sign in sign up
<<
Home , Taro

Archaeol. 38 (2003) 192- 202

Taro planthoppers (Tarophagus spp.) in Australia and the origins of taro ( esculenta) in Oceania

PETER J. MATTHEWS

Abstract

Taro plamhoppers (Taroplwgus spp.) may be associated exclu­ (indigenous) to the region (Matthews 1990, 1995). sively or primarily with taro (Colocasia esculema). and the geo­ Subsequently, in Australia. I found large numbers of graphical distribution orr. proserpina provides circumstantial evi­ another insect, Tarophagus (literally, ' taro-eating'), on dence that taro is native to the Sahul continental region (as well as wild and apparently wildtype taros in the coastal region being native to Sunda). r. colocasiae ( Matsumura) (Asche and near Cairns, Northeast , in wet tropical rain­ Wilson 1989a.b) is reported here for the first time in Australia. and forest. In this paper, I introduce the genus Tarophagus the genus Taroplwgus is reported for the first time on the wildtypc and its species in and the Pacific, report my own fonn of taro (C. esculema). Three species of taro hopper arc findings of Tarophagus in Australia. and then interpret present in Asia and the Pacific. T. proserpina has a relatively nar­ the distributions of Tarophagus species in relation to the row distribution extending from eastern to origins of taro in Oceania. I also discuss plant viruses (Remote Oceania). This distribution adds support to the suggestion associated with taro (the host) and Tarophagus (the vec­ that Polynesian taros are derived from a Melanesian taro gene tor), and recent studies suggesting that there are multiple pool. It is tentatively suggested that different Tarophagus species taro gene pools in Asia and the Pacific. evolved in a~sociation with different taro gene pools. before the In this approach to the history of taro I am crossing domestication of taro in multiple regions within Asia and the disciplinary boundaries, just as I did during 1985 to 1990 Pacific. Plant viruses associated with taro and Taroplwgus are also as a s tudent of Doug Yen, in the Department of discussed. Prehistory headed by Jack Golson, at the Australian National University (ANU). Both scholars encouraged me in this dangerous activity, while making clear the per­ ils of treating any one discipline in too shallow a fashion. Insect associations with cultivated are important Another teacher at ANU was David Shaw, an entomolo­ for many reasons, most notably because insects are often gist who encouraged a healthy respect for the difficulties major pests and can also transmit various diseases. The of biological observation and interpretation. A lthough economic impacts of insects on crops in the past are digging deeply as an archaeologist continues to elude me rarely known or knowable, allhough archaeologists have as an occupation, I am still digging, as best I can, in found storage pests (weevil s) among archaeological other fields that can contribute to our understanding of remains of crops (Yartavan 1990). Spriggs ( 1982) the past. noted the possibility of archaeological preservation of insects associated specifically with taro. giving Papuana spp. (taro beetles) and Tarophagus prosetpina (taro plan­ Tarophagus ecology, , and geographical range thopper) as examples. Both genera are well known as agri cultural pests on cu ltivated taro (G agne 1982). The taro planthopper genus, Tarophagus, has three recog­ Another way to approach agricultural history is to inves­ nised species. and these are considered important insect tigate the biogeography of living insects associated with pests on taro. The entire life history of a taro planthopper living plants. Thi. is the approach that I will introduce occurs on taro , above ground. Taro planthoppers here with respect to taro, Colocasia esculenta (L.) feed on sap and heavy infestations can cause plants to Schott. wilt and become stunted. Feeding and egg-laying punc­ In J uly 1985, near Lae, in , I tures cause sap exudation which forms red encrustations observed an insect pollinator. Drosophilel/a pistilico/a , on the plant. The planthoppers are also vectors of taro on wildtype taro in an area where these drosophilid tlies bobone virus (a rhabdovirus), and this is of concern for had only recently been d iscovered and described by taro growers because the resulting disease can stunt or Carson and Okada ( 1980, 1982). Although sparse, the kill plants . Tarophagus species are widely distributed existing literature on Drosophilella strongly suggested from (including and the Ryukyu Islands that wi ld taro populations in could be nati ve of southern Japan), through to Australi a (N011hern Territory and Queensland), Papua New Guinea, National Museum of Ethnology. Senri Expo Park, Suita-sh i. . and many Pacific island countries Osaka 565- 851 1, Japan. (Gagne 1982, Zettler eta/. 1989. Asche and Wilson

192 Tarophagus belongs to the family Delphacidae, a large 146" insect group in which most species feed on grasses. Banks (1770) Many of them are pests on crops such as , sugarcane, COOKTOWN wheat, rye, corn, and forage crops (Ki simoto 1994, Fletcher and Lariviere 2001). Delphacid planthoppers are characterised by two wing forms (short and long) that I i Queensland appear either in both sexes or in one of the two. The L.l long-winged forms can fly long distances. For example, in 1967, huge numbers of migrating rice planthoppers were discovered on a weather observation boat floating on the Pacific Ocean, at least 500 km distant from the Japanese mainland (Ichikawa 1994). The possibility of transoceanic migration by the taro planthopper has not CORAL SEA been investigated. If taro is the only host for Tarophagus (or in regions where no other hosts are present), effective transoceanic migration by flight would require (or might be made possible) by the prior establishment of natural or introduced populations of taro, at every destination. The known range of the genus Tarophagus is tropical to subtropical, with no reports from northern Japan or from , at the northern and southern limits of cultivated taro in Asia and the Pacific. ln Japan and New Bailey (1899) Zealand, cold temperatures and the winter loss of leaves by temperate-adapted forms of taro might prevent over­ wintering and survival of Tarophagus. The genus has not been reported in Subsaharan Africa (see Smithsonian 200 1), although taro is an ancient and widespread crop in Africa. Asche and Wilson ( 1989a) noted that taro planthop­ pers were identified in most previous reports as T. pros­ erpina (Kirkaldy). After looking at spec imens from throughout Southeast Asia and Oceania, these authors 18" recognised three morpholog ically distinct species, T. colocasiae, T. persephone, and T. prose1pina. The known d istributions of these spec ies (see later) overlap and together span most of the possible natural range of taro (Matthews 1991 , 1995, 1997), as well as extending into Remote Oceania where taro is an entirely introduced crop. So far, Tarophagus has only been reported on taro (Colocasia esculenta). Other species of Colocasia are mostl y non-domesticated, wild species located in main­ 0 50km • land and island Southeast Asia. The presence or absence o f Tarophagus on these close relati ves of taro has not been investi gated. In the fol lowing sections, T. co!o­ Figure I. Map of wildtype taro sites in Queensland, 1987 casiae is reported for the first time in Australia, and the and 1992. Tarophagus was found at the sites numbered association of Tarophagus with wildtype taro, in natural 1- 12 (see Table I for details). Illustration adapted from habitats, is also reported for the first time. Although T. Matthews ( 1990). Early historical reports of taro (dia- persephone (Kirk a ldy) was re po rte d in Northe rn monds) are labe led with name of observer and date. Ten·itory and Queensland by previous authors, the host and habitats were not described.

1989a, b). Eggs are laid in the bases of petioles and the Field work and col!ecrion methods midribs of leaves. Young nymphs are creamy white, and later stages develop predominantly black with white Field surveys in Queensland, Australia, were can·ied out markings. From egg to adult takes about 18 days. These during the dry season, in September 1987 (Matthews details may vary according to species, and have not been 1990) and again in August 1992 (Matthews 1992). The studied with reference to the taxonomic descriptions pro­ primary aim of these surveys was to locate and describe vided by Asche and Wilson ( l989a, b). wild taros and their habitats (Figs. I and 2). Taro plan-

193 form. The insec ts were easily approached with the tube. and were quickly overcome by the chloroform vapour. One tube could be used repeatedly without closing the lid. The tubes were posted to Canberra for cold storage. All Taropftagus specimens were collected from what is now regarded as wildtype taro. a wild form of taro with long stolons. entirely green leaves. and extreme acridity (Mauhews 1997). Spec imens from eight different sites (localities) were sent toM. J. Fletcher and are now lodged in the entomo logy collection of the SW Agricultural Scientific Collections Unit (ASCU). Orange. ~

ldentiji'cation and location records

Two specimens were iden tified for the author as T. colo­ casiae. by J. F. Donaldson (from the Russell River site, 1987) and by M . J. Fletcher (from one of the 1992 sam­ ples and sites. sample not identified). Various juvenile and adu lt forms were seen. but were not studied in detail. Samples from six of eight sites included short-winged adu lts (Cooper Creek tributary; Saltwater Creek; South Mossman Ri ver crossing; Isabella Falls; Kearney's Falls: and Zillie Falls) (M. J. Fletcher pers. comm. 200 1). A nymph of Tarophagus sp. is shown in Fig. l. and an adult ofT. co/ocasiae is shown in Fig. 4. It is difficult to iden- ~ tify nymphs because the taxonomic desc riptio ns fo r Taroplwgus species are based on adult morphology (J. F. Donaldson. pers. comm. 1987). Taroplwgus si tes recor ded by the autho r i n Queensland are mapped in Fig. I and described in Table I . The twelve si tes span large distances from norrh to south and west 10 easr. bur represent only a small fraction of the full range o f wild taros in Quee nsland and (see Fig . I . and Matthews 1990, 199 1) . In Fig. I . the lack of recorded Tarophagus at a w ild taro si te does not imply absence . fo r I d id not always look for insects and the time spent at some sites was very brief. In other words. the search for insects was more opportunistic than systematic.

Addit io n a I records have been prov idcd by J. F. Donaldson (pers. comm. 200 I ): Figure 2. Isabella Falls, Queensland. a typical habitat for T. colocasiae wildtype taro. and collection site for Toroplwgus sp. (4th . Aucust 1992) (s ite 4. near Cairns. Fi12. l ). I. on taro, Darn ley I sl and. Torres S trait (J. F. ~ ~ Donaldson. 29 March 1990) ii. on taro. Bamaga. Qld (J. W. Turner. 28 February 1999) thoppers were mostly seen on the undersides of fully emerged blades. T hey were also seen on the upper T. persephone petioles or heav ily infested leaves. ;\dulls are 2- 3 mm 111 at light. Iron Range. Q ld (R. J. . 13-20 M ay long . When a leaf was lllrned over gently. adu lts and 1995) j uveniles sometime:-- walked with a sideways or diagonal iv on Mimosa pigra, Berrimah. T (C. Wilson. 25 motion across the leaf surface. When disturbed by a jolt September 1989) 10 the leaf. the escape response was a sudden large jump. v on Sida cordifolia . Fogg Darn. NT (C. Wilson, 3 away from the plant. To collect planthoppers. I used a 1.5 Apri l 1985) . ml plastic Eppendo1f tube containing a small twist of tis­ VI in l ig ht trap in Mimosa pigra, C.P.R.$ .. NT (C . sue paper impregnated with two or three drops of chloro- Wi Iso n. 4 October 1988).

194 # ldenti ficat ion Location Approx. latitudeS. Collector & date & longitude E.

T. colocasiae Jiyer Cave. 17 ° 26' Matthews (J.D.) Russell River 147°47' 20.9.87 2 Tarophagus sp. Moochoopa Falls. 17 ° 25' Hinxman (J.D .) upper Russell R. 145 ° 47' 8.10.89 3 Taroplwgus sp. Milmilgee Falls. 16 ° 58' Matthews (P.J.M.) Freshwater Creek 145 ° 40' 2.8.92 4 Tarophagus sp. Isabella Falls. 17 ° 02' Matthews (P.J .M.) Isley Hi lis 145 ° 43' 4.8.92 5 Taroplwgus sp. Kearney's Falls. 17 ° 04' M auhews (P.J .M.) Bellenden Ker N.P. 145 °47' 5.8.92 6 Taroplwgus sp. Elinjaa Falls. 17 () 30. Matt hews (P.J .M.) Atherton Tableland 145 ° 39' 5.8.92 7 Taroplwgus sp. Zillie Falls. 17 ° 29' Manhews (P.J.M.) Atherton Tableland 145 ° 39' 5.8.92 8 Tarophagus sp. Saltwater Creek. 16 ° 24. Mat! hews (P.J.M.) vic. Mossman 145 ° 22' 7.8.92 9 Taroplwgus sp. Cooper Creek trib., 16 ° 11' Matthews (P.J.M. ) vic. Cape Trib · n .P. 145 °25' 7.8.92 I 0 Taroplwgus sp. Stewart Creek, 16 ° 16' Matthews (P.J.M.) Daintree R. trib. 145 ° 19' 8.8.92 II Tarophagus sp. Mart in's Ck Falls, 16 ° 15' Matthews (P.J .M.) Daintree R. trib. 145 ° 18' 8.8.92 12 Taroplwgus sp. S. M ossman River crossing. 16 ° 29' Matthews (P.J .M.) Cook Hwy 145 ° 24' 9.8.92 Table I. Records ofT. colocasiae and Taroplwgus sp. in Northeast Queensland. 1987- 1992. Numbers (#) refer to the sites numbered in Fig. I.

I note that the plants Mimosa pigra and Sida cord(folia cane cultivation. Next, two sites are described in detail in are both major introduced weeds in northern Ausrralia. order to show how habitats varied. Neither species is likely to be a true host for Tarophagus. One of the most isolated sites visit.ed was Jiyer Cave. The insects may have been collected on or near taro in next to the Russell River (site I . Table 2 and Fig. I). Here. habitats that were invaded with weeds. wildtype taro plants were abundant and firmly established by roots and stolons among rocks below a small waterfall. deep in rainforest. A/ocasia sp. (presumably A. brisbanen­ The habitats of wildtype taro sis Hay and Wise 1991 ) was also present in a drier situa­ tion nearby. Isolated clumps of taro of the same phenotype In Queensland there are many stable populations of wild­ were also seen on bends of the· river within a few km type taro in and around waterfalls, where rocky crevices below Jiyer Cave. One isolated clump of an apparently allow strong root-holds. Taro was also found on soft escaped cultivar. with red colouring on the , was banks next to fast streams or rivers. habitats that are also seen opposite the site of a fanner Chinese gold-pan­ unstable because of erosion during periods of heavy ners· camp on the river. also a few kilometers downstream water flow (Mat!hews 1995. 1997). Stable taro popula­ from the cave. Jiyer Cave has been the site of archaeologi­ tions in slow water courses were more common in the cal excavations showing late Holocene occupatio n lowlands in the recent past, within the living memory of ( Horsfall 1996), and was still used as a campsite by local observers. Sugarcane farmers at Ingham ( 1987 sur­ Aborigines in the early 20th century. and by recreational vey) and near Edmonton ( 1992 survey) reported that visitors to Bellenden Ker National Park in recent years. wild taros were more abundant in the lowlands when One of the least isolated sites found was on a bank of fewer swamps had been drained and cleared for - the South Mossman River, at its intersection with the

195 lmm

Figure 4. Dorsal view of Tarophagus colomsiae (Matsumura), adu lt (with short wings), collected in Queensland by Matthews. August 1992. The pointed abdomen indicates that this specimen is a female. Drawing by A. Westcott. SW Agriculture (counesy M . J. Fletcher). Figure 3. ymph of Toroplwgus sp. from a leaf of wild taro collected by R. Hinxman. 8th October 1989. in rain­ forest at M oochoopa Falls. Bellenden Ker ational Park. Northeast Queensland. Upper photo: whole body. Lower been reported in taro. Flowering dates in Queensland span at least seven months. from March (R. Hinxman. photo: front of head. Scale bars I 00 um (0. 1 mm) in both pers. comm. 1989) to October (author's field-work 1987. photos. The antennae, head structure. number of abdomi­ 1992). but the periods for individual plants or sites are nal plates. and foliate spur on the ream1ost leg appear the same as shown for the adult stage of Toroplwgus. in not known. At any one site. drosophilid flies might visit A sche and Wilson ( 1989a.b). Photographed with a inflorescences only briefly or sporadically. A ltematively. Joebel scanning electron microscope by D. Rentz. other pollinators may be involved. or rainfall in this area may facilitate self-pollination of the plants. as Ivancic et a/. ( 1995) reponed in Papua cw Guinea. Whether or not th e insect pollinators and planthoppc r~ Cook Highway. At thi ~ si te. a dense patch of wildtype nrc associated exclusively with taro is not entirely certain. taro grew in so ft mud and accumulated detritus. benea th Efforts to find these insec ts on other plants. in wild and a remnant of riparian broadleaf rainfores t. This remnant cultivated habitats. arc needed to properly test the limits of of forest was surrounded by sugarcane fields. host-range for each insect species. The known geographi­ cal distributions of the three Taroplwgus species overlap in the eastem pa11 of mainland New Guinea (Figure 5). so Taro and insects in Australia this might be a key area for investigating relationships between Tarophagus species and their plant hosts. During the fieldwork in Queensland. I also tried to find Most of the Taroplwgu.\ sites in onhca~t pollinating insects like those seen on taro in Papua ew Queensland were isolated with respect to human se ttle­ Guinea. in 1985. one were found, despite the fact that ment and/or taro cultivation. Taro cu ltivars and cultiva­ fruiting heads with many seeds were present. The pres­ tion are sa id to have been introduced from the Pacific ence o f seeds strong ly suggested that pollination had Islands by labourers employed in the sugarcane industry taken place. since asexual production of seeds has not during the 19th century. and taro is still occasionally

196 <. •. . .I . . I • : * ••• 0 • Q···••.. .,· .. ;· ...... ,..,-._ .., . .. ' . . .· ...... \ ...... ". ' \ ...... ~ ...... ~ ., . ... _. ······· ·······~············~- ·. .. ·······. .. .~ • • •• •• . • .· ~ .... "I • • I ••• •• • «• • • • ~·-~ . "• ' !.- • . ... d: ·~ -~~-. . .:.: \: . .· '· :- .-7·: ~t •••• •••• • • - •:...... ' • • 1: .: •• -~• • • • ,\'. :: ..... ·: ••. .: ·! ( ':" ••• .. . • •• •• •, .. • • . . . T •• • • • ~~·-· · ·.. . · Persephone . / ·;~~ ...:.r .. • •...... o· o· • • • • # • ... • • ...... _ '-rj''' / ::: . .. • 0 • • ' • - ... ··.. ~, • • .. , . _ .·..·· --- . .. ~ '········.._ - ... . • .. -...______-- -... ,. • .... •••••• ...._ t .. " 0 .... ~ -- • , . .. .. ·.•.• .. ' •'..,•. ·· P. . ', .. . . ., ·~...~ ~· .. ' 0 .• -·• ...... \ :- . . - .,, -.. ~·-~. .. • . . ~ . I .., ------~ • ---- T. proserpina------•• "' • • .• ---T. persephone ~,.,...... _. ---- T. proserpina •••••• T. colocasiae ,.•. ~

Figure 5. Tarophagus species, distribution map, adapted from Asche and Wilson ( 1989a,b) with the addition ofT. colo­ casiae in Queensland.

replanted by their descendants, in valleys behind the Taro planthoppers and viruses in Melanesia Gold Coast in South Queensland (D. E. Shaw pers. comm. 1992). For most Australians with European ori­ Mitche ll and Maddison ( 1983) li sted as an gins. taro is an unfamiliar food. In 1987 and 1992, when alternative host forT. proserpina, but gave no source for ~ I carried out my field work, taro was rarely grown as a this information. It is also unlikely that these authors commerc ial crop in Australi a. In Queensland, a few made any distinction between what are now recognised growers had started or were planning commercial pro­ as three s pec ies of Tarophagus. Asche and Wilson duction, and small quantities of fresh taro were sold in (1989a) noted that Tarophagus species are found on taro. shops in Mossman and Cairns. In 1992 I was but made no direct claims concerning host specificity. told that Vietnamese market-gardeners had started grow­ Gagne ( 1982) is perhaps most explicit, stating that the ing taro near Darwin withi n the last five years, but that entire life history of the taro leafhopper occurs on taro there was still no more than a 1/4 acre of taro grown in stems, and reporting other hosts for other pests of taro. all of the Northern Territory. Commercial production in but not for Tarophagus. Australia has increased during the last ten years, judging Despite some uncertainty, lhe existing literature does from recent conve rsations w ith vegetable selle rs in indicate that taro is the sole host for Tarophagus, so it is Sydney, but most taro consumed in Australia is still of historical interest to consider Tarophagus as a host or impo rted. The gene ral lac k of taro cultivation in vector for other forn1s of life. Asche and Wilson ( 1989a. A ustralia makes it likely, though not certain, that the b) discuss a mirid egg predator (Cyrrorhinus fulvus association of taro planthoppers w ith wild taro in Knight) that attacks Tarophagus. and here I will discuss Queensland is prehistoric, and not just the res ult of viruses transmitted by Tarophagus to taro. modern introduction on taro cultivars. Taro bobone virus is apparently endemic to Papua

197 New Guinea and the ( Mitchell and three locations in rainforest. During my own field work. M addison 1983). although it is part of a poorly defined I did not observe symptoms of any viruses among wild complex of viruses or virus strains that is more widely taros in Queensland. but th is does not exclude the possi­ distributed in the Pacific (Pearson et a/. 1999. Jackson bility that symptomless strains of virus were presen t. 1994). This virus has not been reported in taro in Much remains to be learned about the viruses associ­ Australia, with the exception of imported material held ated with taro. and their historical significance is not yet in quarantine (Jones eta/. 1980). According to a 1990 clear. Characterisation of the viruses is being carried out revision by Brunt. in Brunt et a/. ( 1996). taro bobone by Pearson and others (Pearson eta/. 1999). virus is transmitted by T. proserpina in a persistent man­ ncr. and is not transmitted by mechanical inoculation. by contact between plants. by seed. or by pollen. The host Biogeography of the taro plan/hopper and wro range of taro bobone virus also appears to be limited to taro: no natural infection of this virus has been recorded Asche and Wilson ( 1989a, b) did not report Taroplwgus in Alocasia or Cyrtospemw (aroids native to tropical in and adjacent countries (Fig. 5), despite the fact A sia and the Pacific) or and Xantlwsoma that wi ldtype and domcst icated taros and other species of (aroids native to Central and , ancl now Co/o casia are fou nd in and adjacent common in Asia and the Pacific). areas (Matthews 1991. 1995. 1997). T. colocasiae and/or Taropflagus is the on ly insect genus known to transmit undescribed species of Tarophagus might ex ist in these taro bobone virus, but the identification of T. proserpina areas. Reviewing local research literature from India and as vector cannot be relied upon since it was made before adjacent areas might help to define the western limits of the 1989 revi ·ion of the genus. Although taro bobone Toroplwgus more sec urely. In this section. I briefly virus has been reported in both Papua ew Guinea and review the biogeography of each Taroplwgus species in the Solomon Islands, Asche and Wilson ( 1989a) express Southeast Asia and the Pacific. with reference to taro and surprise that they did not find T. proserpina in the its dispersal as a cultivated plant. Solomon Islands. If this species really is absent. then taro ( I )T. co/ocasiae is the most widespread species. ran g­ bobone virus may exist in the Solomon Islands because it ing from the Asian mainland, through and the was introduced in vegetative plan ting materials, and/or to . Australia. New Guinea . because it is transmitted by T. colocasiae or T. perse­ So lomon Islands, and H awaii (Fig. 5). According to plume (both present in the Solomon Islands). For agricul­ Waterhouse and orris ( 1987), Tarophagus was first tural researchers. vegetative transmission of taro bobone established in in 1930. when its population rose ~ virus is considered a high risk. and this has discouraged and caused major crop damage. It is not known if this the export of cultivars from Papua ew Guinea and the statement is supported by any definite evidence. Perhaps Solomon Islands (Jackson 1994 ). Taroplwgus was introduced at an earlier date (pre­ The apparent severity of taro bobone virus in Papua European), but only became obvious as a pes t because of New Guinea and the Solomon Islands might mean that special environmental circumstances. Today. both T. taro in this area was previo usly unexposed, and was colocasiae or T. pro.1·e1pina arc found in Hawaii, and the therefore especially suscepti ble to recent introduction of Iauer is apparently more abundant (Asche and Wi I son the virus. In theory, the virus could ex ist undetected in 1989b).T. colocasiae might have been introduced to other regions. in resistant forms of taro. Alien pathogens Hawaii from or southernmost Japan. on taro culti­ arc often involved in the most severe diseases of plants vars brought by Chinese or Japanese immigrants during (Scheffer 1997). Alternatively. taro bobone vims might the late 19th and early 20th century. There might also have emerged locally from the complex of rhabdoviruses have been an earlier dispersal eastward. on cultivated or rhabdovirus strains (see Pearson et a/. 1999) that is taros carried from Southeast Asia and/or ew Guinea. wide-spread in Micronesia and M elanesia. From current into Micronesia and possibly as far as Hawaii. Whatever ev idence. it is not possible to say when or where taro the dispersal history may be. T. colocasiae may have a bobone virus originated. Nevertheless, if taro is the on ly more western origin than the other planthopper species. host for the rhabdovirus complex, and if Tarophagus is since it is the only species known to be widespread in the the only vector. then the overa ll association of viruses, Asian mainland. plant. and insects may be ancient, in Melanesia or else­ (2) T. persephone has been found in island Southeast where. Unraveling the history of this association is likely A sia. North Australia. and M elanesia. but not in ew to shed light on the domestication and dispersal of taro. Caledonia. M icronesia. Sumatra. Java, and most of main­ Another virus, Dasheen Mosaic Virus (DMV). is com­ land Southeast Asia. Asche and Wi I son ( 1989a:294) mon worldwide among cultivated aroids (Zettler and argued that T. colocasiae and T. proserpina are sister Hartman 1986) . is transmitted by aphids. and is not species. This makes the more central geographical posi­ restricted to taro as a host. Greber and Shaw ( 1986) tion ofT. persephone (Fig. 3) rather curious- why is it found DMV in Queensland in various cultivated spec i­ not more closely related to one or the other species? T mens of Alocasia. Colocosia. Diejfenhacltia, and persephone cou ld have orig inated on either side or Xantlwsoma, and also in wi ld Alocasia macrorrhi:os Wallacca, or within Wallacca. in a primary association (presumably A. brishanensis H ay and Wise 1991) at with natural. wild populations of taro. or possibly in a

198 primary assoc1at 10n w ith Co/ocasia oreshia and/or C. transport on planting materials, followed by establish­ gigamea. These two species are native to Southeast Asia ment in taro gardens. Although the exact eastern limits (Hay 1996), in areas that overlap or lie within the cur­ o f natural populations of w ildtypc taro are not yet rently known range or T. persephone. Speculating about known. the known distribution of T. proserpina may the original distribution and associations of T. perse­ point to eastern mainland ew Guinea as a locus for the phone is difficult because nothing is known about the domestication of taro within ew Guinea. and as a geo­ occurrence of planthoppers on o ther species o r graphical source for cultivars in Polynesia. Colocasia, or on natural. wild populations of taro in any The above suggestions are all very tentative. Whether part of Southeast Asia and Melanesia. o r not they are credible will depend on many as yet (3) T. proserpina is present in the eastern part of unknown details of plant and insect distribution. For Papua New Guinea. but has not been reported in the arc example. we need to know much more about how taro of islands from New Britain to ew Ireland and the planthoppers migrate. in wild and cultivated habitats. Solomon Islands. From eastern Papua ew Guinea, this According to Anon ( 1999a). T. proserpina lays eggs. spec ies i s found eastward through , New often two at a time. into slots cut with the oviposi tor. in Ca ledonia, and across the southern Pacific islands to the midrib of the taro leaf and also in the petioles and the Society group, and also in Hawaii. As indicated petiole bases. T he latter are a major source of infestati on above, an early (pre-European) arrival or Tarophagus in for new gardens since propagation is by means of H awaii may be in doubt, since it has been suggested tops with petiole bases. In Hawaii th e duration of the egg that the genus was es tablished in there in 1930. stage averages about 14 days and that of the five succes­ Nevertheless. the apparent abundance of T. proserpina sive nymphal instars about 4. 3, 3. 4. and 5 days. These in Hawaii (Asche and Wilson J989b) might reflect observations have two implications: ( i) the p lanting early establishment there, before a more recent arrival materials carried on long-distance canoes must have of T. colocasiae. T proserpina has al so been found in included with attached petioles. if not entire plants the Society Islands, on Raiatea in 1927 (Gillespie et a/. with full leaves. and (ii) taro planthoppers could survive 2000) and recently on M oorea (G. R. Roderick pers. long journeys as unhatched or juvenile forms inside their comm. 200 I ). The apparent absence of Tarophagus host. or on their host and protected by surrounding throughout a very large area of central Polynesia may leaves or packing materials. reflect a lack of investigation, and also the presence of In tropical Asia and the Pacific. and other tropical many small and isolated islands with little or no taro. reg ions, taro cultivars are very commonly propagated On such islands, TarophaJ!,liS might have been present using corms (or corm tops) w ith petio les ( leaf stalks) never or intermittently. attached and the blades removed. In contrast, most tem­ perate cultivars are propagated using small side-corms that lie dormant and leafless during winter. Removing Origi11s of raro in Ocea11ia the blades reduces water stress for the plant. and would not prevent the transport of taro planthoppers. As a quar­ From west to east. there may have been two general antine measure to prevent the transport of planthoppers routes for the early movement of taro and taro planthop­ w ith planting materials. Zettler et a/. ( 1989) recom­ pe rs: (i) a northern rou te taking T. colocasiae into mended removing the sheathing petiole bases until they M elanesia. Micrones ia, and possibly Polynesia (from are free of signs of feeding or egg-laying punctures. Sunda to Sahul and to ea r Ocea nia and Remote Anon ( 1999b) stated (in relation to T. persephone) that Oceania}, and ( ii) a southern route carrying T. proserpina most adult planthoppers produced ca nnot fly. so that into Remote Oceania. T he distribution ofT. proserpina is using clean planting stock is a 'useful weapon· in pre­ especially intriguing because it is apparently absent in venting the spread of planthoppers to new plantings. East Asia, Southeast Asia, Micrones ia. and Australia. Then follows a lament that 'all too often in traditional Since al l three species are sympatric in the mainland of plantings it is not possible to have this recommendation Papua New Guinea (Fig. 5). T. proserpina apparenrly can implemented·. The recommendations cited here suggest coexist with other Tarophagus species. though not neces­ that very deliberate and unusual actions would have been sarily on the same plants or in the same habitats. Asche needed to prevent the spread or taro planthoppers with and Wi I son ( 1989a:294) found no morphological evi­ taro, during canoe voyages in the past. Planting materials dence for hybrid belts or clines. The western limits ofT. like those common today undoubted ly permitted inciden­ proserpina might reflect natural and agricultural tal and unintentional transport of planthoppers. processes: limited flying ability, limited ability to coexist Elsewhere ( M atthews 2002) I have argued that to with other planthopper species on plants or in habitats. store taro during voyages. for later planting.- special care and limited transfer of planting materials westward or might have been taken to maximise the number of nonhward. As explained later, taro planthoppers are eas­ healthy buds and growing shoots. Special care may have ily carried on planting materials. inc luded (i) prevention of skin damage during harvest In Polynesia, taro is believed to be an entirely exotic and transport. by not breaking mother corms and side­ plant , introduced by humans. The eastward movement of shoots apart. and by wrapping them in clean materi als. T. proserpina. into Po lynes ia. may be large ly due to and (ii) use of wrapping materials and on-board locat ions

199 that provided the best possible conditions of temperature. useful. since planthopper migration is likely to vary moisture. and aeration. Although there is no reason to according to habitat persistence (cf. Denno eta/. 1991). ass ume that special care was always necessary or Within Australia (Fig. 5), the reports made so far indi­ attempted. there may have been attempts to maximise the cate that T. perseplwne and T. co/ocasiae are not uni­ survival of plants during long voyages of unknown dura­ formly distributed across northern Australia. This is t ion. Wrapping ordinary taro planting materials with intriguing because different patterns of ribosomal DNA leaves or other materials, in bags or baskets. would have were previously found in wildtype taros from Western helped reduce exposure and water loss. An incidental Australia. Northern Territory. and Queensland (M atthews result of wrapping may have been to improve the sur­ and Terauchi 1994). If different planthoppers are closely vival of planthopper eggs, juveniles. and adults. If whole associated with different lineages of wildtype taro, then taro plants were transported with corms, roots. and dispersal o f the insects may have been c lose ly linked leaves intact, then water loss from the leaves would have with dispersal of the plants. been a problem. To prevent dehydration and root dam­ age, the whole plant could have been wrapped in some way, with the corms an d roots kept in damp soil or Conclusions humus. Such transport might have maximised chances for the transf er o f juvenile and adult forms of For historica l purposes. the present observations o f Tarophagus. on lea f blades and petioles, as well as max­ Taroplwgus in Queensland may be most significant imising chances for the transfer o f other organisms because (i) they demonstrate a close association of the closely associated with taro, e.g. land snails (Kirch eta/. insec t with wildtype taro, for the first time, and (ii) on ly J 992). T. colocasiae was found, thus making it more likely that T. proserpina is restricted to New Guinea. The present observations are also significant for agricultural ecology: Tarophagus species and taro gene pools the abundance and wide geographical range of Taroplwgus in Queensland suggests that taro bobone Previously. various lines of ev idence led me to propose virus. or similar insec t-transmitted viruses. could spread that indigenous se lection and dom estication of taro quickly among wild and/or cultivated taros. if the viruses occurred over a wide geographical range, and involved are introduced. M any areas of Queensland are very suit­ genotypically and phenotypically diverse natural w ild able for growing taro. and interest in the crop is increas­ fonns of taro (M atthews 1990). At that time. the genetic ing. It would be unfortunate if natural populations of data was still weak. Since then. isozyme variation has Taroplwgus and its host were to become the targets of been surveyed in many wild and cultivated forms of taro. agricultural pest control. across A sia and the Pacific. Lebot and his colleagues Taroplwgus is of special historical interest because it reported two distinct taro gene pools in Southeast Asia has a very close relationsh ip with taro, a crop that has and M elanesia, and proposed that these reflect natural been central to interpretations of archaeological evidence differentiation of the species on each side of the Wallace for early agriculture in New Guinea (Golson and Hughes Line (Lebot and Aradhya 1991 , Lebot 1999, Lebot eta/. 1980. Bayliss-Smith and Golson 1992). The present evi­ 2002). T hese authors also reported very little genetic dence and interpretations also highlight New Guinea. diversity among Polynesian taros, and suggested that and especially eastern New Guinea, as a possible locus Polynesian taros are derived from the larger M elanesian for the domes tication of taro, and as a likely origin for gene pool. Tahara et a/. ( 1999) found two distinct taro taro in much of Oceania. gene pools in and Yunnan (South China). In sum. these studies suggest the possibility that each species of Tarophagus evolved in association with a different nat­ Acknowledgements ural taro gene pool. before the domestication of taro in multiple locations or regions. The author thanks K. Thiele and E. Takei for field assis­ If the distribution ofT. colocasiae actually extends tance in 1987 and 1992, respectively. For entomological further west than shown in Fig. 5. then this species might advice and iden tifications I thank D. Rentz (CSIRO. have originated in association with taro in the A sian Canberra), Dorothy E. Shaw and J. F. Donaldson mainland. T. persephone might have originated in island ( Quee nsland D epartment of Primary Industries. Southeast Asia or M elanes i a, and T. proserpina in l ndooroopily). and M . J. Fletcher (NSW Agricultural M elanesia. These suggested origins and the present over­ Scientific Collections Unit, Orange). G. Roderick lapping distributions of Taropha{:us species must some­ (University of California, Berkeley) kindly alerted me to how reflect past dispersals of wild and cu ltivated taros, current research on Tarophagus in the Pacific. For practi­ but little more can be said until we know more about cal help in the field I also thank R. Ru ssel (park ranger, planthopper populations and migration in wild and culti­ M ossman), S. M ostachetti (park ranger. Ingham), Tony vated habitats. Matthews ( 1997) reported that wildtype I r vine (Tropic al Fores t Research Centre, CS IRO. taros exi st in both stable and un stable h abitats. A therton). N. Horsfall (archaeologist. National Parks and Investigating the planthoppers in these habitats could be Wildlife Serv ice. Townsville), and R. Hinxman (vic.

200 Cairns). For guidance at the Research School of Pacific Gol on. J. and P. J. Hughes ( 1980). The appearance of plant ~ and As ian Studies, and the Research School o f and animal domestication in ew G uinea. Journal de Ia Biological Sciences, Australian National University, I Societe des Oceanistes 36, 294-303. thank D. E. Yen, Jack Golson, D. Coates, M. Arnold, and Greber, R. S. and D. E. Shaw ( 1986) Dasheen Mosaic Virus in Queensland. Australasian Plant Pathology 15. 29-33. D. Shaw. The Queensland National Parks and Wildlife Hay, A. ( 1996) A new Bornean species of Colocasia Schott Service gave permission to collect insects (Permit No. (1\ raceae-Colocasieae). with a synopsis of the genus in 1562). The Research School of Pacific and Asian Studies Malesia and Australia. Sandalwnia 7. 31-48. (ANU) provided a PhD Scholarship ( 1985- 1988) and a Hay. A. and R. Wise ( 1991) The genus Alocasia () in Visiting Fellowship (1992- 1993). For comments on Australasia. Blumea 35. 499-545. drafts of this paper, I thank J. F. Donaldson (address Horsfall. N. ( 1996). Holocene Occupation of the Tropical above), A. Hay (Royal Botanic Gardens, Sydney), M. N. Rainforests of onh Queensland. In P. Veth and P. Hiscock Pearson (Uni versity of Auckland), and the editors of this (eds.). Archaeology of Northern Australia: Regional volume. K. Fujita ( ational Museum of Ethnology, Perspectil·es. pp.l 75- 190. Universit y of Queensland. St Osaka) assisted with the maps. Lucia, Queensland. Ichikawa, T. (1994) Wing dimorphism and aggressive commu­ nication through vibration signals in the males of the plan­ t hopper, Nilaparvata lugens (Still ). In Proceedings of R eferences Memorial and lntemational Symposium on Dispersal of In sects. its Adoptation and E1·olution. Anon ( 1999a) Tarophagus proserpina. Ecoport Entity Okayama. Japan. June 30- July I. 199./. pp.87-97. ID :20 168 Ecopon (hllp://wwwx.ecopon.org). Ivancic. A.. A. Simin. E. Ososo. and T. Okpul ( 1995) Wild taro Anon ( 1999b) Taroplwgus persephone. Ecoport Entity (Colocasia esculenta (L.) Scholl ) populations in Papua ew ID:20167. Ecopon (http://wwwx.ecopon.org). Guinea. Papua New Guinea J o urnal of Agriculture. Asche, M. and M. R. Wilson ( 1989a) The three taro planthop­ Forestry and Fisheries 38. 3 1-45. pers: species recognition in Tarophagus ( Hemipte ra: Jackson. G. V. H. (1994). Taro and Genetic Resources in De lphac idae). Bulletin of Entomo logical Research 79. thr Pacific and Asia: A Report Prepared for the Australian 285- 298. Centre for International Agricultural Research ond the Asche. M. and M. R. Wilson ( 1989b) Addendum. Bulletin of lntemational Plalll Genetic Resources lnstitwe. Anutech. Entomological Research 79. Canberra. Bayliss-Smith. T. and J. Golson ( 1992). A Colocasian revolu­ Jones. D. R .. D. E. Shaw. and D. H. Gowan1ock ( 1980). tion in the New Guinea Hi ghlands? Insights from Phase 4 at Baci 11 i form virus particles detected in Cyrtosperma sp. Kuk. Archaeology in Oceania 27. 1- 21. imported from Solomon Is lands. Australian Plant Brunt. A. A .. K . Crabtree. M. J. Dallwitz. A. J. Gibbs. L. Patholo~y 9. 5-6. Wat~on and E. J. Zurcher ( 1996). Plant Viruses Online: Kirch. P. V.. J. R. Flenley. D. W. Steadman. and F. Lamont Descr iptions a nd Lists from the VIDE Database. ( 1992). Ancient e nvironmental degradation. National Version:20th August 1996. Geo~:raphic Research and Exploration 8.166-179. (http:/fbi o logy.an u .edu.au /G rou ps/M ES/v ide/) Ki simoto, R. ( 1994) Role of the wi ng-polymorphism in the Carson, H. L. and T. Okada ( 1980) Drosophilidae associated evolutio n of th e life his to ry in De1phac idae. with flowers in Papua New Guinea: I. Colocasia esculenta. Au c he norrhync ha. In Pro cee din ~:s of Mem orial and Kontyu. Tokyo 48. 15-29. International Symposium on Dispersal Polymorphism of Carson. H. L. and T. Okada ( 1982). Ecology and evolution of Insects. its Adaptation and £\'OIIIIion. Okoyama. Japan. flower-breeding pomace flies of New Guinea ( Diptera: June 30-July I. 1994. pp.65- 73. Drosophi lidac). Emomologia Generalis &. 13- 16. Lebot. V. ( 1999) Biomolecular evidence for plant domestica­ Denno. R. F.. G. K. Roderick. K. L. Olmstead, and H. G. ti on in Sahul. Genetic Resouras and Crop £1·olurion 46. Dobel ( 1991 ). Density-related migration in planthoppers 6 19-628. (Homoptera: Delphacidae): The role of habitat persistence. Lebot. V. and K. M. Aradhya ( 1991 ). Isozyme variation in taro The Amerimn Naturalist 138. I 5 13- 1541 . (Coloco.fia esculema (L.) Schott ) from Asia and Oceania. Fle tc her. M. J. and M. -C. La riviere (2001 and updates) Euphytica 56. 55-66. ldentilicati on Keys and Checklists for the leafl10ppers. plan­ Lebot. V.. S. Hartati. N. T. lluc, N. V. Vi et. . H. Nghia. T. th oppers and their relatives occurring in Australia and New Okpul. J. Pardales. M. S. Prana. T. K. Prana, M. Thongjicm. Zealand (Hemiptera: Auchenorrh yncha). C. M. Kri eke. H. Van Eck. T. C. Yap, and A. Ivancic (2002) ( h t tp://w ww.agric .ns w.gov. au/Hon/ascu/st art. htm) Genetic variation of taro (Colocasia csculenta) in South East Gagne. W. C. ( 1982). Staple crops in subsistence agriculture: Asia and Oceania. In M. Nakatani and K. Komaki (eds) their major insect pests. with emphasis on biogeographical Potential of root crops for food ond industrial resources. and ecological aspects. In J. L. Gress ill (ed.). Biogeography Tu·elfth Symposium of the lnternmional Society for Tropical and Ecology of New Guinea. Vol. I. pp.229-259. Dr W. Root Crops (ISTRC). Sept. 10- 16. 2000. Tsukuba. Japan). Junk. The Hague. pp.524-533. ISTRC2000 Organizing Commillee. Tuskuba. Gi ll espie. R. G ..- G. K. Roderick. M.A. Amedo, L. Shapiro. A. Matthews. P. J. ( 1990) The Origi ns, Di spersal a nd Bely. J. E. Garb , and J . Spagna (2000) Spiders. Domestication of Taro. Canberra, Australian National Planthoppers and Flies: Report on Research Conducted in Uni versit y. Unpublished Phd thesis. French Po lynesia. June-A ugust 2000. Unive rs it y of Matthe ws. P. J. ( 199 1) A possib le tropical wildtype taro: Ca Ii fornia. Berk eley. (http://www.c nr. berkeley.edu/evolab/ Colocasia esculema var. aquatilis. Indo-Pacific Prehi.\" IOI)' FP2000/DelphacidSociety. htm) Association Bulletin II . 69- 81.

201 Matthews. P. J. ( 1992) Wild taro (Colocasia esculenta va r. Scheffer, R. P. ( 1997). The Nature of Disease in Plants. aquatilis) and its insect associate Ta rophagus in Cambridge, Cambridge Uni versity Press. Queensland. 2nd- I Oth August 1992. Canberra. Australian Smithsonian. (200 I) A Checklist of the Insects of Suhsalwrcm National Uni versity. Unpublis hed fi e ld report for the Africa - 1/omoptera - Delphacidae. Department o f Queensland National Parks and Wildlife Service. Entomology. Smithsoni an National Museum of Natural Matthews. P. J. and R. Terauchi ( 1994 ). The genetics of agri­ History. (ht tp://www.entomology.si.edu:591/entomology/ culture: DNA vari ation in taro and yam. In J. G. Hather Subsahara/homop/delph.html) (ed.) Tropical Archaeohotany: Applications and new del•el­ Spriggs, M. (1982) Taro cropping systems in the Southeast opments. pp.25 1- 270. Routl edge, London and New York. Asian-Pacific region: Archaeological ev id ence. Matthews, P. J. ( 1995) Aroids and the Austronesians. Tropics Archaeology in Oceania 17, 7- 15. 4. 105- 126. Tahara. M.. V. X. Nguyen, and H. Yoshino (1999) Isozyme Matthews. P. J . ( 1997) Fie ld g uide for wild-type ta ro. analyses of Asian diploid and tri ploid taro. Colocasia escu­ Colocasia esculenta (L.) Schott. Plant Genetic Resources lema (L.) Schott. Aroideana 22, 72- 78. News/euer o. l l 0:41-48. Vartavan, C. d. ( 1990) Contaminated plant-foods from the Matthews, P. J. (2002). Taro Storage Systems. In S. Yoshida and Tomb of Tut ankhamun : A new interpretative system. P. J. Maffhews (eels) Vegeculture in East em Asia and Oceania, Journal ofArclweological Science 17, 473-494. pp.J 35- 163. The Japan Center for Area Studies, Osaka. Waterhouse, D. F. and K. R. Norris ( 1987). Biological Control, Mitchell. W. C. and P. A. Maddison ( 1983). Pests of Taro. In Pacific Prospects. Melbourne, lnkata Press. J.-K. Wang (ed.) Taro: A Review of Colocasia esculema and Zeuler, F. W. and R. D. Hartman ( 1986) Dasheen Mosaic Virus its Potemials. pp.l80-235. University of Hawaii Press. and its Control in Cultivated Aroids. Taipei. Food and Honolulu. Fertilizer Technology Center. Pearson. M. N .. G. V. H. J ackson. J. Saelea. and S. G. Zettler, F. W.. G. V. H. Jackson. and E. A. Frison (eds.) ( 1989) Morar( 1999). Evidence fo r two rh abdoviruses in taro FAOIIBPGR Technical Guidelines for the Safe Mo1•emem of (Colocasia esculenta) in the Pacific region. Australasicm Edible Aroid Germplasm. Rome, FAO and Internati onal Plant Pathology 28. 248-253. Plant Geneti c Resources Institute.

202