campbell and soil temperature rose above 65°C. These verely damage organic tissues. Examples Reece 2006). In the red wavelengths (700– trials also demonstrated that microwaves that highlight the favourable interactions 900 nm), the plant responses are governed interact with the soil rather than with the of photons with organic molecules in liv- by phytochromes which exist in several seeds and that heat must transfer from the ing things include: states – an active (Pfr) and an inactive state soil to the seeds if germination is to be sup-

1. The absorption of photon energy by (Pr). The phytochrome converts between pressed. chlorophyll to start the complex se- the two states depending on the type of Horn antennas can project microwave

quence of reactions known as photo- light it receives. If Pfr receives far red light energy into open space (Connor 1989).

synthesis (Ladiges et al. 2001), it will convert to Pr. If Pr receives red light Heating equations, developed in earlier 2. The absorption of blue light to modulate it will convert to Pfr. Pfr is also unstable studies (Brodie 2006, Brodie 2007a, Brodie

phototropism, slowing of hypocotyl ex- and converts slowly back to Pr in darkness 2007b), can forecast the heating patterns in pansion and stomatal opening (Ladiges (Campbell and Reece 2006). the soil. The temperature distribution can et al. 2001), The requirement for light after seed be described by Equation 1. 3. The absorption of red and far-red pho- burial has been observed in many species Figure 1 shows these patterns. The hot- tons by phytochrome molecules in (Wesson and Wareing 1968, Woolley and test place in the heating pattern was along plants, that regulate plant development Stoller 1978, Froud-Williams et al. 1984, the centre line of the antenna and between including seed germination, fl owering, Scopel et al. 1991, Baskin and Baskin 1992, 2 cm and 5 cm below the surface, depend- branching or apical dominance, (Oh et Corbineau et al. 1992, Milberg and Anders- ing on the soil type and moisture content. al. 2006), and son 1997, Botto et al. 1998, López-Granados Soil type, which depends on the distri- 4. The transformation of retinal molecules and Lutman 1998, Milberg and Andersson bution of particle sizes in the soil mixture,

(C20H28O), found in the light receptor 1998), the seasonal loss of a light require- determines how easily a soil will interact cells of the eye, from the hooked 11-cis ment has only been reported in a few spe- with microwave photons. Table 1 shows isomer to the straight all-trans isomer. cies such as in Bromus rubens L. (Corbineau that dry clay heats more rapidly than dry This transformation disentangling the et al. 1992), Datura ferox L. (Botto et al. 1998) loam, which in turn heats more rapidly retinal molecule from its encapsulating and Lolium rigidum (Steadman 2003). than dry sand. opsin protein and triggers a synaptic Is it possible to use this switching mech- Photons in the microwave energy range signal along the optic nerves to the anism to trigger or stop germination? Pre- interact strongly with polar molecules, brain, which results in vision (Chang vious researchers have tried to use night such as water (Metaxas and Meredith 1998). tillage as a mechanism to stop germina- 1983); therefore the most likely explana- Examples that highlight the damaging ef- tion. This has been successful in some ex- tion for this differentiation in soil heating fect of photons on organic molecules in- periments but not in others. In an experi- between clay, loam and sand probably clude: ment at Dookie (James 2000) showed that involves bound water on the soil par- 1. The ionisation of organic molecules by night tillage did not decrease weed ger- ticles. Usually bound water layers are a very high energy photons such as X- mination. More recently, Watts (2006) and few molecules thick, even when the soil rays and gamma radiation, which leads preliminary results from 2007 (Johansen) is ‘dry’ (Tikhonov 1997). Microwaves to radiation poisoning and tissue death; and 2. The over-heating of organic tissue by Equation 1. − 2 low energy photons, such as micro- 2 ⎡ z ⎤ sπfε ε"τ γα 2 − α ⎛ h ⎞ γ ⎛ π ⎞ waves and radio waves, which leads T = o ⋅ E 2 ⋅ ()e 4 t −1 × ⎢e 2 z + ⎜ + 2α ⎟z ⋅ e 4 t ⎥ × Cos⎜ x⎟ to permanent protein denaturation and 4kα 2 o ⎢ ⎝ k ⎠ ⎥ ⎝ a ⎠ tissue death. ⎣ ⎦

104 Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 interact with this bound water to create 0 heat. Because clay has the smallest par- Distance below soil surface (mm) Temperature °C 15 ticles (followed by loam and then sand), 80 - 90 the surface area per unit volume for clay 30 70 - 80 is very high; therefore there will be more 60 - 70 bound water in a given volume of dry clay, 45 50 - 60 40 - 50 resulting in faster microwave heating. 60 Seed survival depends on the soil tem- 30 - 40 20 - 30 perature (Nelson 1996), seed size and 75 whether the seeds have imbibed water or 90 not (Bebawi et al. 2007, Brodie et al. 2007c). −50 75 −75 −25 0 25 50 −50 75 −75 −25 0 25 Figure 2 shows that wheat seeds, with an 100 50 average mass of 41.7 mg each, are more susceptible to microwave treatment than Distance from centre line of horn (mm) Distance from centre line of horn (mm) wild oats seeds, with an average mass of Figure 1. Comparison of expected heating patterns (left) with measured 7.2 mg. Wild oats are more susceptible pattern (right) in clay/loam soil (Source: Brodie et al. 2007b) than ryegrass seeds, which have an aver- age mass of only 2.1 mg each. Figure 3 shows the temperature re- 100 Survival Data for Dry Ryegrass Dose Response Model for Dry Ryegrass sponse curve for ryegrass seeds in sandy 90 Survival Data for Dry Wild Oats soil. Clearly, dry seeds (dormant seeds in Dose Response Model for Dry Wild Oats 80 dry soil) are much less susceptible to mi- Survival Data for Dry Wheat crowave induced damage than imbibed 70 Dose Response Model for Dry Wheat ⎛ ⎞ 85.7 seeds. 60 ⎜ ⎟ = Survival % Dry Ryegrass 15 ⎝ ⎠ ⎛ T ⎞ Microwave treatment will not be as r2 = 0.75 1+ ⎜ ⎟ 50 ⎝ 77 ⎠ cheap as chemical treatments; however ⎛ ⎞ 100 .0 40 Survival % ⎜ Dry Wheat ⎟ = ⎛ ⎞ 100 .0 its mechanism for killing weeds and their ⎝ ⎠ ⎛ T ⎞15 Survival % ⎜Wild Oats ⎟ = 2 1 + ⎜ ⎟ ⎝ ⎠ 15 r = 0.90 2 ⎛ T ⎞ seeds is different to chemical treatments. 30 ⎝ 64 ⎠ r = 0.80 1 + ⎜ ⎟ ⎝ 69 ⎠

This can deal with herbicide resistant indi- Germination percentage 20 viduals in the weed population; therefore microwave treatments should be consid- 10 ered as part of an integrated weed man- 0 agement plan. 0 20 40 60 80 100 120 140 Soil temperature (°C) References Figure 2. Dose response curve for soil temperature versus seed survival Barker, A.V. and Craker, L.E. (1991). Inhi- bition of weed seed germination by mi- percentage for wheat, wild oats and perennial ryegrass seeds in dry soil. crowaves. Agronomy Journal 83, 302-5. Baskin, J.M. and Baskin, C.C. (1992). Role 100

of temperature and light in the ger- 90 Survival Data for Dry Soil mination ecology of buried seeds of Dose Response Model for Dry Soil 80 Survival Data for Wet Soil weedy species of disturbed forests. I. Dose Response Model for Wet Soil Lobelia infl ata. Canadian Journal of Botany 70 ⎛ ⎞ 85.7 70, 589-92. 60 Survival % ⎜in Dry Soil ⎟ = ⎝ ⎠ ⎛ T ⎞15 Bebawi, F.F., Cooper, A.P., Brodie, G.I., r2 = 0.75 1+ ⎜ ⎟ ⎝ 77 ⎠ Madigan, B.A., Vitelli, J.S., Worsley, 50 K.J. and Davis, K.M. (2007). Effect of 40 microwave radiation on seed mortality 30

of rubber vine (Cryptostegia grandifl ora Germination percentage 20 ⎛ ⎞ 85.7 R.Br.), parthenium (Parthenium hystero- Survival % ⎜in Wet Soil ⎟ = ⎝ ⎠ ⎛ T ⎞15 r2 = 0.70 1+ ⎜ ⎟ phorous L.) and bellyache bush (Jatropha 10 ⎝ 54 ⎠ gossypiifolia L.). Plant Protection Quar- 0 terly, (In press). 0 20 40 60 80 100 120 140 Botto, J.F., Sánchez, R.A. and Casal, J.J. Soil temperature (°C) (1998). Burial conditions affect light re- sponse of Datura ferox seeds. Seed Sci- Figure 3. Dose response curve for soil temperature versus seed survival ence Research 8, 423-9. percentage for perennial ryegrass (Lolium perenne) seeds (Source: Brodie et al. 2007c)

Table 1. Soil heating data for two levels of microwave power. Soil type Average moisture content of air dry Average heating rate for air dry Average heating rate for air dry soil samples soil samples at 2 kW of microwave soil samples at 4 kW of microwave (% of dry soil mass) power (°C per second) power (°C per second) Clay 6.92 a 0.66 a 1.33 d Loam 3.08 b 0.53 b 0.80 e Fine sand 0.38 c 0.30 c 0.68 a LSD (P <0.05) 1.12 0.09 Note: Means with different superscripts are signifi cantly different to each other (Source: Brodie et al. 2007b).

Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 105 Brodie, G. (2006). Modelling microwave annual weed species. Canadian Journal heating in blocks, cylinders and spheres, of Botany 75, 1998-2004. ASABE Paper No. 063006, St. Joseph, Milberg, P. and Andersson, L. (1998). Does Mich.: ASABE. cold stratifi cation level out differences Brodie, G. (2007a). Simultaneous heat and in seed germinability between popula- moisture diffusion during microwave tions? Plant Ecology 134, 225-34. heating of moist wood. Applied Engi- Nelson, S.O. (1996). A review and assess- neering in Agriculture 23 (2), 179-87. ment of microwave energy for soil Brodie, G. (2007b). Microwave treatment treatment to control pests.Transactions accelerates solar timber drying. Trans- of the American Society of Agricultural En- actions of the American Society of Agri- gineers 39 (1), 281-9. cultural and Biological Engineers 50 (2), Nelson, S.O. (2003). Microwave and radio- 389-96. frequency power applications in agri- Brodie, G., Botta, C. and Woodworth, J. culture, paper presented to Third World (2007a). Preliminary investigation into Congress on Microwave and Radio Fre- microwave soil pasteurization using quency Applications, Sydney. wheat as a test species. Plant Protection Oh, E., Yamaguchi, S., Kamiya, Y., Bae, Quarterly 22, 72-5 G., Chung, W. and Choi, G. (2006). Brodie, G., Hamilton, S. and Woodworth, Light activates the degradation of PIL5 J. (2007b). An assessment of microwave protein to promote seed germination soil pasteurization for killing seeds and through gibberellin in Arabidopsis. The weeds. Plant Protection Quarterly (In Plant Journal 47, 124-39. press). Pearson, S., Khattak, A.M. (2006). Spectral Brodie, G., Pasma, L., Bennett, H., Harris, fi lters and temperature effects on the G. and Woodworth, J. (2007c). Evalua- growth and development of chrysan- tion of microwave soil pasteurization themums under low light integral. Plant for controlling germination of perennial Growth Regulation, 2006 49 (1), 61-8, ryegrass (Lolium perenne) seeds. Plant Scopel, A.L., Ballaré, C.L. and Sánchez, Protection Quarterly (In press). R.A. (1991). Induction of extreme light Chang, R. (1998). ‘Chemistry’, 6th edition. sensitivity in buried weed seeds and its (McGraw Hill, New York). role in the perception of soil cultivation. Connor, F.R. (1989). ‘Antennas’, 2nd edi- Plant, Cell and Environment 14, 501-8. tion. (Edward Arnold, London). Steadman, K.J. (2003). Dormancy release Corbineau, F., Belaid, D. and Côme, D. during hydrated storage in (Lolium (1992). Dormancy of Bromus rubens L. rigidum) seeds is dependent on tem- seeds in relation to temperature, light perature, light quality and hydration and oxygen effects. Weed Research 32, status. Journal of Experimental Botany 55, 303-10. 929-37 Dirac, P.A.M. (1927). The quantum theory Tikhonov, V.V. (1997). Dielectric model of the emission and absorption of radia- of bound water in wet soils for micro- tion. Proceedings of the Royal Society. 114 wave remote sensing, paper presented (767), 243-65. to IEEE International Geoscience and Einstein, A. (1951). The advent of the quan- Remote Sensing Symposium, 1997, Sin- tum theory. Science 113 (2926), 82-4. gapore International Convention and Froud-Williams, R.J., Drennan, D.S.H and Exhibition Centre, Singapore. Chancellor, R.J. (1984b). The infl uence Von Hippel, A.R. (1954). ‘Dielectric ma- of burial and dry-storage upon cyclic terials and applications’. (M.I.T. Press, changes in dormancy, germination Cambridge). and response to light in seeds of vari- Watts, A. (2006). Photocontrol of weeds ous arable weeds. New Phytologist 96, in plant production. Third Year Project 473-82. Report Dookie College University of James, A. (2000). Night ploughing to man- Melbourne. age weed emergence. Third Year Project Wesson, G. and Wareing, P.F. (1968). The Report Dookie College University of role of light in the germination of natu- Melbourne. rally occurring populations of buried Ladiges, P., Evans, B. and Saint, R. (2001). weed seeds. Journal of Experimental Bot- ‘Biology’, 2nd edn, (McGraw-Hill, Ro- any 20, 402-13. seville, Australia). Wooley, J.T. and Stoller, E.W. (1978). Light López-Granados, F. and Lutman, P.J. penetration and light induced seed (1998). Effect of environmental condi- germination in soil. Plant Physiology 61, tions on the dormancy and germination 597-600. of volunteer oilseed rape seed (Brassica napus). Weed Science 46, 419-23. Metaxas, A.C. and Meredith, R.J. (1983). ‘Industry microwave heating’. (Peter Peregrinus Ltd, London). Milberg, P. and Andersson, L. (1997). Sea- sonal variation in dormancy and light sensitivity in buried seeds of eight

106 Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 but many of us could name examples where we see them going the other way. The impact of weeds on our roadside secrets So what can we do? Well, as all local government staff know, Karen Jones, Rural City of Wangaratta, PO Box 238, Wangaratta, Victoria the solution needs to be done on the sniff 3676, Australia. of an oily rag, a juggling act of view points needs to be preformed (preferable in a suit of lycra) and you need to think outside the square, especially since roadsides are lin- ear reserves not blocks of land. The lycra is I am here today to share my passion for like me tapping them on the shoulder and a good start to get you out of your comfort roadsides and the challenges of inspir- ask them not to park their machines on zone and it gets you thinking as you have ing people to get out and bond with their tuffs of ‘scrappy, crappy and grey’ native not thought before. roadside before strapping on their spray grasses. No wonder they are left scratch- So where did Wangaratta start with pack. I use the word bond because that is ing their heads!! our thought process? Well, we started what our roadsides need. People need to Fire management personal also bring with how lucky local government staff understand that roadside weed control is another angle to roadside management are. Unusual thought you say? But if you a journey, sometimes of many years, it is where, based on historic practices land- think outside the square we have people not a quick spray and the problem is fi xed. holders spend hours and hours sitting on coming to us, to tell us, what they think Before I jump into the weed focus I would their tractors ploughing away their road- about roadsides. How lucky is this! Oth- like to start at the beginning of the road- sides. If they take a step back they also er land mangers spend huge amounts of side journey, to give you an insight into scratch their heads wondering why they money seeking community input. If there how people understand roadsides and the now have a weed invasion just in their is something local government does well weeds that grow on them. plough lines (in some cases they now have – we attract community input. Obviously, Roadsides are often perceived as that a higher fuel load then the natives species I would love the community input to be forgotten strip of crown land located in they started with) and was it worth the ex- about celebrating the fi rst fl owering of that no mans land between private proper- pense of fuel and their time. There is also their roadside orchids or the nesting of ty and the road. Places where weeds grow the fact that the behaviour and intensity of the grey crowned babbler but realistically in their own merry way, unaware the anxi- recent fi res showed that the majority of fi re we settle for the fact they will be telling us ety and stress they are causing. breaks on roadsides were in-effective. they have the best crop of roadside Pat- Working for Victorian local govern- The adjoining landholder opinion usu- terson’s curse in the district!! ment has given me the ability to realise ally is more sensitive because they try to Once digesting the fact of how lucky we that roadsides are not forgotten, if any- take their paddock focused weed man- are, we then moved on to try and under- thing they are the most thought about agement knowledge and try to adapt it stand what are the weed messages people crown land in Victoria, possibly Australia. the linear roadside. The difference with a have heard and what do they understand. More importantly I have learnt to listen roadside and a paddock is that the road- The positive is that the message is clearly and appreciate that everybody has a view side is usually long and linear in shape, out there that weeds are bad and therefore point about roadsides and in relation to there are a large number of uncontrolla- they must be controlled. Department staff weed management their comments are ble external intrusions and the edge effect and the Landcare movement have done a very rarely positive. provides great stress on the native vegeta- great job on weed education – landholders The focus of a person’s interest in road- tion (making it much easier for weeds to know Patto is bad, they know blackberries side weed management closely refl ects dominate). are the enemy and if you want an interest- the passion and interests of that person. There are also the roadside fi rewood ing discussion ask them about willows! The more general comments relate to the collection and grazing view points that What we found is that the understand- fact that everybody drives past roadsides, need to be considered. ing as to why roadside weeds are bad was we all spend too many hours of our lives To the fl ora and fauna departmental lacking. There is great understanding of watching them, zoning out on them and people, naturalists, friends of groups and weed control in a cropping paddock, a in some cases they are where we interact environmentalists, roadsides are seen as vineyard or alpaca farm but when it comes with our native wildlife – not always in the saviour for our habitat future. Road- to weeds growing within a native area this a positive manner. This understanding sides have been lovingly referred to as the is a different story. If you think about it, of roadside management is what can be window into our past, the source of genet- this knowledge does exist but usually referred to as the 100 km per hour wind- ic material on which the future of our bare- within departmental staff, for example screen opinion. paddock revegetation projects depend on, Parks Victoria Rangers or weed contrac- To local government, VicRoads and as well being the Ned Kelly of the natural tors, not individual land holders. This is roadside contractor staff our appreciation world. Why Ned Kelly? a broad generalisation and I acknowledge and interaction with roadsides can get Roadsides and Ned can be described that there are many friends of people bogged down in the fact there are more as being famous for all the wrong reasons, and Landcare people that will have more than 23 different legislations, strategies yet both are an important part of what it knowledge of weed control in native areas and requirements relating to roadside means to be Australian. As well as the fact that I ever will. management. If you know of any other that many of our endangered ecological Continuing the thought process – we crown land that has as many rules and vegetation classes (EVCs) and endangered worked out that these bushland land man- regulations I would like to know. animals are only found on roadsides and gers and staff have been taught to look at We are also dealing with staff coping like Ned, they are facing their last stand. all elements of the bush not just the ele- with a huge 180 degree change in their This battle is seeing the armour of Mother ment of weed management. work practices. We have engineers and Nature slowly being eroded away and we So, by taking a step back we discov- grader drivers who, as recently as 15 years are seeing the diversity of roadside eco- ered why the message of weed control on ago, were clearing chain widths of road- systems being slowly disappearing. I have roadsides has been missing the mark. So side native vegetation. Now they are hav- yet to see a low conservation roadside be the challenge Wangaratta accepted was ing to dealing with young green things upgraded to medium or a medium to high to fi nd a way to assist landholders to see

Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 107 roadsides through a range of different their roadside before they spray. This is eyes. a huge change in behaviour especially in We discovered the solution existed in areas where Chilean needle grass grows the fact that people come to local govern- in amongst native grasses and historic ac- ment with their knowledge of roadsides cidental kill rates of the native grasses has – again, how luck are we! Another an- been high. There have also been phone gle included in the solution came from a calls from as far a fi eld as NSW, SA and presentation I heard at a Landcare confer- WA. I am always fascinated how network- ence where the speaker challenged us as ing spreads the message! I have also had to why images of salinity and bushfi res calls from many environment offi cers from always get the funding over weeds? The other Victorian local governments where a answer related to the images – salinity member of their community has seen the and bushfi res evoke emotion from people fi lm and are now putting pressure on their as they highlight the devastation. Now Council! think about weed images. Generally they Running in the background of this edu- have been taken to help people identify cation process is the debate of who is actu- the weed not to stir emotion or open the ally responsible for roadside weed con- purse strings. On more than one occasion trol. Personally I do not want to wade into at a weed information stall, a person look- this debate without a nice sunset to watch ing at the weed in a pot has ask me how and bottle of red to share. The outcome much would I like for it as they would of this debate, if it ever comes, will be de- like to plant it in their garden!! I would cided by solicitors and the powers-to-be not think I am the only one to experience in Melbourne, not by me. With guidance this request!! from the Municipal Association of Victoria There is so much jargon and terminolo- (MAV) mention is given to this issue in the gy in the natural resource management in- fi lm without distracting from the key mes- dustry that it does not take long for people sages we want people to hear. to feel overwhelmed and you notice their It is with these key roadside messages eyes glazing over. Therefore we needed I would like to conclude my presentation. something to engage them. To date, roadside weed management mes- We were also conscious that there are sages have focused from the road and many landholders with the attitude that looked toward the property. We have tried they know enough about weed control, to turn this around and looked at them or that they do not seek out information from the property through the roadside to as they are embarrassed by the fact they the road. By understanding the secrets of are suppose to know about weeds but in your own roadside and by looking at the reality they do not. activities undertaken on your roadside a Now if I was to lean closer to the mi- landholder will start to understand why crophone and ask you ‘Can I tell you a weeds grow where they do. For the fi rst secret?’ Would you be interested? Would time in seven years I have had a landhold- you not want to know more? Even better, er approach council wanting to know how once I tell you the secret, many of you will to restore a high conservation roadside he think of someone else you would like to has meandered through with a plough for tell the secret to. his long life time. His enquiry came after Well the secret I would like to share watching our fi lm. While I do not have the with you, is this roadside fi lm, ‘Keeping pulling power of Al Gore it is still a start Roadside Secrets Safe’. in the right direction. For those interested We took many local people out onto in a copy of the roadside fi lm, please come their roadsides and captured their knowl- and take your free copy. Please limit it to edge. Over 30 hours of interviews, road- one copy each but if you would like more side footage and car cam have been con- please let me know. Thank you very much densed into a 25 minute fi lm which cap- for your time and I welcome any ques- tures the wide range of secrets found on tions. our roadsides. This fi lm has been distributed for free to all relevant local community groups, the other 78 Victorian local Councils, NE NRM agencies, local schools and to inter- ested persons. We have even been fortu- nate that the community television station, Channel 31, viewed it to its Melbourne viewers. Never did I think there would be so many Melbournites interested in rural roadsides! There is a twist to this secret and that is we want the secret to get out. The mes- sage has been spreading far and wide and I now receive many phone calls from people wanting to know what else is on

108 Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 Assessment of orange hawkweed in Ballarat Containing orange hawkweed (Hieracium Assessing an infestation of a weed can be described as including the process of de- aurantiacum) infestations in the Ballarat area termining the identity, size, location and density or distribution of a particular spe- Simon Martin, Department of Primary Industries, 402–406 Mair Street, cies. The process of assessment is usually Ballarat, Victoria 3350, Australia. undertaken during a site inspection and can be assisted by the use of photographs and fi eld reporting sheets to record the par- ticulars of the site, including land tenure, Summary The Victorian Department of domestic day trips per year (City of Bal- date and specifi c location coordinates. Primary Industries (DPI) has successfully larat 2007). This has interesting implica- The following is an assessment de- contained and is working towards the tions for the spread of weeds through the scription of known in-ground infestation eradication of orange hawkweed (Hier- vector of human transport and this will be sites of orange hawkweed, found within acium aurantiacum) in the Ballarat area in explored further. the Ballarat area and recorded by the DPI western Victoria. Orange hawkweed is a Weed Alert program. These infestations declared State Prohibited Weed in Victoria Orange hawkweed in Victoria have been recorded on the DPI Integrated and is a signifi cant threat to the State’s ag- Orange hawkweed Hieracium aurantiacum Pest Management System (IPMS) database ricultural industries and natural environ- (F.W.Schultz & Sch.Bip.) was lodged for and are described in Figure 4 using site ment. Several garden and nursery infesta- the fi rst time with the National Herbarium numbers. tions have been located around Ballarat. of Victoria in 1999, from a Falls Creek site These have been treated and are regularly (Williams and Holland 2007). Site 1 monitored. After an assessment of the potential in- An infested site found in the Ballarat area vasiveness of Hieracium species and the was at a property that was initially devel- Introduction history of the plant in other countries, all oped as a nursery and display garden. The Ballarat species of Hieracium were declared as State property which is around two hectares in Ballarat is described as ‘Victoria’s larg- Prohibited Weeds in the State of Victoria, size was once described as being ‘along est inland city’. It is located at 441 metres on the 22nd of May 2003 (Victorian Govern- the lines of the great gardens of Europe above sea level with a population of ap- ment 2003), under the Catchment and Land and England… [with] hidden gardens, proximately 88,000 people. Major indus- Protection Act 1994. dry stone walls, waterfalls, bridges, mani- tries include tourism, retail, manufactur- In particular, hawkweeds were consid- cured hedges and much much more’ (City ing and community services (City of Bal- ered to be a potential threat to ‘pastures of Ballarat 2007). larat 2007). and to the wool, meat and dairy indus- Pots of orange hawkweed were re- The Ballarat area is described in this tries’ (DPI 2003 p. 2). The extent of the area moved from the display garden in which paper as the area of land consisting of the of potential distribution of orange hawk- a 0.005 hectare in-ground infestation was statistical local areas of ‘Ballarat Central’, weed was determined during Phase 1 of recorded on IPMS. The in-ground infes- ‘Ballarat Inner North’ and ‘Ballarat South’ the Noxious Weed Review process and tation occurred along a walking path. (Figure 1). The total area containing ag- was created using CLIMATE® modelling Non-selective herbicide was used to spot ricultural activity is described as being (DPI 2007). The modelling placed the Bal- spray areas of infested lawn on the 17th of 6369 hectares in size. Of the agricultural larat area within a ‘high’ category of po- November 2005. After the initial inspec- land 2753 hectares of land was sown to tential orange hawkweed distribution. tion by Departmental offi cers it was dis- pasture, and there were 14,036 sheep and Orange hawkweed was fi rst recorded, covered that the nursery was to be sold lambs produced during 1996/7 (Austral- outside of the high country, by DPI in 2002 and the new owners were to take over the ian Natural Resource Atlas 2007). at a nursery near Geelong, and in 2003 property from early July 2006. There is a high attraction and infl ux of in a small ‘collectable plant’ nursery on The next inspection on the 29th of Sep- tourists to Ballarat during festivals and the outskirts of Ballarat. It has since been tember 2006 resulted in the discovery of events throughout the year, with one esti- found in West Gippsland at Yallourn in eight orange hawkweed infested pots mate suggesting that there are 1.8 million 2004, and again in Ballarat in 2006. within a disused glasshouse (Figure 2). The neglected and run down glasshouse contained a large number of orange hawk- weed infested pots within which the orig- inal nursery plants had died (Figure 3). Discussions with managers at the site re- vealed that the pots within the glasshouse had been offered to those associated with the property. This led to the creation of a warning notice attached to the glasshouse asking that anyone who had removed pots from the site contact the Department. The eight pots were voluntarily surrendered and taken to the State Government Offi ces for storage prior to disposal. After the initial pot surrender it was considered that the remaining pots within the glasshouse had the potential to con- tain viable seeds, seedlings or stolons of orange hawkweed. Two offi cers returned Figure 1. Statistical local areas of Ballarat. to the site on the 18th of October 2006

Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 109 and worked to manually remove the pot during May 2006 and September 2006 was to fl ower and set seed. These areas would material including potting mix and plant conducted using non-selective and selec- therefore from a priority area for monitor- matter in a hygienic manner. Two hundred tive herbicides within lawn areas. A follow ing and surveillance. and forty litre garbage bags were used to up inspection in February 2007 revealed contain the material. Two bags, one inside no further plants were present. Treatment of orange hawkweed the other, were used to contain the mate- The treatment of orange hawkweed in the rial and bags were tied at the top. Bagged Site 5 Ballarat area has mostly been confi ned to material was stacked into a tandem trailer An orange hawkweed infestation, fi rst re- the expected fl owering period of the spe- and transported to the local land fi ll site corded on IPMS on the 12th of May 2005, cies between January and March (CRC with which an arrangement to bury the was discovered at the Ballarat Botanic Weed Management 2003) but has also in- material had been made. A total of 1.18 Gardens. The infestation in a lawn area cluded the period between September and tonnes of material was buried in an asbes- adjoining garden beds was treated with November. The use of non-selective herbi- tos trench at the land fi ll site. This site was a non-selective herbicide on the 17th of cide by spot spraying has been shown to then recorded on IPMS. November 2005. The infestation was said be an effective treatment method, however to have spread to this location from an it does have the disadvantage of creating Site 2 unknown source some years before (Peter large patches of bare ground and effecting The second infested site of orange hawk- Marquand personal communication 2nd non target species within lawn areas and weed was on a property that formerly of February 2007). A subsequent inspec- garden beds. A shift to spot spraying of traded as a nursery. The infestation was tion on the 1st of May 2006, determined selective herbicide in more recent times recorded on IPMS in the 18th of May 2005. that a repeat treatment was required with has shown to be advantageous. Several garden beds around the nursery spot spraying of non-selective herbicide. An area surrounding a known infesta- site contained plants that were subse- Further inspections on the 23rd of No- tion within a lawn area can be treated using quently treated with non-selective herbi- vember 2006 identifi ed orange hawkweed a selective herbicide resulting in the de- cide on 17th of November 2005. Due to within two separate garden beds to the struction of numerous species of fl at weeds the diligence of the new manager of the east of the initial infestation as well as including orange hawkweed without leav- site, no longer trading as a nursery, the in- around the garden beds previously treat- ing bare soil prone to further weed infesta- fested areas were later staked and marked ed. Spot spraying with selective herbicide tion. The hand pulling of plants within in- with fl agging tape to allow them to be was used on garden paths and rock bor- fested sites such as those at the Ballarat Bo- easily identifi ed. A Departmental offi cer ders as well as within the newly identi- tanic Gardens, where rosettes were found returned to the site on the 23rd of May fi ed infested garden beds. The aim was within structured garden beds amongst 2006 and treated the infested areas with to treat rosette foliage as well as stolons, a variety of other annual and perennial non-selective herbicide. No further germi- which may have regenerated if missed by plants, was used. This treatment method is nations of the weed were found during an manual treatment. An inspection of the limited by the fact that not all of the plant inspection on the 12th of February 2007. gardens on the 2nd of February 2007 led rosette and root system may be accessible, to the discovery of several rosettes with- especially within heavily mulched garden Site 3 in the same gardens, on the soil surface beds, and the network of stolons leading An orange hawkweed infestation record- above Iris plants. Hand pulling was used from a rosette may not be identifi ed and ed on a nature strip in the Ballarat suburb to remove the rosettes and as much of the therefore left intact. Any remaining sto- of Wendouree was reported to DPI by a root material as possible with the material lons may have the capacity to re-shoot and local resident, a student of Ballarat Uni- being bagged and stored prior to hygienic therefore require further treatment. versity. The reporter observed an unusual disposal by incineration. orange fl ower and took it to the University The human aspect of orange to consult a lecturer. Upon receiving ad- Mapping orange hawkweed hawkweed spread vice that the fl ower was most likely that of Maps of weed infestations can be created A large number of human activities have orange hawkweed the reporter contacted from the data recorded within IPMS. The been important to the spread of the weed. the Department. An inspection found that collection of data is usually done as part These activities have included the sale and small rosettes of the plant were present in of the assessment process and is enhanced distribution of the plant as a garden plant, November 2006. These rosettes were treat- with the use of accurate spatial coordi- anecdotally reported throughout various ed by being spot sprayed with selective nates. From the points created to indicate nurseries in the Ballarat area. Two of the herbicide. No further rosettes were found an infestation patterns of infestation can known infested sites described earlier upon inspection on the 2nd of February be interpreted. Maps can help to defi ne the (sites 2 and 4), have been involved in the 2007. Catchment Management Authority Area sale of the plant and it is suspected that the in which the infestation falls and may help plant was also placed within garden beds Site 4 to identify those responsible for manag- at both sites. It may have been planted as A fourth site found to be infested with or- ing the land. This leads to benefi ts in plan- a garden specimen and used to promote ange hawkweed was a private residence ning infestation treatment and may also sales as a display plant. on the south side of Ballarat within a sub- be used as a tool to plan the validation or The fl owering period of the plant may urb with a larger than average property further searching for the location of the also help to create demand for the plant size on the border of State Forest. This same weed within the area. as it fl owers during a period when there property contained a very well main- Due to seed dispersal by wind as well as is limited colour within the garden. There tained garden that has been open to the the possibility of stolons of orange hawk- may also be accidental distribution of the public under the Australian Open Garden weed being distributed by machinery such plant through infested potting material. Scheme. The plant was described by the as lawn mowers, accurate maps may help There are a large number of visitors to owner as having been deliberately planted to identify areas in which a weed has the Ballarat throughout the year for several within a driveway garden bed between potential to infest. For example, areas to very popular festivals and special events. 1998 and 2000. Although the original plant the west and south east of all known in- The special interest events, such as the had been treated, seedlings of orange ground infestations have the potential to ‘Begonia Festival’ are conducted within hawkweed were later found within lawn become infested by wind dispersed seed the expected fl owering period of orange areas adjoining the original site. Treatment if any orange hawkweed plants are able hawkweed. With the assumption that

110 Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 the best time to trade a plant is while it is fl owering, this has implications for the export of orange hawkweed from the Bal- larat area. With visitors potentially having observed orange hawkweed within the Ballarat area during festivals prior to the sites being treated in the past, any number of trades may have occurred resulting in imports and exports of the species.

Conclusion After declaration of orange hawkweed in Victoria as a State Prohibited Weed, fi ve sites have been identifi ed with in-ground infestations within the Ballarat area. The human vector can be considered as being very important to the spread of the weed and continued work with the community will be needed to highlight the potential threat that this weed presents. Continued assessments, monitoring, Figure 2. Nursery pot infested with orange hawkweed. Source: Simon mapping and treatments by DPI have Martin, Department of Primary Industries. been required to ensure containment to date. These efforts which involve working towards eradication of the weed are neces- sary in Ballarat where the potential distri- bution has been assessed as being high. Of greatest threat are the very important agri- cultural enterprises of the Ballarat region, the surrounding area and Victoria. Continued successful containment and eradication of the weed will depend upon successful negotiation with land owners and managers from a range of land ten- ures. The timing of assessments and treat- ments will be critical to ensuring that all in-ground infestations of orange hawk- weed do not spread.

References Australian Natural Resource Atlas (2007). Agriculture – statistics – Cen- tral Highlands, http://audit.deh.gov. au/ANRA/agriculture/agriculture_ frame.cfm?region_type=VIC®ion_ code=220, viewed 21st May 2007. Figure 3. Glasshouse pots at former display garden Ballarat Source: Andrew City of Ballarat (2007). Profi le, http:// Staley, Department of Primary Industries. www.ballarat.vic.gov.au/About_Bal- larat/Profile/index.aspx, viewed on the 17th of July 2007. CRC Weed Management (2003). Weed management guide, orange hawkweed – Hieracium aurantiacum. Department of Primary Industries (2003). Under control pest plant and animal management news, No. 24, June 2003. Department of Primary Industries (2007). Victorian resources online, orange hawkweed potential distribution map: http://www.dpi.vic.gov.au/dpi/vro/ vrosite.nsf/pages/lwm_pest_plants_ orange_hawkweed. Victorian Government Printer (2003). Vic- torian Government Gazette, No. G21 Thursday May 2003. Williams, N.G.S, and Holland, K.D (2007). The ecology and invasion of hawk- Figure 4. A map of the known in-ground infestations of orange hawkweed weeds (Hieracium species) in Australia, in the Ballarat area. Plant Protection Quarterly 22, 76-80.

Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 111 to eradicate or even contain the incur- sions. Awareness of characteristics and Alert and action on Nassella charruana and a range of pathways in which weeds can be intro- duced into Victoria improves the chance other National Alert List species in Victoria of catching the incursion in its infancy, greatly improving the chances of eradica- A B Sarah Partington and Michael Hansford tion. All existing properties that contain A Catchment and Agriculture Services, Department of Primary Industries, or once contained incursions of these spe- 219A Main Street, Bacchus Marsh, Victoria 3340, Australia. cies have been surveyed. Follow-up moni- B Catchment and Agriculture Services, Department of Primary Industries, toring takes place to ensure all necessary Locked Bag 3000, Box Hill, Victoria 3128, Australia. incursion treatments occur seasonally to achieve an eventual eradication of incur- sion sites. As well as knowing the existing sites of the weeds, it is also necessary to have a good knowledge of the growth pat- Summary Six National Environmental hawkweed; Trianoptiles solitaria, subterra- terns of a species. The times of year and Alert List weed species were targeted nean Cape sedge; Piptochaetium montevi- forms of a plant that make identifi cation for follow-up surveys and treatment for dense, Uruguayan rice grass; Cytisus mul- possible must be known before a surveil- eradication in Victoria. The species were: tifl orus, white Spanish broom. This work lance program can be run. The potential Nassella charruana, Acacia karroo, Hieracium was largely a continuation of the work vectors of spread and the habitat in which aurantiacum, Trianoptiles solitaria, Piptocha- achieved by a previous project target- the species will grow are also very impor- etium montevidense, and Cytisus multifl orus. ing the same six species (Hansford 2005), tant to be able to locate outlying popula- New infestation sites were discovered for except that the new project placed more tions. Nassella charruana, Hieracium aurantia- emphasis on Nassella charruana (lobed nee- cum and Cytisus multifl orus. To facilitate dle grass), a species which had shown a 3. Control prompt reporting and effective treatment marked increase in the number of known The control methods used to treat the of new infestations of Nassella charruana the infestation sites due to recent detection ef- weeds in this project were based on those Department of Primary Industries (DPI) forts (Hansford 2006). The latest project, used in the previous project targeting these developed networks with affected land- which ran from July 2006 to June 2007 was species, as described in Hansford (2005). owners and local governments. Neither called ‘Alert and action on Nassella char- These methods are outlined below: Trianoptiles solitaria nor Piptochaetium mon- ruana and a range of other National Alert tevidense were detected during the project List species in Victoria’ and was funded Physical/mechanical A. karroo trees were period. Acacia karroo is a species restricted by the ‘Defeating the Weed Menace Pro- physically removed as follows: to a small number of sites and is consid- gramme’. The project worked with many • All branches were chipped on-site in ered very close to being eradicated from key stakeholders to enable early detection enclosed chippers. Large limbs and Victoria. Negotiations with landowners and treatment of infestations and to raise trunks were cut to smaller size for will progress to ensure the removal of the awareness amongst the general public of ease of handling and loaded with the last remaining Acacia karroo trees. How- these species. Three of the targeted spe- chipped material. ever, much work is predicted in the future cies, N. charruana, A. karroo, and H. auranti- • All tree stumps were cut close to the with Nassella charruana, as it occurs in the acum are declared State Prohibited Weeds ground and swabbed with herbicide. midst of a land development area on the in Victoria under the Catchment and Land • The truck that transported the plant northern outskirts of Melbourne. Caution Protection Act 1994. This status requires material to landfi ll was covered and will need to be taken to ensure that weeds their eradication from the State. secured to avoid any plant material are not spread through the land develop- escaping en route. ment process. DPI will need to continue Materials and methods • All of the plant material was then de- working with land developers to make 1. Framework posited to an excavated hole at a land- certain that suitable hygiene protocols are The ‘Weed Alert Plan Victoria 2006/2007’ fi ll site, and then the hole was covered adopted. describes how the surveillance and re- with soil and other rubbish as part of Keywords National Environmental sponse to potential, new and emerging the landfi ll operation. Alert List, Nassella charruana, Acacia kar- weeds, including the six species targeted • Site revegetation was undertaken using roo, Hieracium aurantiacum, Trianoptiles soli- in this project, is managed in Victoria. non-invasive species. taria, Piptochaetium montevidense, Cytisus Weed Alert is a State Government pro- multifl orus. gram developed to prevent the introduc- Chemical The strategy to control N. char- tion of serious new weeds to Victoria, ruana was to ‘spot spray’ individual plants Introduction and to eradicate the most serious incur- to kill these plants before they set seed. The National Environmental Alert List sions that have naturalised. The program This enables the germination of soil stored (Alert List of Environmental Weeds) was focuses on State Prohibited Weeds and seed which can then be followed-up con- developed in 2001. It consists of 28 weed Victorian Alert Weeds by overseeing the tinuously with ‘spot spraying’ when species currently in the early stages of es- surveillance, collection, identifi cation, as- growth conditions are suitable. This meth- tablishment that have potential to become sessment of and response to these species. od will exhaust the seed bank over time. A a signifi cant threat to biodiversity if they Weed Spotters are enlisted to assist with non-selective chemical is used which may are not managed (Department of Envi- the surveillance part of the process. They promote the appearance of bare ground in ronment and Water Resources 2007). Six assist the Weed Alert team by keeping an the dense infestations sprayed. This will National Environmental Alert List weed eye open for these species (McInerney and encourage further germination from the species were targeted for follow-up sur- Robinson 2007). seed bank. It is expected that once the seed veys and treatment for eradication in Vic- bank is exhausted natural regeneration of toria. The species were: Nassella charruana, 2. Surveillance native or introduced pasture grasses will lobed needle grass; Acacia karroo, Karoo Without ongoing detection new weeds occur, providing suffi cient ground cover thorn; Hieracium aurantiacum, orange can establish in areas making it diffi cult in the future.

112 Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 The strategy to control H. aurantiacum with them to ensure seed-infested soil is program has been. While assessing exist- involves spot spraying individual plants disposed of appropriately. ing properties, neighbouring properties or dense patches before they set seed. This were surveyed to look for new N. charruana enables the germination of seed stored in Results and discussion incursions. If new incursions were found, the soil, which will be followed up with Nassella charruana – lobed needle grass landowner information was obtained and more, spot spraying. In most instances a Even within its native range in South the landowner was contacted. During the selective chemical is used. America, N. charruana is reported to be infestation surveys the following informa- For C. multiflorus, all plants were a weed due to its competitiveness, un- tion was recorded: sprayed during fl owering before seed set. palatability and very sharp, needle like • Global Positioning System (GPS) points C. multifl orus is best targeted during its seeds. Once thought to only infest a small of patches of the weed over the prop- spring fl owering as it can be diffi cult to number of properties on the northern out- erty detect it amongst other vegetation when it skirts of Melbourne (CRC for Australian • The area the infestation covered is not fl owering. Weed Management 2003a), N. charruana • The percentage density of the cover- has now been found on many more prop- age 4. Site monitoring erties in the Whittlesea council region. The • If the plant had been sprayed how All sites were monitored after treatment increase in number of properties infested much had died off from the last treat- to make sure treatment was effective and from 15 (reported in 2004) to 53 (see Table ment. to assess whether more treatment was 1) has been due to more surveillance in All of this information was then recorded required that season. All landowners the region by both Whittlesea City Council in the IPMS. If a known infestation cannot were made aware of the need for follow employees and DPI employees. With more be located, it is followed up each year until up monitoring by DPI offi cers. All treat- surveillance planned for the 2007/08 sea- DPI is confi dent that the weed has been ment and monitoring were recorded on son it is expected that more new sites will eradicated from the site. By carrying out DPI’s Integrated Pest Management Sys- be found. fi eld surveys, DPI can maintain up to date tem (IPMS). Historically this land was predominant- records and also keep up to date on other ly open grasslands with some scattered avenues in which the weed may be moved 5. Machinery hygiene procedures tree cover. After settlement this area was onto other land (i.e. soil removal, slash- There are many vectors of spread for used for grazing and there has also been ing etc.). DPI works closely with property weeds, including wind, water, animals some quarrying in the area. Now this area owners to minimise the spread of weeds. and machinery used for both farming and lies right on the edge of the urban fringe, The success of the N. charruana eradica- land development. While working with with many new housing developments tion program relies on catching new infes- land owners to eradicate a species it is also under way. As the increase in properties to tations in their early stages. By working important to make sure they are not per- be treated continually by contractors has with local councils new infestations have petuating the problem by spreading the increased there is need for coordination of been found and treated during the project weeds to other areas of the property or both spraying and surveying. In addition period. The City of Whittlesea (where most to other properties. Machinery hygiene is to the IPMS database, the Project Offi cer of the infestations occur) has offi cers that import to stop weed spread. It involves maintains a spreadsheet and landowner are trained in the identifi cation of N. char- washing down any part of a vehicle that fi le system that encompasses land owner ruana. Once found the offi cers report the may be harbouring any part of a weed details, size and density of the infestation, incursion to DPI who follow up the report. that may grow. Mapping the infestation times of treatments occurring in the year, Some incursions have been found outside on the property and having knowledge maps, photographs and other relevant in- the boundaries of Whittlesea, therefore the of times of year that the weed is seeding formation. As more properties will be de- councils surrounding have been contacted are also important in the management of veloped in the future it is imperative that and made aware of the weed. An infor- weed spread (Tyers et al. 2004). At known good records are kept and relationships mation and identifi cation session was run infestation sites for N. charruana, DPI has are maintained with developers. at Hume City Council in February 2007. informed land developers about weed All known N. charruana sites were sur- Items covered were what the weed looks spread and the importance of machinery veyed at least once during the project pe- like, its growth habit, how to report an in- hygiene and will continue to work closely riod to assess how effective the spraying cursion and what to tell land owners if it

Table 1. The number of known infestations in 2004 (as reported in Hansford (2005)) compared to 2007. Number of Additional number Total number Number of properties with Locations of infestations of infestations infestations infestations of detected by June detected by treated by Species reported by 2004 infestations 2007 June 2007 June 2007 Nassella charruana 15 Whittlesea, Hume, 38 53 51 Darebin Council regions Acacia karroo 10 Werribee, Bendigo, 0 No new infestations, 1 site 7 trees treated Williamstown only 2 trees left in or removed Victoria Hieracium aurantiacum 27 Falls Creek, Ballarat, 24 51 (33 in Village at Fall 51 Gippsland Creek) Trianoptiles solitaria 1 Balwyn North 0 1 0 Piptochaetium montevidense 0 Altona 0 0 0 Cytisus multifl orus 2 Creswick, Ballarat 1 3 3

Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 113 is found on private land. Around 20 coun- chemical and left standing to die. This tree other brooms, is very invasive and forms cil offi cers attended the training. Training had started showing signs of die-off within dense stands. It can invade in many differ- events in the other councils surrounding a few weeks and will be closely monitored ent habitats and soil types (CRC for Aus- Whittlesea are planned for 2007. DPI is over the next few months to ensure that tralian Weed Management 2003e). The ex- also working with councils to ensure that the tree dies. isting site, Creswick Regional Park, where strict machinery hygiene protocols are The Williamstown and Bendigo trees C. multifl orus has been treated in previous put in place for land developers and their are now the only two known trees of A. years, was surveyed during spring when contractors (particularly in the Whittlesea karroo left in the State. It is hoped these the plant was most identifiable. Some region). trees will be able to be removed in the near sites within the park that had been treated DPI is also working with developers future. This would effectively mean the were mostly clear of the weed; however, to minimise the spread of N. charruana eradication of all known trees of A. kar- probably due to treatment being missed through the movement of soil on the prop- roo from Victoria. However, issues such as in 2005/06, some sites had grown back erty and to other land. It has been proposed their Heritage listing add complexity to or spread to other areas. A school camp that developers scrape seed contaminated the situation. The next step will be for DPI that borders the park was found to have a soil and use as fi ll or for landscaping and to undertake more liaison with a range of medium stand of C. multifl orus within its cap with clean soil and top soil. Burying stakeholders concerning these trees. boundary. Another site in Ballarat (15 km the seed deep in the soil profi le will limit away from Creswick) was also discovered the growth of germinates and eventual- Piptochaetium montevidense – this season. A possible source of the infes- ly the seed will loose its viability. Other Uruguayan rice grass tations on this new site is garden refuse avenues that developers could engage in In its native range in South America, P. being dumped on the edge of parkland. is to take the scraped soil to landfi ll and montevidense forms dense tussocks to about All of the new infestations were part of the have it buried deep beneath other layers 0.5 m high and its growth is stimulated Ballarat Cemetery property in Invermay. of waste. by fi re. P. montevidense may compete well A contractor was engaged to treat all of against native species because it is resist- the known infestations of C. multifl orus. Acacia karroo – Karroo thorn ant to grazing. In its native range it is the All plants were sprayed before they set There have been no new discoveries of most prevalent species of its genus (CRC seed, which greatly reduced the chance A. karroo found within Victoria (see Table for Australian Weed Management 2003c). of spreading. A very large specimen was 1). The known infestations that still exist The original infestation of P. montevidense found while surveying in the Creswick Re- today were all planted many years ago, at Cherry Lake Altona had been inadvert- gional Park this season. This plant had not prior to the 2003 declaration of A. karroo ently completely buried by earthworks at been found in earlier years and is thought as a State Prohibited Weed. Originating in some time in the past. The burial of the to be at least 15–20 years old. South Africa the A. karroo is a large grow- original infestation was noted at the previ- ing acacia with large thorns (CRC for ous inspection of this site in 2004 (Hans- Hieracium aurantiacum – orange Australian Weed Management 2003b). At ford 2005), where P. montevidense was not hawkweed the start of this project, there were only detected. The purpose of the visit to the Hieracium aurantiacum is a perennial plant three known sites: Werribee Open Range site in the current project was to conduct a that can grow up to 400 mm high. It has Zoo (seven trees in enclosures), Bendigo further follow-up inspection. The site was bright orange flowers and dark green and Williamstown. During this project pe- checked twice during the spring/summer hairy steams and leaves. It potentially riod the seven trees at the Werribee Open of 2006, and both times there was no sign threatens alpine regions and temperate Range Zoo were treated by DPI contrac- of the plant. Given that the species has tablelands. Hawkweeds have been found tors. There were issues with timing the not been detected during several previous to be extremely invasive in areas overseas extraction of the trees due to the incon- surveys, it seems very likely that the spe- (CRC for Australian Weed Management venience to the zoo and having to re-de- cies has been eradicated from this site. 2003f). There has also been an increase in sign the Meerkat enclosure once the trees the number of known infestations of H. were removed. The Zoo also wanted to Trianoptiles solitaria – subterranean aurantiacum since the 2004 report. This re- leave one dead A. karroo tree standing in Cape sedge sult could be due to the greater time spent the Cheetah enclosure as a feature. After Trianoptiles solitaria is a small leafy annual in surveying for the weed. The main out- some negotiation, the outcome was suc- herb that can grow to 200 mm high and breaks of H. aurantiacum have been in the cessful, with six trees being removed from has the ability to out compete more de- Alpine region and the Ballarat area. Both the zoo, the remaining tree being poisoned sirable indigenous plants. T. solitaria has outbreaks are believed to have occurred and left standing dead in the Cheetah en- only been reported from Winfi eld Park, a due to escape from previous garden plant- closure (as requested by the Zoo). The zoo small park in the centre of Balwyn North ings. Snow machinery has probably also also received assistance from the project (CRC for Australian Weed Management contributed to the spread in the Alpine re- in the revegetation of the affected Meerkat 2003d). The park was surveyed twice gion. The incursions in Ballarat have most- enclosure. during spring/summer 2006. No plants ly been found in gardens or old nurseries. In May 2007, an arborist was engaged were found during the inspection and it In late 2006, Parks Victoria (PV) em- to remove the trees from the Werribee was concluded that the plant might not ployed a fi eld offi cer to survey and treat all Open Range Zoo. All trees were de-limbed have come up this year due to the severe H. aurantiacum sites in the Falls Creek area. and taken to the chipper that was parked drought conditions experienced in the Due to bushfi res in late 2006, this program close by. Extra large limbs and trunks were spring of 2006. The site will be inspected did not run for the expected timeframe, placed in the truck with the woodchips. again in the spring of 2007 in an attempt however all located H. aurantiacum plants All material was taken to a quarantine to detect it. were treated. New arrangements are now landfi ll site, and was deposited in a hole in place for the next detection season, and at the landfi ll site to be buried under sev- Cytisus multifl orus – white Spanish stakeholder meetings on H. aurantiacum eral meters of landfi ll. The cut stumps of broom detection and management are held each A. karroo were treated with chemical to Cytisus multifl orus is a large shrub that year, with operational participants includ- ensure that re-shooting did not occur. The grows to 3 m high. It has distinctive white ing PV, DPI, Falls Creek Resort Manage- tree that was left standing in the Chee- fl owers with a pink streak and the foliage ment Board, Mt Buller Resort Manage- tah enclosure was drilled and fi lled with is a green grey colour. C. multifl orus, like ment, and the local Falls Creek water

114 Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 authority and ski-lift company. Informa- Finlay and Maarten Hulzebosch (Werribee tion on H. aurantiacum identifi cation has Open Range Zoo); Val Stajsic (National also been sent to a number of stakeholders Herbarium of Victoria); Renée Nicholson and businesses, including ski lodges and (Biosis Research) and Sheree Campbell residents at Falls Creek. This will ensure (Ecology Australia) for reporting infesta- that key messages and identifi cation tech- tions of N. charruana. niques are circulated in the public domain. PV and DPI are also working closely to- References gether on data exchange and management Blood, K. (in publication). Weed Alert Plan techniques for H. aurantiacum. Victoria 2006/2007. Department of Pri- In the Ballarat area, DPI coordinated mary Industries, Bacchus Marsh. all surveys and treatments of H. aurantia- CRC for Australian Weed Management cum. DPI also held two information days (2003a). Weed management guide – featuring H. aurantiacum. These informa- lobed needle grass Nassella charruana. tion days were run at a market in Bal- CRC for Australian Weed Management larat during December 2006 and February (2003b). Weed management guide 2007. The sessions were to alert the public – Karroo thorn Acacia karroo. to the presence of the weed in the area, CRC for Australian Weed Management and to help fi nd any new incursions of this (2003c). Weed management guide weed. – Uruguayan rice grass Piptochaetium montevidense. Conclusion CRC for Australian Weed Management The project was successful. There were (2003d). Weed management guide new detections of several of the targeted – subterranean Cape sedge Trianoptiles species during the project period, namely solitaria. N. charruana, C. multifl orus, and H. auran- CRC for Australian Weed Management tiacum which were subsequently treated (2003e). Weed management guide and the data entered on database. One – white Spanish broom Cytisus multi- infestation site of A. karroo, was treated fl orus for eradication during the project period, CRC for Australian Weed Management bringing this species very close to being (2003f). Weed management guide – or- eradicated in Victoria (there now being ange hawkweed Hieracium aurantiacum only two single tree infestation sites left Department of the Environment and Water in Victoria). P. montevidense was not de- Resources (2007). Weeds on the Nation- tected during the project period, suggest- al Environmental Alert List. Available: ing that this species (combined with the http://www.environment.gov.au/bio- results of past surveys) has most likely diversity/invasive/weeds/alert-list. been eradicated from its only known site html (Accessed: August 2007). of occurrence in Australia. Similarly, T. Hansford, M. (2005). Site management solitaria was not detected at its only previ- strategies for six National Environmen- ously known site of occurrence during the tal Alert List weed species in Victoria. project period. However, this result may Proceedings of the 2nd Victorian Weed have been attributed to the drought, and Conference, Bendigo, pp. 11-14. a further search will be conducted in the Hansford, M. (2006). Lobed needle grass: spring of 2007. There will be a need for management strategies for a Priority continued detection and eradication work Agricultural Sleeper Weed, National for N. charruana for some time, and DPI Environmental Alert List weed and will coordinate this process with affected State Prohibited Weeds in Victoria. Pro- land owners, land developers and others ceedings of the 15th Australian Weeds through the Weed Alert program. Conference, eds C. Preston, J.H. Watts, N.D. Crossman, pp. 645-648. (Weed Acknowledgements Management Society of South Austral- We thank the Defeating the Weed Menace ia, Adelaide). Programme for providing funds to support McInerney, C. and Robinson, V. (2007). the project ‘Alert and action on Nassella Weed spotter engagement plan. Victo- charruana and a range of other National rian Department of Primary Industries, Alert List Species in Victoria’; the Depart- Melbourne. ment of Sustainability and Environment Tyers, G., Grech, C. and Baldyga, N. for providing funds to support the work of (2004). Machinery hygiene – Landcare the Weed Alert program; DPI for support notes, Victorian Department of Primary at all levels in particular the support of Industries, Frankston. other Weed Alert team members; Natasha Baldyga and John Edwards (City of Whit- tlesea) for assistance with the detection and management of N. charruana; Peter Woldarczyk (GAGIN Pty Ltd) for con- trol works on N. charruana; Parks Victoria for assistance with infestation details for C. multifl orus and H. aurantiacum; Simon

Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 115 provide a local contact point for existing Weed Spotters to make reports and receive The Victorian Weed Spotter network training and information.

C.J. McInerney, Weed Spotter Coordinator, Department of Primary Industries, Weed Spotter network information Corner Little Malop and Fenwick Streets, Geelong, Victoria 3220, Australia. Location and recruitment trends The Weed Spotter network rapidly ex- panded during the 2006/07 fi nancial year. Weed Spotter numbers almost doubled from 595 in July 2006 to 1187 at the end of Abstract The Weed Spotter network is a engagement across the Weed Spotter net- June 2007, as indicated by the sharp incline weed surveillance program where indi- work. This primarily occurs through the in Figure 1. viduals are trained to look for and report Weed Spotter Engagement Plan (McIner- The distribution of Weed Spotters has high priority Weed Alert species in Victo- ney and Robinson 2007) which outlines a been represented in Figure 2. Locations ria. They assist in validating the distribu- strategic approach to recruiting new vol- marked with a square indicate gaps where tion of these weeds which pose a serious unteers and maintaining the interest and there were previously no Weed Spotters in threat to the State’s assets and values. Re- effectiveness of existing Weed Spotters the network and the locations marked with ports are made through Weed Alert Con- across Victoria. a triangle represent where numbers have tact Offi cers (WACOs) in the Victorian increased. The number of Weed Spotters Department of Primary Industries (DPI) Role of the Weed Alert Contact Offi cers varies due to a variety of factors including Weed Alert program. Weed Spotters are (WACOs) the population of a catchment, the length individuals recruited from the community, The role of a WACO within the Weed of time a WACO has been present and the government or industry with an interest in Spotter network is to strategically recruit time dedicated to active recruitment. For reporting these priority weeds. new volunteers from their region and to example, during 2006/07 WACOs were The network was offi cially launched and activated in October 2006 and has been expanding since this time. WACOs 1200 are now employed across the State and have been actively engaging with Weed 1000 Spotters, holding training and other events regularly. The network is now supported 800 centrally by a Weed Spotter coordinator. The fi rst year of network data can now be 600 examined to inform future recruitment ac-

tivities and to determine the effectiveness Weed Spotters of Weed Spotters to look out for and report 400 Weed Alert species. 200 Introduction Weed Alert species 0 Weed Spotters are asked to look for and 2001/ 02 2002/03 2003/04 2004/05 2005/06 2006/07 report Weed Alert species which are: Year State Prohibited Weeds Are declared un- der the Catchment and Land Protection Act Figure 1. Increase in registered Weed Spotter numbers. Note that there are 1994 (CaLP) and are illegal to buy, sell, dis- 28 registered Weed Spotters not included in this graph as no recruitment play, propagate, possess for sale, deposit date is available. onto land, bring into or transport around Victoria. They either do not occur in Victo- ria or are already present but it is reason- able to expect that they can be eradicated. Weed Spotter localities New spotters at existing locations Their listing refl ects the very serious harm New locations during 2006–07 these plants cause. Currently there are 25 Existing locations declared State Prohibited Weeds. CMA boundaries

Victorian Alert Weeds Are weeds that potentially pose a serious threat to Victoria’s agricultural and natural assets or could affect human health. They may be naturalised in small numbers but are still eradicable or have not yet reached Victoria, but are believed to present a high risk. Each species will be targeted for sur- veillance by Weed Spotters to determine their distribution throughout Victoria.

Role of the Weed Spotter Coordinator The Weed Spotter Coordinator works with the WACOs to ensure consistent Figure 2. Distribution of registered Weed Spotters across Victoria.

116 Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 recruited later in the year in East and West 250 Gippsland and therefore Weed Spotter 211 Events held in 2006/07 numbers are lower in these regions. 200 Weed Spotters recruited Recruitment methods 150 Information has been gathered on the 128 methods used to recruit Weed Spotters 101 during 2006/07. The most frequently used 100 methods were introductory Weed Spotter 52 50 training sessions, Weed Alert presenta- Spotters Events/Weed 50 27 tions, Weed Alert displays and secondary 17 15 10 12 Weed Spotter training sessions. 0 Introductory Weed Spotter training Introductory Displays Presentations Secondary Other Unknown sessions are planned, organised and pro- training training moted by WACOs for potential and new- ly registered Weed Spotters within each Recruitment method catchment. This is an effective method of recruiting new Weed Spotters with the largest number of Weed Spotters recruited Figure 3. Weed Spotters by method of recruitment for 2006/07. through this approach (Figure 3). Weed Spotters who attend introductory training sessions learn about the purpose of the (eight), garden club presentations (two) Introductory training network, how it operates, what their com- and letters (one). There is also a signifi - only, 243 mitment involves and are given the initial cant number that are unknown which may skills to become an active Weed Spotter. have been recruited through a combination Attendees are also given a Weed Spotter of all methods. This data will be collected Introductory and handbook and a WEEDeck at introductory more accurately in the future through a re- secondary training, 66 training sessions. cently developed Weed Spotter brochure. Secondary training The second most effective method of only, 36 recruiting Weed Spotters during 2006/07 Weed Spotter training Weed Alert was Weed Alert presentations. These are Weed Spotters have been trained to vari- Untrained, 680 presentation, 110 often similar to introductory Weed Spotter ous levels as illustrated in Figure 4. Those training in the content presented, however who have attended both an introductory Figure 4. Number of trained Weed the difference is that a Weed Alert speaker and secondary training session are more Spotters. is invited to present to a pre-existing group highly trained in how to identify and re- that is usually meeting for a purpose other port weeds than those who have only at- than Weed Spotters. For example, Landcare tended a Weed Alert presentation or reg- meetings or University/TAFE lectures. istered their details without receiving any The topic of the presentation is related to training. It is important to encourage all Weed Spotters (39) than other individu- an aspect regarding weeds and contains Weed Spotters to attend training so that als (26), and only two of these have since some Weed Alert and Weed Spotter mate- they can be as effective as possible. registered as Weed Spotters. Discussions rial. This method is similar in effectiveness with WACOs indicate that this is partly to formal introductory training in terms of Weed Spotter skills due to some individuals preferring to numbers recruited; however these recruits The Weed Spotter engagement plan (McIn- keep their contact details confi dential and may not be as committed to the network, erney and Robinson, 2007) suggests that to not formally register with the network. as those that have intentionally given their recruitment efforts can be focused on tar- Others have a good relationship with the time to a session specifi cally designed for geted groups to strategically expand Weed WACO and do not feel that registration is Weed Spotters. Spotters based on requirements from the necessary. This issue has been addressed Weed Alert displays are usually held at network. For example, if there is a need for with the development of a Weed Spotter fi eld days, markets or shows and include more Weed Spotters with skills to locate brochure that includes a registration form. displaying live weed species and printed aquatic Weed Alert species then it is worth It is important that individuals who are material. Weed Spotters who register in recruiting from water authorities, Catch- already reporting Weed Alert species reg- this way are not exposed to as much in- ment Management Authorities (CMAs) ister as a Weed Spotter, so that they receive formation as those who attend dedicated and the WaterWatch program. Table 1 network information to become as effec- training or a presentation, and therefore shows the breakdown of Weed Spotters by tive as possible. may not be fully aware of their commit- affi liation category which is an indication State Prohibited Weeds (39) have been ment and may not be effective Weed Spot- of their weeds interest and skills. reported more often than Victorian Alert ters until they attend training. Weeds (14) and other species (14). This Secondary Weed Spotter training ses- Weed Spotter surveillance and reporting may be due to more readily available in- sions are subsequent training modules that activities formation on State Prohibited Weeds in occur after the introductory session to im- Weed Spotter reports from across the State comparison to Victorian Alert Weeds. For prove the skills of existing Weed Spotters. and all other Weed Alert species reports example, many of the fact sheets and the The purpose is not necessarily to recruit received by WACOs and other Weed Alert WEEDeck which are made available to new Weed Spotters, which is consistent team members during 2006/07 have been Weed Spotters mainly cover State Prohibit- with the trend presented in Figure 3. compiled. Figure 5 shows the number of ed Weeds. It is expected that as Weed Spot- The ‘other’ category is made up of re- reports received and indicates the break- ters receive more training, improve their cruitment via personal communications down according to class. identifi cation skills and more materials on (19), newspaper articles (nine), meetings Information collected suggests that Victorian Alert Weeds are produced, the (nine), the DPI Customer Service Centre more reports were made by registered pattern shown in Figure 5 may change.

Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 117 Table 1. The breakdown of Weed Spotters by affi liation category. The number of look-alikes (Figure 5) re- ported is interesting and indicates a need Affi liation Number of Weed Spotters to ensure information on similar looking 30 June 07 plants is available for both State Prohib- Department of Primary Industries 151 ited and Victorian Alert Weeds. Farming (including Landcare) 93 Conclusion Education 87 The consistent collection and storage of Conservation Group 69 all Weed Spotter network data is essential for accurately determining whether Weed Local Government 68 Spotters are being recruited strategically Department of Sustainability and Environment 53 and whether registered Weed Spotters are reliably monitoring and reporting Weed Waterwatch 43 Alert species. The results from the fi rst Horticulture 35 year of network data collection are prom- Botanic Gardens 24 ising, indicating that recruitment is well underway and that accurate Weed Alert Consultant/Contractor 24 reports are being submitted through the Catchment Management Authority 21 WACOs. The information will be useful as Parks Victoria 19 a benchmark for tracking the Weed Spotter network into the future. Water Authority 17 It is recommended that the Weed Alert Other 16 program should: • Implement the Weed Spotter Engage- Unknown 454 ment Plan (McInerney and Robinson Total 1174 2007), • Encourage all new Weed Spotters to complete the questionnaire in the Weed Spotter brochure, 25 23 • Encourage all Weed Spotters to formal- Registered Weed Spotter reports ly register, 20 Unregistered Weed Spotters • Make information available to Weed Number of species Spotters on Victorian Alert Weeds, 16 • Make information available to Weed 15 Spotters on Weed Alert species look- 10 10 alikes, and 10 9 9 • Collect evaluation information at all 5 5 Weed Spotter training events via the 44 5 3 evaluation survey and incorporate re- 2

Weed Spotter Reports/Species sults into program development and 0 improvement. State Prohibited Victorian Alert Weeds look-alikes non target species Weeds References Weed Alert species McInerney, C.J. and Robinson, V. (2007). Weed Spotter Engagement Plan. De- partment of Primary Industries and Figure 5. Weed Alert species reports made by registered versus unregistered Department of Sustainability and Envi- Weed Spotters including the number of different species reported. ronment, Melbourne.

118 Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 Orbost in Gippsland. Another two sites were identifi ed in dams in the Port Phillip Water hyacinth illegal trade in Victoria: the trace back region in 1994 and 2000. In more recent and forward of a State Prohibited Weed times, water hyacinth has been found in the Botanical Gardens in Williamstown (2002), Darebin Creek (2003), an urban Ryan Melville, Department of Primary Industries – Geelong, PO Box 103, pond system in Blackburn (2005) and Eu- Geelong, Victoria 3220, Australia. roa (late 2006). All infestations are continu- ally monitored by Weed Alert staff. Most locations have shown no signs of water hyacinth following initial or secondary Summary A Victorian Weed Spotter re- particularly State Prohibited Weeds and treatment (DPI 2007). Historically water ported the sale of several water hyacinth Victorian Alert Weeds. hyacinth has been associated with the plant plants (Eichhornia crassipes) on the internet trade, either at nurseries or the aquatic fi sh auction house eBay. Compliance staff from Background trade, although in recent years the nursery the Department of Primary Industries’ Under the CaLP Act 1994 a ‘noxious weed’ industry has shown leadership in remov- Weed Alert team alerted eBay of the plants, is defi ned as any of the following; a State ing these and other State prohibited weeds as the seller was located in Victoria. prohibited weed, a regionally prohibited from sale. Offi cers visited the seller’s residence weed, a regionally controlled weed or a prior to the end of the auction and seized restricted weed. The defi nition of a nox- Discussion all the water hyacinth plants. Through ious weed is important because under the During January 2007, whilst searching for questioning, the offi cers traced the plants CaLP Act 1994 it states in section 71 that items on the internet auction house eBay, back to a previous online auction, where ‘A person must not without a permit from a Weed Spotter noticed a Victoria resident the seller was once again registered in Vic- the Secretary sell or offer to sell in Victo- auctioning water hyacinths. The Weed toria. Offi cers then visited the initial sell- ria a noxious weed’. An offence against Spotter was aware that water hyacinth er and seized additional water hyacinth this section of the Act carries a maximum was a State prohibited weed and referred plants and discovered that plants had penalty of $12,000 per offence (State of Vic- the matter to a Weed Alert Offi cer based been sold to 11 other people from around toria 2006). at Frankston. Australia; one from Queensland, two from Of the four categories of noxious weeds After preliminary investigations the New South Wales, two from South Aus- in Victoria, the highest classifi cation is Departmental Offi cers contacted eBay to tralia, one from Tasmania and fi ve from State prohibited, which is classifi ed for the alert them of a possible offence and to Victoria. entire state of Victoria. A species can only gather intelligence. The following day, of- Officers followed-up by visiting the be considered for classifi cation as a State fi cers visited the sellers Edithvale property residence of each purchaser in Victoria. prohibited weed if it is either not present to ascertain whether the plants were in fact Agencies in other states were alerted and in Victoria, or present in such low num- water hyacinth and begin the process of provided with purchaser information. bers that it can be eradicated from the state tracing the origin of the plants. It was evi- Two court matters relating to the offences (State of Victoria 2006). At present Victo- dent upon inspection that the plants were detected under the Catchment and Land ria only has 25 State Prohibited Weeds, water hyacinths. The seller was a private Protection Act 1994 where heard against one of which is water hyacinth Eichhornia individual seller and not a commercial the two sellers, both resulting in fi nes. crassipes (Martius) Solms-Laub. enterprise, who had bought the plants on Keywords: Catchment and Land Protec- Water hyacinth was one of the very eBay the previous year. DPI’s inspection tion Act 1994, water hyacinth, Eichhornia few species that was declared a noxious occurred prior to the conclusion of the crassipes, eBay, Weed Alert, State prohib- weed in every state prior to the inception auction, and any plants being sold. ited weeds. of Weeds of National Signifi cance (Sainty The Edithvale seller had purchased and Jacobs 2003), and is recognised as one one water hyacinth plant, less than a year Introduction of the worlds worst water weeds (Hus- prior to detection and possessed 41 mature Within the state of Victoria the regulation sey et al. 2007). Water hyacinth is a major water hyacinths on the day of inspection, of noxious weeds is administered by the weed world wide, having originated from not including plantlets attached to parent Department of Primary Industries’ (DPI) South America, and is primarily spread as plants. All of these plants were seized by Landscape Protection branch. The Catch- an ornamental plant because of its showy Departmental Offi cers. ment and Land Protection Act 1994 (CaLP lilac fl owers (Parsons and Cuthbertson Further investigations into the origin of Act 1994) provides guidance on prohibited 2001). This species has been known to the water hyacinth plants revealed that the activities regarding noxious weeds. Pro- multiply from a few plants to covering a original seller was from Seaford, Victoria. hibited activities include the planting or 30 meter dam in one season (Sainty and Offi cers contacted the Seaford seller and propagation, display, movement, spread Jacobs 2003). on inspection of the residence seized a fur- and sale of noxious weeds or any part that Water hyacinth was fi rst introduced ther 60 plants. The Seaford seller had pre- is capable of growing which includes the into Australia in the 1890s, with introduc- viously sold 39 plants to 12 people around seeds. tions into Victoria were prior to the 1900s, Australia during four Dutch auctions. Buy- Within the Landscape Protection however no long term naturalised infesta- ers included one from Queensland, two branch of DPI, the Weed Alert program tions have persisted in Victoria (Parsons from New South Wales, two from South focuses on preventing the introduction and Cuthbertson 2001). Over the last 20 Australia, one from Tasmania and six from of serious new weeds into Victoria and years, most Victorian water hyacinth in- Victoria (including the Edithvale seller). eradicating the highest risk incursions. festations have either been found in resi- The Seaford seller was also a private seller The Weed Alert program has established dential garden ponds or at nurseries and and not a commercial business. a voluntary surveillance network in the pet stores. There have only been a small The description of the plants on the community known as the Weed Spotters number of naturalised infestations discov- Seaford seller’s auction page stated that network. Weeds Spotters provide infor- ered, including one infestation detected in the plants could not be traded to West- mation and intelligence to DPI regard- 1987 at Lochend and another at Brodribb ern Australia or Northern Territory and ing unusual plants in the community, in 1992, both of which are located close to that ‘It is not recommended for dams and

Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 119 waterways, as it can get out of control, but Primary Industries to build awareness of for your pond or water feature it looks fan- the threat that these and other State pro- tastic, and can be easily divided’. hibited weeds pose to the environment Following this inspection, each of the and agricultural industries. The Weed Victorian purchasers were inspected by Spotter network will also provide an im- Departmental Offi cers. Of the remaining portant early detection function enabling fi ve purchasers, two no longer had plants, the Department to activate an early treat- due to the plants being disposed of or dy- ment to achieve eradication. ing. A Mornington purchaser who bought three plants still had two remaining, a Acknowledgements Hastings purchaser had 27 plants from The author would like acknowledge the two original plants purchased and a Gee- work and support of members of the Weed long purchaser had 67 plants after buying Alert in undertaking this investigation 12 plants less than 12 months prior. and the assistance of the Weeds Spotters Agencies in other states were alerted to network. The author would also like to the detection of water hyacinth being im- thank many other Departmental staff for ported into their states and were provided their invaluable input and helpful com- with purchaser information. Investigators ments. in Queensland also found water lettuce (Pistia stratiotes L.) and salvinia (Salvinia References molesta D.S.Mitchell) which are declared DPI (2007). Intergraded Pest Management plants in that state, at the same residence System. of the water hyacinth purchaser. Hussey, B.M.J., Keighery, G.J., Dodd, J., Following these investigations, the two Lloyd, S.G. and Cousens, R.D. (2007). Victorian sellers were both charged under ‘Western weeds: a guide to the weeds the CaLP Act 1994 for ‘without a permit of Western Australia’, Second edition, from the Secretary sell or offer to sell in p. 70. (The Weeds Society of Western Victoria a noxious weed’ (State of Victoria Australia, Victoria Park, WA). 2006). The Edithvale seller pleaded guilty Parsons, W.T. and Cuthbertson, E.G. (2001). in Melbourne County Court and was fi ned ‘Noxious weeds of Australia’, second with costs, whilst the Seaford seller plead- edition, pp. 139-44. (CSIRO publishing, ed guilty in the Frankston Magistrates Melbourne). Court and also received a fi ne with costs. Sainty, G.R. and Jacobs, S.W.L. (2003). ‘Wa- terplants in Australia.’ pp. 6-7. (Sainty Comments and Associates Pty Ltd, Potts Point). Following this case and other similar in- State of Victoria (2006). Catchment and vestigations in other states, eBay has since Land Protection Act 1994, Act No. stated publicly that it has changed its pol- 52/1994, pp. 5-93. (Anstat Pty Ltd, Mel- icy on the sale of noxious weeds and that bourne). if a plant species is declared in any of the Australian states or territories, it will not be permitted to be listed. Despite the best efforts of Weed Alert staff attempting to eradicate State prohib- ited weeds such as water hyacinth from Victoria, infestations are still being found. In the six months following the commence- ment of this investigation, three separate water hyacinth infestations have been dis- covered. Given the ornamental appeal of this species, its distribution through urban garden ponds may be much greater than original estimates.

Conclusion This investigation into the trade of water hyacinth in Victoria may have alleviated signifi cant potential for this plant to es- tablish within waterways throughout Victoria and other states where plants were exported. There may still however, be many more plants fl oating in garden ponds in backyards around the state, wait- ing for someone to throw a plant over their fence into the adjoining drain, creek, dam or river that could potentially devastate Victoria’s water resources. Further educa- tion, awareness and compliance programs are being delivered by the Department of

120 Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 Seed bank spatial dynamics and plant invasions in Victoria’s western basalt plains grasslands

A.J. DoddA,B, N.S.G. WilliamsC,D and M.J. McDonnellC A School of Botany, The University of Melbourne. B Current address: Department of Primary Industries, Bacchus Marsh, Victoria 3340, Australia. C Australian Research Centre for Urban Ecology. D Current address: School of Resource Management, The University of Melbourne, Burnley, Victoria 3121, Australia.

Abstract The soil seed banks of six linear native discrete edge effect related to exotic species grasslands were studied to investigate the invasion. Instead, exotic species were abun- effects of landscape context on the spatial dant throughout the soil seed bank. Simi- patterning of grassland seed banks. Study larity between the seed bank and the above sites were located along a 200 km urban ground vegetation was subsequently low, – rural gradient running west from Mel- and declined as sites became increasingly bourne, Victoria. Exotic annual graminoids rural. The removal of fi re as a disturbance dominated the composition of the soil seed mechanism from grasslands, along with a bank with native species contributing only change in the species composition of the 13% of the seeds present. In contrast to the surrounding landscape matrix, is likely to patterns identifi ed in the vegetation, anal- lead to increased weediness and similarity ysis of the soil seed bank did not detect a of native grassland communities.

Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 121 Practice change by Victorian linear reserve managers

Claire Norris, Department of Primary Industries – Geelong, PO Box 103, Geelong, Victoria 3220, Australia.

Linear reserves (rail and road reserves) Victorian Rail industry organisations to can provide a pathway for weed spread produce vegetation management guide- potentially impacting on thousands of lines for rail corridors, and various projects landholders in Victoria. on weed mapping, weed hygiene, staff Through Victoria’s Tackling Weeds on training and weed management trials by Private Land initiative’s Weed Manage- VicRoads and Municipal Councils. ment Grants, Victoria’s rail companies This poster will share learnings from and VicRoads have been actively improv- the initiative which can assist future prac- ing their capacity for long term, effective tice change programs aiming to reduce the weed management. This has produced im- economic and environmental impacts of proved partnerships developed between weeds on linear reserves.

Practice change within the Victorian fodder industry

Claire Norris, Department of Primary Industries – Geelong, PO Box 103, Geelong, Victoria 3220, Australia.

The Victorian fodder industry is a key projects to assist with minimising weed stakeholder in the challenge to minimise spread across the industry. These projects the spread of weeds. Through the Victori- include development of an industry Code an Tackling Weeds on Private Land initia- of Practice to minimise weed spread and tive’s incentive Weed Management grants a hay bale tagging device (AFIA), for- Program, The Australian Fodder Industry mal training of association members in Association (AFIA) and the Australian Ag- weed hygiene procedures (AACA) and ricultural Contractors Association (AACA) targeted distribution of weed spread have worked collaboratively with the De- risk information at times of wildfi re and partment of Primary Industries on several drought.

122 Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 List of attendees

Robin Adair, Department of Primary Industries, PO Box 48, Frankston Vic 3199 [email protected] Nigel Ainsworth, Department of Sustainability and Environment, PO Box 500, East Melbourne Vic 3002 [email protected] Cameron Alexander, Alpine Shire Council, PO Box 134, Bright Vic 3741 [email protected] Helen Anderson, Department of Primary Industries, Private Bag 105, Hamilton Vic 3300 [email protected] Joanne Anderson, Victorian Catchment Mgt Council, 88 Grahams Track, Lancefi eld Vic 3435 [email protected] Alex Arbuthnot, Victorian Catchment Mgt Council, Level 1, 8 Nicholson Street, East Melbourne Vic 3002 [email protected] Georgie Bailey, Department of Agriculture, Fisheries and Forestry, Edmund Burton Bldg, Barton ACT 2601 [email protected] Joel Benjamin, VicRoads, 60 Denmark Street, Kew Vic 3101 [email protected] Stephen Bitter, Bayer CropScience, Unit 1, 7 Charlotte Street, Newport Vic 3017 [email protected] David Blair, Shire of Yarra Ranges, 17 Kalamunda Terrace, Healesville Vic 3777 [email protected] Kate Blood, Department of Primary Industries, PO Box 7, Beaufort Vic 3373 [email protected] Julio Bonilla, Department of Primary Industries, PO Box 48, Frankston Vic 3199 [email protected] James Booth, City of Darebin, PO Box 91, Preston Vic 3072 Geoff Carr, Ecology Australia, 88B Station Street, Fairfi eld Vic 3078 [email protected] Martin Chatfi eld, South Gippsland Landcare Network, 240 Creamery Valley Road, Toora Vic 3962 [email protected] Hillary Cherry, DECC, Pest Management Unit, PO Box 1967, Hurstville NSW 1481 [email protected] Andrew Clark, Biosecurity Queensland, Level 3 Landcentre, Locked Bag 40, Coorparoo DC Qld 4151 [email protected] Daniel Clements, Department of Primary Industries, PO Box 48, Frankston Vic 3199 [email protected] Harry Combellack, SpraySmart, 7 Michelle Drive, Maiden Gully Vic 3551 [email protected] Tim Connell, City of Whittlesea, Locked Bag 1, Bundoora Vic 3083 [email protected] Roger Cook, Hume City Council, PO Box 119, Dallas Vic 3047 [email protected] Lyn Coulston, Upper Murray Blackberry Action, PO Box 21, Koetong Vic 3704 [email protected] Roger Cousens, University of Melbourne, Burnley Campus, 500 Yarra Boulevard, Richmond Vic 3121 [email protected] Richard Curwell, VicRoads, PO Box 775, Geelong Vic 3220 [email protected] Jamie Davies, Department of Primary Industries, PO Box 48, Frankston Vic 3199 [email protected] Andrew Dawson, Hume City Council, PO Box 119, Dallas Vic 3047 [email protected] Sandra Dillon, RSD 8122, Ballarat Vic 3352 [email protected] Aaron Dodd, Department of Primary Industries, 219A Main Street, Bacchus Marsh Vic 3340 [email protected] Mike Dower, Parks Victoria, PO Box 206, Omeo Vic 3898 [email protected] Tony Dugdale, Department of Primary Industries, PO Box 48, Frankston Vic 3199 [email protected] Fiona Ede, Department of Primary Industries, PO Box 48, Frankston Vic 3199 fi [email protected] Ian Faithfull, Department of Primary Industries, PO Box 48, Frankston Vic 3199 [email protected] Jon Fallaw, Phillip Island Nature Park, PO Box 97, Cowes Vic 3922 [email protected] Mark Farrer, Department of Primary Industries, 23 Patrick Street, Stawell Vic 3380 [email protected] Simon Feillafe, North East Catchment Management Authority, 267 Mahon Lane, Osbornes Flat Vic 3691 [email protected] Penny Gillespie, Department of Sustainability and Environment, 3 Peronne Street, Bittern Vic 3918 [email protected] Charles Grech, Department of Primary Industries, 475 Mickleham Road, Attwood Vic 3049 [email protected] Michael Hansford, Department of Primary Industries, Locked Bag 3000, Box Hill Vic 3128 [email protected] Debbie Harvey, Upper Murray Blackberry Action, PO Box, Koetong Vic 3704 Rosemary Henderson, Protech Consulting, PO Box 851, Eltham Vic 3095 [email protected] Tony Herwerth, Shire of Melton, PO Box 21, Melton Vic 3337 [email protected] Colin Hocking, Institute for Sustainability and Innovation, Victoria University, St. Albans Vic 3021 [email protected] Peter Huhta, North East Catchment Management Authority, PO Box 616, Wodonga Vic 3689 [email protected] Trevor Hunt, Department of Primary Industries, PO Box 48, Frankston Vic 3199 [email protected] Tony Jarvis, Upper Murray Blackberry Action, 61 Ashstead Park Lane, Cudgewa Vic 3705 [email protected] Greg Johnson, Department of Primary Industries, PO Box 303, Wodonga Vic 3690 [email protected] Karen Jones, Rural City of Wangaratta, PO Box 238, Wangaratta Vic 3676 [email protected] Daniel Joubert, Department of Primary Industries, PO Box 48, Frankston Vic 3199 [email protected] Geoff Keech, Macspred Australia, 714 Duri-Dungowan Road, Timbumburi NSW 2340 [email protected] Maree Keenan, City of Greater Dandenong, 20 Bennett Street, Dandenong Vic 3175 [email protected] Tim Kelly, River Restorations, PO Box 473, Guildford Vic 3451 [email protected] Lincoln Kern, Practical Ecology P/L, PO Box 228, Preston Vic 3072 [email protected] John Kershaw, Ecology Australia, 88B Station Street, Fairfi eld Vic 3078 [email protected] Catriona King, Department of Primary Industries, PO Box 879, Seymour Vic 3660 [email protected] Chris Knight, Land Mangement Systems, PO Box 1146, Mitcham North Vic 3132 [email protected] Darren Kriticos, Forest Biosecurity and Protection, Ensis, PO Box E4008, Kingston ACT 2604 [email protected] Annie Lamb, Department of Primary Industries, PO Box 48, Frankston Vic 3199 [email protected]

Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007 123 Stuart Lardner, Department of Primary Industries, 110 Natimuk Road, Horsham Vic 3400 [email protected] Emily Lee, Department of Primary Industries, PO Box 48, Frankston Vic 3199 [email protected] Tony Lovick, Department of Primary Industries, 89 Sydney Road, Benalla Vic 3672 [email protected] Chris Malins, Department of Primary Industries, 110 Natimuk Road, Horsham Vic 3401 [email protected] Bill Mallison, Royal Botanic Gardens Cranbourne, 1000 Ballarto Road, Cranbourne Vic 3977 [email protected] Simon Martin, Department of Primary Industries, 402-406 Mair Street, Ballarat Vic 3350 [email protected] Rachel McFayden, Weeds CRC, Block B, 80 Meyers Rd, Indooroopilly Qld 4068 [email protected] Chris McGrath, Department of Primary Industries, 26 Wellington Street, Kerang Vic 3579 [email protected] Catherine McInerney, Department of Primary Industries, PO Box 103, Geelong Vic 3220 [email protected] Richard McKinnon, VicRoads, PO Box 775, Geelong Vic 3220 [email protected] Fiona McPherson, NSW Department of Primary Industries, PMB 2, Grafton NSW 2460 fi [email protected] Ryan Melville, Department of Primary Industries, PO Box 103, Geelong Vic 3220 [email protected] Lyn Meredith , Manningham City Council, PO Box 1, Doncaster Vic 3108 [email protected] Bronwyn Merritt, Royal Botanic Gardens Cranbourne, 1000 Ballarto Road, Cranbourne Vic 3977 [email protected] Lisa Minchin, 8/9 Kooyong Road, Caulfi eld North Vic 3161 [email protected] Brett Mitchard, Department of Primary Industries, PO Box 48, Frankston Vic 3199 [email protected] Alex Moodie, City of Greater Dandenong, 20 Bennett Street, Dandenong Vic 3175 [email protected] Tim Nitschke, Goulburn-Murray Water, PO Box 264, Kerang Vic 3579 [email protected] Paul Norquay, University of Ballarat, 1027 Yendon-Egerton Road, Mt Egerton Vic 3352 [email protected] Claire Norris, Department of Primary Industries, PO Box 103, Geelong Vic 3220 [email protected] Sarah Partington, Department of Primary Industries, 219A Main Street, Bacchus Marsh Vic 3340 [email protected] Charlie Pascoe, Parks Victoria, PO Box 20, Bright Vic 3741 [email protected] Daniela Pascuzzo, Hume City Council, PO Box 119, Dallas Vic 3047 [email protected] Michael Reid, Department of Primary Industries, 1 McKoy Street, Wodonga Vic 3690 [email protected] Stuart Roberton, Department of Primary Industries, 1 McKoy Street, Wodonga Vic 3690 [email protected] James Robinson, Royal Botanic Gardens Cranbourne, 1000 Ballarto Road, Cranbourne Vic 3977 [email protected] Kristy Roche, Department of Primary Industries, Cnr Taylor Street and Midland Highway, Epsom Vic 3556 [email protected] Rick Roush, University of Melbourne, Parkville Campus, Land and Food Resources, Parkville Vic 3052 [email protected] Leeanne Saffron, Latrobe City Council, PO Box 64, Glengary Vic 3834 [email protected] Jean-Louis Sagliocco, Department of Primary Industries, PO Box 48, Frankston Vic 3199 [email protected] Luke Sandham, City of Darebin, PO Box 91, Preston Vic 3072 [email protected] Marianne Sawyer, Cardinia Shire Council, PO Box 7, Pakenham Vic 3810 [email protected] Amy Scott, Department of Primary Industries, PO Box 48, Frankston Vic 3199 [email protected] Barbara Setchell, 20 Stanley Street, Olinda Vic 3788 [email protected] Adam Shalekoff, City of Darebin, PO Box 91, Preston Vic 3072 [email protected] Ros Shepherd, Weed Society of Victoria, PO Box 987, Frankston Vic 3199 [email protected] Kelly Snell, Department of Primary Industries, PO Box 103, Geelong Vic 3220 [email protected] Val Stajsic, National Herbarium of Victoria, Royal Botanic Gardens, Birdwood Ave, South Yarra Vic 3141 [email protected] Jackie Steel, Department of Primary Industries, PO Box 48, Frankston Vic 3199 [email protected] Matt Stephenson, Victoria Landcare Network, PO Box 272, Cowes Vic 3922 [email protected] Norm Stone, Bayer CropScience, 14 Dorothy Street, Croydon Vic 3136 [email protected] Jason Summers, Hume City Council, PO Box 119, Dallas Vic 3047 [email protected] Liliana Surace, Department of Primary Industries, PO Box 48, Frankston Vic 3199 [email protected] Drew Sutton, Department of Primary Industries, 574 Main Street, Bairnsdale Vic 3785 [email protected] Elspeth Swan, Department of Sustainability and Environment, 402-406 Mair Street, Ballarat Vic 3350 [email protected] Brad Tadday, Manningham City Council, PO Box 1, Doncaster Vic 3108 [email protected] Nicola Thomson, WWF Australia, PO Box 528, Sydney NSW 2001 [email protected] Tamara Threlfall, Department of Primary Industries, 475 Mickleham Road, Attwood Vic 3049 [email protected] Mark Toomey, 8/18A Bloomfi eld Road, Ascotvale Vic 3032 [email protected] Simon Trill, Urbanvirons Group Pty Ltd, 230 Grange Rd., Flinders Park SA 5025 [email protected] Marie-Claire van Besouw, Marie-Claire’s Eco-Works, PO Box 7328, Karingal Vic 3199 [email protected] Alison Vaughan, National Herbarium of Victoria, Royal Botanic Gardens, Birdwood Ave, South Yarra Vic 3141 [email protected] Eugene Vereshanka, Parks Victoria, PO Box 206, Omeo Vic 3898 Bridie Wetzel, Hume City Council, PO Box 119, Dallas Vic 3047 [email protected] Jeremy White, Department of Primary Industries, 402 Mair Street, Ballarat Vic 3350 [email protected] Nicholas Williams, University of Melbourne, Burnley Campus, 500 Yarra Boulevard, Richmond Vic 3141 [email protected] Kylie Woods, Department of Primary Industries, 12 Peart Street, Leongatha Vic 3953 [email protected] Ken Young, University of Melbourne, Dookie College, 14 Ridge Street, Dookie Vic 3647 [email protected] Stephen Young, Department of Primary Industries, 402-406 Mair Street, Ballarat Vic 3350 [email protected]

124 Weed Society of Victoria Third Biennial Conference ‘Earth Wind Fire Water and Weeds’ 3–4 October 2007