18 Plant Protection Quarterly Vol.15(1) 2000

Horehound workshop

Proceedings of a workshop held at the Victorian Institute for Dryland Agriculture in Horsham April 19 and 20 1999. Organized by John Weiss, Emma Wills and Nigel Ainsworth of the Keith Turnbull Research Institute, Victorian Department of Natural Resources and Environment, and sponsored by the Cooperative Research Centre for Weed Management Systems.

Editor: E. Wills, Keith Turnbull Research Institute, Department of Natural Resources and Environment, PO Box 48, Frankston, Victoria 3199, Australia.

Comparative phenological studies of Horehound (): a comparison horehound were carried out at two sites in between European and Australian populations Europe (Cournonterral and La Crau in southern France) and two in Victoria

A (Wyperfeld National Park and Swifts John Weiss and Jean-Louis Sagliocco , Keith Turnbull Research Institute, Creek), Australia during the period of Victorian Department of Natural Resources and Environment, CRC for Weed 1991–1996. Growth, reproductive capac- Management Systems, PO Box 48, Frankston, Victoria 3199, Australia. ity, size, and soil seed banks were meas- A Previous address: CSIRO European Laboratory, Campus International de ured seasonally or biennially. Individu- Baillarguet, 34980 Montferrier sur Lez, Cedex, France. ally marked plants were identified on each visit and details recorded. Soil cores were collected to measure the soil seed bank. These were washed free of soil and Summary horehound seeds checked for viability. Horehound (Marrubium vulgare L.), a (Argentina, Chile, Peru, Uruguay etc.) and Surveys of the fauna of Europe noxious weed throughout southern Aus- in New Zealand (Weiss in press). It is now (Portugal, Spain, France, Italy and Yugo- tralia, was imported from its native naturalized over most of the pastoral ar- slavia) and Morocco were carried out dur- southern Europe during the 1800s. Hore- eas in Australia with rainfall greater than ing 1990–1996. The IIE, CNRS (Toulouse), hound in its native habitat shows weed 200 mm. It occurs in all Australian states potential, but is scattered and occurs in but is of most importance in Victoria and less dense infestations than in Australian South Australia where an estimated 26 infestations where the plant is larger and million hectares (as shown in Figure 2) are more vigorous. The seed production and infested with horehound (Lane et al. 1980, seed bank studies indicate that there is at Carter 1990) least a one hundred-fold difference be- Horehound leaves contain marrubin, a tween the two populations. At Victorian bitter alkaloid, which makes it unpalat- infestations, either space or water appear able for grazing . Horehound to be the limiting factor affecting sur- burrs not only are of nuisance value to vival and recruitment while sustained people, as they catch in clothing and herbivory pressure restricts horehound socks, but they also contaminate wool, re- in its native range. Climate modelling ducing the value of fleece. based on overseas populations indicates Horehound was possibly first intro- that horehound in Australia maybe close duced into Australia from Europe via a to achieving its full distribution. shipment of botanical plants sent by Sir The insect fauna of horehound has Joseph Banks. A record from Sir Joseph Figure 1. Distribution of horehound been examined and a total of twenty- Bank’s diary states that M. vulgare was in Europe. eight have been found feeding on sent to New South Wales on board the the plant in Mediterranean Europe. In ship Porpoise on 11 October 1798 (Frost comparison only a couple of oligo- 1993). It appears to have been introduced phagous insects are on horehound in for use as a garden herb, medicinal and Australia and in consequence four Euro- beer brewing purposes. First recorded as pean insects have been identified as suit- naturalized by the 1840s. able agents for biological control. Two of these have been approved and released Materials and methods into the Australian environment. Potential distribution of horehound in Australia was determined using CLIMEX Introduction for Windows version 1.1 (Sutherst 1996). Horehound (Marrubium vulgare L.) is a The distribution was based on the pro- perennial weed native to temperate vided ‘Mediterranean template’ and the Eurasia, Europe, the Middle East and the parameters for temperature, moisture and Mediterranean region including North stress were expanding upon utilizing Africa (Figure 1). It is also a weed in south- Australian and overseas information Figure 2. Present distribution of ern states of north America, including about stress limits and the distribution of horehound in Australia (from California and Texas, in South America horehound. Parsons and Cuthbertson 1992). Plant Protection Quarterly Vol.15(1) 2000 19 MNHM (Paris) and British National His- Discussion seedlings. The extreme low level of seeds tory Museum identified collected insects. Distribution in Australia and Europe in the soil in Europe is most likely due to The present and potential distribution predation and destruction by the foraging Results maps of horehound in Australia (Figures ant species. Potential distribution 2 and 3) are very similar indicating that Horehound in both Europe and at Figure 3 indicates the potential suitable horehound maybe close to reaching its Swifts Creek, Victoria, has consistent climatic areas for horehound to infest in maximum potential distribution in Aus- populations with very little seasonal or Australia. tralia. Horehound has the potential of in- yearly fluctuations. In comparison, hore- creasing in Western Australia and parts of hound in the drier more drought prone Plant phenology New South Wales, but in South Australia, Mallee at Wyperfeld, undergoes dramatic Table 1 presents a summary of the phenol- Victoria and most of New South Wales, it ‘boom–bust’ cycles. Mature plants die off ogy and demography of two populations will more likely increase the density of during severe droughts and the infesta- of horehound in France and two in Victo- present infestations rather than move into tions rely on the very large seed bank to ria. Horehound in southern France is new localities. recolonize when sufficient rainfall occurs. much smaller with only about one-third In Europe, horehound generally grows This presumably indicates that these areas the number of stems, and produce about on wasteland, on well-drained calcareous are close to the limits of horehound toler- half the seeds that the Victorian popula- souls. The persistence of permanent plant ance and ability to establish. Without the tions do. Most importantly though, the populations depends on low plant compe- large seed bank reservoirs deposited dur- density of plants in Victoria is about 20 to tition, which is favoured by sheep graz- ing more favourable years, and subse- 50 times more, 200 to 1000 times as many ing, and sufficient seed dispersal by ani- quent recruitment, it is most likely that the seedlings occur and 70 to 250 times as mals. In western Europe, large horehound infestation would die out after a series of many seeds are in the soil seed bank than populations are very rare, but are found dry periods. Control strategies in these southern France. where sheep or horse grazing occurs. dry areas should concentrate on removing the seed bank, by fire for example (Weiss Insect fauna in Europe Horehound phenology 1996), or by reducing the seed production Twenty-eight species of insects, predomi- Although Australian plants are signifi- of the plants using biological control nantly from four orders were found on M. cantly larger and have more stems, the agents. vulgare in Europe and Morocco. Figure 4 amount of seed produced per plant is not show the relevant feeding behaviour of dramatically greater than that in Europe. Seed germination the European horehound insect fauna. The number of seeds produced per stem Lippai et al. (1996) reported on the effects One other insect, a rare weevil, and a leaf- is roughly equivalent in both popula- of temperature and water potential on the spotting fungus have been reported tions. However, as the density of plants is germination of Victorian horehound (Sagliocco and Weiss 1996) but were never 20–50 times more, this greatly increases seeds. They indicated that Australian found during the surveys. the seed bank and recruitment potential of horehound is not a drought tolerant ger- minator but more opportunist, germinat- ing whenever sufficient rains occur. In Detritivors contrast horehound seeds from America (1 sp.) Foraging require much more specific environmen- 4% ants (4 spp.) tal factors prior to germination including Sap-suckers 13% fluctuating temperatures of at least 15°C (5 spp.) Root borers or stratification. 17% (4 ssp.) This contrast is possibly due to envi- 13% ronmental differences the seeds are sub- jected to in North America and Australia, Stem borers which is substantiated by increased ger- (2 spp.) Seed mination following moist chilling of 7% feeders North American seeds, but not of Austral- (8 spp.) Leaf feeders ian seeds. It would be interesting to see 29% (5 spp.) what factors influence the germination of 17% European horehound. Anecdotal evidence Figure 3. Potential distribution of Figure 4. The different groups of indicates that horehound seeds survive in the soil seed bank for between 7 and 10 horehound in Australia (as insects recorded feeding on years (Weiss in press). predicted using CLIMEX® Version Marrubium vulgare in Europe. 1.1) Potential biological control agents In Australia prior to 1994 only a couple of Table 1. Demography and seed banks of horehound in Europe and (mites, mealy bugs and one Australia. bug) fed on horehound. The native brightly coloured orange and black hore- France Victoria, Australia hound bug, Agonoscelis rutila, is often seen Demography Cournonterral La Crau Swifts Creek Wyperfeld NP on the weed in great numbers, but does not have any significant impact. -2 Density of plants m 0.4 0.2 6.9 9.9 In comparison twenty eight species Mean height of plant (cm) 14.6 9.6 36.1 27.1 have been found to feed on M. vulgare in Mean no. of stems 6.3 6.8 35.2 18.9 Europe and Morocco. As areas such as Mean no. seed produced per plant 878 174 1746 498 Greece, the Middle East and other coun- Mean no of seed produced per stem 139 26 50 26 tries of northern Africa also rich in -2 Mean no. of seedlings m 0.11 0.32 69.5 196.3 Marrubium genera have yet to be investi- -2 Mean soil seed bank (seeds m ) 116.9 61.2 15 290 7880 gated, the total number of insects feeding 20 Plant Protection Quarterly Vol.15(1) 2000 on M. vulgare is expected to be much biological control agents for horehound in Resources and Environment and the higher. Australia is discussed by Sagliocco and South Australian and Plant Con- So far, through the surveys and the Wills in later papers in this workshop. trol Commission supported the Australian field observations, five insect species have research. been identified as having a potential for Conclusion the biocontrol of horehound in Australia. The reasons for the dense horehound in- References These species are: festations in southern Australia are most Carter, R.J. (1990). Biology and control of • spilodactylus (Curtis) likely a combination of factors. Hore- horehound, Marrubium vulgare. Pro- • Chamaesphecia mysiniformis (Boisduval) hound in Europe has few vectors to trans- ceedings of the 9th Australian Weeds • Meligethes rotroui Easton port the seed calyxes apart from the occa- Conference, p. 382-6. (Crop Science So- • Phytoecia melanocephala Fabricius sional sheep, while in Australia kanga- ciety SA Inc.). • Carcharodes boeticus Rambur roos, sheep, rabbits, farm machinery and Frost, A. (1993). ‘Sir Joseph Banks and the Two of these agents have been host watercourses all aid in the transport of the transfer of plants to and from the South specificity tested and approved for release seed to new areas. Because of different Pacific 1786–1798’. (Melbourne Colony into the Australian environment, the oth- land management techniques competition Press). ers are awaiting further studies. from other plant species maybe higher in Lane, D.W.A., Riches, K. and Combellack, The horehound plume , Wheeleria Europe than in Australia. Horehound J.H. (1980). A survey of the distribution spilodactylus, was first released in 1994 and seeds have a limited ability to germinate of noxious weeds in Australia. Dept of is now established at over 100 localities in conditions where moisture availability, Crown Lands and Survey, Frankston. throughout southeastern Australia (South or where temperatures are low. However, Lippai, A., Smith, P.A., Price, T.V., Weiss, Australia, Victoria, Tasmania and New mature plants can survive dry and J. and Lloyd, C.J. (1996). Effects of tem- South Wales). The caterpillar feeds on the drought conditions effectively as animals perature and water potential on germi- growing tips of the plants and then works preferentially graze more palatable pas- nation of horehound (Marrubium its way down the shoot, progressively de- ture species, reducing competition for vulgare) seeds from two Australian lo- foliating the stem. This weakens the plant available water. In Australia with the high calities. Weed Science 44, 91-9. and reduces the number of seeds and density of feral bees pollinating hore- Parsons, W.T. and Cuthbertson, E.G. flowers produced. Ainsworth discusses hound leading to prolific seeding, the high (1992). ‘Noxious weeds of Australia’, p. research on the impact of this moth in later germination rates, and the large seed bank 496-9. (Inkata Press, Melbourne). papers in this workshop. enable horehound to colonise newly dis- Sagliocco, J-L. and Weiss, J. (1996). The horehound clearwing moth, turbed areas and to recruit new seedlings Biocontrol of horehound (Marrubium Chamaesphecia mysiniformis, was released into the weed infestation when conditions vulgare L.). Work in Europe 1990–1996. in Victoria in March 1997. Larvae feed are favourable. In comparison, horehound Project DCV 14. Final Report to Inter- within the growing tissue of the root and in Europe is restricted by the high specific national Wool Secretariat. lower stems. Larval infection affects the herbivory pressure, which reduces the Sutherst, R.W. (1996). CLIMEX®: Predict- flow of water and nutrients through the ability of mature plants to produce a large ing the effects of climate on plants and plant, weakens it, reduces growth and in- quantity of seeds hence the seed bank is animals. Version 1.1. (CSIRO). creases the likelihood of the plant dying much reduced. The low level of specific Weiss, J.E.R. (1996). Control of hore- (especially when water stressed). As this herbivory in Australia by insects and dis- hound, Marrubium vulgare L., in Spanish insect emerges in early/mid sum- ease by pathogens is probably the major Wyperfeld National Park, Victoria. mer when daily temperatures exceed 30°C reason for the high level of infestation in Proceedings of the Eleventh Australian establishment of this agent may be re- Australia. Weeds Conference, p. 417-19. (Weed stricted to the more arid and warmer sites Science Society of Victoria Inc., in inland New South Wales and the Acknowledgments Frankston). Mallee parts of South Australia and Victo- The Australian Wool Research Develop- Weiss, J. (In press). ‘Best practise manage- ria. ment Corporation provided the funds ment guide: horehound (Marrubium The European horehound insect fauna for the European research of this project, vulgare)’. (CRC for Weed Management and the distribution and establishment of the Victorian Department of Natural Systems).