Lantana camara
(lantana)

L. camara is a highly variable ornamental shrub, native of the neotropics. It has been introduced to most of the tropics and subtropics as a hedge plant and has since been reported as extremely weedy and invasive in many countries. It is generally deleterious to biodiversity and has been reported as an agricultural weed resulting in large economic losses in a number of countries. In addition to this, it increases the risk of fire, is poisonous to livestock and is a host for numerous pests and diseases. L. camara is difficult to control. In Australia, India and South Africa aggressive measures to eradicate L. camara over the last two centuries have been largely unsuccessful, and the invasion trajectory has continued upwards despite control measures. This species has been the target of biological control programmes for over a century, with successful control only being reported in a few instances. 

L. camara is native to Central and South America but its original distribution is unclear due to the introduction of a number of ornamental varieties. The species has also been poorly investigated in its native range, where it is not usually considered to be a serious pest, and the extent of its original native range is unclear. In the West Indies it is found in dry thickets (Adams, 1976). The weed is noted to be present in the Galapagos Islands of Ecuador (Cruz et al., 1986). This species has been widely promoted as an ornamental since the early 1800s and it is widely naturalized throughout the Neotropics. It is present on all continents expect Antartica. It has become very widespread in Australia, India and South Africa, infesting millions of hectares of land (Bhagwat et al., 2012).

The main spread of L. camara into new countries is via the horticultural trade, spreading numerous different varieties of this species. Once introduced into an area the seeds of L. camara are readily dispersed into new areas by birds. L. camara has already spread widely, but there is potential for its range to expand further under future climate change. In Australia it is found in almost all the climatically suitable habitat, but under future climate change it could expand into new areas in Victoria, South Australia and Tasmania (Taylor et al., 2012; Taylor and Kumar, 2013).

L. camara has a wide environmental tolerance and occurs in a variety of habitats. These include wastelands, rainforest edges and beachfronts (ISSG, 2015). It also grows well in disturbed areas such as roads, railways and areas recovering from fire or logging (ISSG, 2015). This species can tolerate some shade but grows best in open unshaded regions. It cannot directly colonise intact forests but instead grows at forest edges and spreads when gaps are created (ISSG, 2015).

L. camara is an agricultural weed that can cause dramatic losses in yields. In Australia, it was reported that L. camara infested 4 million ha of pasture (Parsons and Cuthbertson, 1992). A number of plants affected by L. camara are listed in the "Host Plants/Plants Affected" table below.

Genetics

The known chromosome numbers of L. camara are 2n = 22, 33, 44, 55, but most invasive varieties appear to be tetraploids (Day et al., 2003). Besides variation in chromosome number there is much variation in DNA content, growth rates and toxicity to livestock (Stirton, 1979; Gujral and Vasudevan, 1983; Scott et al., 1997). In the Tamil Nadu region of India there are differences in toxicity of L. camara, with the red flowered variety being more toxic than the pink flowered form (Thirunavukkarasu et al., 2001).

Physiology and Phenology

Flowering and fruiting take place throughout the year with a peak during the first two months of the rainy season.

In the highlands of western Kenya an investigation of leaf decomposition found that after seven days it had decreased to just under a third of the original mass and by the 77th day the leaves had totally decomposed. The percentage of the initial amount of phosphorus and nitrogen remaining in the leaf material after a week was 42 and 54%, respectively. After 21 days 90% of the phosphorus had been released (Kwabiah et al., 2001).

Reproductive Biology

The flowers of L. camara, when yellow, produce nectar and are pollinated by butterflies and thrips. The species is an obligate outcrosser and it is unclear whether apomixis occurs. Fruits mature rapidly and change colour from dark green to black. A number of bird species, and also sheep and goats disperse the seeds, sometimes over long distances, but natural dispersal between oceanic islands has never been demonstrated. Heavy fruit crops are produced yearly, but the thornless forms produce few, if any, seeds. Seeds germinate when sufficient moisture is available, usually at the start of a rainy season. In Australia, Broughton (1999) found that 57-80% of green and ripe fruits tested had one or two viable seeds whereas 12-34% had none, and between 64 and 90% of dried (older) fruits had two nonviable embryos, suggesting that fruit development stage affects germination. She found no difference in viability within sites or between cultivars investigated. Projections of seed survival indicate that L. camara seeds could survive for up to 11 years under natural rainfall conditions in Australia (Vivian-Smith and Panetta, 2009).

In addition to spread by seed, L. camara is able to produce adventitious shoots, especially shallow lateral roots, following mechanical damage. Hence, it is also able to spread and establish dense thickets by vegetative means. The capacity of the species to spread vegetatively and to inhibit both the growth of other vegetation and seed germination, in conjunction with heavy and regular fruiting, is the main reason why L. camara forms long-lasting permanent thickets. In areas where natural fires occur they stimulate thicker regrowth.

For further information, see Mathur and Mohan Ram (1978), Schemske (1983), Sinha and Sharma (1984) and Thaman (1974).

Environmental Requirements

L. camara can grow between the latitudes 45°N and 45°S and an altitude of up to 1,400 m.

The rapid spread of L. camara throughout the tropics is associated with human-induced disturbances. It forms extensive, dense and impenetrable thickets in forestry plantations, orchards, pasture land, waste land and in natural areas. L. camara thrives in open and disturbed areas as well as in open natural vegetation. Being somewhat shade-tolerant it can become the dominant understorey shrub in open forests, but is absent from closed forests. L. camara grows under a wide range of climatic conditions. In Australia it tolerates a mean annual rainfall from 4000 to less than 1000 mm, and as low as 200 mm per annum elsewhere (Gujral and Vasudevan, 1983). It is found between sea level and nearly 1000 m on Hawaii and higher in East Africa, the upper altitudinal limit being determined by frost, which the plant is susceptible to. In Hong Kong, temperature in the range 3-5°C injured L. camara (Corlett, 1992). It tolerates salt spray. Its distribution is affected by soil type. It has a low tolerance for boggy and saline soils but grows well on poor soils. Studies undertaken by Muvengwi and Ndagurwa (2015) on soil seed bank dynamics and soil nutrient concentrations in the wetlands of New Gada in Zimbabwe, suggest that L. camara-invaded soil had significantly higher levels of calcium, magnesium, sodium and ammonium and lower levels nitrate than non-invaded soils. 

L. camara has a marked ability to compensate for herbivory as plants survived experimental defoliation for two years (Broughton, 2000).

Associations

L. camara often occurs in pure stands but can be mingled with a variety of species but emergent shrubs and trees in particular.

The alkaloid-rich leaves of L. camara make it virtually immune to grazing by livestock, although several hundred phytophagous insects have been recorded on it. In the New World, flowers, flower stalks, leaves, shoots and roots are attacked by many insect species and pathogens although their impact on shrub vigour and seed set is poorly understood. The polyphagous pest Phenacoccus parvus severely damages stands of L. camara in Australia and is not, as commonly reported, a pest of potato and aubergine (Solanum melongena), although it has the potential to attack a variety of plant species inclusive of some crops (Marohasy, 1994). In Mexico a stem sap-sucking membracid bug, Aconophora compressa, causes considerable dieback of stems (Swarbrick et al. 1995).

The following fungi have been found attacking the leaves of L. camara: Dendryphiella aspera, Micropustulomyces mucilaginosus, Mycovellosiella lantanae var. lantanae, Septoria sp., Ceratobasidium lantanae, Prospodium tuberculatum and Puccinia lantanae. For further information on fungal natural enemies of L. camara, see Barreto et al. (1995), Breeÿen et al. (2000), Thomas and Ellison (2000), Trujillo and Norman (1995).

Animal feed, fodder, forage

• Fodder/animal feed

Environmental

• Boundary, barrier or support
• Erosion control or dune stabilization

Fuels

• Biofuels
• Fuelwood

General

• Ornamental

Human food and beverage

• Honey/honey flora

Materials

• Pesticide
• Poisonous to mammals

Medicinal, pharmaceutical

• Traditional/folklore

L. camara is conspicuous due to its attractive and multicoloured floral displays and is a well-known species throughout the tropics.

Although large stands of weedy varieties of L. camara are easily recognised, it is in fact a variable polyploid complex of interbreeding taxa resulting from hybridisation with species in the other complexes, such as L. urticifolia (Day et al., 2003). In Florida, USA, it may be confused with the endangered endemic native, L. depressa, with which it has extensively hybridised (Langeland and Burks, 2000).

Due to the variable regulations around (de)registration of pesticides, your national list of registered pesticides or relevant authority should be consulted to determine which products are legally allowed for use in your country when considering chemical control. Pesticides should always be used in a lawful manner, consistent with the product's label.

Cultural Control

Being poisonous to livestock means that the species can not be controlled using large herbivores. In fact, intense grazing by goats and donkeys will favour L. camara infestations by suppressing competition from palatable species (Ashmole and Ashmole, 2000).

Osunkoya et al. (2013) suggest from studies and simulation models in Queensland, Australia that periodic burning could control the weed in forests within 4-10 years if fire frequency is at least every two years. On farms, site-specific control may be achieved by 15 years if the biennial fire frequency is tempered with increased burning intensity.

Mechanical Control

Mechanical control can be effective, particularly where land is cleared, but requires continual follow-up treatment to remove roots and seedlings of L. camara. Slashing and burning stimulate suckering. Both chemical and mechanical control methods are expensive and labour intensive and are only effective in the short term. Cleared areas are rapidly colonised via seeds originating from distant parents or from sprouting roots. Dohn et al. (2013) recommend hand pulling for creating firebreaks or where minimizing damage to native species is paramount.

Chemical Control

The Australian experience in controlling L. camara, reviewed by Swarbrick et al. (1995), indicates that some herbicides are more effective on particular forms of L. camara. The most effective herbicides belong to the phenoxy acid (2,4-D, dichloroprop and MCPA), benzoic acid (dicamba) and pyridine groups. Glyphosate, sulfonylureas (metsulfuron methyl) and imidazolinones (imazapyr) also show good activity. Photosynthetic herbicides (triazine and urea) are not effective. A number of factors affect the effectiveness of the chemical treatment and they include: plant size, time of application, mode of application, and the use of surfactant. Use of herbicide in uncut stands may not be effective in preventing eventual regrowth. Combination of mechanical and chemical control may be the best. The seasonal response of L. camara to applications of fluroxypyr, metsulfuron-methyl, glyphosate and dichlorprop has been reported by Hannan-Jones (1998).

Work carried out in the South African Kruger National Park by Erasmus et al. (1993) showed that chemical control was cheaper and caused less disturbance resulting in higher biodiversity than mechanical control. Chemical control consisted in an application of imazapyr on freshly cut stems and a follow-up operation by spot-spray application of glyphosate. The initial control required 25 man-days per ha and that of the follow-up control 6.8 man-days per ha. Control costs will vary from site to site and will depend on L. camara stem density and cover. Latest South African recommendations are provided by Vermeulen et al. (1996).

In India, eradication of L. camara from sub-watersheds in the Markanda catchment, Himachal Pradesh, was effective and economical using glyphosate sprayed on to regenerated growth, cut four months previously (Rana and Singh, 1999).

In Queensland, Dohn et al. (2013) suggest that foliar spraying with a glyphosate-based herbicide is the most efficient treatment for combating large infestations of L. camara. In Florida, Ferrell et al. (2011) report that this species can be effectively controlled by two applications of fluroxypyr, two applications of fluroxypyr+aminopyralid, or a single application of aminocyclopyrachlor. 

L. camara is resistant to triclopyr, a widely used herbicide for woody weed control (Goodall and Naude, 1998).

Biological Control

Worldwide, well over 200 releases of biocontrol agents have been made (39 different natural enemies have been released in 29 countries), however, in the majority of cases the control agent either failed to become established or became established without achieving control. Despite this limited success, classical biological control is still considered to be the only viable, long-term control option, since it offers a safe, economic and environmentally benign method of suppressing the weed. Most of the releases have been carried out in the Pacific, South Africa and Australia (for historical details see Taylor 1989; Cilliers and Neser, 1991; Denton et al., 1991; Davis et al., 1992; Swarbrick et al., 1995). The most widely established species include Ophiomyia lantanae, Uroplata girardi and Octoma scabripennis. Day et al. (2003) have produced a detailed review of 48 of these control agents.

In Hawaii, Neogalea sunia and Epinotia lantanae contribute to the control of L. camara across the islands. In addition, a combination of Hypena strigata, Octotoma scabripennis, Salbia haemorrhoidalis, Teleonemia scrupulosa and Uroplata girardi provide partial to substantial control in drier areas <1270 mm rainfall), and in wetter areas Plagiohammus spinipennis provides partial control (Julien and Griffiths, 1998).

The release in 1993 of U. girardi on an island of the Russell Island group (Solomon Islands) resulted in the successful control of the 'Hawaiian Pink' form (Swarbrick et al., 1995). U. girardi has proved to be one of the more successful agents and is credited with providing some check on the spread of L. camara in Australia, South Africa and some islands in the Pacific Ocean. In Micronesia seven out of 13 introduced insect species became established and have resulted in acceptable levels of control for current agricultural practices (Denton et al., 1991). As elsewhere the effectiveness of the insect species varied between islands and between varieties of L. camara. Greater success appears to have been achieved in drier areas.

In Uganda, the introduction of T. scrupulosa, which had been widely released after its successful introduction into Hawaii in 1902, was very successful in the area around Serere Research Station in Teso District but it also attacked one of the cultivars of Sesamum indicum grown on the Research Station (Davies and Greathead, 1967). Fortunately, it was unable to breed on that crop and attacks subsided after the L. camara had been controlled. Subsequently other agents for L. camara control were tested on Sesamum and it was found that other Tingidae and the chrysomelid leaf miners would also feed on this crop (Greathead, 1973).

Recent releases of arthropod biocontrol agents in Australia (Queensland and New South Wales) to control L. camara include the treehopper Aconophora compressa (first in 1995), the mirid Falconia intermedia (during 2000-2004), and the leaf miner Ophiomyia camarae (first in 2007). Weather conditions and other factors have resulted in poor establishment or low levels of damage so far (Taylor et al., 2008).

Hersula and Hill (2012) report that populations of F. intermedia released in South Africa to control L. camara had disappeared after initially building up to high densities. It is suggested that some L. camara varieties possess factors enabling them to resist feeding activities after the initial attack. Biological control of L. camara in South Africa is reviewed by Moran et al. (2011), who suggest that it plays a subsidiary role in support of mechanical and chemical control, but that cost benefits justify the continued development of new agents.

Broughton (2000) reviewed biological control programmes of L. camara worldwide and concluded that leaf-, flower- and fruit-feeding species were the most successful feeding groups, and the leaf-mining chrysomelid U. girardi was the most successful control agent. She identified the main factor preventing the establishment of control agents as the number of individuals released and noted that cultivar preferences, parasitism and predation, and climate reduced control. Broughton (2000) concluded that flower- and fruit-feeding species were unlikely to be effective because the seeds of L. camara are only viable for a short period of time and have a low germination, and that defoliating species were likely to be ineffective because of the ability of L. camara to withstand defoliation. Julien and Griffiths (1998) showed that different cultivars display differences in susceptibility to insect herbivores.

A potential pathogen of L. camara (spreading in Hawaii) was identified by Trujillo and Norman (1995) as a leaf-spot fungus, Septoria sp., from Ecuador. Various other pathogens with apparently excellent potential to control a wide range of cultivars have been identified by Barreto et al. (1995) and Thomas and Ellison (2000). In South Africa, permission was granted in 2001 to release the fungus Mycovellosiella lantanae var. lantanae, collected from Florida, USA (Breeÿen et al., 2000, Breeÿen, 2004). In 2001 the rust fungus Prospodium tuberculatum (ex Brazil) was the first pathogen to be released in Australia for biological control of L. camara (Tomley and Riding, 2002; Ellison et al., 2006; Thomas et al., 2006). Drought conditions affected its establishment and spread from the release sites in Queensland and New South Wales and incidence is generally low, although some leaf drop has been observed (Taylor et al., 2008).

In May 2015, the first release of P. lantanae was made on New Zealand’s North Island. The rust had previously been screened by CABI scientists in the UK for its host specificity before being transferred to Landcare’s Plant Pathogen Containment Facility in Auckland. The blister rust is able infect leaves, petioles and stems and can cause systemic infections that lead to stem dieback and defoliation.

In addition, a second rust, P. tuberculatum (the same isolate released in Australia in 2001) was released in New Zealand as part of the same initiative. The leaf rust pathogen causes leaf death and defoliation and as it is subtropical, it is expected to be less dependent on high humidity to compliment P. lantanae in different climatic conditions (Anon, 2015).

07/12/15 Updated by:

Sarah E. Thomas, CABI, UK

23/09/13 Updated by:

Julissa Rojas-Sandoval, Department of Botany-Smithsonian NMNH, Washington DC, USA

Pedro Acevedo-Rodríguez, Department of Botany-Smithsonian NMNH, Washington DC, USA