8

Chromolaena odorata (L.) King and Robinson (Asteraceae)

C. Zachariades, M. Day, R. Muniappan, and G. V. P. Reddy

8.1 Introduction Chromolaena odorata (L.) King and Robinson (Asteraceae), formerly known as Eupatorium odoratum L., is a weedy pioneering shrub native to the Americas from southern USA to northern Argentina (Gautier, 1992). Chromolaena odorata has become one of the worst terrestrial invasive plants in the humid tropics and subtropics of the Old World over the past century (Holm et al., 1977; Gautier, 1992). From its original point of introduction as an ornamental plant in northeastern India in the mid nineteenth century, it has spread throughout Southeast Asia, into parts of Oceania (Muniappan and Marutani, 1988; McFadyen, 1989; Waterhouse, 1994a), and into West and Central Africa (Gautier, 1992; Prasad et al., 1996). A different form of C. odorata (see below), rst recorded as naturalized in the 1940s (Hilliard, 1977), has invaded a large part of the subtropics of southern Africa (Goodall and Erasmus, 1996). Individual C. odorata plants are easily controlled by chemical and/or mechanical means. However, as it is a weed mainly of the tropics and subtropics, many of the countries in which it is a problem do not have the resources to implement comprehensive control programs using conventional methods. Consequently, biological control has become an important management tool (Goodall and Erasmus, 1996; McFadyen, 1996a). Research into the potential of biological control for C. odorata was initiated in the 1960s, when a survey of phytophagous insects on C. odorata, and the host-specicity testing of selected species, was conducted in Trinidad by the Commonwealth Institute for Biological Control, nanced by the Nigerian Institute for Oil Palm Research (Cruttwell, 1972, 1974).

8.2 Taxonomy Chromolaena odorata belongs to the Asteraceae (Compositae), a large, well-dened and highly evolved family (Toelken, 1983; Bremer, 1994; APG II, 2003). Members of the Asteraceae occur throughout the world, and are particularly abundant in the Americas, where the family may have originated (Bremer, 1994). There are many ornamentals and only a few crop plants in the Asteraceae (Toelken, 1983).
Biological Control of Tropical Weeds using Arthropods, ed. R. Muniappan, G. V. P. Reddy, and A. Raman. Published by Cambridge University Press. Cambridge University Press, 2009.

130

Chromolaena odorata

131

Chromolaena odorata is a member of the Eupatorieae, a well-dened, largely New World tribe within the subfamily Asteroideae (King and Robinson, 1987). Eupatorium included over 1200 species before it was split by King and Robinson (1970), following which Chromolaena now includes more than 165 species, all from South and Central America and the West Indies (King and Robinson, 1987). Within the genus, C. odorata, C. ivaefolia (L.) King & Robinson and C. laevigata (Lam.) King & Robinson are widespread and occasionally weedy in the Americas, but only C. odorata has become a serious weed in the Old World (McFadyen, 1989). A minor infestation of C. squalida (DC.) King & Robinson from Brazil was recently found in northern Australia (Waterhouse, 2003), but this species does not appear to be particularly invasive. In addition to C. odorata, several other invasive species were transferred from Eupatorium: Ageratina adenophora (Spreng.) King & Robinson, A. riparia (Regel) King & Robinson, Fleischmannia microstemon (Cass.) King & Robinson, Austroeupatorium inulaefolium (H.B.K.) King & Robinson, Campuloclinium macrocephalum (Less.) DC. and Praxelis clematidea (Griseb.) King & Robinson (Anonymous, 1983; King and Robinson, 1987; Julien and Grifths, 1998; Henderson, 2001; Waterhouse, 2003). Mikania micrantha Kunth., A. adenophora and Ageratum conyzoides L., invasive weeds in the Old World, are also in the Eupatorieae. The two main invasive forms of C. odorata, that occurring in Asia and West Africa and that occurring in southern Africa, differ from one another in morphology, biology, and ecology, and there is little variation within each form (Kluge, 1990; Lanaud et al., 1991; Scott et al., 1998; von Senger et al., 2002; Ye et al., 2004). Thus, they are functionally distinct entities, and have been characterized as biotypes (Zachariades et al., 2004). It is important that these differences be considered when comparing studies on one invasive biotype with the other.

8.3 Morphology, biology, and phenology Chromolaena odorata is a scrambling perennial shrub, with straight, pithy, brittle stems which branch readily, bear three-veined, ovate-triangular leaves placed oppositely, and with a shallow, brous root system (Holm et al., 1977; Henderson, 2001). Capitula are borne in panicles at the ends of the branches and are devoid of ray orets. The corollas of the orets vary between plants from white to pale blue or lilac. Achenes are black with a pale pappus (Holm et al., 1977; McFadyen, 1989). In open-land situations, C. odorata grows to 23 m in height, but it can reach up to 510 m when supported by other vegetation. Within its native range, C. odorata shows marked morphological variability in terms of ower color, leaf shape and hairiness, smell of the crushed leaves, and plant architecture. In some regions, several forms and their intermediates co-occur, while in others, the population appears homogeneous; the basis for this variability presently remains unexplained (Zachariades et al., 2004). In contrast, the biotype invasive in Asia, Oceania, and West Africa is uniform in its morphological features, and has pale bluelilac owers, fairly hairy, dull-green leaves and

132

C. Zachariades et al.

stems, and a lax habit. The southern African biotype of C. odorata is distinct from this more widespread biotype, having glabrous stems and leaves, which in consequence are bright yellow-green when young. The plant has a more upright growth habit, white owers and the smell emitted by the crushed leaves is sharp when compared with that of the AsianWest African biotype. The tropical and subtropical areas in which C. odorata grows as either a native or an invasive species are generally characterized by a dry season with shorter days and a rainy season with longer days. At the start of the wet season, established plants generate new shoots from the crown or from higher, undamaged axillary buds, while seeds in the soil, produced during the previous dry season, germinate (McFadyen, 1988, 1989). In its invasive range, several thousand C. odorata seeds may germinate per square meter, but subsequent high mortality of seedlings (Yadav and Tripathi, 1981; Epp, 1987) and selfthinning of older plants (Witkowski and Wilson, 2001) occurs. Under moist conditions, C. odorata branches trailing on the ground may root (Gautier, 1993), but vegetative reproduction does not contribute signicantly to its pest status. Plants grow vigorously throughout the wet season and owering is initiated by a decrease in both day length and rainfall (Sajise et al., 1974; Gautier, 1993). Flowering peaks in DecemberJanuary in the northern hemisphere and JuneJuly in the southern hemisphere. The onset of owering coincides with the cessation of vegetative growth. Flowering is often prolic, and in both the native and invasive ranges, fertile seed is produced without pollination, as the species is apomictic (Coleman, 1989; Rambuda and Johnson, 2004). rez, However, the owers of C. odorata are also visited by insects (Ghazoul, 2004; Ram 2004) and C. odorata is an important honey plant in Thailand (Thapa and Wongsiri, 1997). Large numbers of seeds (in South Africa, one plant produced over one million seeds in a single season; Liggitt, 1983) are set in less than two months after owering (Erasmus, 1985). In South Africa, seed production increased until plants were about 10 years old and declined dramatically after 15 years, with senescence (Witkowski and Wilson, 2001). Seeds are dispersed by wind, as well as via animal fur, clothing, and vehicles (Gautier, 1992, 1993; Blackmore, 1998). Germinability in the southern African biotype is initially low, but increases over a 6-month period (Erasmus and van Staden, 1987). Most seed loses its viability after a year, although a small proportion of the seed persists in the soil for several years, allowing for rapid recolonization after the removal of a parent population (Witkowski and Wilson, 2001). Once plants have owered and seeded, the terminal, owerbearing parts of the stems die, and in seasonally drier areas, the leaves wither and fall.

8.4 Distribution 8.4.1 Chromolaena odorata in the neotropics In the Americas, C. odorata occurs from USA (southern Florida and Texas) (30 N) to north-western Argentina (about 30 S) (Fig. 8.1) (Gautier, 1992; Kriticos et al., 2005), almost continuously wherever the habitat and climate are suitable. It is present on all the

Chromolaena not reported present Chromolaena introduced

Chromolaena native

Fig. 8.1 Native and reported introduced range of Chromolaena odorata, based on presence records for each country.

134

C. Zachariades et al.

islands of the Caribbean. To the west of the Andes, its range extends as far south as northern Peru (Gautier, 1992). Plants that are similar to the southern African biotype have only been found on the northern Caribbean islands, particularly Jamaica and Cuba (Zachariades et al., 2004). 8.4.2 Chromolaena odorata in the Old World Chromolaena odorata is present in the majority of countries in the humid tropics and subtropics of the Old World (Fig. 8.1). Gautier (1992) and McFadyen (1989, 1996b) have discussed the invasive pathways of C. odorata in the Old World. The rst record of naturalization was in the 1870s in Dacca and the Ganges oodplain (present-day India and Bangladesh). However, the plant was probably introduced into Asia as an ornamental plant in the early 1840s, through the Botanical Gardens in Kolkotta (Calcutta, India). By the early twentieth century, it was widespread in Assam, Bengal (India and Bangladesh), Myanmar (Burma) (Rao, 1920) and present in Thailand (Gautier, 1992). The plant was recorded in Vietnam and Laos in the 1930s (Gautier, 1992). It was introduced as an ornamental to Peradeniya, Sri Lanka, in 1884 and it had naturalized in Sri Lanka by the 1930s (Grierson, 1980). Chromolaena odorata spread eastwards into China and south into Indonesia, especially during World War II (McFadyen, 2002), although it was present in both of these countries in the 1930s. Within Indonesia, C. odorata spread was aided by the transmigration program in the 1960s (McFadyen, 2002). The weed was rst recorded in Nepal in the 1950s, the Philippines and Papua New Guinea in the 1960s (Henty and Pritchard, 1973) and in Timor and Taiwan in the 1980s (Wu et al., 2004; Lai et al., 2006). It was introduced to Guam in the 1960s (Stone, 1966) and it had spread to most of the Micronesian islands by 2000 (Muniappan et al., 2004) (Fig. 8.1). In Australia, C. odorata was found along the Tully River and at Bingil Bay in northern Queensland in 1994 (Waterhouse, 1994a) and at several other localities subsequently (McFadyen, 2004a). Two morphological forms were found in Australia. The more widespread form matched genetically with the AsianWest African biotype, whereas the more localized form matched with material from southern Brazil (Scott et al., 1998). Another species, C. squalida (Waterhouse, 2003) was also found in the vicinity, implying that all three taxa may have been imported with fodder seed from Brazil (Waterhouse and Zeimer, 2002). Chromolaena odorata appeared in West Africa much later than in Asia. It was probably accidentally introduced into Nigeria in 1937, through imported seeds of Gmelina te arborea Roxb. (Verbenaceae) from Sri Lanka (Ivens, 1974). It was recorded in Co dIvoire in the early 1950s and these two nodes of infestation coalesced. Although it is unclear whether separate nodes of infestation appeared in Cameroon and the Central African Republic in the 1930s or whether the plant spread to these countries from Nigeria in the 1960s, the infestation throughout West and Central Africa is represented by the AsianWest African biotype alone. By the mid 1990s, C. odorata had been recorded from

Chromolaena odorata

135

Guinea in the west, south into northern Angola, and east into the central parts of the Democratic Republic of the Congo (Hoevers and MBoob, 1996). Recent records include Chad (Timbilla, 1998), Burkina Faso and The Gambia (CAB International, 2004) and most recently, it has been reported to be common in western Kenya and present in Tanzania (B. le Ru, and Q. Mann, personal communications). In the mid nineteenth century, C. odorata from Jamaica appeared in a list of plants growing in the Cape Town Botanic Gardens, South Africa, and it was found naturalized around Durban in the late 1940s (Hilliard, 1977; Zachariades et al., 2004). From there, it spread rapidly along the coast and is now present from Port St. Johns in the south, where it has probably reached its ecological limit, up into southern Mozambique, and inland through Swaziland into northern South Africa (Goodall and Erasmus, 1996; Macdonald et al., 2003). A single specimen collected from northern Zimbabwe in the late 1960s (Gautier, 1992) appears identical to the AsianWest African biotype (C. Zachariades, unpublished data) and may be linked to an unconrmed C. odorata infestation in northern Mozambique and southern Malawi (J. Findlay, personal communication). If this is true, Mozambique is the only country with both the southern African and AsianWest African biotypes of C. odorata. The weed was introduced to Mauritius before 1949, probably from Asia. 8.4.3 Ecology and impacts In its native range, C. odorata is a common species, growing from the coast to an altitude of 10001500 m (McFadyen, 1988, 1991), although it has been collected at a maximum altitude of almost 3000 m (Gautier, 1992). In Central and South America, C. odorata is common in areas with a rainfall exceeding 1500 mm per annum (McFadyen, 1991). In Trinidad, where the only native-range ecological studies were conducted, the plant prefers well-drained soils, tending to die under waterlogged conditions (Cruttwell, 1972). It grows best in sunny, open areas such as roadsides, abandoned elds, pastures, and disturbed forests, but it tolerates semishade conditions as well. It does not thrive under the shaded conditions of undisturbed forest or in closely planted, well-established orchards. In the Americas, C. odorata is only weedy on occasion, presumably because its natural enemies keep it under control. It acts as a pioneer plant, growing to high densities in recently disturbed (e.g. slashed, overgrazed) areas, but it is soon outcompeted by successional vegetation and disappears after a few years (Cruttwell, 1972; McFadyen, 1988, 1989). The AsianWest African biotype was predicted to invade areas in the Old World with a minimum annual rainfall of 1200 mm (McFadyen, 1989) but has a considerably lower limit (Gautier, 1992; Timbilla, 1998; Kriticos et al., 2005). It does not thrive in areas with extremely high, year-round rainfall, such as the coastal areas of Myanmar (Burma) (Kriticos et al., 2005). It is common up to 1000 m asl in Asia (McFadyen, 1989) and has been recorded at 2000 m in Cameroon (Timbilla, 1998).

136

C. Zachariades et al.

The southern African biotype is more cool tolerant (Kriticos et al., 2005), occurring up to the 31st parallel in frost-free zones with an annual rainfall of 5001500 mm, although at the lower rainfall limits it is restricted to drainage lines and watercourses (Goodall and Erasmus, 1996). In Swaziland, it was found as high as 850 m asl (Goodall et al., 1994). However, it continues to spread northwards in Africa, and thus may also be tolerant of more tropical conditions. In the Old World, C. odorata is an invasive transformer species (sensu Richardson et al., 2000), at least partly because it lacks natural enemies. It grows rapidly and often forms a dense scrambling thicket that grows through and over the existing vegetation. It most readily invades areas of natural or human-induced disturbance, but can invade undisturbed land. Chromolaena odorata affects both subsistence and commercial agriculture, including crops and plantations (e.g. oil palm, rubber, coffee, cacao, coconuts, cashews, cassava, yam, banana, plantain), grazing lands, and silviculture (e.g. teak, pines, eucalypts). Chromolaena odorata invades a wide range of natural vegetation types, from grassland through savanna, bush, open woodland and forest margins and gaps (Goodall and Erasmus, 1996; Prasad et al., 1996). Natural forests are not usually invaded by C. odorata due to its high light requirements, but forest degradation allows the weed to establish (Norbu, 2004), suppressing the recruitment of trees. Forest gaps that naturally develop through tree-fall are colonized rapidly by C. odorata (Epp, 1987; Goodall and Erasmus, 1996). The plant scrambles up through the surrounding trees and emerges on top of the canopy, eventually causing its collapse (Goodall and Erasmus, 1996). However, removal of C. odorata allows rapid regeneration of indigenous forest (Honu and Dang, 2000). In cleared elds infested by C. odorata, indigenous forest was able to regenerate and outcompete the weed if an adequate seed bank of tree species remained (de Rouw, 1991). Chromolaena odorata impacts on biodiversity, the conservation thereof, and ecotourism (Macdonald and Frame, 1988; Goodman, 2003). Thickets of the weed prevent the free movement of livestock and wildlife. In South Africa, growth of C. odorata along riverbanks interfered with the egg-laying of Nile crocodile and altered the sex ratio in the progeny through shading of nests (Leslie and Spotila, 2001). In Thailand, C. odorata attracted butteries away from an indigenous tree, resulting in reduced buttery pollination (Ghazoul, 2004). The weed affects human livelihoods, both through its impacts on agriculture and, in areas with a distinct dry season, because it is a re hazard (Holm et al., 1977; Liggitt, 1983; Macdonald, 1983; Muniappan and Marutani, 1988; Goodall and Erasmus, 1996; Hoevers and MBoob, 1996; McWilliam, 2000). The dry pithy stems and leaves are rich in oils (Moni and Subramoniam, 1960) and burn readily (Hoevers and MBoob, 1996; McFadyen, 1989), although the plants ammability is contested (Goodall and Erasmus, 1996). Dense C. odorata infestations often represent an increased fuel load compared with the native vegetation, resulting in res of increased intensity (McFadyen, 2004b). These cause considerable damage to the surrounding native vegetation and give the resprouting C. odorata plants a further competitive advantage. In South Africa, the native vegetation growing along forest margins is of low ammability and thus protects the

Chromolaena odorata

137

forest interior; once this vegetation is replaced by C. odorata, re is able to penetrate the forest margin, causing progressive erosion of forest edges (Macdonald, 1983). Regular burning of grassland and savanna reduces the establishment of both the Asian West African and the southern African biotypes, although the AsianWest African biotype appears more resistant to re than the southern African one (Macdonald and Frame, 1988; Gautier, 1996; Goodall, 2000). Both biotypes are reported to resprout from the crown after re (Macdonald, 1983; Muniappan and Marutani, 1988), but re-induced mortality rates for southern African plants are likely to be higher. Chromolaena odorata contains a diverse range of secondary chemicals, including avonoids, terpenoids, and alkaloids (Talapatra et al., 1974; Biller et al., 1994). The plant is generally not eaten by vertebrate herbivores, and the high nitrate levels in tender foliage could be the cause of livestock death (Sajise et al., 1974). Pyrrolizidine alkaloids in the owers killed goats which ate the owers (McFadyen, 2004b). Despite the toxic properties of some plant parts, C. odorata leaves appear to be of high nutritive value (Apori et al., 2000). The allelopathic properties of the weed (Sahid and Sugau, 1993) aid it in gaining dominance in vegetation, and in replacing other aggressive invaders such as Lantana camara L. (Verbenaceae) (R. Muniappan, personal observation) and Imperata cylindrica (L.) Beauv. (Poaceae) (Eussen and de Groot, 1974; Ivens, 1974) in Asia and Africa. 8.5 Neotropical arthropods released on C. odorata in its invasive range Cruttwell (1972, 1974) recorded approximately 225 phytophagous insects and mites on C. odorata, mainly in Trinidad but also in other parts of its native range. Of these, a few were identied as being restricted to C. odorata, or to a few close relatives. Following host range testing in Trinidad, Pareuchaetes pseudoinsulata Rego Barros (Lepidoptera: Arctiidae), Mescinia nr. parvula (Zeller) (Lepidoptera: Pyralidae), Apion brunneonigrum Beguin-Billecoq (Coleoptera: Curculionidae) and the mite Acalitus adoratus Keifer (Acarina: Eriophyidae) were sufciently host specic to be recommended as biological control agents (Bennett and Cruttwell, 1973; Cruttwell, 1973a, 1977a, b), whereas Dichomeris n. sp. (Lepidoptera: Gelechiidae) had a wide host range (Cruttwell, 1973b). Following Gagne (1977), Cock (1984) recommended that some of the Cecidomyiidae could be considered as biological control agents. Reviews of agents released and established on C. odorata worldwide are available in Waterhouse (1994b), Julien and Grifths (1998) and Muniappan et al. (2005), while McFadyen (1996a) reviewed the biological control programs that have been implemented against C. odorata. Seven species of biological control agents are currently established, either intentionally or by accident, in the Old World (Fig. 8.2). In exploratory trips to the Americas from the late 1980s to the present, personnel of the South African Agricultural Research Councils Plant Protection Research Institute (ARCPPRI) found several more insect species feeding on C. odorata, some of which were imported into quarantine in South Africa for rearing and host-range testing (Zachariades et al., 1999; Strathie and Zachariades, 2004). In this section, we consider those species of

Distribution of agents No agents present A. adoratus

C. connexa P. pseudoinsulata P. insulata and C. eupatorivora A. adoratus and C. connexa (Palau not visible at map scale) A. adoratus, C. connexa, P. pseudoinsulata, A. anteas and A. thalia pyrrha A. adoratus and P. pseudoinsulata (including Tinian Is, CNMI not visible at map scale)

A. adoratus, C. connexa and P. pseudoinsulata (including FSM, Roti & Saipan Is, CNMI & Guam not visible at map scale)

Fig. 8.2 Countries within the introduced range of Chromolaena odorata with biological control agents established either intentionally or accidentally.

Chromolaena odorata

139

arthropods that have become established on C. odorata outside its native range, through either deliberate introduction as biological control agents or accidental introduction.

8.6 Established biological control agents 8.6.1 Pareuchaetes pseudoinsulata Rego Barros (Lepidoptera: Arctiidae) The native range of P. pseudoinsulata is small, consisting of eastern Venezuela and Trinidad (Cock and Holloway, 1982). Initially identied as Ammalo insulata (Walk.), this butter-yellow moth lays its eggs in batches on the undersides of C. odorata leaves. The hairy caterpillars, which are black with red and white markings, feed on leaves, initially gregariously but later solitarily. Feeding damage by older larvae is characteristic, often leaving only the midrib. The larvae are nocturnal and the larger caterpillars shelter during the day in leaf litter under the plant, where they eventually pupate (Cruttwell, 1972). Until recently, P. pseudoinsulata was the most widely released and established biological control agent on C. odorata (Waterhouse, 1994b; Julien and Grifths, 1998; Muniappan et al., 2005). In the early 1970s, it was released in Ghana, Nigeria, India, Sri Lanka, and Malaysia (Sabah). However, it established only in Sri Lanka and Malaysia and was effective only in Sri Lanka, where it inicted widespread but sporadic defoliation (Dharmadhikari et al., 1977; Waterhouse, 1994b). In the 1980s, the moth was rereleased in India, where it established in the southwest, causing sporadic damage (Joy et al., 1993; Waterhouse, 1994b). It was also released in Guam, where it established and controlled C. odorata to such a degree that the status of the weed could be downgraded (Muniappan et al., 1989; Seibert, 1989). It was released on the islands within the Commonwealth of the Northern Marianas (Rota, Tinian, Saipan, and Aguijan), where it established and caused substantial damage. It failed to establish in Vietnam, Thailand, and South Africa (Julien and Grifths, 1998). The moth was discovered in Brunei and on several islands in the Philippines in the 1980s. Because it had not been released there, it may have spread from Malaysia (Waterhouse, 1994b). In the late 1980s and early 1990s, P. pseudoinsulata was released and established on the islands of Yap, Pohnpei, and Kosrae within the Federated States of Micronesia (FSM). In the 1990s, following successes in other countries, it was rereleased in Ghana (1991 1993), where it established, spread and inicted substantial damage (Timbilla, 1996, te dIvoire, where it again failed to establish (Timbilla et al., 2003), 1998) and in Co despite initial positive results (Julien and Grifths, 1998). Releases in Indonesia resulted in establishment in Sumatra, Kalimantan (although this population may have moved from Sabah), and Sulawesi, causing substantial but sporadic damage, but it did not establish in Java or West Timor (Julien and Grifths, 1998; Desmier de Chenon et al., 2002a). It was released in Papua New Guinea (PNG), where it established in only the Markham Valley (Bofeng et al., 2004) and in South Africa, where it again failed to establish (Strathie and Zachariades, 2002). In the 2000s, it was established on Chuuk (FSM) and released in Palau (Muniappan et al., 2005).

140

C. Zachariades et al.

Pareuchaetes pseudoinsulata has proved a difcult and unpredictable biological control agent, in terms of both establishment ease and subsequent effectiveness in controlling the weed. In some instances, the insect established after small releases; just over 2000 larvae in Sri Lanka (Dharmadhikari et al., 1977) and 500 mated adults on Guam (Seibert, 1989), whereas in others, extremely large releases were necessary: 125 000 at one site in Ghana (Timbilla, 1996). In South Africa, about 350 000 larvae were released at two sites within a year, but the insect failed to establish at either site (Strathie and Zachariades, 2002). Only in a few countries, such as Guam, was there a signicant long-term reduction in the population of C. odorata. In many of the countries or regions where P. pseudoinsulata did establish, including Sri Lanka, the Marianas, India, Ghana, and Sumatra, initial establishment was followed by a spectacular population build-up and widespread, complete defoliation of C. odorata thickets, and high mortality rates of these plants (Dharmadhikari et al., 1977; Seibert, 1989; Timbilla, 1996; Singh, 1998). However, after 12 years, the populations of the moth often declined dramatically and the weed recovered. Subsequent sporadic outbreaks resulting in defoliation occurred, but these were often less spectacular and were unpredictable over time and space. Such population dynamics are typical of many outbreak or epidemic insect species (Wallner, 1987). Tyria jacobaeae (L.) (Lepidoptera: Arctiidae), which feeds on Senecio jacobaea L. (Asteraceae), is one of only two other arctiid species established as a weed biological control agent (Julien and Grifths, 1998). It appears to have a similar impact on its target weed as do Pareuchaetes spp. on C. odorata, with periodic outbreaks causing massive defoliation but an overall unsatisfactory effect (Julien and Grifths, 1998). Native populations of P. pseudoinsulata in Trinidad (Cruttwell, 1972) were low in most areas, with outbreaks only in localized areas during the wet season. The population dynamics also varied from year to year. An initial rapid rise in population at one of these outbreak sites was followed by a population plateau and then a slow decline, coinciding with the appearance of parasitoids and disease. However, Cruttwell (1972) found a correlation between numbers of adults caught in light traps and rainfall levels 16 weeks earlier, and believed that the drop in population each year was due to a loss of plant quality due to cessation of rains, rather than parasitism or disease. Similar population dynamics have been reported for T. jacobaeae within its native range (Dempster, 1971). A number of possible factors, including predation and parasitism, climate, and biotype matching, may explain why establishment of P. pseudoinsulata has been erratic. In areas where P. pseudoinsulata has established, factors affecting its abundance may include weather patterns, natural enemies (predators, parasitoids, disease) and herbivory-induced changes in host-plant quality (Cock and Holloway, 1982; Waterhouse, 1994b; Marutani and Muniappan, 1991). Such factors have been implicated in the population dynamics of other outbreak species (Wallner, 1987). In India and Ghana, predators (particularly ants) were implicated in the nonestablishment of P. pseudoinsulata in the 1970s, although evidence for this was anecdotal (Cock and Holloway, 1982). Ants and other invertebrate predators removed a high

Chromolaena odorata

141

proportion of P. pseudoinsulata eggs placed in the eld in South Africa (Kluge, 1994). A lack of predation may partly explain the relatively high success, both in terms of establishment and subsequent population increase, of releases of smaller numbers of P. pseudoinsulata on islands, where predation pressures are often lower (MacArthur and Wilson, 1967; Seibert, 1989). Although predation of P. pseudoinsulata by both invertebrates and vertebrates (toads, skinks, birds) was observed in Guam, Sri Lanka, and Pohnpei (Seibert, 1989; Waterhouse, 1994b), the effects of this on the moths population were not assessed. The success of releases in some areas on continents illustrates that if Allee effects exacerbated by predation and dispersal are a factor in poor establishment, they can be overcome by the release of sufcient insects over a sufcient duration of time. Levels of parasitism of P. pseudoinsulata remain low in its introduced range, even after many years in the eld, compared with levels of parasitism in its native range (McFadyen, 1997), although a parasitoid caused up to 30% mortality on Guam (Seibert, 1989). An unidentied tachinid has been found depositing eggs on larvae in PNG, but has not prevented outbreak populations of P. pseudoinsulata occurring on a seasonal basis. Pareuchaetes pseudoinsulata is highly prone to diseases such as microsporidia under laboratory conditions, and the release of unhealthy insects may have prevented their establishment in some instances (Singh, 1998; Strathie and Zachariades, 2002). A nuclear polyhedrosis virus caused substantial mortality in Trinidad (Cruttwell, 1972), but the culture was cleared of this on importation into India in the early 1970s (Waterhouse, 1994b). A culture of P. pseudoinsulata from Sri Lanka was released in India and appeared to be doing well until it died out due to a virus (Waterhouse, 1994b). Climate has been viewed as affecting the establishment of many biological control agents (Cock and Holloway, 1982; Day et al., 2003; Zalucki and van Klinken, 2006). A long, severe dry season may have prevented establishment of P. pseudoinsulata in some areas (Cock and Holloway, 1982; Seibert, 1989). The only sites in PNG where P. pseudoinsulata established were in the relatively dry areas of the Markham Valley (Bofeng et al., 2004). South Africa is considerably cooler than Trinidad (Parasram, 2003) or other areas where P. pseudoinsulata has established in the Old World, and this may have contributed to the lack of establishment there. 8.6.2 Pareuchaetes insulata (Walker) (Lepidoptera: Arctiidae) Pareuchaetes insulata occurs from Colombia through Central America to Texas, Florida, and on the larger Caribbean islands (Cock and Holloway, 1982). A culture of P. insulata, collected from C. odorata in Florida, was introduced into quarantine in South Africa in 1989 and was shown to be adequately host specic for release in South Africa (Kluge and Caldwell, 1993a). This species was eventually released from 2001 onwards in KwaZuluNatal (KZN) province (Strathie and Zachariades, 2004), and establishment was conrmed in 2004 at one site (Zachariades and Strathie, 2006). The insect has subsequently spread along the wetter coastal belt and is causing high levels of localized damage to C. odorata. Its biology is similar to that of P. pseudoinsulata. It is thought that this species succeeded

142

C. Zachariades et al.

in South Africa where P. pseudoinsulata did not because of better climate matching (Byrne et al., 2004). Although the form of C. odorata from which P. insulata was collected in Florida differs from the southern African biotype, the host range of the insect was sufciently broad to allow establishment in South Africa. 8.6.3 Cecidochares connexa Macquart (Diptera: Tephritidae) This tephritid y was recorded on C. odorata in Trinidad, Mexico, and Bolivia (Cruttwell, 1974). It lays eggs in the shoot tips of the plant and the larvae develop within galls at the nodes. A strain of y compatible with the AsianWest African C. odorata biotype was imported from Colombia into quarantine at the Indonesian Oil Palm Research Institute in North Sumatra under an Australian Centre for International Agricultural Research program in 1993, and host-range testing indicated its high specicity to C. odorata (McFadyen et al., 2003). The y was released in North Sumatra in 1995, and later on other Indonesian islands (Tjitrosemito, 1998; Desmier de Chenon et al., 2002a; Wilson and Widayanto, 2004) where it established readily. Cecidochares connexa was subsequently released and established in PNG (Orapa and Bofeng, 2004), Palau (Esguerra, 2002), Guam, Saipan, Rota and Pohnpei (Muniappan et al., 2005; Cruz et al., 2006), Chuuk, Kosrae and Yap (Muniappan et al., 2007), East Timor (T. Paul, Charles Darwin University, personal communication), and in India (Bhumannavar and Ramani, 2007). Although it was released in Thailand (McFadyen et al., 2003), it did not establish (Muniappan et al., 2005). Attempts to culture it in quarantine on the southern African C. odorata biotype failed, probably because of insectplant incompatibility (Zachariades et al., 1999). Cecidochares connexa was sent to Ghana in the late 1990s but no culture was established in quarantine (J. A. Timbilla, personal communication). The establishment of C. connexa in West Africa remains a priority, as it will enhance the effects of P. pseudoinsulata there. The y can be established easily from small founder populations. About 100 pairs were released at each of several sites across Indonesia (Wilson and Widayanto, 1998, 2002; Tjitrosemito, 1998), but as few as 26 individuals on Palau (Esguerra, 2002). Once established, y populations can rapidly build up and spread (Desmier de Chenon et al., 2002a; Wilson and Widayanto, 2002). In Sumatra, the y spread over 200 km within 5 years following release (Desmier de Chenon et al., 2002a), while in PNG it spread over 100 km within 4 years (Day and Bofeng, 2007). The y has been so effective, particularly in areas at lower elevations with a short dry season, that McFadyen et al. (2003) and Day and Bofeng (2007) have found successful biological control of C. odorata in many areas. Hundreds of galls per plant (>10 galls/m of stem) have sometimes been recorded, causing plants to become prematurely moribund with few owers or seeds, or to die back (Desmier de Chenon et al., 2002a; Wilson and Widayanto, 2002; McFadyen et al., 2003; Day and Bofeng, 2007). The galls have been shown to act as a resource sink, and to slow the growth of the plant (Cruz et al., 2006). Plant density has subsequently decreased on Lombok, in Irian Jaya (Wilson and

Chromolaena odorata

143

Widayanto, 2002) and in most PNG provinces where C. odorata is present (Day and Bofeng, 2007). The y appears less effective in higher-elevation, cooler areas and those with a longer dry season, as fewer generations are produced per annum and therefore populations build up slowly. Areas with a long, intense dry season are often burned at that time of the year, eliminating the y population locally. Initially, there was concern that parasitism could reduce the effectiveness of C. connexa, as happened with Procecidochares utilis Stone (Diptera: Tephritidae) on A. adenophora in many areas where it was released (Julien and Grifths, 1998). However, parasitism and predation levels have generally remained low, although up to 50% parasitism was recorded at one site in Java (McFadyen et al., 2003). 8.6.4 Actinote species (Lepidoptera: Nymphalidae) Larvae of Actinote feed on the leaves of species largely within the Asteraceae in Central and South America. The host plants of several Actinote species are in the taxa Chromolaena and Mikania, and some Actinote species feed on species within both genera (Ackery, 1988). Identication of Actinote species is difcult and the taxonomy of the group is uncertain (Francini and Penz, 2006). Cruttwell (1974) recorded A. anteas (Doubleday & Hewitson) from C. odorata in Trinidad, and it is also recorded from Ageratum and Mikania there (Desmier de Chenon et al., 2002b). The host range of a culture of A. anteas from Costa Rica was tested in quarantine in South Africa, but the culture soon died out (Caldwell and Kluge, 1993). The same species, also from Costa Rica, was introduced into Indonesia in 1996 where it was subsequently approved for release following host-specicity studies (Desmier de Chenon et al., 2002b). A species initially identied as A. parapheles Jordan and later as A. thalia pyrrha Fabr. was imported from northeastern Brazil into South Africa in 1995. Host-specicity tests indicated that it fed to the same extent on two species of Mikania indigenous to South Africa as it did on C. odorata, and it was therefore not released (Zachariades et al., 2002). A culture of this species was forwarded to Sumatra in 1999 to replace the culture of A. anteas that had been approved for release there, and it was released in the same year (Desmier de Chenon et al., 2002b). In Indonesia, the neotropical vine M. micrantha is as great a problem as C. odorata, and the oligophagous feeding habits of Actinote species were considered desirable, despite the presence of the native M. cordata (Burm. f.) B. L. Rob. Actinote thalia pyrrha is now widely established and common in Sumatra, where it not only feeds on C. odorata and M. micrantha but also the invasive alien Austroeupatorium inulaefolium (Desmier de Chenon et al., 2002a, b; Zhigang et al., 2004; R. Desmier de Chenon, A. Simamora and Nirwanto, Indonesian Oil Palm Research Institute, personal information, 2006). A second species, initially identied as A. thalia thalia L., was collected in Venezuela and tested in South Africa, but it also fed on Mikania indigenous to South Africa and so was not approved for release. It was also forwarded to Sumatra where it was released (Desmier de Chenon et al., 2002b). However, it has only established at two sites in

144

C. Zachariades et al.

Indonesia and is considered not as successful as A. thalia pyrrha (R. Desmier de Chenon, A. Simamora, and Nirwanto, personal information, 2006). 8.6.5 Calycomyza eupatorivora Spencer (Diptera: Agromyzidae) The larvae of this agromyzid y form blotch mines on the leaves of C . odorata. Cruttwell (1974) recorded it in Trinidad, but at that stage it was considered the same species as Calycomyza avinotum Frick, a nearctic species with a wide host range. Calycomyza eupatorivora has also been recorded from Jamaica, Hispaniola, Argentina, Venezuela, and Guadeloupe (Martinez et al., 1993), and it or similar species are widespread in the neotropics (ARC-PPRI, unpublished data). A culture from Jamaica was rst reared in quarantine in South Africa in 1997, and host range testing was completed in 2001, showing the y to be highly host specic (Zachariades et al., 2002). Releases were initiated in 2003 and establishment conrmed the following year at one site on the KGN coast (Zachariades and Strathie, 2006). The insect has since spread at least 40 km along the coast, and has now been widely released on C. odorata in South Africa. However, it is unlikely that C. eupatorivora is having much effect on C. odorata yet. The y displays a preference for young plants in shady conditions, and is likely to be effective on these plants only. It is common in Jamaica and a pest under laboratory conditions, but it is rare on the South American mainland. In South Africa, a congeneric leaf miner on lantana is heavily parasitized, which may contribute to its lack of effectiveness there (Baars and Neser, 1999). 8.6.6 Acalitus adoratus Keifer (Acarina: Eriophyidae) Feeding by this eriophyid mite on the leaves of C. odorata induces abnormal growth of the epidermal hairs, resulting in formation of a white erineum patch (Cruttwell, 1977a). It has been recorded from Brazil, Bolivia, and Trinidad (Cruttwell, 1974). Studies in Trinidad indicated that it was host specic and could be damaging, and it was recommended for release as a biological control agent (Cruttwell, 1977a; Cock, 1984). It was never introduced deliberately, but was found in the eld in the Philippines in 1987, and shortly thereafter in Thailand, Malaysia, and Indonesia (McFadyen, 1995). McFadyen (1995) suggested that it may have been accidentally released into the eld in Sabah, Malaysia, through phoresy on A. brunneonigrum adults imported from Trinidad and released directly into the eld. It now occurs through much of the range of C. odorata in Asia and Oceania (McFadyen, 1995; Muniappan et al., 2005). It is thought not to cause any signicant damage to C. odorata, although this has not been quantied. 8.7 Other arthropod agents A number of other species of arthropods have been released on C. odorata but failed to establish. Some are outlined below.

Chromolaena odorata

145

8.7.1 Apion brunneonigrum Beguin-Billecoq (Curculionidae) The larvae of this weevil develop in the owers of only C. odorata and C. ivaefolia (Cruttwell, 1973a), and it has been recorded from Trinidad, Venezuela, and Argentina (Cruttwell, 1974). Adults were collected in Trinidad and released in Nigeria, Ghana, Sri Lanka, and Malaysia (Sabah) in the 1970s, India between 1972 and 1983, and in Guam in 1984. They persisted in Sabah for one year (Julien and Grifths, 1998). However, A. brunneonigrum has not established anywhere and the reasons for its nonestablishment are unclear.

8.7.2 Mescinia nr. parvula (Zeller) (Pyralidae) ( Phestinia costella Hampson) The larvae of this moth were collected from C. odorata and C. ivaefolia in Trinidad, and similar larvae have been collected from C. odorata in Mexico, Brazil, Jamaica, Cuba, Florida, and Venezuela (Cruttwell, 1974; Zachariades et al., 1999; Strathie and Zachariades, 2004), and from C. hookeriana (Griseb.) King & Robinson in Argentina (Cruttwell, 1974). Solis et al. (2008) re-identied specimens of this moth from Jamaica, Trinidad, Guatemala, and Puerto Rico as Phestinia costella, and expect to occur throughout the native range of C. odorata. Cruttwell (1977b) showed that the moth was highly host specic, but was unable to rear it in the laboratory. Several other attempts to do so were also unsuccessful (Singh, 1998; Zachariades et al., 1999, 2007). Low numbers of insects collected in the eld in Trinidad were released in Guam in 1984, but the moth did not establish (Julien and Grifths, 1998).

8.7.3 Pareuchaetes aurata aurata (Butler) (Arctiidae) This moth has a range covering southern Brazil, Paraguay, Bolivia, and northern Argentina, and was recorded on C. hookeriana in Argentina (Cock and Holloway, 1982). A culture collected from C. hookeriana in Argentina was imported into South Africa, where it was shown to be sufciently host specic to be safely released there (Kluge and Caldwell, 1993b). It was released in South Africa in the early 1990s but did not establish (Zachariades et al., 1999). In addition, several other species have been tested over the years but were found to have too broad a host range for release (Cruttwell 1973b; Zachariades et al., 1999; Strathie and Zachariades, 2002), and several promising agents are currently being investigated or await permission for release (Zachariades and Strathie, 2006). There is still a need for additional agents, particularly in drier parts of the invasive range of C. odorata where the two Pareuchaetes species, C. connexa and C. eupatorivora cannot persist or are less effective. At present, only ARC-PPRI in South Africa is conducting research on new potential agents. This research is focused on (i) agents from the center of origin of the southern

146

C. Zachariades et al.

African C. odorata biotype, in order to ensure insectplant compatibility and (ii) agents with a biology (a distinct diapause and/or a soil-dwelling stage) that will allow them to be effective in seasonally dry, re-prone parts of the invasive range of C. odorata (Zachariades and Strathie, 2006). 8.8 Ecological interactions between the plant and arthropods Few detailed studies have been undertaken so far on interactions between C. odorata and established biological control agents. Only P. pseudoinsulata has been established for any length of time, and many countries using biological control for C. odorata do not have the resources or expertise to conduct such studies. However, the reaction of C. odorata to feeding by P. pseudoinsulata has been well studied in Guam (Marutani and Muniappan, 1991; Raman et al., 2006) and noted in India (Singh, 1998), Indonesia (Desmier de Chenon et al., 2002a), and Ghana (Timbilla, 1996). Typically, after two to three weeks heavy feeding on a plant by P. pseudoinsulata larvae (i.e. under outbreak conditions), all the remaining leaves, both damaged and undamaged, turn from green to yellow and their toughness increases. This reaction is not induced by mechanical damage or feces, but by the feeding action of the larvae (Marutani and Muniappan, 1991), and/or the larval saliva. Chlorophyll in these yellow leaves was much lower than green leaves, while nitratenitrogen levels and water soluble proteins increased. The subcellular physiology of this chlorosis mechanism has recently been described in detail, and involves the accumulation of lipidic materials which act as precursors of compounds that defend leaf cells from the impact of feeding (Raman et al., 2006). In laboratory trials, larvae preferred green leaves over yellow leaves. Larvae only survived on yellow leaves from the third instar onwards, but even then they grew at a slower rate and had a longer development time than those fed on green leaves. In the eld, older larvae feeding on plants with yellow leaves remained on the upper parts of the plant during the day, rather than hiding in the leaf litter as usual. This was probably because they were obtaining insufcient nutrition from the leaves and needed to continue feeding, or to conserve energy, rather than move to the leaf litter each day. This behavior is likely to increase their risk of being preyed upon. About two to three weeks after larval feeding ceased, the leaves returned to their original green color (Marutani and Muniappan, 1991). Yellowing of C. odorata bushes has also been recorded after feeding by A. thalia pyrrha larvae (R. Desmier de Chenon, A. Simamora and Nirwanto, personal information, 2006). In South Africa, the same response by the southern African C. odorata biotype to P. insulata feeding has recently been recorded (C. Zachariades, personal observation).

8.9 Economics of biological control efforts Costs incurred by classical biological control of weeds are reasonably easy to calculate or estimate. These costs include the research stage (identication of the origin of the weed;

Chromolaena odorata

147

surveys for and selection of agents; rearing and testing of agents; release and establishment of agents; measuring their impact) and implementation (mass-rearing and redistribution of agents that have already established at certain sites). The cost-benet ratio of biological control for a given weed, however, is more difcult to measure. It is dependent on (i) how much damage the weed is causing economically, socially and/or in terms of biodiversity, (ii) how effective the agent or suite of agents is, and (iii) the costs and practicalities of other control methods, which is especially relevant in developing countries, where resources are scarce for conventional control and the political will for coordinated, sustained, comprehensive weed control programs is often lacking. The cost of not introducing biological control, in terms of increasing future losses as the weed spreads and increases in density, versus the cost of introducing biological control, is a useful means of estimating the benets of biological control (van Wilgen et al., 2004). We consider two case studies where some costs and benets have been documented: C. odorata in Indonesia and Papua New Guinea (PNG), and in South Africa. An ACIAR-funded project for biological control of C. odorata in Indonesia and the Philippines began in 1993 (McFadyen, 1996a) and in 1998, PNG replaced the Philippines as a partner. The project in PNG was extended twice and ended in June 2007, while the Indonesian project ended in 2002. The total cost of the project was conservatively about AU$2 million. During this time, P. pseudoinsulata, C. connexa, and the two Actinote spp. were introduced into Indonesia and P. pseudoinsulata, C. connexa and C. eupatorivora into PNG. Three of the four agents introduced into Indonesia are having an impact, with the Actinote spp. having an additional benet by helping control M. micrantha and A. inulaefolium (R. Desmier de Chenon, A. Simamora and Nirwanto, personal information, 2006). Desmier de Chenon et al. (2002a) estimated that in plantations in Sumatra, the combined effects of P. pseudoinsulata and C. connexa reduced the cost of maintenance by 75%. In PNG, where a conservative estimate of about AU$1 million was spent, C. odorata is present in 13 provinces (Day and Bofeng, 2007). Calycomyza eupatorivora failed to establish and P. pseudoinsulata is aiding control in only Morobe Province. However, within six years, C. connexa is already controlling C. odorata in most provinces. In several provinces, where C. odorata has been controlled and food gardens re-established, there is already a doubling of income and a 75% reduction in labour costs or time spent weeding. Villagers at Musu in Sandaun Province have reported that income from newly revitalized food gardens has increased by 50 Kina (AU$25) a week for a 1-ha block. In New Ireland and Manus provinces, similar gures have been reported (M. Day and I. Bofeng, PNG National Agricultural Research Institute, unpublished data). While the total area of infestations in PNG is not known, if just 100 ha of food gardens can be revitalized in the 13 provinces following the control of C. odorata, then the cost of the project in PNG can be recovered within eight years. These gures do not include savings in time spent weeding gardens or the social benets of controlling C. odorata (Day and Bofeng, 2007; M. Day and I. Bofeng, unpublished data). In South Africa, C. odorata is mainly a problem for biodiversity. Invasive alien plants have been identied as the major threat to protected areas in KZN, and C. odorata ranks

148

C. Zachariades et al.

high amongst these (Goodman, 2003). Versveld et al. (1998) estimated that 326 139 ha have been invaded by C. odorata in KZN (Marais et al., 2004). In the Kube Yeni Game Reserve in northern KZN, C. odorata infestations were estimated to have reduced the carrying capacity from about 6 ha per large stock unit (LSU 450 kg animal) to more than 15 ha (Goodall and Morley, 1995). In the 2002/3 nancial year, the Working for Water program (WfW) spent about ZAR6.15 million nationally on clearing 3600 condensed hectares of C. odorata, of which just under 50% of the cost was in initial clearing and the remainder in follow-up costs (Marais et al., 2004). Marais et al. (2004) estimated that at the current rates of clearing, it will take 15 years to treat C. odorata at a national level. At costs adjusted to 2006 gures using South African Production Price Index (PPI) (www.statssa.gov.za/keyindicators/ppi.asp), this translates to almost ZAR105 million. However, Marais et al. (2004) acknowledged that this is likely to be an underestimate. Among other assumptions they make, they base their gures on only one follow-up treatment being needed. In addition, van Gils et al. (2004) suggested that the efciency of clearing C. odorata in an indigenous South African forest was low, and suggested that only the use of an alternative plant cover after removal of C. odorata reduced reinfestation substantially. A C. odorata clearing programme that was initiated in the Hluhluwe-iMfolozi Game Reserve in the 2003/4 nancial year (Zachariades and Strathie, 2006) has cost almost ZAR38 million (adjusted to 2006 PPI values ZAR40 million) and has cleared 35 000 ha in the park and surrounds (C. Terblanche, Ezemvelo KZN Wildlife, unpublished data). When we consider the estimated 325 000 ha invaded in 1998 (Versveld et al., 1998), it seems likely that the real clearing costs of C. odorata nationwide will be closer to ZAR400 million. These costs do not take into account the costs that infestations cause to agriculture, ecotourism, or biodiversity. The South African research programme on C. odorata biological control, initiated by ARC-PPRI in 1988, has cost about ZAR16 million, adjusted to 2006 PPI gures. If the research program continues until 2015, with two additional personnel, as is planned, it would cost an additional ZAR14 million in 2006 gures. Therefore, the nal cost will be about ZAR30 million over 28 years. Mass-rearing and release of C. odorata biological control agents (mainly the three Pareuchaetes species) have been conducted by the South African Sugarcane Research Institute (SASRI) and WfW, at an estimated cost of ZAR56 million, bringing the total cost of the biological control project to ZAR36 million. Even if the nal cost for biological control of C. odorata in South Africa runs to ZAR40 million, this represents only 10% of costs for clearing 1998-level infestations of C. odorata nationally. The cost-benet ratio ultimately depends on the effectiveness of biological control. Estimated clearing costs cited above are for reducing C. odorata levels to <5%. Many cost-benet studies of biological control have underestimated benets of biological control because they were conducted too early (McFadyen, 1998) or did not look at the scenario with biological control versus without (van Wilgen et al., 2004). Versveld et al. (1998) estimated that, overall, weed biological control programs in South

Chromolaena odorata

149

Africa had already allowed a 20% saving of up to ZAR1.4 billion on the costs necessary for clearance, and that potential savings would eventually be more than 40%.

8.10 Measures of efciency of biological control A highly efcient biological control program would involve a quick and accurate determination of the origin and biotype of the weed species; a thorough survey of natural enemies in the area of origin, followed by importation, successful breeding and testing of selected potential biological control agents; the successful, easy establishment and spread of the agents, followed by a good level of control of the weed. Factors negatively affecting the success of programs may be intrinsic to the biological system (McFadyen, 1998) (e.g. not many potential agents available, native or valuable plants closely related to the target weed in the country of introduction, poor establishment or impact) and/or extrinsic (e.g. inexperienced researchers, few resources, political). The efciency of C. odorata biological control efforts globally has been substantially enhanced by the IOBC Working Group on Chromolaena odorata, through its regular international workshops, in terms of exchange of information, technology, and biological control agents (Boller et al., 2006). It has also been enhanced by sustained funding at a national level in Ghana (in the 1990s), South Africa, and Micronesia and by international funding by ACIAR for Indonesia, the Philippines, East Timor, and PNG. However, not all international interventions have proved successful. A project funded by the European Economic Community in 19901992 produced limited results due to its short duration, and a UN Food and Agriculture Organization project in West Africa was blocked due to the controversy surrounding the usefulness of C. odorata as a fallow crop (McFadyen, 1996a; Prasad et al., 1996). However, despite these collaborative efforts, the efciency of biological control of C. odorata worldwide has been variable and this can be attributed to both intrinsic and extrinsic factors. We consider two programmes as case studies below. 8.10.1 Biological control of C. odorata in PNG Overall, the biological control of Chromolaena program in PNG has been very efcient, both because the program has been relatively cheap and because the main damaging agent, C. connexa, is easy to establish and very effective. Cecidochares connexa was introduced into PNG in 2001, following testing elsewhere which showed that the agent was host specic. It was only reared in the laboratory in PNG for about two years after initial release, until eld populations were sufciently high to warrant eld collection for redistribution. During the rearing stage of the program, nearly 13 000 galls were released, compared with over 50 000 in the subsequent two years, when the agent was being eld collected for redistribution (Day and Bofeng, 2007). Field release and establishment was very successful, with populations establishing at over 97% of release sites. As few as 50 galls could be released to achieve establishment, so 1 000 eld-collected galls could be

150

C. Zachariades et al.

released over 20 new sites (Day and Bofeng, 2007). This meant that large numbers of new sites could be targeted. As the agent had been tested elsewhere, was easy and cheap to collect and release, and was damaging to C. odorata, there were substantial benets for every dollar spent. 8.10.2 Biological control of C. odorata in South Africa The South African biological control program has faced several obstacles not experienced by other countries where chromolaena is a problem. Firstly, the biotype of chromolaena in South Africa is different from that in other countries and where chromolaena has been controlled. Secondly, much of South Africa where chromolaena is found is cooler than where chromolaena is a problem in other countries so the agents may not be as damaging, or cannot establish due to the different climatic conditions. Cecidochares connexa, which was highly successful in other countries, could not be used in South Africa, due to incompatibility with the southern African C. odorata biotype, and agentplant compatibility problems have also resulted in the failure of the pathogen project to date. Large numbers of the three Pareuchaetes species were mass-reared and released. However, only one of these (P. insulata) established. Several other insect species, most importantly A. thalia pyrrha and Longitarsus sp. (Coleoptera: Chrysomelidae), were tested, found to have broad host ranges and were rejected. The program spent time and resources identifying the small native range of the southern African biotype (the Greater Antilles) and has since begun exploration there for more suitable agents. Some of the recommended agents, for example Melanagromyza eupatoriella Spencer (Diptera: Agromyzidae) and P. costella, have proved difcult to maintain in quarantine for a long enough period for host-range tests to be conducted. One of the current constraints on the program in South Africa is that no insect species with a distinct diapause or soil-dwelling stage have been found in the Greater Antilles (Strathie and Zachariades, 2004). Given the importance of such species for the seasonally dry areas of southern Africa invaded by C. odorata, it has thus been necessary to continue research on potential agents with a suitable biology from regions of South America that are climatically similar to southern Africa. Because the southern African C. odorata biotype is absent from these areas, it is also necessary to ensure that the host range of these insects is broad enough to allow them to establish on infestations of this biotype in South Africa. The difference in the efciencies of the C. connexa program in PNG and the South African program can be explained, in part, through serendipity, as the success of selected species as agents in classical biological control remains highly unpredictable (McFadyen, 1998). In addition, the South African program essentially needed to start de novo because biotype differences make C. odorata in southern Africa a different entity from that which had been studied elsewhere since the 1960s. In addition, South Africa has a cooler climate than many of the areas or regions where C. odorata occurs and has been controlled. As a

Chromolaena odorata

151

result, agents that have worked in the tropics may not work in the cooler regions of South Africa or be suitable to control the southern African biotype.

8.11 Issues of sustainability of the identied biological control efforts Ideally, once an agent is well established and distributed, no further effort should be necessary in redistributing it. The ideal biological control agent would provide fairly consistent control over time and would be able to locate isolated and small infestations of the target weed as they arise. Cecidochares connexa has shown itself to give highly sustainable control, because it spreads quickly and attacks plants consistently. It is good at locating scattered plants and can sustain a population even where C. odorata populations are low. In contrast, P. pseudoinsulata is an outbreak species which is highly unpredictable in terms of where and when it will control C. odorata to adequate levels. In some areas where it was released, initial outbreaks were not matched again, and the insect persists at a low population level. Pareuchaetes pseudoinsulata also tends to do better in large, dense infestations of C. odorata, and is less able to control C. odorata in scattered populations. 8.12 Conclusions Chromolaena odorata is one of the worst weeds in the world, affecting agriculture and biodiversity in the tropical and subtropical regions of the Old World. Its earliest introduction seems to have been in the mid nineteenth century into India, and it has since spread to much of its suitable habitat in Asia, Africa, and Oceania. Some research has been conducted on its biology and ecology, as well as a good number of studies on its allelopathic properties, population dynamics, and impacts on indigenous vegetation. However, signicant research gaps remain. Until recently, studies on the biotype of C. odorata invading Asia, Oceania, and West Africa have been equated with the more localized biotype invading southern Africa. However, the two appear to have different attributes and should functionally be regarded as separate entities. Research on biological control of C. odorata was initiated in the 1960s. The rst insects, Pareuchaetes pseudoinsulata and Apion brunneonigrum, were released in India, Malaysia, Sri Lanka, and West Africa in the early 1970s, but these releases were largely unsuccessful. Releases in various countries continued through the 1970s and 1980s, and P. pseudoinsulata became established in several countries, causing signicant damage in some. In the early 1990s, the IOBC Working Group on Chromolaena odorata was formed, and since then, regular international workshops have been held to discuss the management and biological control of the weed. The ACIAR program in Southeast Asia in the 1990s resulted in the establishment of the highly successful gall y, Cecidochares connexa, throughout Southeast Asia and parts of Oceania. The agent was subsequently released in other countries, including India. The introduction and establishment of C. connexa has allowed a downgrading of the weediness status of C. odorata in many areas with high rainfall.

152

C. Zachariades et al.

In South Africa, biological control research was initiated in the late 1980s and has involved much exploratory work in the Americas. Pareuchaetes insulata and Calycomyza eupatorivora were established in South Africa in the 2000s and are now spreading. South Africa continues to conduct research on several promising insect agents, including species which have a biology which should enable them to persist in areas with severe seasonal drought and res. These agents should also be useful for the drier parts of Southeast Asia such as East Timor. After 40 years of research, C. odorata control is being achieved through the use of biological control, and these gains will be consolidated over the coming decade.

Acknowledgments We thank Ezemvelo KZN Wildlife for permission to use unpublished data on clearing costs in Hluhluwe-iMfolozi park. The rst author thanks the ARC and WfW for funding his time writing this chapter. Jimaima Le Grand (Queensland Department of Primary Industries and Fisheries) is thanked for generating the maps.

References Ackery, P. R. (1988). Hostplants and classication: a review of nymphalid butteries. Biological Journal of the Linnean Society, 33, 95203. Anonymous (1983). Important Weeds of the World. Third edn. Leverkusen, Germany: Bayer AG. APG II (2003). An update of the Angiosperm Phylogeny Group classication for the orders and families of owering plants: APG II. Botanical Journal of the Linnean Society, 141, 399436. Apori, S. O., Long, R. J., Castro, F. B. and rskov, E. R. (2000). Chemical composition and nutritive value of leaves and stems of tropical weed Chromolaena odorata. Grass and Forage Science, 55, 7781. Baars, J.-R. and Neser, S. (1999). Past and present initiatives on the biological control of Lantana camara (Verbenaceae) in South Africa. African Entomology Memoir, 1, 2133. Bennett, F. D. and Cruttwell, R. E. (1973). Insects attacking Eupatorium odoratum in the neotropics. 1. Ammalo insulata (Walk.) (Lep., Arctiidae), a potential biotic agent for the control of Eupatorium odoratum L. (Compositae). Technical Bulletin of the Commonwealth Institute of Biological Control, 16, 105115. Bhumannavar, B. S. and Ramani, S. (2007) Introduction and establishment of Cecidochares connexa (Macquart) (Diptera: Tephritidae) for the biological suppression of Chromolaena odorata in India. In Proceedings of the Seventh International Workshop on Biological Control and Management of Chromolaena and Mikania, ed. P.-Y. Lai, G. V. P. Reddy and R. Muniappan. Taiwan: National Pingtung University, pp. 3848. , M., Witte, L. and Hartmann, T. (1994). Pyrrolizidine alkaloids in Biller, A., Boppre Chromolaena odorata. Chemical and chemoecological aspects. Phytochemistry, 35, 615619. Blackmore, A. C. (1998). Seed dispersal of Chromolaena odorata reconsidered. In Proceedings of the Fourth International Workshop on the Biological Control and

Chromolaena odorata

153

Management of Chromolaena odorata, ed. P. Ferrar, R. Muniappan and K. P. Jayanth. Agricultural Experiment Station, University of Guam, Publication 216, 1621. Bofeng, I., Donnelly, G., Orapa, W. and Day, M. (2004). Biological control of Chromolaena odorata in Papua New Guinea. In Proceedings of the Sixth International Workshop on Biological Control and Management of Chromolaena, ed. M. D. Day and R. E. McFadyen. ACIAR Technical Reports. Canberra, Australia: ACIAR, pp. 55 1416. Boller, E. F., van Lanteren, J. C. and Delucchi, V. (2006). International Organization for Biological Control of Noxious Animals and Plants. History of the First 50 Years (19562006). IOBC, Zurich, Switzerland, 275 pp. Bremer, K. (1994). Asteraceae Cladistics and Classication. Portland, OR: Timber Press. Byrne, M. J., Coetzee, J., McConnachie, A. J., Parasram, W. and Hill, M. P. (2004). Predicting climate compatibility of biological control agents in their region of introduction. In Proceedings of the XI International Symposium on Biological Control of Weeds, ed. J. M. Cullen, D. T. Briese, D. J. Kriticos, et al. Canberra, Australia: CSIRO Entomology, pp. 2835. CAB International (2004). Prevention and Management of Invasive Alien Species: Forging Cooperation throughout West Africa. Proceedings of a workshop held in Accra, Ghana, 911 March 2004. Nairobi, Kenya: CAB International. Caldwell, P. M. and Kluge, R. L. (1993). Failure of the introduction of Actinote anteas (Lep.: Acraeidae) from Costa Rica as a biological control candidate for Chromolaena odorata (Asteraceae) in South Africa. Entomophaga, 38, 475478. Cock, M. J. W. (1984). Possibilities for biological control of Chromolaena odorata. Tropical Pest Management, 30, 713. Cock, M. J. W. and Holloway, J. D. (1982). The history of, and prospects for, the biological control of Chromolaena odorata (Compositae) by Pareuchaetes pseudoinsulata Rego Barros and allies (Lepidoptera: Arctiidae). Bulletin of Entomological Research, 72, 193205. Coleman, J. R. (1989). Embryology and cytogenetics of apomictic hexaploid Eupatorium tica, 12, 803817. odoratum L. (Compositae). Revista Brasileira de Gene Cruttwell, R. E. (1972). The insects of Eupatorium odoratum L. in Trinidad and their potential as agents for biological control. Unpublished Ph.D. thesis, University of the West Indies, Trinidad. Cruttwell, R. E. (1973a). Insects attacking Eupatorium odoratum in the neotropics. 2. Studies of the seed weevil Apion brunneonigrum B.B., and its potential use to control E. odoratum L. Technical Bulletin of the Commonwealth Institute of Biological Control, 16, 117124. Cruttwell, R. E. (1973b). Insects attacking Eupatorium odoratum in the neotropics. 3. Dichomeris sp. nov. (Trichotaphe sp. nr eupatoriella) (Lep.: Gelechiidae), a leaf-roller on Eupatorium odoratum L. (Compositae). Technical Bulletin of the Commonwealth Institute of Biological Control, 16, 125134. Cruttwell, R. E. (1974). Insects and mites attacking Eupatorium odoratum in the neotropics. 4. An annotated list of the insects and mites recorded from Eupatorium odoratum L., with a key to the types of damage found in Trinidad. Technical Bulletin of the Commonwealth Institute of Biological Control, 17, 87125. Cruttwell, R. E. (1977a). Insects and mites attacking Eupatorium odoratum L. in the neotropics. 6. Two eriophyid mites, Acalitus adoratus Keifer and Phyllocoptes cruttwellae Keifer. Technical Bulletin of the Commonwealth Institute of Biological Control, 18, 5963.

154

C. Zachariades et al.

Cruttwell, R. E. (1977b). Insects attacking Eupatorium odoratum L. in the neotropics. 5. Mescinia sp. nr. parvula (Zeller). Technical Bulletin of the Commonwealth Institute of Biological Control, 18, 4958. Cruz, Z. T., Muniappan, R. and Reddy, G. V. P. (2006). Establishment of Cecidochares connexa (Diptera: Tephritidae) in Guam and its effect on the growth of Chromolaena odorata (Asteraceae). Annals of the Entomological Society of America, 99, 845850. Day, M. and Bofeng, I. (2007). The status of biocontrol of Chromolaena odorata in Papua New Guinea. In Proceedings of the Seventh International Workshop on Biological Control and Management of Chromolaena and Mikania, ed. P.-Y. Lai, G. V. P. Reddy and R. Muniappan. Taiwan: National Pingtung University, pp. 5367. Day, M. D., Wiley, C. J., Playford, J. and Zalucki, M. P. (2003). Lantana: Current Management Status and Future Prospects. Monograph 102. Canberra, Australia: ACIAR, 1128. Dempster, J. P. (1971). The population ecology of the cinnabar moth, Tyria jacobaeae L. (Lepidoptera: Arctiidae). Oecologia, 7, 2667. de Rouw, A. (1991). The invasion of Chromolaena odorata (L.) King & Robinson (ex Eupatorium odoratum), and competition with the native ora, in a rain forest te dIvoire. Journal of Biogeography, 18, 1323. zone, south-west Co Desmier de Chenon, R., Sipayung, A. and Sudharto, P. (2002a). A decade of biological control against Chromolaena odorata at the Indonesian Oil Palm Research Institute in Marihat. In Proceedings of the Fifth International Workshop on Biological Control and Management of Chromolaena odorata, ed. C. Zachariades, R. Muniappan and L. W. Strathie. Pretoria, South Africa: ARC-PPRI, pp. 4652. Desmier de Chenon, R., Sipayung, A. and Sudharto, A. (2002b). A new biological agent, Actinote anteas, introduced into Indonesia from South America for the control of Chromolaena odorata. In Proceedings of the Fifth International Workshop on Biological Control and Management of Chromolaena odorata, ed. C. Zachariades, R. Muniappan and L. W. Strathie. Pretoria, South Africa: ARC-PPRI, pp. 170176. Dharmadhikari, P. R., Perera, P. A. C. R. and Hassen, T. M. F. (1977). The introduction of Ammalo insulata for the control of Eupatorium odoratum in Sri Lanka. Technical Bulletin of the Commonwealth Institute of Biological Control, 18, 129135. Epp, G.A. (1987). The seed bank of Eupatorium odoratum along a successional gradient in a tropical rain forest in Ghana. Journal of Tropical Ecology, 3, 139149. Erasmus, D. J. (1985) Achene biology and the chemical control of Chromolaena odorata. Unpublished Ph.D. thesis, University of Natal, South Africa. Erasmus, D. J. and van Staden, J. (1987). Germination of Chromolaena odorata (L.) K. & R. achenes: effect of storage, harvest locality and the pericarp. Weed Research, 27, 113118. Esguerra, N. M. (2002). Introduction and establishment of the tephritid gall y Cecidochares connexa on Siam weed, Chromolaena odorata, in the Republic of Palau. In Proceedings of the Fifth International Workshop on Biological Control and Management of Chromolaena odorata, ed. C. Zachariades, R. Muniappan and L. W. Strathie. Pretoria, South Africa: ARC-PPRI, pp. 148151. Eussen, J. H. H. and de Groot, W. (1974). Control of Imperata cylindrica (L.) Beauv. in Indonesia. Mededelingen Fakulteit Landbouw-Wetenschappen Gent, 39, 451464. Francini, R. B. and Penz, C. M. (2006). An illustrated key to male Actinote from Southeastern Brazil (Lepidoptera: Nymphalidae). Biota Neotropica, 6 (http://www. biotaneotropica.org.br/v6n1/pt/abstract?identication-key+bn00606012006).

Chromolaena odorata

155

Gagne, R. J. (1977). The Cecidomyiidae (Diptera) associated with Chromolaena odorata (L.) K. & R. (Compositae) in the Neotropical Region. Brenesia, 12/13, 113131. Gautier, L. (1992). Taxonomy and distribution of a tropical weed, Chromolaena odorata (L.) R. King and H. Robinson. Candollea, 47, 645662. Gautier, L. (1993). Reproduction of a pantropical weed: Chromolaena odorata (L.) R. King & H. Robinson. Candollea, 48, 179193. Gautier, L. (1996). Establishment of Chromolaena odorata in a savanna protected from te dIvoire. In Proceedings of the Third re: an example from Lamto, central Co International Workshop on Biological Control and Management of Chromolaena odorata, ed. U. K. Prasad, R. Muniappan, P. Ferrar, J. P. Aeschliman and H. de Foresta. Publication 202. Mangilao, Guam: Agricultural Experiment Station, University of Guam, 5467. Ghazoul, J. (2004). Alien abduction: disruption of native plant-pollinator interactions by invasive species. Biotropica, 36, 156164. Goodall, J. M. (2000). Monitoring serial changes in coastal grasslands invaded by Chromolaena odorata (L.) R. M. King and Robinson. Unpublished M.Sc. thesis, University of Natal, South Africa. Goodall, J. M. and Erasmus, D. J. (1996). Review of the status and integrated control of the invasive alien weed, Chromolaena odorata, in South Africa. Agriculture, Ecosystems and Environment, 56, 151164. Goodall, J. G. and Morley, T. A. (1995). Ntambanana vegetation survey and veld improvement plan. Unpublished report submitted to the Mpendle Ntambanana Agricultural Company (Pty.) Ltd. Goodall, J. M., Zimmermann, H. G. and Zeller, D. (1994). The distribution of Chromolaena odorata in Swaziland and implications for further spread. Unpublished report submitted to the Standing Committee of the 17th Meeting of the Southern African Regional Commission for the Conservation and Utilisation of the Soil (SARCCUS), Lesotho and to the FAO/UN, Rome. Goodman, P. S. (2003). Assessing management effectiveness and setting priorities in protected areas in KwaZulu-Natal. BioScience, 53, 843850. Grierson, A. J. C. (1980). Compositae. In A Revised Handbook of the Flora of Ceylon, Vol. 1, ed. M. D. Dassanayake and F. R. Fosberg. Rotterdam, The Netherlands: A. A. Balkema, pp. 111278. Henderson, L. (2001). Alien Weeds and Invasive Plants. Handbook No. 12. Pretoria, South Africa: ARC-PPRI. Henty, E. E. and Pritchard, P. H. (1973). Weeds of New Guinea and Their Control. Botany Bulletin 7. Lae, Papua New Guinea: Department of Forests. Hilliard, O. (1977). Compositae in Natal. Pietermaritzburg, South Africa: University of Natal Press. Hoevers, R. and MBoob, S. S. (1996). The status of Chromolaena odorata (L.) R. M. King and H. Robinson in West and Central Africa. In Proceedings of the Third International Workshop on Biological Control and Management of Chromolaena odorata, ed. U. K. Prasad, R. Muniappan, P. Ferrar, J. P. Aeschliman and H. de Foresta. Publication 202. Mangilao, Guam: Agricultural Experiment Station, University of Guam, pp. 15. Holm, L. G., Plucknett, D. L., Pancho, J. V. and Herberger, P. D. (1977). The Worlds Worst Weeds: Distribution and Biology. Honolulu, HI: University Press of Hawaii.

156

C. Zachariades et al.

Honu, Y. A. K. and Dang, O. L. (2000). Response of tree seedlings to the removal of Chromolaena odorata Linn. in a degraded forest in Ghana. Forest Ecology and Management, 137, 7582. Ivens, G. W. (1974). The problem of Eupatorium odoratum L. in Nigeria. Pest Articles News Summaries, 20, 7682. Joy, P. J., Lyla, K. R. and Satheesan, N. V. (1993). Biological control of Chromolaena odorata in Kerala (India). Chromolaena odorata Newsletter, 7, 13. Julien, M. H. and Grifths, M. W. (1998). Biological Control of Weeds: A World Catalogue of Agents and their Target Weeds. Fourth edn. Wallingford, UK: CAB International Publishing. King, R. M. and Robinson, H. (1970). Studies in the Eupatorieae (Compositae). XXIX. The genus Chromolaena. Phytologia, 20, 196209. King, R. M. and Robinson, H. (1987). The genera of the Eupatorieae (Asteraceae). Monographs in Systematic Botany from the Missouri Botanical Garden, 22, 1581. Kluge, R. L. (1990). Prospects for the biological control of trifd weed, Chromolaena odorata, in southern Africa. South African Journal of Science, 86, 229230. Kluge, R. L. (1994). Ant predation and the establishment of Pareuchaetes pseudoinsulata Rego Barros (Lepidoptera: Arctiidae) for biological control of trifd weed, Chromolaena odorata (L.) King & Robinson, in South Africa. African Entomology, 2, 7172. Kluge, R. L. and Caldwell, P. M. (1993a). Host specicity of Pareuchaetes insulata (Lep.: Arctiidae), a biological control agent for Chromolaena odorata (Compositae). Entomophaga, 38, 451457. Kluge, R. L. and Caldwell, P. M. (1993b). The biology and host specicity of Pareuchaetes aurata aurata (Lepidoptera: Arctiidae), a new association biological control agent for Chromolaena odorata (Compositae). Bulletin of Entomological Research, 83, 8794. Kriticos, D. J., Yonow, T. and McFadyen, R. E. (2005). The potential distribution of Chromolaena odorata (Siam weed) in relation to climate. Weed Research, 45, 246254. Lai, P.-Y., Muniappan, R., Wang, T.-H. and Wu, C.-J. (2006). Distribution of Chromolaena odorata and its biological control in Taiwan. Proceedings of Hawaiian Entomological Society, 38, 119122. Lanaud, C., Jolivot, M. P. and Deat, M. (1991). Preliminary results on the enzymatic diversity in Chromolaena odorata (L.) R. M. King and H. Robinson (Asteraceae). In Proceedings of the Second International Workshop on Biological Control of Chromolaena odorata, ed. R. Muniappan and P. Ferrar. BIOTROP Special Publication 44, 7177. Leslie, A. J. and Spotila, J. R. (2001) Alien plant threatens Nile crocodile (Crocodylus niloticus) breeding in Lake St. Lucia, South Africa. Biological Conservation, 98, 347355. Liggitt, B. (1983). The Invasive Plant Chromolaena odorata, With Regard to its Status and Control in Natal. Rural Studies Series, Monograph 2. Pietermaritzburg, South Africa: Institute of Natural Resources, University of Natal, pp. 141. MacArthur, R. H. and Wilson, E. C. (1967). The Theory of Island Biogeography. Princeton, NJ: Princeton University Press. Macdonald, I. A. W. (1983). Alien trees, shrubs and creepers invading indigenous vegetation in the Hluhluwe-Umfolozi Game Reserve Complex in Natal. Bothalia, 14, 949959.

Chromolaena odorata

157

Macdonald, I. A. W. and Frame, G. W. (1988). The invasion of introduced species into nature reserves in tropical savannas and dry woodlands. Biological Conservation, 44, 6793. Macdonald, I. A. W., Reaser, J. K., Bright, C., et al., eds. (2003). Invasive alien species in southern Africa: national reports and directory of resources. Cape Town: Global Invasive Species Programme. Marais, C., van Wilgen, B. W. and Stevens, D. (2004). The clearing of invasive alien plants in South Africa: a preliminary assessment of costs and progress. South African Journal of Science, 100, 97103. Martinez, M., Etienne, J., Abud-Antun, A. and Reyes, M. (1993). Les Agromyzidae publique Dominicaine (Diptera). Bulletin de la Socie te entomologique de la Re de France, 98, 165179. Marutani, M. and Muniappan, R. (1991). Interactions between Chromolaena odorata (Asteraceae) and Pareuchaetes pseudoinsulata (Lepidoptera: Arctiidae). Annals of Applied Biology, 119, 227237. McFadyen, R. E. C. (1988) Ecology of Chromolaena odorata in the neotropics. In Proceedings of the First International Workshop on Biological Control of Chromolaena odorata, ed. U. Prasad, R. Muniappan, E. Ferrar J. P. Aeschliman and H. de Foresta. Mangilao, Guam: Agricultural Experiment Station, University of Guam, pp. 1320. McFadyen, R. E. C. (1989). Siam weed: a new threat to Australias north. Plant Protection Quarterly, 4, 37. McFadyen, R. E. C. (1991). The ecology of Chromolaena odorata in the neotropics. In Proceedings of the Second International Workshop on Biological Control of Chromolaena odorata, ed. R. Muniappan and P. Ferrar. BIOTROP Special Publication, 44, 19. McFadyen, R. E. C. (1995) The accidental introduction of the Chromolaena mite, Acalitus adoratus, into South-East Asia. In Proceedings of the VIII International Symposium on Biological Control of Weeds, ed. E. S. Delfosse and R. R. Scott. Melbourne: DSIR/CSIRO, pp. 649652. McFadyen, R. C. (1996a). Biocontrol of Chromolaena odorata: divided we fail. In Proceedings of the Ninth International Symposium on Biological Control of Weeds, ed. V. C. Moran and J. H. Hoffmann. Cape Town: University of Cape Town Press, pp. 455459. McFadyen, R. C. (1996b). National report from Australia and the Pacic. In Proceedings of the Third International Workshop on Biological Control and Management of Chromolaena odorata, ed. U. K. Prasad, R. Muniappan, P. Ferrar, J. P. Aeschliman and H. de Foresta. Publication, 202. Mangilao, Guam: Agricultural Experiment Station, University of Guam, 3944. McFadyen, R. E. (1997). Parasitoids of the arctiid moth Pareuchaetes pseudoinsulata (Lep.: Arctiidae), an introduced biocontrol agent against the weed Chromolaena odorata (Asteraceae), in Asia and Africa. Entomophaga, 42, 467470. McFadyen, R. E. C. (1998). Biological control of weeds. Annual Review of Entomology, 43, 369393. McFadyen, R. E. C. (2002). Chromolaena in Asia and the Pacic: spread continues but control prospects improve. In Proceedings of the Fifth International Workshop on Biological Control and Management of Chromolaena odorata, ed. C. Zachariades, R. Muniappan and L. W. Strathie. Pretoria, South Africa: ARC-PPRI, pp. 1318.

158

C. Zachariades et al.

McFadyen, R. C. (2004a). Chromolaena in Australia: new developments. Chromolaena odorata Newsletter, 16, 12. McFadyen, R. C. (2004b). Chromolaena in East Timor: history, extent and control. In Proceedings of the Sixth International Workshop on Biological Control and Management of Chromolaena, ed. M. D. Day and R. E. McFadyen. ACIAR Technical Reports 55. Canberra, Australia: ACIAR, pp. 810. McFadyen, R. E. C., Desmier de Chenon, R. and Sipayung, A. (2003). Biology and host specicity of the chromolaena stem gall y, Cecidochares connexa (Macquart) (Diptera: Tephritidae). Australian Journal of Entomology, 42, 294297. McWilliam, A. (2000). A plague on your house? Some impacts of Chromolaena odorata on Timorese livelihoods. Human Ecology, 28, 451469. Moni, N. S. and Subramoniam, R. (1960). Essential oil from Eupatorium odoratum a common weed in Kerala. Indian Forester, 86, 209. Muniappan, R. and Marutani, M. (1988). Ecology and distribution of C. odorata in Asia and the Pacic. In Proceedings of the First International Workshop on Biological Control of Chromolaena odorata, ed. R. Muniappan. Mangilao, Guam: Agricultural Experiment Station, University of Guam, pp. 2124. Muniappan, R., Englberger, K., Bamba, J. and Reddy, G. V. P. (2004). Biological control of chromolaena in Micronesia. In Proceedings of the Sixth International Workshop on Biological Control and Management of Chromolaena, ed. M. D. Day and R. E. McFadyen. ACIAR Technical Reports 55. Canberra, Australia: ACIAR, pp. 1112. Muniappan, R., Englberger, K. and Reddy, G. V. P. (2007). Biological control of Chromolaena odorata in the American Pacic Micronesian Islands. In Proceedings of the Seventh International Workshop on Biological Control and Management of Chromolaena and Mikania, ed. P.-Y. Lai, G. V. P. Reddy and R. Muniappan. Taiwan: National Pingtung University, pp. 4952. Muniappan, R., Reddy, G. V. P. and Lai, P. Y. (2005). Distribution and biological control of Chromolaena odorata. In Invasive Plants: Ecological and Agricultural Aspects, ed. Inderjit. Basel, Switzerland: Birkhauser Verlag, pp. 223233. Muniappan, R., Sundaramurthy, V. T. and Viraktamath, C. A. (1989). Distribution of Chromolaena odorata (Asteraceae) and bionomics and consumption and utilization of food by Pareuchaetes pseudoinsulata (Lepidoptera: Arctiidae) in India. In Proceedings of the VII International Symposium on Biological Control of Weeds, ed. E. S. Delfosse. Rome: Istituto Sperimentale per la Patologia Vegetale, Ministero dellAgricoltura e delle Foreste, pp. 401409. Norbu, N. (2004). Invasion success of Chromolaena odorata in the Terai of Nepal. Unpublished M.Sc. thesis, International Institute for Geo-information Science and Earth Observation, Enschede, Netherlands. Orapa, W. and Bofeng, I. (2004). Mass production, establishment and impact of Cecidochares connexa on chromolaena in Papua New Guinea. In Proceedings of the Sixth International Workshop on Biological Control and Management of Chromolaena, ed. M. D. Day and R. E. McFadyen. ACIAR Technical Reports 55. Canberra, Australia: ACIAR, pp. 3035. Parasram, W. A. (2003). The role of climate in optimal release strategy design for Pareuchaetes insulata: a new control agent for trifd weed (Chromolaena odorata). Proceedings of the South African Sugar Technologists Association, 77, 210215. Prasad, U. K., Muniappan, R., Ferrar, P., Aeschliman, J. P. and de Foresta, H., eds. (1996). Proceedings of the Third International Workshop on Biological Control and

Chromolaena odorata

159

Management of Chromolaena odorata. Publication 202. Mangilao, Guam: Agricultural Experiment Station, University of Guam. Raman, A., Muniappan, R. N., Silva-Krott, I. U. and Reddy, G. V. P. (2006). Induceddefense responses in the leaves of Chromolaena odorata consequent to infestation by Pareuchaetes pseudoinsulata (Lepidoptera: Arctiidae). Journal of Plant Disease and Protection, 113, 234239. Rambuda, T. D. and Johnson, S. D. (2004). Breeding systems of invasive alien plants in South Africa: does Bakers rule apply? Diversity and Distributions, 10, 409416. rez, N. (2004) Ecology of pollination in a tropical Venezuelan savanna. Plant Ram Ecology, 173, 171189. Rao, Y. R. (1920). Lantana insects in India. Memoirs, Department of Agriculture in India, Entomology Series, Calcutta, 5, 239314. nek, M., et al. (2000). Naturalization and invasion of Richardson, D. M., Pysek, P., Rejma alien plants: concepts and denitions. Diversity and Distributions, 6, 93107. Sahid, I. B. and Sugau, J. B. (1993). Allelopathic effects of lantana (Lantana camara) and Siam weed (Chromolaena odorata) on selected crops. Weed Science, 41, 303308. Sajise, P. E., Palis, R. K., Norcio, N. V. and Lales, J. S. (1974). The biology of Chromolaena odorata (L.) R. M. King and H. Robinson. I. Flowering behavior, pattern of growth and nitrate metabolism. Philippine Weed Science Bulletin, 1, 1724. Scott, L. J., Lange, C. L., Graham, G. C. and Yeates, D. K. (1998). Genetic diversity and origin of Siam weed (Chromolaena odorata) in Australia. Weed Technology, 12, 2731. Seibert, T. F. (1989). Biological control of the weed, Chromolaena odorata (Asteraceae), by Pareuchaetes pseudoinsulata (Lepidoptera: Arctiidae) on Guam and the Northern Mariana Islands. Entomophaga, 35, 531539. Singh, S. P. (1998). A review of biological suppression of Chromolaena odorata (Linnaeus) King and Robinson in India. In Proceedings of the Fourth International Workshop on the Biological Control and Management of Chromolaena odorata, ed. P. Ferrar, R. Muniappan and K. P. Jayanth. Publication 216. Mangilao, Guam: Agricultural Experiment Station, University of Guam, pp. 8692. Solis, M. A., Metz, M. A. and Zachariades, C. (2008). Identity and generic placement of Phestinia costella Hampson (Lepidoptera: Pyralidae: Phycitinae) reared on the invasive plant Chromolaena odorata (L.) R. M. King and H. Rob (Asteraceae). Proceedings of the Entomological Society of Washington, 110, 292301. Stone, B. C. (1966). Further additions to the ora of Guam. 3. Micronesica, 2, 133141. Strathie, L. W. and Zachariades, C. (2002). Biological control of Chromolaena odorata in South Africa: developments in research and implementation. In Proceedings of the Fifth International Workshop on Biological Control and Management of Chromolaena odorata, ed. C. Zachariades, R. Muniappan and L. W. Strathie. Pretoria, South Africa: ARC-PPRI, pp. 7479. Strathie, L. W. and Zachariades, C. (2004). Insects for the biological control of Chromolaena odorata: surveys in the northern Caribbean and efforts undertaken in South Africa. In Proceedings of the Sixth International Workshop on Biological Control and Management of Chromolaena, ed. M. D. Day and R. E. McFadyen. ACIAR Technical Reports 55. Canberra, Australia: ACIAR, pp. 4552. Talapatra, S. K., Bhar, D. S. and Talapatra, B. (1974). Flavonoid and terpenoid constituents of Eupatorium odoratum. Phytochemistry, 13, 284285. Thapa, R. and Wongsiri, S. (1997). Eupatorium odoratum: a honey plant for beekeepers in Thailand. Bee World, 78, 175178.

160

C. Zachariades et al.

Timbilla, J. A. (1996). Status of Chromolaena odorata biological control using Pareuchaetes pseudoinsulata, in Ghana. In Proceedings of the Ninth International Symposium on Biological Control of Weeds, ed. V. C. Moran and J. H. Hoffmann. Cape Town, South Africa: University of Cape Town Press, pp. 327331. Timbilla, J. A. (1998). Effect of biological control of Chromolaena odorata on biodiversity: a case study in the Ashanti region of Ghana. In Proceedings of the Fourth International Workshop on the Biological Control and Management of Chromolaena odorata, ed. P. Ferrar, R. Muniappan and K. P. Jayanth. Publication 216. Mangilao, Guam: Agricultural Experiment Station, University of Guam, pp. 97101. Timbilla, J. A., Zachariades, C. and Braimah, H. (2003). Biological control and management of the alien invasive shrub Chromolaena odorata in Africa. In Biological Control in IPM Systems in Africa, ed. P. Neuenschwander, C. Borgemeister and J. Langewald. Wallingford, UK: CABI Publishing, pp. 145160. Tjitrosemito, S. (1998). Introduction of Procecidochares connexa (Diptera: Tephritidae) to Java island to control Chromolaena odorata. In Proceedings of the Fourth International Workshop on the Biological Control and Management of Chromolaena odorata, ed. P. Ferrar, R. Muniappan and K. P. Jayanth. Publication 216. Mangilao, Guam: Agricultural Experiment Station, University of Guam, pp. 6672. Toelken, H. R. (1983). Compositae. In Flowering Plants in Australia, ed. B. D. Morley and H. R. Toelken. Adelaide, Australia: Rigby, pp. 300314. van Gils, H., Delno, J., Rugege, D. and Janssen, L. (2004). Efcacy of Chromolaena odorata control in a South African conservation forest. South African Journal of Science, 100, 251253. van Wilgen, B. W., de Wit, M. P., Anderson, H. J., et al. (2004). Costs and benets of biological control of invasive alien plants: case studies from South Africa. South African Journal of Science, 100, 113122. Versveld, D. B., Le Maitre, D. C. and Chapman, R. A. (1998). Alien invading plants and water resources in South Africa: a preliminary assessment. Water Research Commission Report No. TT 99/98, C.S.I.R. No. ENV/S-C 97154. von Senger. I., Barker, N. P. and Zachariades, C. (2002). Preliminary phylogeography of Chromolaena odorata: nding the origin of South African weed. In Proceedings of the Fifth International Workshop on Biological Control and Management of Chromolaena odorata, ed. C. Zachariades, R. Muniappan and L. W. Strathie. Pretoria, South Africa: ARC-PPRI, pp. 9099. Wallner, W. E. (1987). Factors affecting insect population dynamics: differences between outbreak and non-outbreak species. Annual Review of Entomology, 32, 317340. Waterhouse, B. M. (1994a). Discovery of Chromolaena odorata in northern Queensland, Australia. Chromolaena odorata Newsletter, 9, 12. Waterhouse, B. M. (2003). Know your enemy: recent records of potentially serious weeds in northern Australia, Papua New Guinea and Papua (Indonesia). Telopea, 10, 477485. Waterhouse, B. M. and Zeimer, O. (2002). On the brink: the status of Chromolaena odorata in northern Australia. In Proceedings of the Fifth International Workshop on Biological Control and Management of Chromolaena odorata, ed. C. Zachariades, R. Muniappan and L. W. Strathie. Pretoria, South Africa: ARC-PPRI, pp. 2933. Waterhouse, D. F. (1994b). Biological Control of Weeds: Southeast Asian Prospects. ACIAR Monograph, 26, 1302.

Chromolaena odorata

161

Wilson, C. G. and Widayanto, E. B. (1998). A technique for spreading the Chromolaena gall-y, Procecidochares connexa, to remote locations. In Proceedings of the Fourth International Workshop on the Biological Control and Management of Chromolaena odorata, ed. P. Ferrar, R. Muniappan and K. P. Jayanth. Publication 216. Mangilao, Guam: Agricultural Experiment Station, University of Guam, pp. 6365. Wilson, C. G. and Widayanto, E. B. (2002). The biological control programme against Chromolaena odorata in eastern Indonesia. In Proceedings of the Fifth International Workshop on Biological Control and Management of Chromolaena odorata, ed. C. Zachariades, R. Muniappan and L. W. Strathie. Pretoria, South Africa: ARC-PPRI, pp. 5357. Wilson, C. G. and Widayanto, E. B. (2004). Establishment and spread of Cecidochares connexa in eastern Indonesia. In Proceedings of the Sixth International Workshop on Biological Control and Management of Chromolaena, ed. M. D. Day and R. E. McFadyen. ACIAR Technical Reports 55. Canberra, Australia: ACIAR, pp. 3944. Witkowski, E. T. F. and Wilson, M. (2001). Changes in density, biomass, seed production and soil seed banks of the non-native invasive plant, Chromolaena odorata, along a 15 year chronosequence. Plant Ecology, 152, 1327. nek, M. (2004). Catalogue of the naturalized ora of Wu, S.-H., Hsieh, C.-F. and Rejma Taiwan. Taiwania, 49, 1631. Yadav, A. S. and Tripathi, R. S. (1981). Population dynamics of the ruderal weed Eupatorium odoratum and its natural regulation. Oikos, 36, 355361. Ye, W. H., Mu, H. P., Cao, H. L. and Ge, X. J. (2004). Genetic structure of the invasive Chromolaena odorata in China. Weed Research, 44, 129135. bel, L. W. and Kluge, R. L. (1999). The South African Zachariades, C., Strathie-Korru programme on the biological control of Chromolaena odorata (L.) King & Robinson using insects. African Entomology Memoir, 1, 89102. Zachariades, C., Strathie, L. W. and Kluge, R. L. (2002). Biology, host specicity and effectiveness of insects for the biocontrol of Chromolaena odorata in South Africa. In Proceedings of the Fifth International Workshop on Biological Control and Management of Chromolaena odorata, ed. C. Zachariades, R. Muniappan and L. W. Strathie. Pretoria, South Africa: ARC-PPRI, pp. 160166. Zachariades, C. and Strathie, L. W. (2006). Biocontrol of chromolaena in South Africa: recent activities in research and implementation. Biocontrol News and Information, 27, 10N15N. Zachariades, C., Strathie, L., Delgado, O. and Retief, E. (2007). Pre-release research on biocontrol agents for chromolaena in South Africa. In Proceedings of the Seventh International Workshop on Biological Control and Management of Chromolaena and Mikania, ed. P.-Y. Lai, G. V. P. Reddy and R. Muniappan. Taiwan: National Pingtung University, pp. 6880. Zachariades, C., von Senger, I. and Barker, N. P. (2004). Evidence for a northern Caribbean origin for the southern African biotype of Chromolaena odorata. In Proceedings of the Sixth International Workshop on Biological Control and Management of Chromolaena, ed. M. D. Day and R. E. McFadyen. ACIAR Technical Reports 55. Canberra, Australia: ACIAR, pp. 2527. Zalucki, M. P. and van Klinken, R. D. (2006). Predicting population dynamics of weed biological control agents: science or gazing into crystal balls? Australian Journal of Entomology, 45, 330343.

162

C. Zachariades et al.

Zhigang, L., Schichou, H., Mingfang G., et al. (2004). Rearing Actinote thalia pyrrha (Fabricius) and Actinote anteas (Doubleday and Hewitson) with cutting and potted Mikania micrantha Kunth. In Proceedings of the Sixth International Workshop on Biological Control and Management of Chromolaena, ed. M. D. Day and R. E. McFadyen. ACIAR Technical Reports, 55, Canberra, Australia: ACIAR, pp. 3638.