ASIAN MYRMECOLOGY Volume 9, e009004, 2017 ISSN 1985-1944 | eISSN: 2462-2362 © James K Wetterer DOI: 10.20362/am.009004

Geographic distribution of the weaver Oecophylla smaragdina

James K Wetterer

Wilkes Honors College, Florida Atlantic University, 5353 Parkside Drive, Jupiter, Florida 33458 USA

Corresponding author's email: [email protected]

ABSTRACT. Weaver (Oecophylla spp.) are conspicuous arboreal ants, which build distinctive nests in . In many regions, people use weaver ants for food, medicine, and/or as biological control agents. There are two recog- nized species of weaver ants: (Latreille, 1802) in Africa and Oecophylla smaragdina (Fabricius, 1775) in Asia, Australia, and the West- ern Pacific. Here, I mapped the geographic distribution ofO. smaragdina based on >2700 site records from 21 countries: Australia, Bangladesh, Bhutan, Bru- nei, Burma (Myanmar), Cambodia, China, India, Indonesia, Laos, Malaysia, Nepal, Palau, Papua New Guinea, Philippines, Singapore, Solomon Islands, Sri Lanka, Thailand, Timor Leste, and Vietnam. The vast majority of O. smarag- dina records come from areas with Tropical (Group A) climates according to the Köppen-Geiger system: rainforest (Af), monsoon (Am), and savanna (Aw). However, >250 records come from areas classified on the map as having a Sub- tropical (Group C) climates, mostly in the Himalayan foothills of India and Ne- pal, southern China, northern Vietnam, and the southern coast of Queensland, Australia. Almost all these sites are classified as dry winter subtropical climate (Cwa). A few O. smaragdina sites are classified as having Arid (Group B) cli- mates, all from warm semi-arid (BSh) areas. This range map based on site re- cords corrects inaccuracies in earlier published range maps.

Keywords: biocontrol, biogeography, entomophagy, geographic range

INTRODUCTION People use weaver ants in many regions for food, medicine, and/or as bio- Weaver ants (Oecophylla spp.) are conspicu- logical control agents. Although Oecophylla ous arboreal ants, well known in the humid species show much geographic variation in tropics and subtropics of Africa, Asia, Aus- color (including shades of green, yellow, tralia, and the Western Pacific. Weaver ants orange, red, and brown), their appearance, build large distinctive nest structures in trees behavior, and nest construction are so dis- by binding together bunches of leaves us- tinctive that the genus can be easily iden- ing a silk-like substance secreted by the lar- tified through written accounts and from vae. Groups of Oecophylla workers hold the photographs. Currently, there are two rec- leaves together while other workers move the ognized species of weaver ants: Oecophylla silk-producing larvae back and forth across longinoda (Latreille, 1802) in Africa, and the gap, effectively weaving the leaves to- Oecophylla smaragdina (Fabricius, 1775) in gether. The name Oecophylla derives from Asia, Australia, and the Western Pacific. Nu- the Greek: oikos (house) and phyllo (leaf). merous papers on weaver ants (Cole & Jones 2 James K Wetterer

1948, Lokkers 1986, Azuma et al. 2002, out in great numbers, and punished the of- 2006, Dlussky et al. 2008, Crozier et al. fender by a much sharper bite than ever we 2010) include basic range maps. In an earli- had felt from the same kind of be- er paper, Wetterer (submitted) examined the fore.” Later, in August 1770, Cook (1773: distribution of O. longinoda in Africa. Here, 627) observed in Australia: “Of the ant there I document in detail the known geographic are several sorts some are as green as a leaf, distribution of O. smaragdina. The epithet and live upon trees, where they build their smaragdina derives from the Latin smarag- nests of various sizes, between that of a dinus (emerald), referring to the green color- man’s head and his fist. These nests are of ation of O. smaragdina queens. a very curious structure: they are formed by The earliest known report on O. sma- bending down several of the leaves, each ragdina can be recognized in the book “Nan of which is as broad as a man’s hand, and Fang Cao Mu Zhuang” (Plants and Trees of gluing the points of them together, so as to the Southern Regions), reputedly written by form a purse; the viscus used for this pur- Ji Han in 304 AD, who described the use of pose, is an animal juice, which Nature has weaver ants in biological control: “In the enabled them to elaborate. Their method of market, the natives of Jiao-zhi [now part of first bending down the leaves, we had not northern Vietnam and southern China] sell an opportunity to observe; but we saw thou- ants stored in bags of rush mats. The nests sands uniting all their strength to hold them are like thin silk. The bags are all attached to in this position, while other busy multitudes twigs and leaves, which, with the ants inside were employed within, in applying the glu- the nests, are for sale. The ants are reddish- ten that was to prevent their returning back. yellow in color, bigger than ordinary ants. In To satisfy ourselves that the leaves were the south, if the Gan [Mandarin orange] trees bent, and held down by the effort of these do not have this kind of ant, the fruits will diminutive artificers, we disturbed them in all be damaged by many harmful and their work, and as soon as they were driven not a single fruit will be perfect” (Li 1979 in from their station, the leaves on which they Huang & Yang 1987). were employed sprung up with a force much Even before Fabricius (1775) de- greater than we could have thought them scribed Formica smaragdina (= O. sma- able to conquer by any combination of their ragdina) based on specimens from India, strength. But though we gratified our curios- James Cook (1773) gave excellent accounts ity at their expense, the injury did not go un- of the species in Australia. Upon making revenged; for thousands immediately threw landfall on 23 May 1770 at a site now called themselves upon us, and gave us intolerable Seventeen Seventy (formerly Round Hill; pain with their stings, especially those which 24.17°S, 151.88°E) in Queensland, Aus- took possession of our necks and our hair, tralia, Cook (1773) wrote: “We found sev- from whence they were not easily driven: eral bogs, and swamps of salt water, upon the sting was scarcely less painful than that which, and by the sides of the lagoon, grows of a bee; but, except it was repeated, the pain the true mangrove, such as is found in the did not last more than a minute.” West Indies, and the first of the kind that we Weaver ants are so conspicuous had met with. In the branches of these man- and recognizable that they have local com- groves there were many nests of a remark- mon names throughout their range. Common able kind of ant, that was as green as grass: names used for both species of Oecophylla when the branches were disturbed they came include tailor ant, red ant, and red ant. Geographic distribution of the Oecophylla smaragdina 3

English common names for O. smaragdina single worker (as there is no range or error include Asian weaver ant, green ant (Austra- in the measurements), and does not actually lia, Papua New Guinea), and muli ant (Papua compare these measurements to those of O. New Guinea). Common names in other lan- smaragdina workers, which are extremely guages for O. smaragdina include fourmi polymorphic and show strong allometry. tisserande (French), kerengga (Malay), dimi- Though formal synonymisation is outside ya (Sinhala), semut rangrang (Indonesian), the scope of this study, I treat O. kolhapuren- hantik (Tagalog), wong mamah (Thailand), sis as conspecific withO. smaragdina. mottdaang (Lao), angkrang or ongkrong I obtained unpublished O. smarag- (Cambodian), kiến vàng (Vietnamese), Khar- dina site records from museum specimens in chin-aon or ‘kâ-gyin (Burmese), kilau lagah the collections of the United States National (Borneo), karrakum (Sepik), and kurakum Museum of Natural History and through per- (Tok Pisin). There are a vast number of local sonal communications from M. Janda (Aus- names in India, e.g., in Arunachal Pradesh, tralia, Bangladesh, Indonesia, Laos, Malay- the ant is known as tonge or babuk (Chakra- sia, Papua New Guinea, Philippines, and vorty et al. 2016). Thailand), D. General (Philippines), R. Re- custodio (Philippines), T. Colvin (Thailand), T. Chan (Laos), S. De Greef and numerous METHODS members of Natural Cambodia (Cambodia), J. Fellowes (China), and H. Bharti (India). I treat all Oecophylla records from outside In addition, I used on-line databases with Africa as O. smaragdina. Oecophylla sma- collection information on specimens by the ragdina has a number of subspecies still Field Museum, Antweb (antweb.org), and considered valid: Oecophylla smaragdina the India Biodiversity Portal (indiabiodiver- fuscoides Karavaiev, 1933 (described from sity.org). Because the genus Oecophylla is so Java, Indonesia), Oecophylla smaragdina easy to recognize, I was able to obtain many gracilior Forel, 1911 (described from Bacan, site records based on photos published on- Indonesia), Oecophylla smaragdina gracil- line at a variety of sites, including iNatural- lima Emery, 1893 (described from Bacan, ist (www.inaturalist.org), Ispotnature (www. Indonesia), Oecophylla smaragdina seleben- ispotnature.org), Project Noah (www.pro- sis Emery, 1893 (described from Sulawesi, jectnoah.org), National Geographic (your- Indonesia), and Oecophylla smaragdina sub- shot.nationalgeographic.com), Flickr (www. nitida Emery, 1892 (described from Morotai, flickr.com), and Instagram (instagram.com). Indonesia and New Guinea). These names, The first three of these sites provided locale however, have not been applied in any con- names and geo-coordinates for all photo- sistent manner, and the status of these taxa graphs, whereas the other sites only provided is unclear. In addition, Kurane et al. (2015) locale names and coordinates for some of the recently described a new Oecophylla spe- photographs. For photographs without locale cies from India, Oecophylla kolhapurensis names or geo-coordinates, I contacted the Kurane, Bhoje, & Satheal, 2015. The de- person who posted the photograph to ask for scription, however, is insufficient to distin- this information. guish O. kolhapurensis from O. smaragdina. Geographic coordinates for sites The description uses imprecise measure- came from published references, from speci- ments (e.g., “Head 2 mm long, 1 mm broad; men labels, geotagged photos, or I looked thorax 3 mm long”) apparently taken from a up the coordinates. For older references and 4 James K Wetterer specimens, some site names were no lon- The single record I found of O. sma- ger in use or are now spelled differently; I ragdina from Palau (Table 1; identified by searched, not always successfully, to deter- S. Shattuck), collected on the small island mine current names. of Peleliu (13 km2) in 1936, may represent Some occurrence records I found for a temporary introduction. Ant researcher O. smaragdina included no specific “point Jesse Czekanski-Moir (pers. comm.) has locales.” The most significant of these are re- spent three years working in Palau, including cords from two provinces in China: Jiangsu seven days collecting on Peleliu, and he has (Hua 2006 in Guénard & Dunn 2012) and Si- never seen Oecophylla in the island group. chuan (Mou 1959), which I mapped to their Czekanski-Moir (pers. comm.) wrote: “Pa- capitals, Nanjing and Chengdu, respectively. lauan people actually have words for several To classify climates, I followed the different types of ants, including Anoplolepis Köppen-Geiger system which considers gracilipes, Solenopsis geminata, and Odon- only precipitation and temperature variation tomachus simillimus. If Oecophylla were through the year (Peel et al. 2007). Of the known to Palauan people, I suspect I would 30 possible climate categories, the ten that have heard about it.” J. Czekanski-Moir occur in and adjacent to areas occupied by (pers. comm.) added “During the last glacial O. smaragdina are as follows: Tropical cli- maximum, Peleliu and most of the ~500 is- mates (Group A; shades of blue in Figs. 1 lands to the North were part of a single is- & 2) = rainforest (Af), monsoon (Am), and land. Thus, if O. smaragdina were native to tropical savanna (Aw); Arid climates (Group Palau, one would expect it to be distributed B; oranges and red in Figs. 1 & 2) = warm more widely in the archipelago.” Much of semi-arid (BSh) and warm desert (BWh); Peleliu’s vegetation was burned during the Temperate/Subtropical climates (Group C; 2.5-month long Battle of Peleliu in 1944 be- greens and yellows in Figs. 1 & 2) = humid tween American and Japanese forces during subtropical (Cfa), dry winter subtropical World War II. If any Oecophylla were living (Cwa), highland subtropical (Cwb), warm on the island, they may have been killed off Mediterranean (Csa), and temperate Medi- at this time. terranean (Csb). For more details on the The single record I found of O. Köppen-Geiger classification system, see smaragdina from Bhutan (Table 1) is Peel et al. (2007). based on a photograph posted by Vanessa The climate of each record site I cat- McLaughlin (www.flickr.com/photos/ga- egorized according to the designations found tos_rojos/8607316210; V. McLaughlin, on climate-data.org. If this site did not list a pers. comm.) taken at Kharbandi Gompa, site by name, I matched each record to the Phuentsholing, Bhutan (26.85°N, 89.39°E; closest listed site with a similar elevation, as 370m asl), <200m from the border with West recorded on google earth. Bengal, India. Research on the ants of Bhu- tan is extremely limited, so it is possible that O. smaragdina is actually widespread in the RESULTS lowlands of southern Bhutan as well in val- leys further north. The entire southern border I mapped the geographic distribution of O. of Bhutan is below 500m asl. smaragdina based on >2700 site records Overall, records of O. smaragdi- (Fig. 1) from 21 countries, including the first na spanned 66.8° of latitude, from Fugou, published records from Bhutan and Palau Henan Province, China (34.1°N; Wang 2008) (Table 1). and Surinsar, Jammu & Kashmir State, India Geographic distribution of the weaver ant Oecophylla smaragdina 5

Table 1. Earliest known records for Oecophylla smaragdina. +no previously published records.

Earliest record Australia 1770 (Cook 1773) Bangladesh ≤1925 (Mukherji & Ribeira 1925) +Bhutan 2013 (V. McLaughlin, Flickr): Kharbandi Gompa Brunei 1983 (P. J. DeVries, Antweb: LACMENT142773): Maura Coast Burma (Myanmar) 1885-1887 (Emery 1889) Cambodia 1914 (Friederichs 1920) China 304 (Huang & Yang 1987) India ≤1775 (Fabricius 1775) Indonesia 1854-1856 (Smith 1857) Laos ≤2005 (Yhoung-Aree & Viwatpanich 2005) Malaysia 1854-1856 (Smith 1857) Nepal 1958 (Kapur 1961) +Palau 1936 (Z. Ono, Antweb: ANIC32-023162): Peleliu Papua New Guinea 1896-1897 (Forel 1901) Philippines 1854-1856 (Smith 1857) Singapore ≤1890 (Ridley 1890) Solomon Islands 1909 (Forel 1910) Sri Lanka ≤1858 (Smith 1858) Thailand ≤1903 (Bingham 1903) Timor Leste 2003 (Trainor & Andersen 2010) Vietnam ≤1879 (Forel 1879)

(32.7°N; H. Bharti; pers. comm.) in the north, having Tropical (Group A) climates: rainfor- to Seventeen Seventy, Queensland, Austra- est (Af), monsoon (Am), and savanna (Aw) lia (24.2°S; Cook 1773) in the south, and (shades of blue in Fig. 2). spanned 98.3° of longitude, from Junagadh, Fewer than 50 O. smaragdina sites Gujarat State, India (70.5°E; Tak & Rathore are classified as having Arid (Group B) cli- 2004) in the west, to Tikopia, Solomon Is- mates. All are from warm semi-arid (BSh) lands (168.8°E; Greenslade 1965) in the areas, and all are from India or Australia, east. Wang (2008) reported O. smaragdina except one, from Mandalay, Burma (Myan- from numerous sites in Zhoukou Prefecture, mar). Many of the Indian O. smaragdina Henan, including farmland, residential areas, records are from urban areas (e.g., Patiala, woodlands, riversides, and wastelands. Dehli, Agra), whereas the Australian records Comparing the documented dis- are almost all from forested river gorges. tribution of O. smaragdina (Fig. 1) with a In addition, >250 records occur in Köppen-Geiger climate map (Fig. 2) indi- areas classified on the map as having a Sub- cates that the vast majority of O. smaragdina tropical (Group C) climate. Almost all of records (>2200) come from sites classified as these sites were classified as dry winter sub- 6 James K Wetterer tropical (Cwa). In southeast Australia, the actually from a different Geraldton, one in southernmost sites (23.1-24.2°S) were clas- Queensland, Australia (16.6°S, 145.3°E). sified as humid subtropical (Cfa), as were While working in Zanzibar, Tanza- most of the sites in China north of 23.0°N. nia, Vanderplank (1959) wrote that he im- Six high elevation sites (>1200m asl) in ported Oecophylla queens from South Amer- southern India were classified highland sub- ica. There are no records of Oecophylla in tropical (Cwb), as were three sites in Dar- the New World, so Vanderplank (1959) must jeeling District, India. have been mistaken about the origin of his On-line photographs from Flickr and alleged South American queens. Instagram alone accounted for almost 1000 of the mapped O. smaragdina site records. Photographs of O. smaragdina on Flickr and PREVIOUS DISTRIBUTIONAL Instagram were particularly common from SUMMARIES areas with higher density of people with smartphones, such as southwestern India, Bingham (1903) wrote that the geographic Java, Bali, peninsular Malaysia, and Austra- range of O. smaragdina included “the whole lia. Unfortunately, both Flickr and Instagram of India, Burma, and Ceylon within our lim- are currently banned in China. Oecophylla is its, except the desert and treeless tracts. The particularly amenable to such crowd-sourced range of this species extends through the Ma- information because it is both photogenic layan subregion to Australia and New Guin- and simple to identify reliably from photo- ea.” This account seems accurate, though it graphs, though I believe the same methods omits several major areas, i.e., Nepal, China, could be used for many other species. and the Solomon Islands. Keuchenius (1914) wrote that O. smaragdina was widespread in tropical Asia and Polynesia, but there are DOUBTFUL RECORDS actually no records known from Polynesia. Vanderplank (1959) reported that Oecophyl- I excluded two doubtful O. smaragdina re- la “is distributed throughout the tropics of cords. Santschi (1919) reported Oecophylla Asia, Africa, America, Australia and is found smaragdina from Apia, ‘Upolu, Samoa. on most Pacific Islands.” Oecophylla is not Wilson & Taylor (1967) listed this record as known from the Americas nor on most Pa- doubtful. Nonetheless, Kami & Miller (1998) cific Islands and is also absent in most arid included this record in their list of ants from tropical areas. Samoa. Given the conspicuousness of this Several Oecophylla studies have in- ant, it is unlikely that the extensive surveys cluded distribution maps for O. smaragdina in Samoa (e.g., see Wetterer & Vargo 2003) (Cole & Jones 1948, Lokkers 1986, Azuma et have overlooked any populations of O. sma- al. 2002, 2006, Dlussky et al. 2008, Crozier ragdina in Samoa. I also excluded a record et al. 2010). All the maps, however, include listed on-line as from Geraldton in Western numerous errors. Cole & Jones (1948) incor- Australia, Australia (28.8°S, 114.6°E; J. rectly included Vanuatu within the range of Clark; Museum Victoria; http://biocache.ala. O. smaragdina, and indicated its presence org.au/occurrences/c5e8b619-59a6-415e- in Australia by a single point in the middle b042-a18a8e7b393d) is a geographic iso- of the continent. Lokkers (1986) incorrectly lated outlier located 4.6° further south than included New Caledonia, Fiji, and Vanuatu any other record. I expect that this record is in the range of O. smaragdina and omit- Geographic distribution of the weaver ant Oecophylla smaragdina 7

Fig. 1. Site records for Oecophylla smaragdina (mapped using carto.com). Records from sites with Tropical cli- mates (Group A) = blue; with Semi-arid climates (Group B) = orange; with Subtropical climate (Group C) = green. ted its range in north India, Nepal, Bhutan, having Tropical climates (Group A) and dry and China. The maps of Azuma et al. (2002, winter subtropical (Cwa) climates (Figs. 1-2). 2006), Dlussky et al. (2008), and Crozier et In the west (in India and Pakistan) al. (2010) correctly omitted New Caledo- and much of the south (in Australia), the nia, Fiji, and Vanuatu from the distribution range of O. smaragdina appears to be largely of O. smaragdina, but incorrectly omitted bounded by lands with Arid (Group B) cli- the range of O. smaragdina in north India, mates. The range of O. smaragdina appears Nepal, Bhutan, and China. Lokkers (1986) to extend into these more arid regions only examined the climatic factors that appear to in riparian forests, irrigated lands, and urban delimit the range of O. smaragdina in Aus- areas. In western India, I found no records of tralia, predicting a range that closely matches Oecophylla from the Thar Desert region that the actual known distribution. covers most of Rajasthan and much of Gu- jarat, in agreement with Bingham’s (1903) evaluation that it is absent from “desert and DISCUSSION treeless tracts.” In much of the north, across Hi- Oecophylla smaragdina site records are malayan India, Nepal, Bhutan, and Burma broadly spread across regions of Asia, Aus- (Myanmar), the range of O. smaragdina ap- tralia, New Guinea, and the Solomon Islands pears to be largely bounded by areas classi- classified by the Köppen-Geiger system as fied as highland subtropical (Cwb). Although 8 James K Wetterer

I found no records of Oecophylla from Paki- nests are purchased in Ts’ingun during the stan, it seems likely that it occurs there in Chinese first and second months of the year areas which have the same dry winter sub- (usually February and March), shortly before tropical climate (Cwa) as neighboring areas the break of spring. The ants have not yet India where O. smaragdina is known (Figs. emerged from the nests, and after the latter 1 & 2). High latitude would seem to limit have been placed in the crutches of the cit- O. smaragdina populations in southeastern rus trees, some candle wax is smeared about Australia at around 24.2°S. Similarly, in the branches to prevent them from running China, I found numerous O. smaragdina site away when they do venture out. At this time records from the four southernmost prov- of year it is very easy to transport the nests, inces of Hainan, Guangdong, Guangxi, and whereas if they wait until the third or fourth Yunnan, as well as Hong Kong and Macau, month, after the weather is warmer, the ants but I found only four isolated areas of O. will escape in transit.” smaragdina in China north of 24.7°N (Fig. In contrast to the extensive subtropi- 1). Ying (1966) reported that O. smaragdina cal populations documented for O. smarag- from Guangdong Province, was twice im- dina (green in Fig. 1), there are very few sub- ported, in 1958 and 1963, into Huang Ling tropical populations of O. longinoda in Africa Miao village (30.8°N) in Hubei Province, (Wetterer submitted). The highest latitude re- China. Ying (1966) noted that O. smarag- cords of O. longinoda (28.4°S), come from dina could not survive outdoors in the winter coastal South Africa, less than one degree of in Hubei. It seems likely that O. smaragdina latitude south of areas classified as having also reached the three other areas in northern Tropical climates (Wetterer submitted). China through intentional introductions and Several studies have begun to ana- must be artificially maintained in the winter. lyze the genetic diversity of Oecophylla. Yet O. smaragdina has been found naturally Azuma et al. (2002, 2006) sequenced DNA occurring at many sites in the India and Ne- from numerous populations of O. smaragdi- pal at latitudes of 27-32°N. This indicates na, using two populations of O. longinoda as that O. smaragdina populations in northern outgroups, and found seven geographically India and Nepal have greater tolerance for distinct O. smaragdina clades (1 = southern cold compared with populations in eastern Indian, 2 = Indochina Peninsula, 3 = Philip- China and Australia. pines, 4 = Flores, Indonesia, 5 = Sulawesi, Collingwood (1970) reported Oeco- Indonesia, 6 = Halmahera, Indonesia, 7 = phylla in Nepal from trees on “the high Australia and New Guinea). These genetic mountainside of Tamba Kosi” (27.4°N; analyses did not include samples from popu- 1150-1450m asl). Collingwood (1970) noted lations in many parts of the known range of that “presumably there the species must un- O. smaragdina, e.g., Solomon Islands, Ne- dergo some form of hibernation for periods pal, northern India, and China. It is possible of at least some weeks in mid winter.” Winter that cold-tolerant Oecophylla from subtropi- inactivity has been reported in China. Groff cal climates in northern Asia are a separate & Howard (1924) wrote: “At Saisha the species, though none of the currently valid growers told us that their region is too cold subspecies of O. smaragdina are known from to winter these ants but that in the Ts’ingun these areas. Genetic analyses are needed. [Quinyuan, Guangdong Province] district a Some populations in China have regular business has developed in the rearing been introduced for biological control (e.g., of these beneficial insects for Szewui [= Si- see above). In addition, Young (2001) wrote hui, Guangdong Province] fruit growers. The that in Papua New Guinea, “unsuccessful Geographic distribution of the weaver ant Oecophylla smaragdina 9

Fig. 2. Climate classification in and adjacent to areas occupied by O. smaragdina based on the Köppen-Geiger system (modified from Peel et al. 2007 used under Creative Commons License). Tropical climates (Group A) = shades of blue; Arid climates (Group B) = oranges and red; Temperate/Subtropical climates (Group C) = greens and yellows. See Methods for more details. attempts were made to introduce the ant nests was unsuccessful on account of high into some islands of the Manus province seas.” This record from Tikopia is far to the where it does not occur naturally (P. Room east of all other records for O. smaragdina pers. comm.).” There are some O. smarag- (Fig 1). I found no further reports on the fate dina specimens from Manus (1963; J. Szent- of this population, so I contacted Thomas Ivany; antweb: ANIC32-043779), but it is Lien, who contacted the Chief Tafua on Tiko- unclear whether these are from a naturally pia and reported “they still have the weaver occurring population. Greenslade (1965) re- ants there, but it’s getting less. Not like in the ported on the intentional introduction of O. 70s and 80s. The problem with them is that smaragdina to the small island of Tikopia (5 they bite, when they want to climb the trees.” km2) in the eastern Solomon Islands to con- Researchers promote the use of trol the Hispine coconut beetle, Promecothe- weaver ants in orchards as an eco-friendly al- ca opacicollis Gestro, 1897: “consignments ternative to chemical , reporting of leaf nests of O. smaragdina were sent to an increase in productivity associated with the island from Guadalcanal in the Solomons this practice (e.g., Van Mele 2008). Using in August and November 1962. The ant was weaver ants to protect crops within the ants’ reported to be established by subsequent ob- native range has great potential benefits. servers, but it is not known what effect it had However, importing weaver ants into locales on Promecotheca populations. An attempt where these ants are not native, these ants by the writer in August 1963 to investigate could have a great negative effect on the na- this and to ship a further batch of Oecophylla tive fauna, particularly in areas where there 10 James K Wetterer are large swaths of land with suitable climate REFERENCES for the ants, such as many Pacific islands and much of the New World tropics. In fact, Azuma N, Kikuchi T, Ogata K and Higashi S, Duncan (1999) proposed introducing weaver 2002. Molecular phylogeny among lo- ants to El Salvador to help control pests on cal populations of weaver ant Oecophyl- plantations, writing that a consultant la smaragdina. Zoological Science 19: 1321 – 1328 proposed that an “avenue of investigation Azuma N, Ogata K, Kikuchi T and Higashi S, should be the introduction and trial of the 2006. Phylogeography of Asian weaver green ant, Oecophylla smaragdina which has ants, Oecophylla smaragdina. Ecologi- proved to be a highly efficient predator of cal Research 21: 126-136. cashew pests in Australia... The pro- Bingham CT, 1903. The fauna of British India, posed trial to give a preliminary evaluation including Ceylon and Burma. Hymenop- of the ant as a biological control agent under tera, Vol. II. Ants and Cuckoo-wasps. El Salvador conditions would take about 4 to Taylor and Francis, London, 506 pp. 5 months. The physical requirements to con- Chakravorty J, Ghosh S, Megu K, Jung C and duct the importation and trial of the ant are Meyer-Rochow VB, 2016. Nutritional and anti-nutritional composition of not difficult, however the approval from of- Oecophylla smaragdina (Hymenop- ficial authorities in El Salvador may be more tera: Formicidae) and Odontotermes sp. complicated.” The catastrophic ecological (Isoptera: Termitidae): Two preferred and economic results of numerous exotic edible insects of Arunachal Pradesh, In- species introduction, both accidental and in- dia. Journal of Asia-Pacific Entomology tentional around the world has still not taught 19: 711 – 720. everyone its cautionary lessons. Cole AC Jr and Jones JW, 1948. A study of the weaver ant, Oecophylla smaragdina (Fab.). American Midland Naturalist ACKNOWLEDGEMENTS 39: 641 – 651. Collingwood CA, 1970. Formicidae (Hymenop- tera, Aculeata) from Nepal. Khumbu I thank M. Wetterer for comments on this Himal, Ergebnisse des Forschungsun- manuscript; T. Schultz and E. Okonski for ternehmens Nepal Himalaya 3: 371 – help at the Smithsonian; M. Janda, D. Gen- 387. eral, R. Recustodio, T. Colvin, T. Chan, S. Cook J, 1773. An Account of a Voyage Round the De Greef, J. Fellowes, and H. Bharti, as well World. Volume III. Book II. in Hawkes- many users of Facebook, Instagram, Flickr, worth, J. (ed.) An Account of Voyages Yourshot, inaturalist, ispotnature, and Proj- Undertaken by the Order of His Present ect Noah, for sharing unpublished records Majesty for Making Discoveries in the and photographs; D.P. Wojcik and S.D. Por- Southern Hemisphere, and Successively Performed by Commodore Byron, Cap- ter for compiling their valuable FORMIS tain Wallis, Captain Carteret, and Cap- bibliography; R. Pasos and W. Howerton of tain Cook, in the Dolphin, the Swallow, the FAU library for processing so many in- and the Endeavour, Drawn Up from the terlibrary loans; Florida Atlantic University Journals Which Were Kept by the Sev- for financial support. eral Commanders, and from the Papers of Joseph Banks, Esq. W. Strahan and T. Cadell, London. 438 pp. Geographic distribution of the weaver ant Oecophylla smaragdina 11

Crozier RH, Newey PS, Schlüns EA and Robson Huang HT and Yang P, 1987: Ancient cultured SKA, 2010. A masterpiece of evolution - ant used as biological control Oecophylla weaver ants (: agent. BioScience 37: 665 – 671. Formicidae). Myrmecological News 13: Kami KS and Miller SE, 1998. Samoan insects 57 – 71. and related : checklist and Dlussky GM, Wappler T and Wedmann S, 2008. bibliography. Bishop Museum Technical New middle Eocene formicid species Report 13: 1 – 121. from Germany and the evolution of Kapur AP, 1961. Zoological results of the In- weaver ants. Acta Palaeontologica Po- dian Cho-oyu Expedition (1958) in lonica 53: 615 – 626. Nepal. Records of the Indian Museum Duncan I, 1999. Second review of progress on ca- 59: 269 – 303. shew production, processing and mar- Keuchenius PE, 1914. De Betekenis van twee keting in Cooperativa Reforma Agraria bekende mieren, in verband met het La Marañonera de R. L. Coralama. U. groene-luizenvraagstuk van de Koffie. S. Agency for International Develop- Tijdschrift Teijsmannia 10: 711 – 716. ment (USAID), Washington D.C. 26 pp. Kurane S, Bhoje PM and Sathe TV, 2015. On a Emery C, 1889. Formiche di Birmania e del new species of the genus Oecophylla Tenasserim raccolte da Leonardo Fea (Hymenoptera: Formicidae). Flora and (1885-87). [part] Annali del Museo Civ- Fauna 21: 37 – 40. ico di Storia Naturale 27: 485 – 512. Lokkers C, 1986. The distribution of the weaver Fabricius JC, 1775. Systema entomologiae, ant, Oecophylla smaragdina (Fabricius) sistens insectorum classes, ordines, gen- (Hymenoptera: Formicidae) in northern era, species adiectis synonymis, locis, Australia. Australian Journal of Zool- descriptionibus, observationibus. Flens- ogy 34: 683 – 687. burg and Leipzig. 832 pp. Mou J, 1959. Ecological observation on leaf Forel A, 1879. Études myrmécologiques en 1879 house ants. [in Chinese] Chinese Jour- (deuxième partie [1re partie en 1878]). nal of Zoology 12: 574. Bulletin de la Société Vaudoise des Sci- Mukherji D and Ribeira S, 1925. On a collection ences Naturelles 16: 53 – 128. of ants (Formicidae) from the Andaman Forel A, 1901. Formiciden aus dem Bismarck- Islands. Records of the Indian Museum Archipel, auf Grundlage des von Prof. 27: 205 – 209. Dr. F. Dahl gesammelten Materials. Peel MC, Finlayson BL and McMahon TA, 2007. Mitteilungen aus den Zoologischen Mu- Updated world map of the Köppen-Gei- seum in Berlin 2: 4 – 37. ger climate classification.Hydrology and Forel A, 1910. Formicides australiens reçus de Earth System Sciences 11: 1633 – 1644. MM. Froggatt et Rowland Turner. Re- Ridley HN, 1890. On Oecophylla smaragdina. vue Suisse de Zoologie 18: 1 – 94. Journal of the Straits Branch of the Roy- Friederichs K, 1920. Weberameisen und Pflanzen- al Asiatic Society, Singapore 1890: 345. schutz. Tropenpflanzer 5-6: 142 – 150. Santschi F, 1919. Cinq notes myrmécologiques. Greenslade PJM, 1965. Promecotheca opacicol- Bulletin de la Société Vaudoise des Sci- lis Gestro (Coleoptera : Chrysomelidae) ences Naturelles 52: 325 – 350. on the island of Tikopia. Pacific Insects Smith F, 1857. Catalogue of the hymenopterous 7: 661 – 664. insects collected at Sarawak, Borneo; Groff GW and Howard CW, 1924: The cultured Mount Ophir, Malacca; and at Singa- citrus ant of South China. Lingnaam Ag- pore, by A. R. Wallace. Journal and ricultural Review 2: 108 – 114. Proceedings of the Linnean Society of Guénard B and Dunn RR, 2012. A checklist of London Zoology 2: 42 – 88. the ants of China. Zootaxa 3358: 1 – 77. 12 James K Wetterer

Smith F, 1858. Catalogue of hymenopterous in- Wetterer JK and Vargo DL, 2003. Ants (Hyme- sects in the collection of the British Mu- noptera: Formicidae) of Samoa. Pacific seum. Part VI. Formicidae. British Mu- Science 57: 409 – 419. seum, London. 216 pp. Wilson EO and Taylor RW, 1967. Ants of Poly- Tak N and Rathore NS, 2004. Insecta: Hymenop- nesia. Pacific Insects Monographs tera: Formicidae. Zoological Survey of 14:1-109. India, State Fauna Series 8: 161 – 183. Yhoung-Aree J and Viwatpanich K, 2005. Ed- Trainor CR and Andersen AN, 2010. The ant fau- ible insects in the Laos PDR, Myanmar, na of Timor and neighbouring islands: Thailand, and Vietnam. In: Ecological potential bridges between the disjunct Implications of Minilivestock (Paoletti faunas of South East Asia and Australia. MG, ed.), Science Publishers, Inc., En- Australian Journal of Zoology 58: 133 field, pp 415 – 440. – 144. Ying R, 1966. Yellow orange ants in Hubei. [in Van Mele P, 2008. A historical review of research Chinese] Chinese Bulletin of Entomol- on the weaver ant Oecophylla in bio- ogy 2: 112 – 113. logical control. Agricultural and Forest Young GR, 2001. A review of Sexava research Entomology 10: 13–22. and control methods in Papua New Vanderplank FL, 1959. The weaver ant. New Sci- Guinea. Northern Territory Department entist 6: 944 – 946. of Primary Industry and Fisheries Tech- Wang YL, 2008. Survey on species diversity, ants nical Bulletin 288: 31 – 43. in Zhoukou Region. [in Chinese] Chinese Bulletin of Entomology 45: 145 – 149.

ASIAN MYRMECOLOGY A Journal of the International Network for the Study of Asian Ants Communicating Editor: Francisco Hita Garcia