2003. The Journal of Arachnology 31:274±277

OBSERVATIONS OF MINUTISSIMUS (ARANEAE, ), A PARTHENOGENETIC

Robert L. Edwards: Box 505, Woods Hole, Massachusetts 02543, USA. Eric H. Edwards: 45 Canterbury Lane, East Falmouth, Massachusetts 02536, USA. Annabel D. Edwards: Massachusetts General Hospital, Department of Anesthesia, Boston, Massachusetts 02114, USA.

ABSTRACT. It has been suggested by several authorities that at least one species of spider of the Theotima, family Ochyroceratidae, occurring in tropical regions in South Africa, the Caribbean and Asia may be parthenogenetic. Theotima minutissimus is particularly abundant in the tropical rainforest leaf litter on El Yunque, Puerto Rico. While many hundreds of specimens have been collected over many years, none has been a male. To examine the possibility that this small species, Ϯ 0.9 mm body length, is parthenogenetic, live specimens were collected and maintained in the laboratory. A second generation spiderling, raised separately, produced viable progeny. Keywords: Parthenogenesis, spider, tropical rainforest, leaf litter, Wolbachia

The six-eyed spider family Ochyroceratidae ples. These samples were taken in all the prin- is found in subtropical and tropical regions cipal habitats on El Junque. The abundance of around the world and is especially species rich T. minutissimus in the samples ranged from in the Indo-Paci®c. There are ten genera found 5±460 individuals mϪ2. An earlier study in the worldwide with four known from the western Tabonuco forest (Dacryodes excelsa Vahl), hemisphere; Fageicera and recorded demonstrated a range of abundance of 43±166 only from Cuba, Ochyrocera in the Caribbean individuals mϪ2, with a mean annual density region and Brazil, and Theotima. Theotima of 74 (Pfeiffer 1996). Only in the Tabanuco minutissimus, originally described as Oono- forest was the abundance of T. minutissimus pinus minutissimus Petrunkevitch 1929, oc- consistently exceeded by another litter spider curs in the Caribbean region, also in the Indo- species, Modisimus montanus Petrunkevitch Paci®c and possibly Africa as well (Platnick 1929. 1997). Females of this family typically carry The number of penultimate and adult Theo- their eggs until they hatch in their . tima minutissimus specimens examined to Two undescribed sympatric species of Ochy- date exceeds 1,000. No male has ever been rocera occur in forest leaf litter in Puerto found. The male palpi of other members of Rico. There is little published information the family Ochyroceratidae are distinctively available on the life history of any of these different from those of the female (see for ex- . ample ®g. 1, p. 83 in Emerit & Lopez 1985) Theotima minutissimus is abundant in the and are readily distinguishable even in the Caribbean National Forest, El Yunque, Lu- penultimate instar. The absence of any T. min- quillo, Puerto Rico. Its preferred habitat is lit- utissimus showing such palpi leads to the sup- ter composed of smaller leaves in second position that this spider could be parthenoge- growth forests with understory shrubs or in netic. Machado (1964) became convinced that those forests with damper and more easily de- a species of the genus Theotima of the femi- cayed leaves such as mahogany (Swietenia nina group in Africa in the Congo and Angola macrophylla King) or bamboo (Bambusa vul- was parthenogenetic. Deeleman-Reinhold garis Schrad.). T. minutissimus prefers eleva- (1995) subsequently reported what appears to tions less than 800 m. From 1992±2001, we be the same species of T. minutissimus from collected more than 700 0.25 m2 litter sam- Panama, the Philippines, Borneo, Indonesia,

274 EDWARDS, ET AL.ÐT. MINUTISSIMUS, A PARTHENOGENETIC SPIDER 275

Sumatra, Thailand and Guam. She noted, in All 22 females were dead by the end of De- addition to not ®nding males, that ``The most cember 2001. salient feature is the paired lightly sclerotized When the eggs are extruded they are ovoid arch on the lateral margin of the epigastric in shape and uniformly translucent. The entire opening: the ``spermathecae'' are very thin clutch is extruded in a matter of hours, usually walled and easily shift position.'' This report during the night. The eggs are more or less implies a potentially faulty genital apparatus. arranged in one layer to a pad of silk which Deeleman-Reinhold also suggested that this the female carries in her chelicerae. For the spider may be parthenogenetic. ®rst 9±11 days the eggs showed no clear sign To determine whether or not Theotima min- of development but became more elongate utissimus was in fact parthenogenetic, on 8 and showed subtle changes in density. Over February 2001, 22 females judged to be either the next week the leg structure became ap- mature or penultimate were collected and parent and development proceeded rapidly placed separately in small petri dishes (Fisher with the abdomen and thorax becoming dis- No. 09-75-53C, 50 mm diameter Petri Dish tinct. The developing spiderlings were with absorbent pad). Each pad was moistened bunched together with their tarsi connected to with two drops of water, and 10 or more en- the pad of silk. The spiderlings were arranged tomobryid springtails (Sinella curviseta facing outwards (Fig. 1). Just before leaving Brook) were added as food. The dishes were the mother, on average three weeks after ex- then placed in a larger plastic container con- trusion, the spiderlings engaged in an activity taining a moistened sponge to help maintain that we describe as doing `pushups'. This con- humidity. The temperature was maintained be- sisted of actively ¯exing their legs, which tween 19±22 ЊC. tended to push their body away from the Unless the spider positioned itself on the mother and appeared to be associated with the top of the petri dish, venter side up, it was not shedding of the ®rst instar exoskeleton. This possible to determine whether or not it was an process was readily observed because at this adult. There appear to be at least 5 instars that time females typically positioned themselves gradually increase in pigmentation from the upside-down in their webbing. Once they sep- third instar on, but with no obvious changes arated from their mother, the spiderlings soon in pigmentation, body structure or the appear- assumed an upside-down orientation in the ance of the epigynum between the penultimate web. On average, the time between egg pro- and adult stage. Twenty adults averaged duction and release of the spiderlings was slightly less than 1 mm (0.79±0.98) and 20 three weeks to one month. Spiderlings re- penultimate 0.76 mm (0.69±0.83) in body mained in the webbing originally made by the length. In the petri dish any exuviae or re- female. The pad of silk containing the rem- mains of deceased spiders were quickly scav- nants of the ®rst instar exuviae remained in enged by the springtails and seldom seen. the web as well. Within two days after the Female Theotima minutissimus carry their spiderlings hatched the female moved else- eggs in a bundle in their chelicerae until the where in the dish and established a new web. spiderlings are fully developed and capable of These relatively dense, irregular sheet webs moving about on their own. The number of usually extended from the top of the dish to eggs varies from 1±9; 4±6 eggs per clutch the side and sometimes to the bottom. were most common. Eggs were produced by The second instar spiderlings are colorless 17 of the 22 captive females with egg num- and virtually transparent. Ten were measured bers varying from 1±9. In three cases eggs averaging 0.39 mm in body length. Within 2 were abandoned before they developed. In or 3 days, all but one spiderling produced three other cases females produced a second from each clutch was separated from the clutch of eggs, and in one case three clutches. mother and placed in separate petri dishes. A In all, ten clutches were produced that resulted few spiderlings were observed attempting to in viable spiderlings. A few females were ob- capture smaller springtails. With disappoint- served to drop their clutch of eggs brie¯y and ing regularity the spiderlings disappeared after return to it. Four females died in the ®rst two about 10±14 days whether they were living months. It was not until November 2001 that alone or with the mother. It could not be de- the remaining females were dying regularly. termined if they had been preyed upon by 276 THE JOURNAL OF ARACHNOLOGY

Figure 1.ÐSeparately raised Theotima minutissimus with clutch of four well developed spiderlings. Spiderlings still attached by legs to a pad of silk carried by the female. The abdomens are directed inward. Space bar ϭ 1 mm. large springtails. We suspect they died after speci®c with Theotima minutissimus. Lake depleting any energy stores they inherited (1986) reported that a female Isopeda insignis, and/or because they lacked appropriate prey. family Heteropodidae, had reproduced parthe- In April 2001, four newly hatched spider- nogenetically. Gruber (1990), following up an lings were transferred into separate petri dish- earlier report by Deeleman-Rhinhold (1986) es with soil. Two individuals survived, one for suggesting that Dysdera hungarica, family 54 days before it died. The second survived Dysderidae, might be able to reproduce parthe- and reached a length of 0.75 mm by January nogenetically, kept female Dysdera hungarica 2002. At this point the spider was judged to isolated from males. The females produced be penultimate. On 11 March this spider pro- egg sacs and a few of the unfertilized eggs duced a clutch of four eggs which developed ultimately became adults which in turn pro- normally. These spiderlings hatched 4 April. duced a further generation. Gunnarsson & An- The female produced a second clutch of four dersson (1992) studied chromosome variation eggs on 14 April, which developed normally in Pityohyphantes phrygianus, family Liny- and all four spiderlings became free of the phiidae. They discovered that 1±2% of em- mother (hatched) on 8 May. byros were haploid which suggested devel- While not the ®rst to suggest that some spi- opment from unfertilized eggs. Shimojana & ders might be parthenogenetic, Machado Nishihira (2000) described a trogloditic spider, (1964) extensively examined the possibility Coelotes troglocaecus, family Amaurobiidae. for a maleless population of Theotima and No males were found and the females had de- concluded that the species had to be parthe- generate copulatory organelles. nogenetic. At this time it is not possible to say It would appear that there may be many that the species he studied was or was not con- roads to achieving parthenogenesis in spiders. EDWARDS, ET AL.ÐT. MINUTISSIMUS, A PARTHENOGENETIC SPIDER 277

Ultimately the question of whether or not an Antilles (Araneae). Revue Arachnologique 6(2): endosymbiotic bacterium such as Wolbachia 81±89. may be involved is worth pursuing now that Gruber, J. 1990. Fatherless spiders. British Arach- it has been demonstrated to exist in Nephila nological Society Newsletter 58:3. Gunnarsson, B. & A. Andersson. 1992. Chromo- clavata (Oh et al. 2000). Weeks et al. (2002) some variation in embryos of a solitary spider, caution that there are other endosymbionts or Pityohyphantes phrygianus, with a skewed sex nuclear effects that could be involved. What- ratio. Heriditas 117:85±91. ever the case with respect to Theotima minu- Lake, D.C. 1986. Possible parthenogenesis in the tissimus, its world wide distribution and suc- huntsman spider Isopoda insignis (Araneae, cess in its chosen habitat points up the Sparassidae). Journal Arachnology 14(1):129. potential evolutionary signi®cance of adopt- Machado, A.B. 1964. Sur l'existence de la partheÂn- ogenese dans quelques espeÂces d'AraigneÂes Ochy- ing the parthenogenetic mode of reproduction. roceÂratides. Comptes Rendus de L'Academie Sci- Voucher specimens are deposited in the ence Paris 2(258):5056±5059. American Museum of Natural History, New Petrunkevitch, A. 1929. The Spiders of Porto Rico. York. Transactions Connecticut Academy of Arts & Science, Pt. 1,(30):1±158. ACKNOWLEDGMENTS Pfeiffer, W.J. 1996. Litter Invertebrates. Pp. 138± 181. In The Food Web of a Tropical Rain Forest. We gratefully acknowledge Kenneth Frank for his (D.P. Reagan, R.B. Waide, Eds.) Univ. Chicago many suggestions and comments on the draft manu- Press. script, and subsequently appreciate those of Paula Platnick, N.I. 1997. Advances in Spider . Cushing and Dan Mott. Jill Thompson (El Verde Field 1992±1995: With Redescriptions 1940±1980. Station, Luquillo, Puerto Rico) helped to make life liv- New York Entomological Society. 976 pp. able in the ®eld. The springtails used to feed the spi- Shimojana, M. & M. Nishihira. 2000. A new cave- ders, Sinella curviseta Brook (Collembola, Entomob- dwelling eyeless spider of the genus Coelotes ryidae) were from a culture originally provided by Dr. (Araneae: Amaurobiidae) from Okinawa Island, Michael L. Draney, University of Wisconsin-Green the Ryukyu Islands, Japan, with notes on possi- Bay. ble parthenogenesis. Acta Arachnology 49(1): 29±40. LITERATURE CITED Oh, H. W., M. G. Kim, S. W. Shin, K. S. Bae & H. Deeleman-Reinhold, C.L. 1986. Dysdera hungarica Park. 2000. Ultrastructural and molecular iden- KulczynskiÐA case of parthenogenesis? Actas ti®cation of a Wolbachia endosymbiont in a spi- X Congress International Arachnology Jaca (Es- der, Nephila clavata. Insect Molecular Biology 9:539±543. panÄa) 1:25±31. Weeks, A.R., K.T. Reynolds & A.A. Hoffmann. Deeleman-Reinhold, C.L. 1995. The Ochyrocerati- 2002. Wolbachia dynamics and host effects: dae of the Indo-Paci®c Region (Araneae). The what has (and has not) been demonstrated? Raf¯es Bulletin of Zoology, Supplement No. 2: Trends in Ecology & Evolution 17(6):257±267. 1±103. Emerit, M. & A. Lopez. 1985. Ochyrocera thibau- Manuscript received 24 June 2002, revised 15 No- di, n. sp., et autres Ochyroceratidae des Petites vember 2002.