26 Plant Protection Quarterly Vol.15(1) 2000 to rear F1 generation offspring failed. It was decided to concentrate time and The release and establishment of two biological funds on other more promising biological control agents of horehound ( L.) control agents. in south-eastern Australia Horehound seed beetle (Meligethes rotroui Easton) Emma Wills, Keith Turnbull Research Institute, Department of Natural Meligethes rotroui feeds on the flower and Resources and Environment, CRC of Weed Management Systems, PO Box 48, pollen of horehound and therefore reduce Frankston, Victoria 3199, Australia. the number of seeds produced. In April 1997 M. rotroui adults collected from Mo- rocco were imported into quarantine at KTRI to initiate a culture for host Summary Introduction specificity testing. Unfortunately all in- In 1991 a survey of the native range of In 1990 to 1992 scientists from the Com- sects were dead on arrival and no further horehound, Marrubium vulgare L. in Eu- monwealth Scientific and Industrial Re- shipments are presently planned. rope was conducted to identify potential search Organisation (CSIRO) Biological biological control agents for Australia. Control Unit in Montepellier, France sur- Horehound plume ( Four were chosen for closer inves- veyed horehound populations in southern spilodactylus (Curtis)) tigation; a plume moth, Wheeleria France for phytophagous insects that (Figure 1) is a spilodactylus (Curtis), a skipper butter- could be used as biological control agents. multivoltine that has up to four gen- fly, Carcharodus boeticus Rambur, a Surveys were also carried out on roadside erations per year and is active from spring clearwing moth, Chamaesphecia mysini- populations in Spain, Portugal, Morrocco, to autumn. The larvae (Figure 2) feed on formis (Boisduval) and a seed beetle, Italy and Yugoslavia (Sagliocco 2000). A the leaves and developing tips of hore- Meligethes rotroui Easton. Host specific- total of 27 different insect species were hound, which weakens the plant and re- ity testing of W. spilodactylus began in found feeding on horehound and of these duces the numbers of flowers and seeds 1991 at Keith Turnbull Research Institute four were chosen to be investigated in produced. In November 1991 W. spilo- in Frankston, Victoria and in 1993 it was more detail. Two of these insects have suc- dactylus was imported into quarantine at approved for release. In 1994 host cessfully established at sites throughout KTRI for host specificity testing. Over the specificity of C. mysiniformis was per- south-eastern Australia. next two years the insect was tested on 56 formed in France and approval for re- plant species (Weiss et al. in press). This lease was gained in 1996. Initial attempts Results testing demonstrated W. spilodactylus was to begin host specificity testing for C. Horehound skipper butterfly host specific to horehound and in Decem- boeticus and M. rotroui failed, and future (Carcharodus boeticus Rambur) ber 1993 it was approved for release in attempts have been halted. In the spring Carcharodus boeticus has two generations Australia. Preliminary releases of W. of 1998 a redistribution program for W. per year and is a voracious nocturnal spilodactylus were made at Wyperfeld Na- spilodactylus was initiated. A total of 32 defoliator that specializes in the bud re- tional Park and Swifts Creek in Victoria, release sites in Victoria, New South gion of young shoots. The larvae build a Murray Bridge in South Australia and Wales, South Australia and Tasmania nest by weaving several leaves together Tamworth in New South Wales. From were visited to determine establishment with silk thread, to protect themselves 1993 until 1996 the rearing and releasing of W. spilodactylus and the potential for from predators. Larvae are nocturnal, of W. spilodactylus was primarily per- redistribution of insects from the sites. feeding on horehound leaves and return- formed in Victoria. In October 1996 the At 90% of the sites visited the biological ing to the nest during the day. In 1994 and Cooperative Research Centre (CRC) for control agent had established and of 1995 C. boeticus was imported into quar- Weed Management Systems began rear- those, 41% were at a stage that the collec- antine at Keith Turnbull Research Institute ing W. spilodactylus at Adelaide University tion and redistribution of W. spilo- (KTRI) for host specificity testing. How- and in September 1997 the Tasmanian In- dactylus to new sites was possible. ever mating conditions were difficult to stitute of Agricultural Research, Hobart, replicate in quarantine and two attempts also began a rearing program. As a result

Figure 1. Adult horehound plume moth, Wheeleria Figure 2. Larva of the horehound plume moth, spilodactylus. Wheeleria spilodactylus. Plant Protection Quarterly Vol.15(1) 2000 27 over 100 000 insects have been released at synchronization of lifecycles present at • Cage. Similar to tent releases except the over 100 sites in the four south-eastern the start of the season. Pupae are more cages only cover one or two plants and states of Australia. difficult to locate on the plants and are used in areas where the horehound stems containing large larvae also con- is scattered. Monitoring and redistribution of W. tain eggs and small larvae that will per- • Container. Pupae are put in a container spilodactylus ish once the stem is cut. that protects them from the elements In spring 1998 a travel grant from CRC for Several methods of releasing W. spilo- and/or predators yet allows the adults Weed Management Systems enabled a dactylus have been trialed over the years. to escape once they emerge. large number of W. spilodactylus nursery • Tent. Pupae and adults are released in- • Free. Larvae, pupae and/or adults are sites in south-eastern Australia to be side 2 × 3 metre by 2 metre high nylon placed on horehound plants in the monitored for insect establishment, distri- shadecloth tents. The insects build up open. bution and suitability for use as harvest- large numbers in a small area while Table 2 shows the different release meth- ing sites for redistribution (Figure 3). protected from predators and the ele- ods used for the 32 sites monitored in 1998 At each site: ments. This was useful for sites with in comparison to insect establishment and • The presence or absence of W. spilo- scattered horehound plants or in dry density (suitability for redistribution). No dactylus was monitored. areas as the tents keep the plants particular release method was more effec- • The distribution and density of the in- greener and healthier than surround- tive for the establishment of W. spilo- sect if present, was measured. ing plants. Tents are removed after the dactylus however it appears that free re- • Site managers were trained in insect insects have completed at least one leases produced populations with higher identification, monitoring and redistri- lifecycle (four weeks later). densities. bution methods. As an operational guideline, a minimum of 200 W. spilodactylus insects is required to set up a new site, redistribution was ➝ only recommended if it was considered N that the original nursery site would not suffer from insect removal. A total of 32 sites in Victoria, South Australia, Tasma- nia and New South Wales were monitored in 1998. At 29 of the sites W. spilodactylus had established and 12 contained populations that would sustain harvest- ing for redistribution (Table 1). There are several methods possible for Key • nursery site harvesting W. spilodactylus from field sites. • 4 number of nursery i. Collection of larvae-infested hore- sites at this locality hound stems in early spring, when lar- vae have recently started to feed and develop again after overwintering as 100 0 100 200 kilometres first instar larvae in the tips of the plant. During this time the population Figure 3. Wheeleria spilodactylus nursery sites monitored in spring 1998 in is synchronized and larvae are at south-eastern Australia. roughly the same stage of develop- ment. Redistribution is recommended when the larvae reach the 4th or 5th Table 1. Summary of evaluation of W. spilodactylus nursery sites monitored instar, prior to pupation. Plant stems in spring 1998. containing larvae are simply cut at the State No. of sites No. of sites No. of sites base and layed across a fresh plant at surveyed with establishment suitable for harvesting the new site. The larvae will crawl onto for redistribution the new plant as the stem dies. It should be noted that stems will wilt Victoria 15 13 5 and die very quickly after being cut so South Australia 11 11 5 they must be stored in a cool container Tasmanian 5 5 2 and released within 24 hours. This New South Wales 1 0 0 method has been used successfully in Total 32 29 12 the field in Tasmania and in laboratory cultures. ii. Collection of pupae in mid spring Table 2. Method of W. spilodactylus release in relation to the establishment when the first wave of larvae have and suitability for harvesting for redistribution at sites monitored in spring started to pupate. Leaves containing 1998. pupae are collected from the plants and sent or taken to a new site. This was the Release No. of sites No. of sites Establishment No. of sites original method of distributing W. method with establishment success (%) suitable for harvesting spilodactylus to new sites from the mass for redistribution rearing cultures. iii. Collection of larvae or pupae later in Tent 13 10 77 4 the season. This method is not as effec- Cage 3 3 100 0 tive or time efficient because the W. Container 4 4 100 0 spilodactylus population has lost the Free 12 12 100 8 28 Plant Protection Quarterly Vol.15(1) 2000

Figure 4. Adult horehound clearwing moth, Figure 5. Larva of the horehound clearwing moth, Chamaesphecia mysiniformis. Chamaesphecia mysiniformis.

Horehound clearwing moth Monitoring of C. mysiniformis work to do before it makes an impact on (Chamaesphecia mysiniformis Monitoring of C. mysiniformis is difficult as the horehound populations in south-east- (Boisduval)) the larvae feed in the roots of horehound ern Australia. Chamaesphecia mysiniformis (Figure 4) is a for around 10 months before emerging as univoltine moth that attacks the roots of adults in the summer. The root of the plant Acknowledgments horehound (Figure 5). The moth can cause must be cut open to view the larva, which This project is partly funded by the Victo- mortality of the plant by disrupting vas- kills the plant, and subsequently the larva. rian Department of Natural Resources and cular flow and indirectly by allowing sec- In January 1998 larvae were recovered Environment, the South Australian Ani- ondary infection by pathogens (Sagliocco from horehound roots at Wyperfeld Na- mal and Plant Control Commission and and Coupland 1995). In April 1994 host tional Park where 150 eggs had been re- was previously funded by the Woolmark specificity testing of C. mysiniformis began leased one year earlier. A month later 800 Company. at the CSIRO European Biological Control eggs were released at the same site and in Unit. After host specificity testing on 60 May 1999 the progress of the site was visu- References different plant species in 1995, it was ap- ally monitored. Frass from larval feeding Sagliocco, J-L. (2000). The insect fauna as- proved for release into the Australian en- in the root was found at the base of a large sociated with horehound (Marrubium vironment. Between 1995 and 1998 a mass number of horehound plants at the site. vulgare L.) in western Mediterranean rearing culture was attempted at KTRI, Destructive sampling of plants at the site Europe and Morocco: potential for bio- however, insufficient emerging adults is to occur in February 2000 to determine logical control in Australia. Plant Pro- were produced due to difficulties in main- if the larvae pupated and emerged as tection Quarterly 15, 21-5. taining prolonged plant health and vigour adults to confirm establishment. A larva Sagliocco, J-L. and Coupland, J.B. (1995). in quarantine conditions. In 1998 the Aus- has also been recovered from a plant at a Biology and host specificity of Chamae- tralian Quarantine Inspection Service South Australian site in October 1999. sphecia mysiniformis (: granted permission for modified non- Sesiidae), a potential biological control quarantine rearing of C. mysiniformis. In Discussion agent of Marrubium vulgare (Lamiaceae) the same year, staff from the CRC for Ideally, eight years since the initial survey in Australia. Biocontrol Science and Tech- Weed Management Systems at Adelaide for insects on horehound in Europe, at nology 5, 509-15. University also set up a C. mysiniformis least three biological control agents would Weiss, J., Lippai, K. and Price, T.V. (In culture. The rearing of both of these cul- now be feeding on horehound press) The biology and host specificity tures proved successful and releases were populations in south-eastern Australia. of the horehound plume moth, Wheel- made in South Australia and Victoria in Unfortunately many factors affect the suc- eria spilodactylus (Curtis) (Lepidoptera: 1999. However, in order for C. mysini- cess of a biological control agent, either at ), a biological control formis to establish at a site, several hun- the collecting, transporting, testing, rear- agent for horehound, Marrubium vul- dred eggs are required. As a result only ing or releasing stage. Once released in the gare L. Australian Journal of Entomology. four sites have been set up (Wyperfeld field many more factors can affect the es- National Park in Victoria and Nurrung, tablishment and success of a biological Disclaimer Monarto and Wilpena in South Australia). control agent; including climate, preda- This publication may be of assistance to The final laboratory generation was tors and human influences. W. you but the State of Victoria and its em- reared in 1999 for release in early 2000. If spilodactylus is having a significant impact ployees do not guarantee that the publica- the opportunity for mass rearing again on some horehound infestations in north- tion is without flaw of any kind or is becomes available in the future it will be east Victoria and Tasmania and is show- wholly appropriate for your particular possible to collect C. mysiniformis adults or ing promise in South Australia. Prelimi- purposes and therefore disclaims all liabil- larvae in the roots from the release sites, nary monitoring of C. mysiniformis sites ity for any error, loss or other consequence rear them in the laboratory and again re- indicate establishment; however, as there which may arise from you relying on any lease as eggs. are so few sites, the insect has much more information in this publication.