AUSTRALIA AND HUANGLONGBING

AUSTRALIA AND HUANGLONGBING GAC Beattie 1, P Holford 1, DJ Mabberley 1,2, AM Haigh 1 and P Broadbent 3 1 Centre for Plant and Food Science, University of Western Sydney, Locked Bag 1797, Penrith South DC, New South Wales 1797, Australia; 2 Royal Botanic Gardens, Kew, Richmond, Surrey, TW9 3AB, United Kingdom; 3 PO Box 46 Mulgoa, NSW 2745, Australia ABSTRACT Preparations are now underway for potential incursions of huanglongbing and its two known vectors, Diaphorina citri Kuwayama and Trioza erytreae del Guercio, into Australia. These preparations, particularly the development of an incursion management plan (IMP), involve extensive reviews of literature related to the origins of Citrus and huanglongbing, and of host records for the disease and its vectors. This paper briefly discusses issues and aspects of the IMP, including pre- and post-incursion management plans. Key words: Incursion Management Plan (IMP), huanglongbing, Australia INTRODUCTION The pathogens that cause huanglongbing (HLB) are not known to occur in Australia neither are the two known vectors of the disease, the Asiatic citrus psyllid Diaphorina citri Kuwayama [Hemiptera: Sternorrhyncha: Psylloidea: Psyllidae] and the African citrus psyllid Trioza erytreae del Guercio [Hemiptera: Sternorrhyncha: Psylloidea: Triozidae]. Early last century, the Asiatic citrus psyllid was found in northern Australia, but was eradicated by chance between 1916-1922, when all introduced species and hybrids of the genus Citrus in the Northern Territory were destroyed during a successful campaign to eradicate citrus canker (Xanthomonas citri subsp. citri (ex Hasse 1915) [Pseudomonadales: Pseudomonadaceae] (Bellis et al. 2005). Australian citrus entomologists and pathologists have been involved in collaborative research and technology transfer projects in Asia since 1979. The two most recent research projects, funded by the Australian Centre for International Agricultural Research (ACIAR) and the Australian Department of Innovation, have focused on HLB research in Indonesia, Vietnam and China. An incursion management plan (IMP) for the disease and its vectors has been prepared for the Australian citrus industry (Beattie and Barkley 2009). Preparation of the plan has involved a thorough review of host records for the disease and its vectors, assessment of likely entry pathways, the biology of the disease and the vectors, and methods to limit the impact of the disease should one or both vectors be introduced to Australia. When the current research on HLB began in 2000-2001, and during initial preparation of the IMP, it was assumed that the systematics of the family Rutaceae, the subfamily Aurantioideae, and genera within the Aurantioideae, including Citrus and Murraya, were accurate. It was also assumed that Citrus was native to East Asia and that HLB originated in Citrus in China. This research questioned the above assumptions. This paper gives a brief overview of the origins of the genus Citrus, the origin and spread of HLB, the origins of D. citri, some aspects of the systematics of the Rutaceae, some aspects of host records of HLB and its vectors, and issues related to the genera Murraya, Merrillia, Bergera and Clausena. Also discussed are issues related to potential incursions of HLB and one or both of its vectors into Australia. These issues include: vulnerability of the Australian citrus industry and indigenous germplasm; potential entry pathways; post-incursion surveys and 1

responses; cost-sharing agreements; and postincursion management of HLB and its vectors. ORIGINS OF CITRUS Beattie et al. (2008) recently hypothesised that the genus Citrus originated in Australia and that it dispersed westward in equatorial currents, to Southeast Asia, when such currents exited to the north of what is now Papua New Guinea, and on island terranes that moved 1000 kilometres eastward across the same region (Hall 1997, 2001, 2002). Some species may also have been dispersed by birds and bats. These hypotheses are based on current knowledge of plate tectonics and the origins of components of Australasia and Asia (e.g., Veevers et al. 1991, Metcalfe 1998, Hall 1997, 2001, 2002, Hartley 2001a, b) 1, the morphology of Citrus and its relatives as circumscribed by Walter Swingle (Swingle and Reece 1967), recent re-classifications by David Mabberley (Mabberley 2004, Zhang et al. 2008), and recent molecular studies (e.g., Samuel et al. 2001, Bayer et al. 2004, Muellner et al. 2007, Bayer et al. 2009). Muellner et al. (2007) estimated the age of Rutacaeae as 91 my, the Aurantioideae as 71 my, C. glauca as 22 my, and C. japonica and C. trifoliata at 18 my. Bayer et al. (2009) showed that the genus comprises two major clades, an Australasian clade comprising plants considered by Swingle to be species of Clymenia and Microcitrus in Papua New Guinea, Eremocitrus and Microcitrus in Australia, and Oxanthera in New Caledonia, and an Asiatic clade comprising well-known commercial Citrus species and hybrids, and species considered by Swingle to be species of Fortunella and Poncirus. The molecular evidence presented by Bayer et al. (2009) also indicates that the citron, C. medica, the first described species of Citrus, and long considered to be native to India, probably originated in Australasia, possibly in Papua New Guinea, and that its closest relative is C. polyandra (syn. Clymenia polyandra), a species native to New Ireland, a relatively small island to the north-east of mainland Papua New Guinea (Bayer et al. 2009). ORIGIN AND SPREAD OF HUANGLONGBING Beattie et al. (2008) recently summarised their views, repeated here, on the origins of HLB 2. The accepted view, as expressed by Zhao (1981), da Graça (1991), Bové (2006), and in many other publications, is that HLB originated in China. It is based on three assumptions: (1) that the disease was present in China in the 1800s (Lin 1956); (2) that Reinking (1919), Lee (1921), Tu (1932), and some other authors in China before 1940, described symptoms of the disease; and (3) that HLB evolved with Citrus. However, presence of the disease in China in the 1800s was based on interviews with farmers and technicians between 1947 and 1955 (Lin 1956; see also Lin and Lin 1990). Symptoms of maladies described by Reinking (1919) and Lee (1921) do not describe HLB, and there is no evidence that HLB evolved with Citrus. It is highly likely that the disease was not widespread in China before the 1940s and that it may not have occurred there before 1930 as: the first substantive record of the disease appears to have been made by Chen Qibao in 1938 (Chen 1943, Lin 1956), four years after Zhou Yuwen collected D. citri on citrus and other hosts at Lingnan University in Guangzhou (see Hoffmann 1936); parasitoids were not associated with these psyllid populations (Hoffmann 1936); and Aubert (1990a) recorded information on Jiaogan (Tankan) tangor production in the Chaoyan district of Guangdong from 1946 and 1990 and noted that the first dramatic HLB epidemic occurred from the late 1950s following use of contaminated trees, and that subsequent epidemics were related to natural spread of the disease by D. citri. 1 Animations of events over the past 55 my can be viewed on the internet (SE Asia Research Group 2006: http://www.gl.rhul.ac.uk/ searg/index.html) 2 These views are being reconsidered in light of recent reports of a fourth species of 'Candidatus Liberibacter' being discovered in potato (Solanum tuberosum L.), tomato (S. lycopersicum L.) [Solanales: Solanaceae] and other solanaceous plants in New Zealand in association with the recently introduced potato/tomato psyllid Bactericera (=Paratrioza) cockerelli (Sulc) [Psyllidae], presumably from North and/or Central America: see http://www.biosecurity.govt.nz/pests-diseases/plants/potato-tomato-psyllid.htm. 2

With the possible exception of China, HLB symptoms were not recorded in Southeast Asia until after the mid 1940s; in Indonesia in 1948 (Aubert et al. 1985); in Taiwan about 1950 (Su and Huang 1990); in the Philippines in 1957 (Martinez and Wallace 1967); in Thailand in the 1960s (Schwarz et al. 1973a, b; Aubert 1990c); and in Malaysia in the 1970s (Ko 1988, 1991, Lim et al. 1990). Human-assisted spread of the disease through the Indonesian archipelago (Tirtawidjaja 1980) to Papua New Guinea (northern Australasia) took more than 50 years. It was first recorded in Papua New Guinea in the northwest at Vanimo in 2002 (Weinert et al. 2004). Early Indian records of symptoms resembling those of the disease seem to have been overlooked. These records suggest that symptoms resembling those of HLB in Citrus were observed in the mid 1700s in the central provinces of India by Roghoji, the Bhonsla Raja of Nagpur (Capoor 1963). Subsequently, relatively early records in northwestern and north-eastern India occurred in the 1800s and early 1900s (Bonavia 1888-1890, Husain and Nath 1927, Pruthi and Mani 1945, Asana 1958, Capoor 1963, Fraser et al. 1966, Chadra et al. 1970). With the benefit of current knowledge, it is clear that the severe symptoms described by Husain and Nath (1927) and Pruthi and Mani (1945) and attributed to D. citri were symptoms now known to be caused by 'Candidatus Liberibacter asiaticus'. Husain and Nath (1927), not Reinking (1919) or Lee (1921), were therefore the first authors to describe symptoms of HLB. Damage caused by D. citri is much more benign and does not lead to death of trees. We think HLB originated in Africa. T. erytreae, the vector of HLB in Africa, is native to sub-saharan Africa, and is the only species of Trioza that is known to feed and develop on Rutaceae (Hollis 1984; Aubert 1987b). T. erytreae has two native rutaceous hosts in Africa on which it can complete its development: Vepris lanceolata (Lam.) G. Don [Toddalioideae: Toddalieae]; and Zanthoxylum capense (Thunb.) Harv. [Rutoideae: Zanthoxyleae] (Moran 1968, Hollis 1984, Aubert 1987b). The psyllid also feeds on the native monospecific Calodendrum capense (L. f.) Thunb. [Rutoideae: Diosmeae], but cannot complete its life cycle on this species (Moran 1968). Synonyms of V. lanceolata include Boscia undulata Thunb., Toddalia lanceolata Lam, Vepris querimbensis Klotzsch and Vepris undulata (Thunb.) Verdoorn and CA Smith (see Mziray 1992). Z. capense is often cited as Fagara capensis Thunb., and sometimes as F. capense. Non-native African hosts on which T. erytreae can complete its development in Africa include Citrus species and hybrids, Murraya paniculata (L.) Jack and Clausena anisata (Willd.) Hook. f. ex Benth. (=Cl. inaequalis (DC.) Benth.) (Moran 1968, Hollis 1984, Aubert 1987b). The latter species is a host of the 'Ca. L. africanus' (Korsten et. al. 1996) and is often stated as being native to Africa (e.g., Moran 1968, Aubert 1987b, OEPP/EPPO 2005). It is, however, native to the Western Ghats of India and the northeast part of the Indian subcontinent through to China, and it has many synonyms (Molino 1994). There are no reports of D. citri feeding or developing on it in Asia. The fact that V. lanceolata is the preferred native host of T. erytreae in Africa (Moran 1968) suggests that it is the original host of both T. erytreae and HLB. HLB was transmitted from V. lanceolata to orange or mandarin trees by T. erytreae in one of the European colonies on the southeast coast of Africa and then taken to the Indian subcontinent in infected plants or budwood some 300-500 years ago. It was then acquired and spread by D. citri. Spread would also have occurred through marcotting and grafting, and enhanced by changes in horticultural practices that through increased use of irrigation and fertilisers within monocultures would have led to more abundant and frequent growth flushes. The latter would have led to far higher populations of D. citri than would have occurred in the insect s original environment. It seems the disease and D. citri may have been introduced to China in the early 1930s, on plants, possibly a lemon variety, shipped directly from India or on plants of Indian origin from Southeast Asia. Determining if this may have been related to plants imported by commercial nurseries or by scientists based at Lingnan University in Guangzhou, is still in process. It has not been determined yet when HLB was first recorded in Vietnam or whether it was present there before it was present in China, but it is of 3

interest to note that Wang (1934) mentioned Citrus magner ex Hort. (sic), a lemon type (C. limonia Osbeck var. nov. wagner), being introduced to Guangdong by George Groff from the garden of a Mr. Wagner, a French resident of Saigon. ORIGINS OF D. CITRI Modern psylloids, including the Diaphorinineae and Triozinae, probably evolved with the Sapindales in Gondwana (Hollis 1985, 1987, White and Hodgkinson 1985). The Diaphorineae have an ecological preference for dry climates (Hollis 1987). D. citri was first recorded as a serious pest of citrus in India by Husain and Nath (1927) who, in describing the damage it caused, were the first to describe what are now known to be symptoms of HLB. In China, the Philippines, Malaya or Indonesia in the early to mid 1930s, the psyllid did not assume such a destructive status, as reported in India by Husain and Nath (1927) (Clausen 1933, Hoffmann 1936). Hoffmann and Clausen were both aware of the destruction wrought by the psyllid in India (Hoffmann 1936). All evidence supports the view of Hollis (1987) who suggested that the psyllid evolved in India in association with a species of Murraya 3. D. citri was described in Taiwan (Kuwayama 1908). Crawford (1919) recorded it as present in the Philippines, Taiwan, Java, Malay Archipelago, Bengal, southern India and southern China, and noted that Frederick Muir collected specimens from Macao and Amboina (Moluccas) in 1906. Clausen (1933) recorded it as present in China, the Philippine Islands, Taiwan, Malaya, Dutch East Indies, Burma, India, and Ceylon. Not verified are the records from Macau and Ambon, and Hoffmann (1936) regarded detection of the psyllid in Guangdong in 1934 as the first record of D. citri in China. If it was present in China, it was not widespread as Hoffmann (1936), who commenced his studies on pests of citrus and other plants in Guangdong in 1926 (Jiang et al. 1935), would presumably have noticed it. Further evidence that the psyllid may not have been present in China before 1930 is provided by Luh (1936). Luh, in a paper on factors, including pests and diseases causing fruit loss in 1935 in Zhejiang, did not mention the presence of D. citri or symptoms resembling HLB. Information about the two primary parasitoids of D. citri also suggests that the psyllid evolved in India. Both parasitoids were first described from India, the ectoparasitoid Tamarixia radiata (Waterston) [Hymenoptera: Eulophidae] by Waterston (1922) and the endoparasitoid Diaphorencyrtus aligarhensis (Shafee, Alam and Agarwal) [Hymenoptera: Encyrtidae] by Shafee et al. (1975). Most records of T. radiata 4 in Southeast Asia appear to be related to intentional introductions (Chiu et al. 1988, Chien et al. 1988, Waterhouse 1998) whereas, records of D. aligarhensis 5 appear to be linked to unintentional movement of parasitised D. citri nymphs on live plants to Taiwan (possibly before 1900) and the Philippines, and to natural spread overland with its host. D. aligarhensis is not, as assumed by Chien et al. (1988, 1989, 1991), native to Taiwan or the Philippines, nor were any parasitoids found in association with D. citri populations when the psyllid was first recorded in Guangzhou (Hoffmann 1936). Tang (1988) mentioned the introduction of T. radiata to Fujian in China, but considered it as possibly indigenous given its wide distribution in Fujian, as recorded in surveys within four years of its introduction. He also considered it possible that the parasitoid occurred in Taiwan before it was released there (Tang 1988) and indigenous in Asia, from Saudi Arabia to China (Tang 1990). Other records do not support this. D. aligarhensis was present in the Philippines in 1968 (Catling 1968, Gavarra and Mercado 1988), but T. radiata was not, and was introduced in 1988 (Catling 1968, Gavarra and Mercado 1988, Gavarra et al. 1990). Both parasitoids were present in Java in Indonesia in 1987 (Nurhadi 1987). Based on comments by Nurhadi (1988), it seems they 3 We assume this view was based on the genus Murraya as circumscribed by Swingle and Reece (1967) within the subtribe Clauseneae. 4 syns Tetrastichus radiatus Waterson (Waterson 1922, Graham 1987, Tang 1990, Tang and Aubert 1990, Waterhouse 1998) and Tetrastichus indicus Khan and Shafee (Hayat and Shahi 2004). 5 syns Aphidencyrtus aligarhensis Shafee, Alam and Agarwal, Aphidencyrtus diaphorinae Myartsera and Tryapitzyn and Psyllaephagus diaphorinae Lin and Tao (Noyes and Hayat 1984, Prinsloo 1985, Tang 1990, Tang and Aubert 1990, Waterhouse 1998). 4

may have been introduced to Indonesia on psyllid infested plants. In addition to D. citri, hosts of D. aligarhensis apparently include Diaphorina cardiae Crawford), Diaphorina auberti Hollis and Psylla sp. (Tang and Aubert 1990): Diaphorina cardiae, is a junior synonym of Diaphorina aegyptiaca Puton (Burckhardt 1984). SYSTEMATICS OF THE RUTACEAE The most widely used taxonomic systems for classifying citrus are those of Walter Tennyson Swingle (Swingle 1943, Swingle and Reece 1967) and Tyozaburo Tanaka (Tanaka 1977). They recognised 16 and 162 species, respectively. Their views have led to widespread confusion in the use of names of cultivar groups, inappropriate species status of hybrids, confusion in the names of true species (Scora 1975, Mabberley 1997), and a profound misunderstanding of generic limits (Mabberley 1998). Confusion and turmoil has been exacerbated by the use of a plethora of species names for apomictic hybrid clones. There has been no consensus on the names of these entities and many dubious synonyms and invalid names are widely used in books, journals and, most recently, in poorly referenced and inconsistent popular and technical internet websites that could perpetuate errors ad infinitum. Some papers, including molecular studies, deal with plants for which claimed taxonomic relationships are invalid or poorly understood and for which verifiable voucher specimens have not been preserved. Many such publications are therefore of limited value and may mislead the unwary. Recent work suggests that the genus Citrus comprises about 25 species (Mabberley 2004). This view is based on recent reunification of Eremocitrus, Fortunella, Microcitrus and Poncirus with Citrus (Mabberley 1998, Zhang et al. 2008) and on molecular studies (Guerra et al. 2000, Samuel et al. 2001, Groppo et al. 2008, Bayer et al. 2009) that support these reunifications and the inclusion of Clymenia polyandra (Tanaka) Swingle from New Ireland in eastern Papua New Guinea and species of Oxanthera from New Caledonia in Citrus. Table 1 lists current names of species (sensu Mabberley), some 50% of which are Australasian. Table 2 lists the names of common hybrids (sensu Mabberley). HOST RECORDS An important component of any IMP for HLB and its vectors is knowledge of the hosts of the disease and its vectors. We have thoroughly reviewed the literature to determine hosts of the 'Ca. Liberibacter' and whether the vectors, both D. citri and T. erytreae feed and multiply on these plants. The format we have followed is based on Halbert and Manjunath (2004) but we have included more details and added what we consider to be valid specific and varietal names for the plants while still citing the names used by authors who published the original records. Some papers have mentioned plants for which claimed taxonomic relationships are invalid or poorly understood and for which verifiable voucher specimens have not been preserved. Many such publications are therefore of limited value and potentially misleading. It would appear that all commercially grown species and hybrids of Citrus are hosts of 'Ca. Liberibacter', but the severity of symptoms varies with the liberibacter form, whether a tree has flushed during a period of psyllid activity (De Lange et al. 1985, Koizumi et al. 1994) and the presence and severity of Citrus Tristeza Virus (CTV) strains (Martinez and Wallace 1967, Martinez 1972, Bhagabati and Nariani 1980, Prommintara 1990). There is some variability in the literature, particularly in sensitivity of mandarin (C. reticulata Blanco) and pomelo (C. maxima) varieties, but in general sweet orange (C. aurantium L.), mandarin and grapefruit (C. aurantium L.) are highly sensitive, with true lemon (C. limon), rough lemon (C. taitensis) and trifoliate orange (C. trifoliata) and its hybrids tolerant and the lime (C. aurantiifolia) group highly tolerant. Table 3 shows an example of the details for the genus Murraya. One of the most contentious issues is whether species or forms of Murraya (sensu stricto) are hosts of 'Ca. Liberibacter' and of D. citri. These issues are addressed elsewhere in the text. It is of interest to note here though that although Mu. paniculata is generally regarded as the preferred host of D. citri, it was not reported 5

Table 1. Specific and common names of 23 species considered to be true species of the genus Citrus (Rutaceae: Aurantioideae: Aurantieae) and their endemic region(s). Widely cultivated species are in bold. Species Common names Native region C. amblycarpa (Hassk.) sambal, djerook leemo, nasnaran Malaysia and Indonesia Ochse (a possible hybrid) mandarin, Celebes papeda C. australis (Mudie) Planchon Australian round lime, dooja Australia C. australasica F. Mueller Australian finger-lime Australia C. cavaleriei H. Léveillé ex Ichang papeda/lime/lemon China Cavalerie (C. ichangensis Swingle) C. fragrans (Montr.) fragrant oxanthera New Caledonia C. garrawayi F. M. Bailey Mount White lime, Garraway s Australian wild lime Australia C. glauca (Lindley) Burkill Australian desert lime Australia C. gracilis Mabb. Humpty Doo lime Australia C. halimii BC Stone sultan lemon, limau kadangsa, West Malaysia and Sumatera in Indonesia limau kedut kera C. hystrix DC kaffir lime, limau purut, limau hantu Southeast Asia (C. macroptera Montrouz.) C. inodora F. M. Bailey Russell River lime, large-leaf Australia Australian wild lime C. japonica Thunb. syn. Fortunella kumquat China C. medica L. citron Assumed to be India but possibly Papua New Guinea C. maxima (Burm.) Merr. pomelo Indo-China C. neo-caledonica Guill. false orange, large-leaf oxanthera New Caledonia C. oxanthera Beauvisage orange flower oxanthera New Caledonia C. polyandra Tanaka syn. Clymenia New Ireland in Papua New Guinea C. reticulata Blanco mandarin, tangerine, satsuma; China includes C. x suhuiensis Hort. ex Tanaka, known as Canton mandarin, sz-wei-kom, som keo wan, siem, xiem, limau langkat C. trifoliata L. syn. Poncirus trifoliate orange China C. undulata Guill. New Caledonia C. warburgiana F. M. Bailey Milne Bay lime, New Guinea wild lime New Guinea C. wintersii Mabb. syn. Brown River finger lime New Guinea Microcitrus papuana H. F. Winters as a host of the psyllid in India until 1975 (Cheema and Kapur 1975), 60, 41 and 13 years, respectively, after it was reported as a host in Taiwan by Maki (1915) and Kuwayama (1931), in China by He and Zhou (1935), and in the Ryukyu Islands of Japan by (Miyatake 1965). Oddly, B. koenigii (cited as Mu. koenigii) was the first Citrus relative to be recorded as a host of the psyllid in India (Fletcher 1917, 1919, Husain and Nath 1927). Murraya, Merrillia and Bergera Murraya sect Murraya, which includes Mu. paniculata, was recently transferred from the 6

Table 2. Common hybrids within the genus Citrus (Rutaceae: Aurantioideae: Aurantieae). Widely cultivated hybrids are in bold. Hybrid Common name(s) C. x aurantiifolia (Christm.) Swingle lime C. x aurantium L. sour, sweet, Valencia and navel oranges; grape (= C. aurantium L. and C. sinensis (L.) Osbeck) fruit, and king orange/tangor/mandarin, the latter sometimes given as C. nobilis Lour. C. x indica Tanaka Indian wild orange C. x insitorum Mabb. ( Citroncirus webberi J. citrange Ingram and H. E. Moore) C. x limon (L.) Osbeck lemon C. x macrophylla Wester alemow C. x microcarpa Bunge calamondin C. x taitensis Risso (= x jambhiri Lush.) rough lemon C. x latifolia (Yu. Tanaka) Tahitian lime subtribe Clauseneae of the Aurantioideae to the subtribe Aurantieae (Citreae) and curry bush, which But et al. (1984) placed in Murraya sect Bergera, is once again Bergera koenigii L., not Mu. koenigii (L.) (Samuel et al. 2001), and remains within the Clauseneae. Bergera may be junior synonym of Clausena, and Samuel et al. (2001) provided molecular evidence to support the return of the monospecific Merrillia caloxylon (Ridley) Swingle to Murraya sect Murraya. Confusion surrounds the status of Mu. paniculata and Mu. exotica L. There may be a single highly variable species (Mu. paniculata), the latter and a hybrid (Mu. exotica), or two species (Mu. paniculata and Mu. exotica), as the validity of Mu. exotica as a species is uncertain (see Stone 1985, Brophy et al. 1994, Mabberley 1998, Ranade et al. 2006). There may be other, undescribed, species. Phytochemical differences between Mu. paniculata and Mu. exotica were reported by Li et al. (1988). In our ACIAR project, significant differences in morphology and molecular biology have been detected between them using RAPD analysis (Inggit Puji Astuti, Universitas Gadjah Mada (UGM), pers. comm.). At this point, what is considered to be Mu. exotica (sensu Stone 1985) is growing and flowering more prolifically than Mu. paniculata (sensu Stone 1985) at one of the ACIAR project s experiment sites in Central Java where the susceptibility of Citrus spp., Citrus hybrids and some Citrus relatives as host of D. citri and 'Ca. L. asiaticus' was being evaluated. D. citri shows a clear preference for Mu. exotica rather than for Mu. paniculata at this site (Achmed Hamiwan, UGM, pers. comm.), and symptoms of HLB have also been observed in Mu. exotica following confirmed inoculation with 'Ca. L. asiaticus' (Siti Subandiyah and Achmed Hamiwan, UGM, pers. comm.). Li and Ke (2002) reported detection of the pathogen in Mu. paniculata, but it is possible that the plants used in their study were Mu. exotica. Likewise, recent extensive records of Mu. paniculata as a host of 'Ca. L. asiaticus' and 'Ca. L. americanus' in Brazil (Lopes 2006, Lopes et al. 2006a,b) may actually pertain to Mu. exotica (Silvio Lopes and Andrew Beattie, pers. discussions, December 2006). Most Murraya observed by Beattie with colleagues in Guangzhou, Guangdong, resemble Mu. exotica (sensu Stone 1985), and plants sent from the Xishuangbana Botanic Gardens in Yunnan to Indonesia for our studies were, although being labelled Mu. paniculata, closely resembled Mu. exotica according to their morphology and molecular biology. Plants from the University of California, Riverside, again labelled Mu. paniculata were also found by the same methods to be Mu. exotica (Inggit Puji Astuti, Indonesian Botanic Gardens, Bogor, pers. comm.). 7

Table 3. Host records for huanglongbing and its vectors on Murraya: Las = 'Ca. L. asiaticus' and Lam = 'Ca. L. americanus'. Host D. citri T. erytreae 'Ca. Liberibacter' spp. Murraya paniculata Murraya paniculata var. paniculata Murraya exotica One of the food-plants in Taiwan (Maki 1915, Kuwayama 1931); noted as an alternative host (He and Zhou 1935); many nymphs and adults observed on plants over several years (Miyatake 1965); new host (Cheema and Kapur 1975); preferred host (Singh and Nimbalkar 1977); all stages frequently observed in Bangladesh (Catling et al. 1978); favourable host on green hedge plant (Tsai et al. 1984); suitable host in Réunion with up to 11 generations in a laboratory and 10 in the field (Aubert and Quilici 1984); complete development and a preferred field host in Asia (Aubert 1987a,b); all stages (Osman and Lim 1992); marked increase in populations (Koizumi et al. 1996); host on which most abundant (Tsai et al. 2000); present in all seasons in Florida (Tsai et al. 2002); preferred host in Florida surveys (Halbert and Manjunath 2004) breeding field populations in the Philippines (Catling 1968); a host plant in China (Huang 1953); all stages present on Kunti or China box (Khan and Borle 1989) adult feeding in laboratory cages (Aubert 1990b); marked increases in populations (Koizumi et al. 1996); a preferred host in southern China (Beattie pers. observation, Guangzhou 2006) where, probably erroneously, only M. paniculata has been recorded as a host (see Yang et al. 2006); a host in Okinawa (Yasuda et al. 2005, Nakata 2006). Common host with occasional complete development (Aubert 1987a,b) attacked to slight extent (van der Merwe, 1941) Las: suspicious symptoms: external and internal symptoms like CVPD after vector transmission (Tirtawidjaja 1981); transmission by D. citri from infected citrus (based on internal phloem degeneration, external mottling and chlorosis, and stunting) (Tirtawidjaja 1981); can harbour greening organism (Aubert et al. 1985). EM negative; no symptoms (Koizumi et al. 1996); not a host (Toorawa 1998); not a host: no graft transmission from infected citrus (Dai et al. 2005); detection of the pathogen by PCR (Li and Ke 2002); in field trees at low rate (Lopes 2006a,b). Lam: In field trees and by graft transmission (Lopes 2006a,b) Las: no symptoms and no detection by dot hybridisation after attempted graft transmission from sweet orange infected with Las and Citrus Tristeza Virus (Hung et al. 2000) Many host records for D. citri in China that have been ascribed to Mu. paniculata (see Yang et al. 2006) should have been recorded as Mu. exotica. Clearly, there is an urgent need to resolve the status of Mu. exotica and, if it is a hybrid, its parentage, as the issue has important implications for management of HLB. Of prime concern is that under field conditions, one or both forms may be an asymptomatic host of 'Ca. L. asiaticus' either permanently after infection or temporarily. Nguyễn Huy Chung, an ACIAR John Allwright Fellow from the Plant Protection Research Institute in Hanoi, is working to resolve these issues as part of his PhD Circumscription of Murraya and Merrillia [Sapindales: Rutaceae: Aurantioideae] and susceptibility of species and forms to huanglongbing at the University of Western Sydney. VULNERABILITY OF THE AUSTRALIAN CITRUS INDUSTRY AND INDIGENOUS GERMPLASM The vulnerability of the Australian citrus industry to HLB and its vectors came from: 8

Australia s proximity to the Indonesian Archipelago, where both HLB and D. citri are endemic, and to Papua New Guinea, where their eastern-most distribution is currently limited to the vicinity of Vanimo in north-eastern Papua New Guinea (about 750 km from the border with northern Australia); the presence of indigenous species of Citrus in Northern Territory, Queensland, New South Wales and South Australia; the widespread occurrence of M. paniculata as a vigorous ornamental plant in gardens; the occurrence of wild forms of M. paniculata in coastal and sub-coastal regions of Western Australia, the Northern Territory, and Queensland; current evidence that suggests that all species and varieties of Citrus are likely to be susceptible to HLB; the location of commercial citrus orchards in regions throughout mainland Australia that have climates that are favourable for 'Ca. L. asiaticus', the pathogen that causes the Asiatic form of HLB, and its principal vector, D. citri; and the location of commercial citrus orchards in regions in southern Australia that have climates that would be suitable in most years for 'Ca. L. africanus', the pathogen that causes the African form of the disease, and its principal vector, T. erytreae. The greatest threat to the Australian citrus industry and indigenous species of Citrus will be from simultaneous incursions of both 'Ca. L. asiaticus' and D. citri and significant spread by both before detection. Entry Pathways The most likely pathways of entry of HLB and its vectors are illegal introductions of budwood; legal importation of infested or infected material that has been inadequately treated and/or inspected; passive transport of adult psyllids in commercial and military aircraft; air currents (e.g., tropical cyclones and jet-streams) with the potential to carry psyllids from the Indonesian Archipelago, Papua New Guinea and South Pacific islands to the southeast of Papua New Guinea; movement of people carrying citrus fruits and other plant material across Torres Strait from Papua New Guinea; and unregulated landings of boats carrying citrus from other areas to the north of Australia. The risk of airborne entry of D. citri from Indonesia, Papua New Guinea and, in the future the Solomon Islands and other Pacific islands, should the psyllid spread to these locations, can be considered high and will increase with the inevitable spread of the psyllid through Papua New Guinea. This risk could be greatest when the psyllid reaches the Milne Bay region on the southeast tip of the Papua New Guinea mainland and nearby islands. Evidence to support this entry risk as high can be gleaned from the introduction of the leucaena psyllid, Heteropsylla cubana Crawford [Hemiptera: Psyllidae] into Australia. Bray and Sands (1987) hypothesised that it probably entered Australia on air currents from the Western Pacific to the northeast of Queensland. It was first recorded in Australia near Bowen in April 1986, shortly after a particularly severe cyclone in the area and within three months the psyllid was recorded near Gympie, more than 10 km from a main road and some 830 km south of Bowen, and by mid-october in Brisbane (Bray and Sands 1987). Bray and Sands (1987) considered it most likely that the insect was dispersed throughout Queensland by wind currents. Others (Sakamaki 2005, Gottwald et al. 2007) have speculated that long-distance movement of D. citri is governed by seasonal winds and through riding lower jet airstreams (geostrophic movement) associated with summer monsoons or related to air masses during hurricanes or tropical storms. Modelling to assess the risk of wind-borne incursions of D. citri could enable the development of an early warning system to alert authorities to a possible incursion. POST-INCURSION SURVEYS AND RESPONSES Draft recommendations for eradication plans, and alternative responses, are being prepared for the IMP for incursions of HLB and/or its vectors. These include: recommendations for the establishment of quarantine zones based on scenarios related to the climate, topography, size 9

and degree of isolation of locations (e.g., towns, cities and orchards) where citrus and/or alternative hosts of the disease and its vectors occur, and regions where native hosts occur naturally; guidelines for delimiting surveys, including trace-back and trace-forward analyses; and responses based on where host plants are located in relation to the initial point of detection. For situations where eradication of D. citri or T. erytreae in the absence of HLB (determined by surveys and PCR-testing of fifth instar and adult psyllids) may be feasible, the following actions can be considered: immediate application of recommended insecticides to all host plants; skeletonisation of all Citrus trees in orchards, home gardens and elsewhere (excluding nurseries); skeletonisation of, and removal of all leaves from, grafted Citrus trees, and destruction of all seedling rootstocks, in nurseries; and elimination of all Murraya, Bergera and Clausena shrubs and trees. Such action should optimise the prospect of eradication, as the psyllid cannot survive in the absence of host canopies. Skeletonising, as an alternative to tree eradication, will minimise costs and allow orchards and nurseries to return to full production sooner. It may be feasible to use defoliants in such circumstances in orchards to achieve rapid removal of foliage on the psyllid. However, currently no registered defoliants for citrus and doses (concentrations and spray volumes) would need to be determined. Outcomes are likely to be variable depending on chemical rate, spray application method, tree age, water relations and environmental conditions at time of application. It may be possible to apply defoliants with insecticides, but the feasibility of such tank mixes would need to be carefully considered. Insecticides available for suppressing psyllid populations include the organophosphates azinphos-methyl, chlorpyrifos and methidathion, the carbamate carbaryl, and the neonicotinoid imidacloprid, all of which are known from overseas reports to be effective against the psyllids. Permits for the use of these chemicals need to be in place before an incursion occurs. Presence of HLB in the absence of its vectors will most probably lead to death of infected host plants and the disease would, therefore, not be viewed as a severe threat to the industry. Nonetheless, tree eradication, most probably confined to a small area, should occur as soon as possible after confirmation of infection. Our recommendations for eradication of HLB following an incursion and in the presence of one or both vectors will depend on the various scenarios that are being considered, particularly the geographical size of the area affected, its degree of isolation, and the distributions of infected plants and psyllid infestations within the area. Cost Sharing In Australia, there is formal cost sharing agreement between the Australian citrus industry and the federal government. The agreement is based on an Emergency Plant Pest (EPP) Response Deed. It allows the Australian citrus industry to be directly involved in decision making, including biosecurity and risk reduction measures. It also reduces delays in securing funding and removes uncertainties and disincentives for growers to report suspected emergency plant pests. It only deals with responses to emergency pests covered by a cost sharing agreement. The relative share of the total cost of the incursion management that will be covered by the industry and the government, respectively, varies according to the public and private benefits to be obtained from eradication. A categorisation group is responsible for determining a cost-sharing category applicable for high priority pests. Only emergency pests that have a high impact or establishment potential are considered for categorisation. The Government and Plant Industry Cost Sharing Deed in respect of Emergency Plant Pest Responses lists HLB as category 2 for which 80% funding for managing an incursion is provided by the government and 20% by the industry, with the industry component comprising 97.1% from Australian Citrus Growers Inc. and 2.9 % from the Nursery and Garden Industry Association (PHA National Citrus Industry Biosecurity Plan: (www.planthealthaustralia.com.au). In the 10

Cost Sharing Categories for Emergency Plant Pests, a key criterion for determining the pest category is the economic impacts on regions or the national economy. The basis for this criterion is that where there are adverse regional impacts from a pest incursion, there is a case for government involvement because of the social costs and disruptions that occur to non-producer stakeholders. Without financial commitment from government, producers could incur the full cost of control measures, whilst other non-producer stakeholders could benefit from control measures without contributing to the cost of the control. Additional information is provided by Barkley and Beattie (2008). POST-INCURSION MANAGEMENT OF HUANGLONGBING AND ITS VECTORS Australia is in a fortunate position of being able to plan and legislate for effective management programs before incursions of HLB and its vectors occur. Biological control will require importation and releases of the primary parasitoids of the vectors 6, conditions that favour the survival and activity of these parasitoids and a range of endemic predators (particularly ladybirds, lacewings and spiders), and possible augmentative releases of predators. Greater use of windbreaks will be required and all nursery trees will need to be propagated in insect-proof screen houses. The timing and synchronisation of flushes and levels of flushing will need to be carefully managed, where possible, by timing and frequency of irrigation, and likewise, fertilisers, the latter without jeopardising nutritional requirements of trees. In the absence of D. citri or T. ertyreae, eradication of the HLB should be, as noted above, easily achieved as the pathogens are most commonly spread by grafting of infected buds, marcotting (air-layering) (see da Graça 1991), from tree to tree by root grafting (see Broadbent et al. 1988). Laboratory experiments reported by Tirtawidjaja (1981), Garnier and Bové (1983), Olfato et al. (1991), Garnier and Bové (1993), Duan et al. (2008) and observations that the pathogen can multiply and spread within infected dodder (Cuscuta spp. [Solanales: Convolvulaceae]) suggest that natural transmission by these parasitic plants, though considered unlikely by Halbert and Manjunath (2004), may be possible. However, dodder-covered beds of citrus seedlings were observed in a nursery in Pakistan in 2006 by one of the authors of this document, and budwood for propagation of the seedlings was sourced from an adjacent block of trees in which HLB was prevalent 7. Transmission via citrus seeds (Tirtawidjaja 1981) at low frequencies may also be possible. This could pose risk for propagation of rootstocks by seeds from infected trees. Nevertheless, such means of transmission would be easily dealt with and would not pose threat to the Australian citrus industry. Information provided by Ke and Xu (1990) suggests that effective management of HLB in the presence of D. citri or T. erytreae will require annual rates of spread (new positive detections) of the disease in orchards by either psyllid to be < 0.2% (1 in 500 trees), preferably zero. In Australia, particularly for D. citri, this will only be achieved costeffectively through area-wide use of pathogenfree rootstocks and scions, regular monitoring of trees for HLB symptoms, and mandatory destruction of infected trees immediately after presence of the pathogen is confirmed. Under these circumstances, it may be possible to continue sustainable production of citrus with integrated pest management (IPM) programs based on: biological control; minimal use of strategically timed applications of mineral oils and systemic and contact insecticides; and cultivation, as intercrops or groundcovers, of plants that produce volatile repellent compounds that slow ingress of D. citri populations, and therefore HLB, into orchards. At this point, it is envisioned that spray programs used in conjunction with mandatory removal of infected trees could comprise one contact-systemic insecticide in winter before spring growth flush on 20% of trees exceeds lengths of more than 5 mm, and 1-4 contact-systemic in summer in conjunction with 3-4, 0.25% to 0.5%, 6 The two primary parasitoids of D. citri will need to be introduced from several regions in Asia. 7 These observations suggest that transmission of liberibacter between species, genera and families of plants may occur naturally or at least be possible. 11

horticultural or agricultural mineral oil (preferably nc24 products) sprays commencing when summerautumn growth flushes 5 mm are present on 20% of trees within blocks of trees of the same age and cultivar; it may be possible to include fungicides in some of these sprays. The aim of such programs would be to kill psyllid eggs, nymphs and adults, and to reduce feeding by adult psyllids and oviposition by adult females. The insecticides registered for use on citrus in Australia and suitable for use against D. citri are limited and include azinphos-methyl, chlorpyrifos, carbaryl and abamectin. These insecticides have withholding periods of 14, 14, 3 and 7 days, respectively. Imidacloprid is registered, with a withholding period of 20 weeks, for soil applications against citrus leafminer Phyllocnistis citrella Stainton [Lepidoptera: Gracillariidae], black citrus aphid Toxoptera citricida (Kirkaldy) [Hemiptera: Aphididae], pink wax scale Ceroplastes rubens Maskell [Hemiptera: Coccidae], and red scale Aonidiella aurantii (Maskell) [Hemiptera: Diaspididae]. Registered pyrethroids are limited to permethrin, bifenthrin and gamma-cyhalothrin. It will not be possible to prevent spread of HLB solely through chemical suppression of D. citri populations. Limiting rates of spread to acceptable levels through such practices will be difficult, probably impossible, to achieve and not economically or environmentally sustainable. Since the 1970s, the clean and green Australian citrus industry has been blessed by very effective biological control and IPM programs based on low pesticide use and relatively stable orchard ecosystems with low pest and disease pressures (Smith et al. 1997). D. citri management programs based on chemical use will require marked increases in the annual number of pesticide applications 8 and significant improvements in spray application (common low-profile airblast sprayers will not achieve required levels of spray coverage and air turbulence associated with their use will enhance dispersal of D. citri and HLB within orchards). Such programs will, through destruction of natural enemies, lead to increased use of pesticides for control of other pests, particularly scales, mealybugs, thrips and mites that currently occur at levels rarely requiring chemical intervention. Increased use of sprays will also: lead to pesticide residues that may exceed processing, local fresh fruit and export MRLs; reduce orchard biodiversity through impacts on non-target organisms (invertebrates and vertebrates); lead to pesticide resistance in a range of pests; and increase risks and liabilities for growers and their employees. In the absence of HLB, it should be possible to rely on biological control, possibly in conjunction with summer-autumn mineral oil sprays and one to two applications of contact-systemic insecticides, for control of D. citri. Less chemical intervention will be required for control of T. erytreae than for D. citri. In the context of the above, we will propose that the Australian citrus industry should adopt the following recommendations for control of HLB, and its vectors, in orchards where both the disease and one or both vectors occur: use of certified pathogen-free buds produced under a certified budwood scheme (no use of uncertified buds or marcotts); mandatory area-wide management practices; quarterly, or more frequent, monitoring by trained personnel of orchards for symptoms of HLB; use of PCR or, alternatively, if validated as appropriate, the iodine-starch test (IST) in conjunction with microscopy to detect phloem degeneration, to confirm HLB infection in symptomatic leaves; mandatory 9 and immediate destruction of HLB-infected trees (those that are PCR or IST positive) by cutting each trunk near its base and applying glyphosate to the stump to kill the roots (this must be done without waiting to harvest any mature fruit); mandatory and immediate destruction of all trees in a block should the percentage of HLB-infected trees in a block reach or exceed 10% of trees within an interval of 12 months; mandatory destruction of abandoned orchards; 8 Spray applications per year could easily exceed 10 in cooler regions and 20 or more in warmer regions. 9 Legislation will be required for mandatory removal of plants. 12

mandatory removal of HLB-infected trees in home gardens; mandatory removal of alternative hosts of the vectors, particularly species of Murraya, Bergera and Clausena, within close proximity (2 km) of orchards; geographical isolation of budwood sources (mother trees), and their maintenance in insect-proof screenhouses; geographical isolation of nurseries from orchards, and of orchards from alternative hosts of the disease and its vectors; nursery production, in HLB-affected regions, in secure insect-proof screen houses; legislation to prohibit transport of HLBinfected or vector-infested plant parts (e.g., nursery trees, budwood, fruit and seeds) to unaffected areas; management of planting densities, canopy dimensions and canopy densities to optimise yields per hectare and effective application of sprays, and to minimise light intensities that favour psyllid infestations; planting of windbreaks to minimise movement of adult psyllids within and between orchards; hedging and pruning practices timed and undertaken to minimise the risk of enhancing vector populations; orchard and nursery management of D. citri and/or T. erytreae with strategically applied insecticides and mineral oils; - timing of sprays should be based on host-plant phenology so as to minimise feeding and oviposition and to maximise mortality of eggs, nymphs and adults; - application of sprays should be even and thorough, with sprays applied to run-off; - use of soil drenches and tree injections should be based on tree size and phenology, and account for potential loss or diminution of active ingredient(s) through leaching, degradation or tree growth; strategies to encourage, where feasible, biological control of the vectors by their natural enemies (e.g., planting of groundcover plants to encourage generalist predators); growing plants (as intercrops and/ or ground-covers within orchards) that produce volatiles that repel the vectors, thereby slowing ingress of HLB into orchards; timing irrigation (where feasible) and fertiliser applications to regulate the timing, number and extent of flush cycles; use of supplementary overhead irrigation to reduce psyllid populations; and education of farm and nursery personnel; pesticide manufacturers, distributors and retailers; consultants, technical advisors, and scouts; research, quarantine, regulatory and advisory staff within government departments. The concept of growing plants that produce volatiles that repel D. citri, and thereby slow ingress of HLB, into citrus orchards stems from observations and studies in southern Vietnam (see Beattie et al. 2006). Evidence to support repellent volatiles as the reason for the impacts that have been observed in experiments, and in field observations by Vietnamese farmers, extension officers and scientists, has been obtained at South China Agricultural University (Cen et al. in prep., Zaka et al. submitted). Visits by scientists (Tim Gottwald and David Hall, from the United States Department of Agriculture) and growers (Michael Stewart, Tim Gast and David McCullough) from Florida to Vietnam has led to interest in developing such technology for use in Florida (see http://www. fcprac.com/final-priorities-list-design.pdf). REFERENCES Asana, R.D. 1958. The citrus die-back problem in relation to citrus cultivation in India. Indian Journal of Horticulture 15: 283-286. Aubert, B. 1987a. Epidemiological aspects of the greening (huanglungbin) disease in Asia. In: The FAO-UNDP Project Coordinator in Fuzhou (ed.), Regional Workshop on Citrus Greening Huanglungbin Disease, Fuzhou, China, 6 12 December 1987. 5 pp. Aubert, B. 1987b. Trioza erytreae Del Guercio and Diaphorina citri Kuwayama (Homoptera: Psylloidea) the two vectors of citrus greening disease: biological aspects and possible control strategies. Fruits 42: 149-162. Aubert, B. 1990a. High density planting (HDP) of Jiaogan mandarine in the lowland area of Shantou (Guangdong China) and implications for greening control. In: B. Aubert, S. Tontyaporn, D. Buangsuwon (eds), Proceedings of the Fourth 13