Ecological basis for selecting biocontrol agents for lantana

Michael D. Day1 and Alan J. Urban2

Summary Over the last century, more than 40 natural enemies have been released against the noxious weed lantana (Lantana camara L.) in over 40 countries or regions. Biocontrol of lantana remains inadequate, however, except on a few islands. Three of the main factors preventing adequate biocontrol of lantana are its unresolved parentage, resilience to established agents and climatic adaptability. These factors form the ecological basis for the current Australian–South African lantana biocontrol research project, which is tackling three main topics: (1) Host plants: the aim is to counteract the effects of the genetic heterogeneity of the weed by selecting agents from (a) the most probable parent species, (b) several species closely related to the probable parents or (c) ornamental cultivars growing in the native range of the probable parents. (2) Agent guilds: we aim to reduce the growth and reproductive vigour and resilience of lantana, by selecting agents that multiply quickly, or feed on the stems or roots. (3) Climatic adaptations: we seek to counteract lantana’s ability to grow in an extensive range of climatic conditions by selecting agents that can bridge periods of plant dormancy and/or leaflessness caused by cold and/or dry conditions. Candidate biocontrol agents, including pathogens and mites, selected on the basis of these ecological considerations, are currently being investigated and are showing considerable promise.

Keywords: climatic adaptation, ecology, genetic heterogeneity, guilds, Lantana.

Introduction over 40 countries or islands worldwide and 27 species established (Julien & Griffiths 1998, Baars & Neser Numerous ornamental forms of lantana were bred in 1999, Day et al., unpublished data) (Table 1). Despite glasshouses in Europe by selection and hybridization of some agents causing sporadic, localized damage, unknown, imported, parental species, probably lantana is still not under adequate control in most obtained from the New World (Stirton 1977). Of the regions. Day & Neser (2000) and Broughton (2000) over 600 named varieties of “Lantana camara L.” that suggested six factors that thwart adequate biocontrol of now exist, different aggregations became environ- lantana: mental and agricultural weeds in different countries. At 1. the plant species from which potential agents were least 40 recognizable weedy lantana varieties are collected present in South Africa, and 29 in Australia, threatening 2. the phenotype of the target plant ecosystem biodiversity and reducing pasture produc- 3. plant biology tivity (Howard 1969, Smith & Smith 1982, Graaff 4. the climate where agents were released 1986, Swarbrick et al. 1998). 5. parasitism of agents Biocontrol of lantana began in 1902, with the intro- 6. release techniques. duction of 23 species into Hawaii. Over the last To improve biocontrol of lantana, efforts have 100 years, 41 biocontrol agents were introduced into recently been made to select candidate agents that address three of these key factors: host plants, agent guilds and climatic adaptations. This paper discusses 1 Alan Fletcher Research Station, PO Box 36, Sherwood, Qld 4075, Australia. each these factors in relation to agent selection, and 2 Plant Protection Research Institute, Private Bag X134, Pretoria 0001, provides information on some of the candidate agents Republic of South Africa. currently being investigated that show considerable Corresponding author: Michael D. Day . promise.

81 Proceedings of the XI International Symposium on Biological Control of Weeds 7 9 1 1 0 0 0 0 1 0 0 1 6 1 0 1 1 6 2 2 0 0 0 0 0 2 2 1 0 1 3 0 15 24 24 29 15 12 No. of countries established in 1 1 2 1 5 2 2 4 3 2 1 3 1 2 7 2 1 5 1 1 1 2 3 2 2 5 3 9 12 26 15 28 10 31 10 15 12 13 No. of countries introduced into requirements Not known Not known coastal Tropical, Not known dry Tropical, Subtropical, coastal Subtropical, coastal tablelands Tropical, temperate All except tablelands Tropical, temperate All except Not known temperate All except coastal Tropical, temperate All except dry Temperate, dry Temperate, temperate All except Not known Subtropical, inland tablelands Tropical, Not known Not known Not known Not known Subtropical, dry Subtropical, coastal Not known Tropical Not known Not known Not known temperate All except Subtropical, coastal temperate All except Not known coastal Tropical, temperate All except species from which they were collected, their feeding guild and climatic species from which they Lantana stem sucker sap sucker sap sucker sap sucker sap sucker feeder flower feeder leaf & flower feeder flower feeder flower sap sucker flower feeder flower seed feeder stem borer root feeder stem borer leaf feeder leaf feeder leaf miner leaf miner leaf miner leaf miner leaf miner leaf miner leaf miner seed feeder stem galler leaf feeder leaf miner feeder flower feeder flower leaf feeder leaf feeder leaf feeder leaf feeder feeder flower leaf feeder leaf feeder feeder flower , the L. urticifolia L. urticifolia L. urticifolia L. urticifolia unknown unknown unknown unknown unknown L. urticifolia L. urticifolia L. urticifolia L. urticifolia L. urticifolia L. hirsuta L. tiliifolia L. fucata L. urticifolia L. urticifolia L. urticifolia L. tiliifolia L. tiliifolia unknown L. urticifolia L. urticifolia L. urticifolia L. fucata L. urticifolia L. urticifolia L. urticifolia unknown L. urticifolia L. urticifolia unknown L. urticifolia L. urticifolia unknown L. urticifolia Lantana camara Brazil N/A pathogen Subtropical, coastal 1 1 Mexico Jamaica Mexico unknown Colombia, Peru Trinidad Brazil Trinidad Brazil Mexico Mexico Mexico Mexico Mexico Mexico Brazil Brazil Costa Rica Mexico Costa Rica Brazil Argentina, Brazil Trinidad Florida Mexico Mexico Brazil Mexico Mexico Mexico Colombia Panama Kenya USA Mexico Mexico Cuba, USA Mexico Brazil N/A pathogen temperate All except 1 1 sp. Ecuador N/A pathogen Not known 1 1 sp. A sp. B Mycovellosiella lantanae Mycovellosiella Aconophora compressa Aconophora intermedia Falconia Orthezia insignis Phenacoccus parvus decora Leptobyrsa bifasciata Teleonemia elata T. harleyi T. prolixa T. scrupulosa T. lantanae Ophiomyia camarae O. lantanae xanthochaeta Eutreta garcia Ectaga lantanella Cremastobombycia Strymon bazochii Tmolus echion illustrata Autoplusia Diastema tigris Hypena laceratalis sunia Neogalea pusillidactyla Lantanophaga santatalis Pseudopyrausta Salbia haemorrhoidalis Epinotia lantana Apion Apion championi xanthomelas Parevander spinipennis Plagiohammus parana Alagoasa Charidotis pygmaea Octotoma championi O. scabripennis fulvopustulata Uroplata girardi U. lantanae U. Prospodium tuberculatum Prospodium Septoria Miridae Ortheziidae Pseudococcidae Tingidae Tephritidae Gracillariidae Lycaenidae Noctuidae Pterophoridae Pyralidae Tortricidae Cerambycidae Chrysomelidae requirements, and the number of countries in which they have been introduced and established. have requirements, and the number of countries in which they A list of agents introduced deliberately or accidentally for the biocontrol Mycosphaerellales Mycosphaerellaceae Hemiptera Membracidae Diptera Lepidoptera Depressariidae OrderColeoptera Family Apionidae Species Origin Host plant Guild Climatic Uredinales Pucciniaceae Coelomycetes Sphaeropsidaceae Table 1. Table

82 Selecting agents for lantana

Host plants have a wide enough host range to be able to utilize the hybrid forms of lantana in the target countries. The first step in any biocontrol program is to know the identity of the target weed (Schroeder & Goeden 1986). For lantana, this has not been achieved. The name Agent guilds “Lantana camara” has been loosely applied by collec- Leaf- and/or flower- and fruit-feeding have been tors and authors to many species and hybrids, with the utilized successfully against many weeds (Julien & result that much of the recorded host-plant information Griffiths 1998). However, they have not been able to in botanical, horticultural and ecological literature is achieve similar results against lantana. Lantana has the unreliable. The introduction of cultivated forms back ability to withstand a wide range of climatic conditions into neotropical regions has allowed interactions with and can survive in hot, dry regions as well as those the native gene pool, producing further complex areas susceptible to frost or prolonged droughts (Swar- morphological variation (Mendez Santos 2002). Even brick et al. 1998). During periods of drought, plants can after decades of considerable effort by morphological lose all their leaves, and populations of biocontrol taxonomists and molecular biologists, lantana remains agents, especially those of leaf- and flower-feeding a taxonomic enigma. However, there have been some insects, can decrease dramatically. Insect populations significant achievements: DNA studies showed that the can increase again once conditions improve and the common pink-flowering lantana variety from Australia, plant becomes healthy (Day et al.2004). As a result, Fiji and Vanuatu has a greater affinity with L. urtici- populations are not maintained at levels sufficient to folia from Mexico than with any other species of control the plant. Rather than leaf- and flower-feeding lantana tested (Scott et al. 2002). insects suppressing plant growth and reproduction, it The DNA studies also suggested that the progenitors appears that they may respond to the health of the plant. of lantana in other regions of the world may be different Historically, most agents selected for the biocontrol from those of lantana in Australia, Fiji and Vanuatu. of lantana were insect members of leaf- or flower- For example, lantana plants from Hawaii and the feeding guilds (possibly as these were the most Solomon Islands are similar to one another, but conspicuous and easiest to test). Other guilds such as different from those from Australia (Scott et al. 2002). stem borers and root feeders and groups such as patho- As yet, the DNA studies have not determined the gens require more effort and were often overlooked. progenitors of lantana from Hawaii, the Solomons or Worldwide, the proportion of the 41 agents introduced many other countries. (and the 27 established) in the different guilds were: In addition, some lantana specimens from Australia leaf feeders 56% (60%); flower or fruit feeders 24% and South Africa have been identified (using morpho- (19%); stem feeders 10% (11%); and root feeders 2.5% logical characters) as L. urticifolia × L. camara hybrids (a single species which did not establish). Three fungal (R. Sanders, Botanical Research Institute of Texas, pathogens have also been introduced and all have estab- pers. comm.). These findings, along with the DNA lished where released. studies, suggest that many of the weedy lantana varie- When considering individual countries or regions, the ties that were assumed to be L. camara could be deriv- effort put into biocontrol varies considerably. Australia atives of L. urticifolia. (30), Hawaii (25) and South Africa (24) have imported Retrospective studies have revealed the following: the most agents, with most other countries importing 3–5 1. most insects and mites collected as potential biocon- agents (Julien & Griffiths 1998, Day et al., unpublished trol agents in surveys in Mexico and Brazil have data). However, some agents have been introduced into discrete hosts (less than 10% were found on more only one country or region and no country has introduced than one Lantana species) (Winder & Harley 1983, all 41 agents. In most countries or regions, only leaf- or Palmer & Pullen 1995) flower-feeding insects have established. In South Africa, 2. a greater proportion (73%) of agents that were 71% of the 14 insect agents that have established feed on collected from L. urticifolia or L. camara estab- leaves, 21% feed on flowers/fruits, while no insects that lished on weedy lantana than agents collected from feed on the stems or roots have established. One path- other species of Lantana (25%) (Day et al. 2002, ogen has established. In Australia, 69% of the 16 insect Day et al., unpublished data) agents established feed on leaves and 19% feed on 3. agents that were found on three or more Lantana flowers/fruits. One stem-feeding insect and one path- species in their native range had a greater establish- ogen have also established. Both countries can claim ment rate (82%) on weedy lantana than those that only partial control of lantana. Other countries, such as were found on only one or two Lantana species Fiji (seven agents established), India (seven) and Feder- (36%) (Day et al., unpublished data). ated States of Micronesia (six) have only leaf- or flower/ Entomologists are now selecting agents that are fruit-feeding agents and adequate control has also not either (a) specific to L. urticifolia or L. camara, as they been achieved. should be better adapted to the weedy hybrids, or (b) In contrast, in Hawaii, where lantana is reported to stenophagous to several Lantana species, as they may be under control in some areas, a smaller proportion of

83 Proceedings of the XI International Symposium on Biological Control of Weeds the established agents feed on leaves (53%), 29% attack specific (Simelane 2001) and very damaging to plants. flowers/fruits and a greater proportion attack stems Adults feed on leaves, but the significant damage is due (12%). One pathogen has also established. The stem- to larvae feeding on secondary roots (Baars & Neser boring cerambycid Plagiohammus spinipennis, and the 1999), disrupting the uptake of water and nutrients by stem-galling tephritid Eutreta xanthochaeta, have the plant, thereby increasing plant stress and reducing established only in Hawaii. Both agents cause substan- plant growth (Simelane 2001). The adults have a tial damage to lantana, especially in the dry areas of the diapause stage enabling over-wintering and the insect is big island and aid control, such that lantana is not relatively easy to rear. considered a major weed there (Davis et al. 1992). The eriophyid budmite Aceria lantanae has been These insects were introduced into Australia and South tested in South Africa and is deemed host-specific. A. Africa, but failed to establish. lantanae causes undifferentiated inflorescence buds to For most countries where only a few agents have form microphyllous galls instead of flowers, fruits and been introduced and established, introducing the main seeds. The galls probably also act as a metabolic sink, damaging agents such as Octotoma scabripennis, debilitating the plant and reducing its competitive Uroplata girardi and Teleonemia scrupulosa would be ability (Baars & Neser 1999, Urban et al. 2001). a priority. However, most perennial weeds can tolerate a single defoliation (Harris 1973). Lantana is adapted to Climatic adaptations repeatedly losing its leaves due to drought and/or frost. This defoliation tolerance is the basis of the resilience Despite the introduction of several new, potentially of lantana to most of its biocontrol agents (Broughton damaging agents, the climatic adaptability of lantana 2000, Day & Neser 2000). For this reason, added pres- may still prove an immense problem to overcome. sure is needed on the plant, to further deplete carbohy- Lantana is found in a wide range of climatic and drate reserves accumulated for regrowth. For countries geographical areas (Henderson 2001, Day et al. 2004) such as South Africa, Australia and India, priority and agents are unlikely to be suitable for all regions. As should be given to the utilization of pathogens and an example, at the limits of the distribution of lantana in obtaining agents that attack stems or roots. Accord- Australia (33°S), only two agents, the leaf-mining ingly, in the past few years, a number of promising new hispine Octotoma championi and the fruit-mining agro- candidate agents has been imported and released into myzid Ophiomyia lantanae are found. Both are found South Africa or Australia. in only low numbers, such that damage to the weed is One major advance in the biocontrol of lantana has negligible, even in late summer (Day et al. 2004). In been the recent introduction of two pathogens, Mycov- addition, few agents are found on lantana under canopy, ellosiella lantanae in South Africa and Prospodium or at altitudes greater than 200 m in temperate areas. tuberculatum in Australia. Only one other pathogen, Agents such as O. scabripennis cease oviposition when Septoria sp. in Hawaii, has been utilized against lantana temperatures drop below 15°C and may complete only (Trujillo & Norman 1995). All pathogens have estab- two generations a year (C. Clech & M. Day, unpub- lished in their respective countries or regions of intro- lished data). This is insufficient for populations to build duction but it is too early to determine their impact on up and be maintained at damaging levels. the weed (Tomley & Riding 2002, A. den Breeÿen, The herring-bone leaf miner, Ophiomyia camarae, pers. comm.). Pathogens induce toxic effects that was released in South Africa in 2001 and has estab- disrupt physiological processes and may complement lished widely (Simelane 2002). Larvae tunnel along the the actions of insects. Surveys by Barreto et al. (1995) leaf veins, disrupting translocation of water to, and found several other pathogens, such as Puccinia photosynthates from, the lamina, and causing prema- lantanae and Ceratobasidium lantanae-camarae, that ture abscission of leaves. Field trials suggest that it indi- are also worthy of further study. rectly reduces plant growth and reproduction (Simelane The stem-boring has 2002). O. camarae appears to perform well in sheltered been released in Australia and Hawaii. It has failed to areas where lantana grows as an understorey. Even establish in both regions, probably due to the small though this agent is a leaf feeder, it may improve numbers released. It is particularly damaging, tunnelling biocontrol in cooler and sheltered areas where few down and killing branches of lantana. In the field, plants agents are present. attacked by A. championi are stunted and less fruitful. So Apart from the diverse climates in which lantana can far, the beetle has proved difficult to rear and establish. If grow, many areas where lantana is a problem have a successful rearing method can be developed to produce distinct wet and dry seasons. During the dry season, large enough numbers of insects for successful establish- lantana may lose its leaves, causing populations of leaf- ment, A. championi may markedly improve biocontrol of feeding agents to crash. It is therefore desirable to select lantana. At present, South Africa and Australia are agents that can bridge periods of leaflessness. The working on new methods to rear the insect. petiole-galling apionine Coelocephalapion camarae is The root-feeding beetle Longitarsus sp. AcSN2440, a small, fast-breeding, host-specific beetle that indi- undergoing quarantine testing in South Africa, is host- rectly stunts root growth (Baars & Heystek 2001).

84 Selecting agents for lantana

Adults feed on the leaf laminae, while the larvae form can have dramatic impact on plant health and popula- galls in the petioles, causing early abscission of leaves, tions (Harris 1973, Blossey & Hunt-Joshi 2003). or occasionally in the peduncles of inflorescences, Furthermore, root feeders have contributed more to reducing flowering and seed set (Baars & Neser 1999). control of weeds than other agents. Over 50% of estab- The petiole galls disrupt the transport of water and lished root feeders contribute to control of weeds, nutrients, causing a reduction in root growth (Baars & compared to only 30% of aboveground feeders Heystek 2001). The long-lived adults have been (Blossey & Hunt-Joshi 2003). There are several Longi- recorded at altitudes of 1600 m (Kissinger 2000), are tarsus species that have been utilized against weeds and clearly adapted to areas where lantana is leafless in they appear to make a valuable contribution to the winter, and should contribute to biocontrol in such control of annual weeds (Julien & Griffiths 1998). The areas. group’s impact on perennials, however, is unknown. The effect of gall-forming agents is often underesti- Discussion mated. Galls can act as metabolic sinks and their effect is often greater than their size or physical damage to the Lantana has been the subject of biocontrol programs for plant would indicate. Galls can disrupt the translocation 100 years and adequate control has not been achieved of water and nutrients to growing shoots and nutrients in most regions. Recent developments have suggested to roots, increasing water stress and reducing plant that better biocontrol could be obtained with the intro- growth. There are a number of stem- and flower-galling duction of better-adapted and more effective agents. agents causing substantial damage to perennial weeds, While it is generally hard to predict which agents or e.g. Cecidochares connexa on Chromolaena odorata in even guilds of agents are likely to be the most damaging Guam and Indonesia, and Trichilogaster acaciaelongi- to a weed, there is sound reason to suggest that targeting foliae on Acacia longifolia and Uromyces tepperianum agents that have certain characteristics would be advan- on A. saligna in South Africa (Julien & Griffiths 1998). tageous. In addition, other Aceria spp. have already been used Sheppard (1992) suggests that genetically variable successfully against weeds (Julien & Griffiths 1998). weeds are more difficult to control through biological Therefore, the potential of gall-forming agents, such as means than weeds that are genetically homogeneous. A. lantanae and E. xanthochaeta, to contribute to the Therefore, potential agents should be collected from the control of lantana would seem to merit further research. most closely related plant species, as those agents Pathogens are still grossly under-utilized. In fact, should be better adapted to the target weed. However, they make up only a small proportion of all agents the most closely related species to the weedy forms of released against weeds. Until recently, there was some lantana were, for many years, just not known. The concern about the use of pathogens, but pathogens have results of recent DNA studies suggest that potential now successfully controlled weeds, such as rubber vine agents should be collected from L. urticifolia and some and noogoora burr (Julien & Griffiths 1998). Pathogens of these agents are showing considerable promise in the may be highly specific and may actually be limited in laboratory. their impact, especially on weeds that are genetically A method for rating the effectiveness of agents, diverse. However, it is possible that not all varieties of using a number of criteria such as host specificity, type lantana would be susceptible to any one pathogen. Path- of damage inflicted, phenology of attack and number of ogens have a number of attributes that are superior to generations per year, was developed by Harris (1973) those of insects. They have different modes of attack to and later modified by Goeden (1983). The benefits of insects and are able to build up into large populations defoliating insects have been well documented, espe- faster and disperse more quickly than insects. Patho- cially in annual weed species. They can reduce plant gens can usually be mass-cultured more cheaply than growth, flowering, seeding and the accumulation of insects while some pathogens can also be prepared and carbohydrate reserves. However, the prime concern in applied as a target-specific mycoherbicide when and controlling perennial weeds is the destruction of where required. existing plants. Defoliating insects do not necessarily While targeting specific guilds may appear benefi- achieve this and their effectiveness is restricted to the cial in theory, the diverse range of climates in which summer months during which the weed is vulnerable to lantana can grow may limit the effectiveness of many insect attack (Harris 1971). potential agents, especially in the cooler regions. Selecting agents that feed on the stems or roots may CLIMEX has been particularly useful in identifying be more effective for suppressing the weed, as they do possible search areas, given that lantana has a wide not rely on plants to be in leaf all year round (Harris distribution in neotropical regions. 1971). Stem-boring or root-feeding agents remove Biocontrol of lantana has been conducted for over a biomass and/or disrupt the transport of water and nutri- century, yet suppression of the weed remains inade- ents and can severely weaken the plant. Stem-boring quate. Today, practitioners have the advantage of being larvae, being internal feeders, can survive independent able to review, analyze and learn from past attempts, of the condition of leaves, while root-feeding agents and to utilize new technology. Applying this knowl-

85 Proceedings of the XI International Symposium on Biological Control of Weeds edge in selecting agents that are ecologically better Harris, P. (1971) Weed vulnerability to damage by biological adapted to the target plant and its environment, and are control agents. Proceedings of the II International Sympo- able to markedly suppress plant growth and reproduc- sium on Biological Control of Weeds, (ed P.H. Dunn), pp. tion, should offer some improvement in the overall 29–39. Commonwealth Agricultural Bureaux, Slough, UK. Harris, P. (1973) The selection of effective agents for the biocontrol of lantana. biological control of weeds. The Canadian Entomologist 105, 1494–1503. Acknowledgements Henderson, L. (2001) Alien weeds and invasive plants. Agricul- tural Research Council, South Africa. We thank members of the lantana biocontrol teams of Howard, R.A. (1969) A checklist of cultivar names used in the NRM-AFRS and ARC-PPRI for their research inputs, genus Lantana. Arnoldia 29, 73–109. Dane Panetta, Joe Scanlan and Stefan Neser for their Julien, M.H. & Griffiths, M.W. (1998) Biological Control of constructive criticism, and the Working for Water Weeds: A World Catalogue of Agents and their Target Programme of the South African Department of Water Weeds, Fourth Edition. CABI Publishing, Wallingford, UK. Affairs and Forestry and HL Hall & Sons of South Kissinger, D.G. (2000) A new species of Coelocephalapion Africa for financial support. Wagner (Brentidae) from Mexico with host Lantana camara L. (Verbenaceae) The Coleopterists Bulletin 54, 365–369. References Mendez Santos, E.I. (2002) A taxonomic revision of Lantana sect. Lantana (Verbenaceae) in the Greater Antilles. Willde- Baars, J-R. & Heystek, F. (2001) Potential contribution of the nowia 32, 285–301. petiole galling weevil, Coelocephalapion camarae Kiss- Palmer, W.A. & Pullen, K.R. (1995) The phytophagous arthro- inger, to the biocontrol of Lantana camara L. Proceedings pods associated with Lantana camara, L. hirsuta, L. urtici- of the 13th Entomological Congress (eds T. Olckers & D.J. folia, and L. urticoides (Verbenaceae) in North America. Brothers), pp. 79–80. Pietermaritzburg, South Africa. Biological Control 5, 54–72. Baars, J.-R. & Neser, S. (1999) Past and present initiatives on Schroeder, D. & Goeden, R.D. (1986) The search for the biological control of Lantana camara (Verbenaceae) in natural enemies of introduced weeds for biological control – South Africa. African Entomology Memoir 1, 21–33. in theory and practice. Biocontrol News and Information 7, Barreto, R.W., Evans, H.C. & Ellison, C.A. (1995) The myco- 147–155. biota of the weed Lantana camara in Brazil, with particular Scott, L.J., Hannan-Jones, M.A. & Graham, G.C. (2002) Affin- reference to biological control. Mycological Research 99, ities of Lantana camara in the Australia–Pacific region. 769–782. Proceedings of the 13th Australian Weeds Conference (eds Blossey, B. & Hunt-Joshi, T.R. (2003) Belowground herbivory H. Spafford Jacob, J. Dodd & J.H. Moore), pp. 471–474. by insects: influence on plants and aboveground herbivores. Plant Protection Society of Western Australia, Perth. Annual Review of Entomology 48, 521–547. Sheppard, A.W. (1992) Predicting biological weed control. Broughton, S. (2000) Review and evaluation of lantana biocon- TREE 7, 290–291. trol programs. Biological Control 17, 272–286. Simelane, D.O. (2001) Laboratory host range of Longitarsus Davis, C.J., Yoshioka, E. & Kageler, D. (1992) Biological sp., a flea beetle on Lantana camara L. Proceedings of the control of lantana, prickly pear, and Hamakua pamakani in 13th Entomological Congress (eds T. Olckers & D.J. Hawai’i: a review and update. Alien Plant Invasions in Brothers), p. 59. Pietermaritzburg, South Africa. Native Ecosystems of Hawaii: Management and Research. Simelane, D.O. (2002) Biology and host range of Ophiomyia (eds C.P. Stone, C.W. Smith & J.T. Tunison), pp. 411–431. camarae, a biological control agent for Lantana camara in University of Hawaii Press, Honolulu. South Africa. BioControl 47, 575–585. Day, M.D., Broughton, S. & Hannan-Jones, M.A. (2004) Smith, L.S. & Smith, D.A. (1982) The naturalised Lantana Current distribution and status of Lantana camara and its camara complex in eastern Australia. Queensland Botany biological control agents in Australia, with recommenda- Bulletin 1, 1–26. tions for further biocontrol introductions into other coun- Stirton, C.H. (1977) Some thoughts on the polyploid complex tries. Biocontrol News and Information (in press). Lantana camara L. (Verbenaceae). Proceedings of the Day, M.D. & Neser, S. (2000) Factors influencing the biological Second National Weeds Conference of South Africa, pp. control of Lantana camara in Australia and South Africa. 321–340. Stellenbosch, South Africa. Proceedings of the X Symposium on Biological Control of Swarbrick, J.T., Willson, B.W. & Hannan-Jones, M.A. (1998) Weeds (ed N.R. Spencer), pp. 897–908. United States Lantana camara L. The Biology of Australian Weeds (eds Department of Agriculture, Agricultural Research Services, F.D. Panetta, R.H. Groves & R.C.H. Shepherd), pp. Sidney, MT and Montana State University, Bozeman, MT. 119–140. R.G. & F.J. Richardson, Melbourne. Day, M.D., Playford, J. & Zalucki, M.P. (2002) The role of Tomley, A.J. & Riding, N. (2002) Prospodium tuberculatum, DNA analyses in the biological control of Lantana camara. lantana rust, a new agent released for the biocontrol of the Proceedings of the IOBC International Symposium on the woody shrub Lantana camara. Proceedings of the 13th role of genetics and evolution in biological control (ed C. Australian Weeds Conference (eds H. Spafford Jacob, J. Silvy), p. 18. Montpellier, France. Dodd & J.H. Moore), pp. 389–390. Plant Protection Society Goeden, R.D. (1983) Critique and revision of Harris’ scoring of Western Australia, Perth. system for selection of insect agents in biological control of Trujillo, E.E. & Norman, D.J. (1995) Septoria leaf spot of weeds. Protection Ecology 5, 287–301. lantana from Ecuador: apotential biological control for bush Graaff, J.L. (1986) Lantana camara, the plant and some lantana in forests of Hawaii. Plant Diseases 79, 819–821. methods for its control. South African Forestry Journal 136, Urban, A.J., Mpedi, P.F., Neser, S. & Craemer, C. (2001) Poten- 26–30. tial of the flower gall mite Aceria lantanae (Cook) (Acari:

86 Selecting agents for lantana

Eriophyidae), for biocontrol of the noxious weed, Lantana Winder, J.A. & Harley, K.L.S. (1983) The phytophagous camara L. (Verbenaceae). Proceedings of the 13th Entomo- insects on Lantana in Brazil and their potential for biological logical Congress (eds T. Olckers & D.J. Brothers), pp. control in Australia. Tropical Pest Management 29, 67–68. Pietermaritzburg, South Africa. 346–362.

87