I The Journal of ARACHNOLOGY OFFICIAL ORGAN OF THE AMERICAN ARACHNOLOGICAL SOCIETY VOLUME 29 2001 NUMBER 1 THE JOURNAL OF ARACHNOLOGY EDITOR-IN-CHIEF : James W. Berry, Butler University MANAGING EDITOR: Petra Sierwald, Field Museum SUBJECT EDITORS: Ecology — Matthew Greenstone, USDA; Systematics — Mark Harvey, Western Australian Museum; Behavior and Physiology — Robert Suter, Vassar College EDITORIAL BOARD: Alan Cady, Miami University (Ohio); James Carrel, University of Missouri; Jonathan Coddington, Smithsonian Institution; William Eberhard, Universidad de Costa Rica; Rosemary Gillespie, University of California, Berkeley; Charles Griswold, California Academy of Sciences; Marshal Hedin, San Diego State University; Herbert Levi, Harvard University; Brent Opell, Virginia Pol3dechnic Institute & State University; Norman Platnick, American Museum of Natural History; Ann Rypstra, Miami University (Ohio); Paul Selden, University of Manchester (UK.); Matthias Schaefer, Universitaet Goettingen (Germany); William Shear, Hampden-Sydney College; Keith Sunderland, Horticulture Research International (UK.); I-Min Tso, Tunghai University (Taiwan). The Journal of Arachnology (ISSN 0160-8202), a publication devoted to the study of Arachnida, is published three times each year by The American Arach- nological Society. Memberships (yearly): Membership is open to all those in- terested in Arachnida. Subscriptions to The Journal of Arachnology 2in& American Arachnology (the newsletter), and annual meeting notices, are included with mem- bership in the Society. Regular, $30; Students, $20; Institutional, $80 (USA) or $90 (all other countries). Inquiries should be directed to the Membership Secretary (see below). Back Issues: Patricia Miller, P.O. Box 5354, Northwest Mississippi Community College, Senatobia, Mississippi 38668 USA. Telephone: (601) 562- 3382. Undelivered Issues: Allen Press, Inc., 1041 New Hampshire Street, P.O. Box 368, Lawrence, Kansas 66044 USA. THE AMERICAN ARACHNOLOGICAL SOCIETY PRESIDENT: Frederick A. Coyle (1999-2001), Department of Biology, Western Carolina University, Cullowhee, North Carolina 28723 USA. PRESIDENT-ELECT: Brent D. Opell ( 1 999-200 1 ), Department of Biology, Virginia Polytechnic Institute and State University, Blacksburg, Virginia 24061 USA. MEMBERSHIP SECRETARY: Norman I. Platnick (appointed), American Museum of Natural History, Central Park West at 79th St., New York, New York 10024 USA. TREASURER: Gail E. Stratton, Department of Biology, University of Missis- sippi, University, Mississippi 3^677 USA. SECRETARY: Alan Cady, Dept, of Zoology, Miami Univ., Middletown, Ohio 45042 USA. ARCHIVIST: Lenny Vincent, Fullerton College, Fullerton, California 92634. DIRECTORS: Bruce Cutler (2000-2002), Paula Cushing (1999-2001), Ann Rypstra (1999-2001). HONORARY MEMBERS: C. D. Dondale, H. W. Levi, A. F. Millidge, W. Whit- comb. Cover photo: Female wolf spider, Rabidosa rabida (Araneae, Lycosidae) with hatchlings that recently emerged from the egg sac still attached to her spinnerets. {Photo by Robert Suter) Publication date: 20 April 2001 0 This paper meets the requirements of ANSI/NISO Z39.48-1992 (Permanence of Paper). 200L The Journal of Arachnology 29:1-10 REVISION OF THE SPIDER GENUS NEOANAGRAPHIS (ARANEAE, LIOCRANIDAE) Richard S. Vetter: Department of Entomology, University of California, Riverside, California 92521 USA ABSTRACT. The spider genus Neoanagraphis consists of two partially sympatric species, K cham- berlini Gertsch & Mulaik 1936 and K pearcei Gertsch 194L Herein I review the genus which is now transferred from the Clubionidae to the Liocranidae, provide a distribution map and describe the females for the first time. Over 55% of the specimens of both species examined in this study came from the Nevada Test Site (i.e., atomic bombing range) in southern Nevada. Collection phenology at this site showed almost non-overlapping temporal activity for the males and habitats for each species within this area. Immatures from the Nevada Test Site also could be separated where the ventral anterior tibia spination and habitat dichotomy matched that of the adults. If this extrapolation holds throughout the distribution and there are no additional species, one should be able to accurately identify immatures of any size to species. Keywords: Arachnida, taxonomy The spider genus Neoanagraphis was erect- ed in 1936 with the naming of the type spe- cies, N. chamberlini, based upon one mature male which was initially assigned to the fam- ily Gnaphosidae (Gertsch & Mulaik 1936). A second species, N. pearcei, was later de- scribed, again from a single male specimen and the genus was transferred to the family Clubionidae (Gertsch 1941). From this point, Neoanagraphis is scarcely mentioned in the literature except in faunal surveys (Allred et al. 1963a; Allred & Gertsch 1976; Allred & Kaston 1983; Jung & Roth 1974; Ryckman & Lee 1956). Neoanagraphis is a rather non-descript looking ‘Alubionoid” spider with the excep- tion of the unique characteristic that the tarsal claws of legs III and IV are extremely long with few teeth at the base (Fig. 1). Here I review this genus of spiders and describe the females of both species for the first time. The acronyms used in this paper are as fol- lows: AMNH “ American Museum of Natu- ral History, N. Platnick; CAS ~ California Academy of Science, C. Griswold, D. Ubick; CDFA ” California Department Food & Ag- riculture, Visalia, California, M. Moody; JLO — J.L. Ortiz, Laguna Niguel, California; NMSU = New Mexico State University Ar- thropod Museum, D. Richman; UCR ~ En- tomology Museum, University Califomia- Riverside; VDR = V.D. Roth, Portal, Arizona; WRI = W.R. Icenogle, Winchester, California. METHODS All specimens were examined under alco- hol with a Wild M5 Microscope with ocular micrometer. Leg measurements and spination pattern were taken from 20 spiders of each gender of each species where possible; spiders used here had at least four legs from the same side of the body intact. (Many specimens were collected in pitfall traps and, hence, were des- iccated with multiple disarticulated limbs). Explanation of spination pattern: 1-1-1 rep- resents surface having 1 basal and 1 distal spine with 1 median spine equidistant between them, 1- 1-0-2 represents surface with 1 basal and 2 distal spines with 1 median spine about twice as far from distal as the basal. Where possible, up to half of each cohort of 20 orig- inated from the Nevada Test Site (NTS). All mature spiders examined in this study were measured for cephalothorax and abdomen widths and lengths as well as cymbium length or epigynal plate width and length. Epigynal plate width was measured as the distance be- tween the median aspects of the base of the lateral spurs; epigynal plate length was mea- sured from the epigynal plate width line to the posterior tip of the plate. All measurements in the paper are presented in millimeters and the limits of the range are presented in parenthe- 2 THE JOURNAL OF ARACHNOLOGY Figure 1. — Leg IV tarsus of Neoanagrapnis showing the long, nearly unarmed claw. ses. Additional methods are described under sections where pertinent. Elevation data not recorded on the specimen label were obtained in correspondence with either the original coF lector or with arachnologists familiar with the collection locale. Live-captured mature specimens (2d, 1 $ ) were boiled in water for 5 min to splay the spinnerets. They were preserved in 70% al- cohol and sent to the AMNH for examination to determine whether the genus would remain in the Clubionidae. KEY 1. Anterior tibia with two pairs of ventral spines (apicals weak if present); tip of em- bolus short, blunt, appearing often as fold- ed flap (Figs. 2, 3); posteriorly-directed epigynal plate usually as long as wide or longer, usually V-shaped and usually ex- tending to the posterior edge of the sper- mathecae or beyond (Fig. 4, 5) . . chamberlini 2. Anterior tibia with three pairs of ventral spines (apicals weak if present); tip of em- bolus long, thin, scythe-like (Figs. 6, 7); posteriorly-directed epigynal plate usually wider than long, U-shaped and extending at most to the midpoint of the spermathe- cae (Fig. 8) ................... . .pearcei TAXONOMY Neoanagraphis Gertsch & Mulaik 1936 Figures 1, 10 Neoanagraphis Gertsch & Mulaik 1936: 11 (Gna- phosidae); Gertsch 1941: 19 (Clubionidae); Com- stock 1948: 327 (Gnaphosidae); Roewer 1955: 559 (Clubionidae, Liocraninae, Liocraneae); Bon- net 1958: 3046 (Drassidae); Lehtinen 1967: 251 (Clubionidae sensu str.); Brignoli 1983: 549 (Clu- bionidae, Clubioninae); Platnick 1993: 605 (Clu- bionidae); Platnick 1998: 701 (Clubionidae) Type species. — Neoanagraphis chamber- lini Gertsch & Mulaik 1936 by original des- ignation. Diagnosis. — Whether one considers Neoanagraphis spiders in the broad sense of all the genera formerly housed in the Clu- bionidae or in the Liocranidae, they can be distinguished from other North American gen- era in either family by the long tarsal claws on legs III and IV that appear almost devoid of teeth. Description. — Small to medium-sized spi- ders, with no bodily pigmentation. Coloration of few live specimens examined similar to those preserved in alcohol: cephalothorax uni- formly pale orange to tan-orange, darkening anteriorly, width about % of length, widest at legs II-III, males slightly wider than females, covered with thin, white hairs with scattered, dark, anteriorly or medially directed setae. Longitudinal row of single, anteriorly-directed setae between eyes and thoracic furrow. Eyes subequal surrounded by black rings. AER re- curved in dorsal view, slightly procurved in anterior view, eyes separated by less than eye diameter. AME dull but not black, all others luminescent. PER straight to slightly recurved in dorsal view, procurved in anterior view, PLE separated from PME by eye diameter, PME slightly farther from each other. PER slightly longer than AER with PLE extended laterally just beyond ALE. Clypeus about height of eye diameter. Undivided chilum. Conspicuous longitudinal thoracic furrow. Chelicerae dusky orange, darker than cepha- lothorax. Teeth: 3 promargin, 2 retromargin, latter separated by 3X width of tooth base. Conspicuous boss. Endites quadrate, labium slightly wider than long. Sternum slightly lon- ger than wide, sometimes darker than legs. Coxa similar in color to legs. Pre-coxal tri- angles lacking. Trochanters notched, III and IV deeply so. Legs similar in color to cepha- lothorax with heavy spination. Leg IV longest, about 15-40% longer than legs I-III which are all subequal in length with minute plumose or feathery hairs. Tarsi lacking claw tufts, dense with white scopulae, sometimes appearing flexible in preserved specimens. Tarsal claws of posterior legs extremely long with few teeth at base, almost hidden (Fig. 1). Tarsal claws of anterior legs shorter, looking more typical. Trichobothria on tarsi and metatarsi of varying lengths, some very long. Spination VETTER— THE SPIDER GENUS NEOANAGRAPHIS 3 pattern virtually non-varyieg for dorsal fem= ora and ventral metatarsi; retrolateral surfaces show greatest variation (patterns that differ between genders/species listed separately un- der species descriptions). Some spines rest in- termediate between two surfaces; these were consistently assigned to one surface although they readily could have been assigned to an- other. Patterns that were most consistent be- tween both sexes and both species are: fem- ora: I, 11 pO-M or 0-M-l; I=IV dl-1-1; tibiae: III pM-1, dl-0-1, v2-2-2, IV pl-1-1, d 1-0-1, v2-2-2; metatarsi: I v2-2-0, II v2-2-0, III pl-1-0-2, rl-1-0-2, V2-2-2, IV pl-1-0-2, dl-l“0, rl- 1-0-2, v2-2-2. Abdomen uniformly cream to tan, in rare instances brown, with scattered dark setae, no conspicuous mark- ings, many long setae on anterior surface, oval in shape, width about % of length in well-pre- served specimens. Occasionally heart can be seen through dorsal integument. In gross ex- amination, ALS occasionally very long ap- pearing gnaphosoid in character in preserved specimens otherwise appearing short and con- ical. Embolus of male palp dorsally projecting on apical-most portion of tegulum, membra- nous conductor dorsal to embolus. Median apophysis on retrolateral surface of tegulum above midline, translucent, elongate and con- cave. Epigynum with median plate, posteri- orly-directed spines lateral where epigynal plate originates anteriorly. Anterior epigynal openings sometimes not readily visible. In dorsal view, epigynum with two simple oval spermathecae, each with duct arching anter- iolaterally to small rotund structure (bursa co- pulatrix?). Natural history. — -Neoanagraphis spiders have been collected in an unidentified mam- mal burrow (Ryckman & Lee 1956), in a ta- rantula (Aphonopelma sp.) burrow and in a kangaroo rat mound; but otherwise, little is known of their natural history. Several collec- tion labels mention that the spiders were found on sand dunes or in washes; one male was live-collected as it crawled around a sandy wash around midnight, at 4 °C. They have been collected often at elevations of 900=1950 m; however, some were taken from below sea level near the Salton Sea to 200 m in Mexico and western Arizona. Jung and Roth (1974) listed it as being found in Zone 1 of their study which is characterized by limestone foothills, alluvial plains and valleys of the Chiracahua Mountains from 4000-5000 feet (1200-1500 m). Only a few live specimens were captured during the course of this study. A female was maintained for >8 mon. She fed on Drosoph- ila flies and small crickets but ignored mos- quitoes, larval waxmoth {Galleria mellonella) and a spider (Drassyllus insularis Banks 1900). Genitalic variation. — In this study, ap- proximately 24 mature females of each spe- cies were available for examination. Despite fewer females relative to males, females showed greater genitalic variation. In both species, the epigyna are covered with hairs which obscure some of its minute features. The epigynal plate in N. chamberlini varied in length such that it could extend past the pos- terior edge of the underlying spermathecae or sometimes would just barely reach the poste- rior edge. Additionally, although the plate was usually V-shaped, the width of the plate varied from narrow to wide, and sometimes was rounded on the posterior edge (Fig. 5) similar to V. pearcei. The plate in N. pearcei was comparatively less variable in length, width and its rounded, U-shaped posterior edge (Fig. 8), however, at least one specimen had a V- shaped plate reminiscent of N. chamberlini. The lateral spurs were rather consistent in size within each species (conspicuous in N. cham- berlini, minute in N. pearcei) and for the few specimens examined here that is a good di- agnostic feature to be used in concert with other features such as anterior tibial spination. Yet they did vary from spike-like to that of an equilateral triangle and could be slightly dif- ferent in form on the right and left sides of the same spider. About half of the females of each species were dissected to inspect the dorsal view of the genitalia, leaving the other females intact for future researchers. There are no consistent internal characters that allow species separa- tion. The small anterior rotund structures (bur- sa copulatrix?) for both species may lie di- rectly on top of the spermathecae or extend laterally (Fig, 9). Likewise, the duct running to it may be thin or thick, and at an acute or obtuse angle of curvature. Possibly with great- er numbers of spiders in the future, diagnostic internal features may become apparent. In contrast, the males were very consistent in their palpal features with little marked var- 4 THE JOURNAL OF ARACHNOLOGY Figures 2-5. — Neoanagraphis chamberlini Gertsch & Mulaik. 2. Male left palp, ventral view; 3. Same, retrolateral view (scale = 0,25 mm); 4. Epigynum, ventral view, hairs removed from ventral surface (scale = 0.1 mm); 5. Schematic drawings of epigyna showing variation of epigynal plate and its position relative to the spermathecae, “e” = dorsally-directed embolus, additional arrows point to dorsal process of retro- lateral tibial apophysis. iation in characters except for differences due to aberrations caused by preservatives which expanded the palp or changed the relative ori- entation of the structures. Neoanagraphis chamberlini Gertsch & Mulaik 1936 Figs. 2-5, 10 Neoanagraphis chamberlini Gertsch & Mulaik 1936: 11-12, fig. 15. Male holotype - White Sands, New Mexico, August 1934, in AMNH, examined. Diagnosis. — The two species in the genus differ consistently in a number of traits and can be readily separated. Neoanagraphis chamberlini is characterized by (1) anterior tibia with two pairs of ventral spines (dis- counting smaller apicals if present), (2) the dorsal process of the retrolateral tibial apoph- ysis (RTA) is thick and projected laterally (Fig. 2), (3) the dorsally-projecting embolus tip is truncate and looks like a folded flap or cresting wave (Fig. 3), and (4) the epigynal plate extends posteriorly past the midpoint, and most often to the posteriormost edge of the spermathecae (visible through the integu- ment) or beyond (Fig. 4, 5). In contrast, N. pearcei has (1) anterior tibia with three pairs of ventral spines (discounting smaller apicals if present), (2) the RTA is forked with the dor- sal process straight, thin and apically-directed (Fig. 7), (3) dorsally-projecting embolus is long, thin and scythe-like (Fig, 7), and (4) the epigynal plate extends posteriorly only to the middle of the spermathecae (Fig. 8). Although there is some overlap of the sizes, in general, the typical N. chamberlini is distinctly larger than the typical N. pearcei. Description.^Ma/e.- Total length 6.9 (3.8- 9.1). Carapace 3.4 (L9-4.5) length, 2,7 (1.6- 3.7) width. Abdomen 3.4 (1. 8-4.9) length, 2,1 (1. 1-3.0) width. Cymbium 1.08 (0,85-1.26) VETTER— THE SPIDER GENUS NEOANAGRAPHIS 5 Figures 6-8. — Neoanagraphis pearcei Gertsch. 6. Male left palp, ventral view; 7. Same, retrolateral view (scale = 0.25 mm); 8. Epigynum, ventral view, hairs removed from ventral surface (scale = 0.1 mm), “e” = dorsally-directed embolus, additional arrows point to dorsal process of retrolateral tibial apophysis. length. Additional spination differing from that presented for genus: femora: I r 0-1- 1=1, II r 0- 1-1 or O-l-l-l, III p O-l-UO-1 or l-l-l-l, IV p O-l-l-O-l, r 0-1-1 or O-l-l-O-l; tibiae: I p (variable with 2 or 3 spines), v 2-2-(2) (apicals weak), II p 1-1-1, r (variable with 0 to 2 spines), v 2-2-(2) (apicals weak). III r 0-1-1 or 1-0-1, IV r 1-0-1 or l-l-O-l; metatarsi: II p (variable with 0 to 2 spines), d 1-1-0. Female: Total length 7.7 (5. 5-9.7). Cara- pace 3.6 (2.9-4. 5) length, 2.8 (2. 2-3. 6) width. Abdomen 4.1 (23-5.7) length, 2.5 (L3-3.8) width. Epigynal plate: 0.20 (0.14-0.24) for both length and width. Epigynal plate bor- dered anteriolaterally by conspicuous, poste- riorly-directed spurs varying in shape from sharp spike to equilateral triangle. At posterior edge, plate varying from smoothly rounded U- shape (rare) to sharp, narrow V-shape (com- mon); if extends posteriorly, plate more likely to be V-shaped. Spination as in male except for: femora: I d l-l-O-l, II d l-l-O-l, III r 0- 1-1; tibiae: I p (variable with 0 to 2 spines). Distribution. — From the mountains around the Central Valley through the southeastern deserts in California, into southern Nevada, the southern half of Arizona, New Mexico and the western edge of Texas (Fig. 10). Also in the state of Sonora in Mexico. Material examined. — ^Holotype male, 112321$, 46 immatures. MEXICO; Sonora: 5 mi. N. Her- mosillo, near sea level, in unidentified rodent bur- row, 16 April 1952, 1$, R. Ryckman & K. Arak- awa (AMNH), S. end Sonoita River, 26 November 1959, 13, V. Roth (AMNH). UNITED STATES; Arizona: Cochise County: Portal, 4800 feet, 8 Au- gust 1965, 1 imm., W. Gertsch (AMNH); 13.5 mi S. Apache, 4330 feet, in kangaroo rat mound, 8 September 1968, 13,1 imm., E. Moore & T. Walk- er (VDR), 5 mi. N. Portal, 4770 feet, 19 April 1977, 1 $ , R. Chew (VDR); Chiricahua Mountains, South- west Research Station, 5400 feet, 16 September 1985, 13, V. Roth (CAS). Graham County: Calva, 3500 feet, 3 November 1955, 1$, V. Roth (AMNH). Pima County: Organ Pipe Cactus National Monu- ment, 1700 feet, on restroom floor, 18 November 1989, 13, W. Icenogle & T. Prentice (WRI); Tuc- 6 THE JOURNAL OF ARACHNOLOGY Figure 9. — Schematic drawings of the dorsal view of female Neoanagraphis genitalia. The top figure is the most common configuration of paired structures for both species. The remaining four fig- ures show the variation among individuals with only one side drawn. son, 2400 feet, 8 October 1953, Id, M. Cazier (AMNH), no date. Id, O. Bryant (AMNH). Yava- pai County: Congress, 3000 feet, 8 August 1948, 1 imm., C. & R Laurie (AMNH). Yuma County: Ca- beza Prieta National Wildlife Refuge, Tule Well, 600 feet, pitfall traps, 9 November 1996, 3d, V. Roth & D. Richman (NMSU); California: Fresno County: N. Kettleman Hills, under boards, 8 De- cember 1993, 1$, WH. Tyson (CDFA). Imperial County: 1 mi W Harper’s Well, San Felipe Creek, — 100 feet, in dunes, probably 11 July 1968 (not 7 November), 1 imm., M.E. Irwin & P.A. Rauch (UCR); 3 mi NW Glamis, sand dunes, 4 March 1972, 1?, A.R. Hardy (UCR). Inyo County: China Lake Naval Air Weapons Station, near S. Coso Vil- lage, 5800 feet, 27 May-8 June 1996, 1 9 ; 22 June- 10 August 1996, 1 penult. 9; in wash, 9 June- 10 August 1996, 3 imm.; near Birchum Springs, 10 August-14 September 1996, 1 d ; 22 June-10 Au- gust 1996, Id; White Hills, in pitfall trap under Joshua trees, 10 August-14 September 1996, 3d; 4 mi N Flight Line & GI roads, 14 September 1996- 15 February 1997, 1 9, G. Pratt & C. Pierce (UCR). Kern County: E. Randsburg, 12 April 1968, 1 imm.. Figure 10. — Southwestern United States and northern Mexico. Distribution of Neoanagraphis chamberlini (o) and N. pearcei (•). J. Cherry (UCR), Edwards Air Force Base, Leuman Ridge, 23 November 1997, in pitfall under Larrea tridentata, Id, C. & M. Breidenbaugh (UCR). Riv- erside County: Rice Dunes, 25 February 1978, 1 9, F. Andrews & A. Hardy (CDFA). San Bernardino County: Cadiz Dunes, 25 April 1978, 1 imm., A. Hardy & F. Andrews (CDFA); Joshua Tree National Monument, Cow Camp, 5800 feet, in pitfall traps, 24 September 1994, 5d, W. Sakai (UCR); Twen- tynine Palms, October 1944, Id, J. Branch (AMNH). Nevada: Nye County: Nevada Test Site (see be- low), 61d5 9, 35 imm., D. Allred (AMNH). New Mexico: Bernalillo County: Albuquerque, Indian Petroglyph State Park, 5000 feet, 7 June 1995, Id, summer 1996, Id, D. Lightfoot (UCR). Dona Ana County: Jornada Experimental Range, 4300 feet, in lowland grass pasture, 21 October 1999, Id, D. Richman (NMSU). Otero County: White Sands, August 1934, Id (holotype), S. Mulaik (AMNH). Socorro County: 20 mi N Socorro, 4500-6500 feet, 1989-1992, 26d7 9, 1 imm., S. Brantley (UCR). Texas: Hudspeth County: 8 mi. W. Sierra Blanca, 5 September 1946, Id (AMNH). Presidio County: in nest of Cratageomys castanops, August 1948, 1 9 , G. Menzies (AMNH). Neoanagraphis pearcei Geitsch 1941 Figs. 6-8, 10 Neoanagraphis pearcei Gertsch, 1941: 19—20, fig. 46. Male holotype - Yermo, San Bernardino Co., California, 28 October 1939, in AMNH, exam- ined. Diagnosis. — See N, chamberlini. Description. — Male: Total length 4.6 (3.2- 6.8). Carapace 2.3 (1. 7-3.0) length, 1.8 (1.3- 2.6) width. Abdomen 2.4 (1.5-3. 8) length, 1.4 (0. 8-2.2) width. Cymbium 0.80 (0.49-1.02) length. Additional spination from that pre- sented for genus: femora: I r l-l-O-l, II r 1- VETTER— THE SPIDER GENUS NEOANAGRAPHIS 1 M or HUGH, III r OAA or 1-Ul, IV r OH- 1; tibiae: I p l-l-O-l or l-l-l-l, r 1-1-1, v 2-2-2-(2) (apicals weak), II p l-l-l-l, r 1-1- 1, V 2”2-2-(2) (apicals weak). III r 0-1-1 or 1- 1-1, IV r 1-1-1; metatarsi I p (variable with 0 to 2 spines), r 0-1-0 or none, Up 1-1-0, r 0- 1-0, III d 0-1-0 or 1-1-0. Female: Total length 6.0 (4.0-8. 1). Cara- pace 2.8 (2. 1-3.5) length, 2.2 (1.6-2. 9) width. Abdomen 3.2 (L8-4.8) length, 2.1 (1.2-2.9) width. Epigynal plate 0.15 (0.10-0.18) length, 0.24 (0.18-0.31) width. Epigynal plate bor- dered anteriolaterally by minute posteriorly- directed spurs; at posterior edge, plate typi- cally rounded, U-shaped, extending only about to midpoint of underlying spermathecae (which can be seen through integument). Spi- natioe as in male except for: femora: I r (var- iable with 0 to 2 spines); tibiae: I p (variable with 0 to 3 spines), r none, II p l-l-O-l, r none, IV r 0-1-1 or 1-1-1; metatarsi: II p none, r none, III d (variable from 1 to 3 spines). Distribution. — ^-Eastem Sierra Range south into the mountains surrounding the Coachella Valley in California, southern portions of Ne- vada and Utah, north and western Arizona (Fig. 10). Material examined, — Holotype male, 866, 27$, 25 imm. UNITED STATES: Arizona: Mo- have County: 1 mi SE Bullhead City, 600 feet, pit- fall trap, 22-26 December 1980, 161$, B. Phelps (CDFA); Virgin River, 3 mi N, 7 mi E Littlefield, in pitfall trap, March-October 1982, 1 $ , D. Giuliani (CAS). Yuma County: near Sheep Tank Mine, 29 October 1958, 1$, V. Roth (VDR). California: Inyo County: E. side Owens Lake, 17 September 1977, imm., F. Andrews & A. Hardy (CDFA); Eu- reka Valley, pitfall traps, November-December 1977, 26, February 1978, 1$, April 1978, 1$, D. Giulani, A.R. Hardy & EG. Andrews (CDFA); N. Eureka Valley, Inyo Mountains, Willow Springs Canyon, 3000-3600 feet, 29 September 1980-18 March 1981, 46; 6 mi E. Independence, 4600 feet, 6 December 1984-20 December 1986, 1$; White Mountains, 5000 feet, 6 mi NE Big Pine, 25 April- 22 July 1982, 1$; 3 mi SW Big Pine, in pitfall, 6 October 1985-13 May 1986, 16; 1 mi W Big Pine, 4100 feet, October 1985-May 1986, 26, D. Giuli- ani (CAS); Death Valley National Monument, Scot- ty's Ranch at Travertine Springs, 2500 feet, 13 Jan- uary 1981, 1$, V. Roth (CAS); China Lake, Mt. Springs Canyon, 4500 feet, crawling on sand dune at eight, 10 October 1997, 16, G. Pratt (UCR). Mono County: 9 mi N Bishop, Fish Slough, 4200 feet, sand dunes, 9 June-9 August 1987, 1$, D. Giuliani (CAS). Riverside County: Joshua Tree Na- tional Monument, pitfall traps. Fried Liver Wash, 30 October 1965, 16; Quail Guzzler, 29-30 Octo- ber 1965, 26; Piny on Wells, 11 November 1965, 16; 0.7 mi S. Squaw Tank, no date, 16; Pleasant Valley, 2 December 1966, 16; 30 October 1968, 16, E.L. Sleeper, S.L. Jenkins (JLO); Boyd Desert Research Center, Coyote Creek, 3.5 mi S Palm De- sert, pitfall trap, 10 May 1975, 1$, W. Icenogle (WRI); Santa Rosa Mountains, Deep Canyon, 0.5 mi S junction Hwy 74 & Piny on Crest turnoff, 3600 feet, in Aphonopelma burrow, 22 July 1976, 1 6 1 $ (collected as immatures, matured late August), W. Icenogle (WRI); Cactus City, 10 mi W Chiriaco Summit off I- 10, 1300 feet, in pitfall trap in wash, 29 April 1999, 1$, 18 December 1999, 161$, R. Vetter (UCR). San Bernardino County: Yermo, 28 October 1939, 16 (holotype), W.M. Pearce (AMNH); Fort Irwin, Avawatz Mountains, 6150 feet, 22 May~17 June 1996, 1$, G. Pratt & C. Pierce (UCR), 4250 feet, 26 May 1997, 1 imm., G. Pratt, W. Savary & D. Ubick (CAS); Pisgah Lava Flats, 24 May 1960, 16, B. Banta (CAS); Pisgah Crater, 11 February 1961, 2$; 11 March 1961, 1$; 12 April 1961, 1$; 11 November 1961, 261$; 19 November 1962, 16, Norris & Heath (AMNH). Ne- vada: Nye County: Nevada Test Site (see below), 6267$, 23 imm., Allred et aL(AMNH); Monitor Summit, 3 mi N, 17 E Tonopah, 6400 feet, March- October 1982, 16, D. Giuliani (CAS). Utah: Wash- ington County: 10 mi N St. George, 21 July 1952, 1 $ , M. Cazier, W. Gertsch, Schrammel (AMNH). Additional locales presented in Fig. 10 are listed in Allred & Gertsch (1976) and Allred & Kaston (1983) but material was not examined. Utah: Kane County: Nipple Bench, Smoky Mountain, Ahlstrom Point. NEVADA TEST SITE In 1960-1961 a comprehensive faunal sur- vey was undertaken (Allred et al. 1963a) to inventory the animal diversity at the Nevada Test Site where atomic bombs were detonated in the 1940’s. Twenty different collection techniques were employed; the fauna collect- ed consisted of invertebrates (insects, arach- nids, chilopods, millipedes) and vertebrates (reptiles, birds, rodents, carnivores, rabbits, artiodactyls); sampling occurred year-round. The survey encompassed about 3367 km^ with exhaustive collection arrays differentiated by the dominant plant species. Inventory results are reported in Allred et al. (1963a) while the cryptic locale data (e.g., each spider label had Mercury, Nevada as the collection locale with a designation such as “4AA5C”) was decod- ed by using the depository amendment (Allred et al. 1963b). 8 THE JOURNAL OF ARACHNOLOGY Figure 1 1 . — Seasonal collection phenology of Neoanagraphis chamberlini (o) and N. pearcei (•) at the Nevada Test Site, Nye County, Nevada, 1960-1961. In the course of this revision, 62% (123 of 200) of the mature males and 61% (193 of 319) of all spiders examined were collected by Allred et al. (1963a) at NTS. Both species of Neoanagraphis were collected at NTS; specimens deposited at AMNH were about equally divided between the two species. Al- most all Neoanagraphis spiders were collect- ed in pitfall traps which explains the prepon- derance of males, most probably as they wandered in search of females. This affords a rather rare opportunity to examine character- istics between the two almost sympatric pop- ulations to compare traits where environmen- tal conditions would be fairly similar. Despite the fact that both species of Neoan- agraphis spiders were collected within the NTS, there are striking differences between them. The temporal collection profiles show little overlap in phenology. Mature males of N. chamberlini were collected most often from mid-August until late September, and mature males of N. pearcei were collected from mid-September to late October (Fig. 11). (Specimens collected outside of NTS corrob- orate this phenology.) Despite being conge- nerics, there was an almost dichotomous sep- aration in size as measured by the cephalothorax length and width of mature qO ^oo 1 2 3 4 5 cephalothorax length (mm) Figure 12. — Comparison of cephalothorax lengths and widths of mature male Neoanagraphis chamberlini (o) and N. pearcei (•) from the Nevada Test Site, Nye County, Nevada, 1960-1961. {n — 58 for each species). males; only two very small N. chamberlini overlapped with the size range of N. pearcei (Fig. 12). (Specimens collected outside of NTS show similar patterns but with greater overlap (data not shown)). The two species also showed habitat differences in that N. chamberlini preferred the flat terrain of Yucca Flats which consisted of communities of Co- leogyne, Grayia-Lycium, Atriplex-Kochia and thistle (i.e., tumbleweed, Salsola kali). In comparison, N. pearcei was found most often 15 km southward in the more montane sec- tions of the NTS with a plant community strictly of creosote {Larrea divaricata) and Franseria sp. Determination of immatures with spina- tion. — Because mature specimens of the two Neoanagraphis species are readily differenti- ated by spination pattern and preferred dis- tinctly different habitats at NTS, the NTS im- matures were also examined to see if anterior tibia spination could be correlated with locale. Immature anterior ventral tibial spine pattern and cephalothorax length was recorded. Spi- nation pattern was correlated to the coded lo- cale data. Immature spiders separated almost dichot- omously into two groups with either 2 pairs or 3 pairs of anterior ventral tibial spines; there were a few immatures with 5 spines which were placed in the 3-pairs-of-spines group because spiders are more likely to add spines with age rather than lose them. (From VETTER— THE SPIDER GENUS NEOANAGRAPHIS 9 previous examination of mature specimens if there were supernumerary spines on N. cham- berlini it usually was double the normal pat- tern, that is, 4 pairs of ventral tibial spines). When immature spination type was matched with locale data, the 2-pair-of-spines imma- tures (cephalothorax length: 0.93-3.67 mm) were found almost entirely in the flatland re- gion (35 of 36) where N. chamberlini adults were most often found. The 3-pairs-of-spines immatures (cephalothorax length: 0.94-2.57 mm) were found almost exclusively in the mountain ranges (22 of 23) where adult N. pearcei was predominant. With this evidence, it is reasonable to state that the spination aspect from the key above will successfully determine both species of Neoanagraphis spiders even as immatures. The generic characteristic of elongate tarsal claws III and IV is evident in the smallest spi- ders examined here. Therefore, even if one has disarticulated limbs and can verify a spi- derling as Neoanagraphis, one can then iden- tify it to species (unless additional species oc- cur) by finding an anterior leg (which has a less elongate claw than the posterior legs) and examine spination pattern on the ventral tibial surface because it is the same for leg I and IT Familial reassignment. — Specimens sent to AMNH were examined microscopically and with a scanning electron microscope. Dr. Norman Platnick has transferred the genus Neoanagraphis to the Liocranidae on the basis that the female has three cylindrical gland spigots on each posterior median spinneret and two on each posterior lateral spinneret (N.I. Platnick pers. comm.). Cylindrical (or tubuliform) glands, used in construction of eggsac silk, are lacking in the clubionids (Ko- voor 1987). In addition, the male palpal struc- ture of Neoanagraphis corresponds well with that of the liocranid genus Agroeca (N.I. Plat- nick pers. comm.) and other liocranids (J. Bosselaers pers. comm.). Current keys. — The spider genus Neoan- agraphis does not appear in any edition of Kaston’s basic spider keys. How to Know the Spiders (Kaston 1953, 1972, 1978). In Roth (1993), under the Clubionidae, N. pearcei will properly key out to Group IV and then further to the Neoanagraphis couplet. In contrast, N. chamberlini does not key out correctly. At couplet 10, it gets shunted to Group III on the basis of its two pairs of ventral macrosetae. Continuing through the Group III key, it will terminate at the Agroeca couplet or not key out at all depending upon one’s degree of dif- ferentiation. ACKNOWLEDGMENTS This paper is dedicated to the memory of Vince Roth with whom I only became ac- quainted too late in life. I thank Dr. Norman Platnick (AMNH) for serving as a patient mentor throughout the course of this study and for making the familial reassignment and Dr. Jan Bosselaers (Musee Royal de TAfrique Centrale, Tervuren, Belgium) for discussion regarding liocranid spiders. W Sakai (Santa Monica College, Santa Monica, California) deserves thanks because his collecting of Neoanagraphis spiders was the catalyst for this study. In addition to those listed above who loaned material. Dr. G. Pratt, C. Pierce (UCR) and Dr. S. Brantley (University of New Mexico) provided additional specimens dur- ing the course of the study. Dr. C. Luke (Sweeney Granite Mountains Desert Research Center, Mojave Desert) assisted by allowing me to deploy pitfall traps at the Granite Mountains reserve, fruitless as they were. Dr. C. Griswold, D. Ubick (CAS) and Dr. H.D. Cameron (Univ. Michigan) offered advice and information which was greatly appreciated. This study was funded by the Theodore Roo- sevelt Memorial Fund (AMNH) and Humbug Mountain Engineering Services P-62. LITERATURE CITED Allred, D.M. & W.J. Gertsch. 1976. Spiders and scorpions from northern Arizona and southern Utah. Journal of Arachnology 3:87-99. Allred, D.M. & B.J. Kaston. 1983. A list of Utah spiders with their localities. Great Basin Natu- ralist 43:494-522. Allred, D.M., D.E. Beck & C.D. Jorgensen. 1963a. Biotic communities of the Nevada test site. Brigham Young University Science Bulletin, Bi- ology Series 2(2): 1-52. Allred, D.M., D.E. Beck & C.D. Jorgensen. 1963b. Nevada test site study areas and specimen de- positories. Brigham Young University Science Bulletin, Biology Series 2(4): 1-15. Banks, N. 1900. Some new North American spi- ders. Canadian Entomologist 32:96-102. Bonnet, P. 1958. Bibliographia Araneorum. Tome II, 4"^® partie: N-S, Douladoure, Toulouse, Pp. 3027-4230. Brignoli, PM. 1983. A catalogue of the Araneae 10 THE JOURNAL OF ARACHNOLOGY described between 1940 and 1981. Manchester Univ. Press, Manchester, United Kingdom. Comstock, J.H. 1948. The Spider Book. Cornell Univ. Press, Ithaca, New York. Gertsch, W.J. 1941. New American spiders of the family Clubionidae. I. American Museum Nov- itates #1147, 20 pp. Gertsch, W.J, & S. Mulaik. 1936. Diagnoses of new southern spiders. American Museum Novi- tates #851, 21 pp. Jung, A.K.S. & V.D. Roth. 1974. Spiders of the Chiricahua mountain area, Cochise Co., Arizona. Journal of the Arizona Academy of Science 9: 29-34. Kaston, B.J. 1953. How To Know The Spiders. Wm. C. Brown Company Publishers, Dubuque, Iowa. Kaston, B.J. 1972, How To Know The Spiders. 2nd ed. Wm. C. Brown Company Publishers, Dubuque, Iowa. Kaston, B.J. 1978. How To Know The Spiders. 3rd ed. Wm. C. Brown Company Publishers, Du- buque, Iowa. Kovoor, J, 1987. Comparative structure and histo- chemistry of silk-producing organs in arachnids. Pp. 160-186. In Ecophysiology of Spiders. (W. Nentwig, ed). Springer- Verlag, Berlin. Lehtinen, P.T. 1967. Classification of the cribellate spiders and some allied families, with notes on the evolution of the suborder Araneomorpha. Annales Zoologici Fennici 4:199-468. Platnick, N.I. 1993. Advances in Spider Taxonomy 1988-1991 With Synonymies and Transfers 1940-1980. New York Entomological Society, New York. Platnick, N.I. 1998. Advances in Spider Taxonomy 1992-1995 with redescriptions 1940-1980. New York Entomological Society, New York. Roewer, C.F. 1955. Katalog der Araneae, 2 Band, Abt, B. Institut Royal des Sciences Naturelles de Belgique, Brussels. Pp. 1-923. Roth, V.D. 1993. Spider Genera of North America. 3rd ed. American Arachnological Society, Gainesville, Florida. Ryckman, R.E. & R.D. Lee. 1956. Spiders and phalangids associated with mammals (Citellus and Neotoma) in southwestern United States and northern Mexico. Annals of the Entomological Society of America 49:406—409. Manuscript received 11 February 2000, revised I September 2000. 2001. The Journal of Arachnology 29:11-15 NOTES ON THE GENUS SYBOTA WITH A DESCRIPTION OF A NEW SPECIES FROM ARGENTINA (ARANEAE, ULOBORIDAE) Cristian J. Grismado: Museo Argentine de Ciencias Naturales “Bernardino Rivadavia,” Av. Angel Gallardo 470 (1405) Buenos Aires, Argentina ABSTRACT. Sybota atlantica new species is described from the Atlantic coast of Buenos Aires Prov- ince, Argentina. The morphology of genitalia and carapace suggests that the new species forms a mono- phyletic group with S. mendozae Opell 1979 and S. rana (Mello-Leitao 1941). The female genitalia of the genus shows an unusual grade of entelegyny, with copulatory and fertilization ducts leading to a common tube. Keywords: Uioboridae, Sybota, taxonomy, Argentina RESUMEN. Sybota atlantica nueva especie es descripta para la costa atlantica de la provincia de Buenos Aires, Argentina. La morfologia genital y cefalica sugiere que la nueva especie forma un grupo monofi- letico con S. mendozae Opell 1979 y S. rana (Mello-Leitao 1941). Los organos genitales femeninos muestran un inusual grado de enteleginia, con los conductos de copulacion y fertilizacion convergiendo en un tubo comun. The genera of the family Uioboridae and their Neotropical species were revised by Opell (1979). In that work, he defined the ge- nus Sybota Simon 1892 and included three species: S. abdominalis (Nicolet 1849) and S. osornis Opell 1979 from Chile, and S. men- dozae Opell 1979 from western Argentina. Fe- males of the genus share with those of Pole- necia Lehtinen 1967 an abdominal projection extending beyond the spinnerets (Figs. 1, 3; Opell 1979: figs. 51, 102, 110, 116). Never- theless, this feature does not reflect a close relationship between both genera. According to Coddington (1990), Sybota is the sister group of the clade Orinomana {Hyptiotes + Miagrammopes), all united by having the pos- terior lateral eyes on conspicuous tubercles. Sybota males have a well-developed conduc- tor and a median apophysis with two or three projections (Figs. 5-7; Opell 1979: figs. 6A, B). In the present paper I describe a new spe- cies, Sybota atlantica, from specimens col- lected in the coast of Buenos Aires Province (courtesy of Martin J. Ramirez, MACN), which seems to be closely related with S. men- dozae because some cephalic and genitalic features (see discussion). Here I also rede- scribe the holotype of S. rana (Mello-Leitao 1941) from Salta province, a species not in- cluded in the OpelTs revision, and describe details of its genitalia, an aspect omitted in the original description (Fig. 11). Although S. rana is known from only a poorly-preserved specimen, apparently collected during the molting process, it can be placed close to the other two species. The homology of the tegular sclerites of the male palps of the Uioboridae is still unclear. Coddington (1990) suggested that the terms median apophysis and conductor, as identified by Opell, should be switched. Nevertheless I maintained OpelTs names only to ease com- parison with previously described species. METHODS Specimens are deposited in the following institutions: Museo Argentino de Ciencias Na- turales “Bernardino Rivadavia,” Buenos Ai- res (MACN, Cristina L. Scioscia), Museo de La Plata (MLP, Luis Pereira), and Institute Ar- gentino de Investigaciones de las Zonas Ari- das, Mendoza (lADIZA, Sergio Roig Junent). The format of descriptions follows Opell (1979). The abbreviations are: C = conductor; CD = copulatory duct; CO = copulatory opening; CY = cymbium; E = embolus; FD = fertilization duct; MA = median apophysis; PP = posterior plate; S = spermathecae; ST = subtegulum; T = tegulum. Abbreviations 11 12 THE JOURNAL OF ARACHNOLOGY Figures 1-4. — Sybota atlantica new species. 1. Female, dorsal view; 2. Male, dorsal view; 3. Female, lateral view; 4. Male, lateral view (palps omitted). Scale bars = 1 mm. for eyes are standard for the Araneae. The fe- male genitalia were cleared with clove oil and observed with compound microscope. Mea- surements are expressed in millimeters. Sybota atlantica new species Figs. 1-10 Types. — Male holotype, and four female paratypes from Argentina, Buenos Aires Prov- ince, Mar del Tuyu, 2 May 1981, M.J. Ra- mirez (MACN No. 9639, 9640 and 9641, re- spectively). Etymology. — The specific name refers to the type locality, on the Atlantic Coast of Ar- gentina. Diagnosis. — Males differ from those of S. abdominalis and S. osornis by having a longer embolus and conductor (Figs. 5-7), and by having the AMEs on a conspicuous tubercle (Figs. 2, 4). Females resemble those of S. mendozae and S. rana by having an elongate carapace, the AMEs on a tubercle, and the convoluted copulatory ducts, but differ by the shape of the epigynum and spermathecae (Figs. 8-10). Description. — Male (holotype): Total length 4.76, carapace length 1.72, sternum length 1.08. Leg I: femur length 3.32, tibia length 3.04, metatarsus length 3.56, tarsus length 0.96. Carapace brown with yellowish median area between median eyes and fovea, margins with diffuse dark dots, more apparent on anterior region. Eyes bordered by dark rings. Sternum dark brown with a reddish me- dian stripe. Legs same color as carapace but with tenuous, darker, longitudinal dorsal bands. Abdomen dorsally whitish with a gray longitudinal band (Fig. 2). Sides of the ab- domen with diffuse longitudinal bands (Fig. 4). Venter pale reddish with a dark central band between pedicel and spinnerets. Palp: Femur with an excavated area where the bulb presumably fits (Fig. 5), tibia with a prolateral translucent prolongation (Fig. 7) covering par- tially the base of cymbium, which has a retro- lateral basal tubercle (Fig. 5, arrow). Copula- tory bulb: retrolateral surface of tegulum with a translucent membrane (Fig. 5); median apophysis with one basal and three distal pro- jections (Figs. 5, 6); conductor long with two prongs: the proximal digitiform and the ter- minal flattened; embolus long, with tip fitting into the terminal prong of conductor. GRISMADO— NEW SYBOTA FROM ARGENTINA 13 Figures 5“- 1 Genitalia of Sybota. 5-10. Sybota atlamtica new species. 5. Left male palp, retrolateral (arrow: cymbial tubercle; asterisk: tegular membrane); 6. Same, ventral; 7. Same, prolateral; 8. Epigynum, ventral view; 9. Same, posterior view; 10. Same, cleared, dorsal view. 11. Sybota rana (Mello-Leitao), cleared epigynum, dorsal view. Scale bars = 0.2 mm. Female (paratype): Total length 7,35, car- apace length 2.00, sternum length 1.32, Leg I: femur length 3.20, tibia length 2.60, metatar- sus length 2,96, tarsus length 0.80. Color: Carapace, legs and eyes as in male, but AME tubercle less pronounced; sternum as in male, but with a stripe restricted to the anterior half. Dorsum of abdomen yellowish with a gray longitudinal band, wider anteriorly and diffuse dark spots, more evident in caudal and lateral 14 THE JOURNAL OF ARACHNOLOGY areas; dorsal and dorsolateral surfaces with aligned bundles of long setae (Figs. 1, 3). Venter yellowish with a brown median stripe between epigastric furrow and spinnerets. Epi- gynum: Lateral lobes flattened with a wide posterior notch (Fig. 8), copulatory openings under two elevated anterolateral margins (Fig. 9). Copulatory and fertilization ducts leading to a common convoluted tube. Natural history. — The specimens were collected in typical uloborid horizontal orb- webs on shrubs and other medium-sized plants near the sandy ground in Mar del Tuyu. The spiders rested with legs I and II extended anteriorly (Martin J. Ramirez pers. comm.). Material examined. — Only the type series. Sybota mendozae Opell 1979 Sybota mendozae Opell 1979: 496 (female holotype and three female paratypes from 7 km W of Men- doza, Argentina, collected in “chaparral” at an elevation of 1200 m, March- April 1958, B. Pat- terson col., in MCZ and AMNH, not examined.) New record. — ARGENTINA: Mendoza, Divisadero Largo, 8 March 1993, Debandi and S. Roig col., 1 penultimate female (lA- DIZA). Note: Although this specimen is sub- adult, the internal genitalia are developed and almost identical to those illustrated by Opell (1979). Sybota rana (Mello-Leitao 1941) Fig. 11 Uloborus rana Mello Leitao 1941: 111 (holotype N° 14635 from Coronel Moldes, Salta, Argentina, in MLP, examined). Roewer 1954: 1344. Sybota rana Lehtinen 1967: 266. Diagnosis. — The female resembles those S. mendozae and S. atlantica by cephalic mor- phology and by the long copulatory ducts, but are distinguished by the reniform spermathe- cae (Fig. 11) and the dorsal design of abdo- men. Description.”— (holotype, poorly preserved): Carapace length, ca. 1.46; abdo- men length, 3.96; leg I, femur length 2.34, tibia length 1.74, metatarsus length 2.00, tar- sus length 0.74. Color: carapace dark brown; legs same color but with light longitudinal ar- eas; chelicerae lighter than carapace. Abdo- men (Mello-Leitao 1941, fig. 10) light brown with a dorsal longitudinal dark band (wider in front), and two large dorsolateral spots. The caudal parallel lines figured by Mello-Leitao are no longer evident, probably faded. Epi- gynum: The poor condition of the specimen makes it impossible to distinguish the main epigynal structures; internally, only the reni- form spermathecae and the distal portion of copulatory ducts remain; the preserved por- tion of them suggests that they were long (Fig. 11). Material examined. — Only the holotype. DISCUSSION Sybota atlantica, S. mendozae and S. rana differ from the Chilean species S. abdominalis and S. osornis by the longer carapace, with the anterior median eyes on a prominent tu- bercle, by the epigynum with a posterior notch, and by the smaller spermathecae, with long and convoluted copulatory ducts (Figs. 1, 8-11; Opell 1979: figs. 115-119). Given that these conditions are not present in other closely related uloborid genera, they seem to be synapomorphies of the three Argentine species. If long copulatory ducts are function- ally correlated with long embolus, the males of S. mendozae and S. rana, which are still unknown, should also have a long embolus. Although Ponella Opell 1979, some Zosis Walckenaer 1837 and some Philoponella Mel- lo-Leitao 1917 (genera which are not closely related with Sybota) also have long and con- voluted copulatory ducts (Muma & Gertsch 1964; Opell 1979, 1981), they differ from Sy- bota by being entelegynes (i.e., the copulatory ducts and fertilization ducts are separated), while Sybota presents an intermediate and pe- culiar grade of entelegyny: the fertilization ducts arise from the proximal part of the cop- ulatory ducts, without a direct conection with spermathecae. As noted by Opell (1983), the Uloboridae show a great diversity in genital features and comprises members both haplo- gyne, entelegyne and some intermediate types. The observation of the web of S. atlantica in the field, and the photograph of an unde- terminated Chilean specimen showed in Figs. 12 and 13 (in American Museum of Natural History, not examined), confirm that these spi- ders rest with legs I and II anteriorly extended, and that they construct typical horizontal orb- webs, as mentioned by Opell (1984) based on a juvenile specimen photographed by Norman Platnick. GRISMADO— NEW SYBOTA FROM ARGENTINA 15 Figures 12-13. — Sybota sp. from Alto de Vilches, Talca, VIII Region, Chile. 12. Web; 13. Living specimen (photographs by Martin J. Ramirez). ACKNOWLEDGMENTS I am greatly endebted to the institutions and curators for loaning the specimens; to Martin J. Ramirez, who collected the specimens stud- ied here, brought them to my attention, and provided the photographs of a Chilean speci- men; and to Jonathan A. Coddington (Nation- al Museum of Natural History, Smithsonian Institution, Washington, D.C.), Brent D. Opell (Virginia Polytechnic Institute and State Uni- versity, Blacksburg), Martin J. Ramirez (MACN) and two anonymous reviewers for helpful comments on a draft of the manu- script. LITERATURE CITED Coddington, J.A, 1990. Ontogeny and homology in the male palpus of orb-weaving spiders and their relatives, with comments on phylogeny (Ar- aneoclada: Araneoidea, Deinopoidea). Smithson- ian Contributions to Zoology 496:1-52, Lehtinen, P.T. 1967. Classification of the cribellate spiders and some allied families, with notes on the evolution of the suborder Araneomorpha. Annales Zoologica Fennici 4:199-468. Mello-Leitao, C.F. 1941. Las aranas de Cordoba, La Rioja, Catamarca, Tucuman, Salta y Jujuy, colectadas por los profesores Biraben. Revista del Museo de la Plata, Seccion Zoologia 2:12, 99-198. Muma, M.M. & W. Gertsch. 1964. The spider fam- ily Uloboridae in North America north of Mex- ico. American Museum Novitates 2196:1-43. Nicolet, H. 1849. Aracnidos. In Historia fis. y pol- ft. de Chile. (C. Gay, ed.). Zoologia 111:319-543. Opell, B.D. 1979. Revision of the genera and trop- ical American species of the spider family Ulo- boridae. Bulletin of the Museum of Comparative Zoology 148:443-549. Opell, B.D. 1981. New Central and South Ameri- can Uloboridae (Arachnida, Araneae). Bulletin of the American Museum of Natural History 170: 219-228. Opell, B.D. 1983. The female genitalia of Hypti- otes cavatus (Araneae: Uloboridae). Transactions of the American Microscopical Society 102:97- 104. Opell, B.D. 1984. Comparison of carapace features in the family Uloboridae (Araneae). Journal of Arachnology 12:105-114. Simon, E. 1892, Histoire Naturelle des Araignees, I. Paris. Manuscript received 1 February 2000, revised 1 July 2000. 2001. The Journal of Arachnology 29:16-20 OKILEUCAUGE SASAKII, A NEW GENUS AND SPECIES OF SPIDER FROM OKINAWAJIMA ISLAND, SOUTHWEST JAPAN (ARANEAE, TETRAGNATHIDAE) Akio Tanikawa: Shichirigahama Senior High School, 2-3-1, Shichirigahama-higashi, Kamakura- shi, Kanagawa, 248-0025 Japan ABSTRACT. Several specimens of a unique spider were collected in Okinawajima Island, Southwestern Japan. They resemble the spiders of the genus Leucauge and its related genera. Phylogenetic analysis was performed to clarify the taxonomic position of the spider, which showed that the focal spider is a sister of a monophyletic group consisting of Tylorida and Mesida. Therefore it is described as a new genus and species under the name Okileucauge sasakii new species. This species lacks the rows of trichobothria on femur IV, which is the synapomorphy of the genera Tylorida and Mesida. Keywords: Okileucauge sasakii, new genus, new species, Tetragnathidae In the spring of 1997, I collected several specimens of a unique spider in Okinawajima Island, Southwestern Japan. This spider, here- after called Okinawa spider, looked like a member of the genus Leucauge because it was hanging at the center of a horizontal orb-web and had a silver colored abdomen. The fea- tures of the male palpal organ, female epigy- num and internal genitalia as well as general appearance show that the Okinawa spider is related to the genus Leucauge. However, the Okinawa spider lacks the rows of trichoboth- ria on femur IV which is a conspicuous fea- ture of the genus Leucauge and its related genera Tylorida and Mesida. On the other hand, American spiders of the genus Metabus, another related genus, also lack the rows of trichobothria on femur IV. I performed a phy- logenetic analysis to clarify the relationship among these spiders. The cladogram shows that the Okinawa spider is a sister of a mono- phyletic group consisting of the genera Tylo- rida and Mesida. The cladogram also shows that Leucauge is a sister of the clade Okinawa + Tylorida + Mesida and that Metabus is a sister of the clade Okinawa + Tylorida + Mesida + Leucauge. These results led me to describe the new genus, though it is monotyp- ic. All the type specimens designated in this paper are deposited in the collection of the Zoological Department of National Science Museum, Tokyo (NSMT). PHYLOGENETIC ANALYSIS Methods. — Taxa used in the analysis: Spi- der in question from Okinawajima Island, Ty- lorida striata (Thorell 1877), Mesida sp, from Taiwan, M. argentiopunctata (Rainbow 1916), Metabus gravidus (O. Pickard-Cambridge 1899), Leucauge subblanda Bosenberg & Strand 1906, L. argentina (Hasselt 1882), L. granulata (Walckenaer 1841), L. sp. from New Guinea Island, Metleucauge chikunii Tanikawa 1992, Meta nigridorsalis Tanikawa 1994, M. reticuroides Yaginuma 1958, and Nephila clavata L. Koch 1878. Nephila clavata was used as an out group judging from the cladogram made by Hormiga et al. (1995). Due to lack of the male speci- men of M. argentiopunctata, the male char- acters of the species were judged from the fig- ures made by Davies (1987) as far as possible. Characters and character states: 1. Tricho- bothria on femur IV of female: none (0); less than 10 pairs (1); 10 pairs and over (2). 2. Depth of the thoracic groove of female: shal- low, bottom visible from above (0); deep, bot- tom invisible from above (1). 3. Cheliceral teeth on posterior margin of fang furrow of female: 4 (0); 5 (1). 4. Booklung cover of fe- male: partly grooved (0); smooth (1). 5. Color of abdomen of female: without silver color (0) ; with silver color (1). 6. Seminal recepta- cles of female: sclerotized (0); not sclerotized (1) . 7. Clypeus height: smaller than one AME diameter (0); equal or larger than one AME 16 TANl¥JK.W A— OKILEUCAUGE SASAKII FROM JAPAN 17 Figure 1.^ — Minimal length cladogram for the data matrix in Table 1. Length: 39; consistency in- dex: 0.615; retention index: 0.712; rescaled consis- tency index: 0.438. diameter. 8. Cheliceral size of the male versus that of the female: same (0); larger (1); small- er (2). 9. Large tooth or modified tooth on fang furrow of male chelicera: absent (0); pre- sent (1). 10. Spur on anterior surface of male chelicera: absent (0); present (1). 11. Projec- tion of cymbium of male palp other than par- acymbium: absent (0); present (1). 12. Lateral eyes of the male: separate (0); touching (1). 13. Course of reservoir within tegulum of male palp in ventral view: without switchback (0) ; with switchback (1). 14. Conductor of male palp: well sclerotized, black or dark brown (0); less sclerotized, almost colorless (1) . 15. Conductor wraps embolus: absent (0); present (1). 16. Metine embolic apophysis: ab- sent (0); present (1). 17. Paracymbium of male palp: small and finger shaped (0); small and flattened (1); large and modified (2). 18. Ma- crosetae on patella of male palp: 2 (0); 1 (1); none (2). 19. Length of tibia of male palp: short, tibia/patella less than 1.2 (0); long, tib- ia/patella 1.2 to 2.0 (1); very long, tibia/patella more than 2.0 (2). 20. Epigynum: well scler- otized (0); weakly sclerotized (1). The data matrix is shown in Table 1. When Hormiga et al. (1995) made a phy- logenetic analysis of tetragnathid spiders, 60 characters were used. Of these, 10 characters were also used in the present study (1, 4, 7, 8, 12, 13, 15, 16, 17, 18). The remaining char- Figure 2. — Okileucauge sasakii new species, fe- male on a leaf. acters were not used because no data were available on the specimens used in the present study or they were uninformative. Analysis: I used PAUP version. 3.1.1 (Swofford 1993) for the analysis. I used the branch and bound search method. I chose the ACCTRAN, accelerated transformation, for the character optimization. Multistate charac- ters were treated as unordered. Results.— As a result I obtained the clad- ogram shown in Fig. 1 (tree length 39; con- sistency index 0.615; retention index 0.712; rescaled consistency index 0.438). The clad- ogram shows that Tylorida is a sister of Mes- ida, the Okinawa spider is a sister of Tylorida -f Mesida, and Leucauge is a sister of Oki- nawa + Tylorida + Mesida, and Metabus is a sister of Okinawa + Tylorida + Mesida + Leucauge. If the Okinawa spider were placed in any of these genera, that genus would not be a monophyletic group. Thus, I conclude that the new genus should be described for this spider. DESCRIPTION Family Tetragnathidae Genus Okileucauge new genus Type species. — Okileucauge sasakii new species by monotypy. Diagnosis. — Okileucauge is a sister of the group consisting of genera Tylorida and Mes- ida. The synapomorphy of the latter group is the presence of rows of trichobothria on femur IV. That is, Okileucauge can be separated from the genera Tylorida and Mesida by the absence of rows of trichobothria on femur IV. The group Okileucauge + Tylorida + Mesida is a sister of Leucauge. The synapomorphies Table 1. — Data matrix (Ok: spider in question from Okinawajima, Ty: Tylorida striata, Mel: Mesida sp., Me2: Mesida argentiopunctata, Mb: Metabus gravidus. Lei: Leucauge subblanda, Le2: Leucauge argentina, Le3: Leucauge granulata, Le4: Leucauge sp.. Ml: Metleucauge chikunii, Mtl: Meta nigridor- salis, Mt2: Meta leticuloides, Ne: Nephila vlavata). 18 THE JOURNAL OF ARACHNOLOGY U o o o o o O (N O O O ^ o ^ d ooinoomooiniooot^o d-!d-5dd-H’^dd— iddd oooooooooooi^oooooocno 0'00ppp'^^p>npoo'0pp»oir)00moo — !d— H’d-H^d^d-2dd-2— ^dd-^ddd U W 0) 2: (N I § § omoooouooooooooooor-oo Ofno>nOOfNiOtnOioiooo»o»r)0^mio fn(N(NC^(NCS(NCN(NCN(N(N(N(N(N(Ncnenm(N OOOOOOOOOOOO-hO-hO-^OOO 000-h000-h^00-h000-h(N^-.^ OOO-HOOO-^’-^OOOOOO^CMf^-^'-' O— 'OOO-^-^OOO-hO-^OO-^O— 'CNO CN-h0-h-h^,-.0000'^-h-^-h00-^-h0 00000-H,-<^^00f^-^0 'CNO 0^0’-^>^e"-0000'-^0’-H-H^00'^^0 ^cNen'^iO'Oi>ooaNO’-H(Nim'^»r>vor'OoaNO TAmKASN \—OKILEUCAUGE SASAKII FROM JAPAN Figures. 3-8. — 3, Female carapace and abdomen, dorsal view (holotype: NSMT — Ar 4301); 4, Male carapace and abdomen, dorsal view (paratype: NSMT — Ar 4305); 5, Male left palp, lateral view (paratype: NSMT — Ar 4305); 6, Epigynum (holotype: NSMT — . seminal receptacle expanded. (Scales: 0.25mm.) of the former group are 1) shallow thoracic groove of female, 2) weakly sclerotized epi- gynum. The genera, Okileucauge, Tylorida, Mesida, Leucauge, and Metabus make a monophyletic group. The synapomorphies of the group are 1) abdomen having silver color; 2) conductor of male palp being weakly scler- otized, 3) conductor wraps embolus, 4) male palp lacks metine embolic apophysis. Description.— Carapace longer than wide, median fovea shallow or bottom visible from above. Median ocular area almost as long as wide; slightly narrower in front than behind. Female chelicera with 3 promarginal and 4 re- tromarginal teeth on fang furrow; male che- licera with a big tooth at the innermost part of posterior margin of fang furrow. Male palp: course of reservoir within the tegulum switch- 4301); 7, Female genitalia, dorsal view; 8, Same, backed; weakly sclerotized conductor wraps embolus. Labium wider than long. Sternum almost as long as wide. Abdomen longer than wide, with silver scales. Seminal receptacle not sclerotized. Booklung cover smooth. Etymology. — Generic name is a coined word made from Okinawa, native island of the type species, and Leucauge. The name is fem- inine. Okileucauge sasakii new species (Figs. 2-8) Specimens examined. — Type series: Ho- lotype female, Kunigami-son, Okinawajima Island, Okinawa Pref., Japan, 1 April 1997, A. Tanikawa leg. (NSMT — Ar 4301). Paratypes: 2?, same data except 30 March 1997 (NSMT— Ar 4302-4303), 1$ Id, same data 20 THE JOURNAL OF ARACHNOLOGY except 1 April 1997 (NSMT— Ar 4304- 4305), 2$, same data except 2 April 1997 (NSMT— Ar 4306). Other specimens examined: 3 9 , Kunigami- son, Okinawa) ima Is., Okinawa Pref., Japan, 1 April 1997, A. Tanikawa leg. 1 9, same data except 2 April 1997. Description. — [Based on the female holo- type and the male paratype; variations among the specimens examined are given in the pa- rentheses.] Measurement (in mm): Total length $ 3.10 (2.77-3.10), S 2.18; carapace length $ 1.15 (1.08-1.16), d 1.07; width $ 0.94 (0.92-0.96), S 0.88; abdomen length $ 2.16(1.61-2.16), d 1.18, width $ 1.64(1.27- 1.64), S 0.90. Length of legs (tarsus + meta- tarsus + tibia + patella + femur = total): $ holotype, I, 0.78 + 2.48 + 1.98 + 0.56 + 2.14 - 7.94, II, 0.65 + 1.83 + 1.48 + 0.51 + 1.75 = 6.22, III, 0.38 + 0.75 + 0.56 + 0.31 + 0.90 = 2.90, IV, 0.45 + 1.18 + 0.98 + 0.34 + 1.39 == 4.34. S paratype, I, 0.74 + 2.40 + 2.09 + 0.51 + 2.11 = 7.85, II, 0.59 + 1.66 + L47 + 0.46 + 1.69 - 5.87, III, 0.34 + 0.64 + 0.53 + 0.27 + 0.80 = 2.58, IV, 0.40 + 1.03 + 0.87 + 0.27 + 1.23 = 3.80. Female and male: Carapace length/width $ 1.22 (1.13-1.24), d 1.22. Length of leg 1/ length of carapace 9 0.93 (0.88-1.00), S 0.90. Male palp (Fig. 5): tibia with one ma- croseta; cymbium with a projection other than paracymbium; weakly sclerotized conductor wraps embolus; reservoir in tegulum switch- backed. Abdomen length/width 9 1.31 (1.12- 1.31), 6 1.31. Female genitalia: epigynum simple and weakly sclerotized as in Fig. 6; seminal receptacles not sclerotized (Fig. 8). Coloration and markings in alcohol: Fe- male and male: Carapace yellow. Abdomen silver with black marking as in Figs. 2-4. Range.^ — Japan (Northern part of Okina- wa) ima Island). Remarks. — The present new species looks like a small sized species of the genus Leu- cauge or its related genera Tylorida and Mes- ida in general appearance. However it can be easily separated from the latters by the ab- sence of rows of trichobothria on femur IV. Etymology, — The species is dedicated to Mr. Takeshi Sasaki, University Museum of the Ryukyus, who supported my field research in Okinawa) ima island. ACKNOWLEDGMENTS I wish to express my hearty thanks to Dr. Tadashi Miyashita, University of Tokyo, and Dr. Jonathan A. Coddington, Smithsonian In- stitution, for critical reading of the manuscript of this paper. I am deeply indebted to Dr. H. W. Levi and Ms. L. Leibensperger, Museum of Comparative Zoology, for loaning valuable specimens. My sincere thanks are also due to Mr. Takeshi Sasaki, University Museum of the Ryukyus, for supporting my fieldwork in Oki- nawa) ima Island. This study was partly supported by the Leg- acy Project (Natural Resources inventory on U.S. Marine Corps Bases in Okinawa). LITERATURE CITED Davies, V. T 1988. An illustrated guide to the gen- era of orb-weaving spiders in Australia. Mem. Queensland Mus., 25:273-332. Hormiga, G., W. G. Eberhard & J. A. Coddington. 1995. Web-construction behaviour in Australian Phonognatha and the phylogeny of nephiline and tetragnathid spiders (Araneae, Tetragnathidae). Australian J. ZooL, 43:313-364. Swofford, D. L. 1993. PAUP: Phylogenetic Anal- ysis Using Parsimony, Ver. 3.1.1. Computer pro- gram distributed by Illinois State Natural History Survey, Champaign, Illinois. 2001. The Journal of Arachnology 29:21-41 REVISION DE LAS ESPECIES DE FREYA DEL GRUPO DECORATA (ARANEAE, SALTICIDAE) Maria Elena Galiano': Museo Argentine de Ciencias Naturales “Bernardino Rivadavia,” Av. Angel Gallardo 470, C1405DJR, Buenos Aires, Argentina ABSTRACT. Eight species of the genus Frey a Koch 1846, closely related with the type species Freya decorata (Koch 1846), are revised and redescribed: F. decorata, F. regia (G. & E. Peckham 1896), F. maculatipes (Cambridge 1901), F. nigrotaeniata (Mello-Leitao 1945) new combination, F. rubiginosa (Koch 1846), new combination. The female of F. nigrotaeniata is described for the first time. Three new species are described: F. dureti from Brazil, F. chapare from Bolivia and F. atures from Venezuela. Diagnostic characters for the genus are given. RESUMEN. Ocho especies del genero Freya Koch 1846, estrechamente relacionadas con la especie tipo Freya decorata (Koch 1846), son revisadas y redescriptas: F. decorata (Koch 1846), F. regia (G. & E. Peckham 1896), F. maculatipes (Cambridge 1901), F. nigrotaeniata (Mello-Leitao 1945) nueva com- binacion y F. rubiginosa (Koch 1846) nueva combinacion. Las hembras de F. nigrotaeniata se desciiben por primera vez. Se describen tres especies nuevas: F. dureti de Brasil, F. chapare de Bolivia y F. atures de Venezuela. Se dan caracteres diagnosticos del genero. Keywords^ Salticidae, Freya decorata group, Neotropica C.L. Koch describio en 1846 veintitres es- pecies nuevas de Salticidae en el genero Euophrys Koch 1834 de las cuales dieciseis proceden del area Neotropica. Posteriormente (1850) creo trece nuevos subgeneros entre los que distribuyo estas especies, evidentemente disimiles. De los seis subgeneros que incluyen especies neotropicales, cuatro han sido ele- vados a la categoria de generos: Aphirape, Corythalia, Frigga y Freya (que a su vez in- cluye a Thore como sinonimo) mientras que Trivia se considera sinonimo de Euophrys. Veinte especies han sido descriptas como Freya, mientras que numerosas otras han sido incorporadas, transferidas desde otros gene- ros: cinco fueron descriptas originalmente como Euophrys, quince como Cyrene Peck- ham & Peckham 1893, una como Eustiro- mastix Simon 1902, una como Phiale Koch 1 846, dos como Attus Walckenaer 1 805 y tres como Heraclea G. & E. Peckham 1896. Con posteriores transferencias y sinonimias, el ge- 'The Journal of Arachnology regrets to note that it has learned of the untimely death of Prof Galiano in an accident on 30 October 2000. nero Freya comprende en la actualidad vein- tisiete especies. En este trabajo se trata un grupo de ocho especies, estrechamente relacionadas con Freya decorata (Koch 1846), especie tipo del genero. La identificacion de la especie tipo se ha hecho a lo largo de los anos sobre la base de la descripcion y dibujo originales, ya que no se ha podido hallar el material tipo, pese a que otros especimenes descriptos en la misma epoca por Koch se encuentran en buen estado en el Zoologisches Museum de Berlin. Las especies del grupo decorata reunen las siguientes caracterfsticas: 1) Quehceros verti- cales, unidentados; 2) Palpo con apofisis tibial retrolateral gruesa, tuberosa, con una punta conica dirigida hacia la cara ventral, el apice o la base de la tibia (Figs. 8-13); 3) Embolo conico, relativamente corto, recto o apenas curvado, acompanado por el conductor para- lelo, membranoso, aproximadamente de la misma longitud que el embolo (Figs. 19-31); 4) Epigino: placa limitada anteriormente por un surco curvo, carenado; borde posterior con bolsillos de anclaje; orificios de copulacion re- lativamente pequenos, ubicados en la mitad anterior del epigino; espermatecas esfericas u ovoideas (Figs. 36-57); y 5) Opistosoma en ambos sexos con tres bandas dorsales longi- 21 22 THE JOURNAL OF ARACHNOLOGY Figura 1. — Freya decorata, macho de Brasil, Para, vista dorsal. Escala = 1 mm. tudinales de pelos blancos; hembras con ban- das radiantes oscuras en la region toracica (Fig. 1). Algunos de estos caracteres aislados pueden encontrarse en otras especies de Freya o de otros generos, pero teniendo en cuenta que un genero se define por su especie tipo y que las ocho especies aqui tratadas reunen los carac- teres mencionados, deben considerarse como un grupo monofiletico. Es posible que gran parte de las especies actualmente clasificadas como Freya deban ser excluidas. No existen suficientes estudios ni evidencias que indiquen las relaciones de Freya con otros generos de Salticidae. Solo futuras revisiones de otros grupos podran aportar argumentos para dis- cusiones filogeneticas. El patron de manchas y bandas es notable- mente uniforme entre las especies del grupo aqui tratadas y algunas son simpatridas en el area de distribucion. Las diferencias residen fundamentalmente en los caracteres de los 6r- ganos copuladores, algunos de los cuales no fueron advertidos por autores anteriores, como la presencia de un conductor paralelo al em- bolo. La identificacion que en el presente tra- bajo se hace de Freya decorata se basa en parte en el material de Guayana Francesa y Guyana estudiado por Caporiacco (1948, 1954) y el colectado por Galiano en Brasil, en los Estados de Para (localidad tipo) y Ama- zonas. Algunas especies estan representadas en las colecciones por material relativamente abundante, mientras que de otras se dispone de un unico ejemplar. La identificacion es po- sible cuando se trata de ejemplares machos, pero se hace dificultosa cuando se trata de hembras, debido a la gran uniformidad de los epiginos. En este trabajo, machos y hembras de cada especie se consideran coespecificos cuando han sido colectados juntos en la mis- ma localidad. Una decision dificil es si Freya rubiginosa (Koch 1846) nueva combinacion es una buena especie o solo una variante de F. decorata. En este trabajo se la considera como diferente, pese a haber sido colectada en Para con machos y hembras de F. deco- rata. El hecho de que cuatro ejemplares coin- cidan totalmente con el holotipo hembra de F. rubiginosa y que se puedan distinguir de las que aqui se determinan como F. decorata jus- tifica esta decision. METODOS El formato de las descripciones es segun Galiano (1963b); la quetotaxia se describe como en Platnick y Shadab (1975) con pe- quenas modificaciones, Todas las medidas se dan en milimetros. Abreviaturas : OMA, OLA, OMP y OLP: ojos medios anteriores, laterales anteriores, medios posteriores y laterales posteriores, res- pect! vamente; RC == region cefalica, RT = re- gion toracica, E = embolo, C = conductor, ATR = apofisis tibial retrolateral, BA = bol- sillo de anclaje, CC = conducto de copula- cion, CF = conducto de fertilizacion, Es = espermateca, OC = orificio de copulacion, ap = apical, b = basal, d = dorsal, p = prola- teral, r = retrolateral, v = ventral, p.p. = pro parte, en parte. Abreviaturas de los Museos: Museo Argentino de Ciencias Naturales, Bue- nos Aires: MACN; Museum of Comparative GALIANO— LAS ESPECIES DE FREYA DEL GRUPO DECORATA 23 Zoology, Harvard: MCZ; Milwaukee Public Museum, Wisconsin: MPM; Zoologisches Museum, Berlin: ZMB; Museo Zoologico de “La Specola,” Florencia: MLS; Natural His- tory Museum, Londres: NHM; Museo de La Plata, Argentina: MLP; Museum National d’Histoire Naturelle, Paris: MNHN; Museu de Zoologia da Universidade de Sao Paulo, Brasil: MZSP; Museu Nacional de Rio de Janeiro, Brasil: MNRJ. Genero Freya C.L. Koch 1850 Euophrys (p.p.): C.L. Koch 1846: 200-203. Roewer 1954: 1698. Freya C.L. Koch 1850: 66 (nuevo subgenero de Euophrys). Bonnet 1956: 1918. Simon 1864: 31; 1902: 412 (= Heraclea); 1903: 723, 730, 733, 739 (p.p.). Petmnkevitch 1911: 651 (p.p.); 1928: 198 (p.p.). Neave 1939: 422. Chickering 1946: 163 (p.p.). Galiano 1963a: 23; 1982: 53. Brignoli 1983: 639; 1985: 416. Platnick 1989: 562; 1993: 760; 1997: 884. Proszyhski 1990: 138. Heraclea Peckham & Peckham 1896: 78 (nuevo ge- nero). Neave 1939: 621. Attus (p.p.) Taczanowski 1871: 70. Cyrene Cambridge 1901: 222 (= Heraclea). Thore C.L. Koch 1850: 66 (nuevo subgenero de Euophrys). Simon 1903: 730 (= Freya). Neave 1940: 477. Bonnet 1959: 4595 (= Freya). Trivia (p.p.) C.L. Koch 1850: 66 (nuevo subgenero de Euophrys). Simon (p.p.) 1864: 314. Neave 1940: 572. Bonnet 1959: 4697 (= Ereya). Especie tipo. — Euophrys decorata C.L, Koch 1846. Diagnosis. — Se diferencia de Phiale Koch 1846 y de Euophrys Koch 1834, por presentar en la division apical del tegulo, un conductor membranoso casi paralelo al embolo, el cual es apenas curvo y no espiralado como en Euophrys o en angulo como en Phiale. La apofisis de la tibia del palpo es gruesa, a veces tuberosa y no larga y delgada como en Phiale o bifida como en Frigga Koch 1846. Femur y patella de pata III son mas largos y gruesos que los de pata IV, mientras que tibia y me- tatarso III son mas cortos o iguales a los de IV. Descripcidn. — Ej emplares relati vamente grandes: longitud total, machos 6,20-9.87, hembras 6.53-11.17. Prosoma con lados sua- vemente redondeados, ancho 74-93% del lar- go, alto 48-58% del largo. Area ocular ocu- pando 32-50% del largo del prosoma. Area ocular 52-58% mas ancha que larga, tercera hilera ocular siempre mas angosta que la pri- mera. Altura del clipeo 30-40% del diametro de OMA. OMP ligeramente mas cerca de OLP que de OLA. Formula de patas: machos I-III-IV-II excepcionalmente III y IV iguales en longitud total; hembras IV-III-I-II, a veces III y IV subiguales. Quetotaxia: (variantes entre parentesis). Machos: Femures I d 1-1- 1, p 2ap (r lap, p 1-2); II d 1-1-1, p 2ap, r 1-2 (r lap); III d M-1, p 1-2 (p 2ap), r 1 (r 2, r 1-2); IV d 1-1-1, p lap (p 2ap), r lap. Patellas I, II p l(r 1); III, IV p 1, r 1. Tibias I V 2-2-2, p 1-1 (p 1-1-1); II V lr-2-2, p 1-1- 1 (p 1-1); III, IV d lb (d lb-2ap, d Ib-lrap), V lp-2, p 1-1-1, r 1-1-1. Hembras: I d 1-1-1, p 2ap (r lap); II d 1-1-1, p 2ap, r lap (r 1-1, r 1-2); III d 1-1-1, p 1-2 (p 2ap), r lap (r 2ap); IV d 1-1-1, r 1. Patellas II (p 1); III, IV p 1, r 1. Tibias I v 2-2-2, p 1-1 (p 1); II v lr-2-2, p 1-1; III, IV v lp-2, p 1-1-1, r 1-1- 1. Metatarsos I, II v 2-2; III v 2-2, p 1-2, r 1-1-2; IV V 2-2, p 1-1-2, r 1-1-2. La presencia de un par de espinas dorsales apicales ade- mas de la dorsal basal media en las tibias III y IV de los machos, no parece ser especifica. A menudo estas espinas no son simetricas y en ejemplares del mismo lote pueden estar ausentes. Palpos: tibia con apofisis retrolate- ral o retrodorsal gruesa, tuberosa, a veces con una punta conica dirigida basal, ventral o api- calmente (Figs. 3-13, 15, 16); cimbio con una depresion retrolateral basal de borde ca- renado, que puede faltar; bulbo con divisio- nes media y basal separadas por profundo surco (Figs. 14, 17, 19); division apical con embolo conico, recto o ligeramente curvo, acompanado por un conductor membranoso, casi paralelo, de base independiente (Figs. 19-31), [excepto en F. maculatipes (Cam- bridge 1901) Fig. 14]. Epigino (Figs. 36-43): placa circundada en el extremo anterior por un surco curvo de hordes ligeramente care- nados; borde posterior con un bolsillo de co- pulacion, a veces con bolsillos de anclaje laterales (Fig. 44); area media longitudi- nalmente algo elevada, en algunas especies claramente en forma de quilla {F. regia, Figs. 42, 43). Orificios de copulacion en la mitad anterior de la placa, relativamente pequenos. Conductos de copulacion en forma de embu- do, espermatecas posteriores, ovoideas o es- fericas (Figs. 44-57). La estructura interna de los conductos de copulacion es compleja y no ha podido comprenderse, pese a intentar dis- tintos tipos de clarificacion (KOH, tripsina. 24 THE JOURNAL OF ARACHNOLOGY Mapa 1. — Distribucion de las especies de Freya del grupo decorata. America Central, sur de Mexico y norte de America del Sur. prolasa). Parece posible que el conductor pe- netre en el conducto, ya que la distension no lo separa del embolo. Freya decorata (C.L. Koch 1846) Figs. 1-3, 12, 13, 19, 20, 27, 33, 36, 44, 49; Mapa 1 Euophrys decorata C.L. Koch 1846: 200, fig. 1248 (macho de Brasil, Para, no examinado). Euophrys trifasciata C.L. Koch 1846: 201, fig. 1249. Simon 1903: 730. Euophrys {Thore) trifasciata: C.L. Koch 1850: 66. Euophrys {Freya) decorata: C.L. Koch 1850: 66. Phyale decorata: C.L. Koch 1850: 59 {sic). Atta {Freya) decorata: Simon 1864: 313. Atta {Parthenia) trifasciata {Thore): Simon 1864: 313 {sic). Freya regia: Simon 1903: 724, 730, fig. 858 (error de identificacion). Attus decoratus: Taczanowski 1871: 70. Freya decorata: Simon 1903: 730, 739. Petrunke= vitch 1911: 653. Caporiacco 1948: 717; 1954: 169, figs. 61, 61a-c. Roewer 1954: 1081. Bonnet 1956: 1919. Platnick 1997: 884. Freya decorata var. dyscrita Penther 1900: 285, fig. 2. Petrunkevitch 1911: 653. Bonnet 1956: 1919. Freya strandi Caporiacco 1947: 32; 1948: 717, fig. 148. [1 hembra, 2 machos y 1 hembra inmaduros, sintipos, en MLS, de British Guiana (Guyana), examinados]. Roewer 1954: 1083. NUEVA SI- NONIMIA. Attus brandtii Taczanowski 1871: 72 {Brandtii) [3 machos de Guayana Francesa (uno de Cayena y dos de St. Laurent de Maroni) no examinados]. Petrunkevitch 1911: 596. Mello-Leitao 1948: 1920 (= Freya brandti). Bonnet 1955: 795. NUEVA SINONIMIA. Descripcidn. — Machos: Longitud total 6.60-7.98 {n = 20, x = 7.30). Ancho del prosoma 75-90% de su largo {n = 21, x = 80%); alto del prosoma 49-58% de su largo {n = 9, X = 53%); largo del area ocular 40- 53% del largo del prosoma {n = 14, x = 44%); largo del area ocular 60-67% del ma- yor ancho del area {n = 14, x = 64%). Altura del cKpeo 26-40% del diametro de OMA {n = 9, X = 32%). Formula de patas: I-III-IV- II {n = 7), I-IV-III-II, IV y III subiguales {n = 3). Quetotaxia: como en el genero. Palpos (Figs. 2, 3, 33): tibia maciza, con gran es- cotadura dorsal, cara retrolateral tuberosa, con horde superior recto, una apofisis trian- gular en el horde inferior de la cara interna, cuyo vertice se dirige hacia la base de la ti- bia. Esta apofisis es visible solamente en vis- ta ventral o retro ventral (Figs. 12, 13). Cim- bio con una depresion transversa en retrodorsal basal, con horde superior ligera- mente carenado. Embolo ligeramente curvo. GALIANO^LAS ESPECIES DE FREYA DEL GRUPO DECORATA 25 ancho en la base y agudo en el apice; el con- ductor es una lamina membranosa, incolora, ligeramente plegada longitudinalmente y con el extremo libre curvado (Figs. 19, 20, 27), acompafia al embolo en todo su trayecto y lo sobrepasa en el apice. Color (Fig. 1): proso- ma negro, con bandas y manchas de pelos blancos de la siguiente manera: anchas ban- das submarginales que atras termiean a cada lado del declive toracico y por delante con- tinuae para formar la deesa barba del clipeo; una gran mancha oval media en el margen anterior, que por detras alcanza la altura del borde anterior de los OLP; desde el borde interno de cada OLA en margen anterior, una banda longitudinal hasta el borde anterior del OLP de su lado, a veces interrumpida en el medio; una mancha oval en la parte media de RC, delante de la estria, separada por algunos pelitos negros de la banda media toracica que se angosta hacia atras y termina en la parte m.edia del declive toracico. Opistosoma ne- gro o pardo negruzco, con pelos al tono; una banda de tegument© amarillo cubierta por pe- los blancos bordea la base y sigue por los lados, afieandose y terminando en el apice o un poco antes; una banda media longitudinal de tegumento amarillo con pelos blancos des- de la banda basal hasta el apice; en algunos ejemplares separada de la basal por una zona con pelos negruzcos, Tuberculo anal blan- quecieo, con algunos pelitos blancos. Vientre pardo claro, con manchitas amarillas. Queli- ceros pardo rojizo oscuro; laminas y labio pardos con hordes amarilleetos. Pata I: femur pardo oscuro, con la mitad dorsal basal pardo claro con escasos pelitos blancos; patella par- do oscuro, mitad dorsal basal amarillenta con pelos blancos y algunos pelos blancos en el borde distal; tibia y metatarso pardos, con el tercio medio amarillento con pelos blancos dorsales; tarso amarillo. Pata II: como I, pero pardo claro. Patas III y IV amarillentas. Pal- pos: femur pardo negruzco, la mitad apical mas Clara, con largos pelos blancos dorsales y en el tercio apical una densa area de pelos blancos; patella con caras dorsal y prolateral con densos pelos blancos; tibia parda con pe- los pardos; cimbio pardo, mas claro hacia el apice con pelos al tono. Hembras: Longitud total 6.53-9.58 {n — 9, X = 8.64). Ancho del prosoma 73-90% de su largo (w — 13, x — 79%); alto del prosoma 48-55% de su largo (w — 8, x — 51%); largo del area ocular 41-50% del largo del proso- ma (« = 8, X — 44%); largo del area ocular 61-66% del mayor ancho del area (« = 8, x — 64%). Altura del clipeo 27-47% del dia- metro de OMA {n = 8, x — 33%). Formula de patas: IV-IILLII (w - 5), IILIV-LII {n = 1). Quetotaxia como en el genero. Epigino (Figs. 36, 44, 49): placa limitada en su borde anterior por ue surco curvo con hordes ca- renados; Imea media longitudinal ligeramen- te elevada; orificios de copulacion en el area anterior, ovales, con el eje mayor longitudi- nal. Borde posterior excavado, con bolsillos de anclaje laterales. Color: prosoma pardo ro- jizo oscuro con RC negruzca, cubierta por pelos escamosos semitransparentes, brillan- tes. En margen anterior, algunos escasos pe- los blancos entre los OMA, entre OMA y OLA y entre OLA y OLR RC limitada atras por una banda pardo claro con forma de V invertida, con vertice en extremo posterior de la estria. En RT, a partir de la estria, dos ban- das oscuras de cada lado, alternadas con ban- das Claras y una banda media longitudinal clara que llega al margen posterior, Sobre las bandas oscuras pelos trasMcidos como los de RC y sobre las claras, pelos blanquecinos, es- casos. Clipeo con pelos largos amarillentos, que no forman barba, Opistosoma pardo cla- ro, con pelos al tono. Bandas como en el ma- cho, pero las laterales se interrumpen antes del tercio apical y son seguidas por una man- chita alargada. La banda media se une a la basal o comienza en el tercio basal y continua hasta el apice. El horde interno de las bandas laterales y ambos hordes de la media, bor- deados por bandas angostas pardo oscuro, con pelos al tono. Vientre pardo claro, la par- te media con manchitas pardas, Pata I pardo claro, femur, patella y tibia con tercio apical mas oscuro, pelos blancos sobre las partes claras, metatarso y tarso amarillentos. Pata II: como I, pero mas clara. Patas III y IV pardo amarillento, mas oscuro en los apices. Pal- pos: femur amarillento, con pelos blancos dorsales; patella amarillenta con una mancha dorsal basal pardo oscuro y pelos blancos en la mitad distal; tibia y tarso pardo claro, con mancha oscura dorsal basal y pelos amarillos dorsales y laterales. Material examlnado. — BRASILj Para: Belem, 8 machos, 5 hembras, 5 inmaduros, N® 9654 MACN, agosto 1971 (Galiano); 1 macho, 2 hem- 26 THE JOURNAL OF ARACHNOLOGY Figuras 2-7. — Palpos izquierdos. 2. Freya decorata, prolateral; 3. El mismo, retrolateral; 4. F. chapare, retrolateral; 5. F. nigrotaeniata, retrolateral; 6. F. atures, retrolateral; 7. F. dureti, retrolateral. Escalas = 0.5 mm. GALIANO—LAS ESPECIES DE FREYA DEL GRUPO DECORATA 27 Figuras 8-13. — ^Tibias de palpos derechos, vista retrolateral. 8. Freya dureti; 9. F. nigrotaeniata; 10. F. atures; 11. F. chapare; 12. F. decorata de Guyana; 13. F. decorata de Para. Escalas 8, 9, 11 = 0.5 mm; 10, 12, 13 = 100 pm. bras, N° 9655 MACN, agosto 1970 (Galiano); 1 macho, N° 9656 MACN, marzo 1953 (Duret); 1 hembra (MPM) (Moeekhaus); 2 machos, (F. regia det. Simon) (MNHN); Amapd: Santana, rio Matapi, 2 machos, N° 9657 MACN, junio 1966 (Galiano); Serra do Navio, 1 macho, 1 hembra, 1 inmaduro, N® 9658 MACN, junio 1966 (Galiano); Amazonas: Manaus, 1 hembra, N° 9690 MACN, agosto 1971 (Galiano). GUYANA: Upper Essequibo, Onoro Re- gion, 1 macho, 1 hembra (MLS), 1-24 diciembre 1937 (Hassler). En MLS, colectado por Drs. Bee- cari y Romiti, determinado por Caporiacco: Con- warook, Potaro, 2 machos, 1 hembra, 18 mayo 1936; Two Mouth, Essequibo, 1 macho, 14 Julio 1936; Tumatumari, 3 machos, 19 setiembre 1936; Garroway Landing, Potaro, 1 macho inmaduro, 20 marzo 1936; Mackenzie, 1 hembra, setiembre 1931; Campo I, Demerara, Baboon Camp, 1 hembra, oc- tubre 1931; Lungo il Cattle fra Campo V Curupu- cari, 1 macho, 8 noviembre 1931; Cannister Falls, Demerara, 1 macho, noviembre 1931. Uni-Con, 2 machos (MPM) (Parrish). GUAYANA FRANCE- 28 THE JOURNAL OF ARACHNOLOGY SA: S. Jean du Maroni, 2 hembras (MLS), 1914; Charvein, 1 macho (MLS), 1916. ECUADOR: Napo: Lago Agrio 270 m, 4 machos, N® 9659 MACN, junio 1976 (Williner); Sacha 240 m, 1 ma= cho, N° 9660 MACN, junio 1976 (Williner); Li= moncocha, 1 macho, 1 hembra, N° 9661 MACN, abril 1984 (A. Roig). Distribucion. — Brasil: Estados de Para, Amapa y Amazonas. Guyana. Guay ana Fran- cesa. Ecuador: Provincia de Napo. Nota: Los caracteres de Freya strandi coin- ciden exactamente con los de las hembras de Freya decorata de Guyana y Guayana Fran- cesa. El patron de coloracion asi como la dis- tribucion de Attus brandti son iguales a los de F. decorata, por lo que se establece la simo- nimia. Los tipos de esta especie no son citados en el Catalog© de Proszyhski (1971:380) por lo que es probable que esten perdidos. Freya regia (Peckham & Peckham 1896) Figs. 17, 18, 30, 31, 32, 42, 43, 46, 52; Mapa 1 Heraclea regia Peckham & Peckham 1896: 77, pi. Ill, figs. 6, 6a-c, pi. IV, figs. 1, la-b (1 macho lectotipo, 1 hembra paralectotipo, 3 hembras, 3 machos paralectotipos, aqui designados, de Gua- temala, N° 807; 4 hembras, 1 macho inmaduro, sintipos, de Guatemala, N° 789 en coleccion Peckham (MCZ), examinados). Cyrene regia: Cambridge 1901: 222, 226, 229, pi. XVIII, figs. 12, 12a-g, 13, 13a-d. Freya regia: Petrunkevitch 1911: 654. Caporiacco 1938: 279. Bonnet 1956: 1921. Roewer 1954: 1082. Diagnosis.— Se diferencia de las otras es- pecies de Freya del grupo decorata por pre- sentar una saliente en angulo recto en el borde intemo del conductor, de modo que el tercio apical es mas ancho que los dos tercios ba- sales; por la apofisis triangular tibial retroven- tral mas horizontal que en F. decorata y con el extreme romo; por carecer de manchas de pelos blancos en margen anterior del prosoma entre OMA y OLA. El epigino se distingue del de todas las otras especies del grupo por el borde posterior curvado hacia atras y por una carena o quilla longitudinal media. Descripcion. — Lectotipo macho: Largo to- tal 7.85. Prosoma largo 4.00, ancho 3.13, alto 1.87. Clipeo, alto 0.33. Area ocular largo 1.70, ancho de hilera anterior 2.70, de hilera posterior 2.47. Distancias OLA-OMP 0.47, OMP-OLP 0.33. Diametro OMA 0.87. For- mula de patas I-III-IV-IL Quetotaxia como en el genero, pero patella I sin espinas y tibia II p 1-1, Palpos (Figs. 17, 18, 32): cimbio sin la depresion retrodorsal basal presente en F. decorata. Conductor ligeramente mas corto que el embolo, laminar, de apice redondeado, con un ensanchamiento en el tercio distal (Figs. 30, 31). Color: los sintipos estan en regular estado de conservacion y depilados en su mayor parte. Solamente algunos ejem- plares conservan restos de pelos. Diferencias con Freya decorata: prosoma pardo oscuro con pelos negros, la RC con pelos rojizos; una mancha de pelos blancos en el margen anterior entre los OMA, una mancha oval de pelos blancos desde extremo anterior de la estria toracica hasta el declive toracico, don- de se bifurca y se une a las bandas laterales. Opistosoma pardo, cubierto por pelos rojo pardusco. Paralectotipo hembra: Largo total 10.37. Prosoma largo 4.27, ancho 3.53, alto 2.21 . Clipeo, alto 0.33. Area ocular largo 1.87, an- cho de hilera anterior 2.87, de hilera posterior 2.73. Distancias OLA-OMP 0.50, OMP-OLP 0.33. Diametro OMA 0.93. Formula de patas IV-III-I-II. Quetotaxia como en el genero. Epigino (Figs. 42, 43, 46, 52): borde poste- rior curvado hacia atras; una carena media longitudinal separa dos depresiones en las que se abren los orificios de copulacion cir- culares. Color: como en F. decorata, excep- to: prosoma pardo rojizo, con escasos pelos blancos laterales. Opistosoma pardo, con pe- los rojos; una banda media longitudinal des- de la parte media hasta cerca del apice, de pelos bianco amarillento; una banda de pelos bianco amarillento circunda la base y se con- tinua por los lados hasta aproximadamente el medio, seguida por dos manchas de cada lado, la apical mas alargada. Patas pardas, to- dos los femures, patellas y tibias II a IV con el tercio medio amarillo con pelos blancos. Palpos: femur pardo claro con manchita api- cal amarilla con pelos blancos; patella y tibia pardo oscuro con tercio apical amarillo con pelos blancos. Material examinado.— Solo la serie de sintipos. Distribucion. — Guatemala; Mexico: Chia- pas. Nota: Para colorido, ver Peckham y Peck- ham (1896) y Cambridge (1901). GALIANO— LAS ESPECIES DE FREYA DEL GRUPO DECORATA 29 Figuras 14-18. — 14-16. Freya maculatipes. 14. Palpo derecho, ventral; 15. El mismo, retrolateral; 16. El mismo, retroventral. 17. F. regia, palpo derecho, ventral; 18. F. regia, palpo izquierdo, retrolateral. Escalas = 0.5 mm. Los especimenes machos procedentes de Brasil, Para, que E. Simon (1903: 724, 730, fig. 858) determino como F. regia, son en rea- lidad F. decorata y las hembras tienen epigino similar a Freya rubiginosa. Este error de iden- tificacion no altera los caracteres diagnosticos que Simon dio para el genero (Simon 1903: 739). Freya rubiginosa (C.L. Koch 1846) nueva combinacion Figs. 38, 45, 50; Mapa 1 Euophrys rubiginosa C.L. Koch 1846: 209, fig. 1255 (hembra holotipo, N° 1803 en ZMB, de Brasil, Para, examinado). Petrunkevitch 1911: 649. Bon- net 1956: 1887; 1959: 4697. Roewer 1954: 1181. 30 THE JOURNAL OF ARACHNOLOGY Figuras 19-24. — 19. Bulbo, vista ventral. 20-24, Embolos y conductores. 19, 20. Freya decorata; 21. F. atures; 22. F. chapare; 23. F. dureti; 24. F. nigrotaeniata. Escalas 19 = 0.5 mm, 20-24 = 100 |jLm. Euophrys (Trivia) rubiginosa: C.L. Koch 1850: 68. Atta (Trivia) rubiginosa: Simon 1864: 314. Diagnosis. — Se diferencia de las otras es- pecies del grupo por tener el horde posterior del epigino con una escotadura apenas exca- vada. Descripcion. — Holotipo hembra: Clipeo, alto 0.27. Area ocular largo 1.73, ancho de hilera anterior 2.73, de hilera posterior 2.60. Distancia OMA-OLA 0.40, OMP-OLP 0.40. Diametro OMA 0.93. Epigino: Fig. 38. Hembra: (N° 9663 MACN comparada con el holotipo). Largo total 9.66. Prosoma largo 4.33, ancho 3.33, alto 2.00. Clipeo, alto 0.23. Area ocular largo 1.77; ancho de hilera ante- rior 2.83, de hilera posterior 2.73. Distancias OLA-OMP 0.50, OMP-OLP 0.43. Diametro OMA 0.90. Formula de patas III-IV-I-II. Que- totaxia como en el genero. Opistosoma largo 5.33. Epigino (Figs. 45, 50): horde posterior con una leve escotadura. Borde anterior con surco profundo, curvo, de hordes carenados. Orificios de copulacion circulares o apenas alargados transversalmente. Conductos de co- GALIANO— LAS ESPECIES DE FREYA DEL GRUPO DECORATA 31 Figuras 25—31, — Palpos: Embolos y conductores. 25, Freya atures, dorsal; 26. Los mismos, proven- tral; 27. F. decorata, ventral; 28, F. dureti, ventral; 29. Los mismos, prolateral; 30. F. regia, retrolate- ral; 31. Los mismos, ventral. Escalas =100 pm. pulacion muy anchos, tocandose en la Imea media o apenas separados. Espermatecas pe- quenas, esfericas. Color: como en F. decorata, con estas diferencias: prosoma pardo, con la RC mas oscura, cubierta por pelos pardo ro- jizo, transparentes, brillantes; algunos pelos bianco amarillento en margen anterior, entre los OMA y entre OMA-OLA; una mancha tri- angular de esos pelos delante de la estria to- racica. Desde extremo anterior de la estria hasta margen posterior de RT, una banda me- dia longitudinal amarilla con pelos blancos. Figuras 32-35. — Tibias del palpo derecho, vista dorsal. 32. Freya regia; 33. F. decorata; 34. F. atu- res; 35. F. chapare. Escalas = 0.5 mm. Opistosoma pardo, las areas entre las bandas blancas con pelos pardos con reflejos rojizos. Variantes. — En una de las hembras las bandas laterales del opistosoma se interrum- pen en la mitad y se contimian con dos man- chas alargadas. Nota. — El holotipo estaba pinchado en aL filer, segun el uso de la epoca. Con autoriza- cion del Dr. Moritz, se procedio a ablandarlo, retirarle el alfiler y se diseco el epigino que se ilustra (Fig. 38). Material examinado. — BRASIL: Para: Belem, 1 hembra, N° 9663, MACN, comparada con el tipo, agosto 1970 (Galiano); 2 hembras, N° 9662 MACN, agosto 1971 (Galiano); 1 hembra determinada como Freya regia, probablemente por Simon (MNHN). Distribucion. — Solo de la localidad tipo. Freya nigrotaeniata (Mello-Leitao 1945) nueva combinacion Figs. 5, 9, 24, 37, 56, 57; Mapa 2 Phiale nigrotaeniata Mello-Leitao 1945: 292, fig. 82 (1 hembra y 1 macho sintipos, inmaduros, N° 16.825 en MLR de Argentina, Corrientes, Goya, examinados), Roewer 1954: 1062. Galiano 1981: 13 (sp. inq.) Freya regia: Mello-Leitao 1941: 203 (identificacion erronea); 1942: 383 (identificacion erronea). 32 THE JOURNAL OF ARACHNOLOGY Figuras 36-39. — Epiginos, vista ventral. 36. Freya decorata; 37. F. nigrotaeniata; 38. F. rubiginosa; 39. F. atures. Escalas = lOOfxm. Diagnosis. — Se diferencia de F. decorata por tener en el palpo dos apofisis tibiales c6- nicas: una retrodorsal, corta y de apice romo y otra retroventral, larga, de apice agudo y granulada en la base, dirigida hacia el apice, El conductor es mas largo que el embolo, an- cho en la base y agudo hacia el apice. El epi= gino se diferencia del de F. decorata por tener una gran escotadura en margen posterior, los orificios de copulacion circulares y los con- ductos de copulacion mas cortos. Descripcidn. — Machos: Longitud total 6.67-9.31 (n = 9, X = 8.26). Macho N° 9666 MACN: Longitud total 9.31. Prosoma largo 4.20, ancho 3.27, alto 2.00. CKpeo, alto 0.27. Area ocular largo 1.60, ancho de hilera ante- rior 2.60, de hilera posterior 2.47. Distancias OLA-OMP 0.43, OMP-OLP 0.40. Diametro OMA 0.80. Palpos (Figs. 5, 9, 24): cimbio sin depresion retrodorsal basal. Formula de patas y quetotaxia como en el genero. Color “in vivo”: prosoma negruzco, con pelos negros; GALIANO— LAS ESPECIES DE FREYA DEL GRUPO DECORATA 33 banda marginal de pelos negros; anchas ban- das laterales submarginales de pelos blancos, desde cada lado del declive toracico hacia adelante, donde forman la barba del clipeo, recta bajo los ojos anteriores, espacio entre los OMA ocupado por pelos negros; banda media longitudinal de pelos blancos desde margen anterior de RC hasta la mitad del declive to- racico, donde se bifurca y se une a las bandas blancas laterales, En cada lado de RC, una bandita de pelos blancos desde margen ante- rior, mitad externa de OLA, hasta borde an- terior de OLR Tanto estas bandas como la me- dia en la RC, estan bordeadas por pelos rojos. Queliceros negros con largos pelos blancos cerca de la base de cara anterior. Piezas labia- les pardo oscuro con hordes amarillos; ester- non pardo claro con margen oscuro. Opisto- soma negro, con banda basal transversa de pelos blancos que se continua por los lados hasta el tercio apical, ensanchandose levemen- te en el ultimo tramo; banda media longitu- dinal de pelos blancos unida a la basal y ter- minando en el apice. Tanto esta banda como las laterales, bordeadas por pelos rojos. El es- pacio entre la banda media y las laterales con pelos negros y rojo apagado mezclados, estos ultimos mas densos en el area central de cada lado. Tuberculo anal amarillo, con pelos blan- cos. Pata I pardo rojizo, femur negruzco en el tercio apical, con algunos pelos blancos en dorso apical; patella con apice negruzco y pe- los blancos en parte media dorsal y prolateral apical; tibia pardo oscuro en los tercios basal y apical, tercio medio pardo claro con pelos blancos; metatarso pardo claro con tercio api- cal negruzco y escasos pelos blancos en el ter- cio basal dorsal. Pata II como I pero mas cla- ra. Patas III y IV pardo amarillento, algo mas oscuras en apices de femures, tibias y meta- tarsos. Palpos pardo muy oscuro, con densi- simos y largos pelos blancos en dorso apical de femur y dorso de patella. Hembras: Largo total 8.11-11.17 {n = 12, X - 10.00). Hembra N° 9665 MACN: Largo total 10.91. Prosoma largo 4.40, ancho 3.47, alto 2.07. Clipeo, alto 0.27. Area ocular largo 1.73, ancho de hilera anterior 2.83, de hilera posterior 2.77. Distancias OLA-OMP 0.40, OMP-OLP 0.40. Diametro OMA 0.90. For- mula de patas y quetotaxia como en el genero. Opistosoma largo 6.33. Epigino (Figs. 37, 56, 57). Color “in vivo”: prosoma pardo negruz- co, con RC negra; RT con cuatro o cinco ban- das radiantes pardo claro; sobre RC pelos ne- gros y rojo bronceado mezclados, brillantes y traslucidos. Sobre las bandas claras de RT pe- los blancos, brillantes, poco densos; sobre las bandas oscuras pelos como en RC. Pelos blan- cos en margen anterior, escasos entre OMA y entre OMA y OLA, formando una manchita delante de la estria y una banda poco densa desde la estria hasta mitad del declive toraci- co. Clipeo sin barba, con largos pelos blancos en el margen. Opistosoma pardo; banda basal transversa amarillenta con pelos blancos, ban- das laterales se adelgazan dos veces antes de llegar al apice; banda media longitudinal ama- rilla con pelos blancos, desde la basal hasta el apice. Estas bandas estan bordeadas por pelos negros; la parte del dorso entre la banda media y las laterales con tegumento pardo y pelos rojo apagado, sin brillo. Vientre con manchi- tas pardas. Pata I: femur con mitad basal ama- rilla y apical pardo oscuro; patella y tibia par- das con mancha media amarilla con pelos blancos; metatarso pardo claro con extremo apical oscuro. Pata II como 1. Patas III y IV pardo claro, todos los artejos oscurecidos en ambos extremos. Palpos: femur, patella y tibia amarillos con mancha basal dorsal pardo os- curo; tarso pardo con mancha dorsal basal ne- gruzca. Material examinado. — En MACN; ARGEN- TINA: Santa Fe: Las Gamas, 20 km de Vera, 1 hembra, N° 9665, 27-30 octubre 1994 (Ramirez & Faivovich); 1 macho, N° 9666; 4 hembras, 2 ma- chos y 7 inmaduros, N° 9664, todos descendientes de la hembra N° 9665; Salta: Quebrada de Piqui- renda, 1 macho, N° 9667, octubre 1966 (Hepper); Aguas Blancas, 2 machos, 1 inmaduro, N° 9668, abril 1984 (E. Maury); 1 macho, 6 inmaduros, N° 9669; marzo 1967 (Galiano); Finca Jakulica 25 km NO de Aguas Blancas, 1 macho, N° 9679, noviem- bre 1994 (Goloboff & Ramirez); El Tabacal, 1 ma- cho, N° 9671, junio-julio 1933 (Daguerre); Urundel, 1 macho, 1 hembra, N° 9672, diciembre 1954 (Bi- raben); La Quena, rio Bermejo, 1 hembra, N° 9673, mayo 1983 (Goloboff). Jujuy: El Cafetal, 1 macho, N° 9674, julio 1978 (Williner); Yuto, El Pantanoso, 1 hembra, N® 9675, noviembre 1966 (Galiano). Chaco: Selva del Rfo de Oro, 2 hembras, N® 9676, enero 1965 (Galiano); Resistencia, 1 hembra, 2 in- maduros, N® 9677, junio-julio 1977 (Carbajal). Cor- rientes: Santiago Alcorta, 2 hembras, N° 9678, ju- nio 1943 (Biraben); Ituzaingo, 1 hembra, N° 9679, noviembre 1963 (Partridge). PARAGUAY: Dpto. Presidente Hayes: Ruta Transchaco km 193, 1 hem- bra, N® 9680, julio 1990 (Ramirez). 34 THE JOURNAL OF ARACHNOLOGY Distribudon. — Argentina: Corrientes, San- ta Fe, Salta, Jujuy, Chaco. Paraguay: Depto. Presidente Hayes. Nota. — La hembra N° 9665 fue mantenida en el laboratorio donde efectuo un desove, se crio la descendencia y obtuvieron adultos, en- tre ellos el macho N° 9666. La observacion de los juveniles durante su desarrollo permitio identificarlos con los tipos inmaduros de la es- pecie. Frey a maculatipes (F.O.P.-Cambridge 1901) Figs. 14-16; Mapa 1 Cyrene maculatipes Cambridge 1901: 225, 234, pi. XIX fig. 12, 12a-d (macho holotipo de Mexico, Chiapas, Teapa, 1905-210 en NHM, examinado). Freya maculaticeps: Simon 1903: 730 (lapsus). Frey a maculatipes: Petrunkevitch 1911: 654, Roe- wer 1954: 1082. Bonnet 1956: 1920. Diagnosis. — Esta especie se diferencia de todas las otras del grupo decorata porque el embolo y el conductor tienen una base comun y se separan aproximadamente en la mitad de su trayecto. Descripcion. — Macho holotipo: Area ocu- lar largo 1.33, ancho de hilera anterior 2.03, de hilera posterior 2.00. Distancias OLA- OMP 0.33, OMP-OLP 0.30. Formula de patas I-III-IV-II. Quetotaxia como en el genero, sal- vo que metatarsos I y II tienen 1 prolateral apical. Palpos (Figs. 14-16): apofisis tibial gruesa, con punta aguda dirigida hacia el cim- bio. Cimbio sin depresion retrodorsal basal. Embolo y conductor con origen comun, el conductor mas corto que el embolo y con el extremo apical truncado. Color: ver descrip- cion original. Aparentemente coincide con el de las especies del grupo decorata, excepto por los pelos anaranjados brillantes en RC y en los femures, y por la banda media dorsal del opistosoma, que en su mitad distal esta formada por chevrons unidos por los vertices. Distribudon. — Solo la localidad tipo. Nota. — Unico ejemplar de la especie co- nocido, esta atravesado por un alfiler ento- mologico que penetra por el estemon, sale por el declive toracico, entra en el opistosoma por el pedicelo y sale por la parte ventral delante de las hileras. No se pueden realizar otras me- didas que las del area ocular. Es con ciertas dudas que se incluye esta especie en el grupo. Pese a que Cambridge (1901) senala que es muy similar a Cyrene curvispina Cambridge 1901, no es asi, puesto que esta ultima especie [actualmente un sinonimo de Nycerella san- guinea (Peckham & Peckham 1896) Galiano 1982] carece de conductor. Tal vez el hallazgo de las hembras de la especie permita una de- finicion en cuanto a su correcta ubicacion. Freya dureti nueva especie Figs. 7, 8, 23, 28, 29, 41, 48, 51; Mapa 1 Tipos. — Macho holotipo, N° 9681 MACN de Brasil, Para, Belem, (elev. 11m; 01°28'S, 48°29'W) agosto de 1953 (P Duret); 1 macho paratipo en MNRJ, de Brasil, Amazonas, Ma- naus (elev. 93 m; 03°08'S, 60°02'W) Reserva Ducke (26 km desde Manaus), agosto 1971 (Galiano); 1 hembra alotipo, N° 9682 MACN, de igual localidad y colector. Etimologia. — La especie se nombra en honor del entomologo Dr. Jose Pedro Duret, quien colecciono abundantes aranas durante sus viajes de estudio. Diagnosis. — Se diferencia de F. decorata y de F. nigrotaeniata por tener el prosoma mas robusto y por carecer de manchas o bandas de pelos blancos en el margen anterior de la RC. De F. decorata se distingue ademas por tener la apofisis tibial palpal retrolateral conica. El conductor, con el extremo distal redondeado, a diferencia de F. nigrotaeniata donde dicho extremo es agudo. Descripcion. — Holotipo macho: Largo to- tal 9.86. Prosoma largo 4.67, ancho 4.00, alto 2.40. Clipeo, alto 0.33. Area ocular largo 1.97, ancho de hilera anterior 2.83, de hilera pos- terior 2.70. Distancias OLA-OMP 0.50, OMP- OLP 0.43. Diametro OMA 0.93. Palpos (Figs. 7, 8, 23, 28, 29): el cimbio carece de la de- presion retrodorsal basal de F. decorata. La apofisis tibial es conica, pero no tiene la base ensanchada y granulada presente en F. nigro- taeniata. Formula de patas I-IV-III-II, IV y III subiguales. Quetotaxia como en el genero. Opistosoma, largo 5.20. Color: prosoma pardo negruzco con RC negra, pelos pardo negruz- co, con algunos reflejos rojizos; no existen tres manchas ni banda transversa de pelos blancos en el margen anterior; banda longi- tudinal media de pelos blancos que comienza en mitad de la RC, pasa sobre la estria y se arnpha formando un rombo en RT, terminando al comienzo del declive toracico; bandas la- terals de pelos blancos casi marginales, muy anchas, bien separadas en el declive de RT, se continuan hacia adelante pero no ocupan todo GALIANO™-LAS ESPECIES DE FREYA DEL GRUPO DECORATA 35 Figuras 40-43. — Epiginos. 40. Freya chapare; ventral. 41. F. dureti, ventral; 42. F. regia, lateral; 43. El mismo, ventral. Escalas =100 jjLm. el espacio bajo los ojos laterales y en el clipeo forman la barba, de borde superior recto, que deja libre un espacio bajo los cuatro ojos an- teriores. Queliceros pardo oscuro; piezas la- biales pardo rojizo, con margenes mas claros. Estemon pardo rojizo. Opistosoma pardo os- curo con manchitas amarillas, cubierto por pe- los pardos con reflejos rojizos; angosta banda basal de pelos blancos, que se continua por los lados y se angosta en la parte media, ter- minando antes del apice; banda media dorsal longitudinal de pelos blancos desde el tercio basal hasta cerca del apice. Tuberculo anal amarillo con escasos pelitos blancos. Vientre pardo, con ancha banda media negruzca. Patas pardo rojizo oscuro con las siguientes areas pardo amarillento: pata I con banda basal en patella y mancha media dorsal en tibia con 36 THE JOURNAL OF ARACHNOLOGY Figuras 44-48. — Epiginos clarificados, vista dorsal. 44. F. decorata; 45. F. rubiginosa; 46. F. regia; 47. F chapare; 48. F. dureti. Escalas = 100 ixm. pelos blancos, metatarso en los dos tercios ba- sales, todo el tarso. Patas II, III y IV, femures con mancha media dorsal, mitad basal de pa- tellas y mancha media dorsal de tibias, todas con pelos blancos; metatarsos pardos, con api- ces negruzcos; tarsos pardo claro. Palpos como en el genero. Alotipo hembra: Largo total 8.11. Prosoma largo 3.87, ancho 2.93, alto 1.87. CKpeo, alto 0.23. Area ocular largo 1.53, ancho de hilera anterior 2.50, de hilera posterior 2.40. Distan- cias OLA-OMP 0.40, OMP-OLP 0.37. Dia- metro OMA 0.77. Formula de patas y queto- taxia como en el genero. Epigino: Figs. 41, 48, 51. Color: como en F. decorata, sin banda media longitudinal de pelos blancos en RT; la banda media longitudinal del opistosoma se- parada de la basal; cada banda lateral llega hasta el tercio apical y es seguida por una mancha de pelos blancos, que no alcanza el apice. Patas amarillas, la I con una mancha retrolateral basal y el tercio apical del femur pardo oscuro; patella y tibia pardas con tercio medio amarillo. Palpos amarillos; patella, tibia y tarso con mancha dorsal basal pardo oscuro. Distribucion. — Brasil: Estados de Amazo- nas y Para. Nota. — Los dos ejemplares machos estu- diados provienen de localidades alejadas pero son sin duda de la misma especie. Como se ha visto en el caso de F. decorata, la distri- bucion desde el Alto Amazonas hasta su de- sembocadura se observa tambien en esa y otras especies. En Manaus se han coleccio- GALIANO— LAS ESPECIES DE FREYA DEL GRUPO DECORATA 37 nado ejemplares femeninos que pertenecen a dos especies diferentes: una hembra que se de- termina como F. decorata y otra que aqui se describe como la hembra de F. dureti. En Para se encuentran F. decorata, F. rubiginosa y F. dureti. Existe la posibilidad de que F. rubi- ginosa sea la hembra de F. dureti. No puede tenerse en este grupo la certeza de la coes- pecificidad de ejemplares de distinto sexo, por lo que se mantiene el status de F. rubiginosa. Freya chapare nueva especie Figs. 4, 11, 22, 35, 40, 47, 53; Mapa 2 Tipos. — Holotipo macho, N° 9683 MACN y 2 machos paratipos, N° 9684 MACN, de Bolivia, Cochabamba, Depto. Chapare, Cris- talmayo, febrero 1971 (A. Martinez); 5 ma- chos paratipos, N° 9685 MACN, y 1 hembra Alotipo, N° 9686 MACN, de Bolivia, Cha- pare, Chimore (coordenadas aproximadas 17°S, 66°W), enero 1972 (M. Fritz). Etimologia.^ — El nombre de la especie de- riva de la localidad tipo. Diagnosis. — Se diferencia de F. decorata por la forma de la apofisis tibial retrolateral, que es menos voluminosa y con la punta tri- angular anterior dirigida hacia la cara ventral de la tibia y no hacia la base, como en F. decorata. Se distingue de F. nigrotaeniata, cuyo embolo y conductor son semej antes, por- que en esta ultima especie la apofisis tibial es conica. El epigino es muy semej ante al de F. nigrotaeniata, del cual se diferencia apenas por tener los conductos de copulacion algo mas largos. Descripcion. — -Holotipo macho: Largo to- tal 7.71. Prosoma largo 3,93, ancho 2.90, alto 1.87. Clipeo, alto 0.27. Area ocular largo 1.43, ancho de hilera anterior 2.43, de hilera pos- terior 2.35. Distancias OLA-OMP 0.40, OMP- OLP 0.37. Diametro OMA 0.80. Formula de patas y quetotaxia como en el genero. Palpos (Figs. 4, 11, 22, 35): cimbio con depresion retrodorsal poco profunda. Apofisis retrolate- ral con borde superior recto, separada de la dorsal por una profunda escotadura. Extreme antero superior con una punta aguda, dirigida hacia cara ventral del palpo. Conductor mem- branoso, ligeramente plegado longitudinal- mente, mas largo que el embolo y terminado en una punta aguda. Color: como en F. de- corata, con estas diferencias: la banda longi- tudinal media de pelos blancos de RT se bi- furca en el declive toracico y cada rama se une con pocos pelos a las bandas submargi- nales. La banda longitudinal media de pelos blancos del dorso del opistosoma, se une a la basal y termina en el apice, asi como las la- terales, El dorso entre estas bandas tiene pelos pardo negruzco, con reflejos rojizos. Alotipo hembra: Largo total 8.11. Prosoma largo 3.67, ancho 3.40, alto 1.80. Clipeo, alto 0.23. Area ocular largo 1.43, ancho de hilera anterior 2.40, de hilera posterior 2.36. Distan- cias OLA-OMP 0.40, OMP-OLP 0.33. Dia- metro de OMA 0.73. Formula de patas y que- totaxia como en el genero. Epigino: Figs. 40, 47, 53. Color: como en F. decorata, con las bandas radiantes toracicas bien evidentes pero sin banda longitudinal media de pelos blancos en RT. Material estudiado. — BRASIL: Goiaz: Faz. Aceiro Yatai, 1 macho, octubre 1962 (Exp. Depto. Zoologia) (MZSP). Distribucion. — Bolivia: Cochabamba. Bra- sil: Goiaz. Nota: — Tres de los machos paratipos pre- sentan d lb-2ap espinas en tibias III y IV. Freya atures nueva especie Figs. 6, 10, 21, 25, 26, 34, 39, 54, 55; Mapa 1 Tipos. — Macho holotipo, N° 9687 MACN, 1 hembra alotipo, N° 9688 MACN, 2 hembras paratipos, N° 9689 MACN, de Venezuela, Te- rritorio Federal de Amazonas, Atures (con- fluencia de los rios Alto Orinoco y Catania- po), junio 1976 (A. Martinez). Etimologia.— -El nombre de la especie de- riva de la localidad tipo. Diagnosis. — -Se diferencia de F. decorata porque en lugar de tres manchas de pelos blancos en margen anterior de RC tiene una banda transversa de pelos blancos desde el borde intemo de un OLA al del otro; el cimbio carece de depresion retrodorsal basal; la apo- fisis tibial retrolateral tiene una prolongacion conica dorsal, el borde superior oblicuo y la punta anterior esta dirigida hacia la base de la tibia; embolo con una envoltura basal mem- branosa, que abarca la base del conductor. Descripcion. — Holotipo macho: Largo to- tal 6.78. Prosoma largo 3.67, ancho 3.07, alto 1.87. Clipeo, alto 0.27. Area ocular largo 1.70, ancho de hilera anterior 2.63, de hilera pos- 38 THE JOURNAL OF ARACHNOLOGY Mapa 2. — Distribucion de las especies de Freya del gmpo decorata. America del Sur, aproximadamente entre los paralelos 15°S y 37°S. tenor 2.50. Distancias OLA-OMP 0.43, OMP- OLP 0.40. Diametro OMA 0.83. Formula de patas y quetotaxia como en el genero; las ti- bias III y IV tienen d lb-2ap. Palpos (Figs. 6, 10, 21, 25, 26, 34): cimbio sin depresion re- trodorsal basal; embolo mas grueso y mas rec- to que en F. decorata; conductor con apice curvado, relativamente ancho; base de embolo y conductor rodeados por un reborde mem- branoso, que forma una punta visible desde cara retrolateral. Color: como en F. decorata; la banda media de pelos blancos en RT puede haber estado unida a las laterales; banda me- dia longitudinal del opistosoma unida a la ba- sal. Alotipo hembra: Largo total 9.26. Prosoma largo 3.73, ancho 2.87, alto 1.93. Clipeo, alto 0.23. Area ocular largo 1.53, ancho de hilera anterior 2.50, de hilera posterior 2.47. Distan- cias OLA-OMP 0.47, OMP-OLP 0.40. Dia- metro de OMA 0.80. Formula de patas III-IV- I-II. Quetotaxia como en el genero. Epigino (Figs. 39, 54, 55): gran escotadura en horde posterior, bolsillos laterales de anclaje bien se- GALIANO— LAS ESPECIES DE FREYA DEL GRUPO DECORATA 39 Figuras 49-53. — Epiginos clarificados, vista ventral. 49. Ereya decorata; 50. E. rubiginosa; 51. E. dureti; 52. E. regia; 53. E. chapare. Escalas =100 |xm. parados. Area anterior media muy granulada, con abundantes rugosidades radiantes; orifi- cios de copulacion elipticos, oblicuamente si- tuados, con margen anterior poco marcado, conductos de copulacion bien separados, es- permatecas esfericas. Color: como en F. de- corata. La banda media longitudinal del opist- osoma no se une a la basal; dorso entre bandas, con pelos pardo dorado o rojizo. Pal- pos amarillos, con tarsos algo mas oscuros, sin manchas dorsales basales pardas en los arte- jos. Distribucion. — Solo de la localidad tipo. AGRADECIMIENTOS Expreso mi agradecimiento a los encarga- dos de colecciones que me enviaron impor- tante material para su estudio: Dr. M. Moritz (ZMB), Mr. P.D. Hillyard (NHM), Ms. L. Lei- bensperger (MCZ), Dr. J.R Jass (MPM), Drs. 1. Berdondini y S. Whitman (MLS), Dr. C. Rollard (MHNP) y Dr. J.L. Moreira Leme (MZSP); al Dr. J.W. Berry por su apoyo y a los revisores Drs. G. Hormiga, J. Coddington 40 THE JOURNAL OF ARACHNOLOGY Figuras 54-57. — Epiginos clarificados. 54. Freya atures, ventral; 55. El mismo, dorsal; 56. F. nigro- taeniata, ventral; 57. El mismo, dorsal. Escalas = 100 |jLm. y C.L. Scioscia sus valiosos comentarios so- bre el manuscrito. Agradezco a las Tecnicas Patricia Sarmiento (MLP) las fotografias de microscopio electronico de barrido y Susana Ledesma (MACN) la atencion de los animales vivos. LITERATURA CITADA Bonnet, P. 1955. Bibliographia Araneorum. Tou- louse, Vol. 2(1):1-918; 1956, 2(2):919-1925; 1959, 2(5):423 1-5058. Brignoli, P.M. 1983. A Catalogue of the Araneae Described Between 1940 and 1981. 755 pp. Manchester Univ. Press. Brignoli, P.M. 1985. On the correct dates of pub- lication of the arachnid taxa described in some works by C.W. Hahn and C.L. Koch (Arachnida). Bulletin of the British Arachnological Society, 6(9):414-416. Cambridge, EO.R 1901. Biologia Central!- Ameri- cana. Arachnida. Araneidea and Opiliones, 2: 193-312. Caporiacco, L. 1938. Aracnidi del Messico, di Guatemala e Honduras Britannico. Atti della So- cieta Italiana di Scienze Natural!, 77:251-282. Caporiacco, L. 1947. Diagnosi preliminari di spe- cie nuove di arachnid! della Guiana Britannica raccolte dal Professor! Beccari e Romiti. Moni- tore Zoologico Italiano, 56(1-6): 16-34. GALIANO— LAS ESPECIES DE FREYA DEL GRUPO DECORATA 41 Caporiacco^ L= 1948. Arachnida of British Guiana collected in 1931 and 1936 by Professors Beccari and Romiti. Proceedings of the Zoological So- ciety of London, 1 18(3):607”747. Caporiacco, L. 1954. Araignees de la Guyane Frangaise du Museum d'Histoire Naturelle de Paris. Commentationes, 16(3):45-193. Pontificia Academia Scientiarum. Chickering, A. 1946. The Salticidae (spiders) from Panama. Bulletin of the Museum of Comparative Zoology 97:1-474. Galiano, M.E. 1963 a. Las variaciones individuales en Euophrys sutrix Holmberg, 1874 (Araneae, Salticidae). Revista de la Sociedad Entomologica Argentina, 24:23-28. Galiano, M.E. 1963b. Las especies americanas de aranas de la familia Salticidae descriptas por Eu- gene Simon. Redescripciones basadas en los ejemplares tfpicos. Physis (Buenos Aires) Sec. C, 23(66):230-470. Galiano, M.E. 1981. Catalogo de los especfmenes tfpicos de Salticidae (Araneae) descriptos por Candido de Mello-Leitao. Segunda parte. Physis (Buenos Aires) Sec. C, 39(97): 1 1-17. Galiano, M.E. 1982, Revision del genero iVycerelto (Araneae, Salticidae). Physis (Buenos Aires) Sec. C, 41(100):53-63. Koch, C.L. 1846. Die Arachniden, 13;L234. Ntim- berg. Koch, C.L, 1850. Uebersicht des Arachnidensys- terns, 5:1-104. Niimberg. Mello-Leitao, C. 1941. Aranas de la Provincia de Santa Fe colectadas por el Profesor Biraben. Re- vista del Museo de La Plata, Zoologfa 2(13): 199-225. Mello-Leitao, C. 1942. Aranas de Chaco y Santi- ago del Estero. Revista del Museo de La Plata, Zoologfa, 2(16):38L426. Mello-Leitao, C. 1945. Aranas de Misiones, Cor- rientes y Entre Rfos. Revista del Museo de La Plata, Zoologfa, 4(29):213-302. Mello-Leitao, C. 1948. Contribugao ao conheci- mento da fauna Araneologica das Guianas. An- naes da Academia Brasileira de Sciencias, 20(2): 151-196. Neave, S.A. 1939-1940. Nomenclator Zoologicus. A list of the names of genera and subgenera in Zoology from the tenth edition of Linnaeus 1758 to the end of 1935. 2:1-1025, 4:1-758. London. Peckham, G.W. & E.G. Peckham. 1893. On the Spiders of the Family Attidae of the Island of St. Vincent. Proceedings of the Zoological Society of London, 692-704 pp. Peckham, G.W. & E.G. Peckham. 1896. Spiders of the Family Attidae from Central America and Mexico. Occasional Papers of the Natural His- tory Society of Wisconsin, 3:1-101. Penther, Dr. 1900. Beschreibung der Euophrys de- corata C.L. Koch var. dyscrita nov. vai. In Sii- damerika gesammelte Myriapoden und Arach- eoideen. Zoologischer Anzeiger 23(615):285- 286. Petrunkevitch, A. 1911. A synonymic index-cata- logue of spiders of North, Central and South America with all adjacent Islands, Greenland, West Indies, Terra del Fuego, Galapagos, etc. Bulletin of the American Museum of Natural History 29:1-790. Petrunkevitch, A. 1928. Systema Aranearum. Transactions of the Connecticut Academy of Arts and Sciences 29:1-270. Platnick, N.L 1989. Advances in Spider Taxono- my. 1981-1987. 673 pp. Manchester Univ. Press. Platnick, N.L 1993. Advances in Spider Taxono- my. 1988-1991. 846 pp. New York Entomolog- ical Society, New York. Platnick, N.L 1997. Advances in Spider Taxonomy 1992-1995. With redescriptions 1940-1980. 976 pp. New York Entomological Society, New York. Platnick, N.L & M.U. Shadab. 1975. A revision of the spider genus Gnaphosa (Araneae, Gnapho- sidae) in America. Bulletin of the American Mu- seum of Natural History 155(1): 1-66. Proszyhski, J. 1971. Catalogue of Salticidae (Ara- nei) specimens kept in major collections of the world. Annales Zoologici, Warszawa 28(17): 367-519. Proszyhski, J. 1990. Catalogue of Salticidae (Ara- neae). Synthesis of quotations in the v/orld lit- erature since 1940, with basic taxonomic data since 1758. Wyzsza Skola Rolniczo-Pedagogicz- na w Sieldcach. 366 pp. Siedlce (Poland). Roewer, C.F. 1954. Katalog der Araneae von 1758 bis 1940, bzw. 1954. 2b:925-175L Brussels. Simon, E. 1864. Histoire Naturelle des Araignees (Araneides). 540 pp. Paris. Simon, E. 1902. Etudes Aracheologiques, 32® Me- moire. Descriptions d'especis nouvelles de la fa- mille des Salticidae. (Suite). Annales de la So- ciete Entomologique de France 71:389-421. Simon, E. 1903. Histoire Naturelle des Araignees 2(4):669-1080. Paris. Taczanowski, L. 1871. Les araneides de la Guyane Frangaise. Trudy rousskago entomologhieskago Obtchiestwa (= Horae Societatis Entomologicae Rossicae) 8:32-132. Walckenaer, C.A. 1805. Tableau des Araneides ou Caracteres essentiels des tribus, genres, families et races que renferme le genre Aranea de Linne, avec la designation des especes comprises dans chacune de ces divisions. Paris, pp. I— XII, 1—88. Manuscript received 16 March 2000, revised 2 Oc- tober 2000. Proofreading was done by C. L. Scioscia. 2001. The Journal of Arachnology 29:42-46 DESCRIPTION OF A NEW SPECIES IN THE NITIDULUS GROUP OF THE GENUS VAEJOVIS (SCORPIONES, VAEJOVIDAE) E. Michelle Capes: Department of Life, Earth, and Environmental Sciences, WTAMU Box 60808, West Texas A&M University, Canyon, Texas 79016 USA ABSTRACT. A new species in the nitidulus group of Vaejovis is described: V. mauryi from Sonora, Mexico. Morphological characters, including the hemispermatophore of the holotype male, are illustrated. The species is compared to Vaejovis decipiens, Vaejovis janssi, and Vaejovis intermedius. Keywords: Scorpion, Vaejovidae, Sonora, Mexico, taxonomy Vaejovis is the most diverse genus of scor- pions in North America, with 66 described species arranged into five species groups (Sis- som 2000). Although a comprehensive revi- sion of the genus is not available at the present time, the genus has recently been catalogued (Sissom 2000). Since the appearance of the catalogue one additional species has been de- scribed from Sonora, Mexico (Hendrixson 2001). The Vaejovis nitidulus group is a moderate- ly diverse group with 15 species found from the southern parts of Texas, USA, and through much of Mexico (Sissom & Francke 1985; Sissom 1991; Sissom 2000). Members of the group share the following characteristics: (1) the anterior margin of the carapace is obtusely emarginate, with a distinct median notch; (2) the genital opercula of the female possess a membranous longitudinal connection on the anterior two-thirds to four-fifths; (3) the pec- tinal teeth of the female are all subequal in size; (4) the ventral submedian carinae of the metasoma are obsolete to moderate and cren- ulate; (5) the cheliceral movable finger bears a well developed serrula on the ventrodistal aspect; (6) the pedipalps are relatively elon- gated, with chela length/width ratios greater than 3.3 and usually greater than 4.0; (7) the pedipalp chela fingers in most species termi- nate in enlarged claw-like denticles bearing an apical white patch; (8) chela trichobothria ib and it are located at the base of the fixed fin- ger; (9) the denticle row of the pedipalp chela fixed finger is divided into six or seven sub- rows; (10) the dorsointemal carina of the ped- ipalp chela is strong and, in most species. bears enlarged, sharp granules; (11) the ven- tral spinule row of the telotarsus is flanked distally by a single pair of larger spinules; (12) the male hemispermatophore bears a two- pronged hook at the base of the distal lamina; and (13) the distal margin of the sperm plug is smooth, i.e., devoid of hooks or spines (Sis- som & Francke 1985; Sissom 1991). The only species in the Vaejovis nitidulus group previously reported from the state of Sonora, Mexico is Vaejovis decipiens Hoff- mann 1931; this record is based on two ju- venile females (Sissom 1991). While perusing the collections at the California Academy of Sciences, W.D. Sissom found three specimens representing an undescribed species in the ni- tidulus group from this state. These speci- mens, which were subsequently made avail- able to me for description, were discovered after years of careful examination of museum material from the United States and around the world (Sissom pers. com., September 1999). The rarity of this species in musem col- lections may be due to its probable lithophilic habits, which make it difficult to collect with conventional rock-rolling techniques. METHODS Terminology for general morphology con- forms to that of Stahnke (1970) with the fol- lowing exceptions: terminology for metaso- mal and pedipalpal carinae is after Francke (1977); and trichobothrial nomenclature fol- lows Vachon (1974), except that the fourth pe- dipalpal segment is considered the patella rather than the tibia, adhering to Stahnke ’s ter- minology. 42 CAPES~NEW SCORPION FROM MEXICO 43 Vaejovis mauryi new species (Figs. 1-11) Type data*—Holotype male, paratype fe^ male, and paratype subadult female from So- nora, Mexico, 28°55'N, 109°45^W, 18 Septem- ber 1982 (V. Roth). Deposited at California Academy of Sciences, San Francisco. Etymology.- — The specific name is a pa- tronym honoring the late Emilio A. Maury for his contributions to the field of aracheology. Distribution.^Knowe only from the type locality. According to maps, this locality lies in the vicinity of Mazatan, Bacaeora and Soy- pa in the state of Sonora, Mexico. Diagnosis.— Within the nitidulus group, Vaejovis mauryi is most similar to Vaejovis decipiens Hoffmann 1931, F. janssi Williams 1980, and V. intermedius Borelli 1915. It can be easily distinguished from V, decipiens and F janssi by (1) the presence in F decipiens and F janssi of ventral submedian carinae on metasomal segments I-II, with these carinae stronger on III-IV; (2) the presence of granu- lation in the ventral median intercarinal space in F. mauryi; (3) weaker digital and external secondary carinae of the pedipalp chelae in F. mauryi; (4) the presence in F. decipiens and F janssi of strong lateral keels on stemite VII; (5) higher pectinal tooth counts in both F de- cipiens (22-25 in males, 21-22 in females) and F. janssi (21-22 in males, 18-21 in fe- males); and (6) the noticeable difference in size, with F. mauryi being smaller. Vaejovis mauryi can be distinguished from Vaejovis intermedius by (1) the sparseness of setation on the pedipalp chelae, metasoma, and stemite VII (in F. intermedius these sur- faces are very hirsute); (2) the dorsolateral ca- rinae of F intermedius are serrate, whereas those of F mauryi are crenulate; (3) the pres- ence of only weak scalloping in the chela fin- gers of F. mauryi (distinct scalloping in F. intermedius); and (4) higher pectinal tooth counts in F. intermedius (21-26 in males, 19- 23 in females). Measurements.— Holotype, in mm: total length 35.90; carapace length 4.60; mesosoma length 9.30; metasoma length 17.20. Metaso- ma: segment I length/width 2.20/2.85; seg- ment 11 length/width 2.60/2.85; segment III leegth/width 2.75/2.75; segment IV length/ width 3.70/2.65; segment V length/width 5.95/2.35. Telson: length 4.90; vesicle length/ width/depth 3.05/1.8/1.4; aculeus length 1.9. Pedipalps: total length 16.10; femur length/ width 4.35/1.25; patella length/width 4,55/ 1.45; chela leegth/width/depth 7.20/1.65/1.9; movable finger length 4.60; fixed finger length 3.80. Description. — -Based on holotype. Colora- tion (in alcohol): Base color of carapace and tergites yellow-brown to orange-brown with an underlying dusky pattern. leterocular area darkly pigmented. Metasoma light orange to dark orange. Telson vesicle orange or reddish- brown. Legs orange, with dusky markings proximally; basitarsi and telotarsi uniformly yellow. Prosoma: Anterior margin of carapace obtusely emarginate. Median notch shallow. Interocular area finely granular with scattered coarse granules. Remainder densely granular. Mesosoma: Median carina on LII obsolete; on III feeble; on IV- VI weak, granular. On VII, mediae carina weak, granular; lateral carinae strong, crenulate to serrate, with distal denticle enlarged. Pectinal tooth count 19-19. Ster- nites III- VI sparsely setose; VII with two weak, finely granular lateral carinae. Metaso- ma: Ratio of segment I length/width 0.76; of segment III length/width LOO; of segment V length/width 2.50. Segments I-IV: dorsolateral carinae strong, finely crenulate, with distal- most denticle of I slightly enlarged, spinoid; on II-IV distinctly enlarged and spinoid dis- tally. Lateral supramedian carinae on I-III strong, finely crenulate; on IV moderate, gran- ular with distalmost denticles on I-III enlarged and spinoid. Lateral inframedian carinae on I complete, strong, irregularly crenulate; on II present on anterior half as isolated granules, on posterior one-half, weak to moderate, gran- ular to crenulate; on III present on posterior one-third, moderate, finely crenulate; on IV absent. Ventrolateral carinae on I-II moderate, smooth to finely granular; on III-IV moderate, irregularly, finely serratocrenulate. Ventral submedian carinae on I and II obsolete; on III weak, feebly granular; on IV weak, irregularly granular. Dorsal and lateral intercarinal spaces sparsely, coarsely granular. Veetromediae in- tercarieal space on IV granulose. Setal count on segments I-IV: dorsolateral setae 0/0: 1/1:1/ 1:2/2; lateral supramedian setae l/l:2/l:2/2:3/ 3; lateral inframedian setae 2/2:l/l:l/l:0/0; ventrolateral setae 3/3:3/3:3/3:4/4; ventral submediae setae 373:4/4:4/4:4/4. Segment V: (Fig. 1) Dorsolateral carinae moderate, cren- 44 THE JOURNAL OF ARACHNOLOGY Figures 1-9. — Morphology of Vaejovis mauryi (all drawings of holotype male). 1. Lateral view of metasomal segments IV and V and the telson; 2. Dorsal aspect of pedipalp femur; 3. External aspect of pedipalp patella; 4. Dorsal aspect of pedipalp patella; 5. Dentition of pedipalp chela fixed finger; 6. Dentition of pedipalp chela movable finger; 7. Ventral aspect of pedipalp chela; 8. External aspect of pedipalp chela; 9. Dorsal aspect of pedipalp chela. ulate basally, granular distally. Lateromedian carinae weak, granular, present on anterior 3/4. Ventrolateral carinae strong, serrate. Ven= tromedian carina strong, crenulate. Intercar= inal spaces with scattered, coarse granules. Segment V setal count: dorsolateral setae 5/6; lateromedian setae 4/4; ventrolateral setae 7/9. Telson: (Fig. 1) Ventral aspect with irreg- CAPES— NEW SCORPION FROM MEXICO 45 ular punctations and granulation. Ventral mid- line with small granules terminating in a sub- tle subaculear tubercule. Nine pairs of large setae, with several smaller setae, Pedipalps: Trichobothrial pattern type C, orthobothriotax- ic. Pedipalpal ratios: chela length/width 4.20; femur length/width 3.38; fixed finger length/ carapace length 0.83. Femur: (Fig. 2) carinae strong, granulose; internal face with 8-10 large, pointed granules, with scattered fine granules. Patella: (Figs. 3, 4) Dorsointemal, ventrointernal, and ventroexternal carinae strong, crenulate. Internal face with oblique longitudinal carina of 8 large, serrated gran- ules and 10 smaller granules. Chela: (Figs. 7, 8, 9) Dorsal marginal carina strong, crenulate. Dorsal secondary carina moderate, smooth. Digital carina moderate, smooth. External sec- ondary carina weak, smooth. Ventroexternal carina moderate, granular. Ventromedian ca- rina vestigial. Ventrointernal carina moderate, smooth. Dorsolateral carina strong, with large, crenulate granules. Dentate margin of fixed finger (Fig. 5) with primary denticle row di- vided into six subrows by five enlarged den- ticles; six inner accessory granules. Dentate margin of chela movable finger (Fig. 6) with primary denticle row divided into six subrows by five enlarged denticles; seven inner acces- sory granules. Fingers without distinct scal- loping. Hemispermatophore: (Figs. 10, 11) Distal lamina slightly longer than trunk, not distinctly tapered. Median lobe relatively large, rounded. Variation.— Only three specimens were available for study. These included one adult male, one adult female, and one subadult fe- male. The adult female is better preserved than the male and may therefore be closer to the actual coloration of the species. Base color of carapace and tergites deep orange-brown to yellow-brown with underlying dusky pattern. Metasomal segment V slightly darker than the preceding segments. Legs yellow-brown with mottling proximally; basitarsi and telotarsi uniformly yellow. Interocular area of female smooth with scattered coarse granules. Re- mainder of prosoma sparsely granular. Ventro- lateral carinae smooth to finely granular on I- II; moderate, finely serrate on III-IV. Ventrolateral carinae in juvenile female para- type moderate, finely serrate on I-II; moderate, irregularly serratocrenulate to finely serrate on Figures 10-11. — Morphology of the hemisper- matophore of Vaejovis mauryi (holotype male). 10. Dorsal aspect of left hemispermatophore; 11. Cap- sular area of left hemispermatophore. III-IV. Pectinal tooth count 17-17 in both fe- male paratypes. Selected measurements (in mm) of the par- atype female are as follows: total length 40.60; carapace length 5.70; mesosoma length 12.25; metasoma length 17.30; metasoma seg- ment III length/width 2.70/3.05; segment V length/width 6.45/2.80; chela length/width/ depth 8.95/1.85/2.10; fixed finger length 4.90; movable finger length 5.95. Setal counts of the adult and subadult fe- males are as follows (L/R): Dorsolaterals: 0/ 0:1/1: 1/1 :2/2 and 0/0:l/l:l/l:2/2. Lateral su- pramedians: 0/l:2/l:2/l:4/4 and 0/0:l/l:2/2:3/ 3. Lateral inframedians: 2/3:l/l:0/l:0/0 and 21 2:l/l:0/0:0/0. Ventrolaterals: 2/2: 3/3: 3/4: 5/5 and 3/3: 3/3: 3/3: 4/3. Ventral submedians: 3/3: 4/4:4/4:4/5 and 3/3: 4/4: 4/4: 4/4. Setal counts on V are as follows: dorsolaterals: 2/5 and 5/ 5; lateromedians: 3/4 and 4/4; ventrolaterals: 8/7 and 7/8. Specimens examined. — MEXICO: Sonora, 28°55'N, 109°45'W, (pine forest which, according to maps of the area, must be located between Ma- zatan, Bacanora, and Soyopa), 18 September 1982 (V. Roth), 1 male holotype, 1 female paratype, 1 juvenile paratype (CAS). 46 THE JOURNAL OF ARACHNOLOGY ACKNOWLEDGMENTS I wish to thank David Sissom for his guid- ance during the course of this study, and for his assistance with the illustration of the cap- sular area of the hemispermatophore. I would also like to thank Charles Griswold of the Cal- ifornia Academy of Sciences (CAS) for the loan of material to Dr. Sissom, who, in turn, made them available to me for study. Dr. Douglas P. Bingham, chair of the Department of Life, Earth, and Environmental Sciences at West Texas A&M University, provided funds to cover reprint costs. LITERATURE CITED Francke, O.F. 1977. Two emendations to Stahnke’s (1974) Vaejovidae revision (Scorpionida, Vae- jovidae). Journal of Arachnology 4:125-135. Hendrixson, B.E. 2001. A new species of Vaejovis (Scorpiones, Vaejovidae) from Sonora, Mexico. Journal of Arachnology 29:47-55. Hoffmann, C.C. 1931. Los scorpiones de Mexico. Primera parte: Diplocentridae, Chactidae, Vejov- idae. Anales del Institute de Biologia de la Univ- ersidad Nacional Autonoma de Mexico 8:291- 408. Sissom, W.D. 1991. Systematic studies on the ni- tidulus group of the genus Vaejovis, with de- scriptions of seven new species (Scorpiones, Vaejovidae). Journal of Arachnology 19:4-28. Sissom, W.D. 2000. Family Vaejovidae Thorell, 1876. Pp. 503-553. In Catalog Of The Scorpions Of The World (1758-1998). (V. Fet, et af, eds.). New York Entomological Society. Sissom, W.D. & O.F Francke. 1985. Redescrip- tions of some poorly known species of the niti- dulus group of the genus Vaejovis (Scorpiones, Vaejovidae). Journal of Arachnology 13:243- 266. Stahnke, H.L. 1970. Scorpion nomenclature and mensuration. Entomological News 81:297-316. Vachon, M. 1974. Etude des caracteres utilises pour classer les families et les genres de Scor- pions (Arachnides). 1. La trichobothriotaxie en Arachnologie, Sigles trichobothriaux et types de trichobothriotaxie chez les Scorpions. Bulletin du Museum National d’Histoire Naturelle, Paris, (Sen 3), 140 (Zool. 104), mai-juin 1973:857- 958. Manuscript received 1 February 2000, revised 10 October 2000. 200 L The Journal of Arachnology 29:47-55 A NEW SPECIES OF VAEJOVIS (SCORPIONES, VAEJOVIDAE) FROM SONORA, MEXICO Brent E« Hendrixson: Department of Life, Earth, and Environmental Sciences, West Texas A&M University, WTAMU Box 60808, Canyon, Texas 79016^0001 USA ABSTRACT. Vaejovis pequeno, a new species of scorpion previously confused with V. waueri Gertsch & Soleglad 1972, is described and illustrated from Sonora, Mexico and is compared to that species. A revised diagnosis and new distributional records for V. waueri, a member of the eusthenura group, are provided. Keywords: Scorpion, Vaejovidae, Sonora, Mexico, Vaejovis is the most diverse genus of scon pions in North America with 66 described species and five recognized species groups (i.e., eusthenura, intrepidus, mexicanus, nitu- dulus and punctipaipi; Sissom 2000). To date, there is no single comprehensive revision of the genus, but a number of partial revisions and regional faunas can be found in the lit^ erature (Williams 1970a, 1970b, 1971, 1980; Soleglad 1973; Sissom & Francke 1985; Sis- som 1991a; Capes 2001). Gertsch & Soleglad (1972) described a small, attractive scorpion, Vaejovis waueri, on the basis of specimens from southern Texas in the USA, and from the states of Nuevo Leon and Sonora in Mexico. This species is well known in the eastern regions (i.e., Texas and Nuevo Leon), but the Sonoran record from Rio Cuchajaqui represents a significant dis- junction. Sissom (1991b) suggested that hu- mans may have accidentally introduced the species to Sonora, supported by the fact that no additional specimens had been collected from the well-sampled Alamos area in the southeastern comer of the state. At the time, only subtle differences in morphology were detected between the specimens from Sonora and Texas including slight variations in the ventrolateral carination of metasomal segment V. My findings, based on subsequent exami- nation of the Rio Cuchajaqui material and some new material that has since accumulat- ed, indicate that these specimens are distinct from V. waueri and represent a new species. The superficial resemblance between the new species and U waueri is striking, so it is understandable that the earlier authors be- systematics lieved these Sonoran specimens to be V, waueri, despite the disjunction. Both of these diminutive species possess relatively lustrous cuticles and exhibit similar color patterns (e.g., dusky, mottled appearance with lighter medial stripe), morphometries, and carination. In addition to describing the new species and comparing it to V. waueri, it is my purpose here to provide a revised diagnosis for V. waueri and to review the species distribution based on many new records. METHODS Nomenclature and mensuration for the most part follow that of Stahnke (1970), with the following exceptions: carinal terminology is after Francke (1977) and trichobothrial ter- minology is after Vachon (1974), except that the fourth pedipalpal segment is considered the patella rather than the tibia, adhering to Stahnke ’s nomenclature. Morphometric char- acters were derived from measurements of a single adult male and nine adult females for the new species, and from 10 adult males and 10 adult females for V. waueri; several (« > 50) additional adult V. waueri were used to provide a revised diagnosis for that species. Hemispermatophore preparation follows that of Sissom et al. (1990). All measurements were taken using an ocular micrometer cali- brated at 20 X, and illustrations were carried out by the use of an ocular grid. Specimen depository designations are as follows: American Museum of Natural His- tory, New York (AMNH); Academy of Nat- ural Sciences, Philadelphia (ANS); Appala- chian State University, Boone, North Carolina 47 48 THE JOURNAL OF ARACHNOLOGY Figure 1. — Map of the southwestern United States and northern Mexico depicting the distribution of Vaejovis pequeno new species (o) and V. waueri (•). (ASU); Florida State Collection of Arthro- pods, Gainesville (FSCA); California Acade- my of Sciences, San Francisco (CAS); per- sonal collection of J. A. Nilsson (JAN); University of Arizona, Tucson (UA); Univer- sity of Arkansas, Fayetteville (UAF); personal collection of W. David Sissom (WDS); West Texas A&M University, Canyon (WTAMU). Vaejovis pequeno new species Figs, 1, 2, 4-15 Vejovis waueri, in part: Gertsch & Soleglad 1972: 607 (misidentification). Vaejovis waueri: Sissom 1991b: 215 (misidentifi- cation). Type data. — Adult male holotype collected 15 mi W Yecora (4000 feet) (1200 m), So- nora, Mexico on 7 August 1986 by V. Roth; deposited at the California Academy of Sci- ences, San Francisco. For paratype data, see “Specimens examined” section. Etymology. — The specific epithet comes from the Spanish pequeno (meaning “little one”) and refers to the minute size of this species; pequeno is regarded as a noun in ap- position. Distribution. — Known from several local- ities in Sonora, Mexico (Fig. 1). Diagnosis. — Vaejovis pequeno (Fig. 2) may be distinguished from V. waueri by being somewhat smaller in size (adults of V. pe- queno to 19.85 mm and V. waueri to 24.8 mm); possessing weak, granular dorsal mar- ginal and dorsointemal carinae on the pedi- palp chelae in males (rather than all keels ob- solete); the inner lobe of the hemispermatoph- oric capsule without barbs (rather than dis- tinctly barbed); the chelicerae with a strongly developed serrula (rather than a weakly de- veloped serrula); one pair of distal spinules flanking the ventromedian spinule row of the tarsi (rather than more than one pair); consis- tently lower setal counts on the metasoma (see Figs. 3,4); and finely granular to weak, serrate ventral submedian carinae on metasomal seg- ment IV (rather than obsolete). The holotype male can be differentiated from V. waueri males {n = 10) by the following morphomet- ric ratios {V. waueri ratios in parentheses; slightly overlapping ratios have been includ- ed): pedipalp femur length/width 2.82 (2.46- 2.82), carapace length/metasomal segment V length 1.00 (0.80-0.90), and chela fixed finger length/carapace length 0.59 (0.50-0.56). The paratype females {n — 9) can be differentiated from V. waueri females (n = 10) by the fol- HENDRIXSON— NEW SCORPION FROM MEXICO 49 Figure 2. — Photograph showing dorsal view of paratype female (left) and holotype male of Vaejov- is pequeno new species. Photograph by W.D. Sis- som. lowing morphometric ratios: pedipalp femur length/width 2.60-3.00 (2.41-2.60), chela movable finger length/chela length 0.60-0.63 (0.56-0.60), carapace length/metasomal seg- ment V length 1.00-1.04 (0.90-0.98), and chela fixed finger length/carapace length 0.58-0.65 (0.50-0.56). The placement of V. pequeno in an existing species group is problematic as its closest rel- ative is unknown. Although superficially sim- ilar to V. waueri, a member of the eusthenura group as defined by Williams (1970b), V. pe- queno clearly does not belong to that group based on the chelicerae possessing a strongly pronounced serrula, one pair of distal spinules flanking the ventromedian spinule row of the tarsi, and the absence of barbs on the capsule of the hemispermatophore. In addition, spe- cies of the eusthenura group that possess strong mottling and fairly robust metasomal segments [e.g., V. waueri, V. bilineatus Po- cock 1898, V. spinigerus (Wood 1863), and V. gravicaudus Williams 1970], all have ex- tremely setose metasomal carinae. Vaejovis pequeno is also superficially sim- ilar to Serradigitus agilis Sissom & Stockwell 1991 of northern Sonora and southern Arizo- na, but does not possess the synapomorphies associated with that genus. In particular, V. pe- queno does not bear serrated denticle rows on the pedipalp chela fingers, enlarged terminal denticles on the chela fingers, or enlarged proximal pectinal teeth devoid of sensilla in the female. Further, S. agilis has higher pec- tinal tooth counts than V. pequeno (14-17 in males instead of 12-13, and 14-15 in females instead of 11-13); and finally, trichobothria ib and it flank the sixth inner accessory denticle on the chela fixed finger in S. agilis, but it is slightly basal to the sixth inner accessory den- ticle in V. pequeno. Description.- Based on adult male holo- type (Fig. 2). Coloration (in alcohol): Base color orange-yellow to orange with underlying dusky pattern. Mesosomal dorsum with dis- tinct medial, longitudinal stripe. Metasomal segments progressively darker distally from yellow-orange to reddish-brown. Telson yel- low-orange; aculeus orange-red. Ventral sur- face cream colored. Pedipalp chela with dusky longitudinal stripes marking positions where keels generally occur. Legs yellow with some dusky mottling. Prosoma: Anterior edge of carapace slightly emarginate. Carapace sur- face densely, finely granular. Mesosoma: Ster- nite VII with weak, serrate lateral keels. Ter- gites densely, finely granular. Pectinal tooth count 13/12 (hr). Metasoma: (Fig. 4) Seg- ments I-III wider than long, IV longer than wide, V 1.76 times longer than wide. Seg- ments I-IV: intercarinal regions finely to coarsely granular. Dorsolateral and lateral su- pramedian carinae strong, crenulate to serrate terminating in an enlarged, spinoid tubercle (except on lateral supramedians of IV, which are widely flared). Lateral inframedian carinae on I moderate, crenulate; on ILIII weak to moderate, granular to crenulate; on IV absent. Ventrolateral carinae on I weak to moderate, crenulate; on II-III moderate, crenulate; on IV strong, crenulate. Ventral submedian carinae on I obsolete; II-III weak, smooth to granular; on IV weak, serrate. Segment V: intercarinal regions finely to coarsely granular. Dorsolat- eral carinae strong, irregularly crenulate. La- teromedian carinae moderate to weak, granu- lar. Ventrolateral and ventromedian carinae strong, crenulate to serratocrenulate. Seg- ments I-IV carinal setation (1/r): dorsolaterals. 50 THE JOURNAL OF ARACHNOLOGY l/0:l/l:l/l:2/2; lateral supramedians, 0/0: 1/1: l/l:2/3; lateral inframedians, l/l:0/0:0/0:0/0; ventrolaterals, 2/2: 3/3: 3/3: 3/3; ventral subme- dians, 3/3: 3/3: 3/3: 3/4. Segment V carinal se- tation: dorsolaterals, 5/5; lateromedians, 3/3; ventrolaterals, 5/5; ventromedians: 5/5. Tel- son: (Fig. 4) Surface smooth to weak, granu- lar; moderately setose. Vesicle dorsoventrally compressed, flattened dorsally. Subaculear tu- bercle minute, rounded and flanked by two large setae. Pedipalps: Orthobothriotaxic, Type C (Vachon 1974). All surfaces densely, finely granular. Femur (Fig. 5) tetracarinate: moderate, granular to crenulate. Patella (Figs. 6, 7) with dorsointemal and ventrointemal ca- rinae moderate, granular; internal carinae moderate, irregularly crenulate; dorsoextemal carinae weak, smooth to irregularly granular; ventroextemal carinae smooth. Chela (Figs. 8, 9) with finely granular dorsal marginal and dorsointemal carinae; others obsolete. Fixed finger (Fig. 10) with primary denticle row di- vided into five subrows by four enlarged pri- mary row denticles; six inner accessory den- ticles present; trichobothria ib and it situated just basal to sixth inner accessory denticle. Movable finger (Fig. 11) with primary denti- cle row divided into six subrows by five en- larged primary row denticles; seven inner ac- cessory denticles present. Cutting margin of both fingers straight (i.e., not scalloped). Hem- ispermatophore: (Figs. 12-15) Distal barb of mating plug without booklets; distal flange present. Measurements of male holotype: (mm) To- tal L, 14.55; carapace L, 2.20; mesosoma L, 3.35; metasoma L, 6.90; telson, 2.10. Meta- somal segments: I L/W, 0.90/1.35; II L/W, 1.10/1.20; III L/W, 1.10/1.25; IV L/W, 1.60/ 1.25; V L/W, 2.20/1. 25. Telson: vesicle L/W/ D, 1.30/1.10/0.60; aculeus L, 0.80. Pedipalps: total L, 6.00; femur L/W, 1.55/0.55; patella L/ W, 1.80/0.70; chela L/W/D, 2.65/0.65/0.65; fixed finger L, 1.45; movable finger L, 1.55. Measurements of female paratype from Ye- cora: (mm) Total L, 17.55; carapace L, 2.65; mesosoma L, 4.65; metasoma L, 7.85; telson, 2.40. Metasomal segments: I L/W, 1.10/1.65; II L/W, 1.25/1.50; III L/W, 1.30/1.50; IV L/W, 1.65/1.40; V L/W,2.55/1.45. Telson: vesicle L/ W/D, 1.50/0.95/0.75; aculeus L, 0.90. Pedi- palps: total L, 7.30; femur L/W, 1.90/0,75; pa- tella L/W, 2.15/0.85; chela L/W/D, 3.20/0.80/ 0.85; fixed finger L, 1.55; movable finger L, 2.00. Variation. — The single adult male differed from nine adult females in the following re- spects: body size is somewhat smaller (17.55- 19.85 mm in females) and the carination of the pedipalp chela is more pronounced (weak, smooth to irregularly granular in females). In addition, the holotype male can be differenti- ated from the paratype females by the follow- ing morphometric ratios (female ratios in pa- rentheses; slightly overlapping ratios have been included): chela movable finger length/ chela length 0,58 (0.60-0.63) and carapace length/metasomal segment V length 1.00 (1.00-1.04), Pectinal tooth counts of the ho- lotype male fell within the range of nine par- atype females: 13/12 (I/r) pectinal teeth in the male, 11-13 (mode = 12) in females. Addi- tional male material is needed to determine if pectinal tooth counts are significantly different between males and females. The following morphometric ratio ranges have been included to indicate intraspecific variation within the females (mean ± one standard deviation): chela length/width 3.81- 4.62 (4.00 ± 0.28), pedipalp femur length/ width 2.60-3.00 (2.74 ± 0.13), metasomal segment III length/width 0.81-0.93 (0.87 ± 0.04), and metasomal segment V length/width 1.61-1.80 (1.67 ± 0.07). Metasomal carination slightly variable in strength of the keels; however, the difference was not determined to be significant. Meta- soma carinal setation proved to be somewhat variable. Variation on segments LIV is as fol- lows: dorsolaterals, 0-1: 1:1:2; lateral supra- medians, 0-2:l-4:l-4:2-4; lateral inframedi- ans, 1-2:0- 1:0- 1:0 (sometimes accessory setae were present where the lateral inframedian keel would be located); ventrolaterals, 2-3:3: 3:3-4; ventral submedians, 3:3:3:3-4. Segment V carinal setation: dorsolaterals, 5; laterome- dians, 3-5; ventrolaterals, 5; ventromedians, 5. One female possessed five (instead of the typ- ical six) inner accessory denticles on the ped- ipalp chela fixed finger. Specimens examined. — MEXICO: Sonora: 15-20 km E Baviacora (29.43N, 1 10.05W), 6 August, no year (V. & B. Roth), 1 paratype female (CAS); Rio Cuchajaqui, E of Alamos, 14 January 1968 (V. Roth), 2 paratype females (AMNH); 7 mi NE Teso Paco (28.50N, 109.40W; thorn forest), 16 Septem- ber 1982 (V. Roth), 3 paratype females +18 first HENDRIXSON— NEW SCORPION FROM MEXICO 51 Figures 3-11. — 3. External morphology of Vaejovis waueri, female from San Angelo, Texas. Metasomal segments III-V and telson, showing carinal setal pattern, lateral aspect. 4-11. External morphology of Vaejovis pequeno new species, holotype male from Yecora, Sonora, Mexico; 4. Metasomal segments III- V and telson, showing carinal setal pattern, lateral aspect; 5. Pedipalp femur, dorsal aspect; 6. Pedipalp patella, dorsal aspect; 7. Pedipalp patella, external aspect; 8. Pedipalp chela, dorsal aspect; 9. Pedipalp chela, external aspect; 10. Pedipalp chela fixed finger, showing dentition and trichobothrial pattern; 11. Pedipalp chela movable finger, showing dentition. 52 THE JOURNAL OF ARACHNOLOGY Figures 12—15. — Morphology of right hemisper- matophore of Vaejovis pequeno new species. 12. Dorsal aspect; 13. Enlarged view of flange; 14. Ventral aspect showing capsule; 15. Mating plug (note the absence of booklets on distal barb). Ab- breviations: db = distal barb of mating plug; dl = distal lamina; fl = flange; mp = mating plug; tr = trunk. instar young (CAS); 15 mi W Yecora (4000 feet), 7 August 1986 (V. Roth), 1 holotype male 1 para- type female (CAS); 3.2 mi NW Huicochi (under rocks; 5200 feet), 11-14 June 1989 (S. Prchal), 1 paratype female +11 second instar young (WDS); Sierra Alamos above La Cieneguilla (1600-2000’), 11 October 1994 (RH. Holm), 1 paratype female (UA). Vaejovis waueri Gertsch & Soleglad Figs. 1, 3 Vejovis waueri: Gertsch & Soleglad 1972: 605, fig. 145, 146. Vaejovis waueri was originally described as a member of the spinigerus group by Gertsch & Soleglad (1972); however, V, spinigerus and its close relatives were actually placed in the eusthenura group by Williams (1980), and are still assigned there (Sissom 2000). Diagnosis. — For characters separating V. pequeno from V. waueri, see diagnosis for V. pequeno. Vaejovis waueri is most closely re- lated to V. bilineatus in the eusthenura group, but can be differentiated from that species by the following characters (V. bilineatus char- acters in parentheses; Yahia & Sissom 1996): somewhat smaller size, adults to 24.8 mm (22-32 mm); ventral submedian carinae of metasomal segments I-IV always obsolete (sometimes weak, crenulate on IV); a single dorsomedial stripe (two or four); modal pec- tinal tooth counts 14 in males, 12 in females (17 in males, 15 in females); metasomal seg- ment V length width 1.67-1.90 in males, 1.62-1.84 in females (2.00-2.38 in males, 1.78-2.18 in females); and male chela fingers with cutting margin straight (moderately scal- loped). Distribution. — Vaejovis waueri was previ- ously recorded from Texas in the USA (Gertsch & Soleglad 1972; Stockwell 1986) and from Nuevo Leon and Durango in Mexico (Sissom 2000). The current distribution, in- cluding old and new records, is shown in Fig. 1. This species has been collected between roughly 1000-2000 m in the Big Bend area of Texas where it is almost always located on rocky, boulder- strewn slopes. It is not known to burrow, but likely inhabits cracks and crev- ices among boulders and rocks during the day. Stockwell (1986), in an unpublished Mas- ter’s thesis, listed the following localities from Texas: Brewster County: Alpine, 8 mi S Al- pine, 4 mi W Marathon, 21 mi S Marathon, Big Bend National Park (N base of Grapevine Mt., base of Nugent Mt., K-Bar Road, Chisos Basin Pass, Chisos Mts., Chisos Basin, Rio Grande Village, Bouquillas Canyon); Crockett County: 10 mi N Iraan, 11 mi N Iraan, 45 mi NW Ozona, 22 mi E Iraan; Crosby County: 24 mi SE Crosbyton; Garza County: 1 mi ENE Justiceberg, 3 mi S Justiceberg; Jejf Da- vis County: Davis Mts. State Park, Fort Davis, 1.2 mi SW Hwy 17 on Hwy 1832; Kinney County: 21 mi N Brackettville, Brackettville; Maverick County: 7 mi S Spofford, 14 mi S El Indio; Pecos County: 12 mi N Ft. Stockton, 4 mi E Sheffield, 4 mi SE Sheffield, 52.5 mi NW Dryden, 15 mi N Sanderson; Starr Coun- ty: Rio Grande City; Terrell County: 19 mi S Sheffield, Chandler Ranch, 5 mi N Sanderson, HENDRIXSON— NEW SCORPION FROM MEXICO 53 72 mi W Pecos River on Hwy 90; Val Verde County: 21 mi N Comstock, 14 mi N Com- stock, 0.5 mi S Langtry; Webb County: 45 mi S El Indio, Laredo. New records. — UNITED STATES: Texas: Brewster County: Big Bend National Park (BBNP), Chisos Basin (W face of Casa Grande), 29 August 1984 (Sissom et al.) Id 1 $ (WDS); BBNP, Chisos Basin, 28 Sept 1950 (Gertsch), ld69 (AMNH); BBNP, Chisos Basin, 5 August 1938 (Mulaik), 1(36$ (AMNH); BBNP, Chisos Basin, 28 Septem- ber 1950 (Gertsch), 2S 1 $ (AMNH); BBNP, Chisos Basin, 28 May 1952 (no collector), 18$ (AMNH); BBNP, Chisos Basin, 26 July 1938 (Mulaik), 1$ (AMNH); BBNP, Chisos Basin, 16 July 1921 (Dun- can), 1 $ + young (AMNH); BBNP, Chisos Basin, 28 September 1950 (no collector), 2$ (AMNH); BBNP, Chisos Basin (6000’), 25 August 1967 (Gertsch & Hastings), 2d 1 $ (AMNH); BBNP, Chi- sos Basin, 22 August 1959 (McAlister), 1$ (AMNH); 10 mi N Hot Springs on Marathon Rd., 21 July 1938 (Mulaik) 1$ (AMNH); BBNP, Hot Springs parking lot and trail to Hot Springs, 24 June 1998 (Henson et ah). Id (ASU, Q-283A); BBNP, Basin, May 1983 (Henson), 1$ (ASU, G-119, 1091); BBNP, Pine Canyon Trail-Grassland, 24 May 1987 (Henson et ah), 1 $ (ASU, A-94, 0092); BBNP, Lost Mine Trail, 23 May 1987 (Henson et ah), 1$ (ASU, A-61, 0059); BBNP, Pine Canyon Road, 24 May 1987 (Henson et ah), 1 $ (ASU, A- 83); BBNP, Pine Canyon, end of wooded area to parking lot, 27 May 1992 (no collector), 1 $ (ASU, L-203); BBNP, Pine Canyon, end of wooded area to parking lot, 27 May 1992 (no collector), 1$ (ASU, L-205, 1916); BBNP, Dugout Wells, 19 May 1987 (Henson et al.), 1 $ (ASU, A-7, 0007); BBNP, Window Trail, 26 May 1987 (Henson et al.), 1 $ (ASU, A-118); BBNP, Glenn Spring Road, 19 May 1988 (Henson), 1$ (ASU, B-2-a-l, 0248); BBNP, Pine Canyon Trail, edge of grassland and pihon pine, 24 May 1987 (Henson et al.), 1$ (ASU, A- 77, 0075); BBNP, Window Trail below group camp- ground, 23 May 1987 (Henson), 1$ (ASU, A-58, 0057); BBNP, Basin, May 1983 (no collector), 1 juv. $ (ASU, G-117, 1089); BBNP, Mine Trail (6350’), 9 June 1991 (Henson et al.), 1 $ (ASU, H- 273, 1459); BBNP, end of Grapevine Hill Road near Grapevine Spring, 31 May 1990 (Henson & David), 1 $ (ASU, D-143, 0519); BBNP, Pine Canyon Trail above parking area, 27 May 1992 (Van Devender), 4$ (ASU, L-292-295, 2009-2012); BBNP, Window Trail, 20 May 1988 (Henson), 1$ (ASU, B-4-d-l, 0333); BBNP, Glenn Spring Road, 19 May 1988 (Henson), 1$ (ASU, B-2-d-l, 0264); BBNP, Win- dow Trail, 20 May 1988 (Henson), 1 $ (ASU, B-4- d-1, 0334); BBNP, end of Grapevine Hills Road near Grapevine Springs, 31 May 1990 (Henson & Davis), 1$ (ASU, D-138, 0514); Mikibbe Springs off Lost Mine Trail, 23 June 1998 (Henson et al.), 1 $ (ASU, Q-260); Pine Canyon Road, 24 May 1987 (Henson et ah), 1$ (ASU, A-82); 9 mi S Black Gap on EM 2627, 27 May 1991 (Davis), Id (ASU, J-97, 1582); Crockett County: Hwy 195 E Iraan, 18 June 1998 (Henson et al.). Id (ASU, Q- 110); Ector County: 24 mi W Odessa, 7 June 1979 (Francke & Merickel) 1 d 1 $ (WDS); Hidalgo County: Edinburg, December 1939 (Mulaik), 2d 1 $ (AMNH); Jeff Davis County: Davis Mountain State Park, behind campground, 16 July 1997 (Henson et al.). Id (ASU, P-262); Fort Davis, 8 June 1902 (no collector), 1 $ (AMNH); Pecos County: Hwy 285, 12.6 mi N Ranch Road 2401, approx. 37 mi S Stockton roadside picnic area, 6 July 1997 (Henson et ah), 1 $ (ASU, P-75); Ranch Road, 4.2 mi from TX 385, 6 July 1997 (Henson et al.), 2d 2$ -h 2 young (ASU, P-67-70); Presidio County (Big Bend Ranch State Park): 3.4 mi W Sauceda (29.28. 55N, 104.00.06W), 18 July 1993 (Henson et al.). Id (WTAMU, SC- 162); 2.2 mi W Sauceda (29.28.30N, 103.59.17W), 18 July 1993 (Henson et ah), Id (WTAMU, SC-159); 1.1 mi W Sauceda (29.28.42N, 103.58.23W), 13 July 1993 (Henson et al.), Id3$ (WTAMU, SC-134); vicinity of Sauceda (west of bunkhouse, 29.28.01N, 103.57.29W), 17 July 1993 (Henson & Sissom), Id (WTAMU, SC- 144); 0.35 mi NE Sauceda (29.28.23N, 103.57.08W), 18 July 1993 (Henson et al.), 2dl$ (WTAMU, SC-150,151); 0.9 mi NE Sauceda (29.28.30N, 103.56.38W), 18 July 1993 (Henson et al.), 2d (WTAMU, SC-156); 1.45 mi E Sauceda (29.28.25N, 103.56. 11 W), 11 July 1997 (Sissom), Id (WTAMU, SC-205); EM 170, 1.1 mi W Lajitas (near boundary of park, 29.15.56N, 103.47. 24W), 28 May 1997 (McWest & Sissom), 1 $ (WTAMU, SC- 183); 7.4 mi inside gate toward Sauceda, 23 June 1999 (Henson et al.). Id (ASU, M-337); 0,35 mi E Sauceda, 18 June 1993 (Henson et al.). Id (ASU, M-116); 0.9 mi E Sauceda, 24 June 1993 (Henson et al.), 4d (ASU, M-340, 342, 347, 347 A); 0.35 mi E Sauceda, 18 June 1993 (Henson et al.), 2d (ASU, M-122, 122A); 0.6 mi from Sauceda, 18 June 1993 (Henson et al.), 1 $ (ASU, M-146); 0.35 mi E Sauceda, 18 June 1993 (Henson et al.), 5d (ASU, M-114, 118-121); 6.4 mi from Sauceda, 23 June 1993 (Henson et al.), 1 juv. (ASU, M-323); Jackson Gate, 11 July 1997 (Zrell et al.). Id (ASU, P-254); 1.1 mi from Sauceda, 13 June 1993 (Hen- son et al.), 1 juv. (ASU, M-21); near Jackson Gate leaving Solitario, 11 July 1997 (Henson et ah), 1 d 1 $ (ASU, P-238-239); 0.3 mi from “Y” on east road out of Solitario, 11 July 1997 (Henson et al.), 3dl$ (ASU, P-200-203); 2.3 mi E of Big Hill, 5 June 1992 (no collector). Id (ASU, L-340); 0.35 mi E Sauceda, 18 June 1993 (Henson et al.), 4d 1 $ (ASU, M-133, 136-138, 151); bottom of Big Hill, 5 June 1992 (no collector), Id (ASU, L-338); 0.6 mi E gate to BBRSNA, 14 June 1993 (Henson et 54 THE JOURNAL OF ARACHNOLOGY al.), 2(5 (ASU, M-69-70); 0.9 mi E Sauceda, 18 June 1993 (Henson et aL), 2d (ASU, M-160-161); 0.3 mi E of gate, 14 June 1993 (Henson et al.). Id (ASU, M-55); 1 mi inside gate, 14 June 1993 (Hen- son et al.), 2d (ASU, M-75-76); Starr County: FM 755, 2 mi N 83 (rock wall), 19 May 1992 (no col- lector), Id (ASU, L-74); FM 755, 2 mi N 83, 19 May 1992 (no collector), 7dl$ (ASU, L-73, 80, 86, 1784, 1791-1797); Kelsay, hill SE Hwy 83, 24 Dec 1984 (Nilsson), 1 9 (WDS); 5 mi E Rio Grande City, 1 June 1939 (Mulaik), 19 + 14 young (AMNH); Rio Grande City, 21 Jan 1939 (Mulaik), 2d 3 9 (AMNH); Rio Grande City, no date (no col- lector), 1 9 (AMNH); Terrell County: Independence Creek/Oak Creek Campground, 7 mi off TX 349, 19 June 1998 (Henson et al.). Id (ASU, Q-114); Independence Creek at SR 349, 26 August 1989 (Van Devender), 19 +4 young (ASU, G-118, 1090); 1 mi S Pecos Co. line (S of Sheffield), 4 June 1986 (Manning), 19 (WDS); 19 mi S Shef- field, 16 May 1958 (McAlister), 19 (AMNH); 19 mi S Sheffield, 16-17 June 1958 (McAlister), 49 (AMNH); Sanderson, 26 May 1952 (Cazier et al.), 1 9 (AMNH); Tom Green County: Gun Club Road at 0.1 mi S Convergence in San Angelo (1900’, 31.25N, 100.30W, under rock on rocky hillside), 24 May 1993 (McWest), 19 (WDS); Upton County: 3 mi S, 5 mi E McCamey, 7 June 1986 (Manning), Id (WDS); Val Verde County: road to Amsted Recreation Area, E of Pecos River, 22 May 1996 (Henson et al.), 1 9 (ASU, D-2, 0376); 4.8 mi from Hwy 90, 4 July 1997 (Henson et al.). Id (ASU, P- 51); Devil’s River State Park, under rock near head- quarters, 13 May 1996 (Henson et al.), 19 (ASU, 0-39); Devil’s River State Natural Area, by old wa- ter tank, 30 June 1997 (Brunner et al.). Id (ASU, P-15); Devil’s River, old water tank, 15 June 1999 (Henson et al.), 7d (ASU, Q-5-6, 8, 10, 12, 14, 14A); Devil’s River State Park, 1 mi from Nature Conservancy line, 17 June 1998 (Henson et al.), 1 d (ASU, Q-88); Dolan Ranch Nature Conservancy, across river from Devil’s River State Natural Area, 17 June 1998 (Henson et al.), 19 + 15 young (ASU, Q-67); Devil’s River, windmill E of old wa- ter tank, 15 June 1999 (Henson et al.), 10d3 9 (ASU, Q-15, 18, 21-24, 26, 28-33); 5.3 mi N Com- stock on TX 1042, under rock outface, 16 May 1996 (Henson et al.), 1 9 (ASU, 0-62); Devil’s Riv- er State Park, across river at Nature Conservancy, 17 June 1998 (Henson et al.). Id (ASU, Q-75); Dolan Ranch W of Devil’s River, 2 July 1997 (Hen- son et al.), 1 juv. (ASU, P-29); Amsted Recreation Area, Pecos River Road to river, 22 May 1990 (Henson et al.), 19 (ASU, D-13, 0387); Seminole Canyon State Park, 30 September 1990 (Henson), 1 9 (ASU, F-283); Devil’s River State Park, under rock, 13 May 1996 (Baldwin et al.), 1 juv. (ASU, 0-38); Devil’s River, 13 May 1996 (Henson et al.), 1 9 (ASU, 0-40); 1 mi SSE Langtry, 7 June 1974 (Drape et al.), 6d29, 1 juv, (WDS); Langtry, 19 March 1960 (Gertsch et al.), 3 9 (AMNH); Webb County: 32 mi E Laredo, 11 Nov 1934 (Mulaik), 5 9 (AMNH); 32 mi E Laredo, 9 Feb 1935 (Mu- laik), 1 9 (AMNH); Zapata County: off US 83 near Environmental Oil and Gas Company, 12 May 1996 (Henson et al.). Id (ASU, 0-32); 2 mi N Zapata (under rock), 16 May 1995 (McWest), Id (WDS). MEXICO: Coahuila: Saltillo, 22 Aug 1947 (Gertsch), 1 9 (AMNH); Durango: Tlahualilo, 1926 (no collector), 29 (UAF); Nuevo Leon: El Ebonito, nr. Mouth Sta. Roen w., no date (Pilsbry), 1 9 (ANS); 9 mi NNW, 2 mi N Mina, 15 July 1975 (Liner), 1 9 (FSCA); 10 km E Villa Aldama (steep cliff to north, much lava rock, 1pm), 18 December 1986 (Nilsson), Id (JAN); Unknown State: 20 mi E San Pedro, 5 July 1936 (Davis), 19 (AMNH). ACKNOWLEDGMENTS I am grateful to Norman 1. Platnick of the American Museum of Natural History, Charles E. Griswold of the California Acad- emy of Sciences, Rowland Shelley of the North Carolina State Museum of Natural Sci- ences, and George Bradley of the University of Arizona for providing specimens of the new species for examination. I express my gratitude to W. David Sissom and Kari J. McWest for reviewing earlier drafts of the manuscript, and for Sissom’s guidance and ad- vice in this study. I also wish to thank Sissom, Richard N. Henson of Appalachian State Uni- versity, James B. Whitfield of the University of Arkansas, and McWest for providing re- cords of Vaejovis waueri from Big Bend Ranch State Park and other localities. On be- half of Henson, who collected specimens of V. waueri from Big Bend National Park, I would like to thank Mike Fleming (Resource Management Specialist of Big Bend National Park) and the National Park Service for grant- ing permission to conduct research in the park. On behalf of Sissom and Henson, I would also like to thank Texas Parks & Wild- life, and in particular David Riskind (Director of the Natural Resources Program) and Luis Armendariz (Superintendent of BBRSP), for permission to collect in Big Bend Ranch State Park. Finally, I would like to thank Dr. Doug- las Bingham, Department Chair of Life, Earth, and Environmental Sciences of West Texas A&M University for providing funds to cover cost of reprints. LITERATURE CITED Capes, E.M. 2001. Description of a new species in the nitidulus group of the genus Vaejovis (Scor- HENDRIXSON— NEW SCORPION FROM MEXICO 55 piones, Vaejovidae). Journal of Arachnology 29: 42-46. Francke, O.E 1977. Scorpions of the genus Diplo- centrus from Oaxaca, Mexico (Scorpionida, Di- plocentridae). Journal of Arachnology 2:107- 118. Gertsch, W.J. & M. Soleglad. 1972. Studies of North American scorpions of the genera Uroc- tonus and Vejovis (Scorpionida, Vejovidae). Bul- letin of the American Museum of Natural His- tory 148:551-608. Sissom, W.D. 1991a. Systematic studies on the ni- tidulus group of the genus Vaejovis, with de- scriptions of seven new species (Scorpiones, Vaejovidae). Journal of Arachnology 19:4-28. Sissom, W.D. 1991b. The genus Vaejovis in So- nora, Mexico (Scorpiones: Vaejovidae). Insecta Mundi 5:215-225. Sissom, WD. 2000. Family Vaejovidae Thorell 1876. In Fet, V, WD. Sissom, G. Lowe, & M. Braunwalder, Catalog of the Scorpions of the World (1758-1998). New York Entomological Society, New York. Sissom, WD. & O.F Francke. 1985. Redescrip- tions of some poorly known species of the niti- dulus group of the genus Vaejovis (Scorpiones, Vaejovidae). Journal of Arachnology 13:243- 266. Sissom, W.D., G.A. Polis & D.D. Watt. 1990. Field and laboratory methods. In The Biology of Scor- pions. (G.A. Polis, ed.). Stanford Univ. Press, Stanford, California. Sissom, W.D. & S.A. Stockwell. 1991. The genus Serradigitus in Sonora, Mexico, with descrip- tions of four new species (Scorpiones, Vaejovi- dae). Insecta Mundi 5:197-214. Soleglad, M.E. 1973. Scorpions of the mexicanus group of the genus Vejovis. Wassman Journal of Biology 31:351-372. Stahnke, H.L. 1970. Scorpion nomenclature and mensuration. Entomological News 81:297-316. Stockwell, S.A. 1986. The scorpions of Texas (Arachnida, Scorpiones). Master's Thesis, Texas Tech University-Lubbock. 193 pp. Vachon, M. 1974. Etude des caracteres utilises pour classer les families es les genres de Scor- piones (Arachnides). Bulletin du Museum Na- tional d’Histoire Naturelle (Paris) (ser. 3) 104: 857-958. Williams, S.C. 1970a. Scorpion fauna of Baja Cal- ifornia, Mexico: Eleven new species of Vejovis (Scorpionida: Vejovidae). Proceedings of the California Academy of Science 37:275-331. Williams, S.C. 1970b. New scorpions belonging to the Eusthenura group of Vejovis from Baja Cal- ifornia, Mexico (Scorpionida, Vejovidae). Pro- ceedings of the California Academy of Science 37:395-418. Williams, S.C. 1971. New and little known scor- pions belonging to the punctipalpi group of the genus Vaejovis from Baja California, Mexico, and adjacent area (Scorpionida: Vaejovidae). Wassman Journal of Biology 29:37-63. Williams, S.C. 1980. Scorpions of Baja California, Mexico and adjacent islands. Occasional Papers of the California Academy of Science 135:1- 127. Yahia, N. & W.D. Sissom. 1996. Studies of the systematics and distribution of Vaejovis bilinea- tus Pocock (Vaejovidae). Journal of Arachnology 24:81-88. Manuscript received 16 March 2000, revised 1 Sep- tember 2000. 2001. The Tournal of Arachnology 29:56-63 THE INFLUENCE OF GROUP SIZE ON DISPERSAL IN THE SOCIAL SPIDER STEGODYPHUS MIMOSARUM (ARANEAE, ERESIDAE) Marilyn Bodasing and Rob Slotowi School of Life and Environmental Sciences, University of Natal, Durban, South Africa Tanza Crouch: Department of Entomology and Arachnology, Durban Natural Science Museum, Durban, South Africa ABSTRACT. The costs and benefits of group living vary with group size, and competition for resources increases with increasing group size. In the social spider, Stegodyphus mimosarum, individuals attain smaller sizes, and survival is lower in larger colonies. In this study we assess whether group size influences the decision to leave a colony — or disperse. Four colony sizes (8, 16, 32 and 64) of S. mimosarum were set up under a proportional feeding regime in a laboratory experiment. We expected more spiders to leave large colonies due to intra-group competition. However, there was no significant increase in the number of spiders leaving with increasing group size. Significantly more spiders left a colony during spring and when spiders were larger (at a more advanced stage of development). Variability in access to resources does not promote dispersal, but season and spider size does influence dispersal. Keywords: Group size, intragroup competition, dispersal, social spiders, Stegodyphus mimosarum The trade-off between the costs and benefits of group living changes with group size (Ran- nala & Brown 1994; Uetz & Hieber 1997). Social animals interact in groups of sizes that maximize the fitness of the individual (Caraco & Wolf 1975; Sibly 1983; Kramer 1985; Gir- aldeau & Gillis 1988; Packer & Ruttan 1988; Aviles & Tufino 1998). There is a stable group size, larger than the optimal group size, where the mean inclusive fitness of joining is larger than if the individual remained alone (Sibly 1983; Giraldeau & Gillis 1985; Zemel & Lu- bin 1995). If the optimal group size cannot be reached, it is preferable for an individual to be in a group larger than optimal rather than a smaller group (Sibly 1983; Giraldeau & Gil- lis 1985), and most groups in nature are larger than optimal (Sibly 1983; Giraldeau & Gillis 1985; Ward & Enders 1985; Zemel & Lubin 1995). An animal should join a group of su- praoptimal size if its fitness would be greater than if it remained alone. Beyond the stable group size, the benefits are too small or the cost levels too high to outweigh the advantag- es of sociality; and individuals should disperse from this group (Kramer 1985). In social spiders, there may be advantages to emigration before reproduction or when there is a large increase in numbers in the col- ony, such as soon after juveniles are bom/ hatch out, and when the predation effects or parasite loads are too high. In addition, the low genetic diversity in social spider colonies may make dispersal imperative (Smith & En- gel 1994). These are the ultimate reasons why animals disperse. However, the proximate reasons driving the decision to disperse from colonies includes access to resources (Ward 1986), season and size of the animal (Miller & Miller 1991). Re- sources in a particular area become depleted, and it is advantageous for an animal or a group of animals to find another location be- fore the resources are completely finished. In social animals there may be increased intra- group competition when resources are dimin- ished (Ward 1986). There are two main aspects to examine with respect to access to resources. First, intra- group competition results in a greater vari- ability in individual access to resources (Ul- brich et al. 1996). In most large social spider nests, competition for resources increased with increasing group size and spiders were less competitive in smaller nests (Ward 1986; Seibt & Wickler 1988a). If the quantity of 56 BODASING ET AL.— DISPERSAL IN SOCIAL SPIDERS 57 prey obtained is proportional to the size of the colony, some individuals may get a higher quantity of food, resulting in a range of indi- vidual body sizes within the colony (Ulbrich et al. 1996; Ward 1986). Although the mean mass of spiders is lower in larger colonies, there is no clear indication whether the vari- ance in body mass correlates with colony size (Seibt & Wickler 1988a; Ward 1986). The de- cision on whether to leave or remain in a group may depend on risk-sensitivity (Uetz 1988), If there is more prey available than the individual needs, remaining in a group reduc- es the risk of starvation by reducing the var- iance in the food intake (i.e., foraging in a risk- averse manner). However, when resourc- es are less than the individual requirements (i.e., there is a negative energy budget), it is preferable to move to improve the chance of obtaining resources (i.e., foraging in a risk- prone manner) (Uetz 1988; Lawes & Perrin 1995). This should also apply when there is less access or more competition for food, as is the situation for the disadvantaged spiders in larger nests. Contest competition gives the larger spiders an advantage over the smaller ones (Ulbrich et al. 1996; Ward 1986; White- house & Lubin 1999). Spiders should then leave the larger nests as competition for re- sources increases, and the smallest spiders should leave. Second, mean access to resources may also trigger dispersal. The mean food intake per spider decreases with increasing group size, spiders take longer to extract the same amount of food (Ward & Enders 1985) and spiders attain smaller sizes in larger nests (Ward 1986; Seibt & Wickler 1988a, b). Ultimately, competition for resources would have an im- pact on adult spider size and time of maturity. This should result in spiders dispersing more from larger nests. Dispersal would then be im- portant since it acts as a stabilizing factor by spreading the risk of starvation (Kuno 1980). In addition, in an experiment to test survival rates, more spiders survived from smaller nests than from larger nests (Ward 1986; Seibt & Wickler 1988a). This also suggests that more spiders should leave the larger nests. We postulated that there would be more in- tra-group competition in larger colonies. Un- der conditions of proportional food availabil- ity per individual, this would result in a range of individual access to food within each col- ony with some spiders being disadvantaged. This variability would be greater in larger col- onies and the more disadvantaged spiders are expected to leave these colonies. In this experiment, we tested the influence of variability in the access to resources on dis- pexsal in different colony sizes. We used four group sizes of S. mimosarum Pavesi 1883 (Ar- aneae, Eresidae) to test if spiders were more likely to disperse from small groups (low var- iability in food intake) or large groups (high variability in food intake). We also examined the influence of spider size and the season at which dispersal occurs by conducting the ex- periment at intervals throughout the year. The influence of mean access to resources will be tested in a subsequent experiment. METHODS Twelve nests of S. mimosarum were col- lected from Weenen Nature Reserve, South Africa (28°50'S, 29°51'E) during March 1997, five in June 1997, six in December 1997 and eight from Itala Game Reserve, South Africa (27°3rS, 3r22'E) in April 1998. Stegody- phus mimosarum are social spiders, with a life cycle of approximately one year; young spi- ders emerge from eggs sacs in late summer (February to March) and the adult spiders are found from spring to midsummer (October to January). Data on the growth rate of S. mi- mosarum from Richmond, Kwazulu-Natal is described elsewhere (Crouch & Lubin 2000). Voucher specimens were deposited at the Dur- ban Natural Science Museum. Nests were maintained in the School of Life and Environmental Sciences, University of Natal, Durban, South Africa under controlled conditions: at 28 °C, on a 12/12 h light/dark cycle to control for seasonal changes in day length. The spiders were fed on a diet of adult mealworms, Tenebrio molitor, and mist- sprayed with water once a week. Nests were housed on Acacia robusta plants in cages of plastic mesh on a metal frame (1 m diameter X 0.5 m or 1 m high). Each cage had a re- movable wooden base on a metal stand. The stand could be immersed in water to prevent predation by ants. A tie-up opening at the top of each cage allowed access for feeding. During preliminary experiments (1996- 1997) we found that groups of two and four spiders either did not survive, or did not pro- duce sufficient silk and had difficulty in the 58 THE JOURNAL OF ARACHNOLOGY Table 1. — Mean body length and mass of spiders for each of the four trials. Note that the spiders used in the April 1998 trial are closer in size to those used in the October 1997 trial than to those used in the April 1997 trial. Trial number Month Season Colony size Mean body length of colony ± SE (mm) Mean body length for trial ± SE (mm) Mean mass (mg) 8 3.44 ± 0.65 1 16 3 24 + 0 79 April 1997 Autumn 32 3.31 ± 0.94 3.32 ± 0.08 6.7 64 3.32 ± 0.72 8 3.96 ± 0.80 z 16 3.85 ± 0.70 July 1997 Winter 32 3.67 ± 0.71 3.62 ± 0.34 6.5 64 3.79 ± 0.59 -3 8 4.55 ± 0.63 J 16 4.4 ± 0.71 October 1997 4.38 ± 0.17 13.8 Spring 32 64 4.16 ± 1.01 4.53 ± 0.93 A 8 3.93 ± 1.39 16 3.71 ± 1.49 April 1998 Autumn 32 3.94 ± 0.71 3.97 ± 0.24 12.6 64 4.29 ± 0.71 capture and immobilization of adult meal- worms. We therefore selected colonies of 8, 16, 32 and 64 spiders for this experiment; to represent small (8), intermediate-sized (16 and 32) and large colonies (64). The selected group sizes of spiders mainly reflected those collected in the field (x ± SE = 43.08 ± 31.42, n = \2) although some field nests con- tained more than 100 spiders. Spiders removed from nests from both lo- calities (Weenen Nature Reserve and Itala Game Reserve) were randomly allocated into groups to eliminate any source effects. Ste~ godyphus mimosarum individuals from differ- ent nests can be combined as they readily ac- cept conspecifics (Seibt & Wickler 1985). At each trial, four replicates of each group size were created, giving a total of 480 spiders in 16 colonies. No spiders were reused in suc- cessive trails. The experiment was repeated four times, in April 1997, July 1997, October 1997 and April 1998, to give a range of sea- sons, spider sizes and levels of maturity. All the spiders used in these trials were immature, i.e., either juvenile or subadult. The total body length of a sub-sample of spiders was measured from every colony. Ev- ery second, third or fourth spider was select- ed, with a total of 4-14 individuals measured. depending on the colony size. The average body length was calculated for each colony (Table 1). The mass for each group was mea- sured to four decimal places, on a Mettler AE240 balance, and the average mass of each spider was calculated (Table 1). We prefer- entially use body length as an indicator of body size (rather than body mass) since it is less affected by the momentary feeding status of the spider. We created a unique color mark- ing for each colony by painting every spider in the colony with two colors of water-based poster paints on the dorsal surface of the ab- domen. Forty-nine A. robusta plants (600-700 mm high) were potted in plastic pots (base diam- eter = 180 mm, top diameter = 240 mm, and height = 205 mm). Each plant was trimmed of all but two or three branches, none of which overhung the pot rim. The plants were arranged in a grid of seven rows, and each row contained seven plants. The pot saucers (outer diameter = 240 mm) were used for the first trial (April 1997), but these were omitted in subsequent trials. The pot centers were 560 mm apart in each row and approximately 820 mm apart diagonally. The windowless experimental room was ar- tificially lit with 14 “daylight” incandescent BODASING ET AL.— -DISPERSAL IN SOCIAL SPIDERS 59 light beibs of 60 W each, mounted on a metal frame suspended from the ceiling (except for Trial 1, where 8 light bulbs were used on a free-standing frame). The allocation of nests on plants was random. However, no nests were placed on the plants adjacent to the walls, to prevent any edge effect from the proximity of the walls. Each colony was placed on a tree, and enclosed with fine net- ting, which was tied onto the branch with string. There was sufficient space inside the netting for the spiders to construct a retreat and capture web. Two days later (i.e., Day 0 of the experiment), the netting was removed. During the experiment, each colony was fed twice weekly—” on days 2, 5, 9, 12, 16 and 19 of each trial. Feeding was proportional to the number of spiders in the colony: colonies of eight were fed one prey item per feeding event, colonies of 16 were fed two prey items, colonies of 32 were fed four prey items and colonies of 64 were fed eight prey items. All movements of spiders were noted daily and each tree or colony was examined for spi- ders and/or silk. Any spiders within a retreat were left undisturbed, although occasionally the retreat was thin enough to estimate the number of spiders present. Information was recorded on the source of the spiders based on color, the number of spiders and their desti- nations. The spiders were removed from their new locations each day. After the first five days, the nests were tak- en apart, the spiders were counted and the number in each colony was recorded. Spiders that had molted were repainted. Some spiders could not be located and the missing individ- uals (excluding any dead spiders, since we could not determine the cause of mortality) were replaced so that the original numbers were re-instated. This initial period was termed the Early Trial (la, 2a, etc.). The col- onies were then covered in netting for a fur- ther two days, after which the netting was re- moved. Fourteen days of daily observations then followed. At the end of this period, the nests were again taken apart, all spiders count- ed and their source noted. This part of the experiment was called Trial lb, 2b, etc., or the Late Trial. The separate early and late parts of each trial were compared using a Wilcoxon Paired Ranks test, and since no influence of early vs. late trials was found (Z ^ -=1.903, P = 0.056), the two sections were combined and averaged. All subsequent analyses were on the combined averaged trials, which increased the internal validity of the data from each colony. The total number leaving each colony was used to calculate the relative number of spi- ders that moved (i.e., total number that moved divided by the number in the colony). The data were normalized using an arcsine [square root] transformation and the transformed data were used for all analyses. An analysis of co- variance, with a post-hoc Bonferroni test, was carried out on each separate section of the ex- periment (i.e., la, 2a, lb, 2b, etc.). ANCOVA was used to remove the effect of trial date or body size. Arcsine [square root] (relative number moving) was the dependent variable, with colony size (8, 16, 32 and 64) as the factor and trial number or body length as the covariate. The assumptions of the ANCOVA were verified using a Kolmogorov-Smimoff test to check that the data and residuals were normally distributed, and a BartletTs Box F- test was used to check for homogeneity of the variances. The assumptions of the parametric tests were met in all cases {P > 0.05). RESULTS We tested the effect of the mean body size of the spiders on dispersal, for the four trials. The relative number of spiders leaving in- creased significantly with increasing body length (Linear Regression: ¥^^^2 ~ 11-45, P ~ 0.001) (Fig. 1), and with increasing spider mass (Linear Regression: Fj ^2 8.21, P = 0.006). The absolute number of spiders moving in- creased with increasing colony size (Fig. 2) (ANOVA: F3,63 = 19.985, P < 0.001). More spiders left the largest colonies (64) compared with the smaller colonies, and this was espe- cially marked during the October 1997 trial. Significantly more spiders left the colonies of 32 in the October 1997 and April 1998 trials compared with the earlier trails. We compared the absolute number of spiders moving with the relative number of spiders moving in each trial (Fig. 3). The relative number of spiders moving increased over the first three trials, (F3,63 = 8.32, P < 0.001). We then tested the relative numbers of spi- ders moving in each colony size. We removed the influence of body length using an AN- COVA, with body length as the covariate (Fig. 4). The trend was for more spiders to leave 60 THE JOURNAL OF ARACHNOLOGY Trial 0 April 1997 n July 1997 1 October 1997 ■ April 1998 Figure 1 . — The influence of body size of spiders on their propensity to move. We plotted the relative number of spiders moving (arcsine square root transformed) against the mean spider body length (mm) for each replicate. The relative number was calculated as the number moving divided by initial colony size. Figure 2. — The influence of colony size on the propensity to move. The absolute number of spiders moving is plotted against trial. Note that all other analyses presented are on the relative number of spiders moving. the smaller group sizes, but these results were not statistically significant (F3 53 = 1.34, P = 0.271). Similar results were obtained using spider mass as covariate (F3 (,3 = 0.82, P = 0.486). We found no influence of colony size on the dispersal of spiders in any of the in- dividual early or late trials or in the combined and averaged early and late trials (in all cases F3 63 < 2.56, P > 0.104). The results for all trials therefore confirm the null hypothesis that group size does not influence dispersal in the group sizes tested. The numbers of spiders leaving increased over the first three trials with more spiders leaving later in the year (Fig. 2, Fig. 5). Trial date had a statistically significant effect (F3 53 - 11.91, P < 0.001) with significantly more spiders leaving during the October trial than either the April or July trials. The first and fourth trials were both run in the same month of different years, i.e., April 1997 and April 1998. The numbers of spiders leaving during the two April trials are significantly different, with more spiders leaving during the April 1998 trial. Despite this difference, when the two April trials are considered as the same season (autumn), there is still a significant seasonal effect (ANOVA: 53 ” 6.64, P = Colony size D 8 016 132 • 64 Time of year for each trial Figure 3. — The influence of colony size on the propensity to move. The effect of mean body length was removed by using the residuals from the re- gression of the relative number moving (arcsine square root transformed) against spider size. The relative number was calculated as the number mov- ing divided by number in the colony. We plotted the residuals against trial date. BODASING ET AL.— DISPERSAL IN SOCIAL SPIDERS 61 Colony size Figure 4. — The influence of colony size on pro- pensity to move. The effect of mean body length was removed by using the residuals of the regres- sion of the relative number moving (actual number moving divided by the number in the colony, arc- sine square root transformed) against spider size. We plotted the residuals against colony size. The results were not statistically different (F3 53 = 1.34, P = 0.271). Sample size for each mean = 16 col- onies. 0.002) with most spiders leaving during the spring (October) trial (Fig. 4). The relative number of spiders leaving for each season was still significantly higher in spring (October) when the effect of body length and mass were removed (ANCOVA: ^2,53 = 3.16, P = 0.050; body length and mass as covariate). We tested the combined effect of colony size and season on the number of spiders moving, in a two-way interaction between the mean number of spiders emigrating in the dif- ferent colony sizes, with season. We used body length as covariate to remove the effect of body length. We found that there was a significant difference in the effect of mean spider size on the relative number of spiders moving in each trial (Fig. 6). In the April 1997 trial, the number of spiders leaving increased with increasing spider size, while this trend reversed in the subsequent trials despite the larger mean size of the spiders in the later trials. There was a significant interaction ef- fect on the mean number of spiders moving (ANOVA: interaction of colony size and trial: Season Autumn Winter Spring April July October Figure 5. — The influence of season on the pro- pensity to move. Variability in spider size was con- trolled by using the residuals from the regression of the relative number moving (arcsine squareroot transformed) against spider size. We present the mean ± 95% confidence intervals for each trial. Significantly more spiders moved during the spring trial. Note that the two autumn trials are combined (i.e., n = 32 colonies; all others n = 16 colonies). ^9^63 = 2.887, P = 0.008, body length as cov- ariate). The size of the colony alone did not influ- ence dispersal but there was a combined effect of colony size and season. The dispersing spi- ders were found on other plants, the walls, ceilings and comers of the experimental room. Most spiders moved during October (spring). Although relative movement from colonies in- creased with increasing spider size, the mean number moving in each of the later trials de- creased. DISCUSSION In most large social spider nests, spider size decreases with increasing group size (Ward 1986; Seibt & Wickler 1988a, b). Under con- ditions of a proportional food supply, intra- group competition results in variability in the individuals’ access to resources. We expected this variability to be greater in larger colonies. This should result in relatively more spiders leaving the larger colonies since ultimately such competition would impact on spider size and time of maturity. We found that spider group size alone did not influence dispersal in the group sizes tested. Other components of fitness (e.g., related- 62 THE JOURNAL OF ARACHNOLOGY trial ° April 1998 » October 1997 0 July 1997 • April 1997 Figure 6. — The influence of the mean size of spi- ders and the time of year on their propensity to move. We plot the mean number of spiders moving against the mean body length for each colony size in each of the four trials. Note the increasing trend in the number of spiders moving with increasing spider size in the April 1997 trial and the decreasing trend in subsequent trials. ness of kin) may make it acceptable to have a larger than optimal group size (Rannala & Brown 1994). Very small spiders would not survive outside the nest (Ward 1986). Even with increased competition, it may benefit an individual to stay in a larger nest since vari- ance in body weight may be less in larger col- onies (Seibt & Wickler 1988a; Ward 1986). Fitness losses are greater on splitting into groups that are smaller than optimal than they are for remaining in a group that is larger than optimal (Giraldeau & Gillis 1985). Dispersal would only replace intra-group competition with inter-group competition (Zemel & Lubin 1995). The costs of dispersal may also dis- courage spiders from moving (Aviles & Tuf- ino 1998). An abundance of insects should be avail- able after the spring rains have fallen and when the trees, on which the spider nests oc- cur, are in flower. Most spiders dispersed dur- ing the October (spring) trial, which repre- sents the time when insects would be abundant. The number of spiders moving increased consistently over the year, with increasing spi- der size. The influence of body size is most important in the October 1997 and April 1998 trials. Spiders mature from October onwards and dispersal may be influenced by the sexual maturity associated with the larger size. Bur- rowing wolf spiders dispersed during spring and autumn and the size of the dispersing spi- ders determined their survival (Miller & Mill- er 1991). Field observations on S. mimosarum showed dispersal by mature males and fe- males during midsummer (Crouch et al. 1998). Also, dispersal of Anelosimus eximius Simon, 1891 (Araneae, Theridiidae) occurs only in inseminated adult females (Vollrath 1982) and S. mimosarum adults occur from October through February. Our results show increased dispersal in spring (October), when spiders are larger and adults occur. The larger size of spiders in the April 1998 trial may be attributed to spiders that were laboratory raised for a few months prior to the experi- ment and hence larger than those in the field at this time. Although there was an overall increase in the number of spiders moving with increasing spider size, in the later trials this trend re- versed. It appears then that for S. mimosarum, the influence of spider body size, level of ma- turity and the time of year (season) with its particular set of environmental conditions, is more important than variability in the access to resources in driving dispersal. The mean amount of food obtained by each spider is less in larger nests (Ward 1986; Seibt & Wickler 1988a). This would influence adult spider size and ultimately, reproduction. It is then preferable to move to improve the chance of obtaining resources (i.e., foraging in a risk- prone manner) if the amount of food obtained is less than the mean requirements (Uetz 1988; Lawes & Perrin 1995). We are presently testing the influence of mean access to food on dispersal in colonies of S. mimosarum, by comparing colonies that have been adequately fed with those that have not been fed. ACKNOWLEDGMENTS Specimens were collected under permit #244/1997 of the Natal Parks Board, to Dr. T. Crouch. We appreciate the field help provided by the Natal Parks Board, Navashni Go vender and Komalan Govender; and the laboratory help provided by Simon Shezi. This study was supported by NRF grant #2037182 to R. Slo- tow. BODASING ET AL.— DISPERSAL IN SOCIAL SPIDERS 63 LITERATURE CITED Aviles, L. & P. Tufino. 1998. Colony size and in- dividual fitness in the social spider, Anelosimus eximius. American Naturalist 152:403-418. Caraco, T. & L.L. Wolf. 1975. Ecological deter- minants of group sizes of foraging lions. Amer- ican Naturalist 109:343-352. Crouch, T, Y. Lubin & M. Bodasing. 1998. Dis- persal in the social spider Stegodyphus mimosa- rum Pavesi, 1883 (Araneae: Eresidae). Durban Museum Novitates 23:52-55. Crouch, T. & Y.D. Lubin. 2000. Effects of climate and prey availability on foraging in a social spi- der, Stegodyphus mimosarum (Araneae, Eresi- dae) Journal of Arachnology 28:158-168. Giraldeau, L. & D. Gillis. 1985. Optimal group size can be stable. A reply to Sibly. Animal Be- haviour 33:666-667. Giraldeau, L. & D. Gillis. 1988. Do lions hunt in group sizes that maximize hunters’ daily food re- turns? Animal Behaviour 36:611-613. Kramer, D.L. 1985. Are colonies supraoptimal groups? Animal Behaviour 33:1031-1032. Kuno, E. 1980. Dispersal and persistence of pop- ulations in unstable habitats. A theoretical note. Oecologia 49:123-126. Lawes, M.J. & M.R. Perrin. 1995. Risk-sensitive foraging behaviour of the round eared elephant shrew (Macroscelides proboscideus). Behavorial Ecology & Sociobiology 37:31-37. Miller, PR. & G.I. Miller. 1991. Dispersal and sur- vivorship in a population of Geolycosa turricola (Araneae, Lycosidae). Journal of Arachnology 19:49„54. Packer, C. & L. Ruttan. 1988. The evolution of cooperative hunting. American Naturalist 132: 159-198. Rannala, B.H. & C.R. Brown. 1994. Relatedness and conflict over optimal group size. TREE 9: 117-119. Sibly, R.M. 1983. Optimal group size is unstable. Animal Behaviour 31:947-948. Seibt, U. & W. Wickler. 1985. Gerontophagy ver- sus cannibalism in the social spiders Stegodyphus mimosarum Pavesi and Stegodyphus dumicola Pocock, Animal Behaviour 35:1903-1905. Seibt, U. & W. Wickler. 1988a. Bionomics and so- cial structure of “family spiders” of the genus Stegodyphus, with special reference to the Afri- can species, S. dumicola and S. mimosarum. Ver- handlungen des Naturwissenschaftlichen vereins in Hamburg 30:255-303. Seibt, U. & W. Wickler. 1988b. Why do “Family spiders,” Stegodyphus (Eresidae), live in colo- nies? Journal of Arachnology 16:193-198. Smith, D.R. & M.S. Engel. 1994. Population struc- ture in an Indian cooperative spider, Stegodyphus sarasinorum Karsch (Eresidae). Journal of Ar- achnology 22:108-113. Uetz, G.W 1988. Risk sensitivity and foraging in colonial spiders. Pp. 353-377, In The Ecology of Social Behaviour (C.N. Slobodchikoff, ed.). Ac- ademic Press, London. Uetz, G.W. & C.S. Hieber. 1997. Colonial web- building spiders: Balancing the costs and benefits of group living. Pp. 458-475, In The Evolution of Social Behaviour in Insects and Arachnids (J.C. Choe & B.J. Crespi, eds.) Cambridge Univ. Press, Cambridge. Ulbrich, K., J.R. Henschel, E Jeltsch & C. Wissel. 1996. Modelling individual variability in a social spider colony in relation to food abundance and its allocation. Proceedings of the International Congress of Arachnology, Revue Suisse de Zool- ogie. Vol. hors Serie 13:661-670. Vollrath, F. 1982. Colony foundation in a social spider. Zeitschrift fur Tierpsychologie 60:313- 324. Ward, PI. 1986. Prey availability increases less quickly than nest size in the social spider Ste- godyphus mimosarum. Behaviour 97:213—225. Ward, PI. & M.M. Enders. 1985. Conflict and co- operation in the group feeding of the social spi- der Stegodyphus mimosarum. Behaviour 94:167- 182. Whitehouse, M.E.A. & Y. Lubin. 1999. Competi- tive foraging in the social spider Stegodyphus mi- mosarum. Animal Behaviour 58:677-688. Zemel, A. & Y Lubin. 1995. Inter-group compe- tition and stable group sizes. Animal Behaviour 50:485-488. Manuscript received 28 January 2000, revised 11 July 2000. 2001. The Journal of Arachnology 29:64-71 SEXUAL SIZE DIMORPHISM AND JUVENILE GROWTH RATE IN LINYPHIA TRIANGULARIS (LINYPHHDAE, ARANEAE) Gary H*P. Lang: Department of Zoology, Goteborg University, Box 463, SE — 405 30 Goteborg, Sweden ABSTRACT. On three separate occasions during the growth season four populations of the sheet web spider Linyphia triangularis were sampled, twice as immatures and once as adults. For the immature specimens, five linear size characteristics (length and width of the cephalothorax, length of tibia of the first leg, and length and height of the abdomen) were measured in the laboratory and compared with fresh weight. The best predictor of weight was abdomen length, closely followed by cephalothorax width. Cephalothorax width was used to compare the size of immatures with the adult size at time of maturity because the abdomen shrinks in the non-foraging adult males. Mean cephalothorax width was larger for males than for females in both immature and adult specimens. The difference increased from the earliest immature population samples to the adult samples. The relationship between cephalothorax width and abdomen length was linear and equal between the sexes over all immature samples. This means that there was no difference in the allocation of resources to body parts important to female fecundity (the abdomen) VS”, body parts important to male fighting ability (the cephalothorax) between males and females. Selection for large male size thus seems to be greater than selection for large female size in this web-building spider, resulting in an overall faster growth rate in males. Males grow >10% larger than females despite the distinct protandry in this species. Keywords: Sexual dimorphism, Linyphia triangularis, growth rate, spider Web-building spiders have normally a sex- ual size dimorphism (SSD) with the female as the larger sex (Vollrath & Parker 1992; Head 1995; Foelix 1996; Coddington et al. 1997). This has been attributed to (1) the develop- ment of dwarf males, favored by a high se- lective adult mortality on males that relaxes sexual selection for large size (Vollrath & Parker 1992), and to (2) the development of giant females (Coddington et al. 1997) in re- sponse to fecundity selection for large females as suggested already by Darwin (Andersson 1994) . Both these views support the idea that nat- ural selection favors larger size in female spi- ders because fecundity is positively correlated with female size in spiders, both intraspecifi- cally (Rubenstein 1987; Suter 1990; Beck & Connor 1992; Higgins 1992) and interspecif- ically (Eberhard 1979; Marshall & Gittleman 1994; Simpson 1995). Females, which expe- rience a constant and relatively low mortality risk throughout their lives, should grow large to maximize fecundity and reproductive suc- cess (Marshall & Gittleman 1994; Head 1995) . Males of web-building spiders run a similar mortality risk as females during juvenile stag- es, but after maturation they leave their pro- tective web and move around in the vegetation searching for females. This behavior is prob- ably associated with a high risk of mortality (Vollrath 1980), as moving around makes the male more vulnerable to predators and reduc- es energy reserves. Small size could give an advantage in avoiding visual predators (Gun- narsson 1998); and, because metabolism is lower for small spiders, more time can be used for reproductive activities (Reiss 1989; Blanckenhom et al. 1995). Linyphiid spiders exhibit all of the charac- ters of other web-builders, except in regard to size dimorphism, where both sexes are of al- most equal size (Vollrath & Parker 1992; Head 1995; Prenter et al. 1997). This probably depends on a more intense sexual selection for large male size through male-male competi- tion over mating opportunities. Protandric mating systems seem to be the rule in liny- phiid spiders (Toft 1989; Gunnarsson & Johnsson 1990; Watson 1990), probably be- cause the first male has precedence in fertil- ization success (Austad 1982; Watson 1991). Males that wait too long for their final molt 64 LANG— SEXUAL SIZE DIMORPHISM 65 risk losing valuable mating opportunities be- cause females reaching sexual maturity mate with the male present in their web. This means that the male growth period is shorter and, assuming an equal growth rate, adult males are expected to be smaller than adult females at maturity. One linyphiid spider that lives on spruce branches and has a bi-annual life cycle, Pi- tyohyphantes phrygianus (C.L. Koch 1836) has been shown to have a male-biased SSD (Gunnarsson 1988). Gunnarsson suggests that this depends on behavioral differences be- tween sexes during winter when males, de- spite high risk of predation, forage more ac- tively than females (Gunnarsson 1998). The species used in this study, Linyphia tri- angularis (Clerck 1757), is univoltine and in- habits low shrubs and bushes in many types of habitats in the southern part of Sweden. Linyphia triangularis overwinters as eggs, hatches in late March/early April, and grows to adult stage in about four months. Males molt to maturity about a week before females (Toft 1989; Stumpf 1990), and in southwest- ern Sweden the first adults are seen at the end of July. The adult males guard subadult fe- males, fighting with other males over access to the female, and mate with the female im- mediately after maturation (Rovner 1968; Nielsen & Toft 1990), leaving mating plugs in the epigynum that impedes further copulation and/or reduce female receptivity (Stumpf 1990; Stumpf & Linsenmair 1995). These studies say nothing about a size dimorphism in either direction, but emphasize the fierce male fights that occur on the web of virgin females, presumably creating a strong sexual selection for large male size. This study was initiated to test whether sexual selection for large male size could overcome fecundity se- lection for large female size in a linyphiid spi- der with a continuous growth period. I used four separate populations to control for local variation in size or size dimorphism between populations. METHODS Four populations of Linyphia triangularis in southwestern Sweden were sampled to con- trol for possible geographic or habitat varia- tion. The NW site (Goteborg) and SW site (Halmstad) were within 5 km from the Swed- ish west coast, whereas the NE site (Skovde) and SE site (Vamamo) were inland sites. The SW site was a sparse pine forest, Pinus syl- vestris, with mostly Vaccinium myrtillus as ground cover. The NW site was dominated by heather, Calluna vulgaris, with some junipers, Juniperus communis, as the only higher veg- etation. The SE site is a pine forest, Pinus sylvestris, with mixed patches of Vaccinium uliginosum and V. myrtillus as ground cover. The NW site is on the edge of a mire, with scattered small (< 5 m) birches, Betula sp., and pines, Pinus sylvestris, as the only higher vegetation, and a ground cover of Calluna vulgaris. Erica tetralix and Empetrum nigrum. The shortest distance between two sites was about 100 km (SW-SE) and the longest dis- tance was about 200 km (SW-NE). The spiders were collected on three separate occasions from each site between 2 June-5 September 1996. The first sampling was planned to occur about four weeks before the expected maturation but, due to bad weather conditions, the sampling was delayed at two locations. The second sampling was done close to the expected maturation date, and the last sampling was made when the reproduc- tive period was over. To ensure that all spiders had put up new webs, sampling was done only after at least 24 hours of relatively still, clear weather. On each occasion, at least 30 speci- mens were collected by “hand-to-jar” sam- pling at random co-ordinates in a selected area of about 10,000 uV with relative homoge- neous vegetation. The spiders were brought to the laboratory and placed in darkness at 4 °C overnight before measuring. On all non-adults I measured the length of the tibia on the first leg, the length (from clyp- eus to pedicel) and maximum width of the cephalothorax, and the length and height of the abdomen to the nearest 0.02 mm with an ocular eye piece on a Wild stereo-microscope. The fresh weight of the spiders was measured using a Sartorius electronic balance to the nearest 0.1 mg. To control the sex of each individual, the spiders were placed in 250 ml plastic jars and reared to maturity. Spiders that could not be unambiguously sexed before they died were excluded. In adults, most males had shrunken and severely distorted abdomens so only the cephalothoracic measurements were noted in these specimens. Statistical analyses were performed with StatView v.5.0 (SAS In- 66 THE JOURNAL OF ARACHNOLOGY Table 1. — Regression statistics of log‘^ weight against the different log'° linear measurements of body size in juvenile stages. Data from all sites are pooled. Measurement (x) Least squares regression on weight (y) r- Cephalothorax length y = 0.095 + 2.833 x 0.861 Cephalothorax width y = 0.464 + 3.307 x 0.928 Tibia 1 length y = 0.080 + 1.963 x 0.714 Abdominal length y = -0.212 + 2.689 x 0.958 Abdominal height y = 0.349 + 1.939 x 0.738 stitute 1998) and Super Anova v.1.11 for Mac- intosh (Abacus Concepts 1989). RESULTS I considered weight as the measure best re- flecting the general size of the animals. How- ever, the contrasting adult lifestyles of males and females make the use of adult weight dif- ficult to accurately assess the actual size at maturity. In order to establish the linear mea- surement that provided the best estimate of general size in L. triangularis, I made a log- arithmic regression of five linear measure- ments against the weight of immature spiders. The adults were not used for this test as the Cephalothorax width (mm) Figure 1 . — The relationship between cephalotho- rax width and abdomen length in immature Liny- phia triangularis, o = females, A = males. Re- gressions are: female abdomen length = -0.37 + 2.18 X cephalothorax width, r^ = 0.84; male ab- domen length = —0.36 + 2.16 X cephalothorax width, r- = 0.80. adult male specimens were more-or-less starved and had severely shrunken abdomens. The correlation was high for all the measures in juveniles (Table 1), but the best predictor of weight in the juvenile specimens was ab- domen length, closely followed by cephalo- thorax width. I chose cephalothorax width as indicator of general size instead of weight or other size measurements including the abdo- men because this study focuses on the com- parison between SSD in juveniles and adults. The deterioration of the males’ abdomen dur- ing their non-feeding, mate-searching adult life inhibits the use of weight or abdomen length for this purpose. In order to examine if there was any dif- ference between males and females in allo- cation of resources to different body parts, i.e., if the relative size of the cephalothorax vs. the abdomen changed during growth, I performed a one factorial ANCOVA on abdomen length with sex as factor and cephalothorax width as covariate. Males and females had very similar expected abdomen length for any given ceph- alothorax width and the only significant factor in the analysis is the covariate, cephalothorax width (F = 934, 3; P < 0.0001). Neither the interaction term, sex*cephalothorax width (F = 0, 029; P = 0, 86), nor the singular factor, sex (P = 0, 007; P = 0, 93), had any effect on abdomen length. The regression lines de- scribing the relationship between maximum cephalothorax width and abdomen length are virtually the same for males and females (Fig. 1). This means that there is no difference be- tween males and females in the relative growth rate of different body parts in juvenile stages. Males were, on average, larger than females in maximum cephalothorax width in all sam- ples except in the second sample at site SE. In the first sampling at the four sites, males were on average 9-14% larger than females. LANG— SEXUAL SIZE DIMORPHISM 67 0 10 20 30 40 50 60 70 0 10 20 30 40 50 60 70 Days Figure 2. — Development of mean cephalothorax width ± S.E.M. in four populations of Linyphia tri- angularis in SW Sweden. The first sampling yielded only immatures. In the second sampling adults found in the populations were excluded and in the third sampling only adults were found. NW, NE, SW, SE = population labels after relative geographic position. Day 1 = July 1.0 = females, A = males. Mean male cephalothorax width ranged from 0.91 mm in the NE site to 1.17 mm in the SE site, while female mean cephalothorax width ranged from 0.81 mm in the NW site to 1.07 mm in the SE site (Fig. 2). Male abdomen length was on average 12-17% larger than fe- male abdomen length in the four sites. In the second sampling, made just before the expected molt to maturity, the mean ceph- alothorax width of males was on average 5- 9% larger than that of females in three sites. In the fourth site (SE), few subadult males were found (n = 2); and they were much smaller than the subadult females. Male ceph- alothorax width at this occasion ranged from 1.07 mm in the SE site to 1.30 mm in the SW site, while female cephalothorax width ranged from 1.16 mm in the NE site to 1.26 mm in the SE site (Fig. 2). The abdomen length var- ied in a similar way, the males being 7-13% larger than females at three sites, but 28% smaller at the SE site. The third sampling was made about three weeks after the normal time for maturation. The average cephalothorax width of adults was larger for males than for females in all areas (8-22%). The absolute difference of the average maximum cephalothorax between males and females width was larger in all pop- ulations compared to the initial collection. Male adult mean cephalothorax width ranged from 1.54 mm in population SE to 1.81 mm 68 THE JOURNAL OF ARACHNOLOGY Table 2. — ANCOVA on cephalothorax width with site and sex as factors and date of sampling as a covariate. Source df Sum of squares Mean square F-value P-value Site 3 1.142 0.381 20.586 <0.0001 Sex 1 0.94 0.094 5.060 0.025 Date of sampling 1 12.262 12.262 663 <0.0001 Site X Sex 3 0.009 0.003 0.154 0.93 Site X Date 3 0.504 0.168 9.085 <0.0001 Sex X Date 1 0.053 0.053 2.865 0.091 Site X Sex X Date 3 0.042 0.014 0.751 0.52 Residual 323 5.971 0.018 in the NW population, while female adult mean cephalothorax width ranged from 1.43 mm in the SE population to 1.52 mm in the NE population. All population means of fe- male cephalothorax width were lower than those of male cephalothorax width (Fig. 2). These results were tested for differences in cephalothorax width with a full interaction, two factor ANCOVA with collection day as a covariate, to control for the time differences between dates of collection, and with sex and site as singular factors (Table 2). There was one significant interaction term in the AN- COVA between site and date of collection. The date of collection, as well as the site of origin, also was highly significant as singular factors; but as the interaction term between these two factors was also significant, these factors cannot be considered independently (Sokal & Rohlf 1995). However, the singular factor sex was not involved in any significant interaction terms and had a significant effect on cephalothorax width (P = 0.0252). That shows that males on average have a larger cephalothorax width than females across all sites and times examined. DISCUSSION The results in this study show that sexual size dimorphism in L. triangularis is generally male-biased in all stages and populations sur- veyed (Fig. 2). Males are larger than females both as juveniles (first sampling, mostly pre- subadult stages) and as adults despite their earlier maturation. The size difference, mea- sured as cephalothorax width, increases from the juvenile stages sampled in early July to the adult stages sampled in August. This sug- gests that males reach a greater weight and overall size at maturity than females do. If so, males have a higher juvenile growth rate than females, as a solution to two apparently op- posing selection pressures on male size in this species. First, males need to mature earlier than females. Female L. triangularis mate with the male present immediately after molt- ing to maturity (Toft 1989) and first male to mate with the female sires most of the off- spring (Stumpf 1995). This gives males a shorter period of time for growth. Second, be- cause of intense male fights for access to un- mated females, males also need to become as large as possible to be able to defend or take over subadult females from other males. The decrease of the SSD in the second sam- pling (compared to the first) in some of the populations is probably due to the low num- bers of subadult males found in the sample, and that these males were probably smaller than the rest of the cohort. This could depend on (1) reproductive success that is increasing with size but decreasing as time of maturity is delayed and (2) the decision to molt to ma- turity that will be a compromise between op- timal size and age. The result of this would be that larger males mature earlier than small males (Steams 1992) and that a late sample would then necessarily underestimate the av- erage size of subadults. Sex-related variation in growth rate was previously described by Wiklund et al. (1991) for the butterfly Pieris napi L. 1758 in south- ern Sweden, This butterfly is partially bivol- tine with discreet generations and must dia- pause as pupae during winter. Males are generally larger than females; but the differ- ence varies depending on the season, as does the mechanism by which the size dimorphism is achieved. Overwintering pupae produce a LANG— SEXUAL SIZE DIMORPHISM 69 first generation of adults in the spring. The eggs laid by these adults can either develop directly, producing a second generation of adults the same year, or they can develop slower into diapausing pupae. Wiklund et al. (1991) shows that the growth rate of the time- stressed larvae that develop directly to second generation adults are higher than for those that develop to diapausing pupae. In the directly developing larvae, males achieve a greater size than females through a higher growth rate. In the diapausing individuals males grow larger than females because they have a longer development time. The authors suggest that growth rate must be considered as a life his- tory trait in its own right amenable to evolu- tionary change and not as a parameter con- trolled passively by temperature and food availability in the environment. The male-biased SSD described in this study is unusual among web-building spiders, and has not been described in other members of the family. Earlier works of SSD in spiders have often used only the total body length as measure of general size, and this may in part explain the difference from the pattern found here. Male-biased SSD in cephalothorax size has previously been described in a study on the orbweaver Metellina segmentata (Clerck 1757) by Prenter et al. (1995). They suggested that the size dimorphism found in the adults of this species depends on sex-specific allo- cation of resources. This hypothesis states that, because eggs are more costly to produce than sperm, the females make a larger invest- ment in reproductive tissues (i.e., the abdo- men). Males can therefore transfer more of their energy to other body parts that assist them in finding and guarding females (such as longer walking legs and a larger, more pow- erful cephalothorax). However, the large re- productive costs for females are the eggs and the yolk associated with the eggs. Even though oocytes are to some extent present at sexual maturity, the largest part of the egg, the yolk, is not added to the oocytes until after copulation (Seitz 1971). Therefore, the costs of reproductive structures that develop before maturity are likely to be similar in males and females and should not have an effect on overall size dimorphism at maturity. Also, if there was a difference in resource allocation prior to maturity between females and males of L. triangularis, one would ex- pect a significant interaction term in the AN- COVA on abdomen length and the slope of the correlation between cephalothorax width and abdomen length (Fig. 2) would be steeper for females than for males. Instead, the slopes between size measurements are virtually iden- tical for males and females. The ANCOVA reveals no difference between males and fe- males in growth of different body parts. This suggests that size dimorphism in L. triangu- laris depends on an overall higher growth rate in males, as opposed to a sex-related alloca- tion of resources between body parts. In Pityohyphantes phrygianus, Gunnarsson (1988) showed that subadult males were sig- nificantly larger than subadult females in both cephalothorax width and abdomen height both before and after winter. A re-analysis of the data shows that the relationship between the two measurements varies considerably be- tween years and time of season, but the chang- es are similar for males and females within each sample. This suggests that overall growth rate is higher for males in this species as well and that allocation of resources does not differ between sexes. Pityohyphantes phrygianus is biannual and overwinters twice before matur- ing in the second spring. In spite of the fe- male-biased primary sex ratio (Gunnarsson & Andersson 1992), it seems to be strong selec- tion on large male size. Males grow larger probably because they forage more actively during winter than females. But foraging ac- tively increases risk of predation, and the re- sult is a more female-biased sex ratio after winter (Gunnarsson 1998). For L. triangularis, which is a species with a continuous growth period from egg to adult, such an explanation is not possible. Neverthe- less, an increased growth rate during juvenile stages would also benefit the female fecundi- ty, unless there is some cost invoked by a high growth rate. Costs could be developmental, for example, if a too-large mass increment be- tween molts would make molting difficult, re- sulting in loss of one or more limbs — or death. Costs could also arise from a risk-prone feed- ing behavior, such as responding indiscrimi- nately to any vibration in the web as if it were prey. This would increase the chance that prey falling onto the web is caught, but it also in- creases the risk to become prey to a larger predator. All of these costs should result in a differential mortality of males, affecting the 70 THE JOURNAL OF ARACHNOLOGY sex ratio of the adult population, and possibly also increase the variance of male sizes com- pared to female size variance. From the data in this study, it is not possible to conclude if there are sexual differences in mortality re- sulting in a skewed sex ratio in the adult pop- ulation. Toft (1989) noted that there is an even sex ratio in L. triangularis just before maturity and concluded that the operational adult sex ratio therefore is male-biased. This suggests that such costs are negligible. Another explanation for this male-biased size dimorphism in both juvenile and adult populations could be that males get a headstart in life. This could occur either because male eggs hatch earlier than female eggs or because male eggs are larger than female eggs. Some egg sacs produced by females in the labora- tory were allowed to hatch, but there was no apparent differentiation in hatching date with- in egg sacs. Other egg sacs were opened and the diameter of the eggs within each egg sac was measured. There was no deviation from an expected normal distribution of egg size within these egg sacs (unpubl. data). No study on spiders that I am aware of has reported a differential investment in male and female eggs nor a differential hatching time between eggs. Differential investment might also be very unlikely as the fertilization occurs to- gether with oviposition when the yolk of the egg already has been supplied (Foelix 1996). In conclusion, this work shows that males are larger than females at maturity in L. triangu- laris and that this depends on differences in behavior or physiology that affect the growth rate during juvenile stages. This result sug- gests that, because of competition for mating opportunities, selection for large size in males is stronger than is selection for large size in females. Since there is no evidence on costs of rapid growth in males, this could mean that there is intraspecific competition that affects males and females asymmetrically. It is also possible that female lifetime reproductive suc- cess does not increase monotonically with size, hence the optimal body size for a female is lower than the maximum attainable size. ACKNOWLEDGMENTS I am grateful to Donald Blomquist for help with statistics. Bengt Gunnarsson, Malte An- dersson and Ingela Danielsson commented on early versions of the manuscript. I also thank Gustavo Hormiga, Jonathan Coddington, Pe- tra Sierwald and Jim Berry for valuable com- ments. Birgit Lundell helped with the food for reared spiders. This study was supported by grants from W.&M. Lundgrens Foundation, the Royal and Hvitfeldtska Foundation, A.&G. Vidfeldts Foundation, Collianders Foundation, and various funds from the Swed- ish Royal Academy of Sciences (to the au- thor). The study also benefited from grants from the Swedish Natural Science Research Council (to Bengt Gunnarsson). LITERATURE CITED Abacus Concepts. 1989. Super Anova. Abacus Concepts, Inc., Berkeley, California. Andersson, M. 1994. Sexual Selection. Princeton Univ. Press. Princeton, New Jersey. 599 pp. Austad, S.N. 1982. First male sperm priority in the bowl and doily spider, Frontinella pyramitela (Walckenaer). Evolution 36(4):777-785. Beck, M.W. & E.E Connor. 1992. Factors affecting the reproductive success of the crab spider Mis- umenoides formosipes: The covariance between juvenile and adult traits. Oecologia (Berlin) 92: 287-295. Blanckenhorn, W.U., R.E Preziosi & D.J. Fairbairn. 1995. Time and energy constraints and the evo- lution of sexual size dimorphism - to eat or to mate? Evolutionary Ecology 9:369-381. Coddington, J.A., G. Hormiga & N. Scharff. 1997. Giant female or dwarf male spiders. Nature 385: 687-688. Eberhard, W.G. 1979. Rates of egg production by tropical spiders in the field. Biotropica 11(4): 292-300. Foelix, R.E 1996. Biology of Spiders (2nd ed.). Ox- ford Univ. Press, Oxford. 330 pp. Gunnarsson, B. 1988. Body size and survival: Im- plications for an overwintering spider. Oikos 52: 274-282. Gunnarsson, B. 1998. Bird predation as a sex- and size-selective agent of the arboreal spider Pityoh- yphantes phrygianus. Functional Ecology 12: 453-458. Gunnarsson, B. & A. Andersson. 1992. Chromo- some variation in embryos of a solitary spider, Pityohyphantes phrygianus, with skewed sex ra- tio. Hereditas 117:85-91. Gunnarsson, B. & J. Johnsson. 1990, Protandry and moulting to maturity in the spider Pityohy- phantes phrygianus. Oikos 59:205—212. Head, G. 1995. Selection on fecundity and varia- tion in the degree of sexual size dimorphism among spider species (Class Araneae). Evolution 49(4):776-781. Higgins, L.E. 1992. Developmental plasticity and lang~sexual size dimorphism 71 fecundity in the orb-weaving spider Nephila cla- vipes. Journal of Arachnology 20:94-106. Marshall, S.D. & J.L. Gittleman. 1994. Clutch size in spiders: Is more better? Eunctional Ecology 8: 118-124. Nielsen, N. & S. Toft. 1990. Alternative male mat- ing strategies in Linyphia triangularis (Araneae, Linyphiidae). Acta Zoologica Eennica 190:293— 297. Prenter, J., W.I. Montgomery & R.W. El wood. 1995. Multivariate morphometries and sexual di- morphism in the orb-web spider Metellina seg- mentata (Clerck, 1757) (Araneae, Metidae). Bi- ological Journal of the Linnean Society 55:345- 354. Prenter, J., W.I. Montgomery & R.W. Elwood. 1997. Sexual dimorphism in northern temperate spiders: Implications for the differential mortal- ity model. Journal of Zoology, London 243:341- 349. Reiss, M.J. 1989. The Allometry of Growth and Reproduction. Cambridge Univ. Press, Cam- bridge. 182 pp. Rovner, J.S. 1968. Territoriality in the sheet- web spider Linyphia triangularis (Clerck) (Araneae, Linyphiidae). Zeitschrift fuer Tierpsychologie 25:232-242. Rubenstein, D.I. 1987. Alternative reproductive tactics in the spider Meta segmentata. Behavioral Ecology and Sociobiology 20:229-237. SAS Institute. 1998. StatView, 2nd ed. SAS Insti- tute Inc., Cary, North Carolina. Seitz, K.-A. 1971. Licht-und elektronenmikrosko- pische Untersuchungen zur Ovarienentwicklung und Oogenese bei Cupiennus salei Keys. (Ara- neae, Ctenidae). Zeitschrift fuer Morphologic der Tiere 69:283-317. Simpson, M.R. 1995. Covariation of spider egg and clutch size: The influence of foraging and parental care. Ecology 76(3):795-800. Sokal, R.R. & EJ. Rohlf. 1995. Biometry (3rd ed.). WH. Freeman & Co., New York. 887 pp. Steams, S.C. 1992, The Evolution of Life Histo- ries. Oxford Univ. Press, Oxford. 249 pp. Stumpf, H. 1990. Observations on the copulation behaviour of the sheet-web spiders Linyphia hor- tensis Sundevall and Linyphia triangularis (Clerck) (Araneae: Linyphiidae). Bulletin of the Society of European Arachnology 1:340-345. Stumpf, H. & K.E. Linsenmair. 1995. Observations on the mating systems of two spiders, Linyphia hortensis Sund. and L. triangularis (Cl.) (Liny- phiidae: Araneae). Revue Suisse de Zoologie (vol. hors serie) 627-634. Suter, R.B. 1990. Determinants of fecundity in Frontinella pyramitela (Araneae, Linyphiidae). Journal of Arachnology 18:263-269. Toft, S. 1989. Mate guarding in two Linyphia spe- cies (Araneae: Linyphiidae). Bulletin of the Brit- ish Arachnological Society 8(2):33-37. Vollrath, F. 1980. Male body size and fitness in the web-building spider Nephila clavipes. Zeitschrift fuer Tierpsychologie 53:61-78. Vollrath, F. & G.A. Parker. 1992. Sexual dimor- phism and distorted sex ratios in spiders. Nature 360(6400): 156-159. Watson, P.J. 1990. Female-enhanced male compe- tition determines the first mate and principal sire in the spider Linyphia litigiosa (Linyphiidae). Behavioral Ecology and Sociobiology 26:77-90. Watson, P.J. 1991. Multiple paternity and first mate sperm precedence in the sierra dome spider, Lin- yphia litigiosa Keyserling (Linyphiidae). Animal Behaviour 41(1): 135-148. Wiklund, C., S. Nylin & J. Forsberg. 1991. Sex- related variation in growth rate as a result of se- lection for large size and protandry in a bivoltine butterfly, Pieris napi. Oikos 60:241-250. Manuscript received 27 July 1999, revised 1 June 2000. r 2001. The Journal of Arachnology 29:72-81 PREDATORY BEHAVIOR OF THREE SPECIES OF SAC SPIDERS ATTACKING CITRUS LEAFMINER Divina M. Amalin', Jonathan Reiskind^, Jorge E. Pena' and Robert McSorley^: 'TREC-IFAS, University of Florida, 18905 SW 280 St., Homestead, Florida 33031 USA; ^Department of Zoology, University of Florida, Gainesville, Florida 32601 USA; ^Department of Entomology and Nematology, University of Florida, Gainesville, Florida 32601 USA ABSTRACT. The predatory habit of three species of sac spiders, Chiracanthium inclusum, Hibana velox, and Trachelas volutus, on citrus leafminer, Phyllocnistis citrella, was investigated. Observation of spider activities during the photophase and the scotophase confirmed that these three species of sac spiders are nocturnal. They detect their prey by sensing vibrations of the substrate induced by the concealed prey. Movements of P. citrella larvae and prepupae appear to create vibrations of the leaf substrate, which then serve as cues for the spiders to locate them. The searching and prey capture behaviors of these spiders are discussed. Two methods of prey attack were exhibited. In one method, the spider punctures the mine, immobilizes the larva and then bites it and sucks the larval body fluid. In the second behavioral pattern, the spider makes a slit in the mine, uses its forelegs to pull the larva or prepupa out of the mine, holds the prey securely, and finally bites it and regurgitates digestive juices into the prey and ingests the pre- digested liquid tissue. The three species of sac spiders were found to start feeding on P. citrella larvae during the 2nd instar stage. Consumption increased as they developed to later instars. Maximum consumption for all species was recorded at the 4th instar. Although C. inclusum and T. volutus can complete their life cycle with P. citrella as their only food, H. velox was not able to develop to the adult stage. Results obtained from this study provide useful data to better understand the role of sac spiders in the overall management of P. citrella. Keywords: Nocturnal, prey capture, feeding stage, Phyllocnistis citrella, sac spiders The citrus leafminer, Phyllocnistis citrella Stainton (1856) (Lepidoptera, Gracillariidae) has become an important pest of Citrus spp. in Florida since its introduction in 1993 (Knapp et al. 1995). The larvae of P. citrella mine in leaf tissues of any citrus and related species (Heppner 1993). Phyllocnistis citrella larval feeding results in citrus plants with dis- torted and reduced young shoots. Severe pres- sure from P. citrella causes decrease in yields and quality (Knapp et al. 1995; Heppner 1995; Burgeons & Constantin 1995). Although many insect parasitoids of this pest have been recorded (Heppner 1993), little attention has been directed towards predators of the larvae. Predatory arthropods are believed to make an important contribution to the mortality of P. citrella (Zhao 1989; Zhang et al. 1994; Argov & Rossler 1996; Browning & Pena 1995; Amalin et al. 1995; Pena & Subramanian un- publ.). In Israel, spiders were observed in the field to prey upon P. citrella (Argov & Ros- sler 1996). Likewise, in south Florida, various species of spiders were considered important in reducing peak populations of P. citrella (Browning & Pena 1995; Amalin et al. 1995). Feeding tests on 14 commonly encountered spider species in lime orchards in Homestead, Florida confirmed that the three species of sac spiders, Chiracanthium inclusum (Hentz 1847) (Clubionidae), Hibana velox (Becker 1879) (Anyphaenidae), and Trachelas volutus Gertsch 1935 (Corrinnidae), fed on P. citrella larvae and, in some cases, on prepupae (Amalin 1999). Apparently these species of spiders are able to detect and attack the larvae through the leaf epidermis. This phenomenon of search and extraction of a cryptic food source has not been reported in this group of spiders. Because of this specialized feeding behavior, these spiders may prove to be im- portant predators of P. citrella. Study of the predatory habits of these spiders on citrus leafminer is of paramount importance to better 72 AMALIN ET AL.— PREDATORY BEHAVIOR OF SAC SPIDERS 73 8 O 8 S Time Intervals Figure 1. — ^Duration of locomotory activities in three species of sac spiders measured in six 4-hour intervals. Lights turned off at 1800 h. understand the potential of these predators as a component of the natural enemy complex of P. citrella. In this paper, investigations of the predatory behavior of three species of sac spiders (C. inclusum, H. velox, and T. volutus) attacking P. citrella is reported. The main objectives of this study were to determine the time of feed- ing activity, to investigate the predation strat- egy, and to identify the developmental stages of spiders that feed most actively on P. ci- trella. METHODS Sources of test organisms. — Egg sacs of C inclusum, H. velox, and T. volutus were collected from citrus orchards in the vicinity of Homestead, Florida. Egg sacs were identi- fied based on descriptions by Amalin (1999). They were brought to the laboratory and maintained in the incubator at 27 °C, 80% RH, and 12:12 h photoperiod and reared on an ar- tificial diet (Amalin 1999). Laboratory-reared 4th instar spiders were used for the experi- ments to determine the time of feeding activ- ity and predatory strategy. Phyllocnistis ci- trella larvae were collected from a culture on lime {Citrus aurantifolia) plants maintained in the greenhouse. Voucher specimens are de- posited at the Tropical Research and Educa- tion Center, Dept, of Entomology, Homestead, Florida. Time of feeding activity. — Feeding times of the three sac spiders were determined from observations over a 24 hour period. A plastic petri plate (10.5 cm diameter X 2.0 cm high) was used as the observational arena. Lime leaves with five P. citrella larvae within the serpentine mines were placed in each petri plate. The number of P. citrella exposed to the spider is based on the result of the pre- dation experiment previously conducted (Amalin 1999), in which a ratio of one spider to 10 P. citrella larvae gave an average of 5.3 P. citrella larvae consumed. A single 4th instar spider was placed in each arena that was lined at the bottom with moistened filter paper to retain the freshness of the leaves. The test spiders were fed with artificial diet (Amalin et al. unpubl.) for 24 h before transferring to each petri plate; in this way the hunger level is controlled. All the pe- tri plates were placed in an incubator with constant temperature (27 °C), relative humid- ity (80%), and a 12:12 h photoperiod. These were the same conditions under which spiders have been previously reared. Observations of spider activities were made every 15 min dur- ing the photophase and during scotophase by the use of a portable red light with an intensity of 5.0 Lux. Some arthropods (Borror et al. 1992; Jackson 1977) cannot recognize red light. Thus, red light is used to observe their nocturnal activities. The set-up was repeated three times on separate dates for each of the three spider species. The circadian rhythm of locomotory activity was observed to deter- mine if they are diumally or noctumally active animals. Duration of movement in six 4 h in- tervals (0600-=-1000, 1000-1400, 1400-1800, 1800-2200, 2200-0200, 0200-0600) was noted. Predatory strategy. — Spider activities were recorded by videotaping, using a video 74 THE JOURNAL OF ARACHNOLOGY Figure 2. — Chiracanthium inclusum touching or sensing the Phyllocnistis citrella larva by its hindleg. Arrow shows the P. citrella larva still within the serpentine mine. time lapse cassette recorder (Panasonic Model AG“6730). A television monitor (Sony Trini^ tron) and a video camera (Javelin Chromachip V, Model JE=3662RGB) were hooked-up to the video recorder. The video camera was held on top of a tripod. The three species of sac spiders included in this observation were placed separately in a small petri plate (3.5 cm diameter X 0.5 cm high). The petri plate was provided with five P. citrella 2nd instar larvae still within their serpentine mines. The petri plate was positioned under the tripod. The exact position of the petri plate was de- termined by looking at the television monitor. The video machine was set to 16 h recording continously. The videotaping was conducted in a room with lights off from 1800 h until 0700 h the next day. To have a clear view of the predatory activity under total darkness, red lights (5.0 Lux intensity) were provided under the tripod. The set-up was repeated five times for each species. After videotaping, each tape was viewed and the following data were gath- ered: retreat period (no locomotion, no body movement, the spider remained inside the re- treat nest); searching time (locomotory activ- ity); and handling time (period from start of attack until prey was consumed). The mean and standard error of the time spent for each activity were calculated. The number of P. ci- trella consumed was counted under a micro- scope the following morning and the average number of P. citrella consumed was calculat- ed. The mean difference for each parameter was compared using Duncan Multiple Range Test (DMRT) (SAS 1989). Active feeding stages. — to determine spi- der developmental stages capable of feeding on P. citrella, the three species of spiders were reared from the 2nd instar to the adult stage using P. citrella larvae their sole food source. Ten P. citrella 2nd instar larvae within the serpentine mines were placed in each petri plate (10.5 cm diameter X 2.5 cm high). In- dividual 2nd instar spiders were introduced into each petri plate. Ten spiders of each spe- AMALIN ET AL.— PREDATORY BEHAVIOR OF SAC SPIDERS 75 Figure 3. — Predation sequence of (A) Chirac anthium inclusum, (B) Hibana velox, (C) Trachelas vol- utus- — searching (1), feeding (2), and after feeding (3). Arrows in A3, B3, and C3 show the empty ser- pentine mines after spider feeding. Table 1. — Duration of time for predation activity and the percent Phyllocnistis citrella consumption in 24 hours by the 4th instar spiderlings of Chirac anthium inclusum, Hibana velox, and Trachelas volutus. All figures are mean ± S.E. of five replications. Means in each column with the same letters are not significantly different according to DMRT Spider species Searching time (min) Handling time (min) % P. citrella consumption Chiracanthium inclusum 22.8 ± 7.6 a 10.8 ± 2.5 a 60.0 ± 24.0 b Hibana velox 8.8 ± 4.8 b 11.3 ± 5.7 a 64.0 ± 16.0 b Trachelas volutus 5.5 ± 1.2 b 6.5 ± 4.7 a 90.0 ± 11.0 a Table 2. — Percent Phyllocnistis citreiia larval consumption by the different instars and adult stage of Chiracanthium inclusum, Hibana velox and Trachelas volutus. All figures are mean ± S.E. of 10 replications. Data for immature stages of female and male spiders for all species are pooled. 76 THE JOURNAL OF ARACHNOLOGY q q ON +1 + 1 S rn CO in m NO +-> 3 < q IT) q U s + 1 + 1 q U d m 00 +1 06 q 1—1 s + 1 'O q 10 CO 'T q 'O 0) I— 1 50 +1 + 1 + 1 un ON U in d (N in w p 00 D q lO cd U NO .-N S B + 1 + 1 + 1 q in CO i> d r- NO 0 NO in ’H + 1 + 1 + 1 m 00 (N NO u ■d- T— ( q c- CO ON TS c -fl + 1 + 1 10 cm were collected). Laboratory conditions varied among the three populations studied (Table 1) and it is not pos- sible to do a single statistical analysis testing for differences between the field and labora- tory collected webs among all populations. Therefore, separate comparisons were made, three testing for an effect of environment (field V5. laboratory) within each population and one testing for differences among the field-collected webs of the three populations. The data were analyzed using multiple anal- ysis of variance (MANOVA) with the GLM module of SYSTAT (Wilkinson 1992). Be- cause percentages are not normally distribut- ed, all data were arc sin (squareroot) trans- formed prior to analysis. Transformed molar percentages of the seven LMW were the de- pendent variables and either location (field vs. laboratory) or population was the independent variable. Similarly, MANOVA was used to compare the LMW composition of webs spun by Chamela spiders in the laboratory in Mex- ico City with those spun in the laboratory in Austin. There are indications that juvenile males of- ten build webs that are chemically distinct from webs of females. However, there were too few males from any one site (Los Tuxtlas, 2S; Chamela, 26; Brazos Bend, 3d) and sex was not included in the analysis as an inde- pendent variable. RESULTS The chemical composition of the aqueous solution of the adhesive spiral varied among individuals both qualitatively, with differences in which compounds were found, and quan- titatively, with differences in the relative amounts of the compounds. Comparisons be- tween field and laboratory web chemistry are based upon analysis of 58 web collections from 29 spiders that spun at least five webs under both field and laboratory conditions. In addition, comparisons between webs spun in Mexico City and Austin laboratories are based upon analysis of webs from eight spiders from Chamela. Below, we present first a description of the qualitative differences found among in- dividuals between treatments and among pop- ulations, then a description of the quantitative differences found when seven major organic components of the aqueous solution are con- sidered. Qualitative variation. — Most of the indi- viduals in all three populations spun webs containing all seven of the organic LMW that we examined quantitatively (A-acetyltaurine, GABamide, glycine, choline, putrescine, gly- cine betaine, alanine). A-acetyltaurine, choline and glycine betaine were invariably detected in web washes. Occasionally, one or more of the other four compounds was not detected by 'H NMR (Table 2). Most notably, GABamide, typically a major constituent, was not detected in nine web collections built by six spiders. Putrescine, glycine and alanine each went un- detected in at least one web collection. A dis- proportionately high percentage of such com- pound-deficient webs were obtained from juvenile males (Table 2). While the seven measured LMW constitute a large percentage of the organic LMW (we estimate about 80-90% typically), they are not the only organic LMW in the viscid coat- ing of N. clavipes adhesive spirals. Two com- pounds observed in some web washes, taurine and 4-aminobutyric acid (GABA) (Table 2), are presumed precursors of A-acetyltaurine and GABamide, respectively. Taurine was present in sizable quantity (9-14 mole %) only in laboratory-collected webs from two male Chamela spiders. These webs were also characterized by relatively low or undetect- able levels of GABamide and glycine. De- tectable amounts of GABA (2-15 mole %) were observed in 9 web collections, all but one from Brazos Bend, Texas. A compound indistinguishable by 'H NMR from acetate occurred in several Chamela web washes in the range of 3-17 mole % (Table 2). All of these web washes contained little or no detectable GABamide. Several other web washes of spiders from Chamela and Los Tux- tlas also appeared to contain small amounts of Table 2. — Qualitative variation in the composition of Nephiia clavipes web washes examined in this study. The identification numbers of the individual spiders exhibiting a given web composition feature are given in parentheses below non-zero values. ' Field and laboratory web collections were obtained from 6 females and 3 males from Brazos Bend. Following collection of webs in the laboratory, the spiders were killed en masse by freezing before the sex of each numbered individual was determined. Thus, we do not know the sex of each individual. ^ F = field-collected; LM = laboratory-collected in Mexico City; LA 86 THE JOURNAL OF ARACHNOLOGY 3 O 3 mole %) (CIO, 14) (CIO) (Cl 2) # with 4.30 ppm 0 0 0 0000000 8 1 compound^ (Bl, 3, 4, 5, (B4) HIGGINS ET AL.— VARIATION IN ORB WEB CHEMISTRY 87 Table 3. — Pearson correlation matrices for each population, including both field and laboratory collected webs. The molar percentage of each compound was arcsin (squareroot) transformed prior to analysis. Abbreviations: gly = glycine; N-tau = N-acetyltaurine; GABam = GABamide; put: putrescine; cho = choline; bet = glycine betaine. Bonferroni-corrected P-values: * P < 0.05, ** p < 0.001. Los Tuxtlas gly N-tau GABam put cho bet N-acetyltaurine -0.941** GABamide 0.735* -0.833** putrescine 0.508 -0.678* 0.341 choline -0.221 0.186 -0.484 0.259 glycine betaine -0.528 0.656 -0.712* -0.490 0.230 alanine 0.538 -0.512 0.399 0.111 -0.149 0.067 Chamela N-acetyltaurine -0.847** GABamide 0.451 -0.529 putrescine -0.214 0.001 -0.401 choline -0.375 0.320 -0.846** 0.496 glycine betaine -0.304 0.449 -0.889** 0.338 0.797** alanine 0.795** -0.804** 0.530 -0.024 -0.425 -0.360 Brazos Bend gly N-tau GABam GABA put cho bet N-acetyltaurine -0.736* GABamide 0.133 -0.562 GABA 0.227 0.070 -0.626 putrescine -0.703* 0.724* -0.597 0.012 choline -0.656 0.855** -0.683* 0.047 0.843** glycine betaine -0.191 0.398 -0.428 0.129 0.058 0.416 alanine 0.362 -0.444 0.178 -0.249 -0.135 -0.252 -0.495 acetate (< 1 mole %). Proline was detected in web washes from individuals of all three pop- ulations. Those web washes containing suffi- cient proline to allow certain identification were from webs built by females in the lab- oratory. Proline accounted for no more than 3 mole % of the organic LMW. Some additional organic LMW have not been identified. Most notable is a compound producing a sometimes prominent singlet (at most, peak area comparable to that of A^-ace- tyltaurine’s singlet) at 4.30 ppm in ’H NMR spectra, observed in all but one of the field- built web collections from the Brazos Bend population. Again with a single exception, this compound was absent from the laboratory- built web collections from this population and in the one exception, it was present in lower relative quantity than was observed in the field-collected webs. It was not observed at all in the two Mexican populations studied (Table 2). Quantitative variation*- — There were some strong correlations among the seven LMW an- alyzed in this study (A-acetyltaurine, GABam- ide, glycine, choline, putrescine, glycine be- taine and alanine; Table 3). Among all three populations, the amount of A-acetyltaurine was negatively correlated with the amount of glycine. There was a tendency for a negative correlation of GABamide with glycine betaine (not significantly for Brazos Bend), choline (not significantly for Los Tuxtlas) and A-ace- tyltaurine (significant only for Los Tuxtlas). Between pairs of chemically similar com- pounds, the amounts of choline and glycine betaine and the amounts of glycine and ala- nine tended to be positively associated, al- though these relationships were significant only for webs spun by Chamela spiders. No significant correlation was found between gly- cine and glycine betaine. There was a non- significant trend toward a negative correlation of the amount of GABA (common only in webs of spiders from Brazos Bend) with the amount of its derivative, GABamide. 88 THE JOURNAL OF ARACHNOLOGY Chamela Field lii Laboratory ft glycine •o E ro m 1 iS c o m £ choline ® c B ® alanine glycine ® ■O E ® .£ 3 V ® c ® choline ® c B ® alanine glycine ® •O E < cd 3 >. < 3 >. < CD Z Ct 05 CD z Q. 05 CD o ^ JZ ® O ^ O) Figure 1. — The average molar percentage of each of the seven studied low-molecular weight organic compounds (± SEM) for each population under field and laboratory conditions. Data from Chamela include observations made in the laboratory in Mexico City (open bars) and in the laboratory in Austin (hatched bars). A-a-taurine: A-acetyltaurine; gly betaine: glycine betaine. ★ = Significant difference (P < 0.05) among populations; § = significant difference (P < 0.05) between field and laboratory conditions within a population (Bonferroni -corrected P values). Testing for significant patterns of variation in these components among populations and between field and laboratory conditions in- volved multivariate analysis of variance of arcsin (square root) transformed relative quan- tities of the seven primary compounds (mole %; Fig. 1). Separate tests were done to ex- amine patterns of variation among webs from different sites and, within a population, be- tween field- and laboratory-spun webs. There were significant quantitative differ- ences in LMW composition among webs col- lected from different field sites (Fig. 1, Table 4). Examination of the univariate F tests for the individual components shows that the mo- lar percentages of A-acetyltaurine and alanine were significantly higher and putrescine was lower in webs spun at Los Tuxtlas compared to the other two sites (A-acetyltaurine: F(2, 26) = 4.274, F = 0.025; alanine: F(2, 26) ~ 3.898, F = 0.033; putrescine: F(2, 26) “ 9.129, F = 0.001). The webs spun at Brazos Bend had higher relative amounts of GAB amide than those from the other two sites (F(2.26) “ 4.878, F = 0.016). Variation in the molar percentages of choline and glycine betaine was nearly sig- nificant {F < 0.06). The relative quantities of these components of the LMW solution changed when spiders were moved from the field to the laboratory (Fig. 1). Analyzing the data for each popula- tion separately (Table 4), the spiders from Chamela and Los Tuxtlas significantly altered the composition of the LMW, and the spiders from Chamela further altered the web chem- istry when they were moved from the labo- ratory in Mexico City to Austin. The spiders from Brazos Bend showed non-significant HIGGINS ET AL.— VARIATION IN ORB WEB CHEMISTRY 89 Table 4. — Multiple analysis of variance: Differences among field-spun webs from spiders in three pop- ulations, and differences between field- and laboratory-spun webs from three populations. The two entries for Chamela field to laboratory comparisons reflect comparisons between field and the laboratory in Mexico (F/LM) and between field and the laboratory in Austin (F/LA). Wilk’s lambda F-statistic Degrees of freedom P value Among field-spun webs 0.190 3.700 14, 40 0.001 Between field- and laboratory-spun webs Brazos Bend 0.335 2.835 7, 10 0.066 Los Tuxtlas 0.130 9.578 7, 10 0.001 Chamela F/LM 0.051 10.73 7, 14 0.018 Chamela F/LA 0.085 21.54 7, 14 <0.001 Between laboratory settings Chamela 0.193 4.767 7, 8 0.022 shifts in the relative amounts of the seven compounds. The changes in LMW composition accom- panying the move from field to laboratory dif- fered among the three populations. Spiders from Los Tuxtlas increased putrescine and de- creased glycine betaine (putrescine: F(i jg) = 10.36, P = 0.005; glycine betaine: F^, ,6) = 22.58, P < 0.001). The spiders from Chamela decreased free alanine when moved from the field into the laboratory in Mexico City and this change persisted when the spiders were moved to Austin (field vs. lab in Mexico: F^, ,0) = 8.92, P = 0.014; field v^. lab in Austin: F(i, 20) = 10.645, P = 0.004). Comparison of the field webs with the webs spun in Austin also showed a decline in the percentage of glycine and an increase in A^-acetyltaurine (glycine: F^, 20) ^ 19.22, P < 0.001; A-acetyl- taurine: F^, 20) == 10.417, P = 0.004). The sig- nificant change in composition between the webs spun by the Chamela spiders in the lab- oratory in Mexico City and in the laboratory in Austin reflects overall trends; no single component changed significantly. In the case of the Brazos Bend population, although the multivariate statistic was nonsignificant, there was a significant increase in the molar per- centage of putrescine when the spiders were moved from the field into the laboratory (F(,,,6) - 14.705, P = 0.001). In addition to the statistically significant changes, three trends are of interest because a majority of individuals from Los Tuxtlas or Chamela exhibited them. Relocation of Los Tuxtlas and Chamela females to the laboratory tended to result in decreased percentages of choline (7 of 7 from Los Tuxtlas, 12 of 13 from Chamela) and increased GAB amide (7 of 7 field/laboratory comparisons from Los Tuxtlas, 12 of 13 from Chamela). Males from these populations (albeit a small sample size) did not exhibit these trends: among males, choline concentration tended to increase and GAB amide tended to decrease with relocation to the laboratory. A-acetyltaurine percentages changed in opposite directions in the webs of individuals from these populations: Los Tux- tlas animals, male and female, tended to de- crease the percentage of this compound (8 of 9) whereas, as mentioned above, the percent- age increased significantly on webs built by male and female Chamela spiders in the lab- oratory relative to webs built in the field (17 of 17). NMR spectral data. — Data for GA- Bamide, A-acetyltaurine, glycine, choline, glycine betaine, and taurine have been pub- lished (Townley et al. 1991). The additional LMW identified in this study yielded the fol- lowing 'H NMR spectral data in D2O (chem- ical shifts in ppm, with the methyl hydrogens of the internal standard, 2-methyl-2-propanol, assigned a chemical shift of 81.2200): acetate, singlet at 81.88; alanine, quartet at 83.75 (J = 7.3 Hz), doublet at 81.45 (J = 7.3 Hz); GABA, triplets at 82.99 (J = 7.5 Hz), 82.27 (J = 7.3 Hz), quintet at 81.87 (J = 7.4 Hz); 90 THE JOURNAL OF ARACHNOLOGY proline, multiplets at 84.10, 83.35 (83.40, 83.30), 82.31, 82.01; putrescine, multiplets at 83.02, 81.73. DISCUSSION The current study extends previous reports on the chemical composition of the organic LMW solution found on ecribellate adhesive spirals by documenting variation in web chemistry within and among populations. Fur- thermore, we observed significant quantitative shifts in LMW composition correlated with changes in environment: the spiders from the two Mexican populations significantly altered relative amounts of some LMW on their webs when moved from the field into the laborato- ry. Examination of webs spun by individuals also revealed patterns of individual qualitative variation in the composition of the LMW so- lution. Some spiders, particularly juvenile males, spun webs in which compounds char- acteristic of TV. clavipes webs were undetected and/or novel compounds were found. Follow- ing a discussion of the extrinsic and intrinsic factors that may singly or in combination re- sult in spiders spinning webs with different LMW composition, we discuss the possible influence of LMW composition on web func- tion. There are four possible factors that may in- fluence the chemistry of the LMW portion of the web: physical environment, diet, web re- cycling, and ontogenetic changes in web chemistry. First, if physical properties of the adhesive spiral (e.g., hygroscopicity, droplet viscosity, extensibility) are influenced by LMW composition, it seems unlikely that a single LMW composition would prove ideal in all environments inhabited by individuals of one species. Thus, there is the possibility that among-population differences in LMW composition and the shift in composition when individuals are moved from one envi- ronment to another may reflect individual spi- ders’ adjustments to the conditions of the physical environment. Among-population dif- ferences may also reflect genetic differences among populations, as selection favors differ- ent chemical and physical properties in dif- ferent physical environments. Second is the possibility that qualitative differences in diet affected LMW composition as spiders were shifted from the field to the laboratory. These spiders eat a variety of prey in the field (Hig- gins & Buskirk 1992), but were kept on a monotypic diet in the laboratory. As prey types vary among these three populations in the field (Higgins pers. obs.), qualitative die- tary differences may contribute to among- population differences as well. Qualitative changes in diet have been found to alter amino acid composition of spider major ampullate silk (Craig et al. in press). Third, we now have evidence that web recycling influences LMW composition (Townley & Tillinghast pers. obs.) and recycling was an uncontrolled vari- able between the field and laboratory portions of the study. Spiders were collected from in- tact orb webs and the first web built in the field portion of the study, also the first web collected, presumably included little recycled material. In contrast, the spiders recycled the orb prior to collection of the first web in the laboratory. Last, there is the possibility of on- togenetic changes in LMW composition, in- dependent of diet and environmental condi- tions. Ontogenetic changes in structural features of orb webs (e.g., number of radii, mesh size, shape) have been documented (Witt et al. 1972; Ramousse 1973; Eberhard 1985 and references therein; Eberhard 1986; Edmunds 1993) and it is possible that changes during development may extend to facets of web composition as well. Indeed, Osaki (1989) has reported changes in the color of major ampullate silk, presumably due to changes in chemical composition, with the ap- proach of maturity in female Nephila clavata. Differences in LMW composition could af- fect various physical properties of the adhe- sive spiral and, thereby, affect the web’s prey- catching ability. Therefore, the possible functional consequences of qualitative and quantitative differences in adhesive spiral composition merit further examination. For example, some of the LMW are hygroscopic (Vollrath et al. 1990; Townley et al. 1991) and the overall hygroscopicity of the LMW mix- ture presumably varies with LMW composi- tion. Differences in hygroscopicity may have an impact on web function because adsorption and retention of water by the adhesive spiral is essential to its adhesiveness, elasticity, and extensibility. Water’s involvement in adhesive spiral functioning may be a combination of direct effects, due to interactions between wa- ter and adhesive spiral components, and in- direct effects, due to interactions between HIGGINS ET AL.— VARIATION IN ORB WEB CHEMISTRY 91 LMW and adhesive spiral macromolecules that require an aqueous medium (Richter 1956; Schildknecht et aL 1972; Vollrath & Ed- monds 1989; Bonthrone et al. 1992; Edmonds & Vollrath 1992; Gosline et al. 1994, 1995; Hayashi & Lewis 1998). Beyond the possible consequences for ad- hesive spiral hygroscopicity, LMW composi- tional differences may also influence the ef- fectiveness of the adhesive spiral by affecting its macromolecular structure more directly. Here we briefly discuss three hypothetical ways in which the organic LMW may accom- plish this: as compatible solutes, through di- rect interaction with macromolecules, and as counteracting solute systems. A wide variety of procaryotic and eucary- otic cells subject to osmotic stress employ cer- tain organic osmolytes to adjust intracellular osmolarity. These osmolytes are sometimes referred to as compatible solutes (Brown & Simpson 1972) because, unlike inorganic ions in most organisms, they can occur at high con- centrations without perturbing, and even while stabilizing, protein structure (Yancey et al. 1982; Le Rudulier et al. 1984; Somero 1986, 1992; Csonka & Hanson 1991; Kinne 1993; Galinski 1993, 1995). Protein stabilization by compatible solutes has been attributed to the tendency of these solutes to be excluded from the immediate vicinity of protein surfaces (so increasing the non-uniform distribution of sol- ute) and to exhibit low specific binding to pro- teins (Arakawa & Timasheff 1985; Timasheff 1992). Thus, these solutes promote processes of protein folding and subunit aggregation that minimize protein surface area and typically favor protein stabTity. Although the web is external, certain of the adhesive spiral’s or- ganic LMW (glycine, glycine betaine, alanine, proline, taurine) are identical to known com- patible solutes (Townley et al. 1991). Thus, these compounds may, by the same mecha- nism, help stabilize the native conformation of adhesive spiral proteins. One important distinction between compat- ible solutes and some of the LMW compounds on the adhesive spiral concerns molecular charge. Compatible solutes are usually un- charged or net neutral molecules, whereas several of the adhesive spiral LMW compo- nents (iV-acetyltaurine, GAB amide, choline, putrescine) carry a net charge and might be expected to show a greater tendency to inter- act with proteins. Such direct interactions be- tween organic LMW and other components of the spiral strand, e.g., the adhesive glycopro- tein(s) or core fiber proteins, may be vital for the proper functioning of these adhesive spiral strand macromolecules (Townley et al. 1991; Gosline et al. 1995). The combination of solutes in the aqueous solution on the web’s adhesive spiral may also function as a counteracting solute system, wherein the perturbing influence to native macromolecular structure by one or more de- stabilizing solutes is offset by the presence of other, stabilizing, i.e., compatible, solutes (So- mero 1986; Timasheff 1992). The best studied example of such a system is the urea (desta- bilizer)/methylamine (stabilizer) system of marine cartilaginous fish, the coelacanth, and mammalian kidneys (Yancey et al. 1982; So- mero 1986, 1992; Garcia-Perez & Burg 1991; Yancey 1992). On the web, the perturbing in- fluence of inorganic ions and/or one or more of the organic LMW (especially net charged organic LMW) could be countered by other, stabilizing LMW. Optimal performance of ad- hesive spiral macromolecules in such a solute system may depend on the various LMW oc- curring at fairly specific concentration ratios to one another (Yancey et al. 1982; Somero 1986, 1992; Yancey 1992). In all three of these chemical models, dif- ferences in LMW composition, such as those documented herein, may directly translate into differences in macromolecular structure, with consequences for the adhesive spiral’s trap- ping ability. However, we must emphasize that at this time the ability of the web’s or- ganic LMW to affect macromolecular struc- ture by any of the aforementioned mecha- nisms is speculative. Whether the observed differences in web chemistry reflect adaptive responses to the environment or simply non- adaptive plasticity (Via 1993), these changes in LMW composition could be important both for web function and for physiological func- tion. Precursors or derivatives of the organic LMW, if not the LMW themselves, play im- portant physiological roles (e.g., as neuro- transmitters and in cell membrane phospholip- ids). However, orb-weaving spiders must invest LMW and other essential and non-es- sential compounds in the synthesis of the orb web because they are completely dependent upon the web for capturing prey. Although 92 THE JOURNAL OF ARACHNOLOGY web recycling allows the spider to recoup a portion of this investment (Breed et al. 1964; Peakall 1971; Townley & Tillinghast 1988), some loss of material is inevitable. Thus, with each web-building event, allocation “deci- sions” must be made; and there is experimen- tal evidence for trade-offs in allocation of lim- ited resources between foraging (the orb) and physiological demands (Higgins & Rankin 1999). Because orb-weaving spiders depend entirely on the web for capturing prey and be- cause the synthesis of the orb web requires an investment of physiologically important com- pounds, this group of spiders could become a model system for investigating resource allo- cation (Benforado and Kistler 1973; Higgins 1990, 1992, 1995; Higgins & Buskirk 1992; Sherman 1994, Blackledge 1998; Herberstein et al. 2000). The full realization of this potential will be facilitated by further investigation of orb web synthesis, recycling, and composition, partic- ularly of the adhesive spiral, which, even ne- glecting water content, often makes a consid- erably greater contribution to web weight than the non-adhesive web elements. ACKNOWLEDGMENTS The Instituto de Ecologia, UNAM, provid- ed logistical support and laboratory space for the portion of this study done in Mexico. Field work at Chamela and Los Tuxtlas was sup- ported by the staff of the Instituto de Biologia, UNAM, and field work in Texas was done by Jerry Drummond. Collecting permits were granted by the National Institute of Ecology in Mexico and Texas Parks and Wildlife. Drs. Craig Hieber and George Uetz kindly provid- ed large samples of Metepeira incrassata orb webs from which putrescine was first identi- fied. We are sincerely grateful to all these peo- ple and organizations. This work was sup- ported by NSF grant #IBN-922094 to MAR and LEH and USDA HATCH grant #370 to EKT. LITERATURE CITED Arakawa, T. & S.N. Timasheff. 1985. The stabili- zation of proteins by osmolytes. Biophysical Journal 47:411-414. Benforado, J. & K.H. Kistler. 1973. Growth of the orb weaver, Araneus diadematus, and correlation with web measurements. Psyche 80:90-100. Blackledge, TA. 1998. Stabilimentum variation and foraging success in Argiope aurantia and Ar- giope trifasciata (Araneae: Araneidae). Journal of Zoology, London 246:21-27. Bonthrone, K.M., F. Vollrath, B.K. Hunter & J.K.M. Sanders. 1992. The elasticity of spiders’ webs is due to water-induced mobility at a molecular level. Proceedings of the Royal Society of Lon- don, B, Biological Sciences 248:141-144. Breed, A.L., V.D. Levine, D.B. Peakall & P.N. Witt. 1964. The fate of the intact orb web of the spider Araneus diadematus Cl. Behaviour 23:43-60. Brown, A.D. & J.R. Simpson. 1972. Water rela- tions of sugar- tolerant yeasts: The role of intra- cellular polyols. Journal of General Microbiolo- gy 72:589-591. Craig, C.L., C. Riekel, M.E. Herberstein, R.S. We- ber, D. Kaplan & N.E, Pierce. In press. Evidence for diet effects on the composition of silk pro- teins produced by spiders. Molecular Biology and Evolution. Csonka, L.N. & A.D. Hanson. 1991. Prokaryotic osmoregulation: Genetics and physiology. An- nual Review of Microbiology 45:569-606. Dadd, R.H. 1985. Nutrition: Organisms. Pp. 313- 390. In Comprehensive Insect Physiology, Bio- chemistry and Pharmacology, Vol. 4, Regulation: Digestion, Nutrition, Excretion. (G.A. Kerkut & L.I. Gilbert, eds.). Pergamon Press, Oxford, UK. Dreesbach, K., G. Uhlenbruck & E.K. Tillinghast. 1983. Carbohydrates of the trypsin soluble frac- tion of the orb web of Argiope trifasciata. Insect Biochemistry 13:627-631. Eberhard, WG. 1985. The “sawtoothed” orb web of Eustala sp, (Araneae, Araneidae) with a dis- cussion of ontogenetic changes in spiders’ web- building behavior. Psyche 92:105-117. Eberhard, W.G. 1986. Ontogenetic changes in the web of Epeirotypus sp. (Araneae, Theridioso- matidae). Journal of Arachnology 14:125-128. Edmonds, D.T & E Vollrath. 1992. The contribu- tion of atmospheric water vapour to the forma- tion and efficiency of a spider’s capture web. Proceedings of the Royal Society of London, B, Biological Sciences 248:145-148. Edmunds, J. 1993. The development of the asym- metrical web of Nephilengys cruentata (Fabri- cius). Memoirs of the Queensland Museum 33: 503-506. Fischer, EG. & J. Brander. 1960. Eine Analyse der Gespinste der Kreuzspinne. Hoppe-Seyler’s Zeitsschrift fuer Physiologische Chemie 320:92- 102. Galinski, E.A. 1993. Compatible solutes of halo- philic eubacteria: Molecular principles, water- solute interaction, stress protection. Experientia 49:487-496. Galinski, E.A. 1995. Osmoadaptation in bacteria. Advances in Microbial Physiology 37:273-328. Garcia, E. 1973. Modificaciones al Systema de Classificacion Climatica de Koppen. Universidad HIGGINS ET AL.— VARIATION IN ORB WEB CHEMISTRY 93 Nacional Autonoma de Mexico Press, Mexico City. Garcia-Perez, A. & M.B. Burg. 1991. Renal med- ullary organic osmolytes. Physiological Reviews 71:1081-1115. Gosline, J., C. Nichols, P. Guerette, A. Cheng & S. Katz. 1995. The macromolecular design of spi- ders’ silks. Pp. 237-261, In Biomimetics: Design and Processing of Materials. (M. Sarikaya & LA. Aksay, eds.). American Institute of Physics Press, Woodbury, New York. Gosline, J.M., C.C. Poliak, PA. Guerette, A. Cheng, M.E. DeMont & M.W Denny. 1994. Elasto- meric network models for the frame and viscid silks from the orb web of the spider Araneus dia- dematus. Pp. 328-341. In Silk Polymers. Mate- rials Science and Biotechnology. (D. Kaplan, W.W Adams, B. Farmer, C. Viney, eds.). Amer- ican Chemical Society, Washington, D.C. Hayashi, C.Y & R.V. Lewis. 1998. Evidence from flagelliform silk cDNA for the structural basis of elasticity and modular nature of spider silks. Journal of Molecular Biology 275:773-784. Herberstein, M.E., C.L. Craig & M.A. Elgar. 2000. Foraging strategies and feeding regimes: Web decoration investment in Argiope keyserlingi Karsh (Araneae: Araneidae). Evolutionary Ecol- ogy Research 2:69-80. Higgins, L.E. 1990. Variation in foraging invest- ment during the intermolt interval and before egg-laying in the spider Nephila clavipes (Ara- neae: Araneidae). Journal of Insect Behavior 3: 773-783. Higgins, L.E. 1992. Developmental changes in barrier web structure under different levels of predation risk in Nephila clavipes (Araneae: Te- tragnathidae). Journal of Insect Behavior 5:635- 655. Higgins, L.E. 1995. Direct evidence for trade-offs between foraging and growth in a juvenile spider. Journal of Arachnology 23:37-43. Higgins, L.E. & R.E. Buskirk. 1992. A trap-build- ing predator exhibits different tactics for differ- ent aspects of foraging behaviour. Animal Be- haviour 44:485-499. Higgins, L.E. & M.A. Rankin. 1999. Nutritional requirements for web synthesis in the tetragnath- id spider Nephila clavipes. Physiological Ento- mology 24:263-270. Kinne, R.K.H. 1993. The role of organic osmolytes in osmoregulation: From bacteria to mammals. Journal of Experimental Zoology 265:346-355. Le Rudulier, D., A.R. Strom, A.M. Dandekar, L.T. Smith & R.C. Valentine. 1984. Molecular biol- ogy of osmoregulation. Science 224:1064-1068. Osaki, S. 1989. Seasonal change in color of spi- ders’ silk. Acta Arachnologia 38:21-28. Peakall, D.B. 1971. Conservation of web proteins in the spider, Araneus diadematus. Journal of Ex- perimental Zoology 176:257-264. Peters, H.M. 1955. liber den Spinnapparat von Ne- phila madagascariensis (Radnetzspinnen, Earn. Argiopidae). Zeitschrift fuer Naturforschung 10b:395-404. Peters, H.M. 1995. Ultrastructure of orb spiders’ gluey capture threads. Naturwissenschaften 82: 380-382. Ramousse, R. 1973. Body, web-building and feed- ing characteristics of males of the spider Araneus diadematus (Araneae: Araneidae). Psyche 80:22- 47. Richter, G. 1956. Untersuchungen iiber Struktur und Funktion der KlebefMen in den Fanggewe- ben ecribellater Radnetzspinnen. Naturwissen- schaften 43:23. Schildknecht, H., P. Kunzelmann, D. KrauB & C. Kuhn. 1972. Uber die Chemie der Spinnwebe, I: Arthropodenabwehrstoffe, LVII. Naturwissen- schaften 59:98-99. Schulz, S. 1997. The chemistry of spider toxins and spider silk. Angewandte Chemie Internation- al Edition in English 36:314-326. Sekiguchi, K. 1952. On a new spinning gland found in geometric spiders and its functions. An- notationes Zoologicae Japonenses 25:394-399. Sherman, P.M. 1994. The orb-web: An energetic and behavioural estimator of a spider’s dynamic foraging and reproductive strategies. Animal Be- haviour 48:19-34. Somero, G.N. 1986. Protons, osmolytes, and fit- ness of internal milieu for protein function. American Journal of Physiology 251:R197- R213. Somero, G.N. 1992. Adapting to water stress: Con- vergence on conunon solutions. Pp. 3-18. /n Wa- ter and Life: Comparative Analysis of Water Re- lationships at the Organismic, Cellular, and Molecular Levels. (G.N. Somero, C.B. Osmond & C.L. Bolis, eds.). Springer- Verlag, Berlin. Tillinghast, E.K. 1981. Selective removal of gly- coproteins from the adhesive spiral of the spiders orb web. Naturwissenschaften 68:526-527. Tillinghast, E.K, 1984. The chemical fractionation of the orb web of Argiope spiders. Insect Bio- chemistry 14:115-120. Tillinghast, E.K. & T. Christenson. 1984. Obser- vations on the chemical composition of the web of Nephila clavipes (Araneae, Araneidae). Jour- nal of Arachnology 12:69-74. Tillinghast, E.K., M.A. Townley, TN. Wight, G. Uhlenbruck & E. Janssen. 1993. The adhesive glycoprotein of the orb web of Argiope aurantia (Araneae, Araneidae). Pp. 9-23. In Biomolecular Materials. (C. Viney, S.T Case & J.H. Waite, eds.). Materials Research Society Symposium Proceedings, Vol. 292. Materials Research Soci- ety, Pittsburgh, Pennsylvania. 94 THE JOURNAL OF ARACHNOLOGY Tillinghast, E.K. & M.A. Townley. 1994. Silk glands of araneid spiders: Selected morphologi- cal and physiological aspects. Pp. 29-44. In Silk Polymers: Materials Science and Biotechnology. (D. Kaplan, W.W. Adams, B. Farmer & C. Viney, eds.). American Chemistry Society Symposium, Series 544. American Chemical Society, Wash- ington, D. C. Timasheff, S.N. 1992. A physicochemical basis for the selection of osmolytes by nature. Pp. 70-84. In Water and Life: Comparative Analysis of Wa- ter Relationships at the Organismic, Cellular, and Molecular Levels. (G.N. Somero, C.B. Osmond & C.L. Bolis, eds.). Springer- Verlag, Berlin. Townley, M.A. & E.K. Tillinghast. 1988. Orb web recycling in Araneus cavaticus (Araneae, Ara- neidae) with an emphasis on the adhesive spiral component, GABamide. Journal of Arachnology 16:303-319. Townley, M.A., D.T. Bernstein, K.S. Gallagher & E.K. Tillinghast. 1991. Comparative study of orb web hygroscopicity and adhesive spiral com- position in three araneid spiders. Journal of Ex- perimental Zoology 259:154-165. Via, S. 1993. Adaptive phenotypic plasticity: Tar- get or by-product of selection in a variable en- vironment? American Naturalist 142:352-365. Vollrath, E & D.T Edmonds. 1989. Modulation of the mechanical properties of spider silk by coat- ing with water. Nature 340:305-307. Vollrath, E, WJ. Fairbrother, R.J.P. Williams, E.K. Tillinghast, D.T. Bernstein, K.S. Gallagher & M.A. Townley. 1990. Compounds in the drop- lets of the orb spider’s viscid spiral. Nature 345: 526-528. Vollrath, F. & E.K. Tillinghast. 1991. Glycoprotein glue beneath a spider web’s aqueous coat, Na- turwissenschaften 78:557-559. Wilkinson, L. 1992. SYSTAT: Statistics, Version 5.2 Edition. Systat, Inc., Evanston, Illinois. Witt, P.N., J.O. Rawlings & C.F. Reed. 1972. On- togeny of web-building behavior in two orb- weaving spiders. American Zoologist 12:445- 454. Yancey, RH. 1992. Compatible and counteracting aspects of organic osmolytes in mammalian kid- ney cells in vivo and in vitro. Pp. 19—32. In Wa- ter and Life: Comparative Analysis of Water Re- lationships at the Organismic, Cellular, and Molecular Levels. (G.N. Somero, C.B. Osmond & C.L. Bolis, eds.). Springer- Verlag, Berlin. Yancey, RH., M.E. Clark, S.C, Hand, R.D. Bowlus & G.N. Somero. 1982. Living with water stress: Evolution of osmolyte systems. Science 217: 1214-1222. Manuscript received 13 May 2000, revised 13 No- vember 2000. 2001. The Journal of Arachnology 29:95-103 PATTERNS OF ABUNDANCE OF FOUR SPECIES OF WANDERING SPIDERS (CTENIDAE, CTENUS) IN A FOREST IN CENTRAL AMAZONIA Thierry R. Gasnierj Depto. de Biologia/ICB, Fundagao Universidade do Amazonas, Av. Gal. R.O. J. Ramos 3000, CEP 69067-000, Manaus, AM, Brazil Hubert Hdferi Staatliches Museum fur Naturkunde Karlsruhe, Postfach 6209, D- 76042 Karlsruhe, Germany ABSTRACT. We studied spatial and temporal patterns of abundance of Ctenus amphora, C. crulsi, C. manauara and C. villasboasi, four syntopic species of medium-to-large sized wandering spiders that forage on the ground in a neotropical rainforest. We found temporal variation, apparently seasonal, in abundance for two of the four species. The four species are sympatric in the study area, but with very distinct spatial patterns of abundance. Ctenus amphora was more abundant in areas with sandy soil but are also common on clay soils, C. manauara and C. crulsi are the dominant species in areas with clay soil and are infrequent in sandy soils, and C. villasboasi had a more homogenous abundance in the study area. Previous studies suggested that a predator, army ants, could have an important impact on the abundance of these spiders. We estimated the frequency of attacks by army ants using pitfall traps in sandy and clay soil areas. The estimated probability of attack by army ants was higher in areas with clay soil (92% per 3 months), where all species are frequently found, than in sandy soil areas (21%), where C. crulsi and C. manauara were almost absent. However, it is still not clear if predation by army ants is a key factor that facilitates coexistence in clay soils, and this factor can not explain the difference on the dominant species between areas with different soil types. We also discuss the description of spatial patterns of abundance as a simple, but powerful, tool seldom used for preliminary studies on the coexistence of spiders. RESUMO. Nos estudamos padroes espaciais e temporais de abundancia de Ctenus amphora, C. crulsi, C. manauara e C. villasboasi, quatro especies sintopicas de aranhas errantes que forrageiam no chao em uma floresta neotropical umida. Nos encontramos uma variagao temporal, aparentemente sazonal, na abun- dafancia de duas das quatro especies. As quatro especies sao simpatricas na area de estudo, mas com padroes espaciais de abundancia muito distintos. Ctenus amphora foi mais abundante em areas de solos arenosos, mas tambem foi comum em solos argilosos C manauara e C. crulsi foram as especies domi= nantes em solos argilosos, e foram infreqiientes em solos arenosos, e C. villasboasi teve uma abundancia mais homogenea na area de estudo. Estudos anteriores sugeriram que um predador, formigas de correigao, poderia ter um forte impacto sobre a abundancia destas aranhas. Nos estimamos a freqiiencia de ataques por formigas de correi^ao usando armadilhas de fosso (pitfall traps) em areas de solo arenoso e argiloso. A probabilidade estimada de ataques por formigas de correi^ao foi maior em ^eas de solo argiloso (92% em 3 meses), onde todas as especies sao freqiientemente encontradas, que em solo arenoso (21%), onde C crulsi e C. manauara foram raras. Entretanto, ainda nao esta claro se a preda9ao e um fator chave para facilitar a coexistencia em solos argilosos, e este fator nao pode explicar a diferenga de especies dominantes entre as &eas com tipos de solo diferentes. Nos tambem discutimos a descrigao de padroes espaciais de abundancia como um ferramenta simples, mas poderosa e pouco usada para o estudo da coexistencia de aranhas. Keywords: Army ants, predation, coexistence, method Studies on the ecology of wandering spi- ders have been performed mainly in temperate regions and mostly with spiders of the family Lycosidae (e.g., Edgar 1971a, b; Ford 1977, 1978; Greenstone 1978, 1979, 1980; Hallan- der 1970 a, b; Suwa 1986; Van Dyke & Low- rie 1975; Wise 1993). Although the family Ctenidae is rich in species and abundant in the tropics, there are few studies on their ecology. One exception is the genus Cupiennius, which 95 96 THE JOURNAL OF ARACHNOLOGY was intensively studied by Barth and collab- orators (e.g., Schuster et al. 1994 and citations in it); however, in this genus only the males are considered wandering spiders (Schmitt et al. 1990). Hofer, Brescovit & Gasnier (1994) studied the wandering spiders of the genus Ctenus (Walckenaer 1805) in “Reserva Florestal Adolfo Ducke” (READ) in central Amazonia. They found seven species, four of which {Ctenus amphora Mello-Leitao 1930; C. crul- si Mello-Leitao 1930; C. manauara Hofer, Brescovit & Gasnier 1994; and C villasboasi Mello-Leitao 1949) are very similar in behav- ior and use of microhabitat. They forage in and on the leaf litter and are the dominant medium-to-large sized wandering spiders on the ground in most parts of this forest. Vieira & Hofer (1994, 1998) studied swarm-raiding army ants {Eciton burchelli and Labidus praedator) in central Amazonia (100 km north of READ) and concluded that the spiders of the genus Ctenus are among the major prey items of these ants. The effect of army ants on ground-living spiders, including Ctenus, was also discussed by Gasnier, Hofer & Brescovit (1995). These ants forage in mas- sive groups of many thousands of individuals, and these authors suggest that they probably have a strong impact on the abundance and on the structure of this guild of wandering spi- ders. However, although Ctenus is an impor- tant prey for army ants, the impact of the ants on the spiders is unknown because there is no estimation of how frequently they pass by a given area. The objective of this study was to describe spatial and temporal patterns of abundance of these four Ctenus species and to evaluate how army ants influence them. METHODS Study area. — The study was conducted in “Reserva Elorestal Adolfo Ducke” (READ), 25 km North of Manaus, Amazon, Brazil. The reserve has 10,000 ha of “terra-firme” pri- mary forest, over poor soils of tertiary origin (Chauvel, Lucas & Boulet 1987). Collections were made in an area of 2 X 5 km (Eig. 1). The northern half of this area is formed by plateaus, slopes and flat valley bottom of the stream “Barro Branco.” In this area clay soils (“latossolo amarelo alico” or “aplic acror- thox”) predominate, mainly in the plateaus and slopes, but there are sandy soils (“podzol Figure 1. — Study area and the trails used in the extensive censuses. B = start and E = end of the main 12 km trail; arrows indicate the direction fol- lowed in this trail. “XI,” “X2” and “X3” are ad- ditional trails. A = administration of the reserve. alico” or “arenic tropaquod”) in some of the lower parts of the slopes and in the valley bot- tom. Close to the streams the hydromorphic soils (“podzolicos vermelho-amarelo latosso- licos” or “epiaquic paleudults”) dominate. This area is covered by a dense forest called “Floresta de Terra-Firme” (in the restricted sense), except for the valley bottom, where the forest is dense, but lower, more humid and with more palms, a vegetation called “Flores- ta de Baixio” (descriptions in Guillaumet 1987). The southern half is the drainage basin of the stream “Acara.” Clay soils are present on the more elevated areas next to the pla- teaus, but most of the work in the “Acara” valley was on sandy areas. The vegetation in this area is lower than the “terra-firme” forest on clay soils, and is generically called “Cam- pinarana” (Guillaumet 1987). Next to the streams in the southern half there is also a predominance of hydromorphic soils. The temperature in the region of Manaus varies little annually, with a mean temperature of GASNIER & HOFER— PATTERNS OF ABUNDANCE IN CTENUS 91 25.8 °C in February and 27.9 °C in September (Salati et al. 1991). Mean annual rainfall in RFAD is 2480 mm, with a dry season from July-November (Marques-Filho et al. 1981). Data presented here were collected from No- vember 1995 to March 1997. Identification of spiders. — The size (mea- sured as prosoma length) of adults varies among species: C. manauara (4-6 mm); C. crulsi (5.5-10.5 mm); C. amphora (5.5-11 mm); C. villasboasi (10-12.5 mm). Identifi- cation of spiders at species level is based mainly on reproductive structures (palp in male and epigynum in females), but inside RFAD all species of Ctenus are known, and the adults and larger juveniles can be identi- fied based on color and design patterns on their body (descriptions and photographs in Hofer, Brescovit & Gasnier 1994). However, in the smaller juveniles the patterns of color and designs may be confused. During a pre- liminary phase of this study, we compared color and design patterns in juveniles of dif- ferent sizes of the four species to provide cri- teria to distinguish the juveniles of each spe- cies in our study area. This was important because it allowed us to get much more data for the evaluation of abundance patterns. We present these criteria below. Vouchers were deposited in the aracnological collection of the Instituto Nacional de Pesquisas da Ama- zonia under the numbers INPA-001 to INPA- 023. On the back side of the opisthosoma of these Ctenus there is a white design similar to an amphora (Greek jar), generally followed by a series of triangles (a very similar pattern is also found in sympatric Centroctenus, which made us believe that this is a primitive char- acteristic for this and related genera). In the adults of C. villasboasi this design almost al- ways disappears completely, but it is still vis- ible in the smaller juveniles. Ctenus villas- boasi can be identified after they reach 2-3 mm because the anterior part of the design is brighter and has the shape of the letter “U.” Besides that, individuals larger than 5 mm have a distinct longitudinal white line on the ventral surface of the prosoma, unique in C villasboasi. In C. amphora the triangles of the design almost always have disappeared com- pletely in juveniles of size 3-4 mm, resulting in the typical amphora-shaped marking. Be- sides that, the pattern of coloration of the body, especially the ventral surface of the op- isthosoma, is generally very dark. Ctenus crulsi and C. manauara have the complete pattern of the design (amphora and triangles) from eclosion from the egg sac through the adult phase. Adults and subadults can be distinguished since adults of C. man- auara are smaller than adults of C. crulsi, and the presence of the external copulatory organ (developed in adults or underdeveloped in subadults) is visible to the naked eye. Fur- thermore, they have some coloration differ- ences, especially on the venter. Adults of C. manauara generally are brighter brown on the ventral surface, sometimes pink like the small- er juveniles of all species, and generally with a black spot near the spinnerets. The other species, including C crulsi, have a large tri- angular design covering most of the ventral surface after they reach 3-4 mm. However, throughout the study, we found this pattern with ventral triangles on the opisthosoma also in adults of C. manauara, and possibly the ventral triangle may be delayed in appearing in the juveniles of C. crulsi. This means that we could have misidentified some juveniles of C. manauara as C crulsi and vice versa. Therefore, we counted as identified only spi- ders larger than 4 mm, which was probably sufficient in avoiding misidentifications. Censuses. — We captured spiders at night only, using rechargeable battery head lamps (Koehler Electric Cap Lamp). With these lamps the reflection of the light by the eyes of the spiders is visible up to 25 m, even for small individuals, but this depends on the po- sition of the spider, the amount of litter and the density of the lower vegetation. Even when the position does not favor the reflection of the light, spiders may be localized by rec- ognizing the spider’s body or part of it. In the censuses, the spiders were captured, identified, and immediately released at their capture site to minimize disturbance of the population. We employed two types of cen- suses: intensive censuses, for evaluation of temporal variation of the abundance of spi- ders, and extensive censuses, for evaluation of spatial patterns of abundance. In the intensive censuses we counted, at intervals of about two months, spiders in 10 small transects (each one of 60 X 1.5 m) per collection for a total of seven collections. The search was inten- sive, and we tried to capture all visible spiders 98 THE JOURNAL OF ARACHNOLOGY with prosoma length larger than 4 mm, which required 20 minutes to one hour per transect. Except for the first collection trip, the tran- sects were in the same 10 defined areas, but in different positions inside the areas at each trip. With this procedure, we tried to include the variability inherent in the system and to obtain similar samples throughout the year, al- though not at exactly the same points. After the first sampling we realized that C. crulsi and C. manauara were nearly absent from all 9 selected areas, so we added an area, where individuals of these species were known to be abundant. An index to standardize the effort was applied to the abundance data for the first trip (abundance per species X 1.11), which compensates for one less transect, but this does not avoid an underestimation for the lat- ter two species because this corrected mean still lacks the area where they are more abun- dant. We identified and recorded the position of spiders on 15 km of trails in the extensive censuses, in June and October of 1995 and in February of 1996. Most of the trails consti- tuted an almost continuous line of 12.5 km (Fig. 1), with small interruptions to avoid dis- turbed areas. The other three trails of about 1 km were outside this continuous line, but were included because they were the spatial ex- tremes of the study area. These three trails are the last three segments in the graphical pre- sentation of the abundance of spiders along the trails. The search for spiders was less in- tensive than in the previous census; and, de- pending on their abundance, we spent 30 min- utes to two hours counting spiders in 1 km. The number of spiders found by this method was low per meter of transect, especially for small spiders, like juveniles and C. manauara (the smallest species), but the total number of spiders, even of C. manauara, was high. The extensive censuses were used to show graphically how the abundance changed along these long transects and to calculate covaria- tion indices among species. The trails were divided into segments of 100 m, which were our sample units. Data used to describe abun- dance graphically were the number of individ- uals of each species in each excursion per sample unit. We excluded from the covaria- tion analyses sample units close to disturbed areas, and half of the sample units, avoiding contiguous segments of 100 m, to minimize the dependency of the data. Data used for the evaluation of covariation were the sum of spi- ders in each sample unit in the three excur- sions. Based on the recapture rate of 100 marked adult spiders, we concluded that the sum of spiders per sample unit in three ex- cursions was not biased by the fact that spi- ders could be counted more than once, be- cause only 1% of the marked animals were found in one transect after one month. Soils type definition in the sample units consisted on a division first in hydromorphic and non- hydromorphic soils, and a division of the non- hydromorphic in clay and sandy soils. This categorization was rough, but sufficient for the analyses, as most places were inside one of the extremes of soil types. We used Pearson’s index (Ludwig & Reynolds 1988) for the de- scription of the interspecific covariation by pairs of species, excluding sample units where both elements of the pair were absent. Estimating the frequency of raid occur- rence of army ants. — We installed 40 pitfall traps, with a minimum distance between each of them of 200 m. Each trap consisted of a plastic container (diameter of opening 8 cm, 20 cm deep) buried flush with the surface, and an aluminum cover 10 cm above the trap for protection against rain, not obstructing the en- trance of invertebrates. We used picric acid inside the trap as a preservative liquid. Two samples were made for periods of 90 days, from March-June 1995, and from October- February 1996. The samples were preserved in 70% alcohol. We counted the number of army ants of the species Eciton burchelli and Labidus predator, which are the swarm-raid- ing species that attack Ctenus. Considering that a spider could be where a trap was, we used the proportion of traps with more than 10 army ants of the same species (a probable raid), as an index of frequency of risky en- counters for Ctenus. RESULTS We recorded 494 individuals of Ctenus dur- ing seven excursions with intensive censuses and 1304 individuals during three extensive censuses. The abundance of Ctenus amphora and C. villas boas i was high in December 1994, gradually decreased until July, and in- creased again after August (Fig. 2), suggesting seasonal variation. The abundance of Ctenus manauara was also low during the dry season. GASNIER & HOFER— PATTERNS OF ABUNDANCE IN CTENUS 99 Figure 2. — -Number of individuals per species collected in the intensive censuses throughout one year. Because the census value for December 1994 was probably underestimated (see methods), it is reasonable to assume that it followed the same pattern of the above species. Data from intensive censuses, which was used to de- scribe the patterns above, were not enough to evaluate temporal variation for C crulsi. However, the totals of C crulsi counted in the three extensive censuses in July, October and February 1995 were respectively 63, 129 and 191, a similar temporal variation compared to C. amphora, 77, 171 and 235, which means that all species have similar temporal patterns of abundance— or at least that they do not dif- fer strongly. Comparing the abundance between pairs of species per sample unit, we found significant negative covariation only in the pairs C am- phora X C. crulsi and C amphora X C man- auara (Table 1). The comparison of spatial patterns of abundance (Fig. 3) facilitated the interpretation of these negative correlations. There were large areas with dominance of C amphora where C. crulsi and C manauara had low abundance, and vice versa. Although the abundances of C crulsi and C. manauara were not positively correlated, the compari- sons revealed very similar spatial patterns of abundance. Apparently, covariance analysis was not appropriate to evaluate patterns of abundance in this case because the number of individuals per sampling unit was low and be- cause there were many points where both spe- cies of the pair were absent. There were three consistent large scale spatial patterns of abun- dance: C. amphora had the highest densities between positions 80 and 150; C crulsi and C manauara had the highest densities be- tween 1-80 and between 150-180; and C vih lasboasi had a relatively homogeneous distri- bution. The total abundance varied through time, but the spatial patterns of abundance within each species were very similar in the three censuses. The positions 80-150, where C am- phora was more abundant, were mostly on sandy (55%) or hydromorphic (38%) soils, while the positions 1-80 and 150-180, where C crulsi and C. manauara were more abun- dant, were mostly on clay (64%) or hydro- morphic {21%) soils. There are two pieces of evidence suggesting that soils could, directly or indirectly, determinate which species dom- inates an area: a) the change in dominance from C. crulsi and C manauara to C ampho- ra between positions 70 and 90 was coinci- dent with a change in predominant soil from clay to sandy; b) C crulsi and C. manauara were almost absent in a large area of white sandy soils, but both appeared in two censuses inside this area, in the only place with a small segment of about 500 m of intermediate sandy-clay soil in the trail. Table L- — Interspecific covariation between pairs of Ctenus species. Pearson = Pearson’s correlation coefficient after excluding sample units where both species in the pair were absent; P = probability associated to the correlation; n = sample size; Sign = significance calculated considering 6 tests: * = significant, ns = non significant. Pairs of species Pearson P n Sign C. amphora X C. crulsi -0.48 <0.001 67 * C. amphora X C. manauara -0.34 0.006 64 C. amphora X C villasboasi 0.00 0.99 62 ns C. crulsi X C. manauara 0.02 0.91 49 ns C. crulsi X C villasboasi -0.23 0.07 61 ns C. manauara X C. villasboasi -0.26 0.05 57 ns 100 THE JOURNAL OF ARACHNOLOGY C amphora C. cruisi June 1995 10- 8- October 1995 12-1 10- February 1996 100 C viiiasboasi C manauara June 1995 2- ^1 June 1995 3- ■q. m O E "•■j October 1995 J- 2- 1- 0- - •"T 10-1 8- October 1995 6“| February 1996 12-j 10- 8- February 1996 Position (units of lOOm) 6- 4-j • 2- 0- — 1 200 50 100 150 Position (units of lOOm) Figure 3. — Number of individuals collected along the extensive censuses trails for each species in three periods. Three patterns of abundance were detected: one for Ctenus amphora, one for Ctenus cruisi and Ctenus manauara, and one for Ctenus viiiasboasi. Considering these differences in relation to soils, we calculated two indices of frequency of risky encounters with army ants, one for clay soil areas and another for sandy soil ar- eas. The traps in hydromorphic soils areas were in insufficient number to be included in the analysis. The proportion of traps with army ants and the amount of army ants were higher in the clay soil areas (Table 2). Clay soil areas had a much higher index of risk of encounters with army ants (0.92) than sandy soil areas (0.21). We suppose that life span for Ctenus is between 6-12 months; therefore, it is highly probable that every Ctenus individ- ual encounters army ants at least once during its life in clay soils, while many Ctenus will GASNIER & HOFER— PATTERNS OF ABUNDANCE IN CTENUS 101 Table 2. — Number of pitfall traps with different amounts of army ants in sandy and clay soils in two periods of the year. Soil Absence of ants 1-10 ants 10-100 ants >100 ants April-June 1995 Clay 1 1 2 7 Sandy 1 2 1 1 December 1 995-February 1996 Clay 0 0 8 7 Sandy 3 5 1 0 probably never encounter army ants on sandy soils. DISCUSSION The four Ctenus species are sympatric throughout the study area. However, there were strong differences in the spatial patterns of abundance of the species, which were sta- ble during the study. Although the amount of spiders changed throughout the year, the gen- eral pattern of locational dominance remained. The predominant species in a given place is probably determined by local characteristics of the environment, which favor one species more than the other. However, it is not clear what these environmental characteristics are, and why the favored species in a given place do not exclude the others. Based on our data, we can conclude that army ants are important predators of Ctenus, but a more detailed study would be necessary to determinate if they are a key factor that facilitates coexistence in this system. One ev- idence that the ants may be important for co- existence is that on clay soils, where army ants are more abundant, the four species are relatively common. However, predation alone would not explain why the dominant species is not the same all over the area. The temporal variation in the abundance of spiders probably reflected seasonal variation of the environment. Evidence of seasonal var- iation in the abundance of ctenid-pisaurid spi- ders was not detected by Gasnier et al. (1995) in this forest; however, this was observed in a year with less pronounced seasons. The cause of variation in the present study is not clear, there are many possible reasons, e.g., a seasonal variation of the amount of leaf litter could restrict the amount of refuges or prey available, or the absence or excess of rain could cause a seasonal variation in mortality. Whatever the cause, seasonality may facilitate the coexistence of species either by maintain- ing the species under a level in which their interactions would influence coexistence or by differentiation in seasonal peaks of activity or abundance. Sympatric species of forest-floor spiders may differ by seasonal peaks of abun- dance (Niemela et al. 1994), but we did not find this difference among the Ctenus species. For a further discussion on how seasonality could affect coexistence in this system it will be necessary to evaluate the mechanism by which environmental seasonality affects the abundance of these spiders. Evaluations of spatial patterns of abundance or distribution (e.g.. Cutler & Jennings 1992; Femandez-Montraveta, Lahoz-Beltra & Orte- ga 1991; Greenstone 1979, 1980) are not fre- quent in the spider literature. Authors defend- ing the use of an experimental approach as the safer mode to evaluate coexistence of species recognize the importance of basic knowledge (e.g., spatial patterns of abundance) to plan and interpret the experiments (e.g., Hairston 1989; Wise 1993). However, there is little dis- cussion on the procedures of how to build this knowledge. The use of interspecific covaria- tion is one of the standard forms to detect pat- terns of coexistence (Ludwig & Reynolds 1988). However, in our study, the evaluation by interspecific covariation did not show any evidence for the similar patterns of abundance of C crulsi and C manauara. We recommend the interpretation of the indices of covariation in conjunction with an evaluation of spatial patterns of abundance. Evaluations based on graphics of abundance along transects (or on bidimensional maps of abundance) may help to detect important factors that affect a spe- cies, especially if the repetition of the census- es indicates that the pattern is stable in time, which probably reflects a local factor. After the pattern is established, and considering the dimension of the areas of higher abundance, the researcher may go back to the field and consider potential factors affecting the abun- dance. During the comparison of spatial pat- terns of abundance among species by super- position of graphs unexpected differences may arise, which makes this a powerful tool in the development of hypothesis for the abundance and coexistence of species. 102 THE JOURNAL OF ARACHNOLOGY ACKNOWLEDGMENTS We are indebted to the workers of Reserva Florestal Adolfo Ducke for their hospitality and help. We thank to the people from the Projeto Flora at READ for the help, especially to Alberto Vicentini for the detailed maps of the trails. The PDBFF project a convenium between the Institute Nacional de Pesquisas da Amazonia (INPA) and the Smithsonian In- stitute supported an excursion to their re- serves, which was important to establish ques- tions and methodology for the study of Ctenus. This paper is part of the thesis of the first author in the post-graduation program of the “Convenio INPA/UFAm.” We thank Lu- cile Anthony, Friedrich Barth, Antonio Bres- covit, Harold Fowler, Christopher Martins, Gary Polis and David Wise for the sugges- tions on this thesis, and Jim Berry and Petra Sierwald for the useful comments on a pre- vious manuscript. Financial support came from a fellowship grant from CAPES and field support grants from the German Science Foundation (DFG -proj. Prof. Dr. L. Beck) the German Society for Technical Cooperation (GTZ-project BE 281), and the Conselho Na- cional de Desenvolvimento Cientifico e Tec- nologico (CNPq-project 400023/98). LITERATURE CITED Chauvel, A., Y. Lucas & R. Boulet. 1987. On the genesis of the soil mantle of the region of Ma- naus, Central Amazonia, Brazil. Experientia 43: 234-241. Cutler, B. & D.T. Jennings. 1992. Habitat segre- gation by species of Metaphidippus (Araneae: Salticidae) in Minnesota. Journal of Arachnology 20:88-93. Edgar, W.D. 1971a. Seasonal weight changes, age structure, natality and mortality in the wolf spi- der Pardosa lugrubis Walck. in Central Scotland. Oikos 22:84-92. Edgar, W.D. 1971b. Aspects of the ecological en- ergetics of the wolf spider Pardosa (Lycosa) lu- gubris (Walckeanaer). Oecologia 7:136-54. Fernandez-Montraveta, C., R. Lahoz-Beltra & J. Ortega. 1991. Spatial distribution of Lycosa tar- entula fasciiventris (Araneae, Lycosidae) in a population from central Spain. Journal of Arach- nology 19:73-79. Ford, M.J. 1977. Metabolic cost of the predation strategy of the spider Pardosa amentata (Clerck) (Lycosidae). Oecologia 28:333-340. Ford, M.J. 1978. Locomotory activity and the pre- dation strategy of the wolf spider Pardosa amen- tata (Clerck) (Lycosidae). Animal Behaviour 26: 31-35. Gasnier, T.R., H. Hofer & A.D. Brescovit. 1995. Factors affecting the activity density of spiders on tree trunks in an Amazonian rainforest. Eco- tropica 1:69-77. Greenstone, M.H. 1978. The numerical response to prey availability of Pardosa ramulosa (McCook) (Araneae: Lycosidae) and its relationships to the role of spiders in the balance of nature. Symposia of the Zoological Society of London 42:189-93. Greenstone, M.H. 1979. A line transect density in- dex for wolf spiders {Pardosa spp.), and a note on the applicability of catch per unit effort meth- ods to entomological studies. Ecological Ento- mology 4:23-29. Greenstone, M.H. 1980. Contiguous allotopy of Pardosa ramulosa and Pardosa tuoba (Araneae: Lycosidae) in San Francisco Bay Region, and its implication for the patterns of resource partition- ing in the genus. American Midland Naturalist 104:305-311. Guillaumet, J.-L. 1987. Some structural and floris- tic aspects of the forest. Experientia 43:241—251. Hairston, N.G. 1989. Ecological Experiments: Pur- pose, Design, and Execution. Cambridge Univ., Cambridge. Hallander, H. 1970a. Environments of the wolf spi- der Pardosa chelata (O.E Miihller) and Pardosa Pullata (Clerk). Ekologia Polska 18:41-72. Hallander, H. 1970b. Prey cannibalism and micro- habitat selection in the wolf spiders Pardosa che- lata O.E Miihller and P. pullata Clerck. Oikos 21:337-340. Hofer, H., A.D. Brescovit & T. Gasnier. 1994. The wandering spiders of the genus Ctenus (Cteni- dae, Araneae) of Reserva Ducke, a rainforest re- serve in central Amazonia. Andrias 13:81-98. Ludwig, J.A. & J.F. Reynolds. 1988. Statistical Ecology. A Primer on Methods and Computing. Willey & Sons. New York. Marques-Filho, A.O., M.N.G. Ribeiro, H.M. dos Santos, J.M. dos Santos, 1981. Estudos clima- tologicos da Reserva Florestal Ducke. IV. Pre- cipita9ao. Acta Amazonica 11:759-768. Manaus. Niemela, J., T. Pajunen, Y. Haila, P. Punttila & E. Halme. 1994. Seasonal activity of boreal forest- floor spiders (Araneae). Journal of Arachnology 22:23-31. Salati, E., M.N.G. Ribeiro, M.L. Absy & B.W Nel- son. 1991. Clima da Amazonia: Presente, Pas- sado e Futuro. Pp. 21-34, In Bases Cientificas para Estrategias de Preservagao e Desenvolvi- mento da Amazonia: Fatos e Perspectivas, Vol 1 . (A.L. Val , R. Figliuolo & E. Feldberg, eds.) In- stituto Nacional de Pesquisas da Amazonia, Ma- naus. Schmitt, A., M. Schuster & EG. Barth. 1990. Daily GASNIER & HOFER— PATTERNS OF ABUNDANCE IN CTENUS 103 locomotor activity patterns in three species of Cupiennius (Araneae, Ctenidae): The males are the wandering spiders. Journal of Arachnology 18:249-255. Schuster, M., D. Baurecht, E. Mitter, A. Schmitt & E Barth. 1994. Field observations on the pop- ulation structure of three ctenid spider {Cupien- nius, Araneae, Ctenidae), Journal of Arachnolo- gy 22:32-38. Suwa, M. 1986. Space partitioning among the wolf spider Pardosa amentata species group in Hok- kaido. Japanese Research in Population Ecology 28:231-252. Van Dyke, D. & D. Lowrie. 1975. Comparative life histories of the wolf spider Pardosa ramulosa and P. sierra (Araneae: Lycosidae). Southwest- ern Naturalist 20:29-44. Vieira, R.S. & H, Hofer. 1994. Prey spectrum of two army ant species in central Amazonia, with special attention on their effect on spider popu- lations. Andrias 13:189-198. Vieira, R.S. & H. Hofer. 1998. Efeito do forragea- mento de Eciton burchelli (Hymenoptera, For- micidae) sobre a araneofauna de liteira em uma floresta tropical de terra-firme na Amazonia Cen- tral. Acta Amazonica 28:345—351. Wise, H.D. 1993. Spiders in Ecological Webs. Cambridge Univ,, Cambridge. Manuscript received 6 November 1998, revised 10 May 2000. 2001. The Journal of Arachnology 29:104-110 ONTOGENETIC CHANGE IN COLORATION AND WEB-BUILDING BEHAVIOR IN THE TROPICAL SPIDER ERIOPHORA FULIGINEA (ARANEAE, ARANEIDAE) Barbara Graf and Wolfgang Nentwig: Zoological Institute, University of Bern, Baltzerstrasse 6, 3012 Bern, Switzerland ABSTRACT. Eriophora fuliginea (Araneae, Araneidae), a tropical orb-weaving spider from Panama, undergoes a dramatic color change in the course of its ontogenesis. The first free instar has an amber opisthosoma, which soon becomes bright yellow, later green. Subadults change to olive, adults are dark brown with a white median stripe. Parallel to this color modification the spider’s behavior changes as well. The main activity phase shifts from day to night and web architecture changes from symmetrical horizontal orb-webs on the upper side of leaves with the spider on the hub, to asymmetrical horizontal orb-webs between shrubs with spiders in a rolled leaf nearby. Keywords: Ontogenetic change, development, coloration, behavior, orb web Quite often in the field, particularly in the tropics, one encounters colorful, juvenile spi- ders. When kept in the laboratory in many cases they change their coloration, frequently in a species-specific ontogenetic sequence. Since juvenile spiders normally cannot be identified, such coloration changes usually are not well documented, but may be very com- mon. Only a few cases have been reported in the literature. The European sparassid spider Micrommata virescens (Clerck 1757) is an ex- ample of a spider of temperate regions where the males change from light green, to yellow, to greenish-yellow with red stripes or dots (Homann 1946). Bonnet (1929, 1930) de- scribes the distinctive changes in color and pattern in Nephila madagascariensis (Vinson 1868). Edmunds & Edmunds (1986) report on two araneid species that show a distinct on- togenetic color change. Araneus rufipalpis (Lucas 1858) changes its color from a juvenile bright green, to a greyish-green or greyish- brown, to an adult dark brown. In Gastera- cantha curvispina (Guerin 1837), the juvenile spiders are flecked white and brown, whereas the adults are white, yellow, orange, brown, red, or even striped in light and dark colors. Different coloration of juveniles and adults within one species could indicate that differ- ent instars use different niches to reduce in- traspecific competition. Such avoidance of competition between adults and their own off- spring would be reasonable (e.g., Begon et al. 1996) and could include selection of different habitats, diurnal activities or food. An onto- genetic color change would support such im- portant changes in a spider’s life, but reports on the coincidence of color change and niche use are rare. During an investigation on the prey com- position of large orb-weaving spiders (Nen- twig 1985), unknown small green juvenile spi- ders encountered during field research in Panama developed in the laboratory into dark brown adult spiders and were later identified as Eriophora fuliginea (C.L. Koch 1839) (Ar- aneae, Araneidae). Later, dark brown females of the same species were caught in the field and built a cocoon in the laboratory from which amber juvenile spiders emerged, A se- quence of the differently colored developmen- tal stages was described in clutches of two females and was correlated to changes in the behavior that the juvenile spiders underwent as they matured. METHODS Two adult female E. fuliginea were caught in Panama (Parque Nacional Soberania, Gam- boa) and brought to our laboratory. They were kept in cages (32 X 35 X 10 cm) made of wood and wire in a climate chamber (25-30 °C, 40-50% RH, L:D = 16:8 h), where they built one cocoon each on days 6 and 9 after 104 GRAF & NENTWIG— ONTOGENETIC COLORATION CHANGE 105 capture. After 29-32 days, the juvenile spi= ders hatched from the cocoons. The spider- lings (n = 200) were kept individually in transparent plastic boxes with lids (20 X 20 X 8.5 cm) which were set up on edge within a climate chamber (25-30 °C, 40-50% RH, L: D — 12.5:11.5 h). The bottom of each box was covered with plaster, which was kept damp to ensure a humidity within the boxes of approximately 100%. A cotton thread was attached to the inside of each box to provide fixing points for webbuilding. Flies of various species {Drosophila melanogaster, Musca do- mestica, CalUphora erythrocephala. Proto- phormia terraenovae) were provided ad libi- tum as food, with the fly size dependent on spider size. Due to high mortality especially in instars III and IV (days 20-40) and due to killing spiders for section preparations, only 40 spiders became adult. The coloration and pattern of the growing spiders were recorded, sketched and photo- graphed from hatching until their death as adult animals. The bases for the colorations and patterns were observed on live animals, on preparations of animals freshly killed with CO2, and on sections. For each main color form, the opisthosoma of 2-3 freshly killed spiders were fixed in parts in 2% osmium te- troxide in potassium dichromate buffer ac- cording to Dalton (Glauert 1974). The prepa- rations were then studied with the optical or electron microscope for the position of pig- ment granula, A white coloration is caused by guanine crystals which are deposited in guan- ocytes and can easily be recognized by the aid of Holl (1987). Yellow, red and brown colors are caused by different ommochromes, which are stored intracellurlarly as membrane- shielded granules. Green colors usually derive from linear tetrapyrroles (Holl 1987). No at- tempt was made to further identify the chem- ical basis of the colors. Apart from recording the web building be- havior with a video camera (one frame per second; night recordings with a lens sensitive to red light), the spiders were observed indi- vidually during the day, and at three different development stages during the night (22 ani- mals on day 52 after hatching, 18 animals on day 90 and 16 adult animals on day 125). The observed activities were logged according to type and duration. The orb-webs were ana- lyzed for differences between juvenile and adult pattern, especially orb web symmetry and position of the spider in hub or retreat. Voucher specimens are deposited in the Nat- ural History Museum, Basle, Switzerland. RESULTS Coloration and pattern.— The opisthoso- ma of freshly emerged juvenile spiders is am- ber, with three large and one small pair of black, slightly curved abdominal stripes on the dorsal side (Fig. 1). The diffuse deposition of guanine in the dorsal integument that be- gins frontally at approximately day 10 after hatching continues until the guanine is con- densed to a tight layer between days 20-40 and causes a general lightening of the color. Parallel to this the opisthosoma turns from amber to bright yellow between days 15-60. The pattern expands with a fifth pair of ab- dominal stripes appearing with the second molting, two to three black median dots (small elevations in the cuticle filled with black pig- ments) and a posterior line. Simultaneously with the deposition of guanine, dark brown pigments begin to accumulate frontally on the opisthosoma, solidifying in the course of the next color change into a black band encircling the opisthosoma latero-dorsally (Fig. 2). In most cases a shift from yellow towards green occurs at this time, which can be more- or-less pronounced: a) the color change either begins between the rear abdominal stripes and expands forwards, leaving the entire width of the frontal opisthosoma yellow and giving the entire opisthosoma a greenish-yellow cast, or b) the whole dorsal opisthosoma changes col- or, giving it a more yellowish-green cast be- fore darkening to green (Fig. 7). The first var- iation can be found between days 17-67, the second between days 36-59. The next color change initiates a darkening of the coloration. A longitudinally oriented V appears, which is first light brown in color, but darkens increasingly. As a result the dorsum becomes divided into a rostral and a caudal area. Based on the preceding color variation, different patterns ensue (Fig. 3): a) a tri-col- ored variation with a yellow area in front of, and a green area behind, the brown V, or b) a bi-colored variation with the brown V in the middle of the uniformly green opisthosoma. Male E. fuliginea show these variations be- tween days 50-63, females between days SO- SO (Fig. 7). 106 THE JOURNAL OF ARACHNOLOGY Figures 1-6. — Color pattern of the dorsal opisthosoma of Eriophora fuliginea. 1. Instar I II; 2. Instar II- V; 3. Instar V-VII; 4. Instar VII (male), instar VII- VIII (female); 5. male, instar VIII-IX (adult); 6. Female, instar VIII-X (adult). Since color changes continuously irrespective of molting, the indication of instars is only approximative. Scale: Opisthosoma length of instar I is about 1 mm, adult females 15-25 mm. The colored version of these figures can be found at http://www.cx.unibe.ch/zos/syn.htm (go to publications, 2001). As the V gets darker the abdominal stripes lose definition and blend into the pattern, changing color from black to a dark reddish- brown. Between the two legs of the V an off- white median stripe begins to form. In all col- or variations the areas in front of and behind the V change color in such a way that they appear olive-colored with either a yellowish, greenish or brownish cast. The anterior area tends to show a pattern of yellow and brown- ish-olive flecks, while the posterior area be- gins to display stripes of olive and brown. These variations appear in females between days 69-99 and in males between days 60- 80 (Figs. 4, 7), Between days 80-120 (females) and days 72-195 (males) the color shifts toward brown. In the males, the dark pattern of the dark brown V expands after approximately day 101, particularly towards the front, decreasing the lighter area (lighter spots on medium dark brown). The contrast in the stripes becomes more pronounced with light (light brown or light greyish-brown) and dark (dark brown or brownish-black) stripes alternating. Addition- ally, short light hairs grow on the lighter stripes and short dark hairs grow on the darker stripes (Fig. 5). In females, the V expands forwards as well, but less distinctly than in the males because of its lighter color. The anterior area stays lighter longer, sometimes flecked or mottled yellow or whitish on brown. The stripes in the posterior part shift towards brown and become less distinct. Most females develop a uniform dark brown to brownish-black coloration after this until the opisthosoma is only marked with a whitish median stripe (Fig. 6). In both females and males the abdominal stripes turn a more reddish-brown during the GRAF & NENTWIG— ONTOGENETIC COLORATION CHANGE 107 Figure 7. — Sequence of color changes in the dorsal opisthosoma of Eriophora fuUginea. Numbers refer to Figs. 1-6. last two or three color changes, becoming thinner and less distinct, until they are only recognizable as hairless borders in adult fe- males. The caudal median black spots are found in all animals except in adult females where they dissolve. In addition to these pat- terns many animals show variations in color, spots and stripes which we consider to be in- dividual modifications. Color origin. — The various colors of E. fu- liginea have different origins. The amber col- or, particularly of the juvenile spiders, is based on the transparent coloration of the cuticle. With increasing thickness of the cuticle the amber becomes darker (dark amber to red- dish-brown), e.g., in the prosoma of older an- imals, where the cuticle is very solid. Thinner cuticle can be grey to greyish-brown, e.g., in the dorsal opisthosoma of older animals. It was not possible to judge how far this is a true coloration of the cuticle. The yellow color originates either in the cu- ticle or in the hypodermis. It was impossible to locate accurately as the yellow color dis- solves quickly in 70% alcohol. All that can be seen in the light microscope in individuals with a yellow opisthosoma is a hypodermis poor in dark pigments. The green color, the reddish-brown color of the abdominal stripes as well as the brown and dark brown colors of the opisthosoma-encircling band, the V, the stripes of the pattern and the adult coloration of the females are based in the hypodermis. The guanine stored in guanocytes provides a white background that causes the colors to be opaque, rather than transparent, and bright 108 THE JOURNAL OF ARACHNOLOGY as in the case of the yellow, yellowish-green and green-colored animals. The guanine is not distributed evenly over the opisthosoma. The lateral and dorsolateral guanocytes differ in their guanine content, creating the striped pat- tern. The opisthosoma’s median stripe is caused by the dorsomedian guanocytes, which contain large deposits of guanine, and by the low density of dark pigments in the hypoder- mis of this area. Web building behavior. — The spectrum of activities observed in E. fuliginea includes walking, construction and destruction of orb- webs, perching in the hub, construction of a hiding place and hiding, catching and eating prey, as well as molting. Although all devel- opment stages of E. fuliginea show these ac- tivities, differences can be found in manner, frequency and timing. Walking constitutes the major part of the moving activities of all age groups. Juvenile spiders were observed walk- ing in short bouts of up to 5 min during the day until day 60, juvenile and adult spiders in bouts from few min to over 1 h during the night. Juvenile E. fuliginea begin orb-web build- ing on day 9 or 10. Up until day 47 they build webs only every second or third day, mostly during the day. Webs of juveniles are small and very variable in form and spatial arrange- ment, either as horizontal orbs or as orbs which are not plane, thus resembling three- dimensional webs. Older animals shift their web building activities to the night and build bigger, mostly two-dimensional, and more and more asymmetric webs, with the hub being placed near the upper end of the web and the lower capture area being enlarged. Destruction of the webs was observed towards the end of the night. Eriophora fuliginea spends 30-70% of its time sitting in the hub of its web. With increasing age, however, fewer individuals are observed sitting on the web during the day: Up to day 60 the yellow and green spiders sit 30-40% of the time in the hub; after day 90 the dark brown spiders spend less then 10% of their time in the hub. DISCUSSION Types of color change. — In E. fuliginea, the changes in color occur gradually and in- dependently of molting. They are irreversible color changes, not temporary, reversible ad- aptations to varying backgrounds as can be the case in species that undergo rapid color changes after disturbances. Examples of re- versible color change are Phonognatha graef- fei (Keyserling 1865) (see Roberts 1936), Cyr- tophora cicatrosa (Stoliczka 1869) (see Blanke 1975) and Argiope flavipalpis (Lucas 1858) (see Edmunds & Edmunds 1986). These species react to disturbances by drop- ping to the ground and changing from a dark color with a distinctly striped white pattern to a darker, indistinctly patterned color within fractions of a second, thus blending into the background. This rapid color change is usually based on the contraction of guanocytes (Blanke 1975; Holl 1987) which diminishes the dimension of the white areas. The original color is usually re-established in less than 1 h. Contrary to this rapid and reversible so-called physiological color change (Holl 1987) examples are known of a different reversible color change that takes place slowly and results in different col- or varieties. This morphological color change (Aechter 1955) is found in animals that adapt their color to their environment once or sev- eral times in their lives. One of the best known examples of this slow color change are the crab spiders. The females of Misumena vatia (Clerck 1757) can slowly but reversibly adapt their body color to a white or yellow background. Individuals of the araneid Cyr- tophora citricola (Forskal 1775) can adapt their color to a new environment over the course of 2-4 wk. In our study, the juveniles of E. fuliginea were observed to be much more frequent in the green color variation in the field than in the laboratory. The varying degrees of the shift from yellow to green at this stage found in the juvenile spiders raised in the laboratory might represent adaptations to the background which are not reversible per se, but lose most of their distinctiveness in the final adult color. Experiments with individuals being raised on different backgrounds might provide insights into this phenomenon. Holl (1987) lists a third type of color change apart from the physiological and the morphological kind, a so-called ontogenetic color change. The resulting coloration is ir- reversible. According to Holl, it appears to be associated with metamorphosis or molting in arthropods. Since the color change in E. fuli- ginea progresses continuously and indepen- GRAF & NENTWIG— ONTOGENETIC COLORATION CHANGE 109 dently from the molts, it does not conform totally to HolTs definition. Since ontogenetic implies, however, that the color change occurs parallel to the ontogenesis of an animal we feel that the color change of E. fuliginea can be classified as such. Different explanations can be considered for the function or purpose of the various col- orations and patterns of E. fuliginea. It is rath- er unlikely that any of the colorations play a role in the thermoregulation of E. fuliginea although that function has been reported for Argiope argentata (Fabricius 1775) (see Rob- inson & Robinson 1978). The opisthosoma of A. argentata, a spider that is exposed to the sun as it sits in its orb-web during the day, is colored silvery-white in large areas, reflecting the sunlight and thus reducing the thermal ef- fect of this exposure. The yellow forms of E. fuliginea, being the lightest color variation of the different forms and therefore possibly playing a similar role as the silvery- white col- oration of A. argentata, are hardly ever ex- posed to the sun in their natural habitat. More often, the various colorations and patterns confer a different kind of protection on their bearers. Many species signal to others their dangerousness and/or inedibility by their apo- sematic colors. Function of color change. — Eriophora fu- liginea, however, belongs to the group of an- imals that render themselves invisible to their prey as well as to their enemies through their coloration and behavior. The most widely spread strategy for chromatic camouflage is an adaptation to the background. Juvenile E. fu- liginea sport colors that are found frequently in the vegetation: amber (light orange-brown), yellow, green. A successful camouflage in front of a like-colored background is easily imaginable. Examples are the diverse color variations of the crab spiders with white, yel- low, green and even red coloring, depending on their host plant. In Panama, E. fuliginea was often observed on the hub of its horizon- tal orb-web constructed on the upper side of leaves during this “plant-colored” phase. The yellow color of the juvenile spiders might have the additional effect of rendering them invisible to UV-sensitive (and thus red-blind) prey, as is the case in Misumena vatia (see Hinton 1976). It is likely that the more con- trastingly colored forms of E. fuliginea do not expose themselves on leaves, at least not in the bright daylight. The results of the labora- tory observations support this assumption, as a shifting of the activities into the night was observed parallel to this color change. The spotted, mottled and striped patterns of E. fu- liginea are similar to those of other spiders. Many araneids have developed such outline- dissolving patterns (Robinson & Robinson 1978). This is especially important for species that do not sit directly in front of their back- grounds, as a purely chromatic adaptation would not suffice for camouflage in this case (Robinson & Robinson 1970). According to observations in the field, adult E. fuliginea construct their large vertical orb-webs more and more between bushes with advancing age and therefore are further removed from the background than the juvenile spiders that con- struct their webs between or on leaves. As the potential predators of E. fuliginea, like noc- turnal lizards, small primates and insectivores, possess much better night vision than humans, this reasoning is still valid in the face of the shift of the activities into the night. Concluding, the main function of the changing coloration of E. fuliginea is likely to be to camouflage the spider adequately during the two different phases of its life. In the ju- venile part, spiders are primarily yellow to green, diurnal, and build small symmetric orb webs in large leaves. In the second life part, spiders change to dark brown, are more noc- turnal, and build large asymmetric orb webs in vegetation gaps. ACKNOWLEDGMENTS We thank the Smithsonian Tropical Re- search Institute in Panama and the Panaman- ian authorities (RENARE) for their kind help, Barbara Keller, Gesa Thies and Maria Schi- wek for technical assistance, Sandra Zingg for editing this manuscript and two referees for valuable comments on an earlier version. LITERATURE CITED Aechter, R. 1955. Untersuchungen iiber die Zeichnung und Farbung der Araneen unter Beriick-sichtigung der Ontogenie und Phylo- genie. Sitzungberichte Osterreichische Akademie der Wissenschaften Mathematisch-Naturwissen- schaften und Erdwissenschaften I 164:545-606. Begon, M., J.L. Harper & C.R. Townsend. 1996. Ecology. Blackwall, Oxford. Blanke, R. 1975. Die Bedeutung der Guanozyten fiir den physiologischen Farbwechsel bei Cyrto- no THE JOURNAL OF ARACHNOLOGY phora cicatrosa (Araneae: Araneidae). Entomo- logia Germanica 2:1-6. Bonnet, R 1929. Les araignees exotiques en Eu- rope; II. Elevage a Toulouse de la grande araig- nee fileuse de Madagascar et consideration sur I’Araneiculture (premiere partie). Bulletin de la Societe Zoologique de France 54:501-523. Bonnet, P. 1930. Les araignees exotiques en Eu- rope; II. Elevage a Toulouse de la grande araig- nee fileuse de Madagascar et consideration sur I’Araneiculture (seconde partie). Bulletin de la Societe Zoologique de France 55:53—77. Edmunds, J. & M. Edmunds. 1986. The defense mechanism of orb weavers (Araneae: Araneidae) in Ghana, West Africa. Pp. 73-89. In Proceed- ings of the Ninth International Congress of Ar- achnology. (W.G, Eberhard, Y.D. Lubin & B.C. Robinson, ed.). Panama 1983, Smithsonian In- stitution. Glauert, A.M. 1974. Fixation, dehydration and em- bedding of biological specimens. Pp. 1-207. In Practical Methods In Electron Microscopy 3 (A.M. Glauert, ed.). North Holland/American El- sevier, Amsterdam. Hinton, G.E. 1976. Possible significance of the red patches of the female crab-spider Misumena va- tia. Journal of Zoology (London) 180:35-39. Holl, A. 1987. Coloration and chromes. Pp. 16-25. In Ecophysiology of Spiders. (W Nentwig, ed.) Springer, Berlin. Homann, H. 1946. Uber die Jugendform von Mi- crommata viridissima (Deg.) (Araneae). Biolo- gisches Zentralblatt 65:82-83. Nentwig, W. 1985. Prey analysis of four species of tropical orb- weaving spiders (Araneae: Aranei- dae) and a comparison with araneids of the tem- perate zone. Oecologia 66:580-594. Roberts, N.L. 1936. Color change in the leaf-curl- ing spider {Araneus wagneri). Proceedings of the Royal Zoological Society of New south Wales 28-29. Robinson M.H. & B.C. Robinson. 1970. The sta- bilimentum of the orb-web spider Argiope ar- gentata (Araneae: Araneidae): An improbable defense against predators. Canadian Entomolo- gist 102:641-655. Robinson, M.H. & B.C. Robinson. 1978. Ther- moregulation in orb- web spiders: New descrip- tions of thermoregulatory postures and experi- ments on the effects of the posture and coloration. Zoological Journal of the Linnean So- ciety 64:87-102. Manuscript received 7 December 1999, revised 6 October 2000. 2001. The Journal of Arachnology 29:111-113 SHORT COMMUNICATION ZOROPSIDAE: A SPIDER FAMILY NEWLY INTRODUCED TO THE USA (ARANEAE, ENTELEGYNAE, LYCOSOIDEA) Charles E. Griswold: Schlinger Curator of Arachnida, Department of Entomology, California Academy of Sciences, Golden Gate Park, San Francisco, California 94118 USA; and Research Professor of Biology, San Francisco State University, 1600 Holloway Avenue, San Francisco, California 94132 USA Darrell Ubick: Senior Curatorial Assistant, Department of Entomology, California Academy of Sciences, Golden Gate Park, San Francisco, California 94118 USA ABSTRACT. The spider family Zoropsidae is newly recorded for the USA, bringing the total to 68 families, Zoropsis spinimana (Dufour 1820), native to the Mediterranean region, has been established in the San Francisco Bay area since at least 1995. The identification and phylogenetic position of this species are provided. Keywords: Spiders, exotic, introduction, Zoropsidae, Lycosoidea Zoropsis spinimana (Dufour 1820) has been en- countered several times over the past five years in the San Francisco Bay region of California. This remarkable species was first brought to the attention of Marge Moody of the California State Department of Agriculture (CDFA) who sent specimens to us at the California Academy of Sciences (CAS) for identification. An additional specimen was sent to us by Rick Vetter of the University of California, Riverside (UCR). Both males and females have been encountered, and specimens have been taken both in houses and in nurseries. Records are from the winter and spring (December through May) and early fall (September and October). The presence of this species in at least five cities in two counties suggests that Zoropsis spinimana (Dufour) has been introduced to and established in the San Francisco Bay area of central California. According to one informant, the spiders were found “high on interior walls or ceilings.” A captive female made no web for prey capture. When that female produced an egg sac in April she surrounded the sac with a wall of cribellate silk; cribellate silk was carded with mo- bile leg IV (Eberhard & Pereira 1993). Captive spi- ders were not aggressive and may not be considered dangerous, though there are possible cases of ev- enomization in France attributed to this species (Emerit & Bonaric 1995). Zoropsidae contains two genera: Takeoa Lehti- nen 1967 and Zoropsis Simon 1878 (Platnick 1993). The family was previously known from the Palearctic region with records from the Canary Is- lands on the west, and east through the circum- Mediterranean region and the Balkans to China, Ja- pan and Korea (Roewer 1954:1284; Brignoli 1983: 591; Platnick 1989:504, 1993:587; 1997:611). Zo- ropsis spinimana has been reported from the cir- cum-Mediterranean (Wunderlich 1994). Zoropsis has been represented in two recent phy- logenetic studies. In a study of the Lycosoidea and their kin, Griswold (1993) confirmed the monophy- ly of a Zoropsidae including Zoropsis and Takeoa. In that study, zoropsid synapomorphies were a membranous process on the papal tegulum (in ad- dition to the conductor and median apophysis) (Fig. 2) and shallowly notched trochanters. Griswold et al. (1999) found that Zoropsis, Acanthoctenus and Psechrus exemplify a monophyletic Lycosoidea (in part) defined by the synapomorphic presence of claw tufts, grate-shaped tapetum in the indirect eyes (with homoplasy in Stiphidiidae), and the presence of a minor ampullate gland spigot nubbin on the posterior median spinnerets. Zoropsis spinimana is easily recognized. It keys to Tengellidae in Roth’s (1993) Spider Genera of North America', p. 37: couplet 5, Section III (eight- eyed spiders). Group I (cribellates). It differs from spiders placed in Roth’s ‘Tengellidae’ (in his paper represented only by Zorocrates, which is now placed in Zorocratidae) by having the body pat- terned (Fig. 1) rather than unicolorous, having the posterior eye row strongly recurved (Fig. 1) rather than straight to weakly procurved, having 6-7 pairs of spines beneath tibia I (Fig. 1) rather than 4-5 111 112 THE JOURNAL OF ARACHNOLOGY Figure L — Dorsal view of female Zoropsis spi- nimana (Dufour) from Sunnyvale, California. pairs, and lacking the inferior tarsal claw from all tarsi (Griswold 1993, fig. 6) rather than retaining the inferior claw on tarsus I. Like other cribellate members of the Lycosoidea and their kin, Zoropsis has an oval calamistrum on metatarsus IV of fe- males and juveniles (Fig. 1; Griswold 1993, fig. 1). The male palpus has a short, blunt embolus, hyaline conductor, hooked median apophysis and an addi- tional, membranous process that cradles the embo- lus (Figs. 2, 5). The epigynum (Fig. 3) is unusual in having a central, digitiform scape, which is hol- low (Fig. 4) and reminiscent of some araneids and some species of the South African lycosoid genus Phanotea Simon 1896 (Griswold 1994, figs. 79, 80). This discovery adds another spider family to the list for the USA. Roth (1993) lists only 59 spider families occurring in the USA, but he maintains broad limits for several families now subdivided (Platnick 1993) including Agelenidae (which also includes Cybaeidae), Clubionidae (also including Corinnidae, Liocranidae, and Miturgidae), Linyphi- idae (also includes Pimoidae), and Amaurobiidae (also includes Titanoecidae). Additionally, Metal- tella is now placed in the Amphinectidae (Davies 1998; Griswold et al. 1999), Zorocrates in the Zo- rocratidae (Griswold et al. 1999), and Liocranoides and relatives in the Tengellidae (Platnick 1999). The addition of Zoropsidae increases the known spider fauna of the USA to 68 families. Marge Moody of the California State Department of Agriculture (CDFA) sent the first specimens to 1.0 mm Figure 2-5. — Genitalia of Zoropsis spinimana (Dufour) from Sunnyvale, California. 2, Left male pe- dipalpus, ventral; 3, Epigynum, ventral; 4, Vulva, dorsal; 5, Left male pedipalpus, retrolateral. GRISWOLD & UBICK— ZOROPSIDAE IN USA 113 us for determination. The habitus illustration and those of the male pedipalpus are by Michelle Schwengel. Her work was supported by the Cali- fornia Academy of Sciences through the Fellows’ Artist Intern Program. The illustrations of the fe- male epigynum and vulva are by Jenny Speckels, who was supported by the Exline-Frizzell Fund of the CAS Department of Entomology. MATERIAL EXAMINED. CALIFORNIA: Alameda County: Oakland, SE side of Lake Merrit, in house, 24 September 1997, K. Lundstrom, 19, CAS. Santa Clara County: Cupertino, inside house, 21 October 1996, M. Beauregard, IS, CAS (CDFA #1174531). Sunnyvale, inside house, January- March 1999, V. Romano, Id 19, CAS, (CDFA #993770). In house, 2 January 1998, J. Ward, Id, UCR; 3 February 1998, M. Murray, 1 penultimate d, CDFA (#1019167). Found in house, 12 February 1996, M. Beauregard, Id, CAS (CDFA #1019036). 27 April 1998, M. Nachand, 19, CDFA (#1174929). In house, 11 October 1995, M. Beau- regard, 19, CAS (CDFA #1019033); 12 October 1995, M. Murray, 19, CAS (CDFA #1141353). In house, 22 October 1992, M. Beauregard, 19, CDFA (#750348). Santa Clara, 31 December 1997, M. Murray, 19, CDFA (#1019023). San Jose, in nursery, 1 April 1999, M. Murray, 19, CDFA (#1162149). LITERATURE CITED Brignoli, P.M. 1983. A catalogue of the Araneae described between 1 940-1 981. Manchester, 755 pp. Davies, V.T. 1998. A revision of the Australian me- tal tellines (Araneae: Amaurobioidea: Amphinec- tidae: Metaltellinae). Invertebrate Taxonomy 12: 211-243. Eberhard, WG. & F. Pereira. 1993. Ultrastructure of cribellate silk of nine species in eight families and possible taxonomic implications (Araneae: Amaurobiidae, Deinopidae, Desidae, Dictynidae, Filistatidae, Hypochilidae, Stiphidiidae, Tengel- lidae). Journal of Arachnology 21:161-174. Emerit, M. & J.C. Bonaric. 1995. Un cas d’envenimation de type loxoscelique attribue a Faraignee Zoropsis spinimana dans le midi de la France. Ann. soc. Hort. Hist. Nat. L’Herault 135: 37-38. (spell out titles) Griswold, C.E, 1993. Investigations into the phy- logeny of the lycosoid spiders and their kin (Arachnida, Araneae, Lycosoidea). Smithsonian Contributions to Zoology 539:1-39. Griswold, C.E. 1994. A revision and phylogenetic analysis of the African spider genus Phanotea Simon (Araneae: Lycosoidea). Ann., Sci. Zook, Mus. Roy. L’Afr. Centr., vol. 273:1-83. Griswold, C., J. Coddington, N. Platnick & R. For- ster. 1999. Towards a phylogeny of entelegyne spiders (Araneae, Araneomorphae, Entelegynae). Journal of Arachnology 27:53-63. Platnick, N.I. 1989. Advances in Spider Taxono- my: A Supplement to Brignoli ’s A Catalogue of the Araneae Described Between 1940 and 1981. Manchester, 673 pp. Platnick, N.I. 1993. Advances in Spider Taxono- my, 1988-1991: with Synonymies and Transfers 1940-1980. New York Entomological Society, 846 pp. Platnick, N.I. 1997. Advances in Spider Taxono- my, 1992-1995: With Redescriptions 1940- 1980. New York Entomological Society, 846 pp. Platnick, N.I. 1999. A Revision of the Appalachian spider genus Liocranoides (Araneae: Tengelli- dae). American Museum Novitates 3285:1-13. Roewer, C.E 1954. Katalog der Araneae von 1758 bis 1940, Vol. 2a: 1-923. Bruxelles. Roth, V.D. 1993. pider Genera of North America with Keys to Families and Genera and a Guide to Literature. American Arachnological Society, available from Dr. Jon Reiskind, Department of Zoology, University of Florida, Gainesville, Florida 32611 USA. Wunderlich, J. 1994. Zur Kenntnis der West-Pa- laarktischen Arten der Gattung Zoropsis Simon 1878 (Arachnida: Araneae: Zoropsidae). Beitrae- ge Aran., 4:723-727. Manuscript received 20 November 1999, revised 1 July 2000. 2001. The Journal of Arachnology 29:114-116 SHORT COMMUNICATION DISPERSAL OF STEGODYPHUS DUMICOLA (ARANEAE, ERESIDAE): THEY DO BALLOON AFTER ALL! Jutta M. Schneider and Jorg Roos: Dept, of Population Biology, Zoological Institute, University of Mainz, 55099, Germany Yael Lubin: Mitrani Department of Desert Ecology, Jakob Blaustein Institute for Desert Research, Ben Gurion University of the Negev, Sede Boqer Campus, 84990, Israel Johannes R. Henschel: Desert Research Foundation of Namibia, RO. Box 20232, Windhoek, Namibia ABSTRACT. There has been some controversy about whether adult females of social Stegodyphus disperse by ballooning. Here we show that adult Stegodyphus dumicola (Eresidae) Pocock 1898 are able to gain up-lift by releasing a very large number of threads. The threads fan out widely from the spider’s body and form a triangular sheet. This previously unknown ballooning mechanism, enables even large spiders to disperse over large distances. Keywords: Eresidae, Stegodyphus, dispersal, ballooning, social spider Dispersal by ballooning appears to be re- stricted to very small spiders and is mainly a strategy of juvenile spiders that disperse short- ly after their emergence from the eggsacs (De- cae 1987; Foelix 1996). The probability of ballooning as a function of spider size quickly approaches zero when the body mass exceeds 1 mg. Suter (1999) stated that large spiders are unlikely to balloon because thermal and climatic conditions are rarely favorable. In ad- dition, unpredictable patch quality and low survival probability of ballooning spiders should make this strategy unattractive for adult spiders. Wickler & Seibt (1986), however, observed a single adult Stegodyphus mimosarum (Er- esidae) Pavesi 1883 ballooning; and Crouch et al. (1998) reported that during a mass dis- persal event, adults of S. mimosarum became airborne and were carried for several meters by strong, gusting winds. In the latter case, it is not clear if the spiders ballooned, i.e., if they lifted off the surface by means of silk, or if they were blown horizontally and used the silk to anchor themselves. Stegodyphus mi- mosarum is one of three social species of the genus and is distributed throughout southern Africa. Social Stegodyphus Simon 1873 are characterized by dispersal of adult females that usually have a body mass larger than 100 mg. Wickler & Seibt (1986) described the sin- gle ballooning individual as flying with 3-4 silk strands that were no longer than 3-4 m in a barely perceptible breeze. Henschel et al. (1995) used these figures in Suter’s (1991) formula and concluded that with the given length of silk and the described wind velocity, an adult spider of that size (80-150 mg) could not become airborne. The apparent contradic- tion can have two possible causes: either the observation was misinterpreted or the param- eters used in the formula did not exactly de- scribe the observed situation. Recent obser- vations now enable us to clarify the issue. Between 25-31 January 2000 we checked 31 colonies of Stegodyphus dumicola Pocock 1898 on a daily basis. (Voucher specimen are deposited at the National Museum in Win- dhoek, Namibia.) The nests were evenly dis- tributed in an area of approximately 70,000 m^ (7 ha). The study site was on the farm Om- draai, located 100 km southeast of Windhoek, 114 SCHNEIDER ET AL„— BALLOONING IN SOCIAL STEGODYPHUS 115 Namibia. The days were hot (28-33 °C), and there was almost no wind. On such days, ris- ing thermals occur characteristically during the warm, calm hours of the day. On 27 Jan- uary, around noon, 20 females of one colony were seen “tiptoeing” on the highest strand of the web. Tiptoeing behavior occurs as a prelude to ballooning: the spider stands on raised legs with the abdomen pointed upwards at an angle to the prosoma. In this position, silk released from the spinnerets will rise even in an almost imperceptible breeze (Foelix 1996). We observed these females releasing silk, and some became airborne. However, silk became snagged on nearby bushes so that the spiders landed between 1-8 m away from their nest of origin. During the late afternoon of the following day the same behavior was observed in three other nests. On one occa- sion, more than six spiders in quick succes- sion were lifted almost straight up and could be observed gaining height for a few sec. We lost sight of the spiders after they reached a height of approximately 30 m. These spiders were the size of adult females, between 7-14 mm body length (see Kraus & Kraus 1988). Perhaps the most important aspect of our observation is that the spiders used a very large number of silk strands to become air- borD.e. At least tens to hundreds of threads were seen silhouetted against the sky. The threads fanned out widely from the spider’s body and formed a triangular sheet with a length and width of about 1 m at its distal end. The silk that was released was not combed, appeared to be produced quickly, and did not tangle or clump once released. In the 31 nests observed, females in 10 col- onies were observed tiptoeing as spiders pre- pared to release silk threads. In two colonies we saw bridging, a common method of dis- persal in Stegodyphus (Heeschel et al. 1995); and in one colony we saw ballooning. We measured the sizes of all nests by using the two largest diagonals. Tee nests that produced dispersers were significantly larger (n = 10; mean ± SE = 155.2 ± 20.23 mm^) than nests without dispersers (n = 21; mean ± SE = 60.5 ± 9.26 mm^; Kruskal- Wallis test: Z = 3.61, P < 0.0003). As nest size is related to colony size (Henschel 1998), this indicates that only larger colonies produced dispersing females. In order to assess the reproductive status of dispersing individuals, nine females were col- lected while tiptoeing, which is indicative of imminent dispersal, and were kept in the lab- oratory without access to males. Seven of these females produced eggsacs within 6 wk after collection. This indicates that most, if not all dispersers are potentially capable of found- ing new colonies after establishing a new nest at their destination. We did not observe males dispersing. In- terestingly, 10 out of 17 collected colonies showed an unusual sex ratio of 25-52% males. Stegodyphus colonies usually have fe- male-biased sex ratios (14-21.2% males) and the sex-ratio bias is primary (Aviles et al. 1999). Thus, the most likely explanation for the reduction in bias is that the majority of the females had left the colony whereas the males did not disperse. Ballooning on multiple strands is a dispers- al mechanism in spiders that has never been described before. By using multiple strands even large adult spiders may be able to dis- perse over great distances and to colonize new habitats. This has to be taken into consider- ation in the future when investigating popu- lation genetics and structure in the social Ste- godyphus. There are a number of questions that remain open. Which glands produce the ballooning silk? The large number of threads would sug- gest that cribellate silk might be used; but if so, it is apparently not combed out. What keeps the multiple threads from collapsing or coalescing? Perhaps electrostatic forces play a role in keeping the threads apart. Is the lift generated by such a large number of threads equal to the sum of the forces acting on each one, or do the threads indeed form a sheet dense enough to be considered as a whole? Finally, is ballooning the regular dispersal mode used by social Stegodyphusl Colonies appear to send out dispersers only in one par- ticular state of development, namely after the majority of females have matured and mated and before egg-laying begins (Henschel et al. 1995; Henschel 1998). Thus, for dispersal by ballooning, weather conditions must be suit- able during a rather narrow time window. Long-term monitoring of dispersal and cli- mate will be required to answer this question. This is publication #309 of the Mitrani De- partment of Desert Ecology. 116 THE JOURNAL OF ARACHNOLOGY LITERATURE CITED Aviles, L., C. Varas & E. Dyreson. 1999. Does the African social spider Stegodyphus dumicola con- trol the sex of individual offspring? Behavioural Ecology and Sociobiology 40:237-245. Decae, A.E. 1987. Dispersal: Ballooning and other mechanisms. Pp. 348-358. In Ecophysiology of Spiders. (W. Nentwig, ed.). Springer Verlag, Hei- delberg. Foelix, R.E 1996. Biology of Spiders, 2nd ed. Ox- ford Univ. Press and Georg Thieme Verlag, New York. Henschel, J.R. 1998. Predation on social and soli- tary individuals of the spider Stegodyphus dum- icola (Araneae, Eresidae). Journal of Arachnol- ogy 26:61-69. Henschel, J.R., J. Schneider & Y. Lubin. 1995. Dispersal mechanisms by the spiders Stegody- phus’. Do they balloon? Journal of Arachnology 23:202-204. Humphrey, J.A.C. 1987. Fluid mechanical con- straints on spider ballooning. Oecologia 73:469- 477. Kraus O. & M. Kraus. 1988. The genus Stegody- phus (Arachnida, Araneae). Sibling species. Spe- cies groups, and parallel origin of social living. Verhandlungen des naturwissenschaftlichen Ver- eins Hamburg 30:151-254. Suter, R.B. 1991. Ballooning in spiders: Results of wind tunnel experiments. Ethology, Ecology & Evolution 3:13-25. Suter, R.B. 1999. An aerial lottery: The physics of ballooning in a chaotic athmosphere. Journal of Arachnology 27:281-293. Wickler, W. & U. Seibt. 1986. Aerial dispersal by ballooning in adult Stegodyphus mimosarum. Na- turwissenschaften 73:628-629. Manuscript received 20 July 2000, revised 15 Oc- tober 2000. 200L The Journal of Arachnology 29:117-118 SHORT COMMUNICATION A TECHNIQUE FOR INDIVIDUALLY IDENTIFYING TARANTULAS USING PASSIVE INTEGRATED TRANSPONDERS Steven B. ReicMing and Chris Tabaka: Memphis Zoo, 2000 Galloway, Memphis, Tennessee 38112 USA ABSTRACT. A surgical technique for implanting passive integrated transponders into theraphosid spi- ders is described. An effecitve procedure for anesthesia was developed. Transponders were implanted in the opisthosomas of 12 spiders. No mortality occurred, and all spiders regained normal behavior. In simulated burrows, tarantulas could be identified to a depth of 16 cm. Keywords: PIT tags, spider marker, Theraphosidae No complete life history study of a theraphosid spider has appeared since the pioneering work of Baerg (1958). A necessary component of such en- deavors is the application of a marker that enables the researcher to permanently distinguish individual spiders. Marking theraphosids is particularly diffi- cult because they molt regularly (Baerg & Peck 1970) throughout their long life of 20 years or more in some females (Marshall 1996). A marker should be internal and identifiable for many years to be useful in long-term life history studies of tarantulas. Widespread use of passive integrated transpon- ders, which are commonly known as PIT tags, in vertebrate studies suggests that an application might be found for tarantulas. These devices are small and can be read by a hand-held reader emitting low- frequency radio waves. The transponder signal is received, decoded, and displayed by the reader as a unique 10-character code. The transponders are her- metically sealed in biocompatible glass and appear to have an unlimited life span. Although widely used by zoo personnel, vertebrate field biologists and veterinariaes (Elbie & Burger 1994), this is the first time- — to our knowledge — that PIT tags have been used in an invertebrate. For large arachnids this technology provides the perfect marker, being permanent, unrecognizable and untransferable to other spiders, benign in its effect on survival, and easy to apply especially under field conditions (Evans & Gleeson 1998). The technique was tested on adults of Aphono- pelma baergi (n = 4; body length 38-47 mm), Bra- chypelma albopilosum (n = 4; body length 38-75 mm), and Grammostola puichra (n = 4; body length 62-68 mm). Aphonopelma were collected as adults, 10 km north of Jessieville, Garland County, Arkansas. Brachypelma and Grammostola were ob- tained as captive-bred juveniles from commercial suppliers in the United States and reared to adult- hood. Transponder implantations were performed in the veterinary hospital at the Memphis Zoo and field trials were conducted on zoo grounds. We used the Trovan® (Grossbuilesheimer, Str. 56, Euskirch- en 16, Germany) reader (Model LID 500) and tran- sponders in all trials. The location for implantation of the transponder was on the dorsolateral aspect of the opisthosoma in an area between the heart and the intestinal tract (Fig. 1). Tarantulas were re- strained by hand during the procedure. A 20-gauge hypodermic needle was used to scrape the setae from a 1.5 X 1.5 mm area of the opisthosoma, and swabbed with a 10% povidone-iodine solution. The sterile needle was used to cut the exoskeleton. The sharp apical edge of the needle was used like a scal- pel rather than creating a puncture wound. The tran- sponder was inserted into the opisthoma with sterile mosquito forceps. The surgical site was then swabbed dry and several drops of n-butyl cyano- acrylate adhesive glue (Vetaboed®, 3M Animal Care Products, St. Paul, Minnesota) were used to close the wound. The entire procedure took 2-3 minutes per spider. Leakage of haemolymph varied. In one instance there was a moderate loss of fluid from the site, but the spider recovered fully. Four of the spiders (A. baergi, n = 2; B. albop- ilosum, n = 2) were anesthetized prior to implan- tation. Spiders were immobilized with isoflurane (Iso-thesia®, Abbott Laboratories, North Chicago, Illinois). A cottonball was soaked in the anesthetic agent and placed in a small plastic container away from the spider. The effect of the anesthetic was monitored by leg movement. As the spiders became 17 118 THE JOURNAL OF ARACHNOLOGY Figure 1 . — Radiograph showing passive integrat- ed transponder (arrow) implanted in a Grammostola pulchra. anesthetized the legs contracted followed by relax- ation. Spiders which were not anesthetized during tran- sponder implants accepted food within several hours, suggesting they had not been severely trau- matized by the procedure. Anesthetized spiders re- quired 2-3 hours post-surgical recovery time before normal movement was exhibited. All spiders which had implants completed ecdysis within 3-7 months. After molting, no evidence of the implants was not- ed. All spiders were preserved after two years. Voucher specimens were deposited in the Field Mu- seum of Natural History, Chicago. To assess the limits of the reader in decoding the transponder signal from implanted spiders in situ, we conducted trials using artificial burrows. A nat- ural burrow replica was prepared by boring a 5 cm diameter hole at an 80° angle to a depth of 20 cm. Trials were conducted in hard-packed humus on a rainy day to simulate typical field conditions for many species of theraphosids. Spiders were identi- fiable in the burrows at a depth of up to 16 cm. This distance approaches the 18-20 cm detection limit of the reader across unobstructed space. This technique is best suited for long term field studies of large, long-lived arthropods such as ther- aphosid spiders, scolopendrid centipedes, and scor- pions. The sensitivity-level of the reader precludes identification of theraphosids resting at the bottom of deep burrows. However, tarantulas could easily be identified at night while they are passively for- aging at their burrow entrance, eliminating the time-consuming process of capture and handling. Anesthesia prior to implantation is unnecessary and not convenient under field conditions, but it is tol- erated by the spiders. Untrained personnel may con- sider anesthetizing specimens until they become more adept at inserting the transponders. Short-term movements such as the migration of male tarantulas during the breeding season can be monitored by radio telemetry (Janowski-Bell & Homer 1999). However, limitations in battery life, durability of transmitter adhesion, and the potential for these transmitters to interfere with normal be- havior make radio telemetry unsuitable for studies conducted over a longer time scale. We believe this new application for PIT tags offers a way to study previously inaccessible aspects of theraphosid spi- der biology such as growth, survivorship, and the movements of individuals over their entire lives. LITERATURE CITED Baerg, W.J. 1958. The Tarantula. Univ. Kansas Press. Lawrence, Kansas. 88 pp. Baerg, W.J. & W.B. Peck. 1970. A note on lon- gevity and molt cycle of two tropical theraphos- ids. Bulletin of the British Arachnological Soci- ety 1:107-108. Elbin, S.B. & J. Burger. 1994. Implantable micro- chips for individual identification in wild and captive populations. Wildlife Society Bulletin 22:677-683. Evans, TA. & P.V. Glee son. 1998. A new method of marking spiders. Journal of Arachnology 26: 382-384. Janowski-Bell, M.E. & N.V. Homer. 1999. Move- ment of the male brown tarantula, Aphonopelma hentzi (Araneae, Theraphosidae) using radio te- lemetry. Journal of Arachnology 27:503-512. Marshall, S.D. 1996. Tarantulas and Other Arach- nids. Barron’s Educ. Ser., Inc., Hauppauge, New York. 104 pp. Manuscript received 12 February 2000, revised 30 June 2000. 2001. The Journal of Arachnology 29:119-124 SHORT COMMUNICATION SPIDERS FEEDING ON EARTHWORMS Martin Nyffeler: Zoological Institute, University of Berne, Baltzerstr. 3, CH-3012 Berne, Switzerland Hans Moor: Neue Kantonsschule Aarau, CH-5000 Aarau, Switzerland Rainer F. Foelix: Naturama, Postfach, CH^SOOl Aarau, Switzerland ABSTRACT. A house spider (Tegenaria atrica C.L. Koch 1843, Agelenidae) was observed, filmed and photographed while feeding on an earthworm. An extensive search in the literature revealed that several arachnologists had noted spiders feeding on earthworms, altogether in 1 1 different families. Earth worm= eating spiders belong mostly to larger sized species dwelling near the ground in woodlands and grasslands. Since earthworms have a high protein content, they could be a welcome supplement to the spider’s usual insect diet. Keywords: Spiders, prey, foraging, diet, earthworm Most spiders are polyphagous predators that prey predominantly on insects and to a lesser extent on other spiders (Riechert & Harp 1987; Nentwig 1987; Nyffeler et al. 1994). Spiders feeding on non-arthropod prey have rarely been reported (see Foelix 1996). That earthworms may be included in a spider's diet has not been recognized so far. However, this is exactly what was noted by one of us (H.M.) in September 1999 in Herznach, Switzerland: a Tegenaria atrica C.L. Koch 1843 (Agelen- idae) was observed, filmed and photographed while feeding on an earthworm of 14 cm length (Figs. 1, 2). Bristowe’s book '"The Co- mity of Spiders'" (1941), which includes a long chapter on ‘The Food of Spiders,’ re- vealed nothing on this peculiar type of feed- ing. Likewise, books on the biology of earth- worms make no reference to spiders as enemies (see Edwards & Lofty 1972; Mac- Donald 1983; Lee 1985). Thus, the question arises whether our observation on Tegenaria was an isolated case or whether similar inci- dences have been noticed elsewhere. An extensive literature search was conduct- ed in order to find any information available on spiders feeding on earthworms. The search was based largely on the “Liste des Travaux Arachnologiques” (1968-1999), published by the International Society of Arachnology (for- merly the C.I.D.A., Paris, France). In addition, an international arachnology discussion group was contacted via Internet. Altogether about 30 reports on spiders consuming earthworms were gathered (Table 1). Spiders from 11 dif- ferent families are known to feed on earth- worms. In two instances an unidentified spe- cies of Tegenaria (possibly agrestis (Walckenaer 1802)) was found by Yann Ev- enou (pers. commun.) preying on earthworms in the field, thus confirming our observation on Tegenaria. Furthermore, Gunter Schmidt (pers. commun.) fed Tegenaria ferruginea (Panzer 1804) in captivity with earthworms of 8-10 cm length. One of the earliest published reports on spi- ders consuming earthworms is that of Ger- hardt & Kaestner (1937). Spiders from the mygalomorph genus Atypus Latreille 1804 (Atypidae), which inhabit silk tubes in the ground, were observed pulling earthworms into the tube and eating them. However, Bris- towe (1958) expressed some reservations: “. . . Some early naturalists thought Atypus must emerge at night to hunt prey, whilst oth- ers were convinced that she subsisted on earthworms ...” and further “the idea that Atypus feeds largely on earthworms gains no support from examination or tests.” Neverthe- less, he admitted: “. . . experiment with 119 120 THE JOURNAL OF ARACHNOLOGY Figures 1, 2. — Spider feeding on an earthworm. 1. Tegenaria atrica trying to pull up its victim, an earthworm, onto its sheet web; 2. Dorsal close-up view of Tegenaria feeding on the front end of an earthworm. worms placed on the surface ‘finger’ has shown that they get tom in the encounter, leaving at most a part of their bodies in the spider’s possession which cannot readily be hauled into the tube. Although Atypus may suck the worm’s juices for a time, she does not appear to finish the meal.” Crome (1967) successfully Atypus afftnis Eichwald 1830 in captivity with earthworms. Hadronyche versuta (Rainbow 1914) (Hex- athelidae), a mygalomorph spider from Aus- tralia that dwells in a silk tube burrow in the ground, also includes earthworms in its diet (Bmnet 1998). Still another case of a myga- lomorph spider feeding on earthworm prey was observed by Ricardo Ott (pers. commun.) in the rainforest of the Amazon: a large Ther- aphosa blondi (Latreille 1804) (Theraphosi- dae) was feeding on an earthworm of 30 cm length. Theraphosidae, representing 12 differ- ent species and 8 genera, have been seen prey- ing on earthworms in captivity (Yann Evenou & Jakob Walter pers. commun.). According to Brunet (1998), insects and earthworms form the staple diet of the mygalomorphs. Large earthworms (up to 20 cm) were also fed in captivity to the fishing spider Dolomedes fim- briatus (Clerck 1757) (Pisauridae) (Schmidt 1957). Eeeding on earthworms is probably a rarity among spiders (Wolfgang Nentwig pers. com- mun.). Spiders that spin a catching web in the higher strata of the vegetation, with which they capture small winged insects from the ae- rial plankton, will rarely, if ever, get in contact with earthworms. Although the orb web spider Araneus diadematus Clerck 1757 accepted earthworms in captivity (Nyffeler unpubl. data; Table 1), it is not expected to show this behavior in the field. During hundreds of hours of field observations, spiders feeding on earthworms were seen very rarely (Nyffeler 1982) or not at all (Zimmermann & Spence 1989). Feeding on earthworms seems to occur NYFFELER FT AL.— SPIDERS FEEDING ON EARTHWORMS 121 Table 1. — Spiders feeding on earthworms (published and unpublished observations). Typical adult body length Species Family (?) Typical habitat Araneomorphae : Tegenaria atrica C. L. Koch 1843 Tegenaria sp. Latreiile 1804 Tegenaria ferruginea (Panzer 1804) AmauroMus ferox (Walckenaer 1830) Amaurobius fenestralis (Stroem 1768) Segestria florentina (Rossi 1790) Araneus diadematus Clerck 1757 Xysticus sp. C. L. Koch 1835 Xysticus sp. C L. Koch 1835 Pardosa sp, C. L. Koch 1847 Trochosa terricola Thorell 1856 Dolomedes fimbriatus (Clerck 1757) Ancylometes rufus (Walckenaer 1837) Ctenus amphora Mello-Leitao 1930 Ctenus crulsi Mello-Leitaeo 1930 My galomorphae : Atypus ajfinis Eichwald 1830 Atypus sp. Latreiile 1804 Atypus affinis Eichwald 1830 Atypus affinis Eichwald 1830 Hadronyche sp. L. Koch 1873 Hadronyche versuta (Rainbow 1914) Theraphosa blondi (Latreiile 1804) Theraphosa blondi (Latreiile 1804) Aphonopelma anax (Chamberlin 1940) Aphonopelma pallidum (F.O.P.-Cam- bridge 1897) Brachypelma albopilosum Valerio 1980 Brachypelma smithi (F.O.P.-Cambridge 1897) Brachypelma vagans (Ausserer 1875) Chromatopelma cyaneopubescens (Strand 1907) Grammostola iheringi (Keyserling 1891) Grammostola pulchra Mello-Leitao 1921 Hysterocrates ederi Charpentier 1995 Lasiodora parahybana Mello-Leitao 1917 Poecilotheria regalis Pocock 1899 15 mm Woodland and gardens, under stones 15 mm Grassland, ground 14 mm Woodland, crevices in tree tranks 15 mm Woodland, under stones and logs 8 mm Woodland, under stones and logs 20 mm Under stones and logs 15 mm Woodland, grassland, bushes 7 mm Grassland, low vegetation or ground 7 mm Grassland, low vegetation or ground 6 mm Marshland, low vegetation or ground 14 mm Woodland, grassland, under stones 20 mm Swampy areas, low vegetation 35 mm Tropical rainforest, ground 17 mm Tropical rainforest, ground 15 mm Woodland slopes, ground 15 mm Slopes with low vegetation, ground 15 mm Woodland slopes, ground 15 mm Woodland slopes, ground 30 mm Subtropical rainforest, ground 30 mm Subtropical rainforest, ground 100 mm Tropical rainforest, ground bur- row 100 mm Tropical rainforest, ground bur- row 60 mm Grassland, scrubland, ground burrow 40 mm Subtropical scrubland, ground burrow 70 mm Tropical rainforest, ground bur- row 70 mm Woodland, grassland, ground burrow 60 mm Subtropical forest, ground bur- row 50 mm Subtropical scrubland, ground 100 mm Tropical rainforest, ground bur- row 60 mm Grassland, ground 70 mm Tropical rainforest, ground bur- row 70 mm Rainforest, ground 60 mm Monsoon forest, hollow trees Agelenidae Ageleeidae Agelenidae Amaurobiidae Amaurobiidae Segestriidae Araneidae Thomisidae Thomisidae Lycosidae Lycosidae Pisauridae Pisauridae Ctenidae Atypidae Atypidae Atypidae Atypidae Hexathelidae Hexathelidae Theraphosidae Theraphosidae Theraphosidae Theraphosidae Theraphosidae Theraphosidae Theraphosidae Theraphosidae Theraphosidae Theraphosidae Theraphosidae Theraphosidae Theraphosidae 122 THE JOURNAL OF ARACHNOLOGY Table 1. — Extended. Web type Location of observation Source Sheet (ecribellate) Field This paper Sheet (ecribellate) Sheet (ecribellate) Field Captivity Yann Evenou (unpubl.) Gunter Schmidt (unpubl.) Sheet (cribellate) Sheet (cribellate) Snare (ecribellate) Orb (ecribellate) None Captivity Captivity Field Captivity Field Gunter Schmidt (unpubl.) Gunter Schmidt (unpubl.) Yann Evenou (unpubl.) Martin Nyffeler (unpubl.) Nyffeler (1982) None Field Jakob Walter (unpubl.) None Field Vogel (1971) None Field Yann Evenou (unpubl.) None None Captivity Field Schmidt (1957) Ricardo Ott & Clarissa Azevedo (unpubl.) None Field Hubert Hoefer (unpubl.) Silk tube burrow Silk tube burrow Field Field? Savory (1926) Gerhardt & Kastner (1937) Silk tube burrow Silk tube burrow Silk tube burrow Silk tube burrow None Captivity Captivity Captivity Field Field Bristowe (1958) Crome (1967) David Rowell (unpubl.) Brunet (1998) Ricardo Ott (unpubl.) None Captivity Yann Evenou (unpubl.) None Captivity Yann Evenou (unpubl.) None Captivity Yann Evenou (unpubl.) None Captivity Yann Evenou (unpubl.) None Captivity Yann Evenou (unpubl.) None Captivity Yann Evenou (unpubl.) None Captivity Yann Evenou (unpubl.) None Captivity Jakob Walter (unpubl.) None None Captivity Captivity Yann Evenou (unpubl.) Yann Evenou (unpubl.) None None Captivity Captivity Yann Evenou (unpubl.) Yann Evenou (unpubl.) NYFFELER ET AL.— SPIDERS EEEDING ON EARTHWORMS 123 among spiders that dwell on the ground — un- der stones and logs, in the leaf litter and moss- covered patches, in cracks in the soil, and in earth burrows and silk tube burrows — or on low vegetation near the ground in woodlands and grasslands (i.e., habitats where earth- worms are abundant) (Table 1). Web-building and nonweb-building spiders alike have been observed eating earthworms. They belong predominantly to larger species (> 10 mm body length, see Table 1), though there are exceptions. For instance, Nyffeler (1982) found a crab spider of the genus Xysticus C.L. Koch 1835, about 7 mm in length, sucking an earthworm of approximately 2 cm in length. Xysticus spp., nonweb-building spiders equipped with powerful front legs and sup- posedly potent venom, are able to subdue prey 2-3 times their own size (see Gertsch 1979; Nentwig & Wissel 1986). Among web-build- ing spiders reported feeding on earthworms (Table 1), species that make sheet webs (i.e., Tegenaria and Amaurobius) or use a silk tube (i.e., Atypus and Hadronyche) dominate. Such webs function as effective traps for the cap- ture of crawling prey organisms. Surprisingly, some nocturnal ground- surface dwellers (e.g., Gnaphosidae and Dysderidae) — expected of- ten to encounter earthworms — are missing in the table. In terrestrial ecosystems, most of the net primary production is used by detritivores and decomposers in the soil, resulting in a huge earthworm biomass which serves a variety of predators as food (see MacDonald 1983; Halaj & Cady 2000). Earthworm tissue has a high protein content («60--70%, dry weight) (Mac- Donald 1983; Lee 1985); thus an earthworm should be a welcome meal to a spider. Table 1 includes, among others, species from the families Pisauridae, Hexathelidae and Thera- phosidae, which exhibit opportunistic feeding (broad diets) (e.g., Zimmermann & Spence 1989; Brunet 1998; Yann Evenou pers. com- mun.). It is not surprising that the diets of these nonspecific feeders also include earth- worms. Such species are adapted to a broad range of prey types that optimizes their sur- vival during periods of food shortage. Preda- tion on earthworms may be of ecological sig- nificance for some larger spiders (e.g., mygalomorphs) by supplementing their insect diets (see Brunet 1998). ACKNOWLEDGMENTS We kindly acknowledge Clarissa Azevedo, Yann Evenou, Hubert Hoefer, Ricardo Ott, David Rowell, Gunter Schmidt and Jakob Walter for communicating their observations on earthworm eating spiders. We thank Chris- tian Kropf for the identification of Tegenaria atrica. Furthermore we are grateful to Jim Berry, Yann Evenou, Robert Jackson, Chris- tian Kropf, Wolfgang Nentwig, and an anon- ymous reviewer for helpful comments. LITERATURE CITED Bristowe, W.S. 1941. The Comity of Spiders. VoL II. Ray Society, London. Bristowe, W.S. 1958. The World of Spiders. Col- lins, London. Brunet, B. 1998. Spiderwatch: A Guide to Austra- lian Spiders. New Holland Publishers, Sydney. Crome, W. 1967. Wirbellose Tiere. Rowohlt Tas- chenbuch Verlag, Hamburg. Edwards, C.A. & J.R. Lofty. 1972. Biology of Earthworms. Chapman & Hall, London. Foelix, R.F. 1996. Biology of Spiders (2nd ed.). Oxford Univ. Press and Thieme Verlag, New York. Gerhardt, U. & A. Kaestner. 1937. 8. Ordnung der Arachnida: Araneae = Echte Spinnen = Web- spinnen. In Handbuch der Zoologie (W. Kiiken- thal & T. Krumbach, eds.). de Gruyter, Berlin. Gertsch, WJ. 1979. American Spiders (2nd ed.). Van Nostrand, New York. Halaj, J. & A.B. Cady. 2000. Diet composition and significance of earthworms as food of harvest- men (Arachnida: Opiliones). American Midland Naturalist 143:487-491. Lee, K.E. 1985. Earthworms, Their Ecology And Relationships With Soils And Land Use. Aca- demic Press, London. MacDonald, D.W 1983. Predation on earthworms by terrestrial vertebrates. Pp. 393-414. In Earth- worm Ecology. (J.E. Satchell, ed.). Chapman and Hall, London, New York. Nentwig, W. 1987. The prey of spiders. Pp. 249- 263. In Ecophysiology of Spiders. (W. Nentwig, ed.). Springer- Verlag, Berlin, New York. Nentwig, W. & C. Wissel. 1986. A comparison of prey lengths among spiders. Oecologia 68:595- 600. Nyffeler, M. 1982. Eield Studies on the Ecological Role of the Spiders as Insect Predators in Agroe- cosystems. Ph.D. dissertation. Swiss Eederal In- stitute of Technology, Zurich. Nyffeler, M., W.L. Sterling & D.A. Dean. 1994. How spiders make a living. Environmental En- tomology 23:1357-1367. 124 THE JOURNAL OF ARACHNOLOGY Riechert, S.E. & J.M. Harp. 1987. Nutritional ecol- ogy of spiders. Pp. 645-672, In Nutritional Ecol- ogy of Insects, Mites, and Spiders. (E Slansky & J.G. Rodriguez, eds.). John Wiley, New York. Savory, T.H. 1926. British Spiders (Their Haunts And Habits). Clarendon Press, Oxford. Schmidt, G. 1957. Einige Notizen iiber fimbriatus (CL). Zoologischer Anzeiger 158:83- 97. Vogel, B.R. 1971. Individual interactions of Par- dosa. Armadillo Papers 5:1-12. Zimmermann, M. & J.R. Spence. 1989. Prey use of the fishing spider Dolomedes triton (Pisauri- dae, Araneae): An important predator of the neuston community. Oecologia 80:187-194. Manuscript received 17 March 2000, revised 7 June 2000. 200 L The Journal of Arachnology 29:125-126 REVIEWERS OF MANUSCRIPTS Volume 28— (2000) Joachim Adis, Max-Planck Institute for Limnology, Ploen (Germany) Robert Bennett, BC Ministry of Forests, Saanichton, British Columbia (Canada) John Brackenbury, Cambridge University (UK) Jason Bond, The Field Museum, Chicago, Illinois (USA) Jan Bosselaers, Beerse (Belgium) Christopher Buddie, University of Alberta, Edmonton, Alberta (Canada) Karen R. Cangialosi, Keene State College, Keene, New Hampshire (USA) James Carpenter, US Dept, of Agriculture- ARS, Tifton, Georgia (USA) James Carrell, University of Missouri, Columbia, Missouri (USA) Jonathan Coddington, Smithsonian Institution, Washington, D.C. (USA) James Cokendolpher, Lubbock, Texas (USA) Thomas S. Collett, University of Sussex, Palmer Brighton (UK) Frederick A. Coyle, Western Carolina University, Cullowhee, North Carolina (USA) Bruce Cutler, University of Kansas, Lawrence, Kansas (USA) Gary Dodson, Ball State University, Muncie, Indiana (USA) Charles D. Dondale, Agriculture Canada, Ottawa (Canada) Mark Dowton, Wollongong University, Wollongong (Australia) Jason Dunlop, Museum fur Naturkunde der Humboldt-Universitat, Berlin (Germany) William G. Eberhard, Universidad de Costa Rica (Costa Rica) Terry L. Erwin, Smithsonian Institution, Washington, D.C, (USA) Theodore A. Evans, CSIRO Entomology, Canberra (Australia) Victor Fet, Marshall University, Huntington, West Virginia (USA) Robert S. Fritz, Vassar College, Poughkeepsie, New York (USA) Rosemary Gillespie, University of California, Berkeley, California (USA) Charles Griswold, California Academy of Sciences, San Francisco, California (USA) Alexander Gromov, Institute of Zoology, Akademgorodok, Almaty (Kazakhstan) Juraj Halaj, Miami University, Oxford, Ohio (USA) Mark Harvey, Western Australian Museum, Perth (Australia) Marshal C. Hedin, San Diego State University, San Diego, California (USA) Linden E. Higgins, University of Massachusetts, Amherst, Massachusetts (USA) Robert Holmberg, Athabasca University, Athabasca, Alberta (Canada) Gustavo Hormiga, George Washington University, Washington, D.C. (USA) Norman Homer, Midwestern State University, Wichita Falls, Texas (USA) Bernhard Huber, American Museum, of Natural History, New York, New York (USA) Larry Hurd, Washington & Lee University, Lexington, Virginia (USA) Robert Jackson, University of Canterbury, Christchurch (New Zealand) Elizabeth Jakob, University of Massachusetts, Amherst, Massachusetts (USA) Mark Judson, Museum National d’Histoire Naturelle, Paris (France) Jacqueline Kovoor, Museum National d’Histoire Naturelle, Paris (France) Tjorbom Kronestedt, Swedish Museum of Natural History, Stockholm (Sweden) 125 126 THE JOURNAL OF ARACHNOLOGY Herbert Levi, Museum of Comparative Zoology, Harvard University, Cambridge, Massachus- setts (USA) Norman C. Leppla, University of Florida, Gainesville, Florida (USA) Gabor Lovei, Danish Institute of Agricultural Sciences, Slagelse (Denmark) Wilson Lourengo, Museum National d’Histoire Naturelle, Paris (France) Yael Lubin, Ben Gurion University, Sede Boqer (Israel) Volker Mahnert, Museum d’Histoire Naturelle, Geneva (Switzerland) Samuel Marshall, J.H. Barrow Field Station, Hiram College, Hiram, Ohio (USA) Gary L. Miller, University of Mississippi, Oxford, Mississippi (USA) Douglass Morse, Brown University, Providence, Rhode Island (USA) Hirotsugu Ono, National Science Museum, Tokyo (Japan) Brent Opell, Virginia Polytechnic Institute and State University, Blacksburg, Virginia (USA) Matthew H. Persons, Susquehanna University, Selinsgrove, Pennsylvania (USA) Norman Platnick, American Museum of Natural History, New York, New York (USA) Thomas Prentice, University of California, Riverside, California (USA) Fred Punzo, University of Tampa, Tampa, Florida (USA) Jonathan Reiskind, University of Florida, Gainesville, Florida (USA) Marianne Robertson, Millikin University, Decatur, Illinois (USA) Ann Rypstra, Miami University, Oxford, Ohio (USA) Ferenc Samu, Hungarian Academy of Sciences, Budapest (Hungary) Nikolaj Scharff, Zoological Museum, Copenhagen (Denmark) William A. Shear, Hampden-Sydney College, Hampden-Sydney, Virginia (USA) W David Sissom, West Texas A&M University, Canyon, Texas (USA) Deborah R. Smith, University of Kansas, Lawrence, Kansas (USA) Song Daxiang, Institute of Zoology, Beijing (China) Christopher Starr, University of the West Indies, St. Augustine (Trinidad & Tobago) Gail Stratton, University of Mississippi, University, Mississippi (USA) Cathy R. Tugmon, Augusta State University, Augusta, Georgia (USA) Cor Vink, Lincoln University, Lincoln (New Zealand) George W Uetz, University of Cincinnati, Cincinnati, Ohio (USA) Erich Volschenk, Western Australian Museum, Perth (Australia) James Wagner, Transylvania University, Lexington, Kentucky (USA) Mary Whitehouse, The University of Queensland, Brisbane (Australia) David Zeh, University of Nevada, Reno, Nevada (USA) Jeremy Zuiko-Miller, George Washington University, Washington, D.C. (USA) INSTRUCTIONS TO AUTHORS (revised August 2000) Manuscripts are accepted in English only. Authors whose primary language is not English may consult the editors for assistance in obtaining help with manuscript preparation. All manuscripts should be prepared in general accordance with the current edition of the Council of Biological Editors Style Manual unless instructed otherwise below. Authors are advised to con- sult a recent issue of the Journal of Arachnology for additional points of style. Manuscripts longer than three printed journal pages should be prepared as Feature Articles, shorter papers as Short Commun- ications. Send four identical copies of the typed mate- rial together with copies of illustrations to the Managing Editor of the Journal of Arachnology: Petra Sierwald, Managing Editor; Division of In- sects, Dept, of Zoology, The Field Museum of Na- tural History, 1400 South Lakeshore Drive, Chicago, IL 60605 USA [Telephone: (312)-665-7744; FAX: (312)-665-7754; E-mail psierwald@fmnh.org]. The Managing Editor will forward your manuscript to one of the Subject Editors for the review process. You will receive correspondence acknowledging the receipt of your manuscript from the responsible subject editor or the managing editor, with the manuscript number of your manuscript. Please use this number in all correspondence regarding your manuscript. Corres- pondence relating to manuscripts should be directed to the Managing Editor or the appropriate Subject Editor. After the manuscript has been accepted, the author will be asked to submit the manuscript on a computer disc in a widely-used word processing program. Indicate clearly on the computer disc the word processing pro- gram and the type of computer (Mac or PC). FEATURE ARTICLES Title page. — The title page will include the complete name, address, and telephone number of the author with whom proofs and correspondence should be exchanged, a FAX number and electronic mail address if available, the title in capital letters, each author’s name and address, and the running head (see belov/). Abstract. — The heading in capital letters should be placed at the beginning of the first paragraph set off by a period. A second abstract, in a language pertinent to the nationality of the author(s) or geographic region(s) emphasized, may be included. Keywords. — Give 3-5 appropriate keywords fol- lowing the abstract. Text. — Double-space text, tables, legends, etc. throughout. Three categories of headings are used. The first category (METHODS, RESULTS, etc.) is typed in capitals and on a separate line. The second category of heading, in bold type, begins a paragraph with an indent and is separated from the text by a period and a dash. The third category of heading may or may not begin a paragraph but is italicized and separated from the text by a colon. Use only the metric system unless quoting text or referencing collection data. All decimal fractions are indicated by the period (e.g.,-0.123). Citation of references in the text: Cite only papers already published or in press. Include within parenthe- ses the surname of the author followed by the date of publication. A comma separates multiple citations by the same author(s) and a semicolon separates citations by different authors, e.g., (Smith 1970), (Jones 1988; Smith 1993), (Smith 1986, 1987; Smith & Jones 1989; Jones et al. 1990). Include a letter of permission from any person who is cited as providing unpublished data in the form of a personal communication. Literature cited section. — Use the following style, and include the full unabbreviated journal title. Lombardi, S.J. & D.L. Kaplan. 1990. The amino acid composition of major ampullate gland silk (drag- line) of Nephila clavipes (Araneae, Tetragnathidae). Journal of Arachnology 18:297-306. Kraffit, B. 1982. The significance and complexity of communication in spiders. Pp. 15-66, In Spider Communications: Mechanisms and Ecological Significance. (P.N. Witt & J.S. Rovner, eds.). Princeton University Press, Princeton, New Jersey. Footnotes. — Footnotes are permitted only on the first journal page to indicate current address or other information concerning the author. These are placed together on a separate manuscript page. Tables and fig- ures may not have footnotes. Running head. — The author’s sumame(s) and an abbreviated title should be typed all in capital letters and must not exceed 60 characters and spaces. The running head should be placed near the top of the title page. Taxonomic articles. — Consult a recent taxonomic article in the Journal of Arachnology for style, or contact the Subject Editor for Systematics. Papers containing the original taxonomic description of the focal arachnid taxon should be listed in the Literature Cited section. Tables. — Each table, with the legend above, should be placed on a separate manuscript page. Only horizon- tal lines (usually three) should be included. Tables may not have footnotes; instead, include all information in the legend. Make notations in the text margins to indi- cate the preferred location of tables in the printed text. Illustrations. — ^Address all questions concerning illustrations to the Editor of the Journal of Arach- nology: James W. Berry, Editor; Dept, of Biological Sciences; Butler University, Indianapolis, Indiana 46208 USA. [Telephone (317)-940-9344; FAX (317)- 940-9519; E-mail: jwberry@butler.edu]. All art work must be camera-ready (mounted and labeled) for reproduction. Figures should be arranged so that they fit (vertically and horizontally) the printed journal page, either one column or two columns, with a mini- mum of wasted space. When reductions are to be made by the printer, pay particular attention to width of lines and size of lettering in line drawings. Multiple photos assembled on a single plate should be mounted with only a minimum of space separating them. In the case of multiple illustrations mounted together, each illus- tration must be numbered sequentially rather than given an alphabetic sequence. Written on the back should be the name(s) of author(s) and an indication of top edge. Indicate whether the illustration should be one column or two columns in width. The overall dimensions should be no more than 1 1 inches (28 cm) X 14 inches (36 cm). Larger drawings present greater difficulty in shipping and greater risks of damage for which the JoA assumes no responsibility. In manu- scripts for review, photocopies are acceptable, and should be reduced to the exact measurements that the author wants to appear in the final publication. Make notations in the text margins to indicate the preferred position of illustrations in the printed text. Color plates can be printed, but the author must assume the full cost, currently about $600 per color plate. Legends for illustrations should be placed together on the same page(s) and separate from the illustrations. Each plate must have only one legend, as indicated below: Figures 1-4. — A-us x-us, male from Timbuktu: 1. Left leg; 2. Right chelicera; 3. Dorsal aspect of geni- talia; 4. Ventral aspect of abdomen. Figures 27-34. — Right chelicerae of species of A-us from Timbuktu: 27, 29, 31, 33. Dorsal views; 28, 30, 32, 34. Prolateral views of moveable finger; 27, 28. A- us x-us holotype male; 33, 34. A-us y-us male. Scale = 1.0 mm. Assemble manuscript for mailing. — Assemble the separate sections or pages in the following sequence; title page, abstract, text, footnotes, tables with legends, figure legends, figures. Page charges and reprints. — There are no page charges, but the author will be charged for changes made in the proof pages. Reprints are available only from the Allen Press and should be ordered when the author receives the proof pages. Allen Press will not accept reprint orders after the paper is published. SHORT COMMUNICATIONS The above instructions pertaining to Feature Articles apply also to Short Communications, which should be prepared in the same manner as regular Feature Articles. Short Communications are usually limited in length to three journal pages, including tables and figures. They will be printed in a smaller (10 point) typeface. CONTENTS The Journal of Arachnology Volume 29 Feature Articles Number 1 Revision of the spider genus Neoana^raphis (Araneae, Liocranidae) by Richard S. Vetter 1 Notes on the genus Sybota with a description of a new species from Argentina (Araneae, Uloboridae) by Cristian J. Grismado 11 Okileucauge sasakii, a new genus and species of spider from Okinawajima Island, southwest Japan (Araneae, Tetragnathidae) by Akio Tanikawa . . 16 Revision de las especies de Freya del grupo decorata (Araneae, Salticidae) by Maria Elana Galiano 21 Description of a new species in the nitidulus group of the genus Vaejovis (Scorpiones, Vaejovidae) by E. Michelle Capes 42 A new species of Vaejovis (Scorpiones, Vaejovidae) from Sonora, Mexico by Brent E. Hendrixson 47 The influence of group size on dispersal in the social spider Stegodyphus mimosarum (Araneae, Eresidae) by Marilyn Bodasing, Rob Slotow & Tanza Crouch 56 Sexual size dimorphism and juvenile growth rate in Linyphia triangularis (Linyphiidae, Araneae) by Gary H.P. Lang 64 Predatory behavior of three species of sac spiders attacking citrus leafrniner by Divina M. Amalin, Jonathan Reisldnd, Jorge E. Pena & Robert McSorley 72 Variation in the chemical composition of orb webs built by the spider Nephila clavipes (Araneae, Tetragnathidae) by Linden E. Higgins, Mark A. Townley, Edward K. TiUinghast & Mary Ann Rankin 82 Patterns of abundance of four species of wandering spiders (Ctenidae, Ctenus) in a forest in central Amazonia by Thierry R. Gasnier «& Hubert Hofer 95 Ontogenetic change in coloration and web-building behavior in the tropical spider Eriophora fuliginea (Araneae, Araneidae) by Barbara Graf & Wolfgang Nentwig 104 Short Communications Zoropsidae: A spider family newly introduced to the USA (Araneae, Ente- legynae, Lycosoidea) by Charles E. Griswold 4& Darrell Ubick 111 Dispersal of Stegodyphus dumicola (Araneae, Eresidae): They do balloon after all! by Jutta M. Schneider, Jorg Roos, Yael Lubin & Johannes R. Henschel 114 A technique for individually identifying tarantulas using passive integrated transponders by Steven B. Reichling 4& Chris Tabaka 117 Spiders feeding on earthworms by Martin Nyffeler, Hans Moor & Ranier F. Foelix 119 List of Manuscript Reviewers for 2000 (Volume 28) 125 Journal of ARACHNOLOGY OFFICIAL ORGAN OF THE AMERICAN ARACHNOLOGICAL SOCIETY VOLUME 29 2001 NUMBER 2 THE JOURNAL OF ARACHNOLOGY EDITOR-IN-CHIEF: James W. Berry, Butler University MANAGING EDITOR: Paula Cushing, Denver Museum of Nature & Science SUBJECT EDITORS: Ecology — Maggie Hodge, Hiram College; Systematics — Mark Harvey, Western Australian Museum; Behavior and Physiology — Robert Suter, Vassar College EDITORIAL BOARD: Alan Cady, Miami University (Ohio); James Carrel, University of Missouri; Jonathan Coddington, Smithsonian Institution; William Eberhard, Universidad de Costa Rica; Rosemary Gillespie, University of California, Berkeley; Charles Griswold, California Academy of Sciences; Marshal Hedin, San Diego State University; Herbert Levi, Harvard University; Brent Opell, Virginia Polytechnic Institute & State University; Norman Platnick, American Museum of Natural History; Ann Rypstra, Miami University (Ohio); Paul Selden, University of Manchester (UK.); Matthias Schaefer, Universitset Goettingen (Germany); William Shear, Hampden- Sydney College; Petra Sierwald, Field Museum; Keith Sunderland, Horticulture Research International (UK.); I-Min Tso, Tunghai University (Taiwan). The Journal of Arachnology (ISSN 0160-8202), a publication devoted to the study of Arachnida, is published three times each year by The American Arach- nological Society. Memberships (yearly): Membership is open to all those in- terested in Arachnida. Subscriptions to The Journal of Arachnology and American Arachnology (the newsletter), and annual meeting notices, are included with mem- bership in the Society. Regular, $40; Students, $25; Institutional, $125 . Inquiries should be directed to the Membership Secretary (see below). Back Issues: Patricia Miller, P.O. Box 5354, Northwest Mississippi Community College, Senatobia, Mississippi 38668 USA. Telephone: (601) 562-3382. Undelivered Issues: Allen Press, Inc., 1041 New Hampshire Street, P.O. Box 368, Lawrence, Kansas 66044 USA. THE AMERICAN ARACHNOLOGICAL SOCIETY PRESIDENT: Brent D. Opell (2001-2003), Department of Biology, Virginia Polytechnic Institute and State University, Blacksburg, Virginia 24061 USA. PRESIDENT-ELECT: Gary Miller (2001-2003), Department of Biology, University of Mississippi; University, Mississippi 38677 USA. MEMBERSHIP SECRETARY: Norman I. Platnick (appointed), American Museum of Natural History, Central Park West at 79th St., New York, New York 10024 USA. TREASURER: Gail E. Stratton, Department of Biology, University of Missis- sippi, University, Mississippi 38677 USA. SECRETARY: Alan Cady, Dept, of Zoology, Miami Univ, Middletown, Ohio 45042 USA. ARCHIVIST: Lenny Vincent, Fullerton College, Fullerton, California 92634. DIRECTORS: Bruce Cutler (2000-2002), Richard Bradley (2001-2003), Frederick Coyle (2001-2003). HONORARY MEMBERS: C. D. Dondale, H. W. Levi, A. F. Millidge, W. Whit- comb. Cover photo: Scorpion {Centruroides sp.) with young. Photo taken by the late M.W. Tyler of Umatilla, Florida about 1953. Publication date: 31 August 2001 @ This paper meets the requirements of ANSI/NISO Z39.48-1992 (Permanence of Paper). 2001. The Journal of Arachnology 29:127-134 AUTOECOLOGY AND DESCRIPTION OF MUMMUCIA MAURYI (SOLIFUGAE, MUMMUCIIDAE), A NEW SOLIFUGE FROM BRAZILIAN SEMI-ARID CAATINGA Eduardo Xavier: Universidade Estadual de Feira de Santana, Laboratorio de Animals Pegonhentos e Herpetologia. Campus Universitario. 44031-460. Feira de Santana, BA, Brazil Lincoln Suesdek Rocha: Museu de Zoologia da Universidade de Sao Paulo. Segao Fntomologia. Cx.P. 42694 - CFP 04299-970. Sao Paulo, SP, Brazil ABSTRACT. The Brazilian solifuge Mummucia mauryi new species (Solifugae, Mummuciidae) from sand dunes of the Sao Francisco River, in semiarid caatinga domain, is herein described, with illustrations of the main taxonomic characters. This is the first species of Solifugae described from the Brazilian caatinga. The specimens were collected in pitfall traps during both the rainy and dry seasons. It exhibits diurnal activity and a clumped distribution (Morisita’s index = 3.32 and 1.38 for rainy and dry season, respectively). Sun-exposed areas were avoided during the dry season, when preference for the cactacean Opuntia inamoena was detected. We suggest this association is related to predator avoidance. RESUMO. O soKfugo Mummucia mauryi (Solifugae, Mummuciidae) e descrito a partir de exemplares coletados nas dunas interiores do Rio Sao Francisco (BA), com ilustragoes dos principais caracteres tax- onomicos. Esta e a primeira especie de Solifugae descrita para o dommio da caatinga semi-arida. Estudos sobre a autoecologia indicam atividade diuma; distribuigao do tipo agregado (mdice de Morisita = 3.32 e 1.38 para as estagbes chuvosa e seca, respect! vamente); preferencia negativa por areas mais expostas a insolagao durante a estagao seca e preferencia pela cactacea Opuntia inamoena, o que sugerimos estar relacionado a protegao contra predadores. Keywords: Arachnida, microhabitat, Solpugida, systematics Knowledge of the order Solifugae in the Neotropical Region is very limited, especially for the portion of South America occupied by Brazil, which includes distinctive environ- ments such as caatinga and cerrado. In this area, studies on Solifugae are scarce and there are a few distributional records, some of which were unfortunately excluded from some world maps (Savory 1964; Punzo 1998). Mau- ry (1984) comprehensively listed and anno- tated all Brazilian records: Gaucha fasciata Mello-Leitao 1924 (Mummuciidae) from Por- to Alegre, State of Rio Grande do Sul (Mello- Leitao, 1924); Metacleobis fulvipes Roewer 1934 (Mummuciidae) from Cuiaba, State of Mato Grosso (Roewer 1934); Ammotrecha friedlaenderi Roewer 1954 (Ammotrechidae) from Mendes, State of Rio de Janeiro (Roewer 1954); an undetermined species of Ammotre- chidae, from State of Roraima, Brazil (Maury 1982); and, at a later date, an undetermined species of Ammotrechidae from Manaus, State of Amazonas (Hofer & Beck 1995). Unfortunately, these records are mere oc- currence registers; and except for the record mentioned by Hofer & Beck (1995), there is no ecological information about the species. Moreover, Ammotrecha friedlaenderi and Me- tacleobis fulvipes are known only from their types and the records from Roraima and Ma- naus are immature individuals and therefore cannot be identified at this time. Recently, new records of Solifugae in Bra- zilian Amazonia and cerrado have been noted (Rocha & Cancello 1997), extending the known distribution and habitats in South America, which is probably greater than pre- sent records indicate. In addition to our poor understanding of the systematics and diversity of Neotropical Solifugae, very little is known about their ecology and behavior. Most of the ecological studies on these arachnids deal 127 128 THE JOURNAL OF ARACHNOLOGY Figure 1 . — Continental sand dunes of Sao Francisco river, Brazil, in the caatinga morphoclimatic domain showing the studied area at Ibiraba. with North American and African species. This is the first study presenting data on the ecology of a Brazilian solifuge. In the present paper Mummucia mauryi new species is de- scribed from the State of Bahia, in the Bra- zilian semi-arid caatinga domain. Ecological data, including circadian activity, spatial dis- tribution pattern and microhabitat preferences are discussed. METHODS Terminology used in the description such as “bristles,” “setae” and “spines” are used ac- cording to Muma (1951). Some of these struc- tures, bearing a bifurcation at the tips, are called “bifid bristles,” etc. Cheliceral teeth are also named according to Muma (1951), where sizes of cheliceral teeth are ordered with Roman numerals, and size I is larger than II and so on. The tarsal spination is repre- sented as in Roewer (1934), Muma (1951) and Maury (1970). The term “ctenidia” is used as in Maury (1984). Type material is deposited in Museu de Zoologia da Universidade de Sao Paulo (MZUSP), State of Sao Paulo, Brazil. The study was carried out on Ibiraba — sand dunes on the northeastern Brazilian caatinga (Fig. 1). The vegetation physiognomy is de- scribed in Rocha (1991). There is a large amount of exposed sandy soil. Nimer (1979) reported an annual mean precipitation of 692 mm. There are two distinct seasons, a dry sea- son from April-September and a rainy season from October-April. A grid with 128 pitfall traps covered two dune summits and two valleys during Febru- ary, and a similar one with 120 traps was set in September 1996. Each trap consisted of a plastic cylindrical receptacle (30 X 40 cm) to which three drift fences (Corn 1994) 1.5 m long were radially attached. The distance be- tween traps was 7 m. No chemicals or baits were used in the traps. The microhabitat around each trap was recorded once each month as seven variables: microgeographic position (summit, talude, plateau or valley), vegetation covering (em^ inside a 3.0 m di- ameter circle centered in each trap) by trees, shrubs, subshrubs, Bromelia antiacantha (Bromeliaceae), Opuntia inamoena (Cacta- ceae) and litter cover. The pitfalls were scanned for solpugids twice a day, around 0600 h and 1700 h. Comparing three methods of collecting solifuges, Muma (1980) sug- gested pitfall trapping as the most suitable method for number of individuals and species composition estimates. Pitfalls were also used XAVIER & ROCHA— AUTOECOLOGY AND DESCRIPTION OF MUMMUCIA 129 by Griffin (1990) in the Namibia desert to study microhabitat preferences, species rich- ness and activity patterns. For spatial distribution pattern analysis, Morisita’s index of dispersion (Id) was applied and its departure from the unity was evaluated by a chi-square test (Brower et al. 1997). The analysis on preferemces by the solpugid on each microhabitat variable on ratio scale was carried out using Mann- Whitney t/-test com- paring the distribution of the values of each variable by the event of capture with the dis- tribution of the same variable obtained on the entire sampling grid. Habitat preference on the cathegoric variable microgeographic position was checked using the goodness of fit test. Family Mummuciidae Mummucia mauryi new species Rocha Types. — Holotype male, MZUSP 16470 (col. R Rocha, 26 February 1996). Paratypes: 161$ MZUSP 15784 and 2$ MZUSP 15932 (col. E. Xavier, February 1996); 16 MZUSP 16471 and 26 MZUSP 16472 (col. P. Rocha, 25 February 1996); 16 MZUSP 16473 (col. E. Xavier, 11 December 1996); 16 MZUSP 16474 (col. E. Xavier, 12 December 1996). All from Ibiraba, western side of Sao Francis- co River, State of Bahia, Brazil. 10°48'S, 42°50'W. Etymology. — The specific name is given in honor to the late Dr. Emilio Maury. Diagnosis. — Mummucia mauryi is a species of Mummuciidae whose anterior tooth of the movable finger is smaller than the intermedi- ate tooth in males and similar to intermediate in females. Description. — Male: Coloration in 80% ethanol: Prosoma. Propeltidium white, central portion brown, dark brown near the lateral lobe grooves. Ocular tubercle black, with a longitudinal white narrow stripe between the eyes. Peltidium white, posterior border brown. Parapeltidium, mesopeltidium and metapelti- dium similar to opisthosomal tergites. Chelic- erae pale brown, three longitudinal white stripes on ectal face joined dorsally above the fondal teeth. Pedipalpi and legs brown, ventral face pale brown. Malleoli pale brown with small brown spots on distal border (Fig. 5). Opisthosoma: Lateral borders of tergites white, with wide dark brown stripe on the cen- tral half, which is darker near the posterior border of the tergites. Brown bifid setae with brown sockets when they are in white area of the tergites, and white sockets when in the dark brown area. Pleurites (Fig. 6) white, dor- sal portion dark brown. Pale brown translu- cent bifid bristles in the white portion have sockets shaped into dark brown spots, which are generally arranged as in Fig. 6. Sternites pale brown, lateral posterior borders brown in the four distal. First to fourth post-spiracular sternites with brown spots which include the sockets of some bifid bristles. All covering bristles and bifid bristles are translucent pale brown. Morphology and chaetotaxy: Prosoma: Propeltidium with some scattered bifid setae, slightly wider than long (Table 1) and sepa- rated from lateral lobes by dorsal grooves. Oc- ular tubercle prominent with bifid setae ante- riorly oriented. Distance between two eyes about twice eye diameter. Peltidium narrow, with a transverse row of bifid setae. Parapel- tidium smooth. Mesopeltidium wider than long, semicircle-shaped, with several bifid se- tae in the posterior border. Metapeltidium wider than long, with several bifid setae. Che- licerae (Figs. 2, 3): stridulatory apparatus on mesal face with seven parallel narrow grooves; ectal face with several short bristles and several setae, bifid or acuminate; movable finger with one anterior, one intermediate and one principal tooth, graded in size from distal to proximal III, II, I; fixed finger dentition: two anterior teeth (the first one may be ves- tigial), one intermediate and one principal tooth, graded in size from distal to proximal II, I, IV, III; five ectal fondal teeth, graded in size, from distal to proximal I, II, III, II, II (the 5‘*’ may be absent); three mesal fondal teeth, graded in size from distal to proximal I, II, II, the first distal separated from the oth- ers by a diastema; in the center of the dorsal face the fixed finger bears one very long seta (about the length of femur IV) with a promi- nent socket; flagellum (Figs. 2, 10) thin, trans- lucent drop-shaped vesicle, laterally flattened and with a longitudinal ectal opening (in the face adjacent to the chelicera), which extends from near the attachment base to the tip of the flagellum. The attachment base of the flagel- lum is a sclerotized ring placed posteriorly in its ectal face. Pedipalp: tarsi immovable, with- out spines, densely covered by differentially sized bifid bristles, with some very long setae in metatarsi and tibiae (about twice the length 130 THE JOURNAL OF ARACHNOLOGY Figures 2-9. — Mummucia mauryi new species. 2. Male right chelicera, mesal view; 3. Male right chelicera, ectal view; 4. Female left chelicera, ectal view; 5. Male right malleolus V; 6. Male left pleurites; 7. Female genital sternite; 8. Female left leg III; 9. Male right leg IV. of pedipalpal tibia). Legs: with several differ- entially-sized bifid bristles and some bifid se- tae. Some very long setae in dorsal surface (about twice the length of metatarsus IV). Leg I thin, without claws and spines. Legs II and III (see female leg III, Fig. 8): tibiae with I or 1 . 1 ventral bifid spines and a distal pair of ventral spines; metatarsus with three retrola- teral spines and 1.1.2 ventral spines; tarsi two- segmented with 1.2. 2/1.2 or 1.2.272,2 ventral XAVIER & ROCHA— AUTOECOLOGY AND DESCRIPTION OF MUMMUCIA 131 Table 1. — Morphometric characters of Mummucia mauryi new species. Measurements are in milli- meters (except propeltidium length/width ratio) and were recorded as described in Muma (1951). Morphometric character Male holotype (MZUSP 16470) Range among males (7 individuals) Female paratype (MZUSP 15932) Range among females (3 individuals) Total length 7.55 7.10-8.10 11.20 8.55-11.20 Cheliceral length 1.71 1.70-1.80 2.60 1.75-2.60 Cheliceral width 0.50 0.50-0.61 0.88 0.60-0.88 Propeltidium length 1.15 1.11-1.30 1.47 0.99-1.47 Propeltidium width 1.35 1.32-1.40 2.05 1.35-2.05 Propeltidium length/ width ratio 0.85 0.84-0.93 0.72 0.71-0.73 Pedipalp 4.80 4.45-4.90 5.70 3.90-5.70 Leg I 4.00 3.20-4.20 4.90 3.40-4.90 Leg IV 7.40 6.20-8.40 9.50 5.50-9.50 spines. Leg IV (Fig. 9): tibia with an anterior row of 1 . 1 . 1 . 1 ventral bifid spines and a distal pair of ventral spines; metatarsus with 1.1. 1.1. 2 ventral spines; tarsi three-segmented, with 2. 2. 2/2/1. 2 or 2. 2. 2/2/2. 2 ventral spines. Malleoli as in Fig. 5. Opisthosoma: Tergites wider than long, with rounded borders, cov- ered by bifid setae and bifid bristles. Sternites wider than long, densely covered by bifid bris- tles. Genital operculum with central longitu- dinal opening. Posterior border of 2"^^ post-spi- racular stemite with a row of about 50 ctenidia, more rigid and slightly longer than the bifid bristles in sternites. Morphometric characters in Table 1. Female: Similar to male, but with the fol- lowing particular features. Coloration in eth- anol 80% similar to male, but with lighter to- nalities. Morphology and chaetotaxy: Prosoma: Propeltidium wider than long with numerous bifid setae and small bifid bristles. Eyes separated by three times the eye diam- eter. Chelicerae (Fig. 4): movable finger with one anterior, one intermediate and one prin- cipal tooth graded in size from distal to prox- imal II, II, L Fixed finger with two anterior teeth, one intermediate and one principal, graded in size from distal to proximal I, I, III, II (the first anterior may be slightly smaller than the 2"^). Five ectal fondal teeth graded in size from distal to proximal I, I, IV, II, III (the 3*^^ may be vestigial). Mesal fondal teeth as in male. Leg III similar to male (Fig. 8). Leg IV: tibia with an anterior row of 1.1.1 or 1.1. 1.1 ventral bifid spines and a distal pair of ventral spines; metatarsus with 1.1.2 or 1.1. 1.2 ventral spines. Opisthosoma. Sternites densely and uniformly covered by bifid bristles, without conspicuous sockets. Genital operculum prominent, fan-shaped, round-bordered, with central longitudinal opening (Fig. 7). Posterior border of 2"^ post-spiracular sternite with a row with several ctenidia, slightly longer than the bifid bristles of the sternites. Morphomet- ric characters in Table 1. Systematic comments. — There is no con- sensus about the number and the systematic position of genera of the family Mummuci- idae. For instance, Muma (1976) recognizes 1 1 genera in Mummuciidae (six of them monotypic), whereas Maury (1984) has trans- ferred three of these genera to the family Am- motrechidae. The typical genus Mummucia Simon 1879 has only three species, which are known only by females (Muma 1976) and one of them {Mummucia patagonica Roewer 1934) should be transferred to family Am- motrechidae, since this species bears spines at the pedipalpal metatarsi, a distinctive charac- ter of this family. According to Maury (pers. comm. 30 December 1997, 1998), there is no good character to distinguish the genera of Mummuciidae, so that the most conservative decision is to consider the new Mummuciidae species herein described as belonging to the typical genus Mummucia, until more precise information about the taxonomy and phylog- eny of the group become available. The shape of the flagellum is a good char- acter for the definition of Neotropical families and in Mummuciidae the flagellum is vesic- ular (Maury 1984). The flagellum of M. mau~ ryi new species bears a longitudinal ectal opening, which has not been reported in other 132 THE JOURNAL OF ARACHNOLOGY Figure 10. — Photomicrograph of the flagellum of Mummucia mauryi new species, showing the lon- gitudinal lateral opening in the face adjacent to chelicerae, indicated by the long arrow. The short arrow indicates the attaching ring. Scale = 0.1 mm. Mummuciidae species. Therefore this charac- ter may be useful in further studies on the sys- tematics of Mummuciidae. AUTOECOLOGY Specimens of Mummucia mauryi new spe- cies represented 5% of all arachnid specimens collected by the pitfall trap method described above (Xavier & Rocha 1998): 22 specimens were collected in February (rainy season) and 88 in September (dry season). The traps ex- amined during the morning showed no soli- fuges. This agrees with Maury (1984), who predicted that mummuciids should be the only South American solifuges with diurnal habits. Indeed, Cloudsley-Thompson (1977, 1978) suggested that Mummucia variegata (Gervais) 1 849 is a diurnal species, and that diurnal ac- tivity is exhibited by smaller and brightly-col- ored solpugid species. On the other hand, Wharton (1987) states that “there are several large, diurnal solifuge species in the arid re- gions of southern Africa.” Mummucia mauryi new species agrees with both Cloudsley- Thompson and Maury predictions, being a brightly-colored, small and diurnal mummu- ciid species. Because solifuges are “unusually tolerant of high temperature (...) and have very low transpiration rates (. . .), it seems probable that the avoidance of predators may be of greater significance than thermal phys- iological requirements in their night-active be- havior” (Cloudsley-Thompson 1991). Follow- ing this idea, we should expect diurnal solifuges to exhibit additional mechanisms of predator avoidance. As discussed below, the analyses performed detected preferences only twice for microhab- itat variables and once for microgeographic region. There was detected a positive prefer- ence for areas covered by Opuntia inamoena (Cactaceae) during the dry season {U = 4592, P = 0.03). The African solpugid Lipophaga trispinosa Purcell 1903 is restricted to low plant cover areas during dry periods (Dean & Griffin 1993). They also found a low diversity of solifuges associated with loose sand soil. For Eremobates marathoni Muma 1951, Pun- zo (1998) detected a preference for sandy soils, open areas and scattered clumps of veg- etation, and he suggested that “scattered clumps of vegetation afford cover and protec- tion from predators (. . .) including night hawks, roadrunners, scorpions, and other so- lifuges.” In fact, Opuntia inamoena is an in- XAVIER & ROCHA— AUTOECOLOGY AND DESCRIPTION OF MUMMUCIA 133 hospitable spiny plant, which may be avoided by many possible predators such as diurnal birds. Nevertheless, a parallel study on lizards at the same area was carried out by Rocha (1998) showing preference for O. inamoena by the lizard Tropidurus psammonastes (Tro- piduridae). The lizard's diet includes mainly ants and insect larvae, and solifuges are rarely preyed on, as one only event was recorded. “The avoidance of open areas devoid of vegetation appears to be a rather common trait in solifuges” (Punzo 1998). This seems to be the case to Mummucia mauryi new species. During the dry season it showed a negative preference for dune summits (x^ = 9.74, 0.025 < P < 0.05) and areas covered by heliophylic subshrubs {U = 4212, P = 0.009), which is here interpreted as avoidance of sun-exposed areas, which may be associated with avoid- ance of predators and/or environmental ex- tremes. Morisita’s index was 3.32 (x^ = 175.818, 0.005 > P > 0.001) in the rainy season and 1.38 (x^ - 151.209, 0.025 > P> 0.01) in the dry season, indicating a clumped distribution through the year. Investigating the eremobatid solifuge Eremobates palpisetulosus Fichter 1941, Punzo (1997) found a clumped disper- sion pattern, without significant differences in adult dispersion as a function of sex or season. ACKNOWLEDGMENTS We are grateful to Dr. Eliana M. Cancello for providing her laboratory and optical equipment used in the description of the new species. We are also indebted to Pedro L.B. Rocha, Eleonora Trajano, Ricardo Pinto-da- Rocha for their critical reading of the manu- script. L.S.R. would like especially to thank in memoriam Dr. Emilio Maury, deceased July 1998, not only for critical reading of the de- scription of Mummucia mauryi new species, but also for the friendship and the invaluable help during the short period he worked at Maury's laboratory. LITERATURE CITED Brower, J.E., J.H. Zar & C. Von Ende. 1997. Field and Laboratory Methods For General Ecology. McGraw-Hill. New York. Cloudsley-Thompson, J.L. 1977. Adaptational bi- ology of Solifugae (Solpugida). Bulletin of the British Arachnological Society 4(2):61-7L Cloudsley-Thompson, J.L. 1978. Biological clocks in Arachnida. Bulletin of the British Arachnol- ogical Society 4(4): 184-191. Cloudsley-Thompson, J.L. 1991. Ecophysiology of Desert Arthropods and Reptiles. Springer- Verlag, Berlin. Corn, RS. 1994. Straight-line drift fences and pit- fall traps. P. 109. In Measuring and Monitoring Biological Diversity: Standard Methods for Am- phibians. (W.R. Heyer, M.A. Donnelly, R.W. McDiarmid, L.C. Hayek & M.S. Foster (eds.). Smithsonian Inst. Press, London. Dean, W.R.J. & E. Griffin. 1993. Seasonal activity patterns and habitats in Solifugae (Arachnida) in southern Karoo. South African Journal of Zool- ogy 28:91-94. Griffin, E. 1990. Seasonal activity, habitat selec- tion and species richness of Solifugae (Arachni- da) on the gravel plains of the central Namib Desert. Transvaal Museum Monograph 7:77-82. Hofer, H. & L. Beck. 1995. Die Spinnentierfauna des Regenwaldreservats “Reserva Ducke” in Zentralamazonien I. Natur und Museum 125(12): 389-401. Maury, E.A. 1970. Sobre la presencia de Gaucha fasciata Mello-Leitao 1 924 en la Argentina. Phy- sis 79(29):357-362. Maury, E.A. 1982. SoKfugos de Colombia y Ven- ezuela (Solifugae, Ammotrechidae). Journal of Arachnology 10(2): 123-143. Maury, E.A. 1998. Solifugae. Pp. 560-568. In Bio- diversdidad de Artropodos Argentinos. (J.J. Mor- rone & S. Coscaron, eds.). Ediciones SUR. La Plata. Maury, E.A. 1984. Las familias de sohfugos amer- icanos y su distribucion geografica (Arachnida, Solifugae). Physis C. 42(103):73-80. Mello-Leitao, C. 1924. A new South- American solpugid. Revista Chilena de Historia Natural XXVIII: 140-143. Muma, M.H. 1951. The arachnid order Solpugida in the United States. Bulletin of the American Museum of Natural History XCVII(2):35-141. Muma, M.H. 1980. Comparison of three methods for estimating solpugid (Arachnida) populations. Journal of Arachnology 8:267-270. Nimer, E. 1979. Climatologia do Brasil. Rio de Ja- neiro, IBGE, 421 pp. Punzo, E 1997. Dispersion, temporal patterns of activity and phenology of feeding and mating be- havior in Eremobates palpisetulosus (Solifugae, Eremobatidae). Bulletin of the British Arachnol- ogical Society 10:303-307. Punzo, E 1998. The Biology Of Camel Spiders (Arachnida, Solifugae). Kluwer Academic Pub- lishers. Rocha, L.S. & E.M. Cancello. 1997. A ordem So- lifugae na America do Sul: novas perspectivas para o seu conhecimento. R 50. In Actas del Primer Encuentro de Aracnologos del Cono Sur. 134 THE JOURNAL OF ARACHNOLOGY (M. Simo, ed.) 1 a 5 Diciembre 1997. Facultad de Ciencias/ 1. NA. PE. Montevideo, Uruguay. Rocha, RL.B. 1991. Ecologia e Morfologia de Uma Nova Especie de Proechimys (Rodentia: Echimidae) das Dunas Interiores do Rio Sao Francisco, Bahia. M.S. thesis. University of Sao Paulo, Sao Paulo, Brazil. Rocha, RL.B. 1998. Uso e Partigao de Recursos Pelas Especies de Lagartos Das Dunas do Rio Sao Francisco, Bahia (Squamata). Ph.D. thesis, University of Sao Paulo, Sao Paulo, Brazil. Roewer, C.E 1934. Solifugae, Palpigradi. Pp. 461- 608, In Klassen und Ordungen des Tierreichs. (H.G. Bronns, ed.). vol 5, div. 4, book 4. Leipzig. Roewer, C.E 1954. Einige neue Opiliones Lan- iatores und Solifugae. Abhandlungen Natur- wissenschaftlichen Verein Zu Bremen 33(3): 377-384. Savory, T.H. 1964. The order Solifugae. Pp. 184- 193. In Arachnida. (W. & J. Mackay, eds.). Ac- ademic Press, London. Wharton, R.A. 1987. Biology of the diurnal Me- tasolpuga picta (Kraepelin) (Solifugae, Solpugi- dae) compared with that of nocturnal species. Journal of Arachnology 14:363-383. Xavier, E. & RL.B. Rocha. 1998. Microhabitat use by an assemblage of cursorial arachnids in sand dunes from Brazilian semiarid caatinga. XIV In- ternational Congress of Arachnology American Arachnological Society (abstract). Manuscript received 29 December 1999, revised 10 October 2000. 2001. The Journal of Arachnology 29:135-140 AN UNUSUAL NEW SPECIES OF MUNDOCHTHONIUS FROM A CAVE IN COLORADO, WITH COMMENTS ON MUNDOCHTHONIUS MONTANUS (PSEUDOSCORPIONES, CHTHONIIDAE) William B. Muchmore: Department of Biology, University of Rochester, Box 270211, Rochester, New York 14627-0211 USA ABSTRACT. Mundochthonius singularis, a troglomorphic species from Fly Cave in Fremont County, Colorado, is described. This is the first cavemicolous pseudoscorpion to be reported from the state. It is compared with M. montanus, the local epigean species, for which an emended description is given. Keywords: Pseudoscorpiones, Chthoniidae, Mundochthonius, cavernicole, Colorado The genus Mundochthonius Chamberlin 1929 is Holarctic in distribution, with two species known from subtropical areas, in Mexico and Hispaniola (see references in Har- vey 1991). Its member species are mostly very small, litter-dwelling creatures; and some have been found in caves. In the United States eight species have been described, two of which are cavemicolous, namely, Mundochthonius cav- ernicola Muchmore 1968 from Illinois and M. holsingeri Benedict & Malcolm 1974 from Virginia. Recently, another cavemicolous spe- cies has been discovered in Colorado, this one more highly troglomorphic than the other two; it is described below. But first, I take this op- portunity to redescribe Mundochthonius mon- tanus Chamberlin 1929, the surface-dwelling species in the area, for comparison with the new cave-dwelling form. The specimens studied here have been dis- sected, cleared, and mounted in Canada bal- sam on microscope slides. They are in the fol- lowing depositories: California Academy of Sciences, San Francisco, California (CAS); Florida State Collection of Arthropods, Gainesville, Florida (FSCA). Some abbrevia- tions are used in the descriptions: L = length; L/B ^ ratio, length/breadth; L/D = ratio, length/depth; m = microseta; T = tactile seta. Genus Mundochthonius Chamberlin Mundochthonius Chamberlin 1929: 64; Beier 1932: 36; Hoff 1949: 436; Hoff 1956: 10; Morikawa 1960: 94; Beier 1963: 18; Muchmore 1973: 48; Benedict 1978: 250; Harvey 1991: 190. Though not reported here in detail, the ho- lotype male of Mundochthonius erosidens Chamberlin 1929, type species of Mundo- chthonius, has been examined and found to support the following diagnosis. Diagnosis. — Mundochthonius is easily di- agnosed. It shares the following characters with several other chthoniid genera, namely, Austrochthonius Chamberlin 1929, Congo- chthonius Beier 1959, Francochthonius Vitali- di Castri 1976, Malcolmochthonius Benedict 1978, and Mexichthonius Muchmore 1975: 1) coxal spines present only on coxae II; 2) con- tiguous teeth on fingers of palpal chela; 3) tri- chobothrium sb (usually much) closer to st than to b on movable chelal finger; 4) epi- stome prominent, serrate; 5) one or more mi- crosetae on anteromedial process (apex) of coxa I, However, it may be distinguished from all of these by the possession of a bisetose intercoxal tubercle between the bases of coxae III and IV. Remarks. — Though not mentioned by Chamberlin (1929), the holotype of Mundo- chthonius erosidens has a small, but distinct, heavily serrate, triangular epistome at the mid- dle of the anterior margin of the carapace. All other species in the genus appear to have a similar, but usually larger, epistome. American species of Mundochthonius (in- cluding M. erosidens) appear to have only two weak eyes or none at all. On the other hand. 135 136 THE JOURNAL OF ARACHNOLOGY one species in Europe, M. alpinus Beier 1947 (and 1963), is described as having four eyes, though other species in Europe and Asia are reported to have only two eyes or none (see references in Harvey 1991). Contrary to the statement of Chamberlin in his “Analytical Key to the Genera of the Ke- wochthonini” (1929: 63), all American spe- cies of Mundochthonius that I have examined (including M. erosidens) possess 1-3 micro- setae on the anteromedial process (apex) of coxa 1. Species from elsewhere are also re- ported to possess these small setae. The coxal spines of the American species of Mundochthonius that I have examined ap- pear as flattened blades, variously indented or incised at the distal ends and along the sides, as described and figured by Chamberlin (1929, 1931), by Hoff (1949, 1952), and by Muchmore (1973). None looks like the cone- shaped, setaceous or spiny structures de- scribed and illustrated for M. alpinus Beier (1947, 1963), and for M. decoui Dumitresco & Orghidan (1970), or those of M. carpaticus Rafalski (1948); nor do they appear as thick and lobe-like as those of M. basarukini Scha- waller (1989). Because of the minute size of the coxal spines in most species of Mundo- chthonius, satisfactory descriptions of their structure will probably be achieved only by following Schawaller (1989) in the use of scanning electron microscopy. In any event, the “coxal spines” of the new species, M. sin- gularis, are unique in being very deeply dis- sected and elongated to resemble, somewhat, the antlers of a deer. Mundochthonius montanus Chamberlin Eigs. 1, 2, 8 Mundochthonius montanus Chamberlin 1929: 65; Chamberlin 1931: fig. 211; Hoff 1952: 40, figs. 1-4; Hoff 1956: 10; Hoff 1959: 26, 33, etc.; Hoff 1961: 420; Harvey 1991: 191. Type data. — Holotype female (JC- 86.01001) from “Manitou [El Paso County] - Colorado. Elev. 8500. In - soil (surface). Coll. E.W. Goldsmith.” (mounted on microscope slide by J.C. Chamberlin; in CAS, Type No. 17445). The original description by Chamberlin (1929) was very brief. Hoff (1952, 1956, 1961) added some observations and measure- ments based on specimens from New Mexico and Colorado, but he did not record several important details. It seems wise to redescribe the holotype in relation to Hoff’s material, in order to firmly establish the species. The ho- lotype has been dissected, the body stained pink, and mounted in Canada balsam; the right palp is missing and the left palpal seg- ments have been somewhat compressed and broken by the cover; many vestitural setae are too faint to see clearly. Redescription of holotype female. — Rep- resentative of the genus as outlined above and with the following particular features. Cara- pace a little longer than broad; epistome small, triangular, serrate; two very small eyes; chaetotaxy 6-4-4-2-2. Coxal area typical, but no microsetae observable on apex of coxa I; coxal spines as shown by Chamberlin (1931: fig. 211); bisetose intercoxal tubercle present. Tergal chaetotaxy 4:4:6:6:6:6:-(others not ob- servable). Sternal chaetotaxy 10:(3)8(3):-(oth- ers not observable). Chelicera about 0.85 as long as carapace; flagellum of nine setae; spinneret a small knob on finger margin; setae on hand not observable, but Chamberlin re- ported six (1929: 64). Palp rather robust (see Eig. 1). Because of damage, the palpal seg- ments are not measurable with accuracy; Chamberlin reported L/B of femur and chela as 4 and 3.8, respectively. Measurements and ratios for some other Colorado specimens are given below. Trichobothriotaxy typical (see Eig. 2). Chelal fingers with numerous but not countable, small, contiguous teeth. Leg IV rather robust (see Eig. 8): L/D of fe- mur+patella 2.45, tibia 3.15. Measurements (mm). — (These given are deemed reliable in spite of some distortion of the body parts). Body L 1.00. Carapace L 0.33. Chelicera L 0.295. Palp: femur 0.31/?; patella 0.155/7; chela 0.495/7; hand 0.185/7; movable finger L 0.34. Leg IV: femur+patella 0.26/0.115; tibia 0.19/0.06. Variation. — As discussed by Hoff (1961), there is considerable variation among the specimens assigned to Mundochthonius mon- tanus. According to the values given by Hoff (1952, 1961), the measurements for M. mon- tanus in New Mexico and Colorado range as follows (females average slightly larger than males): Body L 0.92-1.25. Carapace L 0.32- 0.40. Palpal femur 0.263-0.350/0.068-0.094; patella (New Mexico only) 0.155-0.186/ 0.086-0.105; chela 0.420-0.545/0.100-0.127; hand 0.155-0.211/0.100-0.126; movable fin- MUCHMORE— A NEW CAVE MUNDOCHTHONIUS FROM COLORADO 137 Figures 1~4. — Species of M undoc hthonius, palps. Figs. 1, 2. Mundochthonius montanus Chamberlin, female from Colorado. 1. Right palp, dorsal view; 2. Left chela, lateral view, setae omitted, with teeth from fixed finger. Figs. 3, 4, Mundochthonius singularis new species, holotype female. 3. Right palp, dorsal view; 4. Left chela, lateral view, setae omitted, with teeth from fixed finger. Scale = 0.25 mm. ger L 0.270“0.355. Ratios of palpal segments: L/B of femur 3.3-3.9, patella 1.65-1.85, and chela 3.85-4.5; L/D of hand 1.4-1.65; mov- able finger 1.7-1.85X as long as hand. All the reliable measurements and ratios of the holo- type fall well within these ranges. No microsetae (m) are visible on antero- medial process of coxa I of the holotype, probably because of long exposure to clearing agent. However, two or three such setae do appear on coxa I of all other specimens of M. montanus I have examined from Colorado and New Mexico. The coxal spines are quite varied in shape, from a single broad, incised blade, as figured by Chamberlin for the holotype (1931: fig. 211), to two or three separate, narrower, in- cised or dentate blades (Hoff 1952: figs. 1-4; 1961: 421). There are about 50 marginal teeth on each finger of the palpal chela. They are contiguous basally, uniform in height, and slightly retro- dentate or rounded (Fig. 2). Remarks. — Mundochthonius montanus is a generalized, epigean species, apparently com- mon in the Rocky Mountains of Colorado and New Mexico. According to Hoff (1959, 1961), it is adapted to life in a wide variety of litter and decomposing wood of logs and stumps, at elevations of 6900-11,000 feet (2100-3350 m). Presumably, it represents the population from which the specialized, cav- ernicolous M. singularis was derived. Mundochthonius singularis new species Figs. 3-7 Type data. — Holotype female (WM8097. 01001) from under a rock, about 30 m inside Fly Cave, 20 km N of Canon City, Fremont County, Colorado (see Parris 1973: 89-90), 7 August 1996, P. Beron (slide, in FSCA). Diagnosis. — Differs from other species of 138 THE JOURNAL OF ARACHNOLOGY Figures 5-8. — Species of Mundochthonius, various features. Figures 5-7. Mundochthonius singularis new species, holotype female. 5. Right coxa I (ventral view), showing microsetae on apex; 6. Coxal spines on coxae II (ventral view); 7. Leg IV, setae omitted. 8. Mundochthonius montanus Chamberlin, female from Colorado: Leg IV, setae omitted. the genus in its troglomorphic adaptations: large size (palpal chela 1.07 mm long), atten- uated appendages (L/B of palpal chela = 5.8), and an unique arrangement of coxal spines; also, the palpal chelae are heterodentate, rath- er than homodentate as seen in other species of the genus. Description. — With the characters of the genus as outlined above, and the following particular features. Chelicerae and palps light brown, carapace tan, other parts lighter. Car- apace about as long as broad, narrowed pos- teriorly; epistome large, irregularly serrate; no eyes; chaetotaxy 6-4-4-2-2. Coxal area gen- erally typical; chaetotaxy 2-2-l:2or3m-2-2(l): 2-4(3)-CS:2-5(4):2-3; each coxa I with 2-3 microsetae (m) on medial edge of apex (Fig. 5) ; each coxa II with an unusual complex of coxal spines (CS), consisting of several dif- ferently shaped elements, one on each side somewhat resembling an antler of a deer (Fig. 6) ; a bisetose intercoxal tubercle present. Ter- gal chaetotaxy 4:4:4:6:6:6:6:6:6:6: 1T2T1 :0; sternal chaetotaxy 10:(4)6(4):(2)6(2):1 1:10:9: 10:8:7:0:2. Chelicera 0.85 as long as carapace; hand with 6 setae; flagellum of 10 setae; galea a very small elevation; each finger with 10- 15 small teeth. Palp long and slender (Fig. 3); femur 1.45X and chela 2.05 X as long as car- apace. L/B of trochanter 1.75, femur 6.25, patella 2.35, and chela 5.8; L/D of hand 2.1; movable finger 1.8X as long as hand. Surfaces smooth. Trichobothria as shown in Fig. 4. Fixed finger with about 95 and movable finger with about 85 contiguous, cusped teeth; on each finger, about 12 teeth are conspicuously larger and sharper than adjacent ones (Fig. 4). Legs long and slender: leg I with femur 2.05 X as long as patella; leg IV (Fig. 7) with L/D of femur + patella 3.8 and tibia 4.8. Measurements (mm). — Holotype female (male unknown). Body L 1.59. Carapace L 0.52. Chelicera L 0.45. Palp: trochanter 0.22/ 0.125; femur 0.75/0.12; patella 0.355/0.15; chela 1.07/0.185; hand 0.385/0.185; movable finger L 0.69. Leg I: femur 0.38/0.06; patella 0.185/0.06. Leg IV: femur+patella 0.53/0.14; MUCHMORE— A NEW CAVE MUNDOCHTHONIUS FROM COLORADO 139 tibia 0.385/0.08; basitarsus 0.18/0.06; telotar- sus 0.355/0.045. Etymology. — The species is called singu- laris (Latin, different) for its unusual charac- ters, especially the coxal spines, compared with other members of the genus. Remarks. — Mundochthonius singularis is the first cavernicolous pseudoscorpion record- ed from Colorado. It is certainly troglobitic, being much more highly modified for cave life than any other known species of the genus. It is the largest known species of Mundochthon- ius, with lengths of palpal femur and chela 0.75 and 1.07 mm respectively, compared to the cavernicolous M. cavernicola (0.57 and 0.92 mm) and the epigean M. montanus (0.26-0.35 and 0.44-0.54 mm); and it has very slender appendages, with L/B of palpal femur and chela 6.25 and 5.8, compared to M. cavernicola (4.4 and 4.65) and M. montanus (3. 5-4.0 and 3.8-4.35). Correlated with the elongated chela, trichobothrium sb on the movable finger is much farther removed from St than it is in other species of the genus; but it is still closer to st than to b. The coxal spines are uniquely deeply in- cised and enlarged, apparently an adaptation to some aspect of life in a cave. The coxal spines of the other known cavernicolous spe- cies, M. cavernicola and M. holsingeri, are not so modified. The function of the coxal spines of chthonioid pseudoscorpions is not known with certainty, but Chamberlin (1931: 93) considered it “most probable that they are sensory, perhaps tactile.” Alternatively, Jud- son (1990) reported observations of the coxal spines of three species of chthonioids from West Africa being used in cleaning the legs. Whether one or both of these functions is op- erative in Mundochthonius is still unknown. In Mundochthonius montanus and other species in the genus, the palpal chelae may be characterized as “homodentate” — adjacent teeth of the marginal rows are very much alike, and any differences in size and shape tend to be gradual along the row. In M. sin- gularis, however, there are frequent teeth which are distinctly broader and taller than ad- jacent teeth. It may be supposed that this un- usual dentition is correlated with a diet in the cave different from that of the other species. Fly Cave is named for the many flies en- countered just inside the entrance (Pollard 1954; Parris 1973). According to Ayre (1961a, b), the flies are Neomuscina tripunc- tata (Van der Welp), “a breed common in Mexico and rang[ing] into this portion of the state of Colorado.” Also, Fly Cave is the type locality of the unusual spider Hypochilus bon- neti Gertsch 1964 (and see Vogel & Ayre 1961). Neither the fly nor the spider is trog- lobitic. ACKNOWLEDGMENTS I am much indebted to Petar Beron for col- lecting the type specimen of Mundochthonius singularis, and to David A. Hubbard, Jr. for sending it to me for study. Both David Hub- bard and Donald G. Davis supplied invaluable information about Fly Cave and its fauna. Charles E. Griswold kindly lent the holotypes of M. montanus and M. erosidens from the California Academy of Sciences. I am grateful to V. Mahnert, an anonymous reviewer, and the editors for helpful comments on the man- uscript. LITERATURE CITED Ayre, R.W. 1961a. Bats and bugs. Caving in the Rockies, Colorado Grotto, National Speleologi- cal Society 7:1-3. Ayre, R.W. 1961b. From the biology notebook. Caving in the Rockies, Colorado Grotto, Nation- al Speleological Society 7:13. Beier, M. 1932. Pseudoscorpionidea I. Subord. Chthoniinea et Neobisiinea. Das Tierreich 57:1- 258. Beier, M. 1947. Neue Pseudoscorpione aus der Steiermark. Annalen des Naturhistorischen Mu- seum in Wien 55:296-301. Beier, M. 1959. Pseudoscorpione aus dem Bel- gischen Congo gesammelt von Heirn N. Leleup. Annales du Musee Royal du Congo Beige, Ter- vuren. Sciences Zoologiques 72:5-69. Beier, M. 1963. Ordnung Pseudoscorpionidea (Af- terskorpione). In Bestimmungsbiicher zur Bod- enfauna Europas. Akademie-Verlag, Berlin 1:1- 313. Beier, M. 1971. Ein neuer Mundochthonius (Arachnida, Pseudoscorpionidea) aus der Steier- mark. Mitteilungen des Naturwissenschaftlichen Vereins fiir Steiermark 100:386-387. Benedict, E.M. 1978. A new pseudoscorpion genus Malcolmochthonius n. g., with three new species from the western United States. Transactions of the American Microscopical Society 97:250- 255. Benedict, E.M. & D.R. Malcolm. 1974. A new cavernicolous species of Mundochthonius from the eastern United States (Pseudoscorpionida, Chthoniidae). Journal of Arachnology 2:1-4. 140 THE JOURNAL OF ARACHNOLOGY Chamberlin, J.C. 1929. A synoptic classification of the false scorpions or chela-spinners, with a re- port on a cosmopolitan collection of the same. Part 1. The Heterosphyronida (Chthoniidae) (Arachnida-Chelonethida). Annals and Magazine of Natural History 10(4): 50-80, Chamberlin, J.C, 1931. The arachnid order Che- lonethida. Stanford University Publications, Bi- ological Sciences 7(1): 1-284. Dumitresco, M, & T. Orghidan. 1970. Contribution a la connaissance des Pseudoscorpions souter- rains de Roumanie. Travaux de P Institute de Speologie “Emile Racovitza” 9:97-111. Gertsch, W.J. 1964. A review of the genus Hypo- chilus and a description of a new species from Colorado (Araneae, Hypochilidae). American Museum Novitates 2203:1-14. Harvey, M.S. 1991. Catalogue of the Pseudoscor- pionida. Manchester University Press, Manches- ter, England. Hoff, C.C. 1949. The pseudoscorpions of Illinois, Bulletin of the Illinois Natural History Survey 24:407-498. Hoff, C.C. 1952. Some heterosphyronid pseudo- scorpions from New Mexico. The Great Basin Naturalist 12:39-45. Hoff, C.C. 1956. The heterosphyronid pseudoscor- pions of New Mexico. American Museum Nov- itates 1772:1-13. Hoff, C.C. 1959. The ecology and distribution of the pseudoscorpioes of north-central New Mex- ico. University of New Mexico Publications in Biology 8:1-68. Hoff, C.C, 1961, Pseudoscorpions from Colorado. Bulletin of the American Museum of Natural History 122:409-464. Judson, M.L.I. 1990. Observations on the form and function of the coxal spines of some chthonioid pseudoscorpions from Cameroon (Arachnida, Chelonethida), Acta Zoologica Fennica 190:195- 198. Morikawa, K. 1960. Systematic studies of Japa- nese pseudoscorpions. Memoirs of the Ehime University (2B) 4:85-172. Muchmore, W.B. 1968, A cavemicolous species of the pseudoscorpion genus Mundochthonius (Arachnida, Chelonethida, Chthoniidae). Trans- actions of the American Microscopical Society 87:110-112. Muchmore, W.B. 1973. New and little known pseudoscorpions, mainly from caves in Mexico (Arachnida, Pseudoscorpionida). Association for Mexican Cave Studies Bulletin 5:47-62. Muchmore, W.B. 1975. A new genus and species of chthoniid pseudoscorpion from Mexico (Pseu- doscorpionida, Chthoniidae). Journal of Arach- nology 3:1-4. Muchmore, W.B, 1996. A new Mundochthonius from the Dominican Republic (Pseudoscorpion- ida: Chthoniidae). Insecta Mundi 10:104. Parris, L.E. 1973. Caves of Colorado. Pruett Pub- lishing, Boulder, Colorado. Pollard, G. 1954. Fly Cave. News and notes, Col- orado Grotto, National Speleological Society 3(4): 1-5. Rafalski, J. 1948. Mundochthonius carpaticus sp, nov,, nowy gatunek zaleszczotka (Pseudoscor- pionidea). [Mundochthonius carpaticus, a new species of Pseudoscorpionidea.] Annales Musei Zoologici Polonici 14(3): 13-20. (in Polish, with English summary). Schawaller, W. 1989. Pseudoskorpione aus der So- wjetunion, Teil 3 (Arachnida: Pseudoscorpiones). Stuttgarter Beitrage zur Naturkunde (A) 440:1- 30. Vitali-di Castri, V. 1976 [1975]. Deux nouveaux genres de Chthoniidae du Chili: Chiliochthonius et Francochthonius (Arachnida, Pseudoscorpion- ida). Bulletin du Museum Nationale d'Histoire Naturelle, Paris (3) 334:1277-1291. Vogel, B. & R.W Ayre. 1961, New species of rare spider found in Colorado cave. NSS News 19: 156. Manuscript received 1 October 2000, revised 6 February 200 L 2001. The Journal of Arachnology 29:141-145 SYNONYMY OF CECODITHA (CECODITHINAE) WITH AUSTROCHTHONIUS (CHTHONIINAE) (CHELONETHI, CHTHONIIDAE) Mark L.I. Judson; Museum National d’Histoire Naturelle, Laboratoire de Zoologie (Arthropodes), 61 Rue de Buffon, 75005 Paris, France ABSTRACT. The holotype of Cecoditha parva Mello-Leitao 1939, from Chubut Province, Argentina, is redescribed and shown to be a typical member of the genus Austrochthonius Chamberlin 1929. The monotypic genus Cecoditha Mello-Leitao 1939 is therefore a junior subjective synonym of Austrochthon- ius, while the subfamily Cecodithinae Chamberlin & Chamberlin 1945 (Tridenchthoniidae) is a junior subjective synonym of Chthoniinae Daday 1888 (Chthoniidae). RESUMEN. El holotipo de Cecoditha parva Mello-Leitao 1939, de la provincia de Chubut, Argentina, es redecristo y se demuestra que es un miembro tipico del genero Austrochthonius Chamberlin 1929. El genero monotipico Cecoditha Mello-Leitao 1939 es por lo tanto un sinonimo subjetivo posterior dQ Aus- trochthonius, mientras que la subfamilia Cecodithinae Chamberlin & Chamberlin 1945 (Tridenchthoniidae) es un sinonimo subjetivo posterior de Chthoniinae Daday 1888 (Chthoniidae). Keywords: Pseudoscorpion, taxonomy, Argentina In 1939, Mello-Leitao described a seeming- ly unusual new genus and species of false- scorpion, Cecoditha parva Mello-Leitao, from southern Argentina. Mello-Leitao assigned Cecoditha to the subfamily Dithinae Cham- berlin (now Tridenchthoniidae Balzan), but he did not give his reasons for doing so or dis- cuss its relationships with other genera. Impressed by the incongruity of its char- acters, Chamberlin & Chamberlin (1945: 14) created a new subfamily, Cecodithinae, for C. parva, writing “This species, if correctly de- scribed, is unique in possessing a simple galea (“Galea sencilla”) in the adult stage. This fea- ture, together with the fact that the species also reportedly lacks coxal spines and had the tactile setae IB and ISB placed sub-medially instead of sub-basally on the dorsum of the hand of the chela, sets the species widely apart from all other known Tridenchthoniidae and necessitates its segregation in a separate sub- family.” It is curious that Chamberlin & Chamberlin did not question the assignment of C. parva to the Tridenchthoniidae, despite their doubts about the accuracy of the original description. They might have been more cau- tious had they known that Mello-Leitao was capable of prodigious errors of classification (e.g., see Krantz & Platnick 1995). With the lack of subsequent records of Ce- codithinae, the systematic position of C. parva became increasingly doubtful. In the hope of re- solving the matter, I asked to borrow the unique type of this species from the Museo de La Plata in 1982, but was informed at that time that it could not be found. Fortunately, the specimen was rediscovered in the collection years later by Lie. R.F. Arrozpide, who kindly sent it for study. Although the holotype is in poor condition, it is clearly a member of the chthoniid genus Aus- trochthonius Chamberlin. Family Chthoniidae Daday Subfamily Chthoniinae Daday Chthoniinae Daday 1888: 133. Cecodithiinae [lapsus for Cecodithinae] Chamberlin & Chamberlin 1945: 65; Harvey 1991: 217. NEW SYNONYMY. Cecodithinae Chamberlin & Chamberlin - Harvey 1992: 1401. Austrochthonius Chamberlin Austrochthonius Chamberlin 1929: 68 (type species Chthonius chilensis Chamberlin 1923, by original designation); Beier 1932: 38; Vitali-di Castri 1968: 144-145; Harvey 1991: 139. Paraustrochthonius Beier 1931: 52 (type species Paraustrochthonius tullgreni Beier 1931, by orig- inal designation); Beier 1932: 40; Vitali-di Castri 141 142 THE JOURNAL OF ARACHNOLOGY Figures 1-5. — Austrochthonius parvus (Mello-Leitao), holotype male. 1. Carapace (reconstructed; dots represent gland pores); 2. Genital region; 3. Left chelicera, with detail showing spinneret; 4. Coxal spines of right coxa II; 5. Right leg IV (reticulation only shown in part). Abbreviation: pg = pore group. Divisions of scale lines = 0.1 mm (Figs. 1-3, 5) or 0.01 mm (Fig. 4). 1968: 144-145. Synonymized by Beier 1976: 203. Cecoditha Mello-Leitao 1939: 115-116 (type spe- cies Cecoditha parva Mello-Leitao 1939, by orig- inal designation); Chamberlin & Chamberlin 1945: 65-66; Harvey 1991: 218; Kury & No- gueira 1999: 13. NEW SYNONYMY. Austrochthonius parvus (Mello-Leitao) NEW COMBINATION Figs. 1-9 Cecoditha parva Mello-Leitao 1939: 116-117, figs, la, b; Chamberlin & Chamberlin 1945: 66-67, fig. 17; Harvey 1991: 218; Kury & Nogueira 1999: 13. Material examined. — Holotype 6, ''Cecoditha parva M.L./Mel.-Leit. det.; [Puerto] Madryn, Chu- but, 18.11.1938, [M.] Bir[aben]” (Museo de La Pla- ta, Universidad Nacional de La Plata). Specimen in poor condition — carapace broken into three pieces; right chelicera, both legs I (including coxae) and left leg IV lost — and strongly darkened as a result of storage in corked tube. Diagnosis. — Moderately large species (e.g., movable chela finger length 0.43 mm), male with well developed spinneret tubercle, ter- gites LIV with 4 setae, chelal teeth contigu- ous. Description. — Carapace (Fig. 1) with weak reticulation, laterally hispid (especially ante- riorly); anterior margin strongly serrate me- dially, but without a pronounced epistome; an- terior eyes with strong lens, roughly one ocular diameter from anterior margin, poste- JUDSON— SYNONYMY OF CECODITHA 143 Figures 6-9. — Austrochthonius parvus (Mello-Leitao), holotype male, palps. 6. Trochanter, femur and patella of right palp (reticulation only shown in part). 7. Base of fixed finger of left chela, showing position of ist, 8. Right chela, lateral, with details of dentition and sensilla; 9. Right chela, dorsal. Abbreviation: sa apical sensilla. Scale line = 0.1 mm (0.05 mm for details). rior eyes reduced to weak spots; setae 6:4:4: 2:2 (18); scattered pores present. Tergites re- ticulate, setae 4:4:4:4:6:6:6:6:6:4:1T2T1:0; each tergite with a row of pores just in front of setal row. Coxal setae P 5 (2 on mandu- catory process, subapical long);I ?:II 4:III 5: IV 5; coxa II with 7-8 bipinnate spines, bases almost contiguous (Fig, 4); intercoxal tubercle absent. Anterior genital operculum (Fig. 2) with 12 setae; posterior operculum (Fig. 2) with (3m)6(3m), plus 8-9 setae on each side of notch (total (3m)17(3m)); remaining ster- nites 11:10:8:8:8:8:7:0:2, stemite X with a small, unpaired, median seta; anterior genital 144 THE JOURNAL OF ARACHNOLOGY sternite with three pores grouped on each side {pg. Fig. 2), other sternites with a normal row of pores in front of setal row; stigmata normal, not situated on a separate sclerite; pleural membrane papillate. Genitalia (Fig. 2) typical, with 4 pairs of glandular setae, ace elongate (see Vitali-di Castri 1976). Lateral and median genital sacs not seen. Chelicera (Fig. 3) with hand and base of movable finger scaly-retic- ulate; hand with 6 setae; fixed finger with 10 teeth, moveable finger with 13; flagellum of 11 blades, basal blade short; serrula exterior and serrula interior with about 13 blades; spin- neret in the form of a distinct tubercle. Palp (Figs. 6-9) with femoral setae 3:5:1:2:5:1; pa- tella with ten setae; hand with 3 proximal, 7 medial and 4 distal setae; fixed finger with a single strong seta at base; dorsum of hand moderately hispid distad of ibiisb and slightly depressed behind; fixed finger with 47, mov- able finger with 37 contiguous teeth, those of movable finger generally weaker than those of fixed finger, becoming obsolete proximally; apical sensilla (Fig. 8: sa) small and close to- gether, near tip; proximal sensilla near dental margin, about Vs way from b to sb\ tricho- bothria as illustrated, position of ist variable (compare Figs. 7 and 9). Leg IV (Fig. 5) with scaly reticulation on all segments; setae (tro- chanter to basitarsus) 2:3:7:9:9, basitarsal TS 0.31, telotarsal TS 0.30. Measurements (in mm; ratios in parentheses): carapace (estimat- ed) 0.43 X 0.39; palp femur 0.45 X 0.10 (4.5), tibia 0.20 X 0.11 (1.9), hand 0.26 X 0.13 (2.0), chela 0.67 (5.3), movable finger 0.43 (finger/hand 1.7); leg IV femur 0.20 X 0.17 (1.2), patella 0.26 X 0.15 (1.7), femur+patella 0.40 (2.4), tibia 0.30 X 0.07 (4.2), basitarsus 0.14 X 0.055 (2.5), telotarsus 0.27 X 0.03 (8.6). Remarks. — Although there was no regis- tration number or indication of type status with the specimen, the locality details and the identification leave no doubt that this is the holotype. Kury & Nogueira’s (1999) assump- tion that “syntypes” existed is an error, since Mello-Leitao (1939) only mentioned the male “Tipo.” The original description contains many errors, the most important of which are the statements that A. parvus lacks eyes and coxal spines, and has only six blades in the flagellum. The measurements of the palp are also slightly lower than those given here. This species can be separated from the oth- er members of Austrochthonius by the com- bination of characters given in the diagnosis; the well developed spinneret tubercle of the male is particularly distinctive. The presence of six setae on the hand of the single remain- ing chelicera is also unusual, but more mate- rial is required to determine whether this is anomalous. Vitali-di Castri (1968) found six setae on one chelicera of A. insularis Vitali- di Castri, the normal number being five. ACKNOWLEDGMENTS I am very grateful to Ricardo F. Arrozpide (formerly of Museo de La Plata) for finding the holotype of Cecoditha parva and making it available for study. Helpful comments on the manuscript were provided by Alejandra Ceballos (Univ. Nacional de Cordoba), Mark Harvey (Western Australian Museum) and an anonymous referee. Juan A. Zaragoza Miral- les (Univ. Alicante) kindly translated the ab- stract into Spanish. LITERATURE CITED Beier, M. 1931. Zur Kenntnis der Chthoniiden (Pseudoskorpione). Zoologischer Anzeiger 93: 49-56. Beier, M. 1932. Pseudoscorpionidea I. Subord. Chthoniinea et Neobisiinea. Das Tierreich 57:1- 258. Beier, M. 1976. The pseudoscorpions of New Zea- land, Norfolk and Lord Howe. New Zealand Journal of Zoology 3:199-246. Chamberlin, J.C. 1929. A synoptic classification of the false scorpions or chela-spinners, with a re- port on a cosmopolitan collection of the same. - Part 1. The Heterosphyronida (Chthoniidae) (Arachnida - Chelonethida). Annals and Maga- zine of Natural History (10) 4:50-80. Chamberlin, J.C. & R.V. Chamberlin. 1945. The genera and species of the Tridenchthoniidae (Dithidae), a family of the arachnid order Che- lonethida. Bulletin of the University of Utah 35 (23) (Biol. Ser. 9 (2)): 1-67. Daday, E. 1888. A Magyar Nemzeti Muzeum al- skorpioinak attekintese. Termeszetrajzi Fiizetek 11:111-136, 165-192. Harvey, M.S. 1991. Catalogue Of The Pseudos- corpionida. 726 p. Manchester Univ. Press, Man- chester. Harvey, M.S. 1992. The phylogeny and classifi- cation of the Pseudoscorpionida (Chelicerata: Arachnida). Invertebrate Taxonomy 6:1373- 1435. Krantz, G.W. & N.I. Platnick. 1995. On Bruchar- achne, the spider that wasn’t (Arachnida, Acari, JUDSON— SYNONYMY OF CECODITHA 145 Dermanyssoidea). American Museum Novitates 3151:1-8. Kury, A. & A.L.C. Nogueira. 1999. Annotated check list of type specimens of Arachnida in the Museu Nacional - Rio de Janeiro. I. Scorpiones, Pseudoscorpiones and Solifugae. Publicagoes Avulsas do Museu Nacional, Rio de Janeiro 77: 3-19. Mello-Leitao, C. de. 1939. Pseudoscorpionidos de Argentina, Notas del Museo de La Plata 4 (Zool. 17):115-122. Vitali-di Castri, V. 1968. Austrochthonius insular- is, nouvelle espece de pseudoscorpions de PArchipel de Crozet (Heterosphyronida, Chthon- iidae). Bulletin du Museum National d’Histoire Naturelle, Paris (2) 40:141-148. Vitali-di Castri, V. 1976 (1975). Deux nouveaux genres de Chthoniidae du Chili: Chiliochthonius et Francochthonius (Arachnida, Pseudoscorpion- ida). Bulletin du Museum National d’Histoire Naturelle, Paris (3) 334:1277-1291. Manuscript received 1 December 2000, revised 10 April 2001. 2001. The Journal of Arachnology 29:146-172 TWO NEW SPECIES OF HADOGENES (SCORPIONES, ISCHNURIDAE) FROM SOUTH AFRICA, WITH A REDESCRIPTION OF HADOGENES BICOLOR AND A DISCUSSION ON THE PHYLOGENETIC POSITION OF HADOGENES Lorenzo Prendini: Percy FitzPatrick Institute, University of Cape Town, Rondebosch 7700, South Africa ABSTRACT. The taxonomic status of the endemic South African flat rock scorpion, Hadogenes bicolor Purcell 1899, is reassessed, based on a study of the types and a large series of newly-collected specimens. Specimens identified as H. bicolor by previous authors can be separated into at least three species on the basis of morphology, each of which occupies a discrete, allopatric distributional range. In light of this new evidence, H. bicolor is redescribed and two new species, Hadogenes longimanus and Hadogenes newlandsi, are described. A key is provided for the identification of the three allopatric species, and their ecology and conservation status are discussed. The phylogenetic position of Hadogenes is discussed in light of a recent cladistic analysis, and the monotypic family Hadogenidae Lourengo 2000 is synonymized with the family Ischnuridae Simon 1879. Keywords: Scorpiones, Ischnuridae, Hadogenes Scorpions of the genus Hadogenes Krae- pelin 1894, commonly known as flat rock scorpions, are endemic to the Afrotropical re- gion, where they are distributed from South Africa to Tanzania. Comprising 15 species, Hadogenes is the second most speciose genus in the family Ischnuridae Simon 1879, after Opisthacanthus Peters 1861. With few exceptions, the distributional ranges of Hadogenes species are allopatric or parapatric (Newlands 1980; Prendini 1995), a tendency that appears to be related to the stenotopic ecological requirements of these scorpions. All of the currently recognized spe- cies are obligately lithophilous, inhabiting the narrow cracks, crevices and exfoliations of weathered rock outcrops. Ecomorphological adaptations that facilitate existence in this spe- cialized habitat include extreme dorsoventral compression, elongation of the metasoma and pedipalps, greatly enlarged lateral ocelli rela- tive to the median ocelli, presumably to aid in anterior light perception, and well-developed superciliary carinae to protect the median ocelli from abrasion (Newlands 1972a, b, 1978; Newlands & Prendini 1997). Species of Hadogenes are also characterized by stout, spiniform setae on the ventral surfaces of the telotarsi and highly-curved telotarsal ungues, to provide a vice-like grip on rock surfaces. Such adaptations facilitate locomotion on rock but hinder locomotion across alternative sub- strata. Accordingly, these scorpions are re- stricted to regions of rugged, mountainous to- pography and readily subject to allopatric speciation when mountain ranges become sep- arated through erosion. As part of an ongoing revision of the tax- onomy of Hadogenes, the status of the endem- ic South African flat rock scorpion, Hadoge- nes bicolor Purcell 1899, was reassessed. Purcell (1899: 437, 438) based his original de- scription of H. bicolor on an adult female from ‘Twenty miles east of Pietersburg,” al- though his syntype series contained “several adult and young specimens.” The description made no mention of the characters of the adult male H. bicolor. Hewitt (1918: 160, 161) sub- sequently described an adult male Hadogenes from Doornkop, near Belfast, ca. 200 km south of the type locality, noting “I think [the male] is referable to the same species {H. bi- color^."' In his description of the male, Hewitt (1918) observed that the metasoma was un- usually short for an adult male Hadogenes and that the lobe at the base of the movable finger 146 PRENDINI— TWO NEW SPECIES OF HADOGENES 147 of the pedipalp was “larger, deeper and more acute” than in other species of the genus. Hewitt (1918) also listed two adult females from Woodbush (Pietersburg district), and de- scribed a “half-grown” male, in which the sides of the telson vesicle were finely granu- lated, from the same locality. The evidence supporting Hewitt’s (1918) suggestion that the specimens from Woodbush and Doornkop were conspecific with each other, and with the syntypes of H. bicolor, was inconclusive. Nonetheless, this opinion was adopted by sub- sequent authors (Lawrence 1955; Lamoral & Reynders 1975). Hadogenes bicolor was not reviewed until Newlands (1980) redescribed the species on the basis of newly-collected material (an adult male from Leopard’s Crag and an adult female from Haffenden Heights), and again noted the large basal lobe of the movable finger and the short metasoma of the adult male as diagnos- tic characters. It is unclear whether Newlands (1980) actually examined the syntypes, for he listed the type specimens as “Female holotype and several nymphs housed in the Transvaal Museum (TMSA 4062) from 32 km east of Pietersburg.” During the present investiga- tion, the syntype series, deposited in the South African Museum, was found to comprise an adult male, two adult females, a subadult fe- male, a juvenile male, and a juvenile female. Newlands’ (1980) redescription of H. bi- color was never published. However, New- lands & Cantrell (1985) published electropho- retic and cytogenetic data collected by Newlands (1980), as well as Newlands’ (1980) key to the species of Hadogenes, in which the short metasoma of the adult male was yet again provided as a diagnostic char- acter for H. bicolor. Following Newlands (1980), Newlands & Cantrell (1985) pointed out that the electrophoretic banding patterns of venom proteins from specimens of H. bi- color collected at two localities, viz. Haffen- den Heights (Letaba district. Northern Prov- ince) and Zusterstroom (Bronkhorstspruit district, Gauteng Province), were distinctly different. Specimens from Zusterstroom dis- played a protein component that was absent in specimens from Haffenden Heights, ca. 180 km northeast. Newlands & Cantrell (1985: 42) suggested that these differences might be in- dicative of a cryptic species complex (Pater- son 1991), as “no morphological differences Figure 1 . — Map showing the distribution of Had- ogenes bicolor Purcell 1899 (■), Hadogenes lon- gimanus new species (+), and Hadogenes new- landsi new species (★) in South Africa. The specimens from Doornkop and Steelpoort (?) have been provisionally identified as H. longimanus, but may comprise another undescribed species in this complex. . . . could be detected” between specimens from the two localities. In the present study, specimens from across the distributional range of H. bicolor, includ- ing the material examined by Hewitt and Newlands, and newly collected material, were compared with the syntypes. Since Hadogenes species are notoriously difficult to identify without examination of the adult male, new series of H. bicolor, including adults of both sexes, were collected from several localities in the same and neighboring districts as the type locality, and in the districts from which the other material, examined by Hewitt and Newlands, originated. Some of these localities are as much as 100 km north to 200 km south of the type locality of H. bicolor. Examination of this new material has con- firmed the suggestion of Newlands (1980) and Newlands & Cantrell (1985) that more than 148 THE JOURNAL OF ARACHNOLOGY Table 1. — The currently accepted species of Hadogenes Kraepelin 1894 (Scorpiones, Ischnuridae), with countries of distribution compiled from Prendini (1995). ^ Species of dubious validity. ^ Species complexes. Hadogenes angolensis Lourengo 1999^ Hadogenes bicolor Purcell 1899 Hadogenes gracilis Hewitt 1909 Hadogenes granulatus Purcell 1901 Hadogenes gunningi Purcell 1899 Hadogenes lawrencei Newlands 1972 Hadogenes longimanus new species Hadogenes minor Purcell 1899 Hadogenes newlandsi new species Hadogenes paucidens Pocock 1896^ Hadogenes phyllodes Thorell 1877^ Hadogenes taeniurus (Thorell 1877) Hadogenes tityrus (Simon 1888)^ Hadogenes trichiurus (Gervais 1843)^ Hadogenes troglodytes (Peters 1861) Hadogenes zuluanus Lawrence 1937 Hadogenes zumpti Newlands & Cantrell 1985 Angola South Africa South Africa Botswana, Mozambique, Zambia, Zimbabwe South Africa Namibia South Africa South Africa South Africa Democratic Republic of Congo, ?Tanzania Namibia, South Africa Angola, Namibia Namibia, South Africa South Africa Botswana, Mozambique, South Africa, Zimbabwe South Africa, Swaziland ?Namibia, South Africa one species is involved. However, contrary to the view expressed by these authors, several consistent morphological differences could be identified between specimens from the local- ities at which samples, analyzed electropho- retically by Newlands (1980) and Newlands & Cantrell (1985), were found to differ in venom protein composition. Specimens identified as H. bicolor by pre- vious authors can be separated into at least three species on the basis of morphology, each of which occupies a discrete, allopatric distri- butional range (Fig. 1). In light of this new evidence, H. bicolor is redescribed and two new species, Hadogenes longimanus and Hadogenes newlandsi, are described. As in other closely related species of Hadogenes, adult female specimens of all three species are superficially similar morphologically, whereas adult male specimens differ markedly. How- ever, adult females of all three species can also be reliably identified on the basis of several consistent diagnostic characters. These char- acters are summarized in a key to the identi- fication of the three species. Recognition of the two new species raises the number of cur- rently accepted species of Hadogenes to 17 (Table 1). Lourengo’s (1999, 2000) recent proposals to transfer Hadogenes to the Scorpioeidae La- treille 1802, or provide a monotypic family Hadogenidae Lourengo 2000 are unsupported by cladistic analysis (Prendini 2000). This contribution concludes with a discussion of the phylogenetic position of Hadogenes, in which the Hadogenidae is synonymized with the Ischnuridae. METHODS Material examined, including the type spec- imens of H. bicolor, H. longimanus and H. newlandsi, is deposited in the following col- lections: South African Museum, Cape Town (SAMC); Transvaal Museum, Pretoria, South Africa (TMSA); Albany Museum, Grahams- town. South Africa (AMGS); Natal Museum, Pietermaritzburg, South Africa (NMSA); American Museum of Natural History, New York (AMNH); California Academy of Sci- ences, San Francisco (CASC). Tissue samples of the three species, stored in absolute ethanol at ”20 °C, have been retained separately for DNA isolation and sequencing in the Am- brose Monell Collection for Molecular and Microbial Research at the American Museum of Natural History, New York (AMC). Illustrations of H, bicolor, H. longimanus and H. newlandsi were produced using a Wild stereomicroscope and camera lucida. Mea- surements were made with Mitutoyo® digital calipers. Color designation follows Smithe (1974, 1975, 1981), trichobothrial notation follows Vachon (1974), and mensuration fol- lows Stahnke (1970) and Lamoral (1979). Morphological terminology follows Couzijn PRENDINI— TWO NEW SPECIES OF HADOGENES 149 (1976) for the segmentation of legs, Hjelle (1990) and Sissom (1990) for the segmenta- tion of pedipalps, and Stahnke (1970), La- moral (1979), Newlands (1980), Sissom (1990) and Newlands & Prendini (1997) for remaining features. Key to the identification of Hadogenes bicolor Purcell 1899, Hadogenes longimanus new species and Hadogenes newlandsi new species 1. Pedipalp chela with 5—8 trichobothria in the i series (Fig. 21)............ .Hadogenes longimanus Pedipalp chela with two trichobothria in the i series (Figs. 10, 32). ....................... . 2 2. Pedipalp chela of adult 6 and $ with a pronounced lobe, distal to the notch in the fixed finger (Figs. 8, 9); metasoma of adult 6 length ca. 55% of total length (Figs. 2, 3), with telson smooth and lateral surfaces of metasomal segment V sparsely granular (Fig. 33) ........ Hadogenes bicolor Pedipalp chela of adult S and 9 without a pronounced lobe, distal to the notch in the fixed finger (Figs. 26, 30); metasoma of adult S length ca. 60% of total length (Figs. 22, 23), with telson and lateral surfaces of metasomal segment V densely granular (Fig. 35) .......... Hadogenes newlandsi Hadogenes bicolor Purcell 1899 Figs. 1-10, 33, 36, Table 2 Hadogenes bicolor Purcell 1899: 437, 438. Hadogenes bicolor: Lawrence 1955: 251 (part); La- moral & Reynders 1975: 538 (part); Newlands 1980 (unpublished): 99-105 (part), figs. 48 (part), 49-53; Newlands & Cantrell 1985: 40, 42, 44 (part); Kovank 1998: 132; Fet 2000: 387. Types.- — SOUTH AFRICA; Northern Province: Pietersburg district: Syntypes: d, 2 $ , subadult $ , juv S , juv $ (S AMC 4062), 20 miles east of Pietersburg [23°54'S, 29°47'E]. The S is hereby designated as the lectotype of H. bicolor and the remaining specimens as paralectotypes. Diagnosis. — Hadogenes bicolor is the sis- ter species of H. longimanus. These two spe- cies are both characterized by a pronounced lobe, distal to the notch in the fixed finger of the pedipalp chela of adult S and $, and a relatively short metasoma in the adult S , com- pared with H. newlandsi and other Hadogenes species. Accordingly, these characters are hy- pothesized to be synapomorphic for H. bicolor and H. longimanus. Hadogenes bicolor can be separated from H. longimanus by the presence of two, rather than 5-8, trichobothria on the internal surface of the pedipalp chela. Hadogenes bicolor can be further distinguished from H. newlandsi by the smooth telson of the adult S , and the lon- ger pedipalp segments of adult S and $ . Description. — The following description is based on the lectotype S (SAMC 4062), a paralectotype $ (SAMC 4062), the S from Leopard’s Crag (TMSA 18004) and $ from Haffenden Heights (TMSA 18005) described by Newlands (1980), and a newly collected S (Figs. 2, 3) and $ (Figs. 4, 5) from Jong- manssprait (SAMC C4585). It is intended to complement Purcell’s (1899) original descrip- tion and Newlands’ (1980) unpublished re- description. Color: (SAMC C4585). Pale chelicerae, legs, and telson contrasting markedly with darker carapace, pedipalps, tergites and me- tasomal segments I-V. Stemites also paler than tergites and metasomal segments. Pedipalps, Buff 24 on chela manus and intercarinal sur- faces of patella and femur, Sepia 119 on ca- rinae and chela fingers; cheliceral manus, legs (except prolateral surfaces of femora), telson, sternites, pectines, and genital operculum, Straw Yellow 36; cheliceral fingers, carapace, tergites ( d ) and prolateral surfaces of leg fem- ora, Sepia 119; tergites ($) and metasomal segments I-V, Dark Brownish-olive 129. Carapace: Three pairs of lateral ocelli, equal in size to median ocelli (Fig. 6). Median ocular tubercle with superciliary carinae well developed, protruding above ocelli, and inter- ocular sulcus distinct. Anterior margin of car- apace with median notch well developed, such that triangular inset is situated far back and frontal lobes protrude anteriorly. Anterome- dian sulcus deep, suturiform, furcating ante- riorly around triangular inset. Median longi- tudinal suture distinct, continuous from anterior furcated sutures, through ocular tu- bercle to posterior furcated sutures, which converge on ocular tubercle from posterior carapace margin. Posterior furcated sutures obsolete, discontinuous. Posteromedian and posteromarginal sulci distinct, but shallow. Paired median lateral and posterolateral sulci also distinct, shallow. Carapace entirely gran- 150 THE JOURNAL OF ARACHNOLOGY Figures 2-5. — Hadogenes bicolor Purcell 1899, habitus of S and $ (SAMC C4585). 2. Dorsal aspect, 3. Ventral aspect. S’, 4. Dorsal aspect, ?; 5. Ventral aspect, $. Scale bars = 20 mm. ular, except for surfaces of frontal lobes, me- dian lateral, posterolateral and posteromargin- al sulci, which are smooth. Granulation almost uniformly fine, becoming coarse on antero- ocular and anterolateral surfaces. Chelicerae: Movable finger with distal in- ternal tooth slightly smaller than distal exter- nal tooth, and apposable. Ventral aspect of fin- gers and manus with long, dense macrosetae. Pedipalps: Femur pentacarinate, with four PRENDINI— TWO NEW SPECIES OF HADOGENES 151 6 7 Figures 6-1. — Hadogenes bicolor Purcell 1899, carapace and stemite VII of $ (SAMC C4585), showing carinae, depressions and sulci, 6. Cara- pace; 7. Stemite VII. Scale bar = 4 mm. distinct carinae; ventroexternal carina obso- lete, reduced to a few granules proximally; dorsoexternal and externomedian carinae granular, dorsointemal and ventrointemal ca- rinae costate-granular, composed of very large heavily sclerotized granules; dorsoexternal and ventral intercarinal surfaces finely and uniformly granular; internal intercarinal sur- faces smooth, except for a few scattered spi- niform granules. Femur length 63% (62-64%) greater than width in d, 60.5% (58-63%) greater in 9 (Table 2). Patella with seven carinae, six of which are distinct, whereas the dorsoexternal carina is obsolete; dorsointemal and ventrointemal ca- rinae costate to costate-granular; internome- dian carina costate-granular, composed of very large heavily sclerotized spiniform gran- ules; externomedian and ventroexternal cari- nae granular; dorsoexternal and ventral inter- carinal surfaces finely and uniformly granular, becoming granulo-reticulate on ventral surfac- es; internal intercarinal surfaces smooth; an- terior process strongly developed. Patella length 45% (44-46%) greater than width in <3, 41% (40-42%) greater in 9. Chela pentacarinate, with three distinct ca- rinae; dorsal secondary and digital carinae ob- solete (Figs. 8, 9); external secondary carina strongly developed, costate to costate-granu- lar; ventroexternal carina strongly developed, crenulate, aligned parallel to longitudinal axis of chela, with distal edge disconnected from external movable finger condyle and directed toward a point between external and internal movable finger condyles, but closer to exter- nal condyle (Fig. 10); ventromedian carina ob- solete, reduced to a vestigial granule proxi- mally; ventrointemal carina also obsolete; internomedian and dorsointemal carinae weakly developed, each comprising a series of isolated spiniform granules; dorsomedian ca- rina strongly developed, composed of a con- tinous double row of spiniform granules; dor- sal and ventrointemal intercarinal surfaces smooth, reticulate; dorsointemal intercarinal surfaces with scattered spiniform granules, be- coming finely granular on internal surface of fixed finger; external intercarinal surfaces coarsely granular. Chela with a pronounced, conical lobe on movable finger and corre- sponding notch in fixed finger; fixed finger ad- ditionally with a pronounced, conical lobe dis- tal to the notch, and a smaller, rounded lobe proximally. Dentate margins of chela fingers with double row of denticles, which are fused at the lobe/notch. Chela length along ven- troexternal carina 46.5% (44-49%) greater than chela width in c3, and 36.5% (33-40%) greater in 9; chela width 50.5% (44-57%) greater than chela height in c3, and 52% (49- 55%) greater in 9 ; length of movable finger 9% (5-13%) less than length along ventroex- ternal carina in d, and 4% (1-7%) less in 9. Trichobothria: Neobothriotaxic major, type C (Figs. 8, 10), with the following segment totals (Table 2): femur 3 (1 (/i 0 Q. o o 0 ^ 0 o O ^ n rn r 0 0 ■Q 0 0 2 0 « 0 E fc E 0 0 o 40 0 ^ o 0 “ 0 3 .0 0 |8 ^ 0 w E ^ 0 0 g 2 O Z h- Q 1 n n m m 0 S 2 £0 1 I S g I i 0 0 2 2 0 0 Q. o c o Q. Q. 0 0 2 ‘Ui c. 0 0 0 0 2 0 E o 0 o 2 o CO 1 n r 0 0 2 TD ID 0 O 0 0- 2 'Ic o >. c o E 0 0 1 r 1 r 0 0 ^0 ^ CL ^ 0 ■c ^ E 0 c ^ b P cc m 0 0 g 2 i2 CO K 3: Q li: Q .0 0 0 S- o 0 C 5 o ^ CL .0 .Q-'B ^ O 0 .0 Q, 0 :5 O 3 D> 0 ^ ■§€i 8 "D "0 S 2 0 0 0* O 0 0 2 Q. 0 c o CD § ■0 2 2 0 0 2 0 E E o o 0 0 2 '-a 0 01 c _0 0 1 n r 2 g g § !■§ 0 2 S^OOlIjCoCDltcoJtQ'^ CL X C 0 C 0 ■0 O 0 18'-^ o E o 0 0 2 ‘c 2 CD 1 r 0 0 0 2 0 0 2 0 0 s 0 0 >Q f 2 0 .0 ■Q ■2 0 0 E 0 0 0 0 ■ts 5 2 0 0 0 .0 0 0 0 .5 "0 0 0 0 8 0 2 2 0 0 CL Q ■2] CL CL CO O ^ 2 'd o Owe 2 (U m y. S d d u TO >> 0 3 g eg CO d d d .2 d X p d d "Oh > *g p > d g 0 3 d £ d £ oT 3 ^ r=* CQ i d p d bX) ^ , r2 •S s ^ "S ^ -O (U ’o O X) T (U 2 c -S O M ■*■=* "S tj_. o ^ d 0) o o £ o w E d g H . ^ X) ^ , X g d ^ d + g o ^ S ^ 3 S £ d ^ M 0 d d e •- 5;- d 4-» fs X d ^ .S ^ ^ ^ d 2 § ^ g D. o Q § § ^ d ^ « X 1 ^ i ^ I ^ . I e s = “ I “ I ^ S .X d d w • d -g hJ c ^ .S ^ m d ^ s s'?” = . . d ^ Q M ^ d II ill ^ g I -g d d ^ g X - J 2 ^ d d g § 'N 8 “> > is -2 -S o g- « ;X d d 2 4- d C 3 O I S -g 1 ^ i I § i Si g o S d d CO « d 3 d S d 3 CO ^ o c C ^ eg 3 d d 3 !_, d d c 'O d fi d ^ d ^ .-H g .g S d ^ Cl C 5 Cl s ^ g 3 tzi >> d “I ^ S § ^§-2 ^=i 1 d d ^ d d O CO S d c d -S d ■*■" c^ CL 5 X 196 THE JOURNAL OF ARACHNOLOGY 1995; Cannatella et al. 1998), we performed combined analyses of all partitions. Maximum-parsimony (MP) analyses of all character partitions (all-nucleotides, ntl +2, ntlnoLR + nt2, nt3, amino acids, amino-ac- ids-protpars) recovered the following higher clades within their respective minimal-length trees (bootstrap percentages [BP] in parenthe- ses): Phalangida (100, 100, 100, 95, 100, 100); Laniatores (92, 99, 100, 45, 100, 100); Palpatores (100, 96, 79, 88, 87, 76); Trogu- loidea (93, 97, 94, 45, 94, 97); Caddoidea (100, 100, 100, 100, 100, 100) and Phalan- gioidea (100, 100, 100, 98, 100, 100). Ischy- ropsalidoidea was recovered by all-nucleo- tides (BP 56), ntl + nt2 (BP 73), ntlnoLR + nt2 (BP 43) and amino-acids-protpars (BP 30) but was recovered as a paraphyletic group by nt3 and amino acids. Dyspnoi (= Troguloidea + Ischyropsalidoidea) was recovered by all- nucleotides (BP 24), ntl + nt2, (BP 59), ntlnoLR -f nt2 (BP 54), and amino-acid-prot- pars (BP 47). Eupnoi (= Caddoidea + Phal- angioidea) was recovered by ntlnoLR +nt2 (BP 36) within a subset of 20 MP trees. Gon- yleptoidea was recovered by all-nucleotides, ntlnoLR + nt2, amino acids, and amino-ac- ids-protpars with weak-to-moderate bootstrap support (i.e., 29-74%). Travunioidea was re- covered as a paraphyletic group with Equitius being the sister group to Gonyleptoidea and/ or occurring within Gonyleptoidea. The all-nucleotide and ntl + nt2 data sets significantly preferred a monophyletic Palpa- tores over the MP clade constrained to a monophyletic Laniatores + Dyspnoi (P values ranged from 0.015 to 0.028), according to the test of Kishino & Hasegawa (1989); the other data sets were indecisive. The ML topology recovered by analysis of the ntlnoLR + nt2 data set is shown in Fig. 1. All maximum likelihood analyses (three models of nucleotide substitution and four data subsets, see Methods) recovered the fol- lowing clades in their ML topologies (BP val- ues for analysis of ntl + nt2 and ntlnlLR + nt2, respectively, are in parentheses): Phalan- gida (100, 100), Laniatores (100, 100), Pal- patores (96, 78), Troguloidea (98, 93), Gon- yleptoidea (63, 85), Phalangioidea (100, 100), Caddoidea (100, 100), and Ischyropsalidoidea (53, 42). Eupnoi (37, 55) was recovered by all three nucleotide data sets. Dyspnoi (19, 36) was recovered only by the ntlnoLR + nt2 data set. ML analysis of amino acids recov- ered Phalangida, Laniatores, Palpatores, Tro- guloidea, Caddoidea, Phalangioidea, and Gon- yleptoidea. Combined analyses. — Combined un- weighted MP analyses of 18S + 28S rDNA and EF-la + Pol II sequences strongly and consistently corroborated the monophyly of Palpatores (Fig. 2), the result favored by anal- ysis of EF-la + Pol II alone, not the Dyspnoi + Laniatores hypothesis that tended to be re- covered by the rDNA data analyzed alone (Gi- ribet et al. 1999). The strict consensus of 360 trees (length, 6067; Cl, 0.31; RI, 0.48) derived from analysis of ribosomal and protein-encod- ing DNA included Opiliones, Phalangida, Laniatores, Palpatores, Dyspnoi and all rep- resented opilion superfamilies (except Travu- nioidea) but did not recover Eupnoi (Fig. 2A). The Dyspnoi + Laniatores hypothesis was re- covered in only 5% of bootstrap replicates and required 30 additional steps for recovery by parsimony analysis. Strict consensus of 78 trees (length, 865; Cl, 0.54; RI, 0.75), derived from MP analysis of ribosomal DNA and EF- la + Pol II amino acids included the same major opilion clades listed above but also re- covered Eupnoi (Fig. 2B). The Dyspnoi + Laniatores hypothesis was recovered in only 5% of bootstrap replicates and required 10 ad- ditional steps for recovery by parsimony. DISCUSSION Status and future of higher relationships in Opiliones. — Recent phylogenetic analyses clearly indicate that Opiliones consists of two clades, Cyphophthalmi and Phalangida (Giri- bet et al. 1999; Shultz 1998; Shultz & Regier unpubl. data). Results from analysis of elon- gation factor- la (EF-la) and RNA polymer- ase II (Pol II), both alone and combined with 18S and 28S rDNA, strongly support the monophyly of Palpatores and Laniatores and are inconsistent with the recently proposed Dyspnoi + Laniatores hypothesis (Giribet et al. 1999, Giribet & Wheeler 1999). Molecular data examined thus far have consistently re- covered three palpatorean superfamilies (Cad- doidea, Phalangioidea, Troguloidea) as mono- phyletic groups (Giribet et al. 1999; present study), but monophyly of the remaining su- perfamily, Ischyropsalidoidea, is more prob- lematic. This superfamily was recovered by SHULTZ & REGIER— MOLECULAR SYSTEMATICS OF OPILIONES 197 Ceratoiasma Ischyropsalis Hesperonemastoma Sabacon Taracus Dicranoiasma Trogulus Nipponopsalis Dendrolasma+Centetostoma Odiellus Phalangium+NeHma Astrobunus Caddo agilis Caddo pepperella Sclerobunus Equitius Gonyleptes Progonyleptoidellus Sodreana Promitobates Discocyrtus+Pachyloides Lineosoares Gnidia Pseudobiantes Proscotolemon Maiorerus Scotoiemon Oncopus Siro Parasiro Stylocellus Pseudocellus Eusimonia Androctonus Limulus Carcinoscorpius Phalangida 79 Eupnoi 71 64 62 68 39 88 Opiliones 1 72 Laniatores ggCyphophthalmi \ 100 40 71 98 B Figure 2. — Strict consensus trees derived from combined, unweighted maximum parsimony analyses of 18S rDNA, 28S rDNA, EF-la and Pol IL A. Strict consensus of 360 trees derived from analysis of rDNA and EF-la + Pol II nucleotides. B. Strict consensus of 78 trees derived from analysis of rDNA and EF- la + Pol II amino acids. Numbers above internal branches are bootstrap percentages. EF-la and Pol II in only a few analyses with low bootstrap support (Fig. 1) and in both combined analyses with moderate support (Fig. 2). The traditional grouping of Caddoi- dea and Phalangioidea into Eupnoi and Ischy- ropsalioidea and Troguloidea into Dyspnoi could be neither falsified nor confirmed by EF-la and Pol II, but they are both strongly supported in the combined analysis of 18S and 28S rDNA conducted by Giribet et al. (1999). Based on the present study, it appears that phylogenetic signal within EF-la and Pol II can resolve the deepest nodes within opilion phylogeny but become less useful at lower taxonomic levels. Specifically, relationships among and within superfamilies are largely unresolved by these two genes. This suggests that progress in opilion systematics will re- quire the development of new genes, espe- cially those that have evolved at a greater rate 198 THE JOURNAL OF ARACHNOLOGY than the two examined here as well as re^ newed efforts to search for informative mor- phological characters. Expanded taxon sam- pling is also important, and particular emphasis should be placed on sampling rep- resentatives from basally divergent lineages of each well-defined opilion clade. These in- clude, but are not limited to, Crosbycus from Ischyropsalidoidea, neopilionids from Phal- angioidea, acropsopilionids from Caddoidea and the travunioid families from Laniatores. Relationships within Ischyropsalidoi- dea.— Ischyropsalidoidea is a morphological- ly diverse group encompassing three families (Ischyropsalididae, Sabaconidae, Ceratolas- matidae), but the monophyly of the superfam- ily is not well established. Shear (1986) united the ischyropsalidoids with four synapomor- phies, namely, metapeltidial cones, unseg- mented ovipositor, reduced palpal claw and presence of male cheliceral glands. However, metapeltidial cones are also present in Cad- doidea {Caddo) and the remaining characters occur in some or all members of Troguloidea (Shultz 1998). Further, our molecule-based re- sults do not support Shear’s (1986) system of ischyropsalidoid families (Fig. 3A), but nei- ther do they strongly conflict with it. Specif- ically, our analyses consistently recovered an Ischyropsalis + Ceratolasma grouping and frequently but weakly placed the ceratolas- matid Hesperonemastoma as the sister group to one or both sabaconids (Fig. 1). These re- sults suggest that Ceratolasmatidae and Sa- baconidae are not natural groups, although the evidence is weak. Still, para- or polyphyly of the two families is a hypothesis to be tested and is compatible with several other lines of evidence. First, few morphological characters support monophyly of the morphologically di- verse Ceratolasmatidae (= Acuclavella, Cer- atolasma, Crosbycus, Hesperonemastoma). Shear’s (1986) diagnosis of the family in- cludes no unique features and most characters have multiple states within the family (e.g., “. . . pairs of tubercles on scutum . . . high and acute, or blunt and appressed ... or absent,” “chelicerae with or without glands in males”; “palpi long, with ... or short, without . . . plu- mose setae”) (p. 13). Second, monophyly of Sabaconidae can also be questioned, although it is substantially more convincing than Cer- atolasmatidae. Potential sabaconid synapo- morphies include deep invagination in anterior midline of carapace, reduced sclerotization, and enlarged palpal tibia and tarsus (Shear 1986). The first of these appears to be unique to the family, but the second seems to depend on a priori suppositions of character transfor- mation. The third character is undoubtedly de- rived and clearly expressed in Sabacon, but is less obvious in Taracus, especially when the palps are compared to those of Hesperone- mastoma (original observations). Thus, Saba- conidae is a probable but not unambiguously demonstrated family. Third, in his re-descrip- tion of Ceratolasma, Gruber (1978) noted two phenetic groupings of taxa within Ischyrop- salidoidea that are broadly congruent with our findings. Specifically, Ceratolasma and Ischy- ropsalis share a “prominent sternum, large la- bium, palpi without plumose setae, but with numerous microtrichia, and also a complex midgut anatomy” (p. 109). He also noted that Hesperonemastoma is similar to sabaconids in having “less developed sterna, small labia, palpi with extensive development of plumose setae and reduction of the microtrichial cov- er” and simpler midgut anatomy. Admittedly, both lists are mosaics of primitive and derived traits, and more intensive morphological and molecular analyses are needed to make pro- gress in ischyropsalidoid systematics. The molecular data are open to criticism in that relationships among the relevant taxa are un- stable and Hesperonemastoma appears to have undergone more rapid molecular evolution than other ischyropsalidoids, which may ac- count for the ambiguity. Relationships within Troguloidea. — In contrast to ischyropsalidoids, the troguloids are a well-defined, monophyletic group. Mar- tens (Martens 1980, 1986; Martens et al. 1981; Martens & Suzuki 1966) and Shear & Gruber (1983) proposed several synapomo- phies, including a penis with two longitudinal muscles, unique unsegmented ovipositor, fu- sion of sternum and leg coxae, clavate palpal setae, and reduced palpal claws. Only the lat- ter is open to question, as reduced palpal claws are also present in ischyropsalidoids. The present study included one representative from each of the four troguloid families (Nip- ponopsalididae, Nemastomatidae, Dicranolas- matidae, Trogulidae) and our results strongly supported the monophyly of the superfamily under all character partitions and analytical methods (Fig. 1). However, relationships with- SHULTZ & REGIER— MOLECULAR SYSTEMATICS OF OPILIONES 199 in Troguloidea were ambiguous. Shear & Gruber (1983) regarded Dicranolasmatidae and Nemastomatidae as sister groups on the basis of one character (penis muscles with long tendons) but did not propose relation- ships between this clade, Trogulidae and Nip- ponopsalididae. Shultz (1998) proposed di- cranolasmatids and trogulids as sister groups based on heavy sclerotization and anteriorly projecting eye tubercle or “hood” in these taxa. Our molecular data do not strongly sup- port any phylogenetic arrangement among the troguloid families, although there was a ten- dency to recover the nemastomatid “Dendro- lasma” as the sister group to the remaining representatives. ACKNOWLEDGMENTS We thank James Cokendolpher, Gonzalo Giribet, Jurgen Gruber, Andrew Moldenke, Carles Ribera, Ricardo Pinto-da-Rocha and Nobuo Tsurusaki for specimens, Jurgen Grub- er has been especially helpful in providing in- formation and insights into opilion diversity and systematics. This work was supported by National Science Foundation grants DEB- 9629791 and DEB-9615526, the Center for Agricultural Biotechnology (University of Maryland Biotechnology Institute), and the Maryland Agricultural Experiment Station. LITERATURE CITED Adachi, J. & M. Hasegawa, 1994. Programs for Molecular Phylogenetics. Version 3.3. Institute of Statistical Mathematics. Baker, R.H. & R. DeSalle. 1997. Multiple sources of character information and the phylogeny of Hawaiian drosophilids. Systematic Biology 46: 654-673. Cannatella, D.C., D.M. Hillis, RT. Chippindale, L. Weigt, A.S. Rand & MJ. Ryan. 1998. Phylog- eny of frogs of the Physalaemus pustulosus spe- cies group, with an examination of data incon- gruence. Systematic Biology 47:311-335. Farris, J.S., M. Kallersjo, A.G. Kluge & C. Bult. 1995. Constructing a significance test for incon- gruence. Systematic Biology 44:570-572. Felsenstein, J. 1985. Confidence limits on phylog- enies: An approach using the bootstrap. Evolu- tion 39:783-791. Giribet, G., S. Carranza & C. Ribera. 1995. Mo- lecular phylogeny of Opiliones based on 18S rDNA sequences. P. 219. In Proceedings of the 15 th European Colloquium of Arachnology. (V. Ruzicka, ed.). Institute of Entomology, Ceske Budejovice. Giribet, G., M. Rambla, S. Carranza, M. Riutort, J. Baguna & C. Ribera. 1999. Phylogeny of the arachnid order Opiliones (Arthropoda) inferred from a combined approach of complete 18S, par- tial 28S ribosomal DNA sequences and mor- phology. Molecular Phylogenetics and Evolution 11:296-307. Giribet, G. & W.C. Wheeler. 1999. On gaps. Mo- lecular Phylogenetics and Evolution 13:132-143. Gruber, J. 1978. Redescription of Ceratolasma tri- cantha Goodnight and Goodnight, with notes on the family Ischyropsalidae (Opiliones, Palpa- tores). Journal of Arachnology 6:105-124. Hansen, H.J. & W. Sprensen. 1904. On Two Orders of Arachnida. Cambridge Univ. Press, Cam- bridge. Hasegawa, M., H. Kishino & T. Yano. 1985. Dat- ing of the human-ape splitting by a molecular clock of mitochondrial DNA. Journal of Molec- ular Evolution 22:160-174. Hoheisel, U. 1980. Anatomic und taxonomische Bedeutung der Legerohre der Opiliones. Verhan- dlungen der 8. Internationale Arachnologische Kongress, 1980, Wien 8:315-318. Jones, D.T., WR. Taylor & J.M. Thorton. 1992. The rapid generation of mutation data matrices from protein sequences. Computer Applications in the Biosciences 8:275-287. Juberthie, C. & J.-F. Manier. 1978. Etude ultra- structural comparee de la spermiogenese des op- ilions et son interet phyletique. Symposia of the Zoological Society of London, No. 42:407-416. Kishino, H. & M. Hasegawa. 1989. Evaluation of the maximum likelihood estimate of the evolu- tionary tree topologies from DNA sequence data, and the branching order in Hominoidea. Journal of Molecular Evolution 29:170-179. Macias-Ordonez, R. 1997. The Mating System Of Leiobunum vittatum Say 1821 (Arachnida: Opi- liones: Palpatores): Resource Defense Polygyny In The Striped Harvestman. Ph.D. dissertation, Lehigh University. Maddison, WP. & D.R. Maddison. 1992. Mac- Clade: Analysis of phylogeny and character evo- lution. Version 3. Sinauer Associates: Sunder- land, Massachusetts. Martens, J. 1976. Genitalmorphologie, System und Phylogenie der Weberknechte (Arachnida: Opi- liones). Entomologica Germanica 3:51—68. Martens, J. 1980. Versuch eines phylogenetischen Systems der Opiliones. Verhandelungen der 8. Internationale Arachnologische Kongress, 1980, Wien 355-360. Martens, J. 1986. Die Grossgliederang der Opili- ones und die Evolution der Ordnung (Arachni- da). Actas X Congress International Arachnolo- gy, Jaca/Espana, 1:289-310. Martens, J. 1993. Further cases of paternal care in 200 THE JOURNAL OF ARACHNOLOGY Opiliones (Arachnida). Tropical Zoology 6:97- 107. Martens, J., U. Hoheisel & M. Gotze. 1981. Ver- gleichende Anatomic der Legerohren der Opi- lions als Beitrag zur Phylogenie der Ordnung (Arachnida). Zoologischer Jahrbiicher Anatomic 105:13-76. Martens, J. & S. Suzuki. 1966. Zur systematischen Stellung ostasiatischer Ischyropsalididen-Arten (Arachnoidea, Opiliones, Ischyropsalididae). An- notationes Zoologicae Japonenses 39:215-221. Mora, G. 1987. Male Egg-guarding In The Neo- tropical Harvestman Zygopachylus albomarginis (Opiliones: Gonyleptidae). M.S. thesis. Univer- sity of Florida, Gainesville. Mora, G. 1990. Paternal Care In A Neotropical Harvestman, Zygopachylus albomarginis (Arachnida: Opiliones: Gonyleptidae). Animal Behaviour 39:582-593. Mora, G. 1991. Site-based mating system in a trop- ical harvestman. Ph.D. dissertation. University of Florida, Gainesville. Ramires, E.N. & A. A. Giaretta. 1994. Maternal care in a neotropical harvestman, Acutisoma proximum (Opiliones, Gonyleptidae). Journal of Arachnology 22:179-180. Regier, J.C. & J.W. Shultz. 1997. Molecular phy- logeny of the major arthropod groups indicates polyphyly of crustaceans and a new hypothesis for the origin of hexapods. Molecular Biology and Evolution 14:902-913. Shear, W.A. 1982. Opiliones. Pp. 104-1 10. In Syn- opsis and Classification of Living Organisms, Vol. 2. (S.P. Parker, ed.). McGraw-Hill, New York. Shear, W.A. 1986. A cladistic analysis of the opi- lionid superfamily Ischyropsalidoidea, with de- scriptions of the new family Ceratolasmatidae, the new genus Acuclavella, and four new species. American Museum Novitates, No. 2844:1-29. Shear, W.A. & J. Gruber. 1983. The opilionid sub- family Ortholasmatinae (Opiliones, Troguloidea, Nemastomatidae). American Museum Novitates, No. 2757:1-65. Shultz, J.W. 1990. Evolutionary morphology and phylogeny of Arachnida. Cladistics 6:1-38. Shultz, J.W. 1998. Phylogeny of Opiliones (Arach- nida): An assessment of the “Cyphopalpatores” concept. Journal of Arachnology 26:257-272. Shultz, J.W. 2000. Skeletomuscular anatomy of the harvestman Leiobunum aldrichi (Weed 1893) (Arachnida: Opiliones: Palpatores) and its evo- lutionary significance. Zoological Journal of the Linnean Society 128:401-438. Shultz, J.W. & J.C. Regier. 2000. Phylogenetic analysis of arthropods using two nuclear protein- encoding genes supports a crustacean + hexapod clade. Proceedings of the Royal Society of Lon- don B 267:1011-1019. Silhavy, V. 1961. Die Grundsatze der modernen Weberknechttaxonomie und Revision des bish- erigen Systems der Opilioniden. Verhandlungen der 1 1 . Internationaler Kongress fiir Entomolo- gie, 1960, Wien 1:262-267. Smith, S.W., R. Overbeek, C.R. Woese, W. Gilbert & PM. Gillevet. 1994. The genetic data envi- ronment and expandable GUI for multiple se- quence analysis. Computer Applications in the Biosciences 10:671-675. Staden, R., K.E Beal & J.K. Bonfield. 1999. The Staden package, 1998, In Bioinformatics Meth- ods and Protocols. (S. Misener & S. Krawetz, eds.). The Humana Press, Inc., Totowa, New Jer- sey. 512 pp. Swofford, D.L. 1998. PAUP* 4.0 beta version. Sinauer Associates, Sunderland, Massachusetts. Swofford, D.L., G.J. Olsen, P.J. Waddell & D.M. Hillis. 1996. Phylogenetic inference. Pp. 407- 514. In Molecular Systematics, 2nd ed. (C. Mo- ritz & B.K. Mable, eds). Sinauer Associates, Sunderland, Massachusetts. Weygoldt, P. & H.E Paulus. 1979. Untersuchungen zur Morphologie, Taxonomie und Phylogenie der Chelicerata. Zeitschrift fiir zoologische Syste- matik und Evolutionsforschung 17:85-116, 77- 200. Manuscript received 16 June 2000, revised 9 March 2001. 2001. The Journal of Arachnology 29:201-204 DESCRIPTION OF HAKKA, A NEW GENUS OF JUMPING SPIDER (ARANEAE, SALTICIDAE) FROM HAWAII AND EAST ASIA James W. Berry: Department of Biological Sciences, Butler University, Indianapolis, Indiana 46208 USA Jerzy Proszynski: Muzeum i Instytut Zoologii PAN, uL Wilcza 64 00-679 Warszawa, Poland ABSTRACT. We describe a new genus for a jumping spider that was originally placed in the large genus Menemerus Simon 1868, from which the new genus is clearly different. They were later reclassified as Icius, then as Pseudicius, and still later as Salticus. These initial classifications were repeated by a number of authors. The distinctive features of the male, and somewhat ambiguous features of the female, do not fit any known genus; and this species is here assigned to the new genus Hakka. Keywords: Hakka, Salticus, Menemerus, Hawaii, Salticidae Like many other elements of the Hawaiian Islands, the jumping spiders of the islands are poorly known. Much of the known fauna con- sists of genera whose origin can be traced from either Asia or North America. This paper discusses a species found in Hawaii that was previously known under several different ge- neric names — from a few specimens only — from China, Korea and Japan. One specimen was recorded in Hawaii in 1923, and we have recently collected two more. It is not known whether they are incidental recent arrivals (al- though the three specimens were collected over a period of 74 years) or have populations established there. Hakka new genus Discussion. — Assigning these salticid spi- ders to a genus has always created a problem. Although they have never been described as a separate genus, they were originally placed in the large genus Menemerus Simon 1868 (Doenitz & Strand in Bosenberg & Strand 1906), from which they are clearly different. Proszynski (1976) reclassified them first to Icius Simon 1876, subsequently correcting himself and re-interpreting them as Pseudicius (Proszynski 1987). Wesolowska (1981) inter- preted the structure of the epigynum as resem- bling the genus Salticus Latreille 1804 and de- scribed the female as Salticus koreanus Wesolowska. These initial classifications were later repeated by a number of authors. The fact is that the distinctive features of the male, and somewhat ambiguous features of the fe- male, do not fit any known genus; and this species deserves delimitation to its own ge- nus. There are no direct biological observa- tions confirming the matching of males and females of this species; however, persistent in- terpretation of that matching by a number of authors deserves following until proven oth- erwise. Diagnosis. — Hakka is a unidentate salticid with two prolateral cheliceral teeth, without patellar spines, and without lateral spines on metatarsi I and II. These same characters oc- cur in the genera with which it has been con- fused— Icius, Menemerus, Pseudicius, and Salticus, but they do separate Hakka from many other salticid genera. The absence of stridulatory spines from the carapace and mi- crospines from femur I, and presence of 5-6 ventral spines on tibia I clearly separate Hak- ka from Pseudicius. The latter has the stridu- latory spines and, on tibia I, normally 0-3 spines that are usually much reduced in length and often thickened basally. Pseudicius differs also by having a long, flat, relatively narrow carapace, and large robust first legs with tibia I more-or-less swollen, and with unusually long trichobothria, usually bent at a distinct angle. From Salticus, Hakka is distinguished by the absence of elongate male chelicerae. 201 202 THE JOURNAL OF ARACHNOLOGY the presence of ventral spines on tibiae I and II, the elongate bulb of the male palp overlap- ping the tibia proximally, and the medium- long sinuous embolus (Figs. 3, 4). The epi- gynum is less sclerotized than in Salticus; and the epigynal ducts run forward from the cop- ulatory openings, then turn back to the sper- mathecae (Figs. 6, 7). The typical Salticus col- or pattern of white lines of scale-like hairs is absent. Icius differs by having a proportion- ately longer, somewhat oblong carapace (shorter and more ovate in Hakka) and abdo- men, the palpal bulb narrowing anteriorly, and a distinct color pattern, consisting in part of scale-like hairs. Menemerus, the genus in which H. himeshimensis was originally placed, has a flatter, broader cephalothorax and abdomen. Also, the male palp of Mene- merus has the tibia and patella short and broad, often as broad as the cymbium, and a broad crescentic femur; the RTA is large, the embolar base wide and separated by a groove from the rest of the bulb: the embolus is ac- companied by a membranous conductor-like portion. Epigynal openings lead directly into a bursa connected by a very short thick-walled duct to a second chamber. But in Hakka there is no membranous part in the male palp, and the bulb and epigynal ducts, as described above, differ strikingly. Distribution. — Previously known from China, Japan, North Korea, and now, Hawaii. Etymology. — Named for a group of Chi- nese people, members of which were brought to Hawaii as laborers on sugar cane planta- tions in the middle of 19th century (described in the book “Hawaii” by James Michener). For nomenclatorial purpose the name is con- sidered to be female. Type species. — Menemerus himeshimensis (Doenitz & Strand, in Bosenberg & Strand 1906). Hakka himeshimensis (Doenitz & Strand) (in Bosenberg & Strand 1906) new combination Figs. 1-7 Note: The type specimens, housed in Stutt- gart, were destroyed during World War IT Menemerus himeshimensis Doenitz & Strand, in Bosenberg & Strand 1906: 395-396, table 8, fig. 116; table 14, fig. 309. Menemerus himeshimensis: Yaginuma 1970: 67; 1986a: 234, fig. 130.2. Icius himeshimensis: Proszynski 1976: map 105. Salticus koreanus Wesolowska 1981: 78, figs. 102- 105 (Female holotype from North Korea: Nam- pho, prov. Phyongan-namdo, deposited at Mu- zeum i Instytut Zoologii, PAN, Warsaw, Poland, examined). Icius himeshimensis: Bohdanowicz & Proszynski 1987: 66, 67, figs. 65, 66. Pseudicius himeshimensis: Proszynski 1987: 51 (transfer from Menemerus, Icius). Icius himeshimensis: Chikuni 1989: 151, fig. 22. Pseudicius himeshimensis: Peng, Xie & Xiao 1993: 191, 192, figs. 667, 670. Description. — Male: Measurements {n = 1): total length 6.98, length of eye field 1.32, height of cephalothorax 1.42, width of eye field at eyes I 1.80, width of eye field at eyes III 1.80, width of cephalothorax at eyes III 2.28, maximum width of cephalothorax 2.64, length of flat surface of cephalothorax 0.96, length of abdomen 3.72. Body and legs uni- formly dark brown, without any contrasting pattern. Cephalothorax relatively broad (broadest posteriorly) and low; eye field rect- angular, indistinctly shorter than broad, pos- terior sloping part of cephalothorax short. Covered with sparse, inconspicuous adpressed lighter setae; longer setae, now reddish, stand up diagonally beneath lateral eyes. There is no row of tubercles with spines beneath the lat- eral eyes. Eyes I large, the diameter of median eyes almost twice the size of the lateral eyes. Eyes I surrounded by colorless, slightly red- dish setae; setae above eyes longer; clypeus very narrow with inconspicuous sparse short setae, with a sparse row of brown setae over- hanging cheliceral bases. Chelicerae brown and robust; posterior margin with single con- ical tooth. Abdomen a flattened oval, as broad as cephalothorax and indistinctly longer, densely covered by lighter thin adpressed se- tae. Leg formula: LIV-III-II; but legs of ap- proximately equal length (leg I longest by about 20%); long and thin, their segments of similar width, with femora somewhat wider, but tibia I not broader than neighboring seg- ment and not shortened (in which it differs from Pseudicius). Spines inconspicuous, shorter than sparse upright, reddish setae on the same surfaces; anterior tibia with only ventral spines, on anterolateral edge two short spines located in the anterior one third of seg- BERRY & PROSZYNSKI— A HAWAIIAN SALTICID 203 Figures 1-7. — Hakka himeshimensis new genus. 1. Male, dorso-lateral view of cephalothorax; 2. Male abdomen, laterally; 3. Palpal organ, ventrally; 4. Palpal organ, laterally; 5. Female, general appearance; 6. Epigynum; 7. Internal structure of epigynum. ment, on posterolateral edge two short spines, normally spaced. Palpal organ with bulbus broad anteriorly and with anterior margin curved posteriorly (in which it resembles some Pseudicius of the cinctus group); em- bolus characteristic, elongate conical with wavy outline. Female: Resembles male in appearance and size (Fig. 5); difference from male in leg for- mula (IV-III-II-I) is a secondary sex character. 204 THE JOURNAL OF ARACHNOLOGY found in many genera of Salticidae. Charac- terized by epigynum in a form of a concave plate, with slit-like copulatory openings in the middle, located inside indistinct oval depres- sions, separated by a thin, low ridge (Fig. 6). Internal structures of epigynum consist of a channel running anteriorly, then curving and running back, slightly diagonal and joining the transversely oriented narrow bag-shaped spermathecae, located in the posterior half of epigynum. There is a long chimney-like struc- ture, presumed to be a scent gland pore (see Proszyhski 1998, in press), located at the junc- tion of channel and spermatheca. Walls of channels, spermatheca and scent gland scler- otized and of similar thickness. Interior walls of spermathecae with irregular, transverse ridges. Nutritive pores (see Proszyhski 1998, in press) minute and indistinct, located near the top of conical distal part of spermatheca, near insertion of fertilization channel (Fig. 7). General plan and appearance of these struc- tures superficially resemble those seen in var- ious species of Salticus. Distribution* — Japan, China and North Korea; this is the first record from Hawaii. Material examined, — Hakka himeshimensis {''Pseudicius'’' himeshimemsis], under stones, Necker Island, Hawaii, Id, 14 June 1923 (E.M. Bryan, Jr., AMNH). Hakka himeshimensis [labeled "^Salticus koreanus (Wesolowska 1981) s. Pseudi- cius koreanus'"], on black lava beach, Anaehoom- alu Bay, Hawaii County, Hawaii, 19, 15 February 1997 (J. & E, Berry), Hakka himeshimensis ["'Sal- ticus” koreanus (Wesolowska 1981) s. "Pseudicius koreanus”], among beach rocks near Nailoa, An- aehoomaiu beach, Hawaii County, Hawaii, 19, 17 February 1988 (J. & E. Berry). All specimens iden- tified by J. Proszyhski. ACKNOWLEDGMENTS We are grateful to Butler University for an academic grant to JWB, which permitted this work to be done. We are also grateful to the Indiana Academy of Science for support for travel. Dr. Joe Beatty was of immense help in writing the diagnosis of the genus. LITERATURE CITED Bohdanowicz, A. & J. Proszyhski. 1987. System- atic studies on East Palaearctic Salticidae (Ara- neae), IV. Salticidae of Japan. Annales Zoologici, Warszawa 41 (2):43-151, figs. 1-312, Bosenberg, W. & E, Strand, 1906. Japanische Spin- nen. Abhanglungen der Senckenbergischen Na- turforschenden Gesellschaft 30:93-422, pL IIL XVL Chikuni, Y. 1989. Some interesting Japanese spi- ders of the families Amaurobiidae, Araneidae and Salticidae. Arachnological papers presented to Takeo Yaginuma, Osaka. Pp. 133-152, figs. 1-17. Peng X., L. Xie & X. Xiao. 1993. Salticidae in China. Hunan Normal Univ. Press. 270 pp. 893 figs, [in Chinese, English summary]. Proszyhski, J. 1976. Studium systematyczno-zo- ogeograficzne nad rodzina Salticidae (Aranei) Regionow Palearktycznego i Nearktycznego. Ro- zprawy Wyzszej Szkoly Pedagogicznej. Siedlce: Pp. 1-260, 450 figs., 218 maps. Proszyhski, J. 1987. Atlas rysunkow diagnosty- cznych mniej znanych Salticidae 2. Zeszyty Nau- kowe WSRP, Siedlce, 172 pp., illustr. Proszyhski, J. 1998. Description of new species of Phlegra (Salticidae, Araneae) from Israel. Israel Journal of Zoology 44:159-185. Proszyhski, J. In press. Salticidae of the Levant. Fauna Palaestina: Arachnida. The Israel Acade- my of Sciences and Humanities. Jerusalem. Wesolowska, W. 1981. Salticidae (Aranei) from North Korea, China and Mongolia. Annales Zoologici (Warszawa) 36:45-83; 112 figs. Yaginuma, T. 1970. The spider fauna of Japan (re- vised in 1970). Bulletin of the National Science Museum (Tokyo) 13:639-701. Yaginuma, T. 1986. Spiders of Japan in Color, (new ed.). Hoikusha PubL Co., Osaka. Pp. 1- 350, figs. 1-135, Plate 65. Manuscript received 5 April 2000, revised 30 Jan- uary 200L 2001. The Journal of Arachnology 29:205-219 A REVISION OF THE AFROTROPICAL SPIDER GENUS PALFURIA (ARANEAE, ZODARIIDAE) Tamas Sziits: Department of Ecology, Jozsef Attila University, RO. Box 51. H-6722, Hungary Rudy Jocque: Section of Invertebrates, Royal Museum of Central Africa, B-3080 Tervuren, Belgium ABSTRACT. The African genus Palfuria Simon 1910 is revised. The genus now contains nine species: the type species Palfuria retusa Simon 1910, described on the base of single juvenile, P. gibbosa (Lessen 1936), P. partner Jocque 1991, and six species that are described as new: P. caputlari (69), P. harpago (d), P. helichrysomm ($), P. hir^iita (?), P. gladiator (d ?), P. spirembolus (d ?). The male of Palfuria panner is redesciibed, and ,the female described for the first time. Five species {P. retusa, P. spirembolus, P. gladiator, P. panner, P. harpago) are from the southwestern part of the continent, the other species (P. gibbosa, P. helichrysomm, P. hirsuta, P. caputlari) from the eastern part. The last species is from as far north as northern Tanzania. As in many other genera, there is a tendency for the embolus to increase in length. Both the most basal {Palfuria panner) and the most derived species {Palfuria spirembolus) are found in Namibia. Keywords: Cladistic analysis, complexity, new species Palfuria is a poorly-known genus, recorded only from the southern part of Africa. Its type species {Palfuria retusa) was described on the basis of a single juvenile specimen. Both ad- ditional described species {P. gibbosa and P. panner) were each known from one sex only. Since the revision of Jocque (1991), an im- portant number of specimens representing several new species has become available. The present paper treating these specimens shows that the diagnostic characters identified by Jocque (1991) remain valid; but there is a lot of variation in genitalic characters and, to a lesser degree, in somatic traits. Scanning electron micrographs of some important char- acters are provided, and the distribution of the genus is shown to extend much further north than was known previously. METHODS Male right palps were removed, examined and drawn with a Wild M5 stereomicroscope. Epigyna were removed and cleared in meth- ylsalicylate and temporarily mounted in a mixture of that medium and cedukoL They were observed and drawn with a Leitz Dialux 22 compound microscope. Scanning micro- graphs were made with a JEOL LV 5400 scanning microscope. All measurements are in millimeters. Abbreviations. — a = diameter of PME, b = diameter of RLE, c = diameter of AME, d = diameter of ALE, e distance between RME, f = distance between RME and RLE, g — distance between AME, h = distance be- tween AME and ALE. ALE — anterior lateral eyes, AME = anterior median eyes, AW = anterior width (of the MOQ), L — length of the median ocular quadrangle, MOQ = me- dian ocular quadrangle, RLE = posterior lat- eral eyes, PME = posterior median eyes, PW = posterior width (of the MOQ), PS = pos- terior spinnerets. Institutions. — MHNG = Musee d’Historie Naturelle, Geneve (B. Hauser, P. Schwendin- ger); MNHN = Museum National d’Histoire Naturelle, Paris (J. Heurtault & C. Rollard); MR AC = Musee Royal de FAfrique Centrale, Tervuren (R. Jocque); NMSA = Natal Mu- seum, Pietermaritzburg (P. Croeser, A. Rui- ters); NMZ = National Museum Zimbabwe, Bulawayo (M. Fitzpatrick ); SMNW — State Museum, Windhoek, Namibia (E. Griffin). TAXONOMY Palfuria Simon 1910 Palfuria Simon 1910: 188 (description new genus). Jocque 1987: 143; 1991: 141. Dippenaar Schoeman & Jocque 1997: 327. 205 206 THE JOURNAL OF ARACHNOLOGY Hermippella Lessert 1936: 226 (description new genus); 1938: 432 (formerly included in the Pal- pimanidae). Note: Jocque (1991) provisonally synony- mized Palfuria and Hermippella; this synon- ymy can now be considered as definitive. It is indeed found that in juveniles, and even in some females, that the cephalic lobe is only raised and not slanting back as in Palfuria re- tusa. Type species. — Palfuria retusa Simon 1910. Diagnosis. — Easily recognized by the strongly elevated cephalic part of the cara- pace, slanting back in adults (except P. spi- rembolus female); the abdomen has dorsolat- eral circumferential folds. The genus is part of a large unresolved clade (Jocque 1991) of genera with a femoral organ but the characters listed above unequivocally distinguish Palfur- ia from them. Heradida Simon 1893 is the only genus in that clade with abdominal cir- cumferential folds and must be considered the sister-group of Palfuria. Description. — (slightly modified after Jocque 1991: 141-142.) Small spiders (1.41- 3.4) with slightly to strongly granulated teg- ument. Carapace with strongly raised cephalic lobe, slanting back over the thoracic area in adults; widest between coxae III and IV; nar- rowed in front to about 0.75 X maximum width in females, to about 0.65 X maximum width in males. Color: Carapace and chelic- erae pale to dark brown. Sternum pale yellow to dark brown, often with a darker margin. Legs dark brown to a pale yellow, sometimes with dark stripes; coxae and trochanters pale yellow, femora slightly darker, other leg seg- ments paler. Abdomen pale to dark sepia on dorsum, pale on sides and venter. Eyes: In two strongly procurved rows (anterior one as seen in front, posterior one as seen from above). AME by far the largest up to 4 X diameter of other eyes), dark (except P. spirembolus), cir- cular. Other eyes pale, circular, though PME sometimes slightly ovoid. AME about half their diameter apart, about one diameter from PLE; these almost contiguous with ALE and AME. MOQ subquadrangular. Clypeus: Con- vex, high 3. 5-10 X as high as diameter of ALE. Chilum absent. Chelicerae: Short, fused; without lateral condyle; without teeth, but with cheliceral lamina (Fig. 4). Interchel- iceral triangle most often small. Endites roughly rectangular, strongly converging; with anteromesal scopula. Labium triangular. Ster- num as wide as long in females, longer than wide, slightly rebordered in males. Legs: For- mula 4123. More slender in males than in fe- males. Two claws on short onychium; with 2- 4 teeth, third claw tiny; no claw tufts but spi- niform scopulae present. One dorsal spine in proximal half of femora. Leg segments gen- erally covered with flattened incised hairs (Figs. 1, 2), but femora with 2-4 long rigid hairs, (for example: in P. gladiator, P. hirsu- ta). Femoral organ with 1 or 2 barbed hairs (Fig. 1). Patellae with proximal ring-shaped crack (see Jocque & Dippenaar-Schoeman 1992, fig. 5). Abdomen: Rounded, hardly lon- ger than wide; slightly sclerotized on dorsum in females, more strongly so in males; anterior part of abdomen strongly sclerotized, forming tube around the petiolus; with a number of parallel shallow, circumferential folds. Two spinnerets in males, 4 spinnerets in females, PS minute. Colulus represented by broad field with short setae; a number of modified hairs in front of tracheal spiracle (Fig. 3); spiracle wide with anterior rim sclerotized. Male palp: (Figs. 5-15): Tibia with one or two slender lateral apophyses. Cymbium with distal filed of short hairs and one or two dorsolateral modified hairs. Embolus originating on pos- terior part of tegulum (except P. spirembolus), curved, relatively short. Tegular apophysis fairly short, sometimes bifurcate. Female pal- pus: With finely pectinated claw. Epigynum: (Figs. 16-29) Very simple to relatively com- plex, poorly sclerotized except in P. helichry- sorum. Distribution. — Africa south of 4°S: found in Tanzania, Namibia, Malawi, Zambia, Moz- ambique, South Africa. SZUTS & JOCQUE— REVISION OF PALFURIA 207 KEY TO THE SPECIES Note: Palfuria retusa Simon is not included since it is known only from the juvenile. 1. Males 2 Females 6 2. Embolus long, looped around the tegulum (Figs. 14, 15). Palfuria spirembolus Embolus very short (Figs. 5-10, 12, 13) ........................................... 3 3. Palpal tibia with two apophyses, one dorsal, one retrolateral (Figs. 6, 10-1 1) ................................................ . 4 Palpal tibia with only one apophysis (Figs, 8, 13) .................................... 5 4. Dorsal apophysis almost straight (Fig. 6). ............................. . .Palfuria caputlari Dorsal apophysis harpoon shaped; slightly curved, pointed, with a branch pointing backwards, ending in a few fine, hair like ramifications (Figs. 10, 11). ................. . Palfuria harpago 5. Tibial apophysis straight (Fig. 8). ................. .Palfuria gladiator Tibial apophysis curved (Fig. 13) ...... Palfuria partner 6. Epigynum with well delimited plate (Figs. 16, 17, 19, 20). 7 Epigynum without a plate, but with sclerotized posterior margin (Figs. 18, 21, 22) .......... 10 7. Epigynal plate of different shape, with posterior margin sinuous and indented in the middle (Fig. 19) ....................................................... .Palfuria helichrysorum Epigynal plate ellipsoid (Figs. 16, 17, 20) .......................................... .8 8. Entrance openings situated near posterior margin of epigynum. Spermathecae under plate, lateral margins of epigynum plate angular (Fig. 17). ............ .Palfuria gibbosa Entrance openings — if visible — nearer to anterior margin of the epigynum; spermathecae at var- iable distance from epigynal plate, but never under it. Lateral margins of epigynum plate evenly rounded (Figs. 16, 20). 9 9. Internal structure of epigynum relatively complex, sperm ducts long and wound (Fig. 27) ..... .Palfuria hirsuta Internal structure of epigynum simple, sperm ducts short and slightly curved (Fig. 23) ...... . .............................................................. Palfuria caputlari 10. Epigynum with only a simple, small, sclerotized posterior margin (Fig. 18). ... . .Palfuria gladiator Epigynum with differently shaped sclerotized parts 11 11. Sclerotized margin of the epigynum straight, situated posteriorly, spermathecae rounded; atria large (Figs. 21, 28) Palfuria panner Posterior margin of epigynum accolade shaped, spermathecae oval; atria small, glandular organ present (Figs. 22, 29) Palfuria spirembolus Palfuria caputlari new species Figs. 1-6, 16, 23, 30, 31 Holotype. — Male, Tanzania, Mkomazi Game Res., Ibaya camp, Nov. 1994, Russell- Smith (MRAC 202.528). Paratypes. — 3 $14(3 together with holo- type. Diagnosis.- — ^Males of Palfuria caputlari are easily identified by the long, slender dor- solateral tibial apophysis of the palp and the hook-shaped median apophysis (Figs. 5-6). Females are recognized by the epigynal plate which is much wider than long (length/width 0.3) (Fig. 16). The male shows superficial re- semblance with P. harpago with which it shares the presence of two tibial apophyses; in the latter species the dorsolateral apophysis is harpoon-shaped; the female is similar to P. hirsuta but lacks the long entrance ducts of that species. The sister- species of P. caputlari is P. harpago. Etymology.— The species name is com- posed of two Latin nouns: caput (head) and lari (gen. of Larus: gull), referring to the shape of the median apophysis as seen from the side. Male. — Total length 2.24 (2.24-2.35); car- apace 1.12 long (1.12-1.32), 0.82 wide (0.77- 0.91). Color: Carapace medium to dark brown, with some faint, darker striae in tho- racic area. Cephalic lobe dark brown, with some paler spots. Eye field dark brown. Che- licerae medium brown, fangs yellow, chelic- eral lamina white, sternum medium brown, sometimes with dark margin; legs paler: coxae pale yellow, femora dark brown, other leg segments pale yellow. Abdomen: dorsum dark sepia with yellow folds, contrasting with pale yellow venter. Branchial operculum dark yel- 208 THE JOURNAL OF ARACHNOLOGY Figures 1-4. — Palfuria caputlari, male from Mkomazi Game Reserve. 1, Femoral organ, leg I; 2, Position of femoral organ on right femur I; 3, Modified hairs in front of spinnerets; 4, Cheliceral lamina. low. Carapace: (Figs. 30, 31): Tegument slightly granulated. Chelicerae: setae of che- liceral lamina curved, subequal. Abdomen: Circumferential folds not conspicuous, modi- hed hairs in front of spinnerets stout. Eyes: a: 0.06; b: 0.06; c: 0.13; d: 0.06; e: 0.13; f: 0.14; g: 0.05; h: 0.04; MOQ: AW = 1.35 PW; AW = 1.25 L. Clypeus: 0,39 or 6.5 X diameter of ALE. Legs: All segments covered by flattened incised hairs. Two dorsal spines and three long ventral rigid hairs on all femora. Male palp: (Figs. 5, 6). Tibia with two apophyses; one dorsal, one retrolateral; dorsal apophysis long, thin, almost straight; retrolateral apophysis medium sized, wider; median apophysis pointed, hook shaped; embolus short, blunt. Female* — Total length 2.44 (2.31-2.44); carapace 1.42 long (1.22-1.42), 1.02 wide (0.91-1.02). Color: Carapace medium to dark brown, with some darker striae in thoracic area. Cephalic lobe dark brown, with some paler spots. Chelicerae medium brown, fangs yellow, cheliceral lamina white, sternum pale brown, with dark margin, legs paler: coxae pale yellow, femora dark brown, paler on the ventral side, other leg segments pale yellow. Abdomen: dorsum dark sepia with yellow folds, contrasting with pale yellow venter. Branchial operculum dark yellow. Carapace: Hair cover slightly denser than in males. Ab- domen: Circumferential folds poorly marked, modified hairs in front of spinnerets stout. Epigynum: (Figs. 16, 23). With narrow plate, with dark posterior margin; internal structure of epigynum simple: sperm ducts short, al- most straight, spermathecae rounded. Distribution. — -Only known from type lo- cality. Palfuria gibbosa (Lessert) Figs. 17, 24 Hermippella gibbosa Lessert 1936: 226 (description female); 1938: 432. Palfuria gibbosa: Jocque 1991: 142. Holotype. — Female, Mozambique, Nova Choupanga (? near Chupanga 18°05'S, 35°35'E) (MHNG) (examined). Diagnosis. — The females of Palfuria gib- SZUTS & JOCQUE— REVISION OF PALFURIA 209 Figures 5-11. — Male palps. 5, 6. Palfuria caputlari from Mkomazi Game Reserve. 5, Ventral view; 6, Retrolateral view, 7, 8. P. gladiator, holotype. 7, Ventral view; 8, Retrolateral view. 9-11. P. harpago, holotype; 9, Ventral view; 10, Retrolateral view; 11, Detail of dorsolateral apophysis, dorsal view. bosa can be recognized by the shape of the epigynal plate, the entrance openings near the posterior margin of the plate and the presence of glands. The epigynum vaguely resembles that of P. helichrysorum but lacks the poste- rior median indentation; the epigynal plates of P. caputlari and P. hirsuta both have rounded lateral margins and a sclerotized posterior rim. The closest relatives of P. gibbosa are P. spi- rembolus and P. hirsuta. 210 THE JOURNAL OF ARACHNOLOGY Figures 12-15. — Male palps. 12, 13. Palfuria panner, SMNW42872; 12, Ventral view; 13, Retrolateral view; 14, 15. P. spir embolus, holotype; 14, Ventral view; 15, Retrolateral view. (E = embolus; MA = median apophysis; MH = modified hair; ST = subtegulum; T = tegulum). Female. — Total length 2.30, carapace 1.10 long, 0.64 wide. Color: Carapace medium brown in cephalic area, pale brown in thoracic area; chelicerae medium brown. Sternum yel- low. Legs pale yellow. Abdomen greyish-yel- low on sides and venter. Carapace: (see Jocque, 1991 figs. 354-356). With raised ce- phalic lobe slanting back over thoracic area. Eyes: a: 0.06; b: 0.06; c: 0.1; d: 0.08; e: 0.12; f: 0.07; g: 0.06; h: 0.03. MOQ: AW = 1.04 SZUTS & JOCQUE— REVISION OF PALFURIA 211 Figures 16”22.--”Epigyna, ventral view. 16, Palfuria caputlari from Mkomazi Game Reserve; 17, P. gibbosa, holotype; 18, P. gladiator, paratype; 19, F. helichrysorum, holotype; 20, P. hirsuta, holotype; 21, F. partner from Windhoek; 22, P. spirembolus from Kokerboom forest. PW; AW = LOO L. Legs: only leg II complete. Epigynum: (Figs. 7, 24). Entrance openings situated near posterior margin, spermathecae under epigynal plate, provided with angular lateral margin. Male.— "Unknown. Distribution,' — -Only known from type lo- cality. Palfuria gladiator new species Figs. 7, 8, 18, 25 Holotype. — -Male, Namibia, Karossfontein 19°2rs, 14°3rE, 7 Oct-14 Nov. 1986, pitfall traps, E. Griffin (SMNW 39751). Paratypes. — Namibia: 2d together with holotype; d from Windhoek, wasteland near houses, 14-31 Oct. 1987 pitfall traps, R. Jocque (MRAC 168.421); 6d 1? from Da- maraland, Hobatere Campsite, 19°8'S, 14°7'E, 23-30 April 1996, pitfall traps, E. Griffin (SMNW 43540; Id in MRAC); d and 9 from Hobatere Campsite, 3.2 km from gate, 19°9'S, 14°7'E, 7-17 May 1991, pitfall traps (SMNW 42632); 9 from Wolfsnes, 19°03'S, 15°52'E, 24 March-10 May 1988, pitfall traps, E. Griffin (SMNW 40890). Diagnosis. — -Representatives of this species can be recognized by the strongly granulated 212 THE JOURNAL OF ARACHNOLOGY Figures 23-29. — Epigyna, cleared, dorsal view. 23, Palfuria caputlari from Mkomazi Game Reserve; 24, P. gibbosa, holotype; 25, P. gladiator, paratype; 26, P. helichrysorum, holotype; 27, P. hirsuta, holotype; 28, P. partner from Windhoek; 29, P, spirembolus from Kokerboom forest. tegument of the carapace and by the two (one dorsal, one ventral) long, rigid hairs on tibia lUIII. Males of Palfuria gladiator are char- acterized by the big cymbial claw and the al- most straight palpal tibial apophysis. The fe- males can easily be identified by the epigynum, appearing as a short, sclerotized, transverse line. Males and females are super- ficially similar to those of P. partner, the clos- est relative, but in that species the male palpal tibial apophysis is turned upwards and in the female there is a slight depression in front of the sclerotized epigynal rim. Etymology. — The species name is a noun in apposition and refers the shape of the male carapace and the big tarsal claw on the male palp. Male. — Total length 2.04 (1.41-2.04);, car- apace 1.06 (0.75-1.06) long, 0.71 (0.56-0.71) wide. Color: Carapace dark brown; cephalic area much darker, thoracic area paler, with some darker striae, cephalic lobe of carapace dark brown, lateral part of carapace paler, con- trasting with dark top. Chelicerae brown, fangs yellow, cheliceral lamina poorly devel- oped, sternum pale yellow with dark margin; legs paler: coxae pale yellow, femora I-II dark brown, femora III-IV dark brown on dorsal side, paler on ventral side, tibiae dark yellow, other leg segments pale yellow. Abdomen SZUTS & JOCQUE— REVISION OF PALFURIA 213 dorsum dark sepia with paler folds, venter pale yellow, with contrasting boundary be- tween them on sides. Branchial operculum pale brown. Carapace: Tegument strongly granulated. Chelicerae: Setae on cheliceral lamina small, poorly developed in males. Base of fangs strongly granulated, with long setae. Sternum: with many, fine hairs. Abdomen: Dorsum with some strong hairs. Modified hairs in front of spinnerets strong. Eyes: a: 0.05; b: 0.05; c: 0.11; d: 0.07; e: 0.12; f: 0.10; g: 0.04; h: 0.01; MOQ: AW = 1.18 PW; AW = 1.04 L. Clypeus: 0.3 or 4.2X diameter of ALE. Legs: All segments covered with flat- tened incised hairs. One dorsal spine, three ventral rigid hairs on all femora, one dorsal, one ventral rigid hair on tibia II-III. Male palp: (Figs. 7, 8). Tibia with one almost straight; medium sized prolateral apophysis; median apophysis pointed, hook-shaped; em- bolus long, blunt. Female. — Total length 2.47 (2.04-2.27); carapace 1.03 long (0.9-1.03), 0.92 wide (0.75-0.92). Color: Carapace dark brown, with some darker striae in thoracic area, which paler. Cephalic lobe dark brown. Chelicerae brown, fangs yellow, cheliceral lamina white; sternum pale brown, with dark margin; legs paler: coxae pale yellow, femora dark brown, paler on ventral side, other leg segments pale yellow. Abdomen: dorsum dark sepia with yellow folds, contrasting with pale yellow venter. Branchial operculum dark yellow. Car- apace: Hair cover slightly denser than in males. Abdomen: Circumferential folds not conspicuous; some stout modified hairs in front of the spinnerets. Epigynum: (Figs. 18, 25). Simple; with short, transverse sclerotized line. Internal structure of epigynum similar to that of Palfuria panner, but fertilization ducts turned upward. Distribution. Known only from Namibia. Palfuria harpago new species Figs. 9-1 1 Holotype. — Male, Namibia, Ovamboland, 10 km SE Etunda, 17°26'S, 14°33'E, 20 July- 9 August 1989, pitfall traps, E. Marais (SMNW 41413). Paratype. — 1(3 from Namibia, Ovambo, Mahanene Agric. Res. Sta., 17°26'S, 14°47'E, 5 October-5 December 1993, pitfall traps, B. Wohlleber (SMNW 43396). Diagnosis. — Males of Palfuria harpago are easily identified by the shape of the dorsolat- eral tibial apophysis: almost straight, pointed and with a branch pointing backwards, ending in a few, fine hair-like ramifications. Palfuria caputlari is the only other Palfuria with two palpal tibial apophyses; in P. caputlari, how- ever, the dorsal one is long, straight and spine- shaped. The sister-species of P. harpago is P. caputlari. Etymology. — The species name is a noun in apposition {harpago, Latin for harpoon) re- ferring to the shape of the dorsal tibial apoph- ysis as seen from the dorsolateral side (Fig. 11). Male. — Total length 1.81 (1.81-1.98); car- apace 1.13 long (1.03-1.22), 0.92 (0.66-0.92) wide. Color: Carapace medium to dark brown, with some faint, darker striae in tho- racic area. Cephalic lobe pale brown with dark margin. Eye field dark brown. Chelicerae me- dium brown, fangs dark brown, cheliceral lamina white; sternum pale brown, without darker margin; legs pale brown or yellow. Ab- domen: dorsum shiny, dark sepia with pale circumferential folds, venter dark yellow, con- trasting with dark sides. Branchial operculum dark yellow. Carapace: Tegument slightly granulated on cephalic lobe. Chelicerae: Setae of cheliceral lamina curved, and subequal. Ab- domen: Modified hairs in front of spinnerets fine and long, but few. Ventral side of abdo- men with many hairs. Eyes: a: 0.05; b: 0.05; c: 0.11; d: 0.09; e: 0.19; f: 0.05; g: 0.04; h: 0.02; MOQ: AW - 1.18 PW, AW - 1.36 L. Clypeus: 0.32 or 3.5 X diameter of ALE. Legs: All segments covered with flattened incised hairs. Femora with one dorsal spine and cover of ordinary hairs. Male palp: (Figs. 9, 10). Tibia with two apophyses; one ventral, one dorsolateral. Ventral apophysis short and wide, slightly curved, dorsolateral apophysis long, pointed, harpoon- shaped, with back- pointing branch ending in few thin ramifica- tions. Median apophysis strongly curved, bi- fid, ending in two pointed tips; embolus short, wide, subtegulum present, hidden under cym- bium. Female. — -Unknown. Distribution. — Only known from Ovam- boland, Namibia. Palfuria helichrysorum new species Figs. 19, 26 Holotype.— Female, Malawi, Mt. Mulanje, Lichenya plateau (2000 m), near CCAP hut. 214 THE JOURNAL OF ARACHNOLOGY 15°59'S, 35°32^E, 9 November 1981, under Helichrysum, R, Jocque (MRAC 156.781). Diagnosis. — Palfuria helichrysorum fe- males are recognized by the sclerotized epi- gynum and the shape of the central plate with two frontal lobes covering the entrance open- ings, and indented posterior margin, and by the internal structure of the epigynum with short, thick- walled sperm ducts. The other species with an epigynal plate, P. caputlari, P. gibbosa and F. hirsuta lack the posterior indentation. P. helichrysorum is the sister-tax- on of a group of three species comprising F. hirsuta, P. spirembolus and F. gibbosa. Etymology. — The specific name is derived from Helichrysum, a rosette bearing Astera- ceae, ideal retreat for night active spiders. Female. — Total length 3.06; carapace 1.32 long 0.98 wide. Color: Carapace dark brown, with some darker striae in thoracic area. Ce- phalic lobe dark. Chelicerae brown, fangs yel- low, cheliceral lamina white, sternum pale brown, with wide, dark margin; legs paler: coxae yellow, femora dark brown, femora LII paler on ventral side, other leg segments pale yellow, contrasting with dark femora. Abdo- men: dorsum dark sepia with yellow, circum- ferential folds; venter pale yellow, contrasting with dark sides. Branchial operculum brown. Carapace: Finely granulated. Chelicerae: Se- tae of cheliceral lamina straight, unequal in length. Abdomen: Dorsum with few fine hairs. Modified hairs in front of spinnerets fine. Eyes: a: 0.07; b: 0.07; c: 0.12; d: 0.07; e: 0.16; f: 0.08; g: 0.08; h: 0.08; MOQ: AW = 1.06 PW; AW - 1.15 L. Clypeus: 0.34-4.8X di- ameter of ALE. Legs: Covered with flattened incised hairs. Femora with one dorsal spine and three long ventral rigid hairs. Epigynum: (Figs. 19, 26). Well-sclerotized; central plate with two anterior lobes covering entrance openings, posterior margin indented. Internal structure of epigynum quite simple with short, thick walled sperm ducts. Male. — U nkno wn . Distribution. — Only known from type lo- cality. Palfuria hirsuta new species Figs. 20, 27 Holotype. — Female, Zambia, Wildlives Game Farm, 16°52'S, 26°37'E, B.FA. Study Plot, 8-14 Dec. 1994, E Nyathi (NMZ/ A11862). Diagnosis. — The female of Palfuria hirsuta is recognized by the large epigynal plate with clearly sclerotized posterior rim, and the in- ternal structure of the epigynum with long and winding sperm ducts, but lacking a glandular organ. In the other species with an epigynal plate the shape is clearly different (F, gibbosa; P. helichrysorum) or the entrance ducts are much shorter (F. caputlari). Palfuria hirsuta is the sister species of F. spirembolus and F. gibbosa. Etymology. — The species name refers to the hairy appearance. Female. — Total length 2.32; carapace 1.16 long, 0.85 wide. Color: Carapace brown; ce- phalic area dark, thoracic area paler, with some faint darker striae, cephalic lobe very dark. Chelicerae dark brown, fangs yellow, cheliceral lamina white, sternum yellow, with dark margin, anterior part of sternum darker; legs darker: coxae yellow, femora brown with darker sides, other leg segments slightly paler. Abdomen: sepia on dorsum, with yellow folds, pale yellow on venter, but dorsal dark area narrow. Pale spots on sepia background rounded or irregular. Carapace: Slightly gran- ulated, with many fine hairs. Chelicerae: lam- ina with two straight setae of different length. Sternum: Sternum with fine hairs on anteri- or— darker — part. Abdomen: Dorsum with many fine hairs. Modified hairs in front of spinnerets strong. Eyes: a: 0.06; b: 0,07; c: 0.1; d: 0.06; e: 0.06; f: 0.08; g: 0.07; h: 0.04; MOQ: AW - 1.17 PW; AW = 0.96 L. Clyp^ eus: 0.35-5.8 X diameter of ALE. Legs: Seg- ments covered with flattened incised hairs, but femora, patella, tibia with many rigid hairs. Leg spination: One dorsal spine on all femora, long rigid hairs on femora, patella, tibia, but none on tarsi, metatarsi. Epigynum: (Figs. 20, 27). With simple ellipsoid plate. Internal structure of epigynum complex: sperm ducts long and intricately wound. Male* — -Unknown. Distribution.— Only known from type lo- cality. Palfuria panner Jocque Figs. 12, 13, 21, 28 Palfuria panner Jocqu6 1991: 142 (descrip- tion male, figs. 359-363). Holotype. — Male, Namibia, Panner Gorge, 22°19'S 15°01'E, 11 March-9 April 1985, J. Irish and H. Rust (SMN 38730). SZUTS & JOCQUE— REVISION OF PALFURIA 215 Figures 30-33. — Carapace. 30, 31. Palfuria caputlari, male from Mkomazi Game Reserve. 30, Cara- pace, lateral view; 31, Frontal view. 32, 33. Palfuria spirembolus from Kokerboom forest. 32, Male carapace, lateral view; 33, Female carapace, lateral view. Other material examined. — NAMIBIA :1(3 from Otjiwarongo district, Waterberg Plateau Park, 20°24'S, 17°23'E, 18 May-24 April 1991, pitfall traps, M. Push (SMNW 42465); IS from sand dunes east of Jakkalsputz, SE 2214 Ab, 17-23 April, pitfall traps, 1994, E. Griffin (SMNW 43229); Id from Windhoek district, Richthofen 126, 22°15'S, 17°30'E, 1-31 Oct. 1979, pitfall traps, M.-L. Pentith (SMNW 42872); 1 9 from Win- dhoek, in trunks and leaves of dead Aloe, 15 Oct. 1987, R. Jocque (MRAC 168.482); 1 $ from Frans- fontein 2015 AA, 22. Feb. 1969, B. Lamoral & R. Day (NMSA); 1 subadult S from Liideritz district, 29°59'S, 16°14'E, 22 Nov. 1995, under stones, E. Griffin (SMNW 43479). Diagnosis. — Males of Palfuria panner can be recognized by the upward curved palpal tibial apophysis (Fig. 13) and simple fairly long median apophysis. The females can be recognized by the shape of the shallow epi- gynal depression in front of a sclerotized ridge and the large atria in the epigynum (Fig. 28). The only other species with a simple retrola- teral tibial apophysis is P. gladiator, but in that species the tibial apophysis is almost straight. The female of P. gladiator has a sclerotized line but lacks the depression in front of it. Palfuria panner is closely related to P. gladiator. Note: In the holotype, the tip of the tibial apophysis is broken off; the drawing in Jocque (1991, fig. 361) does not give the normal shape of this apophysis which is here corrected. 216 THE JOURNAL OF ARACHNOLOGY Figure 34. — Distribution map of Palfuria spe- cies, • = F. caputlari; ■ = F. gibbosa; A = P. gladiator; o — p, harpago; ♦ = P. helichrysoriim; A = P. hirsuta; □ P. panner; 0 = P. retusa; ★ = P. spirembolus. Male* — Total length L82 (1.69-2.0); cara- pace 0.90 (0.73-0.98) long, 0.64 (0.58-0.64) wide. Color: Carapace dark brown in cephalic area, medium brown with darker striae in tho- racic area cephalic lobe pale brown, with some pale spots. Eye field darker. Chelicerae medium brown; sternum shiny dark brown; legs dark brown. Abdomen sepia on dorsum and sides, pale yellow on venter. Branchial operculum medium brown. Carapace: Slight- ly granulated. Cephalic lobe low. Abdomen: Circumferential folds well developed. Modi- fied hairs in front of spinnerets stout. Eyes: a: 0.05; b: 0.06; c: 0.09; d: 0.05; e: 0.10; f: 0.04; g: 0.05; h: 0.01; MOQ: AW = 1.09 PW; AW: 1.00 L. Clypeus: 0.26-5. 2 X diameter of ALE. Legs: Segments covered with flattened incised hairs. Femora with one dorsal spine and three ventral, rigid hairs; tibiae with one ventral rig- id. Male palp: (Figs. 12, 13). Cymbium with two modified hairs and one spine. Tibial apophysis curved upward. Median apophysis pointing inward, hook shaped. Female. — Total length 2.23; carapace 1.28 long, 0.92 wide. Color: Carapace brown; ce- phalic area darker, thoracic area pale, with some dark striae. Chelicerae brown, fangs yel- low, cheliceral lamina white, sternum yellow with narrow dark margin; legs paler: coxae yellow; femora dark brown, patellae yellow, tibiae dark yellow, with few brown rings, oth- er leg segments much paler. Abdomen: dor- sum dark sepia with yellow stripes, venter pal- er, contrasting with darker sides. Branchial operculum pale brown. Carapace: Tegument slightly granulated. Cheliceral lamina with two hairs; one stout, short, one finer and lon- ger. Sternum with fine hairs. Abdomen: Dor- sum with few stout hairs. Modified hairs in front of spinnerets stout and strong. Epigyn- um: (Figs. 21, 28) With sclerotized margin. Incurved, anterior edge with many, fine hairs. Internal structure: openings funnel-shaped, sperm ducts short, spermathecae thick-walled. Fertilization ducts curved downwards. Distribution. — Only known from Namibia. Palfuria retusa Simon Palfuria retusa Simon 1910: 188 (description juv. female); Jocque 1991: 142 (figs. 352, 353). Holotype. — Juvenile female. South Africa, Namaqualand, Steinkopf, Shultze (MNHN 1573) (not examined). Diagnosis. — Recognized by the dark stripes on the femora. Since this species is only known from a juvenile it is not possible to discuss its affinities. Subadult female. — Total length: 1.98; car- apace 1.00 long, 0.72 wide. Color: Carapace pale brown with dark margin. Chelicerae pale brown, sternum pale yellow, legs pale yellow: femora with dark stripes. Abdomen dorsum pale sepia with pale stripes in back, remainder cream. Carapace: Finely granulated; cephalic area raised, but not slanting back. Abdomen: Almost globular; parallel circumferential folds poorly marked. Adults. — Unknown. Distribution. — Only known from type lo- cality. Palfuria spirembolus new species Figs. 14, 15, 22, 29, 32, 33 Holotype.— d, NAMIBIA: Keetmanshoop district, Khabus 146, on doleritia hill, east slope, 26°17'S, 18°14^E, 1 Oct.-8 Dec. 1988, pitfall traps, N.G. Olivier (SMNW 42286). Paratypes.— NAMIBIA: 1 S and a juvenile together with holotype; 1 S from Keetman- shoop district, Dassiefontein 87, 27°13'S, 18°35'E, 7-27 Nov. 1992, pitfall traps, E. Marais (SMNW 42767); 1 $ from Kokerboom forest, 26°28^S 18U4'E, 16 Oct 1984, under stones, E. Griffin (SMNW 43179); 1$ from Marieetal district, Berseba 170, 25°12'S, 18°03^E, 7-29 Nov. 1992, pitfall traps, E. Marais (SMNW 42880). SZUTS & JOCQUE— REVISION OF PALFURIA 217 Table 1. — Character matrix for species of Palfuria and the outgroups Heradida and Diores. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Diores 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Heradida 0 1 1 0 0 0 0 0 0 0 0 0 0 0 0 P. caputlari 1 1 0 0 1 0 1 0 1 0 2 0 0 0 0 P. gibbosa 1 1 0 ? ? ? ? ? ? ? 2 0 1 1 1 P. gladiator 1 1 0 0 0 0 1 0 1 1 0 0 0 0 0 P. harpago 1 1 0 0 1 0 1 0 2 1 ? ? ? ? ? P. helichrysorum 1 1 0 ? ? ? ? 7 ? ? 2 0 0 0 1 P. hirsuta 1 1 0 ? ? ? ? ? ? ? 2 1 1 0 1 P. panner 1 1 0 1 0 0 0 0 0 0 0 0 1 0 0 P. spirembolus 1 1 0 2 0 1 2 1 3 1 1 1 1 1 2 Note: The males and the females are ten- tatively attributed to the same species, because both sexes were found in Keetmanshoop dis- trict. Diagnosis. — Males of Palfuria spirembolus are easily identified by the long, slender em- bolus, the complex median apophysis and by the long carapace. Females are recognized by the accolade shape of the sclerotized rim of the epigynum, and by the internal structure of the epigynum: glandular organ present, sperm ducts long and wound, spermathecae oval. Certain of the characteristics of the secondary genital organs of this species are unique and exclude confusion with other species. Palfuria spirembolus appears to be closely related with P. hirsuta and P. gibbosa. Etymology. — The species name is a con- traction of spira (Latin for spiral) and embo- lus, referring to the long large embolus. Male. — Total length 2.22 (2.15-2.45); car- apace 1.22 long (1.03 1.47), 0.88 (0.83-0.91) wide. Color: Carapace medium to pale brown, with some darker, striae in thoracic area. Ce- phalic lobe pale brown. Eye field pale brown. Anterior part of carapace dark brown. Chelic- erae medium brown, fangs dark yellow, che- liceral lamina white, sternum pale brown, with narrow darker margin; legs paler: femora pale brown, other leg segments yellow. Abdomen: dorsum shiny, dark sepia with pale circumfer- ential folds, venter medium brown, contrast- ing with dark sides. Branchial operculum dark brown. Carapace: (Fig. 32). Tegument slight- ly granulated: cephalic part of carapace finely granulated, cephalic lobe with stronger gran- ulations. Chelicerae: Setae of cheliceral lam- ina curved, subequal. Abdomen: With scutum, modified hairs in front of spinnerets fine and long. Eyes: All eyes pale, a: 0.03; b: 0.07; c: 0.13; d: 0.05; e: 0.15; f: 0.11; g: 0.05; h: 0.02; MOQ: AW = 1.45 PW, AW = 1.45 L. Clyp- eus: 0.50-1 OX diameter of ALE. Legs: All leg segments covered by flattened incised hairs. Femora with two dorsal spines, three long, rigid, ventral hairs. Male palp: (Figs. 14, 15). Tibia with one apophysis; median apoph- ysis pointed, funnel shaped; embolus long, slender, subtegulum present. Female. — Total length 2.45; carapace 1.22 long (1.22-1.47), 0.84 wide (0.84-0.91). Col- or: Carapace medium brown, with some dark- er striae in thoracic area. Cephalic lobe brown. Eye field dark brown. Chelicerae medium brown, fangs yellow, cheliceral lamina white, sternum pale brown, with darker margin, legs paler: femora brown, other leg segments yel- low. Abdomen: dorsum dark sepia with pale circumferential folds, venter pale yellow, con- trasting with dark sides. Branchial operculum yellow. Carapace: (Fig. 33). Slightly granu- lated, cephalic area raised, but not slanting back. Abdomen: Circumferential folds not conspicuous, modified hairs in front of spin- Table 2. — Character statistics for consensus tree with length 24, ci 0.87 and ri 0.83. Character 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 steps 1 1 121 1213222222 ci (X 100) 100 100 100 100 100 100 100 100 100 50 100 50 50 50 100 ri (X 100) 100 100 100 100 100 100 100 100 100 50 100 0 66 0 100 218 THE JOURNAL OF ARACHNOLOGY Figure 35. — Cladogram (calculated with Hennig86): strict consensus tree of two trees with length 24, consistency index 0.87 and retention index 0.83, (prepared with WINCLADA 0.99.9, Nixon 1999) under DELTRAN optimization. Numbers indicate characters (above branch) and states (under branch). Black circles: unique gains; white circles: homoplastic gains or reversals. nerets fine and long, Epigynum: (Figs, 22, 29). With sclerotized, accolade-shaped line near posterior margin; internal structure complicat- ed: with glandular organ, sperm ducts long and intricately wound. Spermathecae oval. Distribution. — Only known from Namibia near 18°E, and between 24°-28°S. CLADISTIC ANALYSIS According to the cladogram presented in Jocque (1991) Palfuria is part of an unre- solved clade comprising several Zodariinae with femoral organ and a number of other characters (flattened leg setae, absence of leg spines, presence of patellar crack) which make this clade very robust. Among these, Herad- ida and Palfuria are the only genera with ab- dominal circumferential folds in at least some of the species. This is clear from the drawings in Jocque (1987, fig. 4) which show the pres- ence of these abdominal folds, a synapomor- phy of Heradida and Palfuria. The fact that the genera share a large part of their distri- bution area further supports the assumption that Heradida is the sister-taxon of Palfuria. It is here used as one of the outgroups. The other one is Diores Simon 1893, which is the sister-group of the former clade plus Acanthi- nozodium Denis 1950. The following 15 char- acters were used to analyze the relationships among the species of Palfuria: 1: ’Cephalic lobe’ [0] - not raised; [1] - raised; 2: ’Abd- omen’ [0] - without circumferential folds; [1] - with circumferential folds; 3: ’femoral or- gan’ [0] - no deep alveolus; [1] - single mod- ified hair in deep alveolus; 4: ’retro- lateral tibial apophysis’ [0] - simple, tapered, almost straight process; [1] - srongly curved process; [2] - with broad base, broadly fused to segment; 5: ’dorsal tibial apophysis’ [0] - absent; [1] - present; 6: ’embolus’ [0] - short, rigid; [1] - long, flexible; 7: ’origin of embo- lus’ [0] - far in front on tegulum; [1] - on posterior part of T, base directed retrolaterad; [2] - on prolateral part of T, base directed backwards; 8: ’Tegular swelling near base of embolus’ [0] - absent; [1] - present; 9: ’me- dian apophysis’ [0] - hook-shaped; [1] - slightly curved; [2] - bifid; [3] - complex; 10: ’subtegulum’ [0] - small, invisible on unex- panded palp; [1] - large, visible on unexpand- ed palp; 11: ’epigynum’ [0] - with poorly de- veloped transverse ridge; [1] - with sclerotized transverse ridge; [2] - with plate; 12: entrance ducts’ [0] - short (< 3X diameter spermathecae); [1] - long (> 3X diameter spermathecae); 13: ’atria’ [0] - absent; [1] - present; 14: ’glandular organ’ [0] - absent; [1] - present; 15: ’spermathecae’ [0] - spherical; [1] - narrowed towards centre; [2] - longer than wide. The character-matrix is given in Table 1. Trees were calculated with Hennig86 (Far- ris 1988) and command ie* and with NONA (Goloboff 1994) with mult* 15. All characters were unordered and given equal weight. In both analyses this resulted in two trees of length 24, consistency index 0.87 and reten- tion index 0.83. The only difference between these trees is the position of P. gibbosa and SZUTS & JOCQUE-^REVISION OF PALFURIA 219 P. hirsuta which are either the sister=group of P, spirembolus alone or of P. spirembolus to= gether with the other one. The strict consensus tree (“nelsen”) thus only collapses this ter- minal clade. This cladogram, as optimized un- der DELTRAN, is shown in Fig. 35 (prepared with WINCLADA, Nixon 1999). A number of non-informative characters (2, 4, 6 and 8) were included mainly because the males of three species are still unknown and at least some of these characters are likely to become informative when the missing sex is found. The only effect these characters have on the analysis is a slight increase of the consistency index which drops to 0.84 when these four characters are deactivated. The retention index remains stable, DISCUSSION As in many other genera in the family there is a large range of complexity in male palps and female epigyna. In the male palps this ranges from the basic situation with a simple dorsolateral tibial apophysis and a short, spine-shaped embolus {P. panner), to a tibia with at least two apophyses as in P. gladiator and P. harpago, often combined with a long, filiform embolus as in P. spirembolus. In the epigynum the range is from short-to-long en- trance ducts with the addition of a well sep- arated glandular organ of which the function is unclear. It is remarkable that, here again, the basal arrangement of the secondary geni- talia is more reminiscent of the primitive sit- uation in other genera than in the most derived members of Palfuria itself (Jocque 1998). Re- visions of the genera Storena Walckeeaer 1805 (Jocque & Baehr 1992), Diores Simon 1893 (Jocque 1991), Tenedos O.R-Cambridge 1897 (Jocque & Baert 1996), Asteron Jocque 1991 (Baehr & Jocque 1996) have revealed that in each of these genera the somatic char- acters are very stable whereas there is a wide range in the complexity of the secondary gen- italia. ACKNOWLEDGMENTS Special thanks go to E. Griffin (SMNW), B. Hauser and R Schwendinger (MHNG), and A, RusselLSmith, who sent us specimens for this study. We are indebted to Alain Reygel for the preparation of some drawings and for advice in connection with drawing techniques and to Jan Bosselaers who was so kind to 'print our cladogram in Clados. The first author has profited of a TEMPUS scholarship which gave him the opportunity for a three month stay in the Royal Africa Museum in Tervuren, LITERATURE CITED Baehr, B. & R, Jocque. 1996. A revision of Aster- on, starring male palpal morphology (Araneae, Zodariidae). Revue Suisse de Zoologie. VoL hors serie 1:15-28. Dippenaar-Schoeman, A. & R. Jocque. 1997. Afri- can Spiders. An Identification Manual, Plant Pro- tection Research Institute Handbook #9. 392 pp. Farris, J.S. 1988. Hennig86, ver. 1.5, Computer program distributed by its author. Goioboff, P. 1994. Pee Wee and NONA, version 2.15. Computer program distributed by its au- thor, Jocque, R. 1987. Descriptions of new genera and species of African Zodariidae with a revision of the genus Heradida (Araneae, Zodariidae). Re- vue de Zoologie Africaine 101:143-163. Jocque, R. 1990. A revision of the Afrotropical genus Diores (Araneae, Zodariidae). Annales du Musee Royal de FAfrique Centrale 260:1-81, Jocque, R. 1991. A generic revision of the spider family Zodariidae (Araneae). Bulletin of the Amer- ican Museum of Natural History 201:1-160. Jocque, R. 1998. Female choice, secondary effect of “mate check”? A hypothesis. Belgian Journal of Zoology 128:99-117. Jocque, R. & B. Baehr. 1992. A revision of the Australian spider genus Storena (Araneae, Zo- dariidae). Invertebrate Taxonomy 6:953-1004. Jocque, R &. L. Baert. 1996. Tenedos, an early conquest of America. Revue Suisse de Zoologie. Vol hors serie 1:309-320. Jocque, R. & A. S. Dippenaar-Schoeman. 1992. Two new, termite-eating Diores species (Ara- neae, Zodariidae) and some observations on unique prey immobilization. Journal of Natural History 26:1405 1412. Lessert, R. De. 1936. Araignees de FAfrique or- ientale portugaise, recueillis par MM. P. Lesne et H.-B. Cot. Revue Suisse de Zoologie 43:207- 306. Lessert, R. De. 1938. Araignees du Congo Beige (Premiere partie). Revue de Zoologie et de Bo- tanique Africaines 30:424-457. Nixon, K.C. 1999. Winclada version 0,99.9. Pro- gram and documentation available from the au- thor, Cornell University, Ithaca. Simon, E, 1910. Arachnoidea, Araneae (II). In L. Schultze (ed.), Zoologische und anthropologis- che Ergebnisse einer Forschungsreise im wes- tlichen und zentralen Siidafrika. Denkschriften des medicinisch-naturwissenschaftlichen Gesell- schaft zu Jena 16:175-218. Manuscript received 10 February 1999, revised 10 July 2000. 2001. The Journal of Arachnology 29:220-226 CRIBELLUM AND CALAMISTRUM ONTOGENY IN THE SPIDER FAMILY ULOBORIDAE: LINKING FUNCTIONALLY RELATED BUT SEPARATE SILK SPINNING FEATURES Brent D. Opell: Department of Biology, Virginia Polytechnic Institute and State University, Blacksburg, Virginia 24061 USA ABSTRACT. The fourth metatarsus of cribellate spiders bears a setal comb, the calamistrum, that sweeps over the cribellum, drawing fibrils from its spigots and helping to combine these with the capture thread’s supporting fibers. In four uloborid species (Hyptiotes cavatus, Miagrammopes animotus, Octonoba sinensis, Uloborus glomosus), calamistrum length and cribellum width have similar developmental trajec- tories, despite being borne on different regions of the body. In contrast, developmental rates of metatarsus IV and its calamistrum differ within species and vary independently among species. Thus, the growth rates of metatarsus IV and the calamistrum are not coupled, freeing calamistrum length to track cribellum width and metatarsus IV length to respond to changes in such features as combing behavior and abdomen dimensions. Keywords; Cribellar thread, Hyptiotes cavatus, Miagrammopes animotus, Octonoba sinensis, Uloborus glomosus Members of the family Uloboridae produce cribellar prey capture threads formed of a sheath of fine, looped fibrils that surround par- acribellar and axial supporting fibers (Eber- hard & Pereira 1993; Opell 1990, 1994, 1995, 1996, 1999; Peters 1983, 1984, 1986). Cri- bellar fibrils come from the spigots of an oval spinning field termed the cribellum (Fig. 1; Kovoor & Peters 1988; Opell 1994, 1999), lo- cated on the ventral surface of the spider’s ab- domen, just anterior to its spinnerets. These fibrils are drawn from cribellar spigots by the calamistrum, a setal comb that is formed of a single row of long, slender, curved setae that extends along the proximal of both fourth metatarsi (Fig. 2; Kullmann & Stern 1981; Opell 1979; Peters 1983, 1984). When drawing fibrils from the cribellum, uloborids brace the tarsus of the combing leg on the metatarsus of the opposite fourth leg (Eber- hard 1988; Opell 1979; Peters 1984). Left and right legs are used alternately in short, vig- orous bouts of combing (Eberhard 1988) and the resulting sheet of fibrils is compressed around supporting fibers by adductions of the posterior lateral spinnerets (Peters 1984). In members of the family Uloboridae, spi- derlings hatch from eggs, molt once within the egg sac, and emerge from the egg sac as sec- ond instars (Opell 1979). However, their cri- bella and calamistra are not functional until they molt again to become third instars. Sec- ond instar orb-weaving uloborid species pro- duce a juvenile web that lacks a sticky spiral and has many closely spaced radii (Lubin 1986). Members of the triangle- web genus Hyptiotes Walckenaer 1837 and the simple- web genus Miagrammopes O. Pickard-Cam- bridge 1 869 do not construct capture webs un- til they become third instars. After emerging from the egg sac, second instar Hyptiotes rest on vegetation (Opell 1982a, b), whereas Mia- grammopes cling to the outer surface of their cylindrical egg sac, which is still held by the female (Lubin, et al. 1978; Opell 2001). When spiders mature as sixth instars (Berland 1914; Opell 1982a, 1987), females retain a function- al cribellum and calamistrum, but males do not (Opell 1989, 1995). Complementary structures like the cribel- lum and calamistrum must develop in consort if they are to function throughout a spider’s life. As the cribellum is borne on the abdomen and the calamistrum on the fourth legs, this requires a convergence in the developmental rates of structures on different body regions. If the growth rates of the calamistrum and metatarsus IV are linked, then both must de- 220 OPELL— CRIBELLUM AND CALAMISTRUM ONTOGENY 221 Figures 1, 2. — The cribellum (Fig. 1, scale bar =150 ixm), and calamistrum (Fig. 2, scale bar = 250 fxm) of an adult female Miagrammopes animotus. velop at a rate that equals or exceeds that of cribellum width. If these structures develop at different rates, then only calamistrum length must increase at a rate that equals or exceeds that of cribellum width. I hypothesize that the latter occurs, as this would not compromise other fourth leg functions or require compen- satory changes in the lengths of other fourth leg articles. I tested this hypothesis by comparing the developmental rates of structures within the orb-weaving species Uloborus glomosus (Walckenaer 1841) and Octonoba sinensis (Si- mon 1880) and the reduced- web species Hyp- tiotes cavatus (Hentz 1847) and Miagram- mopes animotus Chickering 1968. The orb is the plesiomorphic web form in the Uloboridae and the triangle-web and simple-web are de- rived forms (Coddington 1990; Opell 1979). METHODS I collected 102 U. glomosus (39 of which were adult) from shrubbery on the Virginia Tech campus in Montgomery County, Virgin- ia during the spring and summer of 1989. Oc- tonoba sinensis is an introduced Asian spe- cies. I collected 79 individuals (17 of which were adults) in greenhouses on the Virginia Tech campus during the spring and summer of 1989. I collected 136 H. cavatus (31 of which were adult) from Giles and Montgomery Counties, Virginia during the spring and sum- mer of 1990. I collected 190 M. animotus (77 of which were female) from the Center for Energy and Environment Research’s El Verde Research Station, Luquillo National Forest, Puerto Rico during the summer of 1990. Only one species of each genus was present at each locality, so there were no problems in deter- mining the species of juvenile specimens. All instars of M. animotus were present at the same time. Successive instars of the other spe- cies were collected as they appeared during the spring and summer. These species were identified using the revisions of Chickering 222 THE JOURNAL OF ARACHNOLOGY Carapace Length jjm Figure 3. — Regressions of metatarsus IV length, calamisirum length, and cribellum width against car- apace length in Uloborus glomosus. Sample size = 102. (1968), Muma & Gertsch (1964), and Opell (1979). Voucher specimens are deposited in the Museum of Comparative Zoology. Specimens were preserved in 80% ethanol. I measured the carapace length of each spider under a dissecting microscope. Each speci- men’s cribellum and fourth leg were then re- moved and mounted in water soluble medium under a cover slip on a microscope slide. Us- ing a compound microscope, I measured cri- bellum width, fourth metatarsus length, and calamistrum length. All features were mea- sured to at least the nearest 20 pm. I measured calamistrum length as the distance separating the proximal- and distal-most setal bases. This approach eliminates problems associated with missing setae and it does not make any as- sumptions about the deflection of calamistrum setae during cribellar fibril combing. Statisti- cal tests were performed with SAS (SAS In- stitute Inc., Cary, North Carolina). P values of <0.05 were considered significant. RESULTS Figures 3-6 plot cribellum, calamistrum, and metatarsus IV lengths against carapace length in the four species studied. Each of these regressions is significant (F = 584- 2951, P — 0.0001). As reflected by values, the variance of these features is greater in H. cavatus and M. animotus than it is in U. glo- mosus and O. sinensis. This may be the result of measurement precision, as H. cavatus and M. animotus are smaller than U. glomosus and O. sinensis. However, the smaller values of the axes of H. cavatus and M. animotus (Figs. 5, 6) tend to exaggerate the scatter of these species’ points. Homogeneity tests show that, for each species, the slope of calamistrum length exceeds that of cribellum width (F == 8.77-80.58, P = 0.0034-0.0001) and the slope of metatarsus IV length exceeds that of calamistrum length (F = 17.45-482.30, P = 0.0001). In M. animotus the intercepts of cri- bellum width and calamistrum length differ (r OPELL— CRIBELLUM AND CALAMISTRUM ONTOGENY 223 Figure 4. — Regressions of metatarsus IV length, calamistrum length, and cribellum width against car- apace length in Octonoba sinensis. Sample size = 79. ~ 2.558, P — 0.015), but in the other three species they do not {t = 0.993-1.698, P ^ 0.100-0.285). Thus, in each species, calam- istrum length and cribellum width are initially very similar, but calamistrum length increases more rapidly than cribellum width. Tests of the homogeneity of the regression slopes of metatarsus IV show the slopes of U. glomosus and O. sinensis do not differ {F = 1.63, P = 0.204). The slope of U. glomosus (the smaller of the two orb- web species’ values) is greater than that of both H. cavatus (F = 70.73, P = 0.0001) and M. animotus {F = 330.98, P = 0.0001). Developmental rates of the calam- istrum also differ but do not mirror differences in metatarsus FV development. If they did, caT amistmm length would increase at a rate less than that of the cribellum in H. cavatus and M animotus. However, as documented above, in all species calamistrum length increases more rap- idly than does cribellum width. The slopes of the calamistrum of U. glo- mosus and O. sinensis do not differ (F — 2.60, P = 0.109). The slope of U. glomosus (the smaller of these values) is less than that of H. cavatus {F — 11.94, P = 0.0007) and greater than that of M. animotus (F = 11.69, P — 0.0007). The metatarsus IV of H. cavatus has a slope that is 0.229 less than that of U. glo- mosus, but its calamistrum has a slope that is 0.066 greater. In M. animotus, these values are “0.446 and —0.071, respectively. This is fur- ther documentation that the slopes of metatar- sus IV and its calamistrum are free to assume different trajectories. DISCUSSION The results of this study support the hy- pothesis that developmental rates of metatar- sus IV and the calamistrum differ. By devel- oping at a slower rate than the leg article that bears it, calamistrum length tracks more close- ly the development of the cribellum, with 224 THE JOURNAL OF ARACHNOLOGY Carapace Length jjm Figure 5. — Regressions of metatarsus IV length, calamistrum length, and cribellum width against car- apace length in Hyptoites cavatus. Sample size = 136. which it is functionally linked. I did not ex- amine possible correlates of metatarsus IV length. However, changes in the lengths of the fourth leg articles may be associated with changes in abdomen length and may serve to maintain the proper alignment of the calam- istrum as it passes over the cribellum. CaL amistrum length and cribellum width are ini- tially very similar, but calamistrum length increases at a slightly greater rate. If a cal- amistrum is to comb all the fibrils from cri- bellum spigots as it passes over the cribellum and is held with its length parallel to the trans- verse axis of the cribellum, then calamistrum length must equal cribellum width. As the an- gle formed by the calamistrum and the trans- verse axis of the cribellum increases, calam- istrum length must increase if it is to completely span the cribellum. For example, at an angle of 15°, the calamistrum must be 4% longer, and at 30°, 16% longer than if it were held at an angle of 0°. If the calamistrum moves laterally as it sweeps across the cribel- lum, further increases in calamistrum length would be necessary to ensure that the calam- istrum completely spans the cribellum. Thus, the developmental increase of calamistrum length relative to cribellum width observed in this study may reflect increases in the angle at which the calamistrum passes over the cri- bellum or the lateral movement of its passage. These changes may be necessary to accom- modate changes in the lengths of fourth leg articles or changes in abdomen length or width that require the fourth legs to assume different postures during the production of cri- bellar thread. Within the family Uloboridae, web reduc- tion is associated with a reduction in the length of metatarsus IV relative to the cara- pace length. In U. glomosus and O. sinensis adult females this ratio is 0.75 and 0.83, re- spectively. In H. cavatus and M. animotus this ratio is 0.63 and 0.56, respectively (Opell, un- publ. obs.). In contrast, the cribella of these two reduced-web species have greater num- bers of spinning spigots than do the orb-weav- ing species (Opell 1994). Thus, unless meta- OPELL— CRIBELLUM AND CALAMISTRUM ONTOGENY 225 Carapace Length pm Figure 6. — ^Regressions of metatarsus IV length, calamistrum length, and cribellum width against car- apace length in Miagrammopes animotus. Sample size = 190. tarsus IV and the calamistrum have different developmental trajectories, increases in cal- amistrum length could not keep pace with the increases in cribellum width that are necessary to accommodate a greater number of spigots and produce wider, stickier cribellar threads (Opel! 1995). The strong ontogenetic linkage of cribellum width and calamistrum length observed in this study contrasts with the weak phylogenetic re- lationship between these features observed by Opell et al. (2000). In that study no correlation between cribellum width and calamistrum length could be demonstrated among repre- sentatives of different families or among gen- era of the family Uloboridae. Only among species of the dictynid genus Mallos O. Pick- ard-Cambridge 1902 was there an association between these features, and this regression had an value of 0,41 compared to a mean value of 0.93 for the developmental studies reported here. As Opell et al. (2000) point out, differences in abdomen dimensions and cri- bellar thread combing behaviors among spe- cies probably explain the weak relationship between cribellum width and calamistrum length at more inclusive taxonomic levels. ACKNOWLEDGMENTS Jonathan Coddington provided helpful comments on this manuscript. This study was supported by N.S.E grants BSR- 89 17935 and IBN-9417803. LITERATURE CITED Berland, J. 1914. Note sur le cycle vital d’une Ar- aignee cribellate, Uloborus plumipes Lucus. Ar- chives de Zoologie Experimentale et Generate. Notes et Revue 54:45-57. Chickering, A.M. 1968. The genus Miagrammopes (Araneae, Uloboridae) in Panama and the West Indies. Breviora (Museum of Comparative Zo- ology) 289:1-28. Coddington, J.A. 1990. Ontogeny and homology in the male palpus of orb-weaving spiders and 226 THE JOURNAL OF ARACHNOLOGY their relatives, with comments on phylogeny (Ar- aneoclada: Araneoidea, Deinopoidea). Smithson- ian Contributions to Zoology 496:1-52. Eberhard, W.G. 1988. Combing and sticky silk at- tachment behaviour by cribellate spiders and its taxonomic implications. Bulletin of the British Arachnological Society 7:247-251. Eberhard, W.G. & F. Pereira. 1993. Ultrastructure of cribellate silk of nine species in eight families and possible taxonomic implications. (Araneae: Amaurobiidae, Deinopidae, Desidae, Dictynidae, Filistatidae, Hypochilidae, Stiphidiidae, Tengel- lidae). Journal of Arachnology 21:161-174. Kovoor, J. & H.M. Peters. 1988. The spinning ap- paratus of Polenecia producta (Araneae, Ulobor- idae): Structure and histochemistry. Zoomor- phology 108:47-59. Kullmann, E & H, Stern. 1975, Leben am seidenen Faden. Bertelsmann Verlag, Miinchen. Lubin, Y.D. 1986. Web building and prey capture in the Uloboridae. Pp. 132-171. In Spiders: Webs, Behavior, and Evolution (W.A. Shear, ed.),. Stanford Univ. Press, Stanford. Lubin, Y.D., W.G. Eberhard, & G.G, Montgomery. 1978. Webs of Miagrammopes (Araneae: Ulo- boridae) in the Neotropics. Psyche 85:1-23. Muma, M.M. & W.J. Gertsch. 1964. The spider family Uloboridae in North America north of Mexico. American Museum Novitates 2196:1- 43. Opell, B.D. 1979. Revision of the genera and trop- ical American species of the spider Family Ulo- boridae. Bulletin of the Museum of Comparative Zoology 148:433-549. Opell, B.D. 1982a. Post-hatching development and web production of Hyptiotes cavatus (Hentz) (Araneae: Uloboridae). Journal of Arachnology 10:185-191. Opell, B.D. 1982b. Cribellum, calamistrum, and ventral comb ontogeny in Hyptiotes cavatus (Hentz) (Araneae: Uloboridae). Bulletin of the British Arachnological Society 5:338-343. Opell, B.D. 1987. Changes in web-monitoring forces association with web reduction in the spi- der family Uloboridae, Canadian Journal of Zo- ology 65:1028-1034. Opell, B.D. 1989. Functional associations between the cribellum spinning plate and prey capture threads of Miagrammopes animotus (Araneida, Uloboridae). Zoomorphology 108:263-267. Opell, B.D. 1990. The material investment and prey capture potential of reduced spider webs. Behavioral Ecology and Sociobiology 26:375- 381. Opell, B.D. 1994, Factors governing the stickiness of cribellar prey capture threads in the spider family Uloboridae. Journal of Morphology 221: 111-119. Opell, B.D. 1995. Ontogenetic changes in cribel- lum spigot number and cribellar prey capture thread stickiness in the spider family Uloboridae, Journal of Morphology 224:47-56, Opell, B.D. 1996, Functional similarities of spider webs with diverse architectures, American Nat- uralist 148:630-648. Opell, B.D. 1999. Changes in spinning anatomy and thread stickiness associated with the origin of orb-weaving spiders. Biological Journal of the Linnean Society 68:593-612. Opell, B.D., J.S. Sandidge & J.E. Bond. 2000. Ex- ploring Functional Associations Between Spider Cribella and Calamistra. Journal of Arachnology 28:43-48. Opell, B.D. 2001. Egg sac recognition by female Miagrammopes animotus (Araneae, Uloboridae), Journal of Arachnology 29:244-248. Peters, H.M. 1983. Struktur und Herstellung der Fangfaden cribellater Spinnen (Arachnida: Ara- neae). Verhandlungen des Naturwissenschaftlich- en Vereins in Hamburg 26:241-253. Peters, H.M, 1984. The spinning apparatus of Ulo- boridae in relation to the structure and construc- tion of capture threads (Arachnida, Araneida). Zoomorphology 104:96-104. Peters, H.M. 1986. Fine structure and function of capture threads. In Ecophysiology of Spiders (W, Nentwig, ed.), Pp. 187-202. Springer Verlag, New York, Manuscript received 5 June 2000, revised 10 Feb- ruary 2001. 2001. The Journal of Arachnology 29:227-237 DOES THE STRUCTURAL COMPLEXITY OF AQUATIC MACROPHYTES EXPLAIN THE DIVERSITY OF ASSOCIATED SPIDER ASSEMBLAGES? Josue Raizer and Maria Eugenia C. Amaral: Departamento de Biologia, Centro de Ciencias biologicas e da Saude, Universidade Federal de Mato Grosso do Sul, C.R 549, Campo Grande, MS 79070-900 Brazil ABSTRACT. Differences in species richness and species composition of spiders associated with aquatic macrophytes of different structural complexities were examined in the Pantanal floodplain of Mato Grosso do Sul, Brazil. The plants studied were Nymphaea amazonum (Nymphaeaceae), Salvinia auriculata (Sal- viniaceae), Echinodorus paniculatus (Alismataceae) and Eichhornia azurea (Pontederiaceae), whose clas- ses of complexity were determined based on their leaf and branch densities, vertical structure, and height. Data were collected from 62 monospecific plant patches in temporary lentic environments. A total of 235 spiders of 33 species in 13 families was collected. Nymphaea amazonum, the plant with the lowest com- plexity class, did not provide adequate sites for the establishment of spiders, and only four individuals of four spider species were found on its patches. Salvinia auriculata and E. paniculatus shared the inter- mediate class of complexity, but showed statistically significant differences in composition and richness of spider species. In E. paniculatus, greater height and lower leaf and branch densities favored the estab- lishment of web weavers, whereas the smaller height and higher density of S. auriculata promoted the occurrence of wandering spiders. Eichhornia azurea, the plant with the highest complexity class, presented the greatest number of unique spider species, differing from the other plants in spider species composition. Results indicate that richness and composition of spider species associated with aquatic macrophytes in the study site are influenced by the structural complexity of these plants. Keywords: Araneae, community structure. South Pantanal, species composition, species richness Habitat structural complexity can affect species diversity (Pianka 1978; Robinson 1981; Gunnarson 1988; Cornell & Lawton 1992; Shorrocks & Sevenster 1995; Balfour & Rypstra 1998). This hypothesis has been supported by studies focusing on different an- imal species in several environments (Pianka 1966, 1967; Murdock et aL 1972; Uetz 1975, 1977; Hatley & MacMahon 1980; Dueser & Porter 1986; Dean & Connell 1987; Pearsons et al. 1992). It has been shown, for example, that the vertical structure of vegetation in North American temperate forests is a better indicator of bird diversity than the diversity of plant species with which the birds are asso- ciated (Mac Arthur & Mac Arthur 1961). Spatial and architectural features of habitat structure can determine diversity, density, and distribution of spider species (Hatley & MacMahon 1980; Balfour & Rypstra 1998). Similarly, environmental physiognomy (for instance, open or closed forest; dense or sparse litter layer) and physical structure can significantly influence spider habitat prefer- ence (Jennings et al. 1988; Uetz 1991). Num- ber and dominance of spider species tend therefore to be highly related to the structure of the plant community on which they occur (Gunnarsson 1990; Uetz 1991; Baur et al. 1996). These relationships among plant and spider communities appear to be determined primarily by the structural complexity of the plant, which can provide, for example, a va- riety of retreats and attachment sites for webs, as well as favorable microclimatic conditions (Hatley & MacMahon 1980; Pulz 1987). Few are the studies performed on the as- sociation of spiders and aquatic macrophytes, and on the effects of the structural complexity of aquatic macrophytes on the community structure of associated spider species. Even these studies, however, have been limited to reporting the occurrence of spiders on those plants (e.g., Merck 1988; Heckman 1994) or to describing new species (Brescovit et al. 2000). The present work examined the influence of plant structural complexity on spider com- 227 228 THE JOURNAL OF ARACHNOLOGY munities associated with aquatic macrophytes in the southern Pantanal, Brazil, by evaluating the variation of spider species composition and richness on four plant species. METHODS Study site. — The Pantanal is a floodplain of ca. 140,000 km^ located in central South America, mainly within Brazil. This area is, in fact, an assemblage of diverse landscapes occupying the hydrographic sub-basins of the Paraguay watershed. Each of these sub-basins has its characteristic hydrologic regimes, soil types, and geologies, which affect fauna and flora distribution (Boggiani & Coimbra 1996). The present study was carried out in a Pan- tanal sub-region known as “Pantanal do Mi- randa e Abobral” {sensu Adamoli 1982), lo- cated in Mato Grosso do Sul state, southwestern Brazil (19°22'-19°33'S; 5T2’- 57°3'W). The climate is characterized by a wet season extending from December-May and a dry one from June-November. All sam- ples were collected from temporary lentic en- vironments located in the vicinity of a 25 km stretch of the MS- 184 road. These lentic en- vironments are formed by depressions in the terrain that remain inundated by nearby rivers or filled with rainwater for most of the wet season, resulting in temporary ponds. When rains end and rivers start to recede, the water level in these ponds begins to fall, and they eventually disappear during the dry season. Because of this cycle, data collection had to be limited to the period when water was not entirely depleted. Aquatic macrophytes. — The plants inves- tigated were Nymphaea amazonum Mart. & Zucc. (Nymphaeaceae), Salvinia auriculata Aublet (Salviniaceae), Echinodorus panicula- tus Mich. (Alismataceae), and Eichhornia azurea (Sw.) Kunth (Pontederiaceae), which are illustrated in Fig. 1. The floating leaves of N. amazonum have glabrous, membranous, or- bicular, laminate limbs and lie flat on the wa- ter surface, forming a discontinuous, thin, flat carpet. Patches of S. auriculata resemble a continuous, thick, curly carpet formed by up- right chains of smaller, floating leaves. This is a herbaceous plant with short petioles and pi- lose, oval limbs. Echinodorus paniculatus, an- other herbaceous plant, has long triangular petioles and leaves that emerge vertically or obliquely. Limbs are glabrous, coriaceous and lanceolate. Eichhornia azurea, also herba- ceous, has leaves that emerge vertically or obliquely; but its petioles are cylindrical, shorter than those of E. paniculatus and have a sheath. Limbs are glabrous, fleshy and ob- ovate. Quantification of the structural complex- ity of aquatic macrophytes. — The structural complexity of monospecific patches of those four aquatic macrophytes was quantified by measuring plant density, height and vertical structure above the water surface, based on the methodology of Balfour & Rypstra (1998). Ten patches of each plant were sampled, em- ploying a 1 m^ floating PVC frame whose sides were numbered at 10 cm intervals, thus defining an orthogonal grid. Numbering ran from 0-10 on one of the sides, then proceeded from 11-21 on the adjacent side (with num- bers 10 and 1 1 coinciding on the comer edge). On the remaining two sides, these integer se- quences (0-10, 11-20) were repeated so as to mirror those parallel to them. In order to measure plant density in each sample, an integer from 0-21 was drawn. A horizontal line was then positioned connecting the same two integers lying on opposite sides of the frame, and the number of branches and leaves touching this line was recorded. For measuring the vertical stmcture of the plants, two integers were drawn, one of them from 0-10 and the other from 11-21. Two horizontal lines were thus determined, at whose intersection a third, vertical line was positioned. The number of leaves and branch- es touching this vertical line was recorded. Plant height was considered as the highest point at which the plant touched this vertical line. To determine the stmctural complexity clas- ses of the plant patches, the differences among the variables considered were tested by anal- ysis of variance (ANOVA) and Tukey test {a = 0.05). Three possible arbitrary values were then assigned to the means of those variables: value 0 to the smallest mean, 1 to the inter- mediate, and 2 to the greatest mean. Data sampling.^ — Data were collected from 13 ponds from November 1994 to April 1997. A total of 62 monospecific patches was sam- pled, namely, 12 of N. amazonum, 18 of S. auriculata, 10 of paniculatus, and 22 of E. azurea. Patch area was estimated and subdi- vided in numbered sub-areas of 1 m^, and one RAIZER & AMARAL— SPIDER DIVERSITY IN AQUATIC MACROPHYTES 229 Figure L— Schematic depiction of the four aquatic macrophyte species investigated. Nymphaea ama- zonum forms a thin mat that lies flat on the water surface. Salvinia auriculata forms a mat that is rich in recesses and projections. Echinodorus paniculatus has stems that emerge vertically or obliquely without ever forming a mat. The carpet formed by Eichhomia azurea is also rich in recesses and projections, but has a series of vertical and oblique emerging stems. of them was randomly chosen as the sampling point for that patch. This sampling area was delimited with the help of a 1 floating PVC frame whose sides were fitted with a 15 cm high nylon-mesh screen to prevent spiders from escaping. At all sampling points, plant species richness had value 1. All spiders vi- sually located were collected for identifica- tion, and their voucher specimens are depos- ited in the Museum of Instituto Butantae, Sao Paulo. Spider species composition and rich- ness*—In order to compare spider species richness among the plant species, the smallest sample size (number of individuals collected associated with one of the plant species) was considered, since species richness is depen- dent on the number of individuals sampled. Spider species richness was then estimated for each plant species by rarefaction, using the software RAREFACT (Krebs 1989). The ex- pected number of species and the standard de- viation for each complexity class were thus obtained. Species richness was considered to differ among complexity classes if no overlap occurred between the intervals generated by the standard deviation of species richness for each sample. The mean numbers of spider species per sample were statistically compared among the plant species by analysis of vari- ance. The mean number of spider species and the number of individuals, both grouped by guild (web or wandering spiders), were also analyzed (ANOVA, Tukey test, and test of independence, a — 0.05). RESULTS Quantification of plant structural com- plexity.— Results obtained for the structural 230 THE JOURNAL OF ARACHNOLOGY complexity of the monospecific patches of aquatic macrophytes are shown in Table 1. Complexity class was lowest for N. amazon- um (class 0) and highest for E. azurea (class 4). No plant species fitted classes 1 or 2. Sal- vinia auriculata and E. paniculatus shared the same complexity class (class 3), despite dif- ferences found in the variables involved in its determination: S. auriculata presented greater density and smaller height, whereas for E. paniculatus density was smaller and height was greater. Eichhornia azurea provided the greatest value for vertical structure, but its val- ues for plant density and height were inter- mediate. Nymphaea amazonum did not show any variation in either height or vertical struc- ture (both at value 0), and its density never exceeded three leaves per meter. Spider species composition and rich- ness.— A total of 235 spiders belonging to 33 species was found (Table 2). Because some of the individuals collected were juvenile, they could only be taxonomically identified down to family or genus level. Only four individu- als— and each of these of a different species — were found on N. amazonum. Their occur- rence was regarded as fortuitous, and the cor- responding data were excluded from the com- position and richness analyses. As for the other three plants, 15 spider species (63 in- dividuals) were found on S. auriculata, 14 species (54 individuals) on E. paniculatus, and 24 species (114 individuals) on E. azurea. Four species of spiders were common to these three plant species. Regarding composition, E. azurea presented the greatest number of unique spider species (9 species; 37.5%), fol- lowed by S. auriculata (5 species; 33.33%) and E. paniculatus (3 species; 21.43%). The smallest overlap in species composition oc- curred between S. auriculata and E. panicu- latus, with only 5 common species {ca. 35%). Four of these five species were also common to E. azurea, accounting for 39.15% of the overall total of individuals collected. Overall, wandering spiders outnumbered web weavers, both in the number of species and of individuals (Table 2). Wandering indi- viduals accounted for 66.23% of the total col- lected (Fig. 2). When only those animals col- lected from E. paniculatus are considered, the proportion of wandering individuals falls to ca. 37% (Fig. 2). The proportions between wandering and web spiders differed signifi- cantly among the three plants analyzed (x^ = 8.05, df= 2, P < 0.02, Fig. 2). The proportion of individuals between guilds was higher for wandering spiders on both S. auriculata and E. azurea, but higher for web spiders on E. paniculatus. As to richness, the expected number of spi- der species arrived at by employing the rare- faction method differed among plant species. By applying this method to the 54 individuals found on E. paniculatus (the smallest number of spiders found on any of the three plants analyzed), the number of spider species could be estimated at 18.89 (SD ± 1.56) for.E. azur- ea, 14.10 (SD ± 0.84) for S. auriculata, and 14 for E. paniculatus (Table 2). Taking into account the intervals generated by the stan- dard deviation of the expected number of spe- cies, E. azurea presented the greatest richness, whereas S. auriculata and E. paniculatus did not differ from each other. Statistical analysis of the data on spider species number per sam- ple for each plant species did not reveal any significant differences among the three plants analyzed (ANOVA, F = 0.598, df = 2, P = 0.545). However, when the numbers of spider species per guild were compared among the plants, E. paniculatus showed the greatest richness of web spiders and the smallest one of wandering spiders, whereas S. auriculata and E. azurea did not differ significantly from each other (Table 3). DISCUSSION Nymphaea amazonum, the plant with the lowest structural complexity, had the lowest richness of spiders, with four species but only one individual of each (Table 2). The occur- rence of these was regarded as fortuitous, since the same four species were abundantly found on all the other macrophytes (Raizer 1997). This is indicative that N. amazonum does not favor the establishment of a com- munity of associated spiders. In fact, because its leaves are smooth and lie flat on the water surface, this plant does not provide microsites for oviposition, molting, or construction of any kind of web or retreat. Spiders living on these leaves would also be directly exposed to solar radiation, which favors dehydration (Pulz 1987). Furthermore, potential prey (such as diptera and orthoptera) are rarely found on N. amazonum (Raizer pers. obs.). Spider species composition varied not only Table 1. — StnictEral comfjlexity of monospecific patches of aquatic macrophytes in southern Pantanafi calculated from the following variables: plant density RAIZER & AMARAL— SPIDER DIVERSITY IN AQUATIC MACROPHYTES 231 eS 1 o m VI m IT) q q q fr d + 1 + 1 +[ O o (T) fr OV q d VO fr CO IT) o C<3 cd Ol O' o O "§ d q 0\ Ol 0 •S .o +1 +1 + i s o o o o q cn CO vd cd d 00 I 2 5 S I i I § < Q o o o q q q csf d „ d . d m VO . , >r) q V VO V O' V OV Q ^ D U n o ^ ^ *0 ^ fNi ^ tn ^ KD ^ cn cT) ^ cT) tM Aj {4=< ^ "Id I I I i o TO q cd q fO VI d q ^=H p- 1 + 1 +1 + 1 o o q TO TO d d d OV OQ o If) o d d +! +[ o o VO o d d CV| Q Q Q ^ A'- OT ^ g q o q g +i q +1 M +! U D c « IS 'S a q q q o o « ^ o ^ « a B d S c S cd 'a, S 'a S ^ S ^ « jag§§jig«3 ‘=^■3 3 I a I o o > > pH 232 THE JOURNAL OF ARACHNOLOGY m X X © "§ -g § a o o X X X X X X X X X XXXXX XX XX fnc4r>.o^ ono t^o o-H or-oovoo o o o cmo-^oo o-h oo oo oO’-'OCNio o o o oo oooooo o o o ^oooo oo oo ^ k ^ M ^ 4) m -o •a e a U K -d ^ "O 60 S 53 3 I o 0 a ^ ^ e a « « 3 « 3 d o 'd o o a -3 a S ^ ^ .a M H 00 O -O « 4) « 3 3 3 see « « 4) o 'd -d >> 3 "o o s O a *« 4_> 4-1 ^ ^ o o o g a d c Q w w w O ^ CNI cn O Qm &( P4 00 C^ 4i i d 'a M M ^ 3 d 'C ^ ^ cd 3 a g i e a W M 3 o 3 ^ u £ m RAIZER & AMARAL— SPIDER DIVERSITY IN AQUATIC MACROPHYTES 233 X X X X I I I X X X X X X X X X TO I I j I r- o o o o o \0 TO C\| T-i o VO o in ^ O TO TO m VO ^ o o o o o o -d © o •d d d © « « © cd cd d a « © d d CN a o. TO a D. 1 1 1 M m M m m I d 'C ’d -c © &i © g X Number of samples 12 18 10 21 Mean number of species ± SD 0.33 ± 0.65 3.50 ± 2.31 5.40 ± 4.30 5.43 ± 5.80 Species richness by rarefaction from 54 individuals ± SD — 14.10 ± 0.84 14.00 ± 0.00 18.89 ± 1.56 234 THE JOURNAL OF ARACHNOLOGY 90 80 - SaMnia Echinodorus Eichhornia In all three auriculata paniculatus azurea species Figure 2. — Percentages of the numbers of individuals of each spider guild (wandering and web spiders) for the aquatic macrophytes analyzed (Salvinia auriculata, Echinodorus paniculatus and Eichhornia azur- ea). among plants of different structural complex- ity classes, but also between those of the same class {S. auriculata and E. paniculatus). Dif- ferences between these two class-3 plants can be explained by the differences in height and density of their emersed parts. Salvinia auri- culata, having high leaf density and small height above the water surface, favored the establishment of wandering spiders, which hunt and build their retreats on the leaves. Among the few web-weaving species found on this plant, only one, Actinosoma pentacan- thum (Walckenaer, 1837), builds webs that are parallel to the water surface (Raizer 1997). This was actually the only spider seen em- ploying the leaf mat of S. auriculata to attach a web with such orientation. On the other hand, E. paniculatus, the macrophyte with the lowest leaf density and greatest height, fa- vored the establishment of spiders that weave large-sized webs — such as those of the fami- lies Araneidae and Tetragnathidae. Similar re- sults were found by Dobel et al. (1990) in a study on a community of spiders associated with the grass Spartina alterniflora Lois. This intertidal salt marsh plant presents three dis- tinct habits: short, intermediate, and tall forms. The short form favors the occurrence of wandering spiders, while the intermediate one, with an architecture similar to that of E. paniculatus, enables the establishment of web-weaving species. The variation in species composition among E. azurea (complexity class 4), S. auriculata Table 3. — Mean proportion of spider species number per guild (web and wandering spiders) on S. auriculata, E. paniculatus, and E. azurea. Analysis of variance (ANOVA) was performed for the arc sine of the square root of the proportions of spider species per sample. On each line, values followed by the same letter do not differ significantly (Tukey test, a = 0.05). Guild Salvinia auriculata (mean ± SD) Echinodorus paniculatus (mean ± SD) Eichhornia azurea (mean ± SD) ANOVA results web spiders 0.19 ± 0.22a 0.54 ± 0.31b 0.24 ± 0.26a F = 4.525, df = 2, P = 0.017 wandering spiders 0.81 ± 0.22a 0.46 ± 0.31b 0.76 ± 0.26a F = 4.525, df = 2, P = 0.017 RAIZER & AMARAL— SPIDER DIVERSITY IN AQUATIC MACROPHYTES 235 and E. paniculatus (both of class 3) is possibly due to the number of spider species that are unique to E. azurea and to the small overlap of species composition between this plant and E. paniculatus (3 species in common) or S. auriculata (5 species in common). In addition, the proportion of web weavers occurring on E. paniculatus was higher than that of wan- dering spiders, whereas the opposite was ob- served for S. auriculata and E. azurea (Fig. 2). These data corroborate the results obtained by Hatley & MacMahon (1980) when com- paring spider species compositions for bushes with differing leaf and branch densities. Ac- cording to their findings, spider species that constructed large- sized webs were found in less structurally complex bushes with lower leaf and branch densities, but not in high-den- sity plants. Our findings support the hypoth- esis that plants of different structural com- plexities favor distinct associations of spider species, thus influencing the species compo- sition of such communities, as also found in other studies (e.g., Hatley & MacMahon 1980; Dobel et al. 1990, and Balfour & Rypstra 1998). The expected number of spider species, as determined by rarefaction, varied among plant species. The greatest richness was the one re- corded for the plant with the highest com- plexity class {E. azurea). However, no signif- icant statistical difference was evidenced when richness was assessed by the mean num- ber of spider species per sample, among plants. Nonetheless, in a third analysis, when rich- ness was assessed separately for each guild, the mean proportions between species num- bers varied significantly (Table 3). Echinodo- rus paniculatus presented the highest propor- tion of web spider species, probably due to the dependence of such spiders on this plant’s architecture, characterized by its high density and great height of leaves. On the other hand, the proportion of web spider species on E. azurea did not differ from that found for S. auriculata. This can be explained on the basis that orb-webs are usually anchored to open sites, which facilitate the capture of flying prey (insects). This feature would render the high leaf density of E. azurea unfavorable to the construction of such webs, which are mainly built on the edges of the patches formed by this plant (Raizer 1997). Nor does S. auriculata offer suitable sites for the at- tachment of orb-webs, except those parallel to the water surface. When wandering spiders alone were con- sidered, E. paniculatus showed the smallest species richness, whereas S. auriculata and E. azurea did not differ from each other. As with web weavers, plant architecture can explain such results: E. paniculatus lacks a suitable architecture for the establishment of various species of wandering spiders since it never forms a mat on the water surface; S. auricu- lata and E. azurea, in turn, with their high leaf and branch densities, do form continuous mats that are rich in recesses and projections that favor wandering spiders. The number of species of a given guild is thus influenced by variables of the structural complexity, such as the density and height of leaves and branches. In the present study, tall plants with low leaf densities displayed a larg- er number of web spider species, whereas a greater richness of wandering spiders was found for short plants with high leaf densities. Our results support the hypothesis that structural complexity of plants also influences spider species richness, and corroborate other studies on the influence of habitat structure on species richness and species composition of spider assemblages (Hatley & MacMahon 1980; Greenstone 1984; Jennings & Hilburn 1988; Uetz 1991; Baur et al. 1996). Habitat structural complexity is in fact one of the main factors used to explain species diversity (e.g., Mac Arthur & Mac Arthur 1961; Pianka 1978; Hatley & MacMahon 1980; Dean & Connell 1987; Shorrocks & Sevenster 1995). Magurran (1988) stated that habitats with high microsite diversity have greater species richness, since different mi- crosites can have characteristic species asso- ciated with them. Other studies testing the re- lationship between structural complexity and species diversity have demonstrated that greater microsite diversity leads to a greater number of niches and can minimize interspe- cific competition (e.g., Pianka 1978; Hatley & MacMahon 1980; Shorrocks & Sevenster 1995). The present study has revealed that structurally dissimilar habitats may show sim- ilar spider species richnesses while differing in species composition. These findings sup- port the suggestions of Jennings et al. (1988) and Baur et al. (1996) that communities of 236 THE JOURNAL OF ARACHNOLOGY spiders or other invertebrates are mainly or- ganized as a function of the structural com- plexity of the environments. Variations in species composition can be explained by habitat preferences resulting from behavioral and morphological character- istics of the spiders (Johnson 1995; Richman 1995). Since richness of aquatic macrophyte spe- cies did not vary in the present investigation, remaining at value 1, it can be concluded that structural complexity is an important factor for the organization of spider communities on these plants, a factor that can affect richness and, even more strongly, composition of the spider species associated with them. ACKNOWLEDGMENTS This work is part of the MSc. thesis of the first author. A graduate student scholarship was granted to J. Raizer by CAPES. Financial support was provided by the State Council of Science and Technology of the State of Mato Grosso do Sul (CECITEC) and Coordenadoria de Pesquisa da Universidade Federal de Mato Grosso do Sul (CPQ-PROPP-UFMS). The au- thors are indebted to Joao Vasconcellos-Neto of Universidade Estadual de Campinas, Ro- gerio Parentoni Martins of Universidade Fed- eral de Minas Gerais, and Erich Fischer and Frederico Santos Lopes, both of Universidade Federal de Mato Grosso do Sul, for their help- ful comments and suggestions. They also thank Kirt Matthew Wackford and Kennedy Francis Roche for their remarks in the first version of the text, Antonio Domingos Bres- covit of Instituto Butantan, Sao Paulo, iden- tified the spiders and provided valuable com- ments, which are gratefully acknowledged. Thanks are also given to Masao Uetanabaro of the Office of Pantanal Studies of Univer- sidade Federal de Mato Grosso do Sul (CEP/ PROPP-UFMS) for making the facilities of UFMS’s Field Station for Pantanal Studies available, and to Vander F. Melquiades de Je- sus for the plant drawings. Restructuring and revision of the final version were provided by Gerson Ferracini. The authors also thank G. Miller, J.W. Berry and M.H. Greenstone for their editorial review. This work was support- ed by CNPq (grants 351235/97.3 and 522616/ 95.0). LITERATURE CITED Adamoli, J. 1982. O Pantanal e suas relagoes fi- togeograficas com os cerrados. Anais do XXXII Congresso Nacional de Botanica 32:109-119. Babcock, C.A. & C.R. Ely. 1994. Classification of vegetation communities in which geese rear broods on the Yukon-Kuskokwim delta, Alaska. Canadian Journal Botany 72:1294-1301. Balfour, R.A. & A.L. Rypstra. 1998. The influence of habitat structure on spider density in a no-till soybean agroecosystem. Journal of Arachnology 26:221-226. Baur, B., J. Joshi, B. Schmid, A. Hanggi, D. Bor- card, J. Stary, A. Pedroli-Christen, G.H. Thorn- men, H. Luka, H. Rusterholz, P. Oggier, S. Led- ergerber & A. Erhardt. 1996. Variation in species richness of plants and diverse groups of invertebrates in three calcareous grasslands of the Swiss Jura mountains. Revue Suisse Zoolo- gie 103:801-833. Boggiani, PC. & A.M. Coimbra. 1996. A planicie e os pantanais. Pp. 18-23. In Tuiuiu. Sob Os Ceus Do Pantanal. Biologia E Conserva^ao Do Tuiuiu (Jabiru nycteria) (P. T.Z. Antas & I.L.S. Nascimento, eds.). Empresa das Artes, Sao Pau- lo. Brescovit, A.D., J. Raizer & M.E.C. Amaral. 2000. Descriptions and notes on the genus Paradossen- us in the neotropical region (Araneae, Trechal- eidae). Journal of Arachnology 28:7-15. Cornell, H.V. & J.H. Lawton. 1992. Species inter- actions, local and regional processes, and limits to the richness of ecological communities: a the- oretical perspective. Journal of Animal Ecology 61:1-12. Dean, R.L. & J.H. Connell. 1987. Marine inver- tebrates in an algal succession. II. Tests of hy- potheses to explain changes in diversity with succession. Jounral of Experimental Marine Bi- ology and Ecology 109:217-247. Dobel, H.G., R.E Denno & J.A. Coddington. 1990. Spider (Araneae) community structure in an in- tertidal salt marsh: Effects of vegetation structure and tidal flooding. Environmental Entomology 19:1357-1370. Dueser, R.D. & J.H. Porter. 1986. Habitat use by insular small mammals: Relative effects of com- petition and habitat structure. Ecology 67:195- 201. Greenstone, M.H. 1984. Determinants of web spi- der species diversity: Vegetation structural diver- sity vs. prey availability. Oecologia (Heidelberg) 62:299-304. Gunnarsson, B. 1988. Spruce-living spiders and forest decline; the importance of needle loss. Bi- ological Conservation 43:309-319. Gunnarsson, B. 1990. Vegetation structure and the abundance and size distribution of spruce-living spiders. Journal of Animal Ecology 59:743-752. RAIZER & AMARAL— SPIDER DIVERSITY IN AQUATIC MACROPHYTES 237 Hatley, C.L. & J.A. MacMahon. 1980. Spider com- munity organization: Seasonal variation and the role of vegetation architecture. Environmental Entomology 9:632-639. Heckman, C.W. 1994. New limnological nomen- clature to describe ecosystem structure in the tropical wet-and-dry climatic zone. Archiv fuer Hydrobiologie 130:385-407. Jennings, D.T. & D.J. Hilburn. 1988. Spiders (Ar- aneae) captured in malaise traps in spruce-fir for- ests of west-central Maine. Journal of Arachnol- ogy 16:85-94. Jennings, D.T, M.W. Houseweart, C.D, Dondale & J.H. Redner. 1988. Spiders (Araneae) associated with strip-clearcut and dense spruce-fir forests of Maine. Journal of Arachnology 16:55-70. Johnson, S.R. 1995. Observation of habitat use by Sarinda hentzi (Araneae, Salticidae) in north- eastern Kansas. Journal of Arachnology 23:71- 74. Krebs, C.J. 1989. Ecological Methodology. Harper & Row, New York. Mac Arthur, R.H. & J.W. Mac Arthur. 1961. On bird species diversity. Ecology 42:594-598. Magurran, A.E. 1988. Ecological Diversity And Its Measurement. Chapman and Hall, New York. Merck, A.M.T. 1988. A fauna do pleuston Salvinia auriculata Aublet (Pteridophyta, Salviniaceae) do Lago Bangui, Curitiba, Parana. 1. Flutuagao anual da densidade. Revista Brasiliera de Biolo- gia 48:991-1004. Murdock, W.W., EC. Evans & C.H. Peterson. 1972. Diversity and pattern in plants and insects. Ecol- ogy 53:819-829. Pearsons, TN., H.W. Li & G.A. Lamberti. 1992. Influence of habitat complexity on resistance to flooding and resilience of stream fish assemblag- es. Transactions of the American Fisheries So- ciety 121:427-436. Pianka, E.R. 1966. Convexity, desert lizards, and spatial heterogeneity. Ecology 47:1055-1059. Pianka, E.R. 1967. On lizard species diversity: North American flatland deserts. Ecology 48: 333-351. Pianka, E.R. 1978. Evolutionary Ecology. Harper & Row, New York. Pulz, R. 1987. Thermal and water relations, Pp. 26-55. In Ecophysiology Of Spiders (W. Nen- twig, ed.). Springer- Verlag, Berlin. Raizer, J. 1997. Spiders Associated With Aquatic Macrophytes In The Pantanal of Mato Grosso Do Sul State. MS Thesis, Universidade Federal de Mato Grosso do Sul. 41 pp. (in Portuguese) Richman, D.B. 1995. A comparison of populations of wolf spiders (Araneae, Lycosidae) on two dif- ferent substrates in southern Florida. Journal of Arachnology 23:151-156. Robinson, J.V. 1981. The effect of architectural variation in habitat on a spider community: An experimental field study. Ecology 62:73-80. Shorrocks, B. & J.G. Sevenster. 1995. Explaining local species diversity. Proceedings of the Royal Society of London, B. Biological Sciences 260: 305-309. Uetz, G.W. 1975. Temporal and spatial variation in species diversity of wandering spiders (Araneae) in deciduous forest litter. Environmental Ento- mology 4:719-724. Uetz, G.W. 1977. Coexistence in a guild of wan- dering spiders. Journal of Animal Ecology 46: 531-541. Uetz, G.W. 1991. Habitat structure and spider for- aging. Pp. 325-348. In Habitat Structure. The Physical Arrangement of Objects in Space (S.S. Bell, E.D. McCoy & H.R. Mushinsky, eds.). Chapman and Hall, London, Manuscript received 14 October 1998, revised 9 November 2000. 2001. The Journal of Arachnology 29:238-243 ON THE DISTRIBUTION AND PHENOLOGY OF ARGYRODES FICTILIUM (ARANEAE, THERIDIIDAE) AT ITS NORTHERN LIMIT OF NORTH AMERICA Pierre Paquin and Nadine Duperre: Departement de Sciences Biologiques, Collection Ouellet-Robert, Universite de Montreal, C.P. 6128, Succ. Centre- ville Montreal, Quebec H3C 3J7 Canada ABSTRACT. Argyrodes fictilium is a rarely collected species whose northern range was thought to be southern Canada. Recent collections in the eastern boreal forests of Quebec extend its distribution range to the north and suggest that A. fictilium might be found anywhere within the boreal forest tree limit. Mature males collected in May indicate a summer-stenochronous type of phenology. Keywords: Distribution, phenology, boreal, araneophagic The genus Argyrodes Simon 1864 is known in Canada from three species easily recogniz- able as members of the genus by their unusual triangular abdomen (Exline & Levi 1962). These are Argyrodes trigonum (Hentz 1850), A. cancellatus (Hentz 1850) and Argyrodes fictilium (Hentz 1850). The elongated palpal tibia and femur of A. fictilium are peculiar, but the male and female genitalia leave no doubt about the species’ identity nor its generic af- finity. According to Exline & Levi (1962), this species is rare but ranges from southern Can- ada to Panama. In a recent revision of the Neotropical species of the genus, Gonzalez & Castro (1996) confirmed the distributional pat- tern of A. fictilium, including its southern Ca- nadian localities and its presence in South America with new records from Argentina. We present new data concerning the distri- bution and phenology of A. fictilium at its northern limit and provide a new illustration of the male palp. The data (Table 1) used to build the distribution map and the phenology graph were gathered from three sources: lit- erature, private and public collections, and our own collection. Genitalia of available speci- mens were studied with a Nikon SMZ-U, and Fig. 1 was done with a camera lucida and a hybrid technique of traditional line drawing with India Ink and computer drawing on Mac- intosh. Males of almost all American species of the genus Argyrodes, except for A. fictilium, have a cephalic projection or protuberance which is characterized by an unmodified, rather flat cephalothorax (Exline & Levi 1962). The gen- italia of Quebec and Ontario specimens were studied (Fig. 1) and the specimens from Can- ada are identical to those occurring from the United States (Exline & Levi 1962) to Argen- tina, including Cuba and Jamaica (GonzMez & Castro 1996). While the genus Argyrodes is well known for its peculiar kleptoparasitic feeding habits (Tanaka 1984; Whitehouse 1988, 1997), A. fictilium has been reported to prey on Araneus sp., Frontinella communis (Hentz 1850), Phil- oponella oweni (Chamberlin 1924), Fronti- nella sp. and an unidentified linyphiid (Archer 1946; Exline & Levi 1962; Trail 1981; W Shear pers. comm.). Trail (1981) has suggest- ed that large Argyrodes species prey on other spiders while smaller species are kleptopar- asites, but Tanaka (1984) showed that small Argyrodes also prey on big spider species. Present knowledge suggests that all species studied so far in the Rhomphaea species group, including A, fictilium, are araneophagic rather than kleptoparasites (Archer 1940; Whitehouse 1987, pers. comm.). The distribution map (Fig. 2) confirms the wide and scattered distribution of the species at its northern limit, with records from the eastern and western regions of Canada. In Ex- line & Levi (1962) and GonzMez & Castro (1996), the northern limit “southern Canada” referred to the records from Vancouver Island (British Columbia) and Lake Temagami (On- 238 PAQUIN & DUPERRE— DISTRIBUTION OF ARGYRODES FICTILIUM 239 Figure 1. — Argyrodes fictilium from Nouveau- Quebec Territories (49°48'N, 78°54'W) Quebec, palpus, ventral view. tario). Some new records presented here show a much more northerly distribution than pre- viously reported. In particular, the records of Paquin and Duperre (Table 1) in the black spruce forest of eastern Canada and the one by Aitchison-Benell & Dondale (1992) in Manitoba extend the species’ distribution to the boreal region. Two hypotheses are formulated here con- cerning the distribution of A. fictilium in the boreal region, this species being of a Neo- tropical origin (GonzMez & Castro 1996). The first explains the northern distribution as the sporadic northern extension of a species usu- ally confined to more southern latitudes. On a smaller geographical scale, the variability of environmental conditions such as temperature and moisture would allow its sporadic pres- ence in more northern latitudes. The instabil- ity of such conditions would explain its rarity as well as the gaps in its distribution in the boreal area. An example of such a dynamic is given by Neochlamisus comptoniae (Brown) (Coleoptera, Chrysomelidae), an insect that is known to have a northern limit around the U.S. and Canada border (LeSage 1984) while its host plant, Comptonia peregrina (L.) Coul- ter, is present in northern Quebec (Marie-Vic- torin 1964). The second hypothesis states that the dis- tribution of A. fictilium may also include the boreal region, which is delimited by the north- Figure 2. — Distribution of Argyrodes fictilium in Canada. The solid line shows the northern limit of the boreal forest. 240 THE JOURNAL OF ARACHNOLOGY Table 1. — Collection data from collections and literature. (CNC: Canada National Collection, DEC: Donald J. Buckle Collection, LLC: Laurent LeSage Collection, CPAD Pierre Paquin and Nadine Duperre Collection). Data source Province and locality Date of collection Sex Notes From collections CNC BRITISH COLUM- BIA: Qualicum beach 23 May 1946 2 F CNC NOVA SCOTIA: Ca- nard 05 Sept. 1956 M Apple trees CNC ONTARIO: Odessa 04 July 1963 F — CNC ONTARIO: Belleville (Field station) 09 June 1960 F Pine CNC NEW BRUNSWICK: Fredericton 02-03 July 1969 F Balsam fir CNC ONTARIO: Markdale 17 June 1988 F Cedar fen beating CNC ONTARIO: Ottawa 03-09 July 1989 M Damp acer/betula wood DEC NOVA SCOTIA: Ma- hone Bay 24 Sept. 1996 Juv. Balsam fir LLC QUEBEC: Pontiac, Lake Davis (N of Fort-Coulonge) 24 August 1991 Juv. Mixed forest, beating LLC QUEBEC: Pontiac; Lake Davis (N of Fort-Coulonge) 01 Sept. 1990 Juv. Mixed forest, beating CPAD QUEBEC: Nouveau- Quebec Territories; 49°48'N, 78°54'W 06-13 July 1997 M Burned Black spruce forest, flight inter- ception trap CPAD QUEBEC: Nouveau- Quebec Territories; 49°48'N, 78°54'W 29 June-6 July 1997 M Mature Black spruce forest, flight inter- ception trap CPAD QUEBEC: Abitibi; Lake Duparquet; 48°30'N, 79°13'W 24-30 June 1997 M Mature Jackpine forest, soil emerging cage From literature Emerton (1920) QUEBEC: Outaouais; Hull — — — Kurata (1943) ONTARIO: Nipissing Co.; Lake Temagami 20 August 1937 F — Exline & Levi (1962) ONTARIO: Nipissing Co.; Lake Temagami — — — BRITISH COLUMBIA: Pender Harbour BRITISH COLUMBIA: Wel- lington BRITISH COLUMBIA: Nanai- mo Aitchison-Bennel & Dondale MANITOBA: STSO'N, 95°00'W (C4) — — Boreal forest; decidous woods; tree foliage (1992) PAQUIN & DUPERRE— DISTRIBUTION OF ARGYRODES FICTILIUM 241 Table 1. — Continued. Data source Province and locality Date of collection Sex Notes LeSage & Hutch- inson (1992) QUEBEC: Maski- nonge; St-AngMe 12 July 1990 Juv, Mixed forest, beating QUEBEC: Gatineau; Aylmer 08 Sept. 1990 Juv. Maple-beech forest, beating QUEBEC: Pontiac; Lake Davis (N of Fort-Coulonge) 01 Sept. 1990 Juv. Mixed forest, beating QUEBEC: Pontiac; Lake Davis (N of Fort-Coulonge) 30 Sept. 1990 6 Juv. Mixed forest, beating ern tree limit (Danks & Footitt 1989) (Fig. 2). This hypothesis partially relies on the records of Aitchison-Benell & Dondale (1992) from Manitoba, and Paquin and Duperre (Table 1) from Quebec, both from the boreal region but in two different ecological contexts. As men- tioned by Scudder (1979), the boreal ecolog- ical zone forms a vast transcontinental belt and the largest continuous vegetation associ- ation in North America. This wide ecological zone is roughly divided into a southern and a northern part. The southern part (also called mixed-boreal) is dominated by deciduous (as- pen and birch) and coniferous forest (white spruce, balsam fir, and white cedar) while the northern part is mainly dominated by black spruce (Grandtner 1966; Rowe 1972). In east- ern Canada, the boreal belt is under the strong climatic influence of Hudson and James Bays. In this area, northern conditions are met at a lower latitude than anywhere else in the coun- try (Danks 1979). This distribution pattern is shown by the black spruce distribution across Canada (Rowes 1972). Very little is known about the biological traits of A. fictilium and its ecological prefer- ences. Its distribution pattern might, however, reflect a wide range of potential prey rather than a limited distribution of a specific host. The known prey of A. fictilium are likely to find suitable web substrate everywhere within the tree limit; and the records of A. fictilium indicate an association with forest habitats, particularly coniferous forest (Table 1). This suggests that food would not be a limiting fac- tor. Other abiotic factors that might limit its northern distribution are largely unknown. There is little evidence to favor either of the two hypotheses for the distribution of A. fic- tilium. Nevertheless, the hypothesis for a dis- tribution that covers the boreal area rather then a sporadic extension seems to be better supported when the records from Quebec are considered. Even though the record of A. fic- tilium from Manitoba is the most northern for the species in regards to the latitude, it does not indicate an extension of the species range into the northern boreal region because it is reported from the mixed-boreal forest. How- ever, the records from the black spruce forest of Quebec allow such a range extension into the northern boreal area. Despite the lower lat- itudes of the localities, these records come from an ecological zone that is more repre- sentative of northern conditions as shown by the vegetation belt and climatic data (Danks 1979). The presence of A. fictilium in the black spruce forests of Quebec clearly indicate that the species occurs in the northern boreal area; and, according to the second hypothesis, its presence can be expected throughout this wide ecological region. Figure 3 shows periods of collection for the known Canadian specimens of A. fictilium. Mature females are active from the end of May until the beginning of July while males seem to appear later in the season, from the end of June until September. Juveniles are mainly reported in September, and we assume these to be the overwintering stages. Accord- ing to Tretzel (1954) there are three types of phenology: 1) stenochronous where adults are present in a definite period of the year, 2) eu- rychronous where adults are present all year long (with or without a definite reproduction period), and 3) winter-mature. The male peak of abundance is considered to be the indicator of the maturation period. Aitchison (1984), 242 THE JOURNAL OF ARACHNOLOGY 6 -r 5 - Number 4 of individuals 3 - 2 - 1 May H Female n Male ■ Juvenile June a n = 24 M July August U September Figure 3. — Seasonal distribution of immature, male, and female of Argyrodes fictilium in Canada. however, refined that terminology by dividing the stenochronous type into three classes: spring-, summer- and autumn-stenochronous. Despite the fact that it is difficult to confirm a phenological type with so few specimens, the summer-stenochronous class fits our data, mature males being collected mainly in July. Argyrodes fictilium is a rare species within its northern limits and it is difficult to study its biology. Only 30 specimens are known from Canada, most of them from Quebec and Ontario where the collection intensity may have been higher than in other parts of the country. It is also difficult to see a clear pat- tern in its distribution because of the rarity of the species and the lack of collections. How- ever, the present state of knowledge allows a hypothesis that may link the forested portion of the territory to its distribution, including the boreal forest. It is surprising to see a spe- cies with such wide range of habitat occurring from South America to the boreal forest. Fu- ture collections, especially in poorly studied regions such as the northern boreal forest in central and western part of Canada, are likely to yield more specimens and confirm its phe- nology and distributional pattern. ACKNOWLEDGMENTS We are grateful to D. Buckle, J.H. Redner, W. Shear and L. LeSage for sharing time, data and knowledge. We are also grateful to C.D. Dondale, Grace Hall, Mary Whitehouse and an anonymous reviewer for their constructive comments on the manuscript. LITERATURE CITED Aitchison, C.W. 1984. The phenology of winter- active spiders. Journal of Arachnology 12:249- 271. Aitchison-Benell, C.W. & C.D. Dondale. 1992 (1990). A checklist of Manitoba spiders (Ara- neae) with notes on geographic relationships. Naturaliste Canadien 117:215-237. Archer, A.F. 1940. The Argiopidae or orb- weaving spiders of Alabama. Alabama Museum of Natu- ral History, paper 14:1-77. Archer, A.F. 1946. The Theridiidae or comb-footed spiders of Alabama. Alabama Museum of Natu- ral History, paper 22:1-67. Danks, H.V. 1979. Physical basis for Canadian re- gional diversity. Pp. 7-30. In Canada and Its In- sect Fauna (H.V. Danks, ed.). Memoirs of the Entomological Society of Canada, No. 108. Danks, H.V. & R.G, Footitt. 1989. Insects of the boreal zone of Canada. Canadian Entomologist 121:625-690. Emerton, J.H. 1920. Catalogue of the spiders of Canada known to the year 1919. Transactions of the Royal Canadian Institute 12:309-338. Exline, H. & H.W. Levi. 1962. American spiders of the genus Argyrodes (Araneae: Theridiidae). Bulletin of the Museum of Comparative Zoology 127:75-204. Gonzalez, A. & D. del C, Castro. 1996. Neotrop- ical spiders of the genus Argyrodes Simon (Ar- aneae, Theridiidae). Bulletin of the British Ar- achnological Society 10:127-137. Grandtner, M.M. 1966. La vegetation forestiere du Quebec meridional. Les Presse de I’Universite Laval, Quebec, Canada. 216 pp. Kurata, T.B. 1943. The spiders of the Lake Nipis- sing and Lake Temagami regions, Ontario. Ca- nadian Field Naturalist 57:9-13. PAQUIN & DUPERRE— DISTRIBUTION OF ARGYRODES FICTILIUM 243 LeSage, L. 1984. Immature stages of Canadian Neochlamisus Karren (Coleoptera: Chrysomeli- dae). Canadian Entomologist 116:383-409. LeSage, L. & R. Hutchinson. 1992. Une des plus extraordinaires araignees du Quebec Argyrodes fictilium (Hentz) (Araneae: Theridiidae). Pirata 1: 127-129. Marie- Victorin. 1964. La Flore Laurentienne. 2eme Edition. Les Presses de I’universite de Montreal, Montreal. 925 pp. Rowe, R.S. 1972. Forest Region of Canada. En- vironment Canada. Publ. No. 1300. Ottawa, On- tario, Canada. 172 pp. Scudder, G.G.E. 1979. Present patterns in the fauna and flora of Canada. Pp. 87-179. In Canada and Its Insect Fauna (H.V. Banks, ed.). Memoirs of the Entomological Society of Canada, No. 108. Trail, D.S. 1981 (1980). Predation by Argyrodes (Theridiidae) on solitary and communal spiders. Psyche 87:349-355. Tanaka, K. 1984. Rate of predation by a klepto- parasitic spider, Argyrodes fissifrons upon a large host spider, Agelena Umbata. Journal of Arach- nology 12:363-367. Tretzel, E. von. 1954. Reife- und Fortpflanzung- szeit bei Spinnen. Zeitschrift fuer Morphologic und Okologie der Tiere 42:634-691. Whitehouse, M.A.E. 1987. Spider eat spider: The predatory behaviour of Rhomphaea sp. indet. from New Zealand. Journal of Arachnology 15: 355-362. Whitehouse, M.A.E. 1988. Factors influencing specificity and choice of host in Argyrodes an- tipodiana (Theridiidae). Journal of Arachnology 16:349-355. Whitehouse, M.A.E. 1997. The benefits of stealing from a predator: Foraging rates, predation risk and intraspecific aggression in the kleptoparasitic spider Argyrodes antipodiana. Behavioral Ecol- ogy 8:663-667. Manuscript received 25 October 1999, revised 22 January 2001. 2001. The Journal of Arachnology 29:244-248 EGG SAC RECOGNITION BY FEMALE MIAGRAMMOPES ANIMOTUS (ARANEAE, ULOBORIDAE) Brent D. Opell: Department of Biology, Virginia Polytechnic Institute and State University, Blacksburg, Virginia 24061 USA ABSTRACT. After producing a cylindrical egg sac, a female Miagrammopes animotus holds it until spiderlings emerge and disperse. When sacs were taken from females, these females exhibited a putative searching behavior and, upon contacting either their sacs or those of conspecifics, exhibited a putative recognition behavior. These responses would cause a female to search for and reclaim her sac if it were temporarily abandoned during feeding or web construction. Females with sacs did not respond positively to sacs from which spiderlings had emerged. Females that did not have sacs did not respond positively to viable sacs. Females separated from their sacs for increasing time periods exhibited a decline in positive responses to their sacs. Thus, contact with the sac appears necessary to maintain an affinity for the sac during the development of spiderlings. Keywords: Maternal care, spider Females of the simple-web species Mia- grammopes animotus Chickering 1968 pro- duce cylindrical egg sacs consisting of two columns of eggs surrounded by two thin lay- ers of silk (Lubin et al. 1978; Opell 1984). During the day, a female attaches her egg sac along one of the web’s non-sticky lines and aligns herself with the egg sac, her abdomen touching the egg sac, her legs I and II extend- ing directly anteriorly, and her legs III and IV extending directly posteriorly (Fig. 1; Lubin et al. 1978; Opell 1989a). This posture en- hances the twig-like appearance of both the female and her egg sac. Miagrammopes ani- motus females range in color from light tan to dark, reddish-brown and produce egg sacs whose wrapping silk is similar in color to their bodies (Opell 1989a). This makes it even more difficult to distinguish a spider and her egg sac and further enhances the cryptic ap- pearance of each. At dusk a female detaches her egg sac from the line and holds the sac with her first leg as she monitors her web (Fig. 2; Lubin et al. 1978; Opell unpubl. obs.). This suggests that her linear, day-time posture is a defense against visually hunting predators like insects and birds. When spiderlings emerge from an egg sac as second instars, they lack functional cribella and cling to the egg sac for several days until they molt to the third instar (Lubin et al. 1978; Opell 1989b), at which time they disperse and begin constructing capture webs. Females continue to tend their egg sacs until spiderlings leave (Fig. 3). While collecting M. animotus in conjunc- tion with studies of their cribellar threads and cribella, I routinely collected females with egg sacs. When I attempted to remove a female’s egg sac so that I could weigh and measure her, she held tenaciously to her egg sac. Unless all of her legs were removed from the egg sac, she quickly regained her firm grasp. After I separated a female from her egg sac and placed her on a horizontal surface, she walked rapidly and made broad, rapid sweeping movements with her first legs. When she con- tacted her egg sac, she held it tightly with her first two pairs of legs, immediately pressed her chelicerae to its surface for a few seconds (although I could not determine if she bit the egg sac’s silk covering), and then immediately firmly grasped the egg sac with her first and second legs. When a female with an egg sac was placed on a horizontal surface, she often attached a dragline to her egg sac and walked away, possibly searching for a secure site to climb. When either the egg sac or dragline was touched, the female ran quickly to her egg sac, pressed her chelicerae to its surface and then grasped the egg sac. I interpreted this rapid walking and leg waving behavior as searching behavior, the cheliceral contact as egg sac evaluation be- 244 OFELL^MIAGRAMMOPES EGG SAC RECOGNITION 245 Figures 1“3. — Miagrammopes animotus females tending egg sacs during the day (1), at night (2), and after spiderlings have emerged (3). havior, and the subsequent grasping of the egg sac as positive egg sac recognition. To better understand this behavior, I evaluated the re- sponses of females that had no egg sacs, in- vestigated the specificity of egg sac recogni- tion, and determined if this behavior was expressed after a female had been separated from her egg sac for different lengths of time. METHODS Spiders and egg sacs were collected at the Center for Energy and Environment Re- search’s El Verde, Puerto Rico field station. Voucher specimens are deposited in Harvard University’s Museum of Comparative Zoolo- gy. All observations were made in a window- less laboratory that was illuminated by fluo- rescent lights and had a temperature of 20-24 °C and a relative humidity of 60-63%. After removing a female’s egg sac, I placed each in a separate, clean glass vial stoppered with a cotton plug. A small piece of moist cotton was placed with each spider. Vials were numbered so that I could identify each spi- 246 THE JOURNAL OF ARACHNOLOGY der’s egg sac, though only by referring to a record book. Observations were conducted by placing a female in a clean, 40 mm diameter aluminum weighing pan and observing her re- sponse to an object or an egg sac. I define a positive response to an object or an egg sac as the female clinging tightly to the object af- ter contacting it with her chelicerae so that she could not be easily dislodged by repeated prodding with a small artist’s brush. Occa- sionally, a female rested momentarily on an egg sac or object, but did not bring her che- licerae to its surface. In these cases a light prod with a brush caused her to leave the ob- ject and continue walking and I scored this as a negative response. I examined the duration of egg sac recog- nition in 82 females that were collected with egg sacs. After removing their egg sacs, I kept these spiders for periods of 5-25 h. At the end of each of these 21 periods the responses of 2-10 females to their own egg sacs were ob- served. Thirty-four of the females that re- sponded positively to their egg sacs were used for a second trial after each was kept in a vial with her egg sac for 12-24 h to permit pro- longed contact with her sac. RESULTS Eleven mature females that had no egg sacs, including five whose abdomens were clearly swollen with eggs, did not cling to egg sacs or press their chelicerae to their surfaces when each was presented with 7-10 egg sacs of conspecifics, nor did they cling tightly to pieces of grass or cotton that I placed in their collecting vials. In contrast, 26 of 32 females whose egg sacs were removed 5-6 h earlier responded positively to their reintroduced egg sacs. The response of each group differed from a null model in which 50% of the indi- viduals responded positively to egg sacs (x^ — 11.0 and 12.5, respectively, P < 0.001). Two females that responded positively to their own egg sacs also responded positively to egg sacs produced by nine other females. Seven females that responded positively to their own egg sacs also responded positively to an egg sac of the closely related species M. pinopus Chickering 1968 from St. John, U.S. Virgin Islands. However, they did not respond positively to a piece of cotton, a section of wooden applicator stick similar in size to an egg sac, or to an egg sac of Uloborus glo- mosus (Walckenaer 1841) from Blacksburg, Virginia. Another five females that responded positively to egg sacs that contained eggs re- sponded negatively to egg sacs from which spiderlings had emerged, walking over these egg sacs and continuing to exhibit searching behavior. Figure 4 shows the percentages of females that responded positively to their egg sacs af- ter times of increasing separation. In each of the first six time intervals, the responses of females used in a second trial did not differ from that of females that were used in only one trial (x" < 0.425, I df, P > 0.50). How- ever, in the last interval (23-25 h) five of the six females used in a second trial responded positively to their egg sacs, a greater number than predicted by the responses of females used in only one trial (x^ = 16.000, I df, P < 0.001). The median time of these seven peri- ods regresses against the percent positive re- sponse (Y = -2.06X + 92.92; F = 8.20, P = 0.035, = 0.62), indicating that a female’s ability to identify and respond to her egg sac decays with increasing separation. Additional support for this decay comes from a compar- ison of the responses when grouped into short, intermediate, and long periods of egg sac sep- aration (5-9 h, n = 35; 10-20 h, n = 48; and 21-25 h, n = 33; respectively). For this com- parison, I determined the mean positive re- sponse for each of the 21 hourly trials and then compared the grand means of responses in the three periods. A Kruskal- Wallis test showed that the values of these periods (X ± 1 SD: 78.7% ± 19.7, 66.0% ± 27.4, and 38.9% ± 7.9, respectively) differed (x^ = 6.47, P = 0.039). DISCUSSION These observations indicate that sensory stimuli from one or several sources unique to oviposition or egg sac construction cause per- sistent but reversible changes in a spider’s physiology that alter its response to egg sacs. Possible sources of these stimuli include the pressure exerted on the oviducts as eggs pass through them and the activation of tubuliform silk glands and spigots that, in uloborids, ap- pear to be used only for egg sac production (Kovoor 1977; Foelix 1996; Opell 1984). Copulation is an unlikely source of stimuli, as females store sperm and mating may occur long before eggs are fertilized and deposited. OmUL—MIAGRAMMOPES EGG SAC RECOGNITION 247 Hours of Egg Sac Separation Figure 4. — Positive egg sac recognition following periods of increasing separation of a female and her egg sac. Bars represent the means of the three-hour periods. The sample size is given within each bar. The number of females used in a second trial is given in parentheses. Once a positive response to an egg sac has been established it appears to be general and does not permit a female to distinguish her egg sac from those of conspecifics. However, unlike wolf and fishing spiders, M. animotus females can not be tricked into accepting sub- stitute objects (Gertsch 1979; Foelix 1996). Egg sac recognition persists for varying pe- riods of time, but lasts long enough to cause a female to search thoroughly for her egg sac if she left it. The regression formula for the decay of positive responses to egg sacs sug- gests that egg sac recognition disappears in all females after they are separated from their egg sacs for about 45 h. Continual or frequent con- tact with an egg sac appears to be necessary to maintain a female's positive response to her egg sac during the approximately 20 days re- quired for the eggs to develop and the spider- lings to emerge from the egg sac (Opell 1979, 1982; Peaslee & Peck 1983). Unlike members of other uloborid genera, Miagrammopes females have no permanent attachment site for their egg sacs. At night fe- males probably temporarily attach the egg sac to a line while renewing their capture webs and feeding. This behavior has not been ob- served in M. animotus; but Lubie et al. (1978) report that females of an unidentified Mia- grammopes species hung their egg sacs from threads at night and resumed prey capture ac- tivity until dawn, at which time they again tended their egg sacs. Egg sac recognition would cause a female to search for her egg sac if she were forced to abandon it during the day or to anchor it while building or repairing her capture web at night. I did not investigate cues that might permit 248 THE JOURNAL OF ARACHNOLOGY females to distinguish viable egg sacs from those that the young have abandoned. The masses of viable and empty egg sacs differ; but, in my laboratory observations, females did not appear to lift egg sacs to assess their masses. Second instar spiderlings may deposit silk on the egg sac’s outer surface and this may mask the egg sac silk that allows a fe- male to identify an egg sac. Alternatively, the female may identify and respond negatively to the silk of these spiderlings. The loss of egg sac recognition has advan- tages for M. animotus females that appear to produce several egg sacs during a lifetime and often reach high densities. A female’s nega- tive response to an egg sac from which spi- derlings have emerged and dispersed allows her to shift her activity from egg sac tending to web building and foraging so that another egg sac can be produced. This negative re- sponse may also reduce intraspecific conflict. At El Verde, I often observed 3-5 mature fe- males on a single under-story plant or cluster of vegetation occupying a space of about 1 m^. Under these high densities, the loss of egg sac recognition would eliminate contests be- tween females that were tending egg sacs and those from whose egg sacs spiderlings had re- cently dispersed. In the absence of individual- specific egg sac recognition, the failure of fe- males to respond to old egg sacs also assures that they will not mistakenly abandon viable egg sacs for these discarded egg sacs that sometimes remain suspended from vegetation. LITERATURE CITED Foelix, R.G. 1996. Biology of Spiders. 2^^ ed. Ox- ford Univ. Press, New York. Gertsch, WJ. 1979. American Spiders. Van Nos- trand Company, Inc., New York. Kovoor. J. 1977. Lappareil sericignene dans le genera Uloborus Latr. (Araneae, Uloboridae). I. Anatomic Revue Arachnologique 1:89-102. Lubin, Y.D., W.G. Eberhard & G.G. Montgomery. 1978. Webs of Miagrammopes (Araneae: Ulo- boridae) in the Neotropics. Psyche 85:1-23, Opell, B.D, 1979. Revision of the genera and trop- ical American species of the spider family Ulo- boridae. Bulletin of the Museum of Comparative Zoology 148:433-549. Opell, B.D. 1982, Post-hatching development and web production of Hyptiotes cavatus (Hentz) (Araneae, Uloboridae). Journal of Arachnology 10:185-191, Opell, B.D. 1984, Eggsac differences in the spider family Uloboridae (Arachnida: Araneae). Trans- actions of the American Microscopical Society 103:122-129. Opell, B.D. 1989a. Do female Miagrammopes an- imotus (Araneae: Uloboridae) spin color-coordi- nated egg sacs. Journal of Arachnology 17:108- 111. Opell, B.D. 1989b. Functional associations be- tween the cribellum spinning plate and prey cap- ture threads of Miagrammopes animotus (Ara- neida, Uloboridae). Zoomorphology 108:263- 261, Peaslee, J.E. & W.B, Peck. 1983, The biology of Octonoba octonaria (Muma) (Araneae, Ulobor- idae). Journal of Arachnology 11:51-67, Manuscript received 20 June 2000, revised 28 No- vember 2000. 2001. The Journal of Arachnology 29:249-252 EGG COVERING BEHAVIOR OF THE NEOTROPICAL HARVESTMAN PROMITOBATES ORNATUS (OPILIONES, GONYLEPTIDAE) Rodrigo Hirata Willemart: Departamento de Zoologia, Institute de Biociencias, Universidade de Sao Paulo, Caixa Postal 11461, 05422-970, Sao Paulo, SP, Brazil ABSTRACT. The egg covering behavior of the laniatorid harvestman Promitobates ornatus was studied. Females of this species laid eggs isolated, on soil. After laying an egg, the female started scraping the substrate next to the egg, picking up debris, and attached the earth particles to the egg. After she scraped one area, she rotated around the egg, stopped turning, and restarted the collection of debris from another site. Alternation of scraping and changing body position was repeated twice or more until the female completed the egg covering. Data on egg size, duration of egg laying and egg covering, and duration of embryonic development are also provided. Keywords: Laniatores, Mitobatinae, biology, care, maternal investment In the Laniatorine suborder of Opiliones, females lay eggs that are either clustered (Ca- nals 1936; Capocasale & Bruno-Trezza 1964; Mitchell 1971; Juberthie & Munoz-Cuevas 1971; Matthiesen 1975; Goodnight & Good- night 1976; Pinto-da-Rocha 1993; Ramires & Giaretta 1994; Gnaspini 1995; Machado & Oliveira 1998) or isolated (Canals 1936; Jub- erthie 1965, 1972; Cokendolpher & Jones 1991), on a large variety of substrates, such as leaves, moss, rocks, bark crevices and soil. Among the Laniatores (unless otherwise in- dicated, all species mentioned below belong to the family Gonyleptidae), different forms of parental investment have been described in the literature, ranging from the oviposition site selection to egg guarding. Egg guarding has been observed in one species of Cosmetidae and one of Stynopsidae and in seven species of Gonyleptidae (see Gnaspini 1995 for ref- erences), and is usually performed by females. Paternal care has seldom been reported (Rod- riguez & Guerrero 1976; Mora 1990; Martens 1993) and there is no record of biparental care in harvestmen, although Machado & Oliveira (1998) reported males of Goniosoma longipes (Roewer 1913) near the eggs, and taking care of eggs when the female was experimentally removed. Scotolemon lespesi Lucas 1860 (Juberthie 1965), Cynorta cubana (Banks 1909) (Cos- metidae) (Juberthie 1972), Pachylus quina- mavidensis Munoz-Cuevas 1969 (Juberthie & Munoz-Cuevas 1971), Vonones sayi (Simon 1879) (Cosmetidae) (Cokendolpher & Jones 1991) as well as two other species of cosme- tids and six species of gonyleptids (Canals 1936) are known to cover eggs with debris. The behavior of covering eggs has never been described in detail. The only mention of how egg covering occurs was by Canals (1936), who reported “scraping of the sub- strate with the anterior legs” by the female. Again, he did not specify which species did this. This paper provides the first detailed de- scription of egg covering behavior in harvest- men, based on data from Promitobates orna- tus (Mello-Leitao 1922) (Mitobatinae). Three female P. ornatus were used for this study. One of them (identified as Po\) was collected on 24 January 1999 in Carlos Bo- telho State Park, Sao Miguel Arcanjo county. The other two (identified as Pol and Pol) were collected on 27 July 1999 in Paranapia- caba (= Alto da Serra), Santo Andre county. Both localities are representative of tropical rain forest in Sao Paulo state, southeastern Brazil. I maintained Pol with a conspecific male at room temperature in a terrarium with damp soil, a wet piece of cotton, and hard surfaces such as stones and plastic blocks. Pol and Pol were kept in a second terrarium under the same conditions, but with six other conspecifics including males and females. In both of the cases, the artificial light : dark pe- riods were irregularly distributed throughout 249 250 THE JOURNAL OF ARACHNOLOGY Figure 1. — Drawing of an egg of Promitobates ornatus after the covering was completed, showing soil particles (black spots) and a fragment of root (arrow) attached to it. the day. The harvestmen were fed once a week with dead arthropods such as isopods, mosquitoes, drosophilids, pieces of Tenebrio obscurus larvae and a variety of plant items (papaya, sugar beet, boiled carrots, beans and rice) and industrial food (cream cheese, cooked ground beef, and bread). They ac- cepted all the items mentioned. All observa- tions were conducted between August 1999 and December 1999. Females of P. ornatus laid isolated eggs over soil surfaces. During oviposition, the fe- male P. ornatus stood at legs III and IV, with legs I and II extended forward. The ovipositor extended forward to the genital operculum, at 20° below the horizontal body axis, and the egg slid slowly along it, until the distal part of the ovipositor was reached. At this mo- ment, the female bent the ovipositor bringing it close to the substrate and deposited the egg. Only one egg was laid in each event. In the two cases in which I observed nearly the en- tire act of oviposition, the times spent for one egg to be laid were 3.4 and 3.5 min. The mean egg length was 1.29 ± 0.16 mm (ji = 8, range = 1.05-1.40 mm), approximately 25% of the female body length (5.10, 5.15, and 5.20 mm). Females laid eggs in the morning {n = 5\ one not included in Table 1), afternoon {n = 3) and at night {n = 2), and so apparently did not favor a particular time of day for ovipo- sition. The time spent by P. ornatus to lay one egg was similar to that in other laniatorean spe- cies— e.g., 4-12 min for Pachylus quinamav- idensis (Juberthie & Munoz-Cuevas 1971). The general egg-laying behavior was also similar among the species studied so far, and follows the general description of Juberthie & Munoz-Cuevas (1971). However, after laying an egg, P. ornatus waved legs I over the egg occasionally touching it. Thereafter, the fe- male started scraping the substrate next to the egg with alternate movements of legs I, pick- ing up debris. She then raised legs I and strongly pressed them simultaneously or one at a time against the egg, leaving earth parti- cles attached to it. While scraping, some big- ger particles were occasionally brought near the egg, without adhering to it. After she scraped the substrate from one area, she ro- tated around the egg, stopped turning, and re- started the collection of soil particles from an- other site. The female’s rotation was either clockwise or counterclockwise, with no ap- parent rule concerning direction or angle of rotation. Alternation of scraping and changing body position was repeated twice or more un- til the egg covering was complete (Table 1). The mean time spent during egg covering was 37 ± 1 1 min {n = 9, range = 20-50 min). Occasionally, between two events of scraping, the female would pass her legs I between the chelae of her chelicerae. This explains why the total time is greater than the sum of partial time periods in Table 1. Before leaving the site, the female tapped the substrate around the egg with the first pair of legs. In one case, 2.3 h elapsed between covering one egg and laying the next one. Promitobates ornatus apparently does not always choose an appropriate site for collec- tion of soil particles. Female Pol twice laid an egg in sites where she was unable to turn herself around the egg, although she tried to, because the egg was laid too close to a vertical substrate. In addition, females Pol and Pol were observed scraping stones instead of earth surfaces, using the same behavioral patterns described earlier. Thus, the quality of the sub- strate used for collection of soil particles is probably not the factor that determines the time spent in egg covering. It should be noted, however, that the females always laid their eggs on soil, indicating that they probably rec- ognized and selected soil surfaces for ovipo- sition. Females of P. ornatus did not abort egg lay- ing and egg covering when disturbed by light WILLEMART— EGG COVERING BEHAVIOR OF HARVESTMAN 251 Table 1. — Change of body position during nine covering events by three females of Promitobates ornatus {Pol, Pol, and Po?>). The second column represents the positions adopted by the female. In all cases, 0° is horizontally at left and the angles of rotation have to be counted clockwise. Partial time periods follow the sequence of the location of the female relative to the egg. The lines are organized by animal and hour. Female Location of female relative to egg Partial time periods (min) Total time (min) Hour when egg covering started Pol 07225707607071 107180° 4/1 1/4/1/4/9/2 36 0830 Pol 0°/2257180°/270° 13/9/6/5 34 0919 Pol 0°/225°/0°/657135° 9/10/7/10/5 45 1058 Pol 0°/270°/457135° 7/12/6/19 45 1135 Pol 0°/907l35° 12/6/5 24 1500 Pol 0°/90°/225° 6/7/6 20 0010 Pol 07180°/340° 24/10/16 51 1836 Po3 0727071 80°/90° 15/10/4/19 49 1515 Po3 0°/45°/290° 10/11/7 29 1950 (n “ 5) or by the approach of other harvest- men of approximately the same body size [a conspecific male (« = 1) and Ilhaia cuspidata Roewer 1913 male introduced in the terrarium {n = 1)]. On one occasion, a female stopped egg covering and remained motionless when touched on the dorsum with a thin paintbrush. In this case, she waved her second pair of legs searching for the stimulus. Fleeing only oc- curred when she touched the paintbrush with her second pair of legs. This reluctance to abandon the eggs has been described in two other Gonyleptidae. Light did not cause females of Goniosoma proximum (Mello-Leitao 1922) with eggs to flee (Ramires & Giaretta 1994) and females of Acanthopachylus aculeatus (Kirby 1819) guarding eggs fled only under very intense light (Capocasale & Bruno-Trezza 1964). However, in contrast with the behavior dis- played by P. ornatus, females of Pachylus quinamavidensis, while laying an egg, reacted to approaching conspecific males springing with the palps extended towards the male (Juberthie & Munoz-Cuevas 1971). No droplets of exocrine gland secretion were noticed on P. ornatus' body while laying or covering an egg, but it could be that secre- tions are added to the legs as they are passed between the chelicerae. Clawson (1988) noted females of two species of Palpatores would rub the exocrine gland openings over their oviposition sites, and suggested this behavior was to mark the sites. An average of 30 ± 4.97 days {n = 5; range = 23-36) of embryonic development was necessary for nymphs of P. ornatus to hatch. less than the 30-60 days found for Gonioso- ma spelaeum (Mello-Leitao 1933) (Gnaspini 1995) and the 45-64 days for Goniosoma lon- gipes (Machado & Oliveira 1998). These dif- ferences are tentative since temperature great- ly influences the duration of embryonic development, and as mentioned above, the laboratory temperature was not controlled during this study. Egg development took 16- 27 days for Cynorta cubana at 20-28 °C (Jub- erthie 1972), 13 days at 26 °C and 23-27 days at 20 °C for Erginulus clavotibialis (Cam- bridge 1904) (Goodnight & Goodnight 1976), and 20-38 days for Vonones sayi with the temperature ranging from 5-20 °C (Coken- dolpher & Jones 1991). Several invertebrates are known to feed on harvestmen eggs — from conspecifics to flat- worms, ants, reduviid bugs, staphylinid beetle larvae, and crickets (Capocasale & Bruno- Trezza 1964; Juberthie & Munoz-Cuevas 1971; Mora 1990; Gnaspini 1995; Machado & Oliveira 1998). In that context, the energy invested in parental care may be justified be- cause of the resulting (presumed) protection from predation (Alcock 1993). Egg covering in P. ornatus lasts an average of 37 min, and the process repeats for each egg laid. Never- theless, this investment is certainly not as costly as the investments made by females of Goniosoma spelaeum and G. longipes, which lay clustered eggs and stay with their off- spring until the dispersion of the nymphs (sometimes 60-80 days between laying eggs and dispersion) (Gnaspini 1995; Machado & Oliveira 1998). 252 THE JOURNAL OF ARACHNOLOGY By laying isolated eggs P. ornatus avoids the risk of losing several eggs if an egg hap- pens to be noticed by a predator. Covering eggs with debris is interpreted as a way to hide them from predators (Canals 1936; Jub- erthie 1972; Cokendolpher & Jones 1991), thus increasing the chances of the embryo’s survival. I believe that, in addition to the fact that camouflage makes the eggs difficult to be seen, it may also be effective against predators that use tactile clues. A wandering predator may pass over the egg without noticing it be- cause of the soil particles adhered to the egg. In theory, the greater the number of particles attached and the more uniformly they are dis- tributed on the egg surface, the more effective would be the protection. I believe that there is a strong relationship between the act of chang- ing the body position radially around the egg and the effectiveness of the process. As sug- gested by Mitchell (1971) for eggs laid in crevices, egg guarding would be of little se- lective value if the egg is difficult to find. ACKNOWLEDGMENTS This paper has been greatly improved by the suggestions of my advisor R Gnaspini, to whom I am deeply grateful. I am also indebted to J. Berry, J. Cokendolpher, G. Machado, R. Suter and an anonymous referee for helpful comments. I thank M.R. Hara for rearing some animals used in this study and R. Pinto- da-Rocha for identification of the harvestmen. E Garcia provided assistance with computer matters. LITERATURE CITED Alcock, J. 1993. Animal Behavior: An Evolution- ary Approach. 5th ed., Sinauer Associates, Sun- derland. 626 p. Canals, J. 1936. Observaciones biologicas em ar- acnidos del orden Opiliones. Revista de Chilena de Historia Natural 40:61-63. Capocasale, R. & L. Bruno-Trezza. 1964. Biologia de Acanthopachylus aculeatus (Kirby, 1819) (Opiliones: Pachylinae). Revista de la Sociedad Uruguaya de Entomologia 6:19-32. Clawson, R.C. 1988. Morphology of defense glands of the opilions (daddylonglegs) Leiobun- um vittatum and L. flavum (Arachnida: Opili- ones: Palpatores: Phalangiidae). Journal of Mor- phology 196:363-381. Cokendolpher, J.C. & S.R. Jones. 1991. Karyotype and notes on the male reproductive system and natural history of the harvestman Vonones sayi (Simon) (Opiliones, Cosmetidae). Proceedings of the Entomological Society of Washington 93:86- 91. Gnaspini, P. 1995. Reproduction and postembry- onic development of Goniosoma spelaeum, a cavernicolous harvestman from southeastern Brazil (Arachnida: Opiliones: Gonyleptidae). In- vertebrate Reproduction and Development 28: 137-151. Goodnight, M.R. & C.J. Goodnight. 1976. Obser- vations on the systematics, development and habits of Erginulus clavotibialis (Opiliones, Cos- metidae). Transactions of the American Micro- scopical Society 95:654-664. Juberthie, C. 1965. Donnees sur I’ecologie, le de- veloppement et la reproduction des Opilions. Re- vue d’Ecologie et de Biologie du Sol 2:377-396. Juberthie, C. 1972. Reproduction et developpe- ment d’un opilion Cosmetidae, Cynorta cubana (Banks), de Cuba. Annales de Speleologie 27: 773-785. Juberthie, C. & Muhoz-Cuevas 1971. Sur la ponte de Pachylus quinamavidensis (Opilion, Gonylep- tidae). Bulletin de la Societe d’Histoire Naturelle de Toulouse 107:468-474. Machado, G. & PS. Oliveira. 1998. Reproductive biology of the Neotropical harvestman (Gonio- soma longipes) (Arachnida, Opiliones: Gonylep- tidae): Mating, oviposition behaviour, brood mortality, and parental care. Journal of Zoology 246:359-367. Martens, J. 1993. Further cases of paternal care in Opiliones (Arachnida). Tropical Zoology 6:97- 107. Matthiesen, EA. 1975. Sobre a postura de Disco- cyrtus pectinifemur Mello-Leitao, 1937 (Opili- ones, Gonyleptidae). Ciencia e Cultura (Sao Pau- lo) 27 suppl:372. Mitchell, R.W. 1971, Egg and young guarding by a Mexican cave-dwelling harvestman, Hoplobu- nus boneti (Arachnida). Southwestern Naturalist 15:392-394. Mora, G. 1990. Paternal care in a Neotropical har- vestman, Zygopachylus albomarginis (Arachni- da, Opiliones: Gonyleptidae). Animal Behaviour 39:582-593. Pinto-da-Rocha, R. 1993. Invertebrados cavemi- colas da porgao meridional da Provmcia Espe- leologica do Vale do Ribeira, sul do Brasil. Re- vista Brasiliera de Zoologia 10:229-255. Ramires, E.N. & A. A. Giaretta. 1994. Maternal care in a Neotropical harvestman, Acutisoma proximum (Opiliones, Gonyleptidae). Journal of Arachnology 22:179-180. Rodriguez, C.A. & S. Guerrero. 1976. La historia natural y el comportamiento de Zygopachylus al- bomarginis (Chamberlin) (Arachnida: Opiliones: Gonyleptidae). Biotropica 8:242-247. Manuscript received 20 March 2000, revised 30 November 2000. 2001. The Journal of Arachnology 29:253-262 COMPARISON OF THE SURVIVAL OF THREE SPECIES OF SAC SPIDERS ON NATURAL AND ARTIEICIAL DIETS Divine M. Amelin: TREC-IFAS, University of Florida, 18905 SW 280 St., Homestead, Florida 33031 USA Jorge E. Pefie: TRFC-IFAS, University of Florida, 18905 SW 280 St., Homestead, Florida 33031 USA Jonethen Reiskind: Department of Zoology, University of Florida, Gainesville, Florida 32601 USA Robert McSorley: Department of Entomology and Nematology, University of Florida, Gainesville, Florida 32601 USA ABSTRACT. Three species of sac spiders were reared under laboratory conditions to investigate their survival and development. First, the effects of three artificial diets, milk + egg yolk, soybean liquid, and a combination of them, on the survival and development of Hibana velox were evaluated. Results over a 10 wk rearing period showed that the percentages of survival of H. velox reared on soybean liquid and combination diets did not differ significantly. However, the survival of H. velox on the milk + egg yolk diet was significantly lower than on the other two artificial diets. More molts and instars occurred in spiders raised on milk + egg yolk and on the combination diet than on the soybean liquid diet. Second, the development and percent survival of three sac spiders (Chiracanthium inclusum, H. velox, and Trach- elas volutus) on artificial diet (i.e., the combination diet) and natural diets (i.e., citrus leafminer larvae and Drosophila adults) were compared. The three sac spiders developed into the adult stage on the combination diet. Similarly, all three sac spiders reared on Drosophila adults were able to develop to the adult stage. Chiracanthium inclusum and T. volutus reared on citrus leafminer larvae developed to the adult stage, whereas H. velox did not. Females of these three species that matured using combination diet and were fertilized in captivity produced 1-3 egg masses. Oviposition took place 2-7 days after mating. Chiracanthium inclusum had an average of 57 eggs per egg mass, whereas H. velox and T. volutus had an average of 110 and 56 eggs per egg mass, respectively. Keywords: Laboratory rearing, sac spiders, Chiracanthium inclusum, Hibana velox, Trachelas volutus, citrus leafminer The diversity of spiders in almost all agroe- cosystems suggests their importance as pred- ators of insect and other arthropod pests (Whitcomb et al. 1963; Yeargan & Dondale 1974; Carroll 1980; Mansour et al. 1982; Mansour & Whitcomb 1986; Riechert & Bish- op 1990; Barrion & Litsinger 1995). Baseline information on life history and biology is fun- damental for ecological work and also impor- tant to further investigate the potential of spi- ders as biological control agents. However, life history studies have been done on very few species of spiders. One reason is the lack of reliable rearing methods to determine life histories and other biological data directly from laboratory cultures (Peck & Whitcomb 1968; Whitcomb 1967). Another reason is the lack of appropriate artihcial diets. Since spi- ders are primarily carnivorous, they require behavioral cues from the prey to initiate attack and feeding. This makes the rearing and main- tenance of spiders in the laboratory very la- borious. Moreover, it appears that most spi- ders must feed on a variety of insect prey species to obtain the optimum nutrition for survival and reproduction (Greenstone 1979; Uetz et al. 1992). The need to rear different insect prey species makes it especially diffi- cult to culture spiders in the laboratory. For- mulation of artificial diets would greatly fa- cilitate laboratory rearing of spiders; however, knowledge of the nutritional requirements for spiders is necessary. It was reported that some species of wan- 253 254 THE JOURNAL OF ARACHNOLOGY dering spiders are facultative nectar feeders (Taylor & Foster 1996). This finding inspired us to compare the survival of spiders on dif- ferent artificial diets. Preliminary results of our previous study on the survival of the sac spider Hibana velox (Becker 1879) showed that spiders reared on soybean diet had a high- er survival rate but slower rate of develop- ment than spiders reared on milk + egg yolk or on sugar solution alone (Amalin et al, 1999). This present study is a follow up of our previous experiment on the survival of H. velox raised on different artificial diets. We investigated the effect of different artificial di- ets including the previously tested (milk + egg yolk and soybean liquid) diets and a new diet (combination diet) on the survival and de- velopment of H. velox. The nutritional com- position of each diet is evaluated in relation to the survival and development of the spider. Also, the various degrees of survival of the three species of sac spiders, Chirac anthium inclusum (Hentz 1847), H. velox, and Trach- elas volutus Gertsch 1935 are compared when they were reared on artificial and natural diets. These sac spiders actively hunt their prey at night and during the day they hide in tubular silken capsules that they construct (hence the name “sac spiders”). For this study, the three species of sac spiders mentioned above were selected as the test organisms because they were found associated with citrus leafminer, which is one of the major insect pests of lime in south Florida (Amalin 1999). METHODS Comparison of artificial diets. — The three different artificial diets evaluated with respect to the survival of H. velox were soybean liq- uid (non-dairy beverage, Edensoy® Eden Eoods Inc.), milk + egg yolk mixture (100 ml homogenized whole milk + 1 fresh chicken egg yolk), and their combination(l:l soybean and milk + egg yolk mixture). One ml of green food color (McCormick®) was added to each diet to serve as an indicator of whether the spiders fed on the liquid diet since the col- or of their abdomen changes depending on the color of the food they consumed. The nutri- tional composition for each diet is shown in Table 1. A single first instar spiderling of H. velox was placed in a glass vial (15 mm di- ameter X 60 mm long) (Eig. lA). The mouth of the vial was closed with a cotton swab sat- Figure 1. — Set-up for rearing spiders. (A). On the combination diet. (B). On adults of Drosophila. (C). On larvae of the citrus leafminer. urated with the liquid diet. A one-inch long stick was impaled in the swab with the end pointing to the interior of the vial. The spider perched on the stick as it fed on the diet. The diet was replaced with fresh ingredients on a cotton swab every two days. The treatments for each artificial diet were replicated three times with 20 spiderlings per replication. All vials were kept in an incubator at 27 °C, 80% RH and a L:D 12:12 photoperiod. Spider mor- tality and molting were recorded every two days for 10 wk. The rate of development and growth in the different artificial diets was compared using one-way analysis of variance (ANOVA) (SAS institute 1989). Nature/sources of artificial and natural AMALIN ET AL.— SAC SPIDERS ON NATURAL AND ARTIFICIAL DIETS 255 Table 1. — Nutritional composition of the different diets based on the manufacturers’ nutritional analyses per 100 ml. Nutrient composition Milk + egg yolk Soybean Combination diet Total fat 3.04 g 1.30 g 4.34 g Saturated fat 1.30 g 0.0 g 1.3 g Cholesterol 97.83 mg 0.0 mg 97.83 mg Sodium 82.61 mg 39.13 mg 121.74 mg Total carbohydrates 6.09 g 10.87 g 17.04 g Sugars 5.22 g 6.52 g 12.99 g Protein 6.04 g 2.61 g 8.65 g Potassium 0.0 mg 126.09 mg 126.09 mg Vitamin A 348.00 lU 0.0 lU 348 lU Thiamin (Bl) 0.0 mg 0.05 mg 0.05 mg Riboflavin (B2) 0.0 mg 0.03 mg 0.03 mg Niacin (B3) 0.0 mg 0.52 mg 0.52 mg Pantothenic acid (B5) 0.0 mg 0.35 mg 0.35 mg Pyridoxine hydrochloride (B6) 0.0 mg 0.05 mg 0.05 mg Folate (B9) 0.0 mg 0.02 mg 0.02 mg Vitamin C 0.52 mg 0.0 mg 0.52 mg Vitamin D 43.48 lU 0.0 lU 43.48 lU Biotin (Vitamin H) 0.0 0.0003 g 0.003 g Calcium 0.14 g 0.03 g 0T7 g Iron 0.31 mg 0.31 mg 0.62 mg Phosphorus 0.11 g 0.04 g 0.15 g Magnesium 0.0 mg 17.4 mg 17.4 mg Zinc 0.0 mg 0.26 mg 0.26 mg diets. — The composition of the artificial diet for rearing of spider colonies was similar to the combination diet (Table 1) to which 5 ml honey was added. The natural diets consisted of either adults of the fruit fly, Drosophila me- lanogaster Meigen 1830, or larvae of the cit- Table 2. — Percentages of survival of Hibana ve- lox during 10 weeks on three different artificial di- ets under laboratory conditions at 27° C and 80% RH. Means followed by the same letters in the same row are not significantly different {P < 0,05) ac- cording to Duncan’s Multiple Range Test. Week Artificial diets Milk + yolk Soybean Combination 1 86.67 a 95.53 a 95.57 a 2 86.63 a 95.53 a 95.57 a 3 68.87 b 93.30 a 95.57 a 4 60.00 b 86.70 a 88.87 a 5 46.67 b 77.77 a 86.67 a 6 39.97 b 73.33 a 66.67 a 7 26.67 be 53.33 ab 57.77 a 8 22.33 b 51.10 a 51.10 a 9 17.80 b 42.20 a 42.23 a 10 17.80 b 42.20 a 42.23 a rus leafminer, Phyllocnistis citrella Stainton 1856. The fruit flies were mass reared in the laboratory at 25-27 °C and 70-80% RH. The initial population of fruit flies was obtained by exposing over-ripe bananas in a glass jar. The Figure 2. — Mean weekly survival of Hibana ve- lox reared on different artificial diets. Bars with the same letters are not significantly different according to Duncan’s Multiple Range Test {P ^ 0.05). 256 THE JOURNAL OF ARACHNOLOGY Number of weeks Figure 3. — Weekly percentages of survival and molting of Hibana velox reared on soybean liquid diet. fruit fly adults trapped in the jar were reared using the banana medium described by Yoon (1985), with some modifications. Approxi- mately 10 ml of the medium were poured from a sterilized beaker into each sterilized glass vial (15 mm diameter X 60 mm long). After the medium cooled, a sterilized strip of filter paper (1 cm X 5 cm) was inserted into each vial. The mouth of the vial was plugged with a sterilized cotton ball. The vials were stored for a day at 20 °C before use or stored at 4 °C until needed. Adult fruit flies from the initial population were immobilized by plac- ing them in a freezer (~0 °C) for 30-40 s and transferred to an empty glass jar. Ether was used for a longer period of immobilization. Next, five males and five females were trans- ferred to each glass vial containing banana medium. The fruit fly cultures were kept in an incubator at 27 °C and 80% RH. Adults from the succeeding cultures were mass reared. To avoid inbreeding, adult fruit flies from differ- ent stock cultures were mixed for re-culturing. A 1 wk-old fruit fly culture was utilized for spider rearing. This provided a continous sup- ply for the spider in the rearing cage for 3-4 wk. Citrus leafminer larvae were collected from an unsprayed lime orchard or from a green- house culture on lime shoots. Test spiders. — Egg sacs of C. inclusum, H. velox and T. volutus were collected in the field, brought to the laboratory and maintained in an incubator at 27 °C and 80% RH until spiderling emergence. Egg sacs were identi- fied based on the description by Amalin et al. (1999). First instar spiderlings were used in the experiment. Voucher specimens are de- posited at the Tropical Research and Educa- tion Center (TREC), Entomology Division at Homestead, Florida. Experimental protocol. — Three different containers were used for artificial and natural rearing. These rearing containers were used based on the preliminary trial we conducted on different rearing containers for each diet. They were selected for ease of handling of the spiders with minimal injury and disturbance, and also for speed in replacing the diet. The cage size might have an effect on the feeding behavior of the spiders, but this was not con- sidered in this study. The spiders fed with ar- tificial diet were placed in glass vials (15 mm diameter X 60 mm long) (Fig. lA) as ex- plained in the experiment on the comparison of artificial diets. To rear the spiders on D. AMALIN ET AL.— SAC SPIDERS ON NATURAL AND ARTIFICIAL DIETS 257 100 80 60 40 I 20 0 Number of weeks Figure 4. — Weekly percentages of survival and molting of Hibana velox reared on milk + yolk diet. melanogaster, newly emerged spiderlings were placed individually in a translucent plas- tic box (70 cm high X 70 cm long X 20 cm wide) (Fig. IB) with two circular openings (2 cm diameter) in the opposite sides. A 30 ml vial containing 10 ml of water was plugged with cotton and inserted into the one opening of the box. A glass vial containing a 1 wk-old culture of D. melanogaster adults on banana medium was introduced through the second circular opening. The spiders fed with 10 sec- ond instar P. citrella larvae were placed with these prey inside a plastic petri dish (10 cm in diameter X 1 cm high) lined with moist- ened filter paper (Fig. 1C). Thirty spiders of each species were reared from egg to maturity on the different diets. All spiders were maintained in an incubator at 27 °C and 80% RH with a L:D 12:12 photope- riod. The artificial diet and P. citrella larvae were replaced every 2 days, whereas D. me- lanogaster cultures were checked every 2 wk and replaced as needed. The survival and the developmental rates of the three sac spiders reared on artificial and natural diets were com- pared using Duncan Multiple Range Test (SAS Institute 1989). RESULTS Comparison of artificial diets. — Spiders raised on milk + egg yolk diet had a signifi- cantly lower weekly percentage survival than those reared on soybean liquid diet and the combination diet (Fig. 2). Table 2 shows the percentage spider survival weekly for a period of 10 wk. During wk 1 and 2, percentages of spider survival did not differ significantly among the three artificial diets. From wk 3 to wk 10, the percentages of spider survival were significantly higher for soybean liquid and combination diets than the milk + egg yolk mixture diet. In wk 7, percentages of spider survival on soybean liquid did not differ sig- nificantly from that of milk + egg yolk and combination diets. However, percentage sur- vival was significantly higher on combination diet than on milk + egg yolk diet. The developmental growth of H. velox reared on the different artificial diets was re- corded based on weekly survival and per- centages of molting (Figs. 3-5). The growth of spiders differed on the various artificial di- ets. Only 2% of the spiders raised on soybean liquid diet underwent one single molt during 258 THE JOURNAL OF ARACHNOLOGY Figure 5. — Weekly percentages of survival and molting of Hibana velox reared on combination diet. the second and third weeks of rearing (Fig. 3). The peak of molting was observed during week 9. Percentage survival of H. velox on soybean liquid diet was relatively high. More than 90% survived from week 1 to week 3 and 80% from from week 4 to week 6. During weeks 7 to 9, percentage survival decreased to less than 50% and by week 10 all of the spiders raised on this diet died (Fig. 3). Spiders reared on milk + egg yolk mixture and combination diets underwent more molts than those reared on soybean liquid diet (Figs. 4, 5). Spiders raised on milk T egg yolk diet molted three times (Fig. 4). A mean of 34.4% of the total spiders molted in week 2 and the percent that molted increased as the rearing progressed. The frequency of first molts peak- ed in week 4 and remained level. Eighty per- cent of the surviving spiders molted at least once. Almost 50% of the surviving spiders underwent a second molt between weeks 4 and 10, third molts started on week 6 until week 10. Percentage survival drastically de- creased from week 1 to week 10, from 85% to 9%. Spiders raised on the combination diet underwent as many as five molts (Fig. 5). First molts started during week 2, with a mean of almost 40%, and reached 93.3% by week 4. Second molts started during week 3, which was a week earlier than on milk + egg yolk diet; however, the percentage of second molts rose during week 4 and peaked in week 6. Third molts peaked during week 10. However, very few spiders reared on combination diet underwent the fourth and fifth molts. The trend of the percentage survival of spiders reared on the combination diet was similar to that on the soybean liquid diet except on week 10 in which almost 40% were still surviving. Comparison of artificial and natural di- ets*— The developmental stages and percent- ages of survival for each developmental stages of the three sac spider species reared on arti- ficial and natural diets are shown in Figs. 6- 8. On the combination diet, all three species were able to develop into the adult stage (Fig. 6), after the sixth and seventh molts. In gen- eral, percentages of survival of H. velox were relatively higher than those of C. inclusum and T. volutus in all developmental stages on AMALIN ET AL.— SAC SPIDERS ON NATURAL AND ARTIFICIAL DIETS 259 40 nj 30 tn +1 c m & £ 20 li > £ □ m ^ 10 0 Figure 6. — Survival of developmental stages of Chiracanthium inclusum, Hibana velox, and Trachelas volutus reared on combination diet. I I Hibana veiox ■■ Chiracanthium inciusum ^Z77\ Tracheias voiutus 2 3 1 m 4 5 6 7 Developmental stages male female the combination diet. Moreover, percentage survival to adulthood was higher for H. velox than for C. inclusum and T. volutus. From the surviving spiders in the sixth and seventh molt stages, 31% and 32% of H. velox developed into male and female adults, respectively. Sur- vival was less than 10% for C. inclusum and T. volutus. Females that matured on combi- nation diet and were fertilized in captivity pro- duced one to three egg masses. Oviposition took place 2-1 days after mating for all three species. The number of eggs laid by female H. velox ranges from 96-120 with an average of 110. The number of eggs per egg mass of C inclusum reared in the laboratory varied from 36-86 with an average of 57. Edwards (1958) reported 112 eggs in a single egg mass and Peck & Whitcomb (1970) reported a range of 17-86 eggs per egg mass. Trachelas volutus produced 47-66 per egg mass with an average of 56. The three species of sac spiders reared on Drosophila were able to develop into the adult stage (Fig. 7), In general, percentage survival of T. volutus was relatively higher than those of H. velox and C inclusum for the immature stages. However, C. inclusum and H. velox had higher percentages survival to the adult stage. Less than 10% from the seventh and eighth molt stages survived into male and fe- male adults for T. volutus. For C, inclusum, 30% and 38% from the seventh and eighth molt stages developed into male and female adults, respectively. For H. velox, 17% devel- oped into male and 12% into female adults. In general, percentage survival of T. volutus was relatively higher than those of H. velox and C. inclusum when reared on P. citrella. Trachelas volutus and C inclusum success- fully developed into the adult stage but H. ve- lox did not (Fig. 8). This finding suggests that P. citrella is deficient in one or more nutrients required by H. velox. However, the consump- tion of P. citrella by immature H. velox was relatively higher than by C. inclusum and T. volutus. 260 THE JOURNAL OF ARACHNOLOGY 40 UJ +1 c fO 0) E 15 > t 3 CO 30 20 I I Hibana veiox WKM Chiracanthium inclusum X//A Trachelas volutus l / it] male female Developmental stages Figure 7. — Survival of developmental stages of Chiracanthium inclusum, Hibana velox, and Trachelas volutus reared on adults of Drosophila. DISCUSSION Each single diet has an important nutrient for the growth and survival of spiders under laboratory conditions. For instance, on soy- bean liquid and combination diets the follow- ing nutrients are available in relatively higher amounts: carbohydrates, sugar, potassium, magnesium, and zinc. Among these nutrients, carbohydrates are known to be the major en- ergy source important for survival or longev- ity of any arthropod species, whereas the min- erals potassium, magnesium, and zinc are necessary for optimal growth (Singh 1984). Vitamin B-complex is also required in artifi- cial diets and is available in soybean^ liquid but absent in milk + egg yolk diet. This find- ing suggests that if enough carbohydrates, and possibly the other nutrients mentioned above are available, mortality at an early stage of spider development will be avoided. However, results from our previous experiment (Amalin et al. 1999) revealed that the development of spiders on soybean liquid was delayed but progressed normally on milk + egg yolk diet. The main nutrient that is available in milk + egg yolk diet, which is absent in soybean liq- uid, is cholesterol. Cholesterol is a common sterol and a precursor of ecdysone, the molt- ing hormone (Foelix 1982; Singh 1984). This may explain the delayed development of spi- ders on soybean diet. Other nutrients available in relatively higher amounts in milk + egg yolk diet are total fat, saturated fat, sodium, protein. Vitamin A, Vitamin D, calcium, and phosphorous. These nutrients may also con- tribute to the complete development of spiders on milk + egg yolk diet. According to House (1961) an artificial diet should contain a bal- ance of proteins, carbohydrates, lipids, and vi- tamins for normal growth, development, re- production and other life processes. All of these important nutrients are available in the combination diet with more concentrated val- ues (Table 1), which probably explains the higher percent survival and normal develop- ment of spiders on the combination diet than on the soybean and milk + egg yolk diets. The completion of development of the three AMALIN ET AL.— SAC SPIDERS ON NATURAL AND ARTIFICIAL DIETS 261 Developmental stages Figure 8. — Survival of developmental stages of Chiracanthium inclusum, Hibana velox, and Trachelas volutus reared on larvae of citrus leafminer. species of sac spiders on the combination diet suggests that they are also nectar feeders as reported by Taylor & Foster (1996). This fur- ther suggests that the combination diet pro- vided more complete nutrional or dietary re- quirements for sac spiders. Nevertheless, different species of spiders even under the same guild or group could have different re- quirements of the proportion of all the nutri- ents pertinent to survival and development. This might be one reason why there was a higher percentage survival of H. velox fed with combination diet than C. inclusum and T. volutus. Thus, we recommend that different proportions of the nutrients of the combina- tion diet should be tried to determine the best proportion of each one for the survival and development of these three species of sac spi- ders. Behavioral and ecological differences among these three species of sac spiders should not be ruled out as possible reasons for differences in percentage survival; however, any such differences were not observed in this particular study. Of the natural diets tested. Drosophila provided a suitable diet for the three species of spiders particularly for H. ve- lox and C. inclusum; whereas, citrus leafminer seems to be less suitable as diet of the spiders under laboratory conditions. Results from this experiment reveal that an artificial diet is adequate for these spiders un- der laboratory conditions. Attempts towards the mass rearing of these spiders using artifi- cial diets should be pursued. Clearly the pro- portions of the various ingredients in combi- nation diet must be evaluated to optimize spider survival and reproduction. Advance- ments in mass-rearing spiders on artificial di- ets may enable their use in agriculture for aug- mentation of field populations. ACKNOWLEDGMENTS We thank Drs. Waldemar Klassen and Nor- man Leppla for review of the manuscript. We also thank Michelle Codallo, Jose Alegria, and Ivan Toledo for their help in rearing the spiders. Florida Agricultural Experiment Sta- tion Journal Series No. R-07222. 262 THE JOURNAL OF ARACHNOLOGY LITERATURE CITED Amalin, D.M. 1999. Evaluation of Predatory Spi- ders as Biological Control Agents of Citrus Leaf- miner, Phyllocnistis citrella, in Lime Orchards at Homestead, Florida. Ph.D. dissertation, Univer- sity of Florida, Gainesville. Amalin, D.M., J. Reiskind, R. McSorley & J. Pena. 1999, Survival of the hunting spider, Hibana re- lax (Arneae, Anyphaenidae), raised on different artificial diets. Journal of Arachnology 27(2): 692-696. Barrion, A.T & J.A. Litsinger. 1995. Riceland Spi- ders of South and Southeast Asia. Centre for Ag- riculture and Biosciences International, Walling- ford, Oxon, England. 700 pp. Carroll, D.P. 1980. Biological notes on the spiders of some citrus groves in central and southern California. Entomological News 91:147-154. Edwards, R.J. 1958. The spider subfamily Clu- bioninae of the United States, Canada, and Alas- ka (Araneae: Clubionidae). Bulletin of the Mu- seum of Comparative Zoology 118:365-434. Foelix, R. 1982. Biology of Spiders. Harvard Univ. Press, Cambridge, Massachusetts. 306 pp. Greenstone, M.H. 1979. Spider feeding behaviour optimises dietary essential amino acid composi- tion. Nature 282:501-503. House, H.L. 1961. Insect nutrition. Annual Review of Entomology 6:13-26. Mansour, F. & W.H. Whitcomb. 1986. The spiders of a citrus grove in Israel and their role as bio- control agents of Ceroplastes floridensis (Ho- moptera: Coccidae). Entomophaga, 31:269-276. Mansour, F, J.W. Ross, G.B. Edwards, W.H. Whit- comb & D.B. Richman. 1982. Spiders of Florida citrus groves. Florida Entomologist 65:514-522. Peck, W.B. & W.H, Whitcomb. 1968. Feeding spi- ders on artificial diet. Entomological News 79: 233-236. Peck, W.B. & W.H. Whitcomb. 1970. Studies on the biology of a spider, Chiracanthium inclusum (Hentz). Florida Agricultural Extension Station Bulletin 753:1-76. Riechert, S.E. & L. Bishop. 1990. Prey control by an assemblage of generalist predators in a garden test system. Ecology 71:726-736. SAS Institute. 1989, SAS/STAT User’s Guide, ver- sion 6, 3rd ed., Vol. 2. SAS Institute Inc., Cary, North Carolina, 846 pp. Singh, P. 1984. Insect diets. Historical develop- ments, recent advances, and future prospects. Pp. 32-44. In Advances and Challenges in Insect Rearing. (E.G. King & N.C. Leppla, eds.). Ag- ricultural Research Service, Southern Region, U.S. Department of Agriculture, New Orleans, 306 pp, Taylor, R.M. & WA. Foster. 1996. Spider nectar- ivory. American Entomologist 82-86, Uetz, G.W, J. Bischoff & J. Rovner. 1992. Sur- vivorship of wolf spiders (Lycosidae) reared on different diets. Journal of Arachnology 20:207- 221. Whitcomb, WH. 1967. Wolf and lynx spider life histories. In Terminal Report to National Science Foundation. Univ. of Arkansas. Div. of Agricul- ture, Dept, of Entomology. Fayetteville, Arkan- sas. 142 pp, Whitcomb, W.H., H. Exline & R.C. Hunter. 1963. Spiders of the Arkansas cotton field. Annals of the Entomological Society of America 56:653- 660. Yeargan, K.V. & C.D. Dondale. 1974. The spider fauna of alfalfa fields in northern California. An- nals of the Entomological Society of America 67:681-682. Yoon, J.S. 1985. Drosophilidae: Drosophila me- lanogaster. Pp.75-91. In Handbook of Insect Rearing, Vol. II (P. Singh & R.R Moore, eds.). Elsevier Science Publishing Company, New York, 514 pp. Manuscript received 20 May 2000, revised 1 De- cember 2000. 2001. The Journal of Arachnology 29:263-266 SHORT COMMUNICATION POST-MATURATION MOLT FOUND IN A WOLF SPIDER, PARDOSA ASTRIGERA (ARANEAE, LYCOSIDAE) Yasuhiro Fujiii Department of Biology, The Nippon Dental University, U9-20 Fujimi, Chiyoda-ku, Tokyo 102=8159, Japan ABSTRACT. An adult female Pardosa astrigera (Araneae, Lycosidae) died failing to finish an addi- tional molt in the laboratory. Its maturity was morphologically ascertained by SEM examination. Keywords: Lycosidae, Pardosa astrigera, post-maturation molt Post-maturation molt is well known in fe- males of the primitive spiders (Liphistiomor- phae and Mygalomorphae) which continue to grow for several years after sexual maturation (B^aerg & Peck 1970; Main 1976; Stradling 1978; Stewart & Martin 1982; Yoshikura 1987; Maki 1989; Miyashita 1992). Post-maturation molt, however, is very rare among entelegynes and has been reported only for six females in three species: three in Latrodectus mactans (Fabricius 1775) (Theridiidae) (Kastoe 1968), two in L. hesperus Chamberlin & Ivie 1935 (Kaston 1968), and one in Heteropoda vena- toria (L. 1758) (Sparassidae) (Kayashima 1981). This report documents one more entel- egyne post-maturation molt found in a female wolf spider (Lycosidae), Pardosa astrigera L. Koch 1878. This female was captured on 11 June 1983 at Hidaka, in northwestern Kanto Plain, central Japan. It molted to maturity on 30 June, and died on 3 August of that year, after failing to extract its extremities during the additional molt. It was then preserved in 70% ethanol. Pardosa astrigera is common on the sunny ground with sparse vegetation (Fujii 1998) and is found also in Korea and China (Tanaka 1993). Its ecophysiological character- istics were heavily studied (Miyashita 1968a, 1968b, 1969a, 1969b, 1998; Fujii 1974, 1978, 1980; Tanaka & ltd 1982), though this spider had been identified with a closely related spe- cies, Pardosa (Lycosa) T-insignita until the ex- amination by Tanaka (1980). Sclerification and lengthening of epigyea in females of Schizocosa ocreata (Hentz 1844) (Lycosidae) begin by the third or fourth iestar prior to maturation (Amaya & KlawinsM 1996). Thus, one may sometimes confuse an immature female with a mature one when ob- serving it with a magnifying glass or the na- ked eye. The females reported by Kaston (1968) and Kayashima (1981) had undoubt- edly matured because they copulated and laid fertile eggs before the additional molt. On the other hand, this P. astrigera female refused the courtship of a male and killed it. This fe- male left no evidence of egg-cocoon construc- tion, which sometimes occurs even in virgin females. Its maturity was ascertained by the morphological observations described below. In many iycosid species, knob-tipped hairs (knobbed hairs) peculiar to adult female ab- domens were found (Graefe 1964; Rovner et al. 1973). Also in P. astrigera, I found the hairs on adult females (Fujii 1983), but not on subadult females nor on males. If the P. as- trigera female actually had matured before the final molt, a well-developed epigynum (with genital openings) and knobbed hairs should be found on the old (molted) exuvium of its ab- domen. The specimen was observed with a digital optical microscope (Keyence VH-Z05) (Fig. 1), thee was examined with a scanning electron microscope (Hitachi S-4000) after critical point drying and ion-beam sputter coating with Pt-Pd (Figs. 2-7). A well-devel- oped epigynum was seen in the area of old exuvium (Figs. 1-3), and its external features coincided with those of the standard epigyn- um illustrated in Tanaka (1980, 1993). Many knobbed hairs were also detected on both old and new exuviae (Figs. 6, 7). From these re- 263 264 THE JOURNAL OF ARACHNOLOGY Figures 1-7. — A female of Pardosa astrigera that died at an additional molt after maturation. 1. The female in 70% ethanol before treatments for electron microscopic observation; 2-7. Scanning electron micrographs of the ventral side of the abdomen. 2. The whole abdomen; 3. The epigynum on the old exuvium; 4, 5. The old and new exuviae in the mid-dexter portion (5 corresponds to 4a); 6, 7. Knobbed hairs both on the old and new exuviae (6 and 7 correspond to 5b and 5c, respectively). Abbreviations: Ca = carapace, Ep = epigynum, KH = knobbed hairs, NEx = new exuvium, OEx = old exuvium. Scale bars: Figures 1-4 = 1 mm. Figures 5-7 = 0.1 mm. FUJII— POST-MATURATION MOLT IN A WOLF SPIDER 265 suits it can be said that post-maturation molt occurred in this lycosid. Renewal of the epi- gynum at this molt could not be seen in this specimen as well as in the Kaston’s females. This specimen was deposited as the voucher in the collection of the Department of Zool- ogy, National Science Museum, Tokyo (NSMT~Ar 4321). If the post-maturation molt of entelegynes were part of a reproductive strategy, it would be expected to occur only in extremely old or small females. But the female of P. astrigera molted only 34 days after maturation, while females of this species usually live for a lon- ger period (143 days is the longest known). Moreover, its carapace width reached to 3.2 mm at maturity. This size is not small com- pared to the range of 2.2-3. 5 mm in adult fe- males of P. astrigera collected in the field (Fujii unpubl. data). This additional molt could not be found in the other 368 females of 18 lycosid species (50 females of P. astri- gera), which had matured in the field (216 females) or the laboratory (152 females) from 1981-1987 and were reared until the death to examine their life cycles. This molt may be an accidental phenomenon occurring at very low frequency (0.27% for the total lycosids) and seems to occur also in natural lycosid populations. ACKNOWLEDGMENTS I wish to thank Dr. Tamotsu Nagumo for operating the scanning electron microscope, and Dr. Koichi Tanaka for critically reading the manuscript. My thanks also go to Dr. Hoz- umi Tanaka and Dr. Kazuyoshi Miyashita for giving helpful information, Dr. Hirotsugu Ono for depositing the voucher specimen, and to Dr. Sadashi Komiya for continual encourage- ment. LITERATURE CITED Amaya, C.C. & P.D. Klawinski. 1996. A method for assessing gender in immature wolf spiders (Araneae, Lycosidae). Journal of Arachnology 24:158-160. Baerg, W.J. & W.B. Peck. 1970. A note on the lon- gevity and molt cycle of two tropical theraphos- ids. Bulletin of the British Arachnological Soci- ety 1:107-108. Fujii, Y. 1974. Hunting behaviour of the wolf spi- der, Pardosa T-insignita (Boes. et Str.). Bulletin of the Nippon Dental College General Education 3:135-148. Fujii, Y. 1978. Examinations of the maternal care of cocoon in Pardosa astrigera L. Koch (Ara- neae, Lycosidae). Bulletin of the Nippon Dental University General Education 7:221-230. Fujii, Y 1980. Analytical study of maternal behav- iour in Pardosa astrigera L. Koch (Araneae, Ly- cosidae). Bulletin of the Nippon Dental Univer- sity General Education 9:233-245. Fujii, Y. 1983. Four Arctosa lycosids lacking the abdominal knobbed hairs and their pulli’s post- emergent behaviour (Araneae, Lycosidae). Bul- letin of the Nippon Dental University General Education 12:177-188. Fujii, Y. 1998. Ecological studies on wolf spiders (Araneae: Lycosidae) in a northwest area of Kan- to Plain, central Japan: Habitat preference ob- served by hand-sorting. Acta Arachnologica 47: 7-19. Graefe, G. 1964. Die Bmtfiirsorge bei Pardosa lu- gubris (Walckenaer 1802) (Araneae, Lycosidae). Ph.D. thesis, Ludwig-Maximilians-Universitat, Munich. Kaston, B.J. 1968. Remarks on black widow spi- ders, with an account of some anomalies. Ento- mological News 79:113-124. Kayashima, 1. 1981. A report on long-term rearing of Heteropoda venatoria (Linne) (1). Kishidaia (47):57-64. (in Japanese) Main, B.Y 1976. Spiders. Collins, Sydney. Maki, T. 1989. Life history of the trapdoor spider Latouchia typica (Kishida). Atypus 94:18-25. (in Japanese) Miyashita, K. 1968a. Growth and development of Lycosa T-insignita Boes. et Str. (Araneae: Ly- cosidae) under different feeding conditions. Ap- plied Entomology and Zoology 3:81-88. Miyashita, K. 1968b. Quantitative feeding biology of Lycosa T-insignita Boes. et Str. (Araneae: Ly- cosidae). Bulletin of the National Institute of Ag- ricultural Sciences (Japan), Series C 22:329-344. Miyashita, K. 1969a. Seasonal changes of popula- tion density and some characteristics of overwin- tering nymph of Lycosa T-insignita Boes. et Str. (Araneae: Lycosidae). Applied Entomology and Zoology 4:1-8. Miyashita, K. 1969b. Effects of locomotory activ- ity, temperature and hunger on the respiratory rate of Lycosa T-insignita Boes. et Str. (Araneae: Lycosidae). Applied Entomology and Zoology 4: 105-113. Miyashita, K. 1992. Postembryonic development and life cycle of Atypus karschi Donitz (Araneae: Atypidae). Acta Arachnologica 41:177-186. Miyashita, K. 1998. Egg sac production and nymphal development of Pardosa agraria Ta- naka and Pardosa astrigera L. Koch under a seminatural condition. Journal of the Natural History Museum and Institute, Chiba 5:41-45. (in Japanese) 266 THE JOURNAL OF ARACHNOLOGY Rovner, J.S., G.A. Higashi & R.F. Foelix. 1973. Ma- ternal behavior in wolf spiders: The role of ab- dominal hairs. Science 182:1153-1155. Stewart, D.M. & A.W, Martin. 1982. Moulting in the Tarantula, Dugesiella hentzi. Journal of Com- parative Physiology 149:121-136. Stradling, D.J. 1978. The growth and maturation of the “tarantula,” Avicularia avicularia L. Zoolog- ical Journal of Linnean Society 62:291-303. Tanaka, H. 1980. Notes on four type- specimens of Japanese wolf spiders of the Museum fur Natur- kunde, Humbolt Universitat, Berlin. Acta Arach- nologica 29:47-55. Tanaka, H. 1993, Lycosid spiders of Japan XL The genus Pardosa C.L. Koch - paludicola-gmwp. Acta Arachnologica 42:159-171. Tanaka, K. & Y. Ito. 1982. Decrease in respiratory rate in a wolf spider Pardosa astrigera (L. Koch), under starvation. Researches on Popula- tion Ecology 24:360-374. Yoshikura, M. 1987. The Biology of Spiders. Japan Scientific Societies Press, Tokyo, (in Japanese) Manuscript received 27 February 2000, revised 22 December 2000. 2001. The Journal of Arachnology 29:267-269 SHORT COMMUNICATION DESCRIPTION OF THE EGG SAC OF MIMETUS NOTIUS (ARANEAE, MIMETIDAE) AND A CASE OF EGG PREDATION BY PHALACROTOPHORA EPEIRAE (DIPTERA, PHORIDAE) HANK GUARISCO: P.O. Box 3171, Lawrence, Kansas 66046 USA ABSTRACT. The eggsac of the pirate spider, Mimetm notius, is described and compared with eggs of other members of the genus. The phorid fly egg predator, Phalacrotophora epeirae, was reared from a M. notius eggsac. Keywords: Mimetus notius, egg sac, Phalacrotophora epeirae, predation Members of the family Mimetidae have jus- tifiably earned the comin.on appellations of “assassin” and “pirate” spiders because of their interesting feeding habits. Armed with a series of long, slightly curved spines on the promargieal areas of the tibiae and metatarsi of legs I and II (Kastoe 1981), they enter spi^ der webs, especially those of comb-footed spi- ders (Theridiidae) and orbweavers (Araeei- dae), and prey upon the occupants (Gertsch 1979). Jackson & Whitehouse (1986) ob- served mimetids using vibratory aggressive mimicry to lure spiders within striking range. Although these spiders are poorly repre- sented in systematic collections, a recent sur- vey of the genus Mimetus Hentz 1832 in Kan- sas has revealed the presence of four species (Guarisco & Mott 1990). Mimetus notius Chamberlin 1923 is the second most com- monly encountered member of the genus in northeastern Kansas. It has been taken in sweep samples of understory vegetation, mostly coralberry (Symphoricarpos orbicula- tus Moeech.), in woodland on the Fitch Nat- ural History Reservation (FNHR) and on the eaves and outer walls of Reservation Head- quarters (FNHR) in Douglas County, Kansas. The FNHR is a 590 acre (239 ha) tract of land which comprises all but the southwestern 50 acres (20 ha) of Section 4 (T12S, R20E) in Douglas County. It is located at the ecotone between the Eastern Deciduous Forest and Tallgrass Prairie Biomes, and ranges from 880—1080 feet (268—329 m) in elevation (39°00\ 95°ir) (Fitch & Kettle 1988). Two adult females were collected from eastern red cedar (Juniperus virginiana L.) in Montgom- ery and Greenwood counties, on 4 and 26 May 1991, respectively. The egg sac of M. notius is fluffy, translu- cent, spherical to subspherical, and is com- posed of a 1 mm thick outer layer of sparse, curly, brown silk strands surrounding a dense white central section containing the brown eggs. Completely surrounding the egg sac is a thin, subspherical to elliptical net of silk. The egg sac is suspended within this net by several thick silk strands which extend from the sac to the net (Fig. 1). The shape of the net appears to be determined by the amount of space available near the egg sac. Two sacs, each laid within the confines of a petri dish, were surrounded by elliptical silk nets, 30 X 23 and 50 X 20 mm in diameter. Each net was 15 mm in height, which equalled the height of the petri dish. An egg sac discovered on the underside of a wooden door leaning against the outer wall of a laboratory building on the FNHR had a silk net with the following dimensions: 30 X 30 X 7 mm. Several egg sacs located in the comers between the eaves and outer walls of buildings possessed nets with similar dimensions. A female collected on 4 May produced an egg sac 7X5 mm on 16 May containing 24 eggs. On 26 May, this individual produced a second egg sac (6X8 mm) containing an un- determined number of eggs. A second female obtained on 26 May produced a total of 5 egg sacs during the following four weeks. The first 267 268 THE JOURNAL OF ARACHNOLOGY two egg sacs were laid on 28 May and 6 June, were 6X6 and 5X6 mm in diameter, with surrounding nets 20 and 25 mm in diameter, and contained a total of 29 and 37 eggs, re- spectively. The last three sacs were laid on 13, 21 and 27 June and contained 21, 12, and 25 eggs, respectively. These sacs were laid in a small vial, and the silk nets surrounding them were 15 X 25 mm. The last two egg sacs were probably infertile because they became cov- ered with mold in a few days. Observing the structure of egg sacs pro- duced in the laboratory enabled me to recog- nize them in the field. One egg sac discovered on the FNHR on 18 June yielded 47 spider- lings ten days later. An egg sac collected from the eaves of a building in the Topeka Zoo on 21 June produced 35 spiderlings and one in- fertile egg on 4 July. Two egg sacs collected from the eaves of a residence in Lawrence, Douglas County, on 13 August contained 14 empty shells and 12 infertile eggs, and 25 empty shells and 7 infertile eggs, respectively. The average number of eggs per sac = 28.9, SD = 9.71 {n = 10). On 3 July, I collected a M. notius egg sac from the eaves of a house in Jefferson County, Kansas which contained several brown phorid fly pupae. Two days later, an adult fly, Phal- acrotophora epeirae (Brues 1902)(Diptera, Phoridae), emerged from the egg sac. The re- maining flies matured over the next few days. The structure of the egg sac of M. notius resembles those of other members of the ge- nus, except for the unique net of silk which surrounds it. Mimetus puritanus Chamberlin 1923 and M. hesperus Chamberlin 1923 both construct spherical, bright orange, loosely wo- ven egg sacs (Guarisco & Mott 1990; Icenogle 1972). Lawler (1972) described the egg sacs of both wild and captive M. eutypus Cham- berlin & Ivie 1935. They were loosely woven and those produced in captivity ranged from white to dark rust-brown, while the wild egg sacs were all pale yellow-white. The egg sacs of Old World mimetids are generally globular GUARISCO— EGG SAC OF MIMETUS 269 with many loops of silk on the surface and contain only 5-20 eggs (Heimer 1986). The tufted nature of the silk may protect the eggs from mechanical damage, predation, or para- sitism. Although it is tempting to speculate upon the probable functions of the unique structure of the egg sac of M. notius, further observations are needed to explore the rela- tionships of various aspects of egg sac design and relative survivability. The present observation on the emergence of Phalacrotophora epeirae from an egg sac of M. notius is the first recorded instance of egg predation or parasitism upon any member of the genus Mimetus. Phalacrotophora epei- rae is a well known larval egg predator which has been reared from spider egg sacs of the following species: Linyphiidae: Pityohyphan- tes costatus (Hentz 1850)(see Manuel 1984); Araneidae: Larinioides Caporiacco 1934 (= Nuctenea Simon 1864 — Epeira Walckenaer 1805) sp. (see Brues 1902, 1903), Larinioides sclopetarius (Clerck 1757) (see Auten 1925), Gasteracantha cancriformis (Linnaeus 1785) (see Muma & Stone 1971); and Salticidae: Phidippus audax (Hentz 1845) (see Jones 1940). ACKNOWLEDGMENTS I thank the University of Kansas Depart- ment of Entomology for providing laboratory space; Paul Liechti, Kansas Biological Sur- vey, for providing valuable equipment; Brian V. Brown of the Natural History Museum of Los Angeles County, California for identifi- cation of the phorid fly; and William Bell, University of Kansas for permission to collect specimens on his property. Eor critically re- viewing the manuscript, I thank Bruce Cutler of the University of Kansas and Daniel Mott of Dickinson State University. I thank Will Bouchard of the University of Kansas for pro- viding the art work. Voucher specimens were deposited in the Smithsonian Collection (USNM). This paper is dedicated to the mem- ory of the late William Bell. LITERATURE CITED Auten, M. 1925. Insects associated with spider nests. Annals of the Entomological Society of America 18:240-250. Brues, C.T. 1902. Notes on the larvae of some Tex- an Diptera. Psyche 9:351-354. Brues, C.T. 1903. A monograph of the North American Phoridae. Transactions of the Ameri- can Entomological Society 29:331-404 + 5 pi. Fitch, H.S. & W.D. Kettle. 1988. Kansas Ecolog- ical Reserves (University of Kansas Natural Ar- eas). Transactions of the Kansas Academy of Science 91(l-2):30-36. Gertsch, W.J. 1979. American Spiders (second edi- tion). Van Nostrand Reinhold, New York. 274 pp. Guarisco, H. & D.J. Mott. 1990. Status of the ge- nus Mimetus (Araneae: Mimetidae) in Kansas and a description of the egg sac of Mimetus pur- itaniis Chamberlin. Transactions of the Kansas Academy of Science 93(3-4):79-84. Heimer, S. 1986. Notes of the spider family Mi- metidae with description of a new genus from Australia (Arachnida, Araneae), Entomologische Abhandlungen des Staatlichen Museums fiir Tierkunde in Dresden 49(7): 1 13-137. Icenogle, W.R. 1972. Notes on the biology of the genus Mimetus. Notes on Arachnology of the Southwest 3:19. Jackson, R.R. & M.E.A. Whitehouse. 1986. The biology of New Zealand and Queensland pirate spiders (Araneae, Mimetidae): Aggressive mim- icry, araneophagy and prey specialization. Jour- nal of Zoology (London)(A) 210:279-303. Jones, S.E. 1940. An annotated list of the spiders of an East Central Illinois forest (Wm. Trelease Woods, University of Illinois). Transactions of the Illinois Academy of Science 33(2):2 16-220. Kaston, B.J. 1981. Spiders of Connecticut (revised ed.). State Geology and Natural History Survey of Connecticut, Bulletin. No. 70. 1020 pp. Lawler, N. 1972. Notes on the biology and behav- ior of Mimetus eutypus Chamberlin and Ivie (Ar- aneae: Mimetidae). Notes on Arachnology of the Southwest 3:7-10. Manuel, R.L. 1984. The egg sac of Pityohyphantes costatus (Hentz) (Araneae, Linyphiidae) and its phorid parasite. Journal of Arachnology 12:371- 372. .• Muma, M.H. & K.J. Stone. 1971. Predation of Gasteracantha cancriformis (Arachnidae [sic]: Araneidae) eggs in Florida citrus groves by Phal- acrotophora epeirae (Insecta: Phoridae) and Ar- achnophaga ferruginea (Insecta: Eupelmidae). Florida Entomologist 54(4):305--310. Manuscript received 11 April 1997, revised 9 March 2001. 2001. The Journal of Arachnology 29:270-272 SHORT COMMUNICATION REVIEW OF THE SOUTH AMERICAN SPECIES OF THE GENERA AULONIA AND ALLOCOSA (ARANEAE, LYCOSIDAE) Roberto M. Capocasale: Division Zoologia Experimental, Institute de Investigaciones Biologicas C. Estable, Ave. Italia 3318, 11600 Montevideo, Uruguay ABSTRACT. Aulonia bergi (Holmberg 1876) and Aulonia macrops Simon 1897 are considered nomina dubia. Agalenocosa luteonigra (Mello-Leitao 1945) new combination (= Aulonia luteonigra Mello-Leitao 1945) is illustrated. GliescMella senex (Mello-Leitao 1945) is illustrated and synonymized und&i Allocosa brasiliensis (Petrankevitch 1910). RESUMEN. Aulonia bergi (Holmberg 1876) y Aulonia macrops Simon 1897 se consideran nomina dubia. Agalenocosa luteonigra (Mello-Leitao 1945) nueva combinacion (= Aulonia luteonigra Mello- Leitao 1945) se ilustra. GliescMella senex (Mello-Leitao 1945) se illustra y sinonimiza bajo Allocosa brasiliensis (Petrankevitch 1910). Keywords: Araneae, Lycosidae, Aulonia, Allocosa, GliescMella Twenty-five species from South America were originally listed in the subfamily Hip- pasinae (Roewer 1954; Capocasale 1982, 1990). Originally I assigned 16 of these spe- cies (Capocasale 1990) to three different sub- families according to the definitions given by Dondale (1986) for each subfamily. However, there are four taxa for which the systematic position is unknown. They are: Allocosa senex (Mello-Leitao 1945) and the three species of Aulonia from South America. The purposes of this note are to convey the results of the study of those four taxa and to clarify their system- atic position. In this form the author com- pletes his review of the lycosoid subfamily Hippasinae from South America. Abbreviations used: MACN = Museo Ar- gentino de Ciencias Naturales, Argentina; MLP ” Museo de La Plata, Argentina; MHNP = Museum National d’Histoire Naturelle, Par- is; MNRJ — Museu Nacional de Rio de Ja- neiro, Brazil. Illustrations were made with the aid of a camera lucida. Aulonia bergi (Holmberg 1876) Lycosa {Aulonia) Bergii Holmberg 1876: 176. Aulonia gergi (sic): Bonnet 1955: 822. Aulonia bergi: Mello-Leitao 1944: 321; Roewer 1954: 234 Comments. — Mello-Leitao (1944) consid- ered A. bergi as a nomen nudum. Holmberg’s description is: . . /Tomada en Las Conchas, y muy semejante a Aulonia albimana K. pero de doble longitud.” Holmberg 's types are lost; consequently it is impossible to make any de- finitive identification. As I cannot identify this species I consider it as a nomen dubium. Aulonia macrops Simon 1897 Aulonia macrops Simon 1897: 329, 1898: 30; Roewer 1954: 234; Lehtinen & Hippa 1979: 21. Comments. — Lehtinen & Hippa (1979) considered this species as a Lycosinae and not congeneric with Aulonia albimana. I have ex- amined the female holotype from Rio de Ja- neiro, Brazil, deposited at MHNR It is im- mature. I agree with Lehtinen & Hippa that this specimen could be a Lycosinae, but I can- not identify it as Aulonia. I consider it as a nomen dubium. Agalenocosa luteonigra (Mello-Leitao 1945) new combination Figs. 1^7 Aulonia luteonigra Mello-Leitao 1945: 247; Roew- er 1954: 234. 270 CAPOC AS ALE— SOUTH AMERICAN AULONIA AND ALLOCOSA 271 Figures 1-10. Agalenocosa and Allocosa. 1-7. Agalenocosa luteonigra (Mello-Leitao). 1. Eyes, frontal view; 2. Epigynum; 3. Spermathecae; 4. Palpus of male, ventral view; 5. Retrolateral apophysis of tibia, ventral view; 6. Terminal apophysis, ventral view; 7. Median apophysis, ventral view. 8-10. Allocosa bra- siliensis (Petmnkevitch) (= Glieschiella senex Mello-Leitao, holotype MLP, Entre Rios, Colon, Argentina); 8. Palpus of male, ventral view; 9. Terminal apophysis, ventral view; 10. Median apophysis, ventral view. Comments. — Mello-Leitao (1945: 248) said the types are in MLP (N° 16480). How- ever this number today does not exist in the collection of this museum. In this institution there is only an immature specimen (N° 16678) and two specimens without number (collection Biraben) one male and one female from Misiones, Pindapoy, Argentina. Al- though Pereira et al. (1999) suggest that they are syntypes, I cannot accept this conclusion. At MNRJ there are two specimens (a female and male) from Misiones, Pindapoy, Argen- 272 THE JOURNAL OF ARACHNOLOGY tina, labelled by Mello-Leitao as “typus.” Judging by the measurements, etc., these are the holotype and female paratype. The following apomorphic characters of Aulonia luteonigra Mello-Leitao — retrolateral apophysis in the male palpal tibia (Figs. 4, 5), terminal apophysis and lateral apophysis on the male palp (Figs. 4, 6, 7) — lead me to de- duce that A. luteonigra share them with Aga- lenocosa singularis Mello-Leitao 1944 and Agalenocosa punctata Mello-Leitao 1944. For this reason, it must be established as a new combination. Distribution. — Argentina: Misiones, Pin- dapoy; Santa Maria. Specimens examined. — Six specimens: 1 d 1 $ from Pindapoy, Argentina (holotype and female paratype) at MNRJ labelled by Mello-Leitao as ''typus''; one immature from Misiones, Pindapoy, Argentina, at MLP (N° 16678) labelled by Mello Leitao; one male and one female from Misiones, Pindapoy, Ar- gentina at MLP (collection Biraben) labelled by Mello-Leitao as “Cotipo”; one female from Argentina, Misiones, Santa Maria at MACN. Allocosa brasiliensis (Petrunkevitch 1910) Figs. 8-10 Moenkhausiana brasiliensis Petrunkevitch 1910: 223, figs. 26-29. Allocosa brasiliensis: Capocasale 1990: 133. Glieschiella senex Mello-Leitao 1945: 254. New synonym. Comments. — The holotype is a male, not a female as Mello-Leitao said. I have examined this specimen from Entre Rios, Colon, Argen- tina, deposited at MLP. Capocasale (1990) synonymized Glieschiel- la Mello-Leitao 1932 and Moenkhausiana (Petrunkevitch 1910) with Allocosa Banks 1900; and since Glieschiella halophila Mello- Leitao 1932 and Moenkhausiana argentinen- sis Mello-Leitao 1938 were immatures he considered them nomina dubia. However, Ca- pocasale (1990: 137) omitted this conclusion. Consequently, Glieschiella senex Mello-Lei- tao = Allocosa senex (Mello-Leitao); Glies- chiella alticeps Mello-Leitao = Allocosa al- ticeps Mello-Leitao and Moenkhausiana brasiliensis Petrunkevitch = Allocosa brasi- liensis (Petrunkevitch). In this study the apomorphic characters, ter- minal apophysis and the median apophysis of the male palp (Figs. 8-10) of Allocosa senex confirm it is a new synonym of Allocosa bra- siliensis (Petrunkevitch). Thus Allocosa brasi- liensis (Petrunkevitch) and Allocosa alticeps (Mello-Leitao) are the only two good species. ACKNOWLEDGMENTS The following individuals and institutions are gratefully acknowledged for loan of spec- imens and their cooperation: C. Rollard (MHNP), the late M.E. Galiano (MACN), R. Pinto-da-Rocha (MZUSP), C. Sutton and L.A. Pereira (MLP) and the two anonymous re- viewers for their intelligent suggestions. LITERATURE CITED Bonnet,?. 1955. Biblographia Araneorum 2: 1-918. Toulouse. Capocasale R.M. 1990. Las especies de la subfam- ilia Hippasinae de America del Sur (Ara- neae,Lycosidae). Journal of Arachnology 18: 131-141. Dondale, C. 1986. The subfamilies of wolf spiders (Araneae: Lycosidae). Actas X Congress Inter- national Aracnologia, Jaca/Espana 1:327-332. Holmberg, E.L. 1 876. Aracnidos Argentinos. Ana- les de Agricultura Argentina 4:1-30. Lehtinen, P. & H. Hippa. 1979. Spiders of the Ori- ental-Australian region. I. Lycosidae: Venoniinae and Zoicinae. Annales Zoologica Fennici 16:1- 22. Mello-Leitao, C. 1944. Aranas de la Provincia de Buenos Aires. Revista del Museo de La Plata (N.S.) (Zool. 3) (24):31 1-393. Mello-Leitao, C. 1945. Aranas de Misiones, Cor- rientes y Entre Rios. Revista del Museo de La Plata (N.S.) (Zool. 4) (29):213-302. Pereira, L.A., C. Sutton & M.J. Ramirez. 1999. Ca- talogo de tipos de Araneae (Archnida) del Museo de La Plata 45 (113-1 14):77-100. Petrunkevitch, A. 1910. Some new of little known American spiders. Annals of the New York Academy of Science 19:205-224. Roewer, C. 1954. Katalog der Araneae von 1788- 1954. 2a: 1-923. Verlag von Natura. Bremen. Simon, E. 1897. Histoire naturelle des Araignees 2(2): 1-380. Enc. Roret. Paris. Simon, E. 1898. Description d’Arachnides nou- veaux des families des Agelenidae, Pisauridae, Lycosidae et Oxyopidae. Annales de la Societe Entomologique du Belgique 42:5-34. Manuscript received 20 November 1999, revised 30 November 2000. 2001. The Journal of Arachnology 29:273-275 SHORT COMMUNICATION HARVESTMEN AS COMMENSALS OF CRAB SPIDERS Douglass H. Morse: Department of Ecology and Evolutionary Biology, Box G-W, Brown University, Providence, Rhode Island 02912 USA ABSTRACT. Harvestmen Phalangium opilio regularly feed upon the carcasses of bees and moths dis- carded by crab spiders Misumena vatia Clerck 1757 hunting on flowers. I report one observation of a harvestman unsuccessfully attempting to secure a bee still being fed on by a crab spider. Keywords: Phalangium opilio, Misumena vatia, Opiliones, scavenger, kleptoparasite Harvestmen (Opiliones) are important consumers that feed upon a wide variety of animal and vege- table items, including both dead and live small in- vertebrates (Bishop 1949; Cloudsley-Thompson 1968). Although largely confined to the understory and litter during sunny conditions, a consequence of their limited ability to withstand desiccation (Bishop 1949; Cloudsley-Thompson 1968), har- vestmen may wander widely over the vegetation, including large flowers and inflorescences, at night and at other times of high humidity (Todd 1949; Edgar 1971). The latter sites, such as inflorescences of common milkweed Asclepias syriaca also attract many insects, as well as their predators, one of the most common in northeastern North America being the crab spider Misumena vatia Clerck 1757 (Ara- neae: Thomisidae), a sit-and-wait predator (Morse 1981). Adult female M. vatia capture large prey, including bumble bees, honey bees, wasps, and noctuid moths (Morse 1986). Since the spiders feed without masticating these prey, the released car- casses drop intact, frequently lodging on lower parts of the vegetation in the process and providing a resource for scavengers. In our study site, an old field in Bremen, Lincoln County, Maine, USA (Morse 1981), these discarded carcasses sometimes accumulated on the broad milkweed leaves below the inflorescences, often remaining there for several days. However, when collecting carcasses for an un- related analysis (Fritz & Morse 1985), we noted that many of them disappeared within a day of be- ing dropped by the spiders, even during clear, calm periods when they were unlikely to be blown away by wind or washed off by rain. A few of these car- casses were also dismembered, probably where they had been dropped. These observations suggested the work of harvestmen, since harvestmen often carry food items away from a site and also process large items by dismembering them (Cloudsley- Thompson 1968; Edgar 1971). Also, the harvest- man Phalangium opilio Linnaeus 1758 (Phalangi- idae, Palpatores) commonly visited flowering milkweeds in the study area during the night (re- corded on 42.7% of censuses between 2100-0130 h of 100 inflorescences over a 10 year period, n = 157, unpubl. data). To establish why these carcasses disappeared quickly, we collected eight sets of eight carcasses {n = 64) discarded by M. vatia, dusted them with red micronite dye, placed them under milkweed stems in the middle of the study site at three-day intervals (one set for each interval) and monitored them. Sixty of the 64 dyed carcasses were partly or totally dismembered or removed on the night after they were placed in the field. In 10 instances we subsequently found dyed carcasses, or parts of them, as high as 45 cm above the ground on leaves of the milkweed plants, and as far as ^95 cm away from where we had placed them, indicating that they had been actively carried there. Edgar (1971) reported that harvestmen might carry food to sites of low disturbance, including tree trunks. Using pit- fall traps, we subsequently captured seven P. opilio with mouthparts and facial regions covered by red dye, which they could have obtained only from the insect carcasses. Ants (Formicidae), the other im- portant scavengers in the study area, were unlikely to have removed or processed the carcasses because the carcasses were manipulated only at night. Ant activity in this area is largely diurnal (Fritz & Morse 1981). To explore the scavenging habits of P. opilio fur- ther, we confined 12 of them in 4-liter glass con- tainers, one per container, and presented them with discarded spider prey. All fed heavily on these car- casses, in the process tearing them into smaller 273 274 THE JOURNAL OF ARACHNOLOGY parts similar to those found near the sites of the “food” caches and on the lower leaves of the milk- weed. Nine of the 12 attacked the carcasses during the day (164.4 ± 75.4 min after they were placed in the containers), while the remaining three at- tacked the carcasses during the following evening. The substantial lag time for exploitation of these carcasses in the field (none taken until the following evening) is thus probably a consequence of the low activity levels of harvestmen during the middle of warm, dry days, conditions not experienced in the shade and relatively high humidity of the labora- tory. Given the apparent high frequency of scavenging by the harvestmen, it is of interest to know whether they kleptoparasitize the spiders while the latter are feeding on their large prey, an act that could pro- vide considerably larger rewards for the harvestmen than the spent carcasses. Here I report an observa- tion of a P. opilio attempting to wrest control of the honey bee Apis mellifera prey of an adult fe- male M. vatia. We made this observation under il- lumination of a battery-driven headlamp covered by a red filter. Neither the harvestman nor the spider showed any sign of being affected by the resulting red light. At 2125 h on 21 July 1982, shortly after darkness, we observed an adult P. opilio on an in- florescence of a common milkweed in full flower. Within this inflorescence, approximately 3 cm away, an adult female M. vatia was still feeding on a honey bee it had captured at 1430 h the preceding afternoon. Although the spider was largely buried within the flowers of the inflorescence, its prey was located on the outside of that inflorescence in a con- spicuous and seemingly vulnerable position. The harvestman initially moved to the end of the bee opposite the spider (the bee’s abdomen) and at- tempted to grasp it with its mouthparts three dif- ferent times within a few seconds. Each time the spider responded aggressively to the approaches of the harvestman by rearing and rapidly moving its large forelimbs forward. In response, the harvest- man quickly retreated backward for one to two bee lengths, simultaneously lowering its body so that it was situated immediately behind the bee. Rapidly following each thrust and retraction by the spider, the harvestman lunged forward in an apparent at- tempt to secure the bee carcass. In a final effort the harvestman quickly advanced on top of the bee, but during the rapid subsequent response by the spider the harvestman fell off the umbel and dropped into the grass about 80 cm below its previous location. We observed it there for 10 min; but it did not at- tempt to climb back up the plant, and eventually it wandered away from the plant, Misumena vatia would be unlikely to take the harvestman as prey, although spiders are often list- ed as regular predators of harvestmen (Edgar 1971). We have never seen M. vatia with harvestman prey, although logging thousands of hours of field obser- vations on them in over 20 years (1977-2000), both by day and night, and documenting a wide variety of other prey taken by this spider (pers. obs.). Fur- ther, as implied, P. opilio is common in the study area and regularly recorded in censuses (unpubl. data). Therefore the danger inherent in this act, of- ten cited as an important tradeoff of kleptoparasi- tism (Whitehouse 1997), seems low and unlikely to inhibit the harvestman’s effort to secure the food item. In common with other Palpatores, P. opilio possesses large, anterolateral exocrine glands (Bish- op 1949), which contain noxious secretions that ap- pear to deter many potential predators (Edgar 1971). However, we have found the brown crab spi- der Xysticus emertoni Keyserling 1880, a far less frequent visitor to flowers than M. vatia, feeding on P. opilio (pers. obs.). Xysticus emertoni regularly feeds on putatively noxious prey that we have never seen M. vatia exploit (Morse 1983). The ready exploitation of spent prey by P. opilio strongly suggests that the interaction between this harvestman and the spider was an extension of nor- mal scavenger behavior, though the repeated at- tempts to wrest the bee carcass from the spider were consistent with predatory behavior by the harvest- man. Phalangium opilio is well known to prey on small invertebrates (Bristowe 1949). Sabini & Gnaspini (1999) have recently reported an instance of a tropical gonyleptid species taking a moth from a ctenid spider hunting on a tree trunk, which they believe to be the first reported instance of klepto- parasitism by a harvestman. Nearly all instances of probable kleptoparasitism involving spiders have been reported from web-building species, probably because of the relatively high availability of prey there and the web owner’s difficulty of patrolling all parts of large webs (Vollrath 1987; Cangialosi 1997; Grostal & Walter 1997). As we have observed this interaction only once, it seems unlikely to be common, although it would not be observed routinely, given the time of day at which it occurred. It would appear unlikely to result in a major loss of resources to the spiders, espe- cially if such attacks took place after the carcass had been almost completely processed, as in the present instance, when the spider had already re- tained the bee considerably longer than usual (Morse & Fritz 1982). Even then, the spider showed no tendency to give up its prey to the harvestman, so it remains unclear whether P. opilio would often succeed in appropriating such food items before they were discarded by their original exploiters. It thus seems premature to consider P. opilio to be a kleptoparasite of M. vatia. However, P. opilio clear- ly benefits as a commensal of M. vatia, obtaining a resource that would otherwise be unavailable to it. MORSE— HARVESTMEN AS COMMENSALS 275 ACKNOWLEDGMENTS I thank J.M. Kraus and E.L. Leighton for comments on the manuscript and W.R Morse for assistance in the field. These data were gathered incidental to work performed with the support of NSF DEB 80-08501 -AO 1 and IBN98- 16692. Voucher specimens have been deposited in the National Museum of Natural History, Smithsonian Institution (harvestman) and the American Museum of Natural History (spiders). I thank W.A. Shear for identifying the harvestman. LITERATURE CITED Bishop, S.C. 1949. The Phalangida (Opiliones) of New York. Proceedings of the Rochester Acad- emy of Science 9:159-235. Bristowe, W.S. 1949. The distribution of harvest- men (Phalangida) in Great Britain and Ireland, with notes on their names, enemies and food. Journal of Animal Ecology 18:100-114. Cangialosi, K.R. 1997. Foraging versatility and the influence of host availability in Argyrodes tri- gonum (Araneae, Theridiidae). Journal of Arach- nology 25H 82-193. Cloudsley-Thompson, J.L. 1968. Spiders, Scorpi- ons, Centipedes and Mites. Oxford: Pergamon. Edgar, A. L. 1971. Studies on the biology and ecol- ogy of Michigan Phalangida (Opiliones). Mis- cellaneous Publications of the Museum of Zo- ology, University of Michigan 144:1-64. Eisner, T, D. Alsop & J. Meinwald. 1978. Secre- tions of opilionids, whip scorpions and pseudo- scorpions. Pp. 87-99. In Arthropod Venoms. (S. Bettini, ed.), Springer- Verlag, Berlin. Fritz, R.S, & D.H. Morse. 1985. Reproductive suc- cess, growth rate and foraging decisions of the crab spider Misumena vatia. Oecologia 65:194- 200. Grostal, P. & D.E. Walter. 1997. Kleptoparasites or commensals? Effect of Argyrodes antipodianus (Araneae: Theridiidae) on Nephila plumipes (Ar- aneae: Tetragnathidae). Oecologia 111:570-574. Morse, D.H. 1981. Prey capture by the crab spider Misumena vatia (L.) (Thomisidae) on three com- mon native flowers. American Midland Natural- ist 105:358-367. Morse, D.H. 1982. The turnover of milkweed pol- linia on bumble bees, and implications for out- crossing. Of^^^iogia 53:187-196. Morse, D.H. 1983. Foraging patterns and time bud- gets of the crab spiders Xysticus emertoni Key- serling and Misumena vatia (Clerck) (Araneae: Thomisidae) on flowers. Journal of Arachnology 11:87-94. Morse, D.H. 1986. Predatory risk to insects for- aging at flowers. Oikos 46:223-228. Morse, D.H. & R.S. Fritz. 1982. Experimental and observational studies of patch-choice at different scales by the crab spider Misumena vatia. Ecol- ogy 63:172-182. Sabino, J. & P. Gnaspini. 1999. Harvestman (Op- iliones, Gonyleptidae) takes prey from a spider (Araneae, Ctenidae). Journal of Arachnology 27: 675-678. Todd, V. 1949. The habits and ecology of the Brit- ish harvestmen (Arachnida, Opiliones), with spe- cial reference to those of the Oxford district. Journal of Animal Ecology 18:209-229. Vollrath, E 1987. Kleptobiosis in spiders. Pp. 274- 286. In Ecophysiology of Spiders. (W. Nentwig, ed.). Springer- Verlag, Berlin, Whitehouse, M.E.A. 1997. The benefits of stealing from a predator: Foraging rates, predation risk, and intraspecific aggression in the kleptoparasitic spider Argyrodes antipodiana. Behavioral Ecol- ogy 8:663-667. Manuscript received 12 April 2000, revised 30 No- vember 2000. 2001. The Journal of Arachnology 29:276-278 SHORT COMMUNICATION DIFFERENCES IN THE ACTIVITY OF JUVENILES, FEMALES AND MALES OF TWO HUNTING SPIDERS OF THE GENUS CTENUS (ARANEAE, CTENIDAE): ACTIVE MALES OR INACTIVE FEMALES? Fabiola M.D. Salvestrini and Thierry R. Gasnier: Depto. de Biologia/ICB, Fundagao Universidade do Amazonas, Av. Gal. R.O.J. Ramos 3000, CEP 69067-000, Manaus, AM, Brazil ABSTRACT. The difference in activity levels between adult male and female spiders has been attributed to a more sexually motivated searching behavior by males, but the possibility that females reduce their activity when they reach maturity has not been considered, which may be evaluated by comparing adults and late instar juveniles behavior. We recorded the displacements during 15 min periods for 137 males, females and juveniles of Ctenus amphora and C. crulsi, two similar-sized syntopic hunting spiders species which search for prey on the leaf litter in central Amazonian tropical rainforests. For both species, males were significantly more active than females and juveniles. Ctenus amphora females were less active than juveniles, but the C. crulsi female activity did not differ from the juvenile activity. There were no signif- icant differences in activity between these species for males and females, but the juveniles of C. amphora where more active than the juveniles of C. crulsi. Therefore, differences in activity between sexes are not always restricted to changes in male behavior, and the degree of decrease in female activity may depend on how active juveniles are. Keywords: Amazonia, behavior, movements, foraging mode Hunting spiders actively move about in search of prey (Uetz et al. 1999). However, activity levels may differ between adult males and females (e.g., Schmitt et al. 1990). This difference has been attributed to a more sex- ually motivated searching behavior of males (Rovner & Barth 1981); but, as far as we know, the possibility that females reduce their activity when they reach maturity has not been considered. To test this hypothesis it is nec- essary to compare the behavior of adults with that of late instar juveniles. The objective of the present paper is to evaluate whether the differences in activity between sexes in two species of hunting spiders {Ctenus amphora Mello-Leitao 1930 and C. crulsi Mello-Leitao 1930) may be attributed to a more active be- havior of adult males, a less active behavior of adult females, or both. Both species forage on the leaf litter, do not have fixed retreats, have similar size (both with prosoma length of 5.5-1 1 mm) and are sympatric in the study areas. The observations were made in Adolfo Ducke Forest Reserve, a 10,000 ha “terra-fir- me” primary forest reserve 25 km north of the city of Manaus, Brazil, where the ecology of this genus has been intensively studied (Hofer et al. 1994; Gasnier 1996; Gasnier & H5fer 2001) and on the campus of the Univ- ersidade do Amazonas, a forest fragment in Manaus. Using head lamps, we observed the spiders during their nocturnal activity, in the dry (2 nights in June and 3 nights in October of 1998) and wet seasons (6 nights in January and 3 nights in April of 1999). We memorized the trajectory of the movements for 15 min of activity and recorded the total displacement (including curves) with a tape measure. This was possible because of the low activity levels and the tendency of the spiders to move in straight lines. We tried to minimize the effect of our presence on the behavior of the spider by using a red filter on the lamp and by avoid- ing movements. Voucher specimens for these 276 SALVESTRINI & GASNIER— ACTIVITY IN CTENUS 211 studies are deposited in the arachnological collection of the Institute Nacional de Pes- quisas da Amazonia under the numbers INPA- 001 to INPA=023. We used non-parametric statistics (Mann Whitney U-test and Kruskal- Wallis //“test), for all comparisons. Our sig- nificance level was a — 0.05; however, we adjusted a when multiple comparisons were performed following Rice (1989). When we compared juveniles, males and females, we had three pairs of comparisons per species: in these cases, we used the significance levels of 0.017, 0.025 and 0.05 from the greatest to smallest P value. We observed 137 Ctenus crulsi individuals (28$, 19(3 and 37 juveniles) and C. amphora (15$, IAS and 24 juveniles). There were no significant differences between the species in the displacements of males (f/1419 — 136, P = 0.91) (Fig. 1) or females = 186, P - 0.44). However, C. amphora juveniles were significantly more active than C. crulsi juve- niles (C/24.37 = 579, P = 0.03). There were significant differences in dis- placement among males, females and juve- niles within each species (C. amphora: H = 14.45, P < 0.001; C. crulsi: H = 19.71, P < 0.001). For both species, the males were sig- nificantly more active than females (C. am- phora: C/i4 15 = 176, P < 0.001 and C. crulsi: ^19.28 ~ 436, P < 0.001) and significantly more active than juveniles (C. amphora: C/24.14 ” 94.50, P = 0.02 and C. crulsi: C/37 19 = 143, P < 0.001). However, the species differed when we compared the activity of females and juveniles. Ctenus amphora females were sig- nificantly less active than juveniles (C/24.14 “ 267, P = 0.01), and C. crulsi female activity did not differ from juvenile activity (C/3728 = 564.50, P = 0.48). Our results support the hypothesis that males become more active when they reach maturity. However, at least for C. amphora, females activity does decline. The reason for this, and for the absence of a decline in Ctenus crulsi, is not clear, Gasnier (1996) found no evidence that these species differed in repro- ductive cycle and apparently they reproduce continuously throughout the year, so males probably seek females in all seasons. Extreme sedentary behavior appears to be the foraging strategy adopted by adult females of this ge- nus. However, juveniles of these species have different foraging strategies and this may re- 150 §100 I On S 50 Q 0 C. amphora Juveniles •• C. crulsi 700 ^600 _ • |500 - «400 ■|300 • •5200 ik • 100 - • 0 C. amphora Males •• C. crulsi 120 80 d 100“ Oi B V 5 60- 6 .S 40 ® 20 C. amphora Females ...* — I C. crulsi Figure 1. — Displacement (cm) of juveniles and adults of Ctenus amphora and Ctenus crulsi during a 15 minute period. suit from differences in the species’ diets and the resources available. It will be necessary to study diet and resource availability to deter- mine whether these factors affect the activity of males, females and juveniles of these spe- cies, and why some females become less ac- tive and others maintain the same level of ac- tivity as juveniles. We thank William Magnusson for sugges- tions that improved the manuscript. Financial support came from a fellowship grant from CAPES (PET program) for the first author, and from CNPq (Project 400023/98) for field work. LITERATURE CITED Gasnier, TR. 1996. Ecologia comparada de quatro especies de aranhas errantes do genero Ctenus (Walckenaer) (Araneae, Ctenidae) em uma flo- 278 THE JOURNAL OF ARACHNOLOGY resta na Amazonia Central: Bases para um mo- delo integrado de coexistencia. Dr. Thesis, CAPES/INPA/FUA, Manaus, Brasil. 77 pp. Gasnier, TR. & H. Hofer. 2001. Patterns of abun- dance of four species of wandering spiders (Ctenidae, Ctenus) in a forest in central Ama- zonia. Journal of Arachnology 29:95-103. Hofer, H., A.D. Brescovit & T. Gasnier. 1994. The wandering spiders species of the genus Ctenus (Ctenidae, Araneae) of Reserva Ducke, a rain- forest reserve in central Amazonia. Andrias 13: 81-98. Rice, W.R. 1989. Analysing tables of statistical tests. Evolution 43:223-225. Rovner, J.S. & EG. Barth. 1981. Vibratory com- munication through living plants by a tropical wandering spider. Science 214:464-466. Schmitt, A., M. Schuster & EG. Barth. 1990. Daily locomotor activity patterns in three species of Cupiennius (Araneae, Ctenidae): The males are the wandering spiders. Journal of Arachnology 18:249-255. Uetz, G.W., J. Halaj & A.B. Cady. 1999. Guild structure of spiders in major crops. Journal of Arachnology 27:270-280. Manuscript received 21 July 2000, revised 8 Jan- uary 2001. 2001. The Journal of Arachnology 29:279-280 BOOK REVIEW Forest Spiders of South East Asia. Christa L. Deeleman-Reinhold. 2001. K. Brill NV. L eiden, The Netherlands, xii + 592 pages. ISBN 90- 04-11959-0. US$200. (US and Canadian orders to cs@brillusa.com) The title of this book suggests a broad cov- erage of the south east Asian fauna, but the subtitle quickly reveals that its main subject matter is a revision of just six families from the area. Whether additional volumes on other families are projected is not stated. The book is well made, the printing, paper and binding all of high quality, print clear and easily readable, illustrations well laid out and of generous size, and distribution maps (50 of them) large and clear. The author is to be com- plimented on this culmination of 20 years of research in many parts of south east Asia. An introductory section of 26 pages deals briefly with objectives, definition of the area covered (Thailand, Malaysia, and Indonesia exclusive of Irian Jay a), history of araneology of the area, some aspects of spider natural his- tory, taxonomy and zoogeography, and meth- ods of collecting and study. A six page glos- sary follows the introduction and is succeeded by a 40 page illustrated key to araneomorph spider families (liphistiids and mygalomorphs are keyed only to order-group level). The key runs to only 48 couplets, 33 of the 40 pages are taken up by 131 figures. (These figures are numbered separately from and in addition to the figures in the revisioeary section.) The bulk of the book, almost 500 pages, consists of a study of the forest spiders of six families - Clubionidae, Corinnidae, Liocrani- dae, Gnaphosidae, Prodidomidae and Tro- chanteriidae. A seventh family, Miturgidae, is mentioned only to be dismissed as not occur- ring in south east Asia. (The author has re- turned the Butichurinae to the family Clu- bionidae.) A new subfamily, 18 new genera and over 100 new species are described, the great ma- jority of them in the first three of the families listed above. The illustrations (989 of them) are mostly of palps and epigyna or habitus (minus legs or legs shown on one side), with various other structures shown as appropriate. Lists of species from other tropical Asian areas are presented, and unidentified speci- mens are mentioned with brief notes, and shown on the distribution maps. Occasional genera from outside the area are included. The author states that “type species of all genera ostensibly present in the tropical Asian region have been included.” Diagnoses and descriptions are given for both new and previously described taxa. The species descriptions include primarily mea- surements, short notes on coloration, chelic- eral teeth, leg spination and special structures such as abdominal scuta. Genital structures are discussed largely in the diagnoses. Other information includes collection data, type lo- cality, habitat, distribution and (usually) et- ymology of new names. The taxonomic sec- tion is followed by acknowledgements (two pages), a list of aracheological periodicals and societies (one page), a list of references (six pages), index (nine pages), and eight photographic color plates with 19 illustrations, including the remarkable ant-mimicking cor- ieeid Pranburia, This is, so far as I know, the first and only high-quality major taxonomic work on south Asian spiders so far produced. The author is surely to be congratulated on a huge job well done. The criticisms below (and some criti- cisms are always called for by a work of this size) do not detract significantly from the work. The most obvious, and at first rather jarring, defect is in the illustrations, a large number of which are quite noticeably asymmetrical. Pro- ducing symmetrical drawings can be difficult, but a simple trick solves the problem. (Some 279 280 THE JOURNAL OF ARACHNOLOGY specimens, of course, really are asymmetrical, but they are few.) Details seem fuzzy in some drawings, but this cannot be judged properly without comparison with specimens. There are frequent minor errors of spelling and grammar, and awkwardnesses, of English usage. I assume this results from the author’s writing in what is not her first language. The publisher should bear some of the responsi- bility for this. Two terms used frequently in the text acti- vate two of my pet peeves. The words “chi- tinized” and “sclerotized” are both used for describing hard and rigid structures. Chitini- zed is not appropriate for this meaning, though it was commonly so used in older lit- erature. Chitin is a soft, flexible substance. Hardening is by sclerotization. The term “vul- va” is used for internal female genitalic struc- tures, a common usage in Europe, but entirely inappropriate. The word vulva, simply trans- ferred from vertebrate anatomy, refers specif- ically to external, not internal structures. I see no problem in simply referring to the whole secondary genitalic apparatus, external and in- ternal, as the epigynum, and referring to, e.g., “ventral view” and “dorsal view, cleared.” It is astonishing to have two major works on tropical Asian spiders appear nearly si- multaneously, especially two so well done. This volume, in conjunction with Frances and John Murphy’s, should surely stimulate inter- est in south east Asian spiders. It is unfortu- nate that the very high price of the present volume will probably severely limit its avail- ability. The contrast in prices of the two works could scarcely be greater. Joseph A. Beatty: Dept, of Zoology, South- ern Illinois University, Carbondale, Illinois 62901 USA 2001. The Journal of Arachnology 29:281-282 BOOK REVIEW An Introduction to the Spiders of South East Asia, With Notes on All the Genera. Frances and John Murphy. 2000. Malaysian Nature Society, RO. Box 10750, 50724 Kuala Lumpur, Malaysia, vii + 625 pp., ca. US$34 (yes, $34!) plus postage. This is a most remarkable book, quite un- like anything previously produced on spiders. It isn’t exactly a field guide, although it con- tains over 250 color images of spiders taken by one of the most talented photographers ever to grace our field, Frances Murphy, and its discussions of taxa are organized in ways to make them maximally useful to collectors and field biologists. It isn’t exactly an identi- fication manual, even of the ''How to Know the Spiders'' ilk, for there are no dichotomous keys (or genitalic illustrations for species identification) to be found anywhere between its covers. But it is exactly what the title in- dicates - a superb introduction to the spiders of a significant chunk of the world. That chunk is somewhat curiously defined; the book covers, as one would expect, Viet- nam, Cambodia, Laos, Thailand, and Myan- mar. But it also covers Sumatra, Java, Borneo, and even the Philippines. Even more surpris- ingly, a large part of southern China, and Tai- wan, are included (although the authors admit that, in retrospect, some of the Chinese prov- inces covered have faunas with more northern affinities and don’t fit well). In any case, the volume will be of interest to arachnologists everywhere, for it includes a number of unique features. Perhaps most ob- vious is the plethora of habitus drawings; I doubt that there is any other work that in- cludes so many fine drawings of the entire bodies of such a varied cast of characters (many of these drawings are by the world- renowned spider illustrator Michael Roberts, and were especially commissioned for the book). In some cases, these are probably the first habitus drawings ever to appear for given taxa (such as the family Cithaeronidae). Also unique is the organization, especially in the treatments of the larger families. Take, for example, the araneids. First discussed are a number of rare genera that include only one or two species; in most cases, no information on these taxa has appeared since their original description (often a century ago), and little can be said about these animals. For more modem taxa, likely errors are often pointed out; so, for example, a Chinese species de- scribed in the New World genus Eustala in 1990 is suggested to be closer to Cyclosa in- stead (and in this case, a transfer of the species to that genus has actually been published, in an obscure Chinese journal). For the genera more likely to be recog- nized, the treatments are arranged by where the spiders’ webs are most likely to be found: on vegetation, on dead twigs or bare branches, or at the ground layer. One genus, the curious Chorizopes, is even separated out as being found in leaf litter - I was not aware of that, or that these spiffy animals spin no webs and instead prey on other spiders! The authors have spent a considerable amount of time col- lecting and observing spiders in southeast Asia, and the book is chock-full of such tidbits of natural history information. For that reason, as well as an often delightfully sly turn of phrase, even the most detailed parts of the text are quite readable. With regard to Cithaeron, for example, we’re told that “When disturbed, their main defence against even the most ex- perienced collectors is an unreasonable turn of speed.” As introductory material, to help newcom- ers to the field, the book includes brief ac- counts of the other arachnid orders (even those which don’t occur in southeast Asia), 281 282 THE JOURNAL OF ARACHNOLOGY spider anatomy, natural history, and collecting techniques. The liphistiids, mygalomorphs, and araneomorphs are treated separately; but within the two large infraorders, the families are listed alphabetically, which leads to some strange juxtapositions (anapids are thus found between amaurobiids and anyphaenids, rather than with mysmenids or symphytognathids). But tables of “field hints for families” will help the novice navigate through this huge compendium of information. After the family discussions, there is a full checklist of species recorded from the area, comprising over 80 pages, with detailed geo- graphic data. The bibliography is extensive (another 35 pages), and there are useful lists of societies, periodicals, a glossary, and a de- tailed index. The book is capped by 32 gor- geous color plates of photographs, mostly by Frances Murphy. Those of us fortunate enough to have known Frances regret im- mensely that she did not survive to see this publication, but it is a most impressive tribute to her unflagging enthusiasm, and to her de- sire to communicate that enthusiasm to others. Indeed, as John Murphy aptly phrases it, he “became an arachnologist by marriage” (a fate with which I can readily sympathize, since I became one by courtship instead)! As with any project this large, there are al- ways items about which one could carp (Cy- clocosmia is known from Thailand as well as China; the accounts of Crassignatha on pp. 83 and 221, for example, imply that there is some controversy about the family-level re- lationships of the genus, when there is only a difference in the relative ranking of groups in- volved). But, on the whole, typographical and other errors are quite uncommon, and they fade into total insignificance when one con- siders that this enormously useful volume has been made available at a price that seems im- possibly low. There is surely no better bargain to be had, for the selling price is an entire order of magnitude lower than that of some similarly large volumes ! For this feat, both the authors and the publisher (Mr. Henry Barlow of the Malaysian Nature Society) are to be congratulated heartily (potential purchasers may wish to contact Mr. Barlow at hsbar@pc.jaring.my, or RO. Box 10139, 50704 Kuala Lumpur, for details on exchange rates, postage options, and payment methods). I suspect that the inexpensive availability of such a remarkably useful volume will lead to a substantial increase in interest in, and work on, the southeast Asian arachnid fauna, and in the end, there could be no more appropriate tribute to Frances Murphy than that! Norman I. Platnick: Division of Inverte- brate Zoology, American Museum of Nat- ural History, Central Park West at 79th St., New York, New York 10024 USA INSTRUCTIONS TO AUTHORS (revised August 2000) Manuscripts are accepted in English only. Authors whose primary language is not English may consult the editors for assistance in obtaining help with manuscript preparation. All manuscripts should be prepared in general accordance with the current edition of the Council of Biological Editors Style Manual unless instructed otherwise below. Authors are advised to con- sult a recent issue of the Journal of Arachnology for additional points of style. Manuscripts longer than three printed journal pages should be prepared as Feature Articles, shorter papers as Short Commun- ications. Send four identical copies of the typed mate- rial together with copies of illustrations to the Managing Editor of the Journal of Arachnology. Paula Cushing, Managing Editor, Dept, of Zoology, Denver Museum of Nature & Science, 2001 Colorado Blvd., Denver, CO 80205. [Telephone (303)-370-6442; FAX (303)-33 1-6492; E-mail pcushing@dmms.org]. The Managing Editor will forward your manuscript to one of the Subject Editors for the review process. You will receive correspondence acknowledging the receipt of your manuscript from the responsible subject editor or the managing editor, with the manuscript number of your manuscript. Please use this number in all correspondence regarding your manuscript. Corres- pondence relating to manuscripts should be directed to the Managing Editor or the appropriate Subject Editor. After the manuscript has been accepted, the author will be asked to submit the manuscript on a computer disc in a widely-used word processing program. Indicate clearly on the computer disc the word processing pro- gram and the type of computer (Mac or PC). FEATURE ARTICLES Title page. — The title page will include the complete name, address, and telephone number of the author with whom proofs and correspondence should be exchanged, a FAX number and electronic mail address if available, the title in capital letters, each author’s name and address, and the running head (see below). Abstract. — The heading in capital letters should be placed at the beginning of the first paragraph set off by a period. A second abstract, in a language pertinent to the nationality of the author(s) or geographic region(s) emphasized, may be included. Keywords. — Give 3-5 appropriate keywords fol- lowing the abstract. Text. — Double-space text, tables, legends, etc. throughout. Three categories of headings are used. The first category (METHODS, RESULTS, etc.) is typed in capitals and on a separate line. The second category of heading, in bold type, begins a paragraph with an indent and is separated from the text by a period and a dash. The third category of heading may or may not begin a paragraph but is italicized and separated from the text by a colon. Use only the metric system unless quoting text or referencing collection data. All decimal fractions are indicated by the period (e.g.,-0.123). Citation of references in the text: Cite only papers already published or in press. Include within parenthe- ses the surname of the author followed by the date of publication. A comma separates multiple citations by the same author(s) and a semicolon separates citations by different authors, e.g., (Smith 1970), (Jones 1988; Smith 1993), (Smith 1986, 1987; Smith & Jones 1989; Jones et al. 1990). Include a letter of permission from any person who is cited as providing unpublished data in the form of a personal communication. Literature cited section.^ — ^Use the following style, and include the full unabbreviated journal title. Lombardi, S.J. & D.L. Kaplan. 1990. The amino acid composition of major ampullate gland silk (drag- line) of Nephila clavipes (Araneae, Tetragnathidae). Journal of Arachnology 18:297-306. Kraftt, B. 1982. The significance and complexity of communication in spiders. Pp. 15-66, In Spider Communications: Mechanisms and Ecological Significance. (P.N. Witt & J.S. Rovner, eds.). Princeton University Press, Princeton, New Jersey. Footnotes. — Footnotes are permitted only on the first journal page to indicate current address or other information concerning the author. These are placed together on a separate manuscript page. Tables and fig- ures may not have footnotes. Running head. — The author’s sumame(s) and an abbreviated title should be typed all in capital letters and must not exceed 60 characters and spaces. The running head should be placed near the top of the title page. Taxonomic articles. — Consult a recent taxonomic article in the Journal of Arachnology for style, or contact the Subject Editor for Systematics. Papers containing the original taxonomic description of the focal arachnid taxon should be listed in the Literature Cited section. Tables. — Each table, with the legend above, should be placed on a separate manuscript page. Only horizon- tal lines (usually three) should be included. Tables may not have footnotes; instead, include all information in the legend. Make notations in the text margins to indi- cate the preferred location of tables in the printed text. Illustrations. — ^Address all questions concerning illustrations to the Editor of the Journal of Arach- nology: James W. Berry, Editor; Dept, of Biological Sciences; Butler University, Indianapolis, Indiana 46208 USA. [Telephone (317)-940-9344; FAX (317)- 940-9519; E-mail: jwberry@butler.edu]. All art work must be camera-ready (mounted and labeled) for reproduction. Figures should be arranged so that they fit (vertically and horizontally) the printed journal page, either one column or two columns, with a mini- mum of wasted space. When reductions are to be made by the printer, pay particular attention to width of lines and size of lettering in line drawings. Multiple photos assembled on a single plate should be mounted with only a minimum of space separating them. In the case of multiple illustrations mounted together, each illus- tration must be numbered sequentially rather than given an alphabetic sequence. Written on the back should be the name(s) of author(s) and an indication of top edge. Indicate whether the illustration should be one column or two columns in width. The overall dimensions should be no more than 1 1 inches (28 cm) X 14 inches (36 cm). Larger drawings present greater difficulty in shipping and greater risks of damage for which the JoA assumes no responsibility. In manu- scripts for review, photocopies are acceptable, and should be reduced to the exact measurements that the author wants to appear in the final publication. Make notations in the text margins to indicate the preferred position of illustrations in the printed text. Color plates can be printed, but the author must assume the full cost, currently about $600 per color plate. Legends for illustrations should be placed together on the same page(s) and separate from the illustrations. Each plate must have only one legend, as indicated below: Figures 1-4. — A-us x-us, male from Timbuktu: 1. Left leg; 2. Right chelicera; 3. Dorsal aspect of geni- talia; 4. Ventral aspect of abdomen. Figures 27-34. — Right chelicerae of species of A-us from Timbuktu: 27, 29, 31, 33. Dorsal views; 28, 30, 32, 34. Prolateral views of moveable finger; 27, 28. A- us x-us holotype male; 33, 34. A-us y-us male. Scale = 1.0 mm. Assemble manuscript for mailing. — Assemble the separate sections or pages in the following sequence; title page, abstract, text, footnotes, tables with legends, figure legends, figures. Page charges and reprints. — There are no page charges, but the author will be charged for changes made in the proof pages. Reprints are available only from the Allen Press and should be ordered when the author receives the proof pages. Allen Press will not accept reprint orders after the paper is published. SHORT COMMUNICATIONS The above instructions pertaining to Feature Articles apply also to Short Communications, which should be prepared in the same manner as regular Feature Articles. Short Communications are usually limited in length to three journal pages, including tables and figures. They will be printed in a smaller (10 point) typeface. CONTENTS The Journal of Arachnology Volume 29 Feature Articles Number 2 Autoecology and description of Mummucia mauryi (Solifiigae, Mummuciidae), a new solifuge from Brazilian semi-arid caatinga by Eduardo Xavier & Lincoln Suesdek Rocha 127 An unusual new species of Mundochthonius from a cave in Colorado, with comments on Mundochthonius montanus (Pseudoscorpiones, Chthoniidae) by William B. Muchmore 135 Synonymy of Cecoditha (Cecodithinae) with Austrochthonius (Chthoniinae) (Chelonethi, Chthoniidae) by Mark L.I. Judson 141 Two new species of Hadogenes (Scorpiones, Ischnuridae) from South Africa, with a redescription of Hadogenes bicolor and a discussion of the phylogenetic position of Hadogenes by Lorenzo Prendini 146 Further additions to the scorpion fauna of Trinidad and Tobago by Lorenzo Prendini 173 Phylogenetic analysis of Phalangida (Arachnida, Opiliones) using two nuclear protein-encoding genes supports monophyly of Palpatores by Jeffrey W. Shultz & Jerome C. Regier 189 Description of Hakka, a new genus of jumping spiders (Araneae, Salticidae) from Hawaii and east Asia by James W. Berry & Jerzy Proszyriski . . 201 A revision of the Afrotropical spider genus Palfuria (Araneae, Zodariidae) by Tamas Sziits & Rudy Jocque 205 Cribellum and calamistrum ontogeny in the spider family Uloboridae: Linking functionally related but separate silk spinning features by Brent D. Opell 220 Does the structural complexity of aquatic macrophytes explain the diversity of associated spider assemblages? by Josue Raizer & Maria Eugenia C. Amaral 227 On the distribution and phenology of Argyrodes fictilium (Araneae, Theridiidae) at its northern limit of North America by Pierre Paquin & Nadine Duperre 238 Egg sac recognition by female Miagrammopes animotus (Araneae, Uloboridae) by Brent D. Opell 244 Egg covering behavior of the Neotropical harvestman Promitobates ornatus (Opiliones, Gonyleptidae) by Rodrigo Hirata Willemart 249 Comparison of the survival of three species of sac spiders on natural and artificial diets by Divina M. Amalin, Jorge E. Pena, Jonathan Reiskind & Robert McSorley 253 Short Communications Post-maturation molt found in a wolf spider, Pardosa astrigera (Araneae, Lycosidae) by Yasuhiro Fujii 263 Description of the egg sac of Mimetus notius (Araneae, Mimetidae) and a case of egg predation by Phalacrotophora epeirae (Diptera, Phoridae) by Hank Guarisco 267 Review of the South American species of the genera Aulonia and Allocosa (Araneae, Lycosidae) by Roberto M. Capocasale 270 Harvestmen as commensals of crab spiders by Douglass H. Morse 273 Differences in the activity of juveniles, females and males of two hunting spiders of the genus Ctenus (Araneae, Ctenidae): active males or inac- tive females? by Fabiola M.D. Salvestrini & Thierry R. Gasnier . . 276 Book Reviews Forest Spiders of South East Asia, written by Christa L. Deeleman- Reinhold. reviewed by Joseph A. Beatty 279 An Introduction to the Spiders of South East Asia, With Notes on All the Genera, written by Frances & John Murphy, reviewed by Norman I. Platnick 281 The Journal of ARACHNOLOGY OFFICIAL ORGAN OF THE AMERICAN ARACHNOLOGICAL SOCIETY VOLUME 29 2001 NUMBER 3 THE JOURNAL OF ARACHNOLOGY EDITOR-IN-CHIEF: Daniel J. Mott, Lincoln Land Community College MANAGING EDITOR: Paula Cushing, Denver Museum of Nature & Science SUBJECT EDITORS: Ecology — Maggie Hodge, Hiram College; Systematics — Mark Harvey, Western Australian Museum; Behavior and Physiology — Robert Suter, Vassar College EDITORIAL BOARD: Alan Cady, Miami University (Ohio); James Carrel, University of Missouri; Jonathan Coddington, Smithsonian Institution; William Eberhard, Universidad de Costa Rica; Rosemary Gillespie, University of California, Berkeley; Charles Griswold, California Academy of Sciences; Marshal Hedin, San Diego State University; Herbert Levi, Harvard University; Brent Opell, Virginia Polytechnic Institute & State University; Norman Platnick, American Museum of Natural History; Ann Rypstra, Miami University (Ohio); Paul Selden, University of Manchester (UK.); Matthias Schaefer, Universitset Goettingen (Germany); William Shear, Hampden- Sydney College; Petra Sierwald, Field Museum; Keith Sunderland, Horticulture Research International (UK.); I-Min Tso, Tunghai University (Taiwan). The Journal of Arachnology (ISSN 0161-8202), a publication devoted to the study of Arachnida, is published three times each year by The American Arach- nological Society. Memberships (yearly): Membership is open to all those in- terested in Arachnida. Subscriptions to The Journal of Arachnology and American Arachnology (the newsletter), and annual meeting notices, are included with mem- bership in the Society. Regular, $40; Students, $25; Institutional, $125 . Inquiries should be directed to the Membership Secretary (see below). Back Issues: Patricia Miller, P.O. Box 5354, Northwest Mississippi Community College, Senatobia, Mississippi 38668 USA. Telephone: (601) 562-3382. Undelivered Issues: Allen Press, Inc., 1041 New Hampshire Street, P.O. Box 368, Lawrence, Kansas 66044 USA. THE AMERICAN ARACHNOLOGICAL SOCIETY PRESIDENT: Brent D. Opell (2001-2003), Department of Biology, Virginia Polytechnic Institute and State University, Blacksburg, Virginia 24061 USA. PRESIDENT-ELECT: Gary Miller (2001-2003), Department of Biology, University of Mississippi; University, Mississippi 38677 USA. MEMBERSHIP SECRETARY: Norman I. Platnick (appointed), American Museum of Natural History, Central Park West at 79th St., New York, New York 10024 USA. TREASURER: Gail E. Stratton, Department of Biology, University of Missis- sippi, University, Mississippi 3S677 USA. SECRETARY: Alan Cady, Dept, of Zoology, Miami Univ, Middletown, Ohio 45042 USA. ARCHIVIST: Lenny Vincent, Fullerton College, Fullerton, California 92634. DIRECTORS: Bruce Cutler (2000-2002), Richard Bradley (2001-2003), Frederick Coyle (2001-2003). HONORARY MEMBERS: C. D. Dondale, H. W. Levi, A. F. Millidge, W. Whit- comb. Cover photo: Western Black widow, Latrodectus hesperus, molting. From New Mexico. Photo by Bryan E. Reynolds. Publication date: 28 December 2001 @ This paper meets the requirements of ANSI/NISO Z39.48-1992 (Permanence of Paper). 2001. The Journal of Arachnology 29:283-303 GROSS MUSCULAR ANATOMY OT LIMULUS POLYPHEMUS (XIPHOSURA, CHELICERATA) AND ITS BEARING ON EVOLUTION IN THE ARACHNIDA Jeffrey W. Shultz: Department of Entomology, University of Maryland, College Park, Maryland 20742=4454 USA ABSTRACT. Due to their widespread use as model systems and their reputation as living fossils, horse- shoe crabs (Xiphosura) have been studied intensively by physiologists and paleontologists. The close phylogenetic relationship between horseshoe crabs and arachnids might also have been expected to inspire studies of xiphosurans by comparative arachnologists, but surprisingly few have been undertaken. Here, the first exhaustive survey of muscular anatomy of the Atlantic horseshoe crab is conducted as part of an on-going study of the evolutionary morphology and phylogeny of arachnids. Dissections of adult and immature individuals established 113 muscle groups comprising over 750 individual muscles, with several being recognized or correctly described for the first time. New insights into skeletomusciilar evolution and phylogeny of arachnids were derived primarily from the axial muscle system. Specifically, it is argued that Limulus retains a box-truss axial muscle system like that of plesiomorphic members of other arthropod groups, that this is also a plesiomorphic condition for Chelicerata, and that arachnids are united by the loss of one component of this system, the anterior oblique muscles. Combined with comparative morpho- logical and molecular evidence from previous studies, this study adds greater weight to the widely held view that, among extant chelicerates, Xiphosura and Arachnida are monophyletic sister groups and coun- ters recent speculation that scorpions are more closely related to xiphosurans than to spiders, whipscorpions and other arachnids. Keywords: Horseshoe crab, morphology, phylogeny, muscles Due to its large size, availability to inves- tigators and reputation as a “living fossil,” the Atlantic horseshoe crab, Limulus poly- phemus (Linneaus 1758) (Xiphosura, Cheli- cerata), is one of the most intensively studied invertebrates (e.g., Cohen 1979; Bonaventura et al. 1982; Sekiguchi 1988), and aspects of its external and internal anatomy are routine- ly depicted in textbooks. Investigations of its skeletomuscular anatomy were undertaken repeatedly in the nineteenth and early twen- tieth centuries (e.g., Milne-Edwards 1873; Owen 1873), with the works of Lankester (1881, 1885; Lankester et al. 1885) and Pat- ten (1893, 1912; Patten & Redeebaugh 1899-1900; Patten & Hazen 1900) being the most influential. Lankester's work led to the conclusion that Limulus is a chelicerate (orig- inally “arachnid”) rather than a crustacean, and subsequent workers have tended to adopt his terminology and muscle numbering sys- tem (e.g., Snodgrass 1952; Manton 1964; Wyse & Dwyer 1973). Although Patten’s studies generally surpassed Lankester’s in quality and detail, the reputation of Patten’s descriptive work may have suffered from its association with his failed attempt to dem- onstrate the origin of vertebrates from chel- icerates (Patten 1912). Later workers often based their phylogenetic and functional in- ferences on these early descriptions (e.g., Versluys & Demoll 1922; Stprmer 1944; Page 1949; Manton 1964; Wyse & Dwyer 1973), and cursory empirical work has tend- ed to corroborate the classic treatments. Despite the influence of the early anatomi- cal studies, important discoveries were made whenever detailed exploratory surveys were undertaken, especially those focusing on em- bryonic or larval stages. For example, Iwanoff (1933) showed the chilaria in Tachypleus gi- gas (Muller 1785) (formerly Limulus moluc- cans) to be appendages of postoral somite VII and refuted morphological speculations of Versluys & Demoll (1922), who hypothesized the derivation of xiphosurans from arachnids. Scholl (1977) showed that the dorsal hinge be- tween the cephalothorax and abdomen is a ter- gal specialization of postoral somite VIII, dis- covered muscles associated with the pedal 283 284 THE JOURNAL OF ARACHNOLOGY coxae inserting on the walls of the preoral chamber, and showed that postoral somite I (cheliceral somite) migrates rearward during ontogeny and thereby distorts the metameric pattern in the adult. Interestingly, most of these features are also evident in adults and subadults and, as demonstrated here, would have been revealed by more detailed descrip- tions of these stages. Given that new skele- tomuscular features have been discovered by each careful survey of skeletomuscular anat- omy, and the absence of a single treatment encompassing all known skeletal muscles, I chose to undertake an exhaustive survey of skeletal muscles in Limulus, with the specific goal of integrating the resulting data with those obtained from original dissections of arachnids (Shultz 1993, 1999, 2000). The present survey documents 113 muscle groups which encompass over 750 individual muscles. Most have been described in earlier studies, but several evolutionarily significant muscles were undescribed, described impre- cisely or incorrectly, or described correctly in poorly known publications. These features in- clude 1) two sets of three cheliceral muscles arising on the carapace that are generally de- picted as one muscle; 2) four intrinsic chelic- eral muscles; 3) a preoral sphincter elaborated from the coxosternal region of the prosoma; 4) muscles that originate on the coxae of legs 1-4 and insert on the preoral chamber; 5) ex- trinsic chilarial muscles; 6) muscles associated with the chondrites of the opisthosomal ap- pendages (i.e., chilaria and opercula); 7) evi- dence that the axial muscles of the abdomen are homologous with the box-truss muscle system of other plesiomorphic arthropods; 8) evidence that the so-called “branchio-thorac- ic” muscles are axial muscles rather than ex- trinsic muscles of the opercula; 9) evidence that the first six pairs of dorsal endostemal suspensor muscles are members of a single metameric series rather than two different se- ries; 10) thin, sheet-like ligaments connecting the ventral surfaces of the endosternite to the pliable intercoxal cuticle; and 1 1) the absence of the ventral esophageal dilator muscle de- scribed by Lankester et al. (1885). Combined with information currently available for arach- nids, these and other new perspectives on the muscular anatomy of Limulus provide insights in the phylogeny and evolutionary morphol- ogy of arachnids. Notably, this information confirms the widely held view derived from morphological and molecular evidence that Arachnida is monophyletic with respect to Xiphosura and contradicts recent proposals that scorpions are more closely related to xiphosurans than to other arachnids (e.g., van der Hammen 1989; Dunlop 1998). METHODS Preserved specimens of Limulus polyphe- mus were obtained from Ward’s Natural Sci- ence Establishment, Inc. and Carolina Biolog- ical Supply Co. and ranged in mid-sagittal carapace length from 2-12 cm. Dissections were performed using a Leica MIO dissecting microscope at magnifications of 1 X to 1280X. Only standard dissection equipment and tech- niques were used. Drawings were made with a drawing tube, scanned electronically and transformed into black-and-white bitmaps, which were then enhanced and labeled using a variety of graphics software. RESULTS This study does not provide a complete re- description of skeletomuscular anatomy Lim- ulus but surveys the skeletal muscles in an attempt to 1) corroborate descriptions of ear- lier workers (i.e., Lankester et al. 1885; Patten 1912; Snodgrass 1952; Manton 1964; Scholl 1977), 2) clarify ambiguities and correct er- rors in earlier descriptions, 3) describe “new” features, and 4) document gross muscular anatomy in a manner comparable to descrip- tions being generated for other chelicerates (Shultz 1993, 1999, 2000). The “Results” section highlights novel observations, but a brief description of all muscles is provided in Table 1 and sources for detailed descriptions are provided in the text. Italicized Arabic nu- merals refer to the muscles listed in Table 1, and Roman numerals refer to postoral somites. General anatomy. — The chelicerate body consists of two tagmata, prosoma and opistho- soma. The prosoma includes a preoral region and six embryologically postoral appendage- bearing somites (I-VI) (Damen et al. 1998; Tel- ford & Thomas 1998). The primitive opistho- soma was probably composed of 11 or 12 somites and a terminal peri- or post-anal struc- ture, the telson (Weygoldt & Paulus 1979; Shultz 1990; Anderson & Selden 1997). Lim- ulus and other extant xiphosurans depart sec- ondarily from this organization in several SHULTZ— MUSCLES OF LIMULUS 285 ways. Specifically, the first opisthosomal so- mite (VII), along with its appendages (chilaria), and medial tergal elements of the second (VIII) are incorporated into the prosoma to form a “cephalothorax” (Scholl 1977) (Fig. 3B). The remaining opisthosomal somites are consoli- dated dorsally, laterally and posteriorly into a heavily sclerotized abdomen (Figs. 1, 2) which bears six opercular appendages (i.e,, the ante- rior genital operculum with paired genital openings and lacking book gills and five open cula with book gills) inserted into a pliable ventral cuticle (Fig. 2). The hard dorsal surface of the abdomen is here termed the tergum and the hard posterior ventral surface is termed the postopercular sternum. There are seven well- defined opisthosomal somites (VIILXIV) with- in the abdomen which are indicated by the ar- rangement of dorsal entapophyses (i.e., internal projections formed by invagination of the ter- gal exoskeleton), marginal spines, and append- ages (Figs. 1, 3). There is internal evidence of another somite (XV) in the form of a crescent- shaped site of muscle attachment on the ab- dominal tergum (Fig. lA: 36, 37), The mus- culature of somite XV is more complicated in larval stages and resembles that of the more anterior abdominal somites but lacks a dorsal entapophysis (Scholl 1977: fig. 5). Appendicular muscles.- — The skeletomus- cular anatomy of appendages in Lirnulus has been described by several workers, notably Lankester et al. (1885), Patten (1912), Vachon (1945), Snodgrass (1952), Manton (1964), Wyse & Dwyer (1973) and Shultz (1989). Their observations were largely confirmed in the present study (Table 1, Figs. 1-8), and an- other detailed description of this skeletomus- cular system is not provided here. However, the literature contains persistent errors and omissions regarding muscles of the chelicerae, opercular chondrites, chilaria and preoral cox- ostereal apparatus. Consequently, these sys- tems are described in greater detail here and in the following section on the preoral appa- ratus. Cheliceral muscles: The chelicerae com- prise three articles: protomerite, deutomerite and tritomerite (Fig. 5). Each chelicera is bor- dered medially by the epistome and posteri- orly to laterally by the procurved epistomal horns (Fig. 4). Each chelicera is operated by four extrinsic muscles (Fig. 5), three carapace- protomerite muscles {45-47) that originate from a small region on the anteromedial sur- face of the carapace (Fig, 1) and one endos- ternite-protomerite muscle (Fig. 5) that origi- nates on the medial surface of the anterior endosternal horn. Lankester et al. (1885) were mistaken in regarding muscles 45-47 as a sin- gle muscle and in considering muscle 15 a second endosternite-protomerite muscle. The intrinsic muscles are described here for the first time. The pro tomerite-deutomerite joint is equipped with an extensor {49) and a flexor (50), and the deutomerite-tritomerite joint is operated by a closer {51) and an opener (52) (Fig. 5). Opercular chondrites and associated mus- cles: Muscles of the abdominal appendages have been treated in detail elsewhere (e.g,, Lankester et al. 1885; Patten 1912) and these are described briefly in Table 1 and illustrated in Figs. 1 and 8. Only those associated with the opercular chondrites will be described here. Opercular chondrites are paired, roughly cylindrical structures composed of a pliable cartilaginous material (Patten & Hazee 1900; Fahreebach 1999). Each choedrite attaches ventrally at an oval region on the anterior sur- face of each operculum (Fig. 8: bchdt) and projects anterodorsally, but it terminates be- fore reaching the abdominal tergum and ad- heres medially to the lateral surface of the dorsal entapophysis of the same somite as the appendage from which the chondrite origi- nates (Fig. 2C). Longitudinally adjacent chon- drites interconnect dorsally through a thin hor- izontal sheet (Fig. 7C) composed of the same cartilaginous tissue. The shaft of each choe- drite has three muscles whose fibers pass downward and attach to the anterior opercular surface near the base of the chondrite (Fig. 8: 97, 99, 100). These muscles presumably func- tion in compressing and/or flexing the chon- drite, which may act, in turn, as an elastic and/ or hydrostatic skeleton for operating the operculum. Chilarial muscles: Xiphosurans are the only extant chelicerates to retain distinct append- ages (chilaria) (Figs. 2, 4) on somite VII as adults. Each chilarium has a cartilaginous bar (chilarial chondrite) that originates near its posterior margin (Fig. 4: cht) and extends dor- sally to attach to the posterior margin of the endosternite (Patten & Redenbaugh 1899: “capsuliginous bars”). This structure is com- posed of the same material as the opercular 286 THE JOURNAL OF ARACHNOLOGY Table 1. — Muscles of Limulus poIyphemus, numbered and described. Abbreviations: O, origin; I, in- sertion; H: homologs in other studies. References to homologs consist of abbreviated author name and the reference number or name that author used in denoting the muscle. Author abbreviations: LBB, Lan- kester, Benham & Beck (1885); M, Manton (1964); P, Patten (1912); Sch, Scholl (1977); Sh, Shultz (1989); Sn, Snodgrass (1952). Axial and gut muscles 1 Hundreds of loosely packed strands passing through hepatopancreas and other viscera. O: cephal- othoracic carapace, dorsal surface of marginal fold. I; cephalothoracic carapace, ventral surface of marginal fold (not illustrated). H: Sch, IDvm. 2 Many long strands. O: cephalothoracic carapace, dorsal anterior surface of marginal fold. I: ce- phalothoracic carapace, ventral anterior surface of marginal fold (Fig. 2). Probably enlarged com- ponents of 1. H: LBB 66? (Because these muscles pass along the lateral surfaces of the crop, LBB may have mistaken these for pharyngeal muscles.) 3 Transverse, unpaired. O: base of right epistomal horn. I: base of left epistomal horn (Figs 2, 4). May be a component of 5. 4 Hundreds of short, tightly packed strands. O: abdominal tergum, dorsal surface of marginal fold. I: abdominal tergum, ventral surface of marginal fold (not illustrated). 5 Five elements (Sn-Syi). Fibers span crests of adjacent and subadjacent intercoxal folds and epistomal horns to form preoral sphincter (Figs 2, 4). 6 O: subendosternal subneural plastron. I: anterior surface of endostoma, interdigitates with 5 (Figs 2, 4). 7 O: medial surface of endosternal horn and lateral ventral surface of endosternite. I: dorsal and dorsolateral surfaces of esophageal portion of foregut (Figs 2, 1C). H: LBB, 67; Sch, MuVd. 8 Encircles foregut, well developed around crop and gizzard (Figs 2, 1C). H: LBB, S. 9 O: postopercular sternum, anterior margin. I: rectum, ventral wall (Figs 2, 1C). 10 Sheetlike. O: tergum, posteromedial dorsal surface. I: rectum, dorsolateral surface (Figs 1, 2, 1C). Separates 36 and 37 (Fig 7B,C). 11 Nine serial groups. O: carapace and tergum, dorsal medial surfaces. I: dorsal pericardium (Fig. 1). 12 “Veno-pericardiac muscles.” Nine serial members. O: ventral pericardium. I: lateral surface of endosternite (2) or ventral venous sinus (7) (Fig. 2). They span the abdominal space between muscle groups 22 and 26, and interdigitate with the members of 22. H: LBB, 68; Sch, Vpkm. 13 Six serial members, 13,-13v,. O: endosternite, dorsal surface. I: central carapace. (Figs 1, 7A). Probably serial homologs of 17. H: LBB, 49-52, 57-59; Sch, SuE 1-6. 14 O: endosternite at base of tendinous process shared with 13vi. L first dorsal entapophysis (Fig. 7A, C). May be anteriormost member of series 22 (i.e. 22vi). H; LBB, 53; Sch, SuE; M: Fig. 16. 15 O: endosternal horn. I: anterior half of epistomal horn and anteriorly adjacent series of sclerites (Figs 2, 4, 9B). May be anteriormost member of series 16. H: LBB, 30 (Error: not a cheliceral muscle). 16 Three or four serial members, 16m?, 16,v— 16vi. Sheetlike. O: lateral surface of endosternite and associated marginal membrane. I: pliable cuticle between coxae of legs 1-5 (Fig. 4). See 15. 17 Seven serial members, 17viii— 17xiv but the “trilobite larva” has \lxw (Scholl 1977: fig. 5). O: subneural plastrons (17vni-17xiii) or postopercular sternum (17xiv)- I‘- tergum medial to dorsal en- tapophyses (Figs 1, 7C). See 13. H: LBB, 12 (errors); Sch, Dvm. 18 Thin sheet of connective tissue. O: postopercular sternum, anterior medial surface. I: tergum, medial surface posterior to 7th dorsal entapophysis (somite XIV) (Figs 1, 2). 19 Seven serial members, 19vii-19xni- O: endosternite (19vii) or subneural plastron (19viii-19xiii). I: posteriorly adjacent subneural plastron (19vii-19xii) or postopercular sternum (19xiii) (Figs 7C, 9B). H: LBB 5. 20 Variably expressed. O: endosternite or subneural plastron. I: subneural plastron or postopercular sternum, two or more somites posterior to origin (Figs 1C, 9B). 21 Variably expressed. O: subneural plastrons. I: dorsal entapophysis of more posterior somite. Ar- rangement of one individual depicted in Fig. 1C. H: LBB 13-17. 22 Seven serial members, 22vm-22xiv, but the “trilobite larva” has 22xv (Scholl 1977: fig. 5). O: endosternite, posterior dorsal surface. I: 1st to 7th dorsal entapophyses. Anterior-to-posterior se- quence of insertions reflected in medial-to-lateral sequence of origins (Figs 7A, 9B). H: LBB, 1, 2, 54-55, 83-87, 103-107; Sch, Ent. SHULTZ— MUSCLES OF LIMULUS 287 Table 1. — Continued. 23 Seven serial members, 23viii-23xiv O: dorsal surface of endosternite; anterior origins smaller, more medial. I: subneural plastrons of somites (23viii-23xni) and anterior margin of postopercular sternum (23xiv) (Fig- 7C). H: LBB, 3 (in part). 24 Five serial members, 24,x-24xiii. O: medial margin of 2nd to 6th ventral entapophyses (somites IX- XIII). I: postopercular sternum, anterior margin (Fig. 7A). H: LBB, 3, 16 (in part). 25 Variable, up to four serial members. O: medial margin of 1st to 4th ventral entapophyses (somites VIII-XI); if fewer, absent posteriorly. I: 7th dorsal entapophysis (somite XIV) (Fig. 7A). 26 Seven serial members, 26vni-26xiv- O: lateral margins of cardiac lobe of carapace and second dorsal entapophysis (26xiv in part). I: ventral entapophyses (somites VIII-XIV) (Figs 1, 7A, B, 9B). Origin of posterior six members in quasi-concentric pattern with anteriormost element (26, x) located “cen- trally” (Fig. 1). H: LBB, 18, 19; Sch, Btm. 27 O: carapace, cardiac lobe. I: hollow apodeme formed by invagination of hinge between carapace and tergum (Figs 1, 7B, 9B). H: LBB, 78; Sch, Dim 7. 28 O: carapace, cardiac lobe. I: anterior margin of hinge between carapace and tergum (Figs 1, 7B, 9B). H: LBB, 78; M, “retractor dorsalis”; Sch, Dim 7. 29 O: 1st dorsal entapophysis (somite VIII). I: hinge between carapace and tergum (Fig. 7B). 30 Four serial members, 30ix-30xn. O: 1st dorsal entapophysis (somite VIII). I: 2nd through 5th dorsal entapophyses (somites IX-XII) (Figs 7B, 9B). H: LBB, 4a. Telson muscles 31 O: abdominal tergum, medial to 6th dorsal entapophysis (somite XIII). I: telson, dorsal process (Figs 1, 7B). H: LBB, 6, 120? 32 O: abdominal tergum, medial to 7th dorsal entapophysis (somite XIV). I: telson, dorsal process (Figs I, 7B). H: LBB: 6, 101? 33 O: abdomen, 5th dorsal entapophysis (somite XII). I: telson, dorsal process (Fig. 7B). H: LBB, 93. 34 O: abdomen, 6th dorsal entapophysis (somite XIII). I: telson, dorsal process (Fig. 7B). H: LBB, 92. 35 O: abdomen, 7th dorsal entapophysis (somite XIV). I: telson, dorsal process (Fig. 7B). H: LBB, 91. 36 O: abdomen, tergum posterior to 7th dorsal entapophysis (somite XIV) medial to 37. I: telson, dorsal process (Figs 1, 7B). H: LBB, 7. 37 O: abdomen, tergum posterior to 7th dorsal entapophysis (somite XIV) lateral to 36. I: telson, dorsal process (Figs 1, 7B). H: LBB, 8. 38 O: abdomen, postopercular sternum. I: telson, dorsal process (Fig. 7B). H: LBB, 94-97. 39 O: abdomen, tergum lateral to 7th dorsal entapophysis (somite XIV). I: telson, ventrolateral process (Figs 1, 7A). Sometimes with second part originating between 36 and 37 (somite XV?). 40 O: abdomen, 5th dorsal entapophysis (somite XII). I: telson, ventrolateral process (Fig. 7A). H: LBB, 88. 41 O: abdomen, 6th dorsal entapophysis (somite XIII). I: telson, ventrolateral process (Fig. 7A). H: LBB, 89. 42 O: abdomen, 7th dorsal entapophysis (somite XIV). I: telson, ventrolateral process (Fig. 7A). H: LBB, 90. 43 O: abdomen, postopercular sternum medial to 44. I: telson, ventrolateral process (Fig. 7A). H: LBB, 10. 44 O: abdomen, postopercular sternum lateral to 43. I: telson, ventrolateral process (Fig. 7A). H: LBB, II. Cheliceral muscles 45 Long, thin. O: carapace, anteromedial; sometimes with additional fibers originating on endosternal horn. I: protomerite, anterior margin (Figs 1, 5). H: LBB, 24; Sch, MuCh. 46 Long, thin. O: carapace, anteromedial. I: protomerite, medial margin (Figs 1, 5). H: LBB, 24; Sch, MuCh. 47 Long, thin. O: carapace, anteromedial. I: protomerite, lateral margin (Figs 1, 5). H: LBB, 24; Sch, MuCh. 48 O: endosternal horn, medial surface. I: protomerite, posterior process (Fig. 5). H: LBB, 31. 49 O: protomerite, dorsolateral and ventroproximal surfaces. I: deutomerite, dorsal margin (Fig. 5). 50 O: protomerite, dorsomedial and ventroproximal surfaces. I: deutomerite, ventral margin (Fig. 5). 51 O: deutomerite. I: tritomerite, medial margin (Fig. 5). 52 O: deutomerite, lateral surface. I: tritomerite, lateral margin (Fig. 5). 288 THE JOURNAL OF ARACHNOLOGY Table 1. — Continued. Leg muscles (legs 1-5 are appendages of somites II-VI) 53 All legs. O: carapace, near extrinsic muscles of anteriorly adjacent appendage. I: coxa, anteromedial margin (Figs 1, 3A; see also figs 14, 15 and 17 in M). H: LBB, M, Sn, 27. 54 All legs. O: carapace, near extrinsic muscles of anteriorly adjacent leg. I: coxa, anterolateral margin (Figs 1, 3A; see also figs 14, 15 and 17 in M). H: LBB, M, Sn, 26. 55 All legs. O: carapace, near extrinsic muscles of posteriorly adjacent leg. I: coxa, posteromedial margin (Figs 1, 3A; see also figs 14, 15 and 17 in M). H: LBB, M, Sn, 29. 56 All legs. O: carapace. I: coxa, posterolateral margin (Figs I, 3A; see also figs 14, 15 and 17 in M). H: LBB, M, Sn, 28. 57 All legs. O: carapace. I: coxa, posterolateral process (Figs 1, 3 A; see also figs 14, 15 and 17 in M). H: LBB, M, Sn, 25. 58 Leg 5. Small, thin. O: carapace, posterior. I: coxa, posterolateral (Fig. 1;- see also fig. 15 in M). H: M, “dorsal coxal muscle”. 59 All legs. O: eedosternite, ventral surface. I: coxa, anteromedial margin (Fig. 3A; see’ also figs 14, 15 and 17 in M). H: LBB, M, Sn, 34, 37, 40, 43, 46. 60 All legs. O; eedosternite, ventral surface. I: coxa, anterolateral margin (Fig. 3 A; see also figs 14, 15 and 17 in M). H: LBB, M, Sn, 32m, 35o, 38q, 41s, 44y. 61 Ail legs. O: endosternite, ventral surface. I: coxa, posteromedial margin (Fig. 3A; see also figs 14, 15 and 17 in M). H: LBB, M, Sn, 33, 36, 39, 42, 45. 62 All legs. O: endosternite, ventral surface. I: coxa, posterolateral margin (Fig. 3A; see also figs 14, 15 and 17 in M). H: LBB, Sn, 32n, 35p, 38r, 4 It, 44z. 63 Leg 5. O: endosternite, ventral surface. I: coxa, anterior margin (not illustrated but see figs 15 and 17 in M). H: LBB, M, 47. 64 Leg 5. O: endosternite, ventral surface. I: coxa, posterior margin (not illustrated but see figs 15 and 17 in M). H: LBB, M, 60. 65 Legs 1-4. O: inner wall of preoral chamber. I: coxa, posteromedial margin with 66 in legs 2-4 and corresponding region in leg 1 (Fig. 4; see also fig. 14 in M). H: M, “sternite muscle”. 66 Legs 2-4. O: coxa, posteromedial margin. I: moveable endite, anterior surface (Fig. 4; see also fig. 14 in M). H: M, “coxal endite muscle”. 67 All legs, O: coxa, proximal anterior and posterior margins. I: trochanter, dorsal margin and arthrodial membrane (Fig. 6). H: Sn, 1 Leg 5. O: tarsus, ventral surface. I: apotele, ventral margin (Fig. 6). H: Sh, 2; Sn, 21. 68 All legs. O: coxa, ventral anterior surface. I: trochanter, anteroventral margin (Fig. 6). H: Sn, 2. 69 All legs: O: coxa, dorsal posterior and dorsal anterior surfaces. I: trochanter, ventral margin via heavily sclerotized tendons (Fig. 6). H: Sn, 2+3. 70 All legs. O: coxa, ventral posterior surface. I: trochanter, posteroventral margin (Fig. 6). H: Sn, 3. 71 All legs. O: dorsal arthrodial membrane of trochanter-femur joint. I: femur, proximal half with anterior, dorsal and posterior parts (71a-71c) (Fig. 6). H: Sh, 12; Sn, 7. 72 Ail legs. O: trochanter, posterior and ventroposterior surfaces. I:- femur, posterior ventral margin (Fig. 6). H: Sh, 11; Sn, 4. 73 All legs. O: trochanter, distal anterior surface. I: femur, proximal posterior surface (Fig. 6). H: Sh, 10; Sn, 6. 74 All legs. O: trochanter, anterior and ventral surfaces. I: patellar sclerite, proximal end (Fig. 6). H: Sh, 8d; Sn, 10. 75 All legs. O: femur, middle dorsoanterior surface. I: patellar sclerite, distal shaft (Fig. 6). 76 All legs. O: femur, middle ventroposterior surface. I: patellar sclerite, distal shaft (Fig. 6). H: Sh, 8c. 77 All legs. O: femur, distal anterior surface. I: patellar sclerite, anterior arm (Fig. 6). H: Sh, 8a; Sn, 8. 78 All legs. O: femur, distal posterior surface. I: patellar sclerite, posterior arm (Fig. 6). Sh: 8b, Sn: 8. 79 All legs. O: femur, distal dorsal surface, and patella, anterior and anteroventral surfaces. I: tibia, ventral margin (Fig. 6). H: Sh, 6; Sn, 16. 80 All legs. O: femur, distal dorsal surface, and patella, posterior and posteroventral surfaces. I: tibia, ventral margin (Fig. 6). H: Sh, 7; Sn, 17. 81 All legs. O: patella, anterodorsal surface. I: tibia, anterodorsal process (Fig. 6). H: Sh, 4a; Sn, 12. 82 All legs. O: patella, posterodorsal surface. I: tibia, posterodorsal process (Fig. 6). H: Sh, 5a; Sn, 13.- 83 All legs. O: patella, anterior surface. I: tibia, anterior margin (Fig. 6). H: Sh, 4b; Sn, 14. 84 All legs. O: patella, posterior surface. I: tibia, posterior proximal margin (Fig. 6). H: Sh, 5b; Sn, 15. SHULTZ— MUSCLES OF LIMULUS 289 Table 1. — Continued. 85 Leg 5. O: tibia, anterior surface. I: tarsus, anterior margin (Fig. 6). H: Sh, 3b; Sn 18. 86 Leg 5. O: tibia, anterior and posterior surfaces. I: tarsus, posterior margin (Fig. 6). H: Sh, 3a; Sn 19. 87 All legs. Legs 1-4: O: tibiotarsus, dorsal surface (not illustrated). I: apotele, dorsal margin. Leg 5: O: tarsus, dorsal surface. I: apotele, dorsal margin (Fig. 6). H: Sh, 1; Sn, 20. 88 All legs. Legs 1-4: O: tibiotarsus. I: apotele, ventral margin (not illustrated). Leg 5: O: tarsus, ventral surface. I: apotele, ventral margin (Fig. 6). H: Sh, 2; Sn, 21. Chilarial muscles 89 Long, thin. O: carapace, medial posterior region. I: chilarium, lateral flange (Figs 1, 4). 90 O: endosternite, posterior margin. I: chilarium, lateral flange (Fig. 4). 91 O: endosternite, posterior ventral surface near attachment of chilarial chondrite. I: chilarium, an- teromedial margin (Figs 2, 4). 92 O: subendostemal subneural plastron, posterior margin. I: chilarium, medial surface (Figs 2, 4). 93 O: subendostemal subneural plastron, posterior margin. I: chilarium, lateral surface (Fig. 4). 94 Transverse. O: right chilarium at base of chondrite. I: left chilarium at base of chondrite (Figs 2, 4). Opercular muscles 95 Genital operculum only. 0: carapace, posteromedial surface (Fig. 1). I: chondrite, dorsal surface near attachment to 1st dorsal entapophysis (not illustrated). 96 Postgenital opercula, 96viii-96xiii. O: posteromedial surface of carapace (96viii) (Fig. 1) or posterior margin of dorsal entapophysis (96ix-96xni)- h anterior margin of operculum of posteriorly adjacent somite near 100 (Fig. 8). H: LBB, 21, 22 + 23?; Sn, pmcl. 97 Opercula, 97viir-97xin. O: chondrite, anterior surface. I: operculum, anterior surface with 96 (Fig. 8). 98 Opercula, 98vin-98xiii. O: tergum, near dorsal entapophysis (Figs 1, 8). I: operculum, anterior surface proximal to transverse ridge (Fig. 8). H: LBB, 20; Sn, rmcl. 99 Opercula, 99vin-99xin. O: chondrite, posteromedial surface. 1. operculum, anterior surface medial to longitudinal ridge (Fig. 8). 100 Opercula, lOOvin-lOOxin. O: dorsomedial surface of chondrite and adjacent regions of tergum; tergal part smaller posteriorly (Figs 1, 8). I: operculum, anterior surface distal and lateral to intersection of longitudinal and transverse ridges (Fig. 8). H: LBB, 20, 113. 101 Opercula, lOlvm-lOlxin. O: carapace (lOlym) (Fig. 1) or dorsal entapophysis (IX-XIII). I: anterior surface of operculum of posteriorly adjacent somite (Fig. 8). H: LBB, 21-23? 102 Opercula, 102vin-102xni- O: subneural plastron, ventral. I: chondrite, lateral face (Fig. 8). H: LBB, 48? 103 Opercula, 103viii-103xin. O: subneural plastron, ventral surface. I: basal posterior face of chondrite and/or adjacent region of operculum (Fig. 8). H: LBB, 115. 104 Opercula, 104vin-104xni- O: chondrite, posterior surface. I: exopod, medial margin of posterior plate (not illustrated). 105 Opercula, 105viii-105xiii- Many separate fibers. O: anterior surface. I: posterior surface. More de- veloped near lamellae (not illustrated). 106 Opercula, 106viii-106xiii. O: operculum, medial anterior margin. I: telopod (Fig. 8). H: LBB, 114. 107 Opercula, 107vin-107xiii. O: operculum, anterior surface. I: telopod base (Fig. 8). H: LBB, 112. 108 Postgenital opercula, 108ix-108xiii. O: telopod protopodite. I: telopod deutomerite (Fig. 8). H: LBB, 114. 109 Postgenital opercula, 109ix-109xiii. Telopod. O: protomerite, distal margin. I: deutomerite, proximal margin (Fig. 8). 110 Postgenital opercula, 1 lOjx-l lOxm. Telopod. O: deutomerite. I: tritomerite (Fig. 8). 111 Postgenital opercula, 1 1 Ijx-l 1 Ixm- Telopod. O: deutomerite. I: tritomerite (Fig. 8). 112 Opercula, 1 12|x-l 12xi]i. O: telopod, base. I: exopod lobe, proximal margin (Fig. 8). 113 Opercula, 1 13ix-l 13xni- O: operculum, anterior surface. I: exopodial lobe, proximal margin (Fig. 8). 290 THE JOURNAL OF ARACHNOLOGY Figure 1. — Dorsal muscle attachments of an immature Limulus polyphemus. A, Cephalothoracic cara- pace, abdominal tergum and base of the telson. Muscles with attachments depicted in black arise from the endosternite or siibneural plastrons (compare Fig. 7). B, Details of muscle attachments on the right side of the cardiac lobe of the carapace showing quasi-concentric arrangement of attachments of muscle series 26. Arabic numerals correspond to muscles listed in Table 1. Roman numerals correspond to the postoral somite with which the indicated muscle or structure is associated. Abbreviations: af, axial furrow; de, invagination associated with dorsal entapophysis. chondrites (Patten & Redenbaugh 1899) and is probably serially homologous with them. A transverse muscle (94) passes from the base of one chilarial chondrite to the base of the other (Figs. 2, 4). Each chilarium also has two muscles that originate on the subendosternal subneural plastron (92, 93) and two that orig- inate on the endosternite {90, 91) (Figs. 2, 4). A long, thin muscle {89) originates on the car- apace among extrinsic muscles of the append- age of somite VI (i.e., leg 5) (Fig. 1), passes ventrally lateral to the axial muscles (26), and inserts on the lateral flange of the chilarium (Fig. 4). Axial muscles. — Endosternite: The endos- ternite of Limulus has been described and il- lustrated especially well by Patten & Reden- baugh (1899) (see also Lankester et al. 1885; Pocock 1902; Snodgrass 1952; Manton 1964; Firstman 1973; Yamasaki et al. 1988) and will SHULTZ— MUSCLES OF LIMULUS 291 carapace tergum gizzard crop pgop post telson fs eh chel Figure 2. — Medial view of mid-sagittal section of an immature Limulus polyphemus. All extrinsic ap- pendicular and axial skeletal muscles, including the endosternal suspensors, have been removed. Arabic numerals correspond to muscles listed in Table 1. Abbreviations: aa, anterior aorta; chel, chelicera; chil, chilarium; e, endostoma; eh, epistomal horn; est, endosternite; fs, frontal sclerite; gop, genital operculum; hng, hinge between carapace and tergum; Ibm, labrum; me, medial eye; mg, midgut with openings to digestive caeca; pgop, postgenital opercula; rctm, rectum; post, postopercular sternum; s, subendosternal subneural plastron; snp, subneural plastron; vs, venous sinus (collapsed). therefore be treated here in a general manner. The endosternite (Figs. 2, 3, 7, 9) is a meso- dermally derived endoskeleton composed of a tough, fibrous connective tissue (Fahrenbach 1999). It is a roughly rectangular, horizontal sheet with a pair of anterior projections, the anterior horns (Figs. 2, 7); a pair of sheetlike posterolateral projections; and a posterome- dial projection. It serves primarily as an at- tachment for extrinsic leg muscles (67-72) (Fig. 3 A; see also Lankester et al. 1885; Man- ton 1964); axial muscles that attach to various sites in the abdomen {14, 19, 22, 23) (Figs. 7, 9B); pharyngeal dilator muscles (7) (Figs. 2, 7), which originate from the concave ventral surface of the endosternite; and the first two “veno-pericardiac” muscles (72) (Fig. 2). It is suspended from the carapace by six paired muscles that attach to dorsal projections which are continuous with the body of the en- dosternite {13) (Figs. 1, 7, 9B). The endosternite is also connected to the body wall by a less well-understood system of ventrolateral muscles (75) and ligaments (76). The dorsal lateral margin of the endosternite is modified into a flexible marginal membrane which extends from the attachment of the first “venopericardiac” muscle (12) posteriorly to the attachment of the second “venopericar- diac” muscle (Fig. 2). Posteriorly, the mem- brane becomes bilayered, merges with a large lateral venous sinus and continues rearward to form the ventral sinus of the abdomen. The floor of the sinus is firmly connected to the abdominal floor and is attached dorsally to the pericardium via the seven remaining venoper- icardiac muscles (Fig. 2: 72). The endosternal attachment points of the first two venoperi- cardiac muscles have membranous ligaments (76) that pass ventrolaterally from the margin- al membrane to the pliable cuticle between the leg coxae. Specifically, there are two liga- ments associated with the anterior “veno-per- icardiac” muscle, one attaching between the appendages of somites III and IV (legs 1 and 2) and the other between appendages of so- mites IV and V (legs 2 and 3) (Fig. 4). The ligament associated with the posterior “veno- pericardiac” muscle is especially well devel- oped and attaches between legs 3 and 4 (Fig. 4). An apparent ligament was observed ex- tending from the lateral surface of the endos- ternal horn and inserting on the intercoxal cu- ticle between legs 1 and 2, but this was not confirmed in all individuals. Muscle 75 arises from the ventral surface of the anterior en- 292 THE JOURNAL OF ARACHNOLOGY Figure 3. — A, Anterior view of cross section through cephalothorax showing arrangement of ex- trinsic muscles of the third leg (after Wyse & Dwyer 1973). B, Dorsal view showing approximate locations of “muscular somites” composing the carapace and tergum. Note the posterior displace- ment of somite I (cheliceral somite) and- that the lateral portions of the hinge between the carapace and tergum are specializations of somite VIII. Ar- abic numerals correspond to muscles listed in Table 1 . Roman numerals correspond to the postoral seg- ment with which the indicated muscle is associated. Abbreviations: ap, apotele or moveable finger of chela; ca, carapace; cx, coxa; e, endostoma; est, en- dostemite; fe, femur; men, moveable endite; mf, marginal fold; le, lateral eye; pa, patella; pp, pivot point between ventral surface of marginal fold and coxa; tita, tibiotarsus; tr, trochanter. dosternal horn and inserts on the epistomal horn that is embedded in the flexible cuticle between the chelicera and leg 1. Muscle 15 was mistaken for an extrinsic cheliceral mus- cle by Lankester et al. (1885), Subneural plastrons: The subneural plas- trons are metamerically arranged endoskeletal elements composed of material similar to that of the endostemite (Patten & Redenbaugh 1899), Despite apparent similarities in com- position,- it is unlikely that the plastrons and endostemite are serial homologs, because the endostemal element and subneural plastron of somite VII are both present (Figs. 2, 9B), and the central nervous system passes ventral to the endostemite and dorsal to the plastrons. The anteriormost subneural plastron is sus- pended from the ventral surface of the endos- temite by processes of connective tissue (Figs, 2, 4: s) and is an attachment site for muscles of the preoral chamber and chilaria (Figs. 2, 4: 6, 92, 93). It appears to be associated with somite VII, the chilarial somite. The remain- ing six subneural plastrons are located on the floor of the abdomen (Figs. 2, 7, 8: snp) and appear to belong to somites VIII to XIII (Figs. 2, 7, 9B). Each abdominal plastron spans the crests of two folds in the pliable cuticle of the ventral body wall (Fig. 2) and thereby forms a series of transverse “tunnels.” Muscles {102, 103.) arising- from the opercula pass me- dially through these tunnels to insert on the ventral surface of the plastron (Fig. 8). A bi- lateral pair of connective-tissue processes pro- ject dorsally from each plastron and serves as attachment sites for a variety of axial muscles that are described' in detail below. Axial muscles of the abdomen: A notable incongruity in Limulus is a complex axial muscle system in the abdomen (Figs. 7, 9B), a tagma that lacks dorsal mobility between its constituent somites. The system is apparently used in flexing the dorsal hinge between the carapace and tergum during defensive “en- rollment” and in' moving the opercula. Lan- kester et al. (1885) provided a description of the axial muscle system, but it is confusing, im.precise and- s'ometimes incorrect. The axial muscles- are here categorized into four groups, 1 )■ the dorsal longitudinal complex, 2) ventral longitudinal complex, 3) the posterior oblique complex, and 4) the anterior oblique complex. Muscles in the dorsal longitudinal complex pass from one dorsal cuticular attachment to another. Seven muscles span the dorsal hinge between the carapace and tergum. Three paired muscles insert on the pliable fold of the dorsal hinge, two originating from the cardiac lobe of the carapace (Figs. 1, 7B: 27, 28) and SHULTZ— MUSCLES OF LIMULUS 293 chil Figure 4. — Semi-diagrammatic dorsal view of the ventral surface of the cephalothorax showing the preoral apparatus. The esophagus has been cut and reflected posteriorly to better show the arrangement of the skeletomuscular elements (compare Fig. 2). Note also that the inter-coxal bands (icb) would not normally be visible dorsally in the region of the preoral sphincter (5), because they are covered by muscle fibers that pass from each band to bands that are not directly adjacent. Note that dilator muscles (65) pass from the leg coxae and attach to the walls of the preoral chamber; their site of attachment may represent the position of the true mouth. Arabic numerals correspond to muscles listed in Table 1. Roman numerals correspond to the postoral somite with which the indicated muscle or structure is associated. Abbreviations: al5, attachment of muscle 15; al6,i,-al6vi, attachments of 16,ji-16vi; chel, attachment site of chelicera; chf, lateral flange of chilarium; chil, chilaria; cht, base of chilarial chondrite; e, endostoma; eh, epistomal horn; fs, frontal sclerite; icb, inter-coxal band; s, subendosternal subneural plastron with tendinous attach- ments to the endosternite cut (compare Fig. 2). one from the first dorsal entapophysis (Fig. 7B: 31). The remaining four pairs originate on the posterior -margin of the first dorsal enta- pophysis and insert sequentially on the next four pairs of dorsal entapophyses (Fig. 7B: 30ix—30xii)- Muscles of the ventral longitudinal complex pass from one ventral attachment to another, either a cuticular structure (ventral entapo- physis, postopercular sternum) or an endo- skeletal structure (endosternite or subneural plastron) (Fig, 7). The complex can be divided into three bilateral longitudinal series: a me- dial series, a middle series, and a lateral series. Members of the medial series are thin, strap- like muscles that originate anteriorly on one endoskeletal element and pass posteriorly to insert on another (Fig. 7C: 19, 20). In contrast, the middle series {23) is best described as a collection of parallel muscle fibers with dif- ferent posterior attachments. The fibers origi- nate on the dorsal posterior surface of the en- dosternite and pass posteromedially to insert on the subneural plastrons and anterior margin 294 THE JOURNAL OF ARACHNOLOGY of the postopercular sternum, with those fibers inserting more anteriorly originating more me- dially on the endosternite (Fig. 7C: 23). The lateral series (24) is also composed of “fiber tracts” rather than distinct muscles. In fact, the fibers of the middle series intermingle with those of the lateral series, but the two groups can be distinguished by tracing fibers to their respective attachments. Fibers in the lateral series originate on the postopercular sternum and insert on the cuticular folds of the ventral body wall just medial to the ventral entapo- physes of somites IX to XIII (Fig. 7 A: 24). Muscles of the posterior oblique complex attach dorsally to some element of the dorsal body wall (carapace, tergum, dorsal entapo- physis) and ventrally to the ventral body wall or to an endoskeletal structure. This complex can also be divided into medial, middle and lateral series, and these appear to be linked morphologically to the three series of the ven- tral longitudinal complex. The medial series (14?, 17, 21) consists of strap-like muscles that originate on the endosternite or a sub- neural plastron and insert on the tergum (Figs. 1, 7C: /7) or a dorsal entapophysis (Fig. 7C: 14, 21). Members of the middle series (22) arise from the dorsal surface of the endoster- nite and pass dorsoposteriorly to insert on each of the dorsal entapophyses (Fig. 7A). Like the corresponding series in the ventral longitudinal complex, those muscles inserting more anteriorly originate more medially on the endosternite. The lateral series is com- posed of small muscles (25) that originate from the folds of the ventral body wall near the ventral entapophyses along with the fibers of the lateral series of the ventral longitudinal complex (Fig. 7A). Muscles of the anterior oblique complex (2(5vni-2bxiv) originate on the carapace at the extreme lateral portion of the cardiac lobe (Fig. 1), and, in one case (26xiv), on the sec- ond dorsal entapophysis (Fig. 1C). The first member of this complex (26vin is a small, thin muscle that inserts on a small infolding, or ventral entapophysis, on the ventral body wall at the attachment of the posterior margin of the endosternite with pliable ventral cuticle anterior to the genital operculum (Fig. 7A: vevni)- The remaining muscles originate in a quasi-concentric pattern, with those having a more anterior insertion originating nearer to the center of the origin and muscles with more posterior insertions originating more periph- erally (Fig. 1). These muscles insert at the ends of long, hollow tendons which are ex- tensions of infoldings (ventral entapophyses) of the pliable cuticle between adjacent oper- cula (Fig. 7). Feeding apparatus. — The basic anatomy of the gnathobasic feeding apparatus and di- gestive tract of Limulus and other xiphosurans is well known and has been described in con- siderable detail by previous workers (Lankes- ter et al. 1885; Manton 1964; Wyse & Dwyer 1973; Clarke 1979; Yamasaki et al. 1988; Fah- renbach 1999; etc.). Consequently, muscles associated with the feeding and digestive sys- tems have been listed and briefly described in Table 1 but are not described in detail here. However, several features of the anterior di- gestive tract have been overlooked or inade- quately described by previous workers, and these are treated in more detail. Preoral apparatus: The preoral chamber is shaped like an inverted funnel (Figs. 2, 4) sur- rounded by the leg coxae (Figs. 3, 4). Its walls are formed by lobes of pliable cuticle, with the moveable and fixed endites of the leg cox- ae projecting between them (see Manton 1964: figs. 14, 16). The furrows between the lobes are continuous laterally with those formed by the flexible inter-coxal cuticle. The anterior wall of the chamber is an unpaired lobe that is continuous with the labrum and epistome (Fig. 2), and the posterior wall con- sists of an oblong plate, the endostoma, com- posed of stiffer but still flexible cuticle (Figs. 2-4). The chamber narrows as it passes deep- er into the body and bends anteriorly to pass through the brain and to become the ‘esoph- ageal’ region of the foregut (Fig. 2). There is no gross cuticular feature demarcating the “true mouth,” that is, the junction of the preoral chamber and foregut. The musculature of the preoral chamber is described in detail here for the first time. The inner surface of the walls of the preoral cham- ber are surrounded by a roughly circular meshwork of muscle fibers (5) that forms a large sphincter (Figs. 2, 4). The sphincter has a radially arranged ‘skeleton’ formed by strips of connective tissue that begin laterally on the pliable inter-coxal cuticle medially adjacent to the attachments of the endosternite-intercoxal muscles (21) and pass centripedally along the ventral body surface onto the walls of the SHULTZ— MUSCLES OF LIMULUS 295 preoral chamber (Fig. 5). These bands occupy the internal crests of the inter-coxal folds be- tween the lobes of the preoral chamber and may act to maintain the shape of the preoral chamber. Muscle fibers (5) arise from each cartilage-like band and associated cuticle and pass to adjacent and subadjacent bands. It is noteworthy that the inter-coxal portion of the anteriormost pair of bands appears to have been modified, or replaced, by the epistomal horns, which pass between the chelicerae and the coxae of the first leg pair (Figs. 2, 4). In fact, like the connective tissue bands, each ep- istomal horn is associated with a muscle (20) that arises from the endosternite (Figs. 2, 4). The epistomal horns wrap around the poste- rior margin of the chelicerae and join the body of the epistome, whereupon the sclerite as- sumes the appearance of a connective tissue band, gives rise to muscle fibers (5), and de- fines the lateral margins of the labrum or an- terior lobe of the complex (Fig. 4). Apparent dilator muscles (73) arise from the postero- medial coxal margins of legs 1-4 and pass centripedally to insert in a ring around the deeper, narrower region of the preoral cham- ber (Fig. 4). The attachment of these muscles separates the roughly circular fibers of the preoral sphincter from those of the foregut (8) (Figs. 2, 3) and may indicate the site of the true mouth. Foregut: The foregut is the cuticle-lined re- gion of the digestive tract that connects the preoral chamber and midgut (Figs. 2C, 3). The foregut is essentially C-shaped in lateral per- spective (Fig. 2). The narrow “esophageal” portion of the foregut passes anteriorly through the brain. The longitudinally folded lumen is surrounded by circular constrictor muscles (8) and is supplied with one well-de- veloped dilator muscle arising from the en- dosternite (Figs. 2, 7: 7). A second dilator de- scribed and illustrated by Lankester et al. (1885) was never observed and was probably based on a misinterpretation of dorsoventral muscles that arise form the anteroventral sur- face of the cephalothorax, pass dorsally along the lateral surfaces of the crop, but continue dorsally to insert on the carapace (Fig. 2: 2). The foregut then expands, both in the diam- eter of the lumen and thickness of circular muscles, and turns dorsally and then posteri- orly (Figs. 2, 5C). The lumen walls become progressively more heavily sclerotized as the foregut approaches the midgut, and the last portion (proventriculus or gizzard) is appar- ently specialized for grinding. The foregut is separated from the midgut by a strongly de- veloped valve (Fig. 2). DISCUSSION Evolutionary morphology of axial mus- cles in Chelicerata. — The box-truss axial muscle system: Comparative anatomical stud- ies of crustaceans (e.g., Cephalocarida, Ma- lacostraca, Mystacocarida, Branchiopoda: Hessler 1964), myriapods (e.g., Pauropoda: Verhoeff 1934, Tiegs 1947), hexapods (Diplu- ra: Manton 1972; Microcoryphia: Birket- Smith 1974) and, perhaps, trilobites (Cisne 1981) have revealed a common box-truss ax- ial muscle system (Fig. 9A). This system con- sists of bilateral dorsal and ventral longitudi- nal elements that attach to each somite, a bilateral set of dorsoventral muscles within each somite that passes from the tergite to the ventral body wall, a bilateral set of posterior oblique elements that arises ventrally in as- sociation with dorsoventral elements and in- serts dorsally on a more posterior somite, and a bilateral pair of anterior oblique elements that also arises ventrally with a dorsoventral elements but inserts dorsally on a more ante- rior somite. The ventral longitudinal, dorso- ventral, anterior oblique and posterior oblique elements all attach to a transverse endoskele- tal bar within each somite (Fig. 9A). Axial muscles of arachnids appear to correspond to elements of the box-truss system (Fig. 9C); that is, the dorsal longitudinal muscles and en- dosternite plus ventral longitudinal muscles correspond to the dorsal longitudinal and ven- tral longitudinal elements, respectively; dorsal endosternal suspensors and dorsoventral mus- cles of the opisthosoma appear to correspond to the dorsoventral elements; and the dorso- posterior endosternal suspensors and, perhaps, “transverse” muscles of the opisthosoma (e.g., Amblypygi: Shultz 1999: muscle 22; Ar- aneae: Whitehead & Rempel 1959: muscle 89; Scorpiones: original observation) can be ho- mologized with posterior oblique elements. No muscles corresponding to the anterior oblique elements are known in arachnids (Fig. 9C). However, it is unclear from these com- parisons whether the relative simplicity of arachnids is a primitive antecedent of the box- 296 THE JOURNAL OF ARACHNOLOGY Figure 5. — Skeletomuscular anatomy of the che- licerae. A, Medial view of the right chelicera show- ing insertions of extrinsic muscles (45-48) and proximal intrinsic muscles (49, 50). B, Distal arti- cles of the right chelicera showing intrinsic muscles that operate the chela (51, 52). The figure shows a dorsal view of a fully flexed chelicera. Numbers correspond to muscles listed in Table 1. Abbrevia- tions: pt, protomerite or first cheliceral article; dt, deutomerite or second cheliceral article; tt, trito- merite or third cheliceral article. truss system or a derived reduction of the box- truss system. Based on information obtained in the pre- sent study, I propose that the abdominal axial muscle system in Limulus retains all essential components of the box-truss system, including the anterior oblique elements (Fig. 9B), and that the box-truss system is the plesiomorphic condition in Chelicerata. Specifically, the dor- sal longitudinal elements of the box-truss sys- tem are retained as muscles 27-30 and per- haps 31-35 (Fig. 7); the ventral longitudinal elements are retained as muscles 19, 20, 23 and 24 (Fig. 7); the dorsoventral elements are retained as muscles 13 and 77 (Fig. 7); the posterior oblique muscles are retained as mus- cles 21, 22 and perhaps 25; and the anterior oblique muscles are retained as muscle series 26. Most of these comparisons are probably non-controversial, except for the anterior oblique muscles (26). Specifically, all relevant muscles in Limulus other than 26 are already accepted as axial muscles, and the necessary evolutionary transformation of a portion of the posterior oblique elements into muscle se- ries 22 (i.e., anterior migration of the ventral attachments from each abdominal somite to the dorsal surface of the endosternite) has been documented in ontogeny (see Scholl 1977: Figs. 3, 5). However, muscle series 26 is generally considered a group of extrinsic opercular muscles, not axial muscles, and this inconsistency must be addressed. Lankester et al. (1885) and many subse- quent authors have referred to muscle series 26 as “branchio-thoracic” muscles and con- sidered them to be extrinsic opercular mus- cles, although the anteriormost member of this series (26vin) was recognized for the first time in the present study. My conclusion that these are actually axial muscles is based on the fol- lowing lines of evidence: 1) One muscle (26xiv) is associated with a somite that lacks appendages during all stages of development (Scholl 1977). 2) The ventral entapophyses, which give rise to the long tendons on which these muscles insert, are invaginations of the body wall rather than the appendages. This is particularly evident in the ventral entapophys- is associated with 26xiv which is continuous with the postopercular sternum (Fig. 7). 3) The bases of the ventral entapophyses and the adjacent region of the abdominal floor serve as attachment sites for muscles that are clearly axial rather than appendicular in origin (Fig. 7: 24, 25). 4) The muscles insert on the car- diac lobe medial to the axial furrow, not lat- eral to the furrow like all appendicular mus- cles (Fig. 1: 53-57), including those associated with the genital operculum (Fig. 1: 96vni, /67viii). I suggest that members of mus- cle series 26 may have been misinterpreted as extrinsic opercular muscles due to the erro- neous assumption that the postgenital opercula are formed by medial fusion of the paired ap- pendages, an evolutionary process that would be expected to obliterate the ventral body wall between them and its associated axial mus- cles. In fact, however, the abdominal floor is present externally between each postgenital operculum as a triangular fold (e.g., Snodrass 1952) and internally serves as an attachment site for the unambiguous axial muscles de- scribed above. Given these arguments, I hypothesize that Limulus retains the abdominal elements of a box-truss system like that observed in other arthropod groups, although certain elements have been modified (Fig. 9). The principal evolutionary transformations required by this hypothesis are anterior migrations of portions SHULTZ— MUSCLES OF LIMULUS 291 Figure 6. — Muscular anatomy of the intrinsic muscles of leg 5 from anterior and posterior perspectives. Insets show deeper muscles with superficial muscles removed. The lines near the coxae indicate the mid- sagittal plane of the intact animal. Abbreviations: ap, apotele; apr, anterior process of tibia; cx, coxa; fbm, flabellum; fe, femur; pa, patella; ppr, posterior process of tibia; ps, patellar sclerite; ta, tarsus; ti, tibia; tr, trochanter. of the ventral attachments of the posterior oblique elements to the dorsal surface of the endosternite and anterior migration of the dor- sal attachments of the anterior oblique ele- ments to the carapace. These transformations would likely be associated with the evolution- ary elimination of all tergal articulations ex- cept the carapace-tergum hinge (summarized by Anderson & Selden 1997), as they would allow the muscles to retain a function in mov- ing the body. If this scenario is correct, the box-truss axial muscle system should proba- bly be regarded as synapomorphic for Ar- thropoda and plesiomorphic for Chelicerata (Edgecombe et al. 2000), and the losses re- sulting in simplification of the box-truss sys- tem, especially the loss of all anterior oblique elements, would be synapomorphic for Arach- nida. Endosternal evolution in Chelicerata: First- man (1973) proposed a primitive axial muscle system for Chelicerata that was remarkably similar to the box-truss model. He hypothe- sized that the primitive system was ladderlike with a “rung” of connective tissue (transverse endoskeletal bar) positioned transversely within each somite and connected to longitu- dinally adjacent transverse bars by ventral longitudinal muscles. Each transverse bar was also suspended from the exoskeleton by bilat- 298 THE JOURNAL OF ARACHNOLOGY Figure 7. — Dorsal view of the endosternite (est), ventral surface of the abdomen, and the base of the telson. The appendages, carapace, tergum and most other heavily sclerotized structures have been removed, but the dorsal entapophyses (de) have been cut at their attachments to the tergum and are depicted as “floating” at their intact anatomical positions. The ventromedial surface of the abdomen is composed of pliable cuticle aiTanged in transverse folds. A series of subneural plastrons (snp) bridge the crests of a pair of adjacent folds (compare Fig. 2). A, The left side shows the middle tract of the posterior oblique muscles (22). The right side shows the lateral tract of ventral longitudinal muscles (24) and posterior oblique muscles (25). Only one member of the anterior oblique tract (26xi) is shown, the others have been cut near their ventral attachments. The posterior end shows the arrangement of muscles that insert on the ventrolateral processes of the telson (39-44). B, Same perspective as in A, but the medial invaginated portion of the carapace-tergum hinge (if) “floats” above the other elements to show the attachments of muscles 27-29. The left side shows the arrangement of the posteriormost anterior oblique muscle (26xiv), the right side shows muscles that span the dorsal entapophyses (30), and the posterior end shows the arrangement of muscles that insert on the dorsal process of the telson (31-38). C, Same perspective showing the position of the gut and its muscles (7-10). The crop and rectum have been cut horizontally and the intermediate portions of the gut have been removed to show the underlying endoskeleton and muscles (compare Fig. 2). The right side depicts the dorsoventral muscles (17) and the medial tracts of ventral longitudinal muscles (19, 20) and posterior oblique muscles (21). The left side shows the middle tract of ventral longitudinal muscles (23) and the dorsal interconnections between opercular chondrites (diac). Arabic numerals correspond to muscles listed in Table 1. Roman numerals correspond to the postoral somite with which the indicated muscle or structure is associated. Abbreviations: de, dorsal SHULTZ— MUSCLES OF LIMULUS 299 eral dorsoventral suspensor muscles and from the lateral exoskeleton by transverse suspen- sor muscles. As in the box-truss model, the endosternite v/ould have evolved by fusion of the transverse bars and the tendinified longi- tudinal muscles of the first seven postoral so- mites. Thus, Firstman’s model departed from the box-truss model only in predicting a series of intra- segmental transverse muscles in chel- icerates rather than intersegmental posterior oblique muscles and, apparently, in regarding the “braechio-thoracic” muscles (26) as ex- trinsic appendicular muscles rather than axial muscles. Upon applying his model to the axial sys- tem in Limulus, Firstman (1973) concluded that the endosternite retained six pairs of en- dostereal suspensor muscles, namely, the dor- sal, transverse and ventral suspensors of so- mites III, the dorsal and transverse suspensors of somite IV, and the dorsal suspensors of so- mite V. However, observations from the pre- sent study are not consistent with Firstman’s (1973) interpretation. First, he overlooked or omitted several endosternal and axial muscles in Limulus, specifically, one dorsal suspensor (iJj) (Figs. 1, 7), all axial muscles that arise from, the endosternite and insert on more pos- terior structures (Fig. 7: 14, 19, 20, 22-24), and the anterior oblique muscles (26) (Figs. 1, 7). Second, Firstman implied in his figures that transverse suspensors were present in the abdominal somites of Limulus, although he recognized in the text that these muscles (probably 102 and 103: Fig. 8) inserted on the opercula rather than the lateral body wall and doubted their homology with the transverse suspensors of the endosternite. Third, he did not provide specific criteria for assigning mus- cles to particular somites (e.g., position with respect to other muscles). Fourth, he appealed to assign suspensor muscles to the “dorsal” and “transverse” series based on whether the muscle had a dorsoventral or transverse ori- entation rather than some more precise crite- rion, such as placement with respect to other muscles that could be assigned unambiguous- ly to specific somites. [Recent studies of 98 99 100 106 113 112 Figure 8. — Dorsal view of the first postgenital operculum with the posterior surfaces and respira- tory lamellae removed. The appendage is shown in its fully retracted position, so the inner surface of the anterior surface faces dorsally. The subneural plastron (snp) of somite VIII is the only other struc- ture depicted; the pliable ventral cuticle in which the opercula are embedded has been removed. The anatomical relationship between the appendage and the other abdominal structures can be envisioned by superimposing this figure on appropriate elements of Fig. 7. Numbers correspond to muscles listed in Table 1. Abbreviations: bchdt, site where the base of the appendicular chondrite attaches to the ante- rior plate; chdt, dorsal terminus and shaft of chon- drite; ep, inner surface of the anterior plate; epl, exopodial lobe; Ir, longitudinal ridge; snp, subneural plastron of somite IX; tp, telopod; tr, transverse ridge. arachnid anatomy suggest that muscles First- man homologized as “transverse suspensor muscles” represent different kinds of muscles in different chelicerate taxa, such as ventral suspensors in scorpions (pers. obs.) and pos- terior oblique muscles in Pedipalpi (Shultz 1993, 1999)]. In contrast, current evidence from Limulus suggests that the endosternal suspensors ac- knowledged by Firstman are members of a single metameric series representing somites II through VI. This is consistent with embry- ological evidence (Scholl 1977) and with the pattern of suspensor insertions on the cara- pace, which shov/s one suspensor associated with each set of extrinsic appendicular mus- cles (Fig. 1: iJn-iJvi)- Firstman apparently overlooked one dorsal suspensor muscle that appendicular chondrites; dprt, dorsal process of telson; esph, esophageal portion of foregut; est, endoster- nite; if, intersegmental fold of carapace-tergum hinge; post, postopercular sternum; snp, subneural plastron; ve, ventral entapophysis; vlprt, ventrolateral process of telson. 300 THE JOURNAL OF ARACHNOLOGY Somites I II III IV V VI VII VIII IX X XI XII XIII XIV XV I [ 1 1 1 1 [ 1 1 r— 1 1 1 [ r — i Figure 9. — Diagrammatic medial views of hypothetical axial muscle systems in chelicerates showing how the box-truss axial muscle system of other arthropods may have been modified in Limulus and arachnids. A, Hypothetical ancestral chelicerate condition showing primitive box-truss axial-muscle system and the chelicerate endosternite. B, Anangement of axial muscles in Limulus labeled to show proposed homologies with the box-truss axial-muscle system in “A.” Note simplification of prosomal elements and anterior displacement of dorsal attachments of aom and ventral attachments of pom. The dotted lines at the posterior indicate muscles present in the larva but not in adult. C, Hypothetical ancestral arachnid condition showing loss or modification of anterior elements, loss of aom in all somites, and displacement of opisthosomal pom attachments from tergites to pleural regions. Abbreviations: aom, anterior oblique muscle; dvm, dorsoventral muscle; est, endosternite; pom, posterior oblique muscle; teb, transverse en- doskeletal bar; vim, ventral longitudinal muscle. appears to be associated with the cheliceral somite (Figs. 1,7: 13^). This muscle arises and inserts more posteriorly than would be ex- pected from its metameric position due to pos- terior migration of the entire cheliceral somite during development (Scholl 1977) (Fig. 3). With the recognition of iJj, Limulus appears to have a single metameric series of six dorsal endosternal suspensors {13y-13y^, one for each of the six original prosomal somites. When these conclusions are interpreted in the framework of the box-truss model, several evolutionary and phylogenetically relevant in- sights emerge. First, anterior oblique muscles are absent from the endosternite in both xiph- osurans and arachnids, and this might repre- sent a synapomorphy for these two lineages. However, given the extreme reduction of all axial muscles in pycnogonids (Firstman 1973), it is possible that this feature is syna- pomorphic for Chelicerata. Second, posterior oblique muscles are absent from the first five SHULTZ— MUSCLES OF LIMULUS 301 somites of the eedosternite in Limulus but are present for somites III-V in at least some arachnids (i.e., Araeeae, Amblypygi, Thely= phonida) (Shultz 1991, 1993, 1999). Thus ab- sence of posterior oblique eedosternal sespen- sors in Limulus and all Asian species of horseshoe crabs (Yamasaki et al. 1988) is probably a synapomorphy of extant Xipho- sura. Third, the apparent absence of endoster- nal components ’associated with somites I and II is a possible synapomorphy of Arachnida, given that they are present in Limulus. It should be noted, however, that these eedos- temal components may have been retained but modified and incorporated into the epiphar- yngeal complex of arachnids in a variety of ways (Shultz 1993, 2000). Fourth, the pres- ence of a postcerebral pharynx supplied with dilator muscles that arise from the eedoster- nite may be a plesiomorphic condition for Arachnida rather than a derived feature of Ar- aneae and Amblypygi alone as is widely thought (e.g., Wheeler & Hayashi 1998). A muscularized postcerebral pharynx is clearly present in Limulus (Figs. 2, 7C) and has been reported but not confirmed in a palpigrade by Rucker (1901). An unmuscularized postcere- bral pharynx (cuticle-lined “esophagus'’ only) and eedosternal foramen instead of pharyn- geal dilator muscles are present in Uropygi (Millot 1949; Shultz 1993). Moeophyly of Arachnida.- — The vast ma- jority of phylogenetic analyses of Chelicerata have concluded that Xiphosura and Arachnida are sister groups among extant chelicerates and that each group is moeophyletic. This conclusion is strongly and consistently sup- ported by phylogenetic analyses of morpho- logical evidence (Weygoldt & Paulus 1979; Shultz 1990; Edgecombe et al. 2000), molec- ular evidence (Regier & Shultz 1997, 1998; Shultz & Regier 2000; but see Colgan et al. 1998; Giribet & Ribera 2000) and combined evidence (Wheeler & Hayashi 1998; but see Edgecombe et al. 2000). Morphological syn- apomorphies supporting the m.onophyly of Arachnida include 1) reduced pleural fold (doublure) in the prosomal carapace (Shultz 1990); 2) slit sensilla (Weygoldt & Paulus 1979; Shultz 1990); 3) anterodorsal rotation of anterior prosoma resulting in anteroven- trally directed mouth (Weygoldt 1979); 4) ab- sence of appendages on somite VII in adults (Shultz 1990); 5) absence of cardiac lobe or glabella on carapace (original observation); 6) single medial genital opening rather than bi- laterally paired genital openings (paired gen- ital openings in all extant xiphosurans: Ya- masaki et al. 1988; single median opening in arachnids: Clarke 1979, original observa- tions); 7) absence of appendages on somite XIII (Shultz 1990; but see Dunlop 1998); 8) loss or reduction of postcerebral crop and pro- ventriculus (present study); 9) absence of an- terior oblique axial muscles (present study); 10) pleural rather than tergal attachments of opisthosomal posterior oblique axial muscles (present study); and 11) eedosternal suspen- sors of somites I and II absent or detached from the endosternite (present study). How- ever, alternative phylogenetic systems have been proposed. Van der Hammen (1985, 1989) suggested that Xiphosura, Scorpiones and Opiliones be placed within a clade (My- liosomata) based on presence of a “coxister- nal” feeding apparatus, a feature that actually appears to be primitive for all extant arthro- pods, including myriapods and hexapods (Po- padic et al. 1998; Scholtz, Mittmann & Ger- berdieg 1998). Dunlop (1998) suggested that scorpions are more closely related to xipho- surans than to tetrapulmoeate arachnids (Ar- aneae, Amblypygi, Uropygi), because the te- trapulmoeates retain primitive lamellate respiratory structures on the genital somite but scorpions and xiphosurans have lost them (see Weygoldt 1998 for an alternative view). Each dissenting view is derived from interpretation of a single character system and the devalua- tion or dismissal of all characters that do not support its conclusions. Given the explicit enumeration of arachnid synapomorphies of- fered here and elsewhere, workers who main- tain that xiphosurans be placed among arach- nids should provide explicit justification for their position and specific reasons for reject- ing the accumulating evidence that excludes xiphosurans from Arachnida. ACKNOWLEDGMENTS This work was supported by the Maryland Agricultural Experiment Station and the Na- tional Science Foundation (Grant DEB- 9615526). LITERATURE CITED Anderson, L.I. & P.A. Selden. 1997. Opisthosomal fusion and phylogeny of Palaeozoic Xiphosura. Lethaia 30:19-31. 302 THE JOURNAL OF ARACHNOLOGY Bonaventura, J., C. Bonaventura & S. Tesh, eds. 1982. Physiology and Biology of Horseshoe Crabs: Studies on Normal and Environmentally Stressed Crabs. A.R. Liss, New York. Birkett-Smith, S.J.R. 1974. On the abdominal mor- phology of Thysanura (Archaeognata and Thy- sanura s. str.). Entomologica Scandinavica Sup- plement 6:1-67. Cisne, J.L. 1981. Triarthrus eatoni (Trilobita): Anatomy of its exoskeletal, skeletomuscular, and digestive systems. Palaeontographica Americana 9(53):1-142. Clarke, K.U. 1979. Visceral anatomy and arthro- pod phylogeny. Pp. 467-549. In Arthropod Phy- togeny. (A.P Gupta, ed.). Van Nostrand Rein- hold, New York. Cohen, E., (ed.) 1979. Biomedical Application of the Horseshoe Crab (Limulidae). A.R. Liss, New York. Colgan, D.J., A. McLauchlan, G.D.E Wilson, S.R Livingston, G.D. Edgecombe, J. Macaranas, G. Cassis & M.R. Gray. 1998. Histone H3 and U2 snRNA DNA sequences and arthropod molecular evolution. Australian Journal of Zoology 46: 419-437. Damen, W.G.M., M. Hausdorf, E-A. Seyfarth & D. Tautz. 1998. A conserved mode of head seg- mentation in arthropods revealed by the expres- sion pattern of Hox genes in a spider. Proceed- ings of the National Academy of Sciences USA 95:10665-10670. Dunlop, J.A. 1998. The origins of tetrapulmonate book lungs and their significance for chelicerate phylogeny. Pp. 9-16. In Proceedings of the 17th European Colloquium of Arachnology, Edin- burgh 1997. (PA. Selden, ed.). British Arach- nological Society, Burnham Beeches, Buckingh- amshire, England. Edgecombe, G.D., G.D.E Wilson, D.J. Colgan, M.R. Gray & G. Cassis. 2000. Arthropod cla- distics: Combined analysis of histone H3 and U2 snRNA sequences and moiphology. Cladistics 16:155-203. Page, L. 1949. Classe des Merostomaces. Pp. 219- 262. In Traite de Zoologie, Vol. 6. (P.-P. Grasse, ed.). Masson, Paris. Fahrenbach, W.H. 1999. Merostomata. Pp. 21- 115. In Microscopic Anatomy of Invertebrates, Vol. 8A. Chelicerate Arthropoda. (EW. Harrison & R.E Foelix, eds.). Wiley-Liss, New York. Firstman, B. 1973. The relationship of the cheli- cerate arterial system to the evolution of the en- dosternite. Journal of Arachnology 1:1-54. Giribet, G. & C. Ribera. 2000. A review of arthro- pod phylogeny: New data based on ribosomal DNA sequences and direct character optimiza- tion. Cladistics 16:204-231. Hammen, L. van der. 1985. Functional morphology and affinities of extant Chelicerata in evolution- ary perspective. Transactions of the Royal Soci- ety of Edinburgh: Earth Sciences 76:137-146. Hammen, L. van der. 1989. An Introduction to Comparative Arachnology. SPB Academic Pub- lishing, The Hague. Hessler, R.R. 1964. The Cephalocarida: Compara- tive skeletomusculature. Memoirs of the Con- necticut Academy of Arts & Sciences 16:1-97. Iwanoff, PR 1933. Die embryonale Entwicklung von Limulus moluccanus . Zoologische Jahrbiich- er Anatomie 56:163-348. Lankester, E.R. 1881. Limulus an arachnid. Quar- terly Journal of Microscopical Science, New Se- ries 21:504-548, 609-649. Lankester, E.R. 1885. A new hypothesis as to the relationship of the lung-book of Scorpio to the gill-book of Limulus. Quarterly Journal of Mi- croscopical Science 25:339-342. Lankester, E.R., W.B.S. Benham & E.J. Beck. 1885. On the muscular and endoskeletal systems of Limulus and Scorpio’, with some notes on the anatomy and generic characters of scorpions. Transactions of the Zoological Society of Lon- don 11:311-384. Manton, S.M. 1964. Mandibular mechanisms and the evolution of arthropods. Philosophical Trans- actions of the Royal Society of London B 7(737): 1-183. Manton, S.M. 1972. The evolution of arthropodan locomotory mechanisms. Part 10. Locomotory habits, morphology and evolution of the hexapod classes. Journal of the Linnean Society (Zoolo- gy) 51:203-400. Millot, J. 1949. Ordre des Uropyges. Pp. 533-562. In Traite de Zoologie, Vol. 6. (P.P. Grasse, ed.). Masson, Paris. Milne-Edwards, A. 1873. Recherches sur Fanatomie des Limules. Annales des Sciences Naturelles. Zoologie, 5th Series 17:1-67. Owen, R. 1873. On the anatomy of the American king crab {Limulus polyphemus Latr.). Transac- tions of the Linnean Society of London 28:459- 506. Patten, W 1893. On the morphology and physiol- ogy of the brain and sense organs of Limulus. Quarterly Journal of Microscopical Science 35: 1-96. Patten, W 1912. The Evolution of the Vertebrates and their Kin. P. Blakiston’s Son and Co., Phil- adelphia. Patten, W. & A.P. Hazen. 1900. The development of the coxal gland, branchial cartilages, and gen- ital ducts of Limulus polyphemus. Journal of Morphology 16:459-490. Patten, W. & W. A. Redenbaugh. 1889-1900. Stud- ies on Limulus. Journal of Morphology 16:1-26, 91-200. Pocock, R.I. 1902. Studies on the arachnid endos- SHULTZ— MUSCLES OF LIMULUS 303 temite. Quarterly Journal of Microscopical Sci- ence 46:225-262. Popadic, A., G. Paeganibae, D. Rusch, W.A. Shear & TC. Kaufman. 1998. Molecular evidence for the geathobasic derivation of arthropod mandi- bles and for the appendicular origin of the la- brum and other structures. Development Genes and Evolution 208:142-150. Regiep J.C. & J.W. Shultz. 1997. Molecular phy- logeny of the major arthropod groups indicates polyphyly of crustaceans and a new hypothesis for the origin of hexapods. Molecular Biology and Evolution 14:902-913. Regier, J.C. & J.W. Shultz. 1998. Molecular phy- logeny of arthropods and the significance of the Cambrian “explosion” for molecular systemat- ics. American Zoologist 38:918-928. Rucker, A. 1901. The Texan Koenenia. American Naturalist 35:615-630. Scholl, G. 1977. Beitrage zur Embryonalentwick- lung von Limulus polyphemus L. (Chelicerata, Xiphosura). Zoornorphologie 86:99-154. Scholtz, G., B. Mittmann & M. Gerberding. 1998. The pattern of Distal-less expression in the mouthparts of crustaceans, myriapods and in- sects: New evidence for a gnathobasic mandible and the common origin of Mandibulata. Inter- national Journal of Developmental Biology 42: 801-810. Sekiguchi, K. (ed.) 1988. Biology of Horseshoe Crabs. Science House Co., Tokyo. Shultz, J.W. 1989. Morphology of locomotor ap- pendages in Arachnida: Evolutionary trends and phylogenetic implications. Zoological Journal of the Linnean Society 97:1-56. Shultz, J.W. 1990. Evolutionary morphology and phylogeny of Arachnida. Cladistics 6:1-38. Shultz, J.W. 1993. Muscular anatomy of the giant whipscorpion, Mastigoproctus giganteus (Arach- nida, Uropygi), and its evolutionary significance. Zoological Journal of the Linnean Society 108: 335-365. Shultz, J.W. 1999. Muscular anatomy of a whip- spider, Phrynus longipes (Pocock) (Arachnida: Amblypygi), and its evolutionary significance. Zoological Journal of the Linnean Society 126: 81-116. Shultz, J.W. 2000. Skeletomuscular anatomy of the harvestman Leiobunum aldrichi (Weed 1893) (Arachnida: Opiliones) and its evolutionary sig- nificance, Zoological Journal of the Linnean So- ciety 128:401-438. Shultz, J.W. & J.C. Regier. 2000. Phylogenetic analysis of arthropods using two nuclear-protein encoding genes supports a crustacean + hexapod clade. Proceedings of the Royal Society of Lon- don B 267:1011-1019. Snodgrass, R.E. 1952. A Textbook of Arthropod Anatomy. Comstock Publishing Co., Ithaca, New York. Telford, M.J. & R.H. Thomas. 1998. Expression of homeobox genes shows chelicerate arthropods retain their deutocerebral segment. Proceedings of the National Academy of Sciences USA 95: 10671-10675. Tiegs, O.W. 1947. The development and affinities of the Pauropoda, based on a study of Pauropus sylvaticus. Quarterly Journal of Microscopical Science, 3rd Series 88:165-336. Vachon, M. 1945. Remarques sur ies appendices du prosoma des Limules et leur arthrogenese. Ar- chives de Zoologie Experimentale et Generate 84:271-300. Verhoeff, K.W. 1934, Symphyla und Pauropoda. Pp. 1-200, In Bronns Klassen und Ordunugen des Tierreichs, VoL 5. Arthropoda, Part 2. Myr- iapoda. Book 3. (H.G. Bronn, ed.). Akadmische Verlagsgesellschaft m.b.H., Leipzig. Versluys, J. & R. Demoll. 1922. Das Limulus- Problem, Die Verwaedtschaftsbezeihungen der Merostomen und Arachnoiden unter sich und mit anderee Arthropoden. Ergebnisse und Fortschrit- te der Zoologie 5:67-388. Weygoldt, R 1979. Significance of later embryonic stages and head development in arthropod phy- logeny. Pp. 107-135, In Arthropod Phytogeny. (A.P. Gupta, ed.). Van Nostrand Reinhold, New York. Weygoldt, P. 1998. Evolution and systematics of the Chelicerata. Experimental & Applied Aca- rology 22:63-79. Weygoldt, P. & H.F. Paulus. 1979. Uetersuchungen zur Morphologie, Taxonomic und Phylogenie der Chelicerata. Zeitschrift fur Zoologicshe Syste- matik und Evolutionsforschung 17:85-116, 77- 200. Wheeler, W.C. & C.Y Hayashi. 1998. The phylog- eny of the extant chelicerate orders. Cladistics 14:173-192. Whitehead, W.E & J.G. Rempel. 1959. A study of the musculature of the black widow spider, Lat- rodectus mactans (Fabr.). Canadian Journal of Zoology 37:831-870. Wyse, G.A . & N.K. Dwyer. 1973. The neuromus- cular basis of coxal feeding and locomotory movements in Limulus. Biological Bulletin 144: 567-579. Yamasaki, T., T. Makioka & J. Salto. 1988. Mor- phology. Pp. 69—132. In Biology of Horseshoe Crabs. (K. Sekiguchi, ed,). Science House, To- kyo. Manuscript received 20 June 2000, revised 19 March 2001. 2001. The Journal of Arachnology 29:304-311 A NEW SPECIES OF DIPLOCENTRUS (SCORPIONES, DIPLOCENTRIDAE) FROM TEXAS Scott A. Stockwell': Medical Zoology Branch, Department of Preventive Health Services, Academy of Health Sciences, Fort Sam Houston, Texas 78234 USA Andrew S. Baldwin: University of Texas at Arlington, Department of Biology, Box 19498, Arlington, Texas 76019 USA ABSTRACT. Diplocentrus Undo new species, from west Texas, USA, central Nuevo Leon and northern Coahuila, Mexico is described. This description is based on the morphological examination of 199 spec- imens from nine Texas counties and the Mexican states of Coahuila and Nuevo Leon. This species rep- resents the third Diplocentrus known from the state of Texas, has a wider distribution than D. diablo and D. whitei, and it exhibits a marked range in adult size. Keywords: Diplocentrus Undo, new species, scorpion The genus Diplocentrus Peters 1861 is a poorly understood assemblage of burrowing scorpions known primarily from meso-Amer- ica reaching its northern range limit in Ari- zona, New Mexico, and Texas in the United States. Species of Diplocentrus reported from Texas include D. whitei Gervais 1844, D. dia- blo Stockwell & Nilsson 1987, and D. key- serlingii Karsch 1880. Diplocentrus whitei is common in the Big Bend region of Texas and adjacent Mexico (Stockwell & Nilsson 1987; Sissom & Fet 2000). Diplocentrus diablo is known from the lower Rio Grande Valley from Webb County to Starr County (Stockwell & Nilsson 1987). Diplocentrus keyserlingii is found only in Oaxaca, Mexico (Sissom 1994). Records of this species from Texas (Ewing 1928; Gertsch 1939; Rowland & Reddell 1976; Fet et al. 2000) are referable to the new species described below. METHODS The measurements and terminology follow those of Stahnke (1970), except for tricho- bothriotaxy, which follows that of Vachon (1974), metasomal and pedipalpal carination, which follows that of Francke (1978), and hemispermatophore structure, which is modi- fied from Vachon (1952). The measurements reported herein differ from Stahnke (1970) as ^The views of the author do not represent the views of the Department of the Army or the Department of Defense. follows: in measuring the pedipalp chela, the depth is the greatest measurement between the dorsomarginal carina and ventromedian Cari- na, chela width is the narrowest measurement between the digital carina and the interno- median carina, and chela length is measured from the basal-most edge of the external face at the proximal end of the digital or external secondary carina to the distal tip of the fixed finger. All measurements were made to the nearest 0.05 mm using a dissecting micro- scope equipped with an ocular micrometer. Paraxial organs were dissected from males using iris scissors and forceps as described by Lamoral (1979). The hemispermatophores were dissected from the surrounding tissues and observed in 70% ethanol. Letter codes used in the text to indicate the collections from which specimens were ob- tained are given in the acknowledgments. Specimens from the author’s (SAS) collection are listed SAS. Diplocentrus Undo new species Figs. 1-9 Diplocentrus whitei Banks 1900: 424 (in part); Po- cock 1902: 3, 4 (in part); Rowland & Reddell 1976: 5 (in part) (all misidentified). Diplocentrus keyserlingi Ewing 1928: 5, 6; Gertsch 1939: 17; Rowland & Reddell 1976: 5 (all mis- identified). Diplocentrus linda Brown & Formanowicz 1996: 41, 42, 44, 45; Kovarfk 1998: 130; Fet et al. 2000: 597 (NOMEN NUDUM). 304 STOCKWELL & BALDWIN— NEW DIPLOCENTRUS FROM TEXAS 305 Figure 1. — Diplocentrus Undo, adult male paratype (left) from 5 mi. N of Sanderson, Terrell County, Texas, USA, and adult female paratype (right) from 0..5 mi. S of Langtry, Val Verde County, Texas, USA, dorsal aspect. Type data* — Holotype male from 5 miles north of Sandersoe, Terrell 'County, Texas, 15 June 1974 -(Linda Draper, Mont A. Cazier, Os^ car E Francke),, deposited in the- American Museum of Natural History, New York.. Par- atypes are listed under specimens examined. Etymology.— “The specific epithet is Span- ish for “pretty,”' and 'is used as a noun in ap- position. In their paper on reproductive investment in this species, Brown' &■ Fo.rmaeowicz (1996) used the manuscript name, "'Diplocentrus Un- da” believing it to be an available name. Au- tho-rship - was attributed to - Stockwell with no date. The authors did not include a formal de- scription of the species since it was not their intention to-- describe the species as new. The name thus fails to conform to. the requirements of Article 1 3 of the International Code of Zoo- logical Nomenclature, third edition (Interna- tional Co-mmission on Z.ooiogical Nomencla- ture 1985) and is therefore, a nomen nudum by definition (Fet et al. 200-0). Diageosisw- — Diplocentrus Undo- is' distin- guished from its -m-ost morphologically similar congener, D\.diablo Stockwell & Nilsson 1987 from Texas, USA and Tamaulipas, Mexico, by its slightly higher telotarsal spine formula (4/ 5:5/5-6:6/7:6/7 in D. Undo, 4/4:475:5/6:5/6 in D. diablo) and slightly longer fixed chela fin- ger length (chela length/fixed fi.nger length 2.29-2.50 in males and 2.33-2.59 in females in D. Undo, 2.31-2.37 in males and 2.21-2.33 in females in D. diablo). Diplocentrus Undo and D. diablo are also widely allopatric in Texas. It differs from D. colwelli Sissom 1986 from central Nuevo Leon, Mexico by its slightly longer pedipalp chelae (chela length/ depth 1.92-2.16 in D. Undo, 1.78-1.88 in males- and 1.86-1.96 in females in D. col- welli), weaker reticulations on the pedipalp chelae, slightly lower telotarsal spine formula (4/5:5/5-6:6/7:677 in D. Undo, 5/5-6:576:6/7: 6/7 in D. colwelli), and m^oderately dentate lat- eral margin of median lobe of the hemisper- matophore (vestigially dentate in D. colwelli). Diplocentrus Undo is distinguished from D. ferruginous Fritts & Sissom 1996 (from south- ern Nuevo Leon) by having a lower telotarsal spine formula (4/5:5/5-6:6/7:677 in D. Undo, 515:616:1/1:7/1—^ in D. ferruginous), a higher chela leegth/depth ratio in males (mean chela length/Mepth 2.00 in both sexes for D. Undo, 2.19-2.28 in D. ferruginous males, 2.02 for 306 THE JOURNAL OF ARACHNOLOGY Figures 2-9. — Diplocentnis Undo, adult male holotype from 5 miles north of Sanderson, Terrell County, Texas, USA. 2-8. Right pedipalp; 2. Femur, dorsal aspect; 3. Patella, dorsal aspect; 4. Patella, external aspect; 5. Patella, ventral aspect; 6. Chela, external aspect; 7. Chela, ventral aspect; 8. Chela, internal aspect. 9. Right hemispermatophore, lateral aspect. Abbreviations: dl = distal lamella, ml — median lobe, tr = trunk. Not to scale. females), and is much lighter in color than in D. Undo. Diplocentrus Undo is easily distin- guished from D. whitei Gervais 1844 from Texas, USA and Coahuila, Mexico, by its low- er telotarsal spine formula (4/5: 5/5-6: 6/7: 6/7 in D. Undo, 5— 6/7:6/7— 8:7/8:7/8 in D. whitei), its lower pectinal tooth counts (males 11-15, females 9-13, for D. Undo’, males 16-20, fe- males 14—18, for D. whitei), and shorter, stouter pedipalps (mean chela length/depth 2.00 in both sexes for D. Undo, 2.60 in males and 2.17 in females for D. whitei). Diplocen- trus Undo is generally darker in color than D. spitzeri Stahnke 1970 from Arizona, USA and Sonora, Mexico, and D. peloncillensis Francke 1975 from Arizona and New Mexico, USA. Diplocentrus Undo also has lower telo- tarsal spine formulas, especially on telotarsi I and II (4/5:5/5-6:6/7:6/7 in D. Undo, 616:616- 1:111:111 in D. spitzeri, 5/6: 6/6-7: 6/7: 6/7 in D. peloncillensis). Description. — Male: Color brown with variable dark brown marbling. Carapace smooth to minutely granular interspersed with a few larger granules anteriorly and laterally; prosomal venter lustrous, punctate; pectinal tooth count 11 to 15 (mode 14). Mesosomal tergites moderately granular; tergite VII acar- inate, moderately bilobate, coarsely granular. Sternites lustrous; sternite VII with submedian and lateral carinae weak, smooth on posterior one-half to one-fourth. Metasoma moderately hirsute; intercarinal spaces lustrous, finely punctate. Dorsolateral STOCKWELL & BALDWIN— NEW DIPLOCENTRUS FROM TEXAS 307 carinae weak, vestigially granulose on seg= ments LIV. Lateral supramediae carinae mod= erate, granulose on segments LIV. Lateral ie= framedian carinae moderate on segment I; weak on II and III; vestigial on IV; granulose on all segments. Ventrolateral carinae moder- ate, granulose on segments LIII; weak, gran- ulose on IV Ventral submedian carinae mod- erate, vestigially granulose on segments LII; weak, vestigially granulose on III; obsolete on IV Metasomai segment V dorsolateral carinae weak, smooth; lateromedian carinae weak, granular; ventrolateral, ventromedian and ven- tral transverse carinae moderate with a single row of large tubercles; anal subterminal carina moderate, tuberculate; anal terminal carina weak, crenulate. Telson smooth with a few tu- bercles on ventral anterior margin; moderately setose. Pedipalps robust, orthobothriotaxic C (Va- chon 1974, figs. 11-17). Femur with dorsal face sparsely granular; internal face with mod- erately dense tubercles; ventral and external faces weakly granular to smooth; dorsointer- nal carina moderate, granulose; dorsoexternal carina moderate with large granules proxi- mally, obsolete distaliy; ventroexternal carina obsolete; ventrointernal carina weak, irregu- larly tuberculate. Patella with ventral and ex- ternal faces weakly to vestigially reticulate; internal face moderately granular; basal tuber- cle moderately strong, rounded, moderately granular; dorsomedian carina moderate, smooth; ventroexternal carina moderate, smooth; ventrointernal carina weak to mod- erate, tuberculate; other carinae obsolete. Che- la with dorsal and external faces moderately reticulate; internal and ventral faces weakly reticulate; dorsomarginal carina weak to mod- erate, granular; dorsal secondary carina ves- tigial; digital carina strong; external secondary carinae weak; ventroexternal carina obsolete; ventromedian carina strong; ventrointernal ca- rina moderate; three internal carinae weak to vestigial, granular. Legs typical for genus. Ex- ternal faces with weak to moderately dense granulation. Modal telotarsal spine formula 4/ 5: 5/5-6: 6/7: 6/7. Distal lamella of hemisper- matophore not noticeably elongate. Lateral external margin of median lobe moderately dentate (fig. 18). Measurements of holotype male (L = length, W = width, D = depth). Total L, 43.30. Carapace L, 5.65. Mesosoma L, 14.00. Metasoma L, 23.65. Metasomai seg- ments L/W/D: I, 2.90/3.10/2.30; II, 3.20/2.70/ 2.20; III, 3.50/2.60/2.15; IV, 4.10/2.55/2.10; V, 5.30/2.05/1.90. Telson: L, 4.65; vesicle L/ W/D, 3.65/2.10/1.75; aculeus L, LOO. Chelic- era: chela L/W, 2.85/1.20; fixed finger L, 1.05; movable finger L, 1.70. Pedipalp: femur L/W/ D, 4.50/2.05/1.70; patella L/W./D, 4.65/2.25/ 2.50; chela L/W/D, 9.90/2.80/4.80; fixed finger L, 4.30; movable finger L, 6.10. Female: Similar to male except as follows. Tergites sparsely granular to smooth; pedipal- pal and metasomai carinae weaker, reticulate pattern vestigial to obsolete; pectinal tooth counts 9 to 13 (m.ode 10). Sexes are morpho- metrically similar; carapace wider than long; pedipalp chela length/depth ratio 1.9-2. 2; me- tasomai segment I wider than long; remaining segments longer than wide. Measurements of a paratype female from 19 mi. S of Sheffield, Terrell County, Texas: Total L, 38.55. Carapace L, 5,20. Mesosoma L, 13.95. Metasoma L, 19.40. Metasomai seg- ments L/W/D: I, 2.35/2.90/2.05; II, 2.60/2.50/ 2.10; III, 2.85/2.40/2.05; IV, 3.25/2.25/1.90; V, 4.25/1.95/1.85. Telson: L, 4.10; vesicle L/ W/D, 3.20/2.15/1.70; aculeus L, 0.90. Chelic- era: chela L/W, 2.65/1.30; fixed finger L, 1.00; movable finger L, 1.60. Pedipalp: femur L/W/ D, 3.75/1.80/1.55; patella L/W/D, 4.15/1.90/ 2.05; chela L/W/D, 8.55/2.60/4.25; fixed finger L, 3.45; movable finger L, 5.05. Variatioe. — Diplocentrus Undo displays a marked difference in overall size across its geographic range. In higher, cooler areas, such as the Davis Mountains of Texas, adults may be nearly half the size of individuals from lower, warmer localities. It was initially sus- pected that the two size classes represented different species. However, except for the dif- ference in size, we have found no discernible variation between the groups. We compared the ranges, means, and standard deviations of six taxonomically important morphometric ra- tios from 20 small adult males and 20 large adult males and found no significant differ- ences between the two groups, despite the dif- ference in overall size. We subsequently pooled the data for all males. The ranges, means, and standard deviations of these six taxonomically important morphometric ratios from 40 adult males and 20 adult females are as follows: pedipalp chela length/depth: males 1.92-2.16, 2.00, 0.05; females 1.92-2.13, 2.00, 0.06. Pedipalp chela length/carapace 308 THE JOURNAL OF ARACHNOLOGY length: males 1.63-1.80, 1.70, 0.04; females 1.53-1.68, 1.62, 0.04. Pedipalp chela length/ pedipalp fixed finger length: males 2.29-2.50, 2.40, 0.06; females 2.33-2.59, 2.41, 0.07. Car- apace length/pedipalp fixed finger length: males 1.33-1.53, 1.41, 0.06; females 1.39- 1.69, 1.48, 0.08. Pedipalp fixed finger length/ pedipalp femur length: males 0.85-0.98, 0.90, 0.03; females 0.84-1.00, 0.94, 0.04. Pedipalp fixed finger length/metasomal segment V length: males 0.68-0.81, 0.76, 0.04; females 0.74-0.92, 0.85, 0.04. Specimens {n = 199) varied in pectinal tooth counts as follows: in males, one comb had 11 teeth, 11 combs had 12 teeth, 78 combs had 13 teeth, 153 combs had 14 teeth, 27 combs had 15 teeth, and three combs were damaged or missing; in females, one comb had nine teeth, 70 combs had ten teeth, 37 combs had 11 teeth, 13 combs had 12 teeth, one comb had 13 teeth, and three combs were damaged or missing. As in all Diplocentrus species, there is var- iation in telotarsal spine counts of D. Undo. We report the telotarsal spine counts from 1 83 specimens. Spine counts from both the right and left legs are reported for each specimen. For each leg, we report the number of spines in the row followed by the number of legs exhibiting that count (in parentheses). Missing or damaged telotarsal rows are indicated by X. Leg I: prolateral row - 2 (2), 3 (7), 4 (327), 5 (25), X (5); retrolateral row - 3 (4), 4 (42), 5 (312), 6 (2), 7 (1), X (5). Leg II: prolateral row - 3 (1), 4 (10), 5 (344), 6 (4), 7 (1), X (6); retrolateral row - 2 (1), 3 (1), 4 (10), 5 (121), 6 (226), X (7). Leg III: prolateral row - 4 (2), 5 (48), 6 (303), 7 (10), 10 (1), X (2); retrolateral row - 5 (1), 6 (53), 7 (303), 8 (6), 11 (1), X (2). Leg IV: prolateral row - 4 (1), 5 (29), 6 (308), 7 (28), X (3); retrolateral row - 4 (1), 5 (1), 6 (53), 7 (290), 8 (18), X (3). Distribution. — This species is widely dis- tributed throughout west Texas and is record- ed from Culberson, Reeves, Jeff Davis, Pecos, Upton, Crockett, Brewster, Terrell, and Val Verde counties. This species is also known from the states of Coahuila and Nuevo Leon in Mexico. Diplocentrus Undo is usually found on rocky slopes. Individuals may be found in burrows beneath large stones and other surface objects. Specimens examined. — USA: TEXAS: Brewster County, Alpine, 5 June 1942 (E.S. Ross), 3 males 1 female (CAS); Alpine, 16 July 1949 (B.H. War- nock), 1 male (CAS); Alpine, 23 April 1964 (J. Scudday), 1 male (CAS); Alpine, 27 April 1964 (C. Babcock), 1 female (CAS); Alpine, 26 November 1964 (J. Scudday), 1 female (CAS); Alpine, 14 June 1965 (J. Scudday), 1 female (CAS); 5 mi. S Alpine, 12 July 1955 (S.A. Minton), 1 male (CAS); 5.5 mi. S Alpine, 19 August 1968 (S.C. Williams, M.M. Bentzien, J. Bigelow), 2 females (CAS); 6 mi. S of Alpine, 30 June 1965 (M.H. Muma), 1 female (CAS); 6 mi. S of Alpine, 27 July 1978 (O.E Francke, J.V. Moody), 4 females (AMNH); 8 mi. S Alpine, Hwy. 1 18, 25 September 1964 (J. Scudday), 2 females (CAS); 8 mi. S Alpine, 26 November 1964 (J. Scudday), 7 females (CAS); 10-12 mi. S Alpine, 4/11 October 1964 (J. Scudday), 5 males 2 females 2 immatures (CAS); 10 mi. SW of Alpine, 11 August 1966 (A. Jung, K. Horn), 1 male (AMNH); 10 mi. SW of Alpine, 11 August 1966 (T. Briggs, A. Jung, K. Horn), 6 females (CAS); 12 mi. S Alpine, Hwy. 118, 5/10 October 1965 (Rog- ers, Freels), 2 females (CAS); 22 mi. S Alpine, Babcock Ranch, 25/26 April 1964 (S. Sikes), 2 fe- males (CAS); 22 mi. S Alpine, Babcock Ranch, 27 April 1964 (C.E. Babcock), 1 male 2 females (CAS); 23 mi. S Alpine, Babcock Ranch, 15 March 1964 (C.E. Babcock), 1 male 2 females (CAS); 25 mi. S Alpine nr. Calamity Cr., 15 April 1964 (T. Watson), 1 female (CAS); 5 mi. E Lajitas, 1 August 1986 (R.W. Manning, R. Hollander), 1 male (WDS); Paisano (Biological Expeditions, U. S. Dept, of Agriculture), 7 July 1890 (W Lloyd), 1 female (USNM); Black Gap Wildlife Area, 30 July 1955 (W.G. Degenhardt), 1 male (AMNH); Black Gap Refuge, Norton Tank, 30 August 1960 (WG. Degenhardt), 1 female (CAS); 41 mi. N Panther Jet., Hwy. 385, 17 August 1968 (S.C. Williams, M.M. Bentzien, J. Bigelow), 3 males 4 females (CAS); Big Bend N.P., 1959 (H.L. Stahnke), 1 male (CAS); Big Bend N.P., Grapevine Hills, 17 August 1968 (M.A. Cazier, J. Bigelow), 1 male (AMNH); Big Bend N.P., 13. 2 mi. SE Panther Jet., 17 August 1968 (S.C. Williams, M.M. Bentzien), 2 males (CAS); Big Bend N.R, Kibbe Spn, 24 July 1956 (H.L. Stahnke), 2 females (CAS); Big Bend N.R, Oak Spr., 30 July 1956 (H.L. Stahnke), 1 female (CAS); Big Bend N.R, Window Trail, 19 July 1956 (R. Curbow), 1 female (CAS); Big Bend N.R, Chi- sos Mts., no date (no collector), 2 females (AMNH); Big Bend N.R, Chisos Basin, CCC Camp, 25-30 July 1937 (Necker), 1 female (AMNH); Big Bend N.R, Laguna, Mt. Emory, 12 April 1937 (no collector), 1 female (AMNH); Big Bend N.R, Chisos Mts., foot of Emery Mts., 6 July 1938 (E. Shaw, J. & R. Schmidt), 1 female (CAS); Big Bend N.R, Chisos Mts., foot of Mt. Emery, 6 July 1938 (B.H. & E. Shaw), 1 male 1 female (CAS); Big Bend N.R, Chisos Mts., 28 May 1952 STOCKWELL & BALDWIN— NEW DIPLOCENTRUS FROM TEXAS 309 (M.A. Cazier, W. J.. Gertsch, R. Schrammel), 2 fe- males (AMNH); Big Bend N.P., Chisos Mts., Green Gulch, 5- April 1955 (S.A. Minton), Big Bend N.R, Chisos Mts. Basin, 1 July 1956 (E. Steele), 1 male (CAS); Big Bend N.R, Chisos Mts. Upper Basin, 30 July 1955 (H.L. St-ahnke), 1 male (CAS); Big Bend- N.R, Upper Basin, 21/25 July 1956 (H.L. Stahnke), 1 male 1 female (CAS); Big Bend N.P., Chisos Basin, 5- August 1962 (C.A. Triplehom), 1 male (CAS); Big Bend' N.P., Chisos Basin, 26 May 1965 (K.W. Haller), 2 males (AMNH); Big Bend N.P., Chisos Basie, 27 August 1965 (C. Parrish), 1 male 2 females (CAS); Big Bend N.P., Chisos Basin Pass, 28 July 1978 (O.F, Francke, J.V. Moody), 1 female (AMNH); Big Bend N.P., Chisos Basin,. 29 July 1978 (O.F. Francke, J.V. Moody), 1 female (AMNH); Big Bend N.P., CMs.os Basin, 9 August 1979 (O.F Francke), 1 female (AMNH); Big Bend N.P., Juniper Canyon, July 1921 (no collector), 1 male 5 females (AMNH); Big Bend N.P., Pine Can- yon, 10 August 1979 (Francke, Moody, Merickel), I mate 1 female (AMNH); Big Bend N.P., Pine Canyon, 10 August 1979 (Francke, Moody, Mer- ickle), 1 female with young (AMNH); Big Bend N.P., Boquillas Canyon, 27 January 1973 (C. McConnell), 1- female (AMNH). Crockett County: II mi. N of Iraan, 29 September 1985 (S.A. Stock- well), 1 male 1 female (SAS); 10 mi. N of Iraan, 15 September 1985 (S.A. Stockwell), 6 females (SAS); 5 mi. N, 4 mi.. W Iraan, 30 June 1986 (Man- ning-, Hollander), 2 males (WDS); 15 mi. E of Iraan, 14 September 1985 (S.A. Stockwell), 1 male (SAS); 45 mi. NW of Ozona, 21 March 1978 (O.F Francke, T.B. Hall, J.V.- Moody), 2 females (AMNH).. Culberson County: 8 January 1981 (G. Zolnerowich), 2 females (MWSU); 4 mi. NNE of Kent, 14 March 1981 (N.V Homer), 2 mates 1 fe- male (MWSU); 6 mi. N of Kent, 20 March 1985 (no collector), 1 female (MWSU); 9 mi. N of Kent, 19 April 1980 (W.W. Dalquest), 1 female (MWSU); 31.8 mi. NE of Van Horn, 2 July 1978 (Francke, Hall, Moody), 1 male (AMNH), Jeff Davis County: 15"April. 1968 (F. Home), 2- males 6 females (CAS); Davis- Mts., Fort Davis Quad., Cottonwood Springs, 27 May 1916 (F.M. Gaige), 1 female (UMMZ); Da- vis Mts,, Fort Davis Quad., Cottonwood Springs, 5 June 1916 (F.M. Gaige), 1 female (UMMZ); Davis Mts., Fort Davis Quad.,- Cottonwood Springs, 7 June 1916 (EM. Gaige), 1 male 4 females (UMMZ); Davis Mts., Fort Davis Quad., 12 June 1916 (F.M Gaige), 1 female (UMMZ); Davis Mts., Fort Davis Quad., Two Spring Canyon, 28 June 1916 (F.M. Gaige), 1 male 1 female (UMMZ); Da- vis Mts., Fort Davis Quad., 6 July 1916 (F.M. Gai- ge), 1 mate (UMMZ); Davis Mts., Fort Davis Quad., Maple Canyon, 8 July 1916 (EM, Gaige), 5 females 2 immatures (UMMZ); Davis Mts., Fort Davis Quad., Cherry Canyon, 9 July 1916 (F.M. Gaige), 1 female (UMMZ); Davis Mts., Fort Davis Quad., 14 July 1916 (FM. Gaige), 1 female (UMMZ); Davis Mts., 9 May 1951 (O. Bryant), 1 male (CAS); Davis Mts. State Park, no date (O.F. Francke, J.V. Moody), 1 male (AMNH); Davis Mts. State Park, 5 mi, N Ft. Davis, 26 April 1964 (C. Babcock), 1 mate (CAS); Davis Mts. State Park, 20 June 1970 (M.A. Ca-zier, L. Draper, O.F. Francke), 46 males 7 females (AMNH); Davis Mts. State Park, Limpia Canyon Campground, 5 June 1 974 (L. Draper, M.A. Cazier, O.F. Francke), 37 males 10 females (AMNH); Davis Mts. State Park, 9 June 1978 (O.F. Francke), 3 males (WDS); Davis Mts. State Park, 1 March 1985 (S.A. Stockwell, J.M. Steele), 1 male 1 females (SAS); 2 mi. W of Fort Davis, 22 April 1970 (A. Schoenhor, W.L. Minck- ley), 1 male 2 females (AMNH); 4 mi. W of Fort Davis, 6 June 1978 (O.F. Francke) (WDS); 9 mi. N of Fort Davis, 22 June 1970 (W. Seifert), 1 male (MWSU); 8 mi. E of McDonald Observatory, no date (N-.V. Horner), 1 male (MWSU); 20 km S of Toy ah vale, 13 March 1977 (D. Holub, K. Douglas), 1 male, 1 female (MWSU). Pecos County: 10 mi. N of Ft. Stockton on Will Banks Ranch, 27 Decem- ber 1966 (B-. Winokur), 4 females (CAS); 30 mi. S of Ft. Stockton,. Glass Mts., 7 June 1974 (L. Draper, M.A. Cazier, O.F. Francke), 2 males (AMNH); Sheffield, Pecos River Bluff, 7 July 1968 (M.H. and E.U. Muma), 1 female with three first instar young (CAS); 4 mi. E of Sheffield, Pecos River, 7 June 1974 (M.A. Cazier, L. Draper, O.F. Francke), 8 mates 7 females (AMNH); 15 mi. N of Sanderson, 3 June 1970 (W. Seifert), 1 female (MWSU); 20 mi. W of Sanderson, 2 September 1983 (W.D. & J.C. Si'ssom), 1 female (WDS). Reeves County: 22 mi. SW of Toyah, 3 October 1983 (D. Foster), 1 male (WDS); Balmorhea State Park, 26 August 1971 (K., M.,, and M.A. Cazier), 2 males 1 female (AMNH); Balmorhea State Park, June 1979 (Moody, Merickel), 2 females (AMNH). Terrell County: Sheffield, Pecos R. Bluff, 7 July 1968 (M.H.. and E.U. Muma), 1 female (CAS); 19 mi. S of Sheffield, Blackstone Ranch, 16 May 1958 (W.H. McAlister), 1 female (TMM), 19 mi. S of Sheffield, 8 June 1974 (O.F. Francke), 6 immatures (AMNH); 19 mi. S of Sheffield, 15 June 1974 (L. Draper, M.A. Cazier, O.E Francke), 2 males 2 females (AMNH); Pecos River and Independence Creek, Chandler Ranch, 27-28 June 1968 (W.L. Minck- ley), 2 males (AMNH); 1 mi. S Pecos County line, 4 June 1986 (Manning, Hollander), 1 male 1 juve- nile (WDS); 6.3 mi. NW Sanderson, 20 November 1960 (D. Campbell, H. Harris), 1 female (AMNH); 5 mi. N of Sanderson, 8 June 1974 (L. Draper, M.A. Cazier,. O.F. Francke), 3 males (AMNH); 5 mi N of Sanderson, 15 June 1974 (L. Draper, M.A. Cazier, O.F. Francke), 10 males (AMNH); 4 mi. E of Dry- den, 4 September 1939 (D. and S. Mulaik), 3 fe- males (AMNH); 21 mi. N of Dryden, 2 July 1970 (W. Seifert), 1 male (MWSU); Upton County, 3 mi. 310 THE JOURNAL OF ARACHNOLOGY S, 5 mi. E McCamey, 7 June 1986 (Manning, Hol- lander), 2 males 1 juvenile (WDS). Val Verde County: 20 mi S of Juno, 2 May 1970 (W. Seifert), 2 females (MWSU); 21 mi. N of Comstock, 14 Sep- tember 1985 (S.A. Stockwell), 2 males 1 female (SAS); 19 mi. N of Comstock, 14 April 1973 (J. Cooke), 1 male 2 females (AMNH); 15 mi. N Com- stock, 22 June 1971 (no collector), 1 male 2 females (CAS); 10 mi. N Comstock, 22 June 1971 (no col- lector), 1 male 2 females (CAS); 7 mi. N Comstock, 12 July 1986 (Manning, Hollander), 1 female (WDS); 0.5 mi. NW of Comstock, 8 May 1968 (T. Walker), 1 female (AMNH); 10 mi. W of Com- stock, Pecos River, 2 September 1983 (W.D. & J.C. Sissom), 1 male (WDS); 1 1 mi. W of Comstock, 28 August 1970 (E & J.M. Davidson), 2 males (CAS); 21 mi. N of Langtry, 14 April 1973 (T.R. Mollhagen), 1 female (AMNH); 5 mi. N of Langtry, 15 April 1973 (J. Cooke), 1 male 1 female (AMNH); 3 mi. N of Langtry, 3 November 1984 (J. Reddell, M. Reyes), 1 female (TMM), Langtry, 26 June 1971 (E. Tombellin), 1 female (AMNH); 0.5 mi. S of Langtry, 14 June 1974 (L. Draper, M.A. Cazier, O.E Francke), 5 males 7 females (AMNH); 2 mi. SSE Langtry, 7 June 1974 (L. Draper, M.A. Cazier, O.E Francke), 1 female (AMNH); 3 mi. W of Langtry, Rattlesnake Canyon, 30 April 1983 (EL. Rose), 1 female with young (AMNH). MEXICO! COAHUILA: 10 km SE Mus- quiz, 24 June 65 (J. Reddell), 1 female (TMM). NUEVO LEON: 4 mi. S of Bustamante, 26 March 1964 (B. Russell), 1 male (AMNH); 9 mi. E of Mex. 57 on Galeana Road, March 1968 (T. Walker), 1 male (AMNH); 2 mi. NE of Villa de Garcia, 19 August 1984 (Sissom, Myers, Born), 1 male (WDS). ACKNOWLEDGMENTS The senior author thanks Dr. Oscar E Francke for providing specimens and his per- sonal notes for use in this study. Thanks are also due Dr. W. David Sissom (WDS), Mr. James C. Cokendolpher, and Dr. Steven W. Ta- ber, all of whom provided advice and assis- tance during the course of this study. The helpful consideration of the following persons and their respective institutions for the loan of material on which this contribution based is greatly appreciated: Dr. Norman I. Platnick, American Museum of Natural History (AMNH); Dr. Norman Penny and Mr. Vincent F. Lee, California Academy of Science (CAS); Dr. Jonathan Coddington and Mr. Scott Larch- er, United States National Museum (USNM); Dr. Norman V. Horner, Midwestern State Uni- versity (MWSU); Mr. James R. Reddell, Texas Memorial Museum (TMM); Dr. T. Moore, University of Michigan (UMMZ). The Na- tional Park Service at Big Bend National kindly provided me (SAS) with permits to col- lect on federal lands. We would also like to acknowledge two anonymous reviewers for their suggestions to improve this manuscript. LITERATURE CITED Banks, N. 1900. Synopses of the North American invertebrates. IX. The scorpions, solpugids and pedipalpi. American Naturalist 34:421-427. Brown, C.A. & D.R. Formanowicz, Jr. 1996. Re- productive investment in two species of scorpi- on, Vaejovis waueri (Vaejovidae) and Diplocen- trus linda (Diplocentridae), from west Texas. Annals of the Entomological Society of America 89:41-46. Ewing, H.E. 1928. The scorpions of the western part of the United States, with notes on those occurring in northern Mexico. Proceedings of the United States National Museum 73:1-24. Fet, V., W.D. Sissom, G. Lowe & M.E. Braunwald- er. 2000. Catalog of the Scorpions of the World (1758-1998). The New York Entomological So- ciety, New York, 690 pp. Francke, O.E 1978. Systematic Revision of Diplo- centrid Scorpions (Diplocentridae) from Circum- Caribbean Lands. Special Publications of the Museum, Texas Tech University, No. 14, 92 pp. Gertsch, W.J. 1939. Report on a collection of Arachnida from the Chisos Mountains. Contri- butions to the Baylor University Museum, Waco, Texas 24:17-26. Kovafik, F. 1998. [Stiri (Scorpions). Madagascar.] Jihlava. 175 pp. (in Czech). Lamoral, B.H. 1979. The scorpions of Namibia (Arachnida: Scorpionida). Annals of the Natal Museum 23:497-784. Pocock, R.I. 1902. Arachnida: Scorpiones, Pedi- palpi, et Solfugae. In: Biologia Centrali-Ameri- cana (Zoologia). London, Taylor & Francis. 71 pp. Rowland, J.M. & J.R. Reddell. 1976. Annotated checklist of the arachnid fauna of Texas (exclud- ing Acarida and Araneida). Occasional Papers of the Museum, Texas Tech University 38:1-25. Sissom, W.D. 1994. Systematic studies on Diplo- centrus keyserlingii and related species from cen- tral Oaxaca, Mexico (Scorpiones, Diplocentri- dae). Mitteilungen aus dem Hamburgischen Zoologischen Museum und Institut Berlin 70: 257-266. Sissom, W.D. & V. Fet. 2000. Family Diplocentri- dae Karsch, 1880. Pp. 329-354. In: Catalog of the Scorpions of the World (1758-1998) (Fet, V, W.D. Sissom, G. Lowe, & M.E. Braunwalder, eds.). The New York Entomological Society, New York. STOCKWELL & BALDWIN— NEW DIPLOCENTRUS FROM TEXAS 311 Stahnke, H.L. 1970. Scorpion nomenclature and mensuration. Entomological News 81:297-316. Stockwell, S.A. & J.A. Nilsson. 1987. A new spe- cies of Diplocentrus Peters from Texas (Scorpi- ones, Diplocentridae). Journal of Arachnology 15:151-156. Vachon, M. 1952. Etudes Sur Les Scorpions. Pub- lications de ITnstitut Pasteur d’Algerie, Algiers. 462 pp. Vachon, M. 1973 (1974). Etude des caracteres uti- lises pour classer les families et les genres de scorpions. 1. La trichobothriotaxie en Arachnol- ogie. Sigles trichobothriaux et types de tricho- bothriotaxie chez lez Scorpions. Bulletin du mu- seum national d’histoire naturelle (Paris), sen 3, 104: 857-958. Manuscript received 1 October 2000, revised 19 March 2001. 2001. The Journal of Arachnology 29:312-329 NOTES ON THE GENUS SCYTODES (ARANEAE, SCYTODIDAE) IN CENTRAL AND SOUTH AMERICA Antonio D* Brescovit and Cristina A* Rheims: Laboratorio de Artropodes, Institute Butantan, Av. Vital Brasil, 1500, Butanta, CEP 05503-900, Sao Paulo SP, Brazil ABSTRACT. In this study we present a redescription of Scytodes championi, S. romitii and S, guttipes. Seven species are newly described: S. panamensis from Panama; S. vaurieorum and S, chiquimula from Guatemala; S. cogu and S. armata from Costa Rica; S. tegucigalpa and S. zamorano from Honduras. Four of these were described as variations of 5'. championi in a recent revision of the species of Central America. New records are presented for S. championi, S. romitii, S. guttipes, S. gertschi and S. cubensis. Keywords: Araneae, Scytodidae, Scytodes, Neotropical region, systematics The genus Scytodes Latreille 1 804 has been intensively studied in the Neotropical region during the last two decades (Brignoli 1976; Alayon 1977, 1985, 1992; Valerio 1981; Bres- covit & Hofer 1999; Brescovit & Rheims 2000; Rheims & Brescovit 2000). The genus has a worldwide distribution with several syn- anthropic species (Brescovit & Rheims 2000). To date, at least 42 species in the Neotropical region are considered valid. During a preliminary study of the Brazilian Scytodes, we observed that Scytodes romitii, described by Caporiacco (1947) from Guyana, was very common in the north and northeast of the country. This species, herein rede- scribed, is very similar to S. championi EO.R- Cambridge 1899, previously known from Central America, differing only slightly in the morphology of the male and female genitalia. The similarity between these species was so accentuated that it became necessary to recon- sider earlier records, for the State of Amazon- as, of S, cf. championi (see Hofer 1990) and V championi (see Brescovit & Hofer 1999). Scytodes championi was originally de- scribed by EO.P.-Cambridge (1899) for Chi- riqui, Panama, and more recently redescribed by Valerio (1981). Valerio presented a series of variations for S. championi together with a revision of the Central American scytodid species. He also examined specimens identi- fied as 5. guttipes Simon 1893, by Banks (1929) and considered them identical to S. championi. Nevertheless, he kept the name championi for the Central American forms due to the lack of type examination and gen- eral revisions. Nentwig (1993) followed Val- erio’s identifications and considered Banks’s specimens as misidentifications. Based on Valerio’s paper and on the study of material from Central and north of South America we concluded that the S. championi sensu Valerio is, in fact, a group of four dif- ferent species, based on morphological differ- ences of male and female genitalia as well as carapace basic coloration pattern. Although Brignoli (1976, figs. 20-25) argues that S. thoracica (Latreille 1802) and S. strandi Spas- sky 1941 present a high degree of genitalic variation we do not consider this applicable to S. championi sensu Valerio (1981, figs. 16- 1 8) since we observed a very constant pattern of genitalic morphology in all Central Amer- ican species of what we could call the "'charn- pionf" group. In addition, we found that S. championi oc- curs in the Brazilian states of Amazonas, Ro- raima and Para and is sympatric with S. rom- itii at least in the state of Amazonas. A redescription of S. guttipes is presented, con- firming it as a valid species. Seven new Scy- todes species are described for Central Amer- ica and new records and illustrations are presented for Scytodes gertschi Valerio 1981 and for the male of S. cubensis Alayon 1977. 312 BRESCOVIT & RHEIMS— THE GENUS SCYTODES 313 The material examined belongs to the fob lowing institutions: AMNH, American Mu- seum of Natural History, New York (NT. Plat- nick); BMNH, The Natural History Museum, London (J. Beccaloni); CEPLAC, Centro de Pesquisas do Cacau, Itabuna, Bahia (PS. Ter- ra); IBSP, Instituto Butantan, Sao Paulo (A.D. Brescovit); INPA, Instituto Nacional de Pes- quisas da Amazonia, Manaus (C. Magalhaes); MCN, Museu de Ciencias Naturals, Fundagao Zoobotanica do Rio Grande do Sul, Porto Ale- gre (E.H. Buckup); MZS, Museo Zoologico de La Specola, Firenze (S. Whitman); MCTP, Museu de Ciencias e Tecnologia, Pontificia Universidade Catolica do Rio Grande do Sul, Porto Alegre (A. A. Lise); MCZ, Museum of Comparative Zoology, Cambridge, Massachu- setts, (L. Leibensperger); MNHN, Museum National de Histoire Naturelle, Paris (C. Rol- lard); MNRJ, Museu Nacional do Rio de Ja- neiro, Rio de Janeiro (A.B. Kury); MZSP, Mu- seu de Zoologia da USP, Sao Paulo (E. Cancello); SMNK, Staatliches Museum fiir Naturkunde Karlsruhe (H. Hofer). Despite our efforts, it was not possible to obtain Valerio’s scytodid material, deposited in the Museo de Zoologia, Universidade de Costa Rica (MZUCR). Descriptions and terminology follow Bres- covit & Rheims (2000). All measurements are in mm. The female genitalia were submerged in lactic acid to study internal structures. Mi- crographs were obtained with a JEOL (JSM 840A) scanning electron microscope from the “Laboratorio de Microscopia Eletrdnica do Departamento de Fisica Geral do Instituto de Fisica da Universidade de Sao Paulo (USP).” Scytodes championi EO.P -Cambridge (Figs. 1, 2, 13-17) Scytodes championi EO.P.-Cambridge 1899: 51 (male lectotype and female paralectotype, here designated, from Chiriqui, Panama deposited in BMNH, examined); Roewer 1942: 329; Valerio 1981: 87, only figs. 7-9. Diagnosis. — The males of S. championi re- semble those of S. romitii and S. panamensis by the dorsal groove on the distal area of the palpal bulb, but differ by the strong median narrowing of the bulb and greater depth of the dorsal groove (Figs. 1, 2, 14). The female dif- fers from females of the other species by the widely separated and almost transversally ori- ented positioning ridges (Fig. 16) and anterior pair of subtriangular seminal receptacles sep- arated from the smaller posterior pair (Fig. 17). Male. (MCTP 1828). — Carapace yellow with double U-shaped dark brown pattern and a pair of internal parallel light brown stripes (Fig. 13). Pedipalps yellow with longitudinal dorsal brown stripe. Labium and endites yel- low with brownish margins. Sternum yellow with brown margins at base of legs and ex- tending towards center along slight grooves. Legs yellow with three longitudinal brown stripes along ventral face of femora and lon- gitudinal stains along tibiae and metatarsi. Ab- domen cream colored with dorsal scattered black spots and pair of transversal posterior black bands (Fig. 13). Total length 3.50. Car- apace, 1.75 long, 1.50 wide. Eye diameters: PME 0.12, ALE 0.12, PLE 0.14. Lateral eyes on tubercle. Chelicerae with subapical hyaline keel. Labium 0.20 long, 0.22 wide. Sternum 0.96 long, 0.80 wide. Leg measurements: I - femur 2.63/ patella 0.50/ tibia 3.00/ metatarsus 3.50/ tarsus 0.63/ total 10.26/ II - 2.13/ 0.50/ 2.25/ 2.50/ 0.38/ 7.76/ III - 1.50/ 0.38/ 1.25/ 1.50/ 0.50/ 5.13/ IV - 2.00/ 0.50/ 2.25/ 2.25/ 0.50/ 7.50. Palpal femur presenting stridula- tory pick long and slender with rounded and projected socket. Cymbium with single apical slender spine (Fig. 14). Bulb 0.38 long. Distal area ventrally with anterior slightly sclerotized membrane (SM) followed by subtriangular pocket. (P; Figs. 2, 14). Abdomen 1.75 long, 1.50 wide, rounded, covered with slender hairs. Female. (MCTP 1827). — Coloration with same basic pattern as male. Total length 3.38. Carapace, 2.00 long, 1.75 wide. Eye diame- ters: PME 0.12, ALE 0.14, PLE 0.12. Lateral eyes and chelicerae as in male. Labium 0.16 long, 0.22 wide. Sternum 1.12 long, 0.82 wide. Leg measurements: I - femur 2.00/ pa- tella 0.50/ tibia 2.13/ metatarsus 2.50/ tarsus 0.50/ total 7.63/ II - 1.63/ 0.50/ 1.50/ 2.00/ 0.50/ 6.13/ III - 1.13/ 0.38/ 1.13/ 1.25/ 0.38/ 4.14/ IV - 1.75/ 0.50/ 1.75/ 1.75/ 0.50/ 6.25. Epigynal fovea very narrow. Positioning ridge semicircular (Fig. 16). Internal genitalia with two pairs of seminal receptacles, the smaller ones globose (Fig. 17). Abdomen 1.38 long, 1.25 wide, as in male. Variation. — Males: Total length 3.00- 4.63; carapace 1.63-3.25; femur I 2.38-6.25; 314 THE JOURNAL OF ARACHNOLOGY Figures 1-6. — 1-2. Scytodes champioui FO.P.-Cambridge, male palp, prolateral view; 2. Retrolateral view. 3-6. Scytodes romitii Caporiacco. 3. Male palp, prolateral view; 4. Retrolateral view; 5. Distal area, retrolateral view; 6. Stridulatory pick (P = pocket, SM = sclerotized membrane). bulb 0.32-0.50 (n = 15). Females: Total length 3.63-5.75; carapace 1.88-2.63; femur I 2.00-5.25 {n = 20). Coloration pattern and genitalic morphology constant. Distribution. — Central America and north- ern South America. Material examined. — NICARAGUA: Jinote- ga: Masawas (Waspuc River), 19, 17-30 Septem- ber 1955, B. Malkin (AMNH); GUATEMALA: Peten: Tucuru, 262 9, 12-13 July 1947, C. & P. Vaurie (AMNH); Panzos: 14-17 July 1947, C. Vau- rie & P. Vaurie (AMNH); EL SALVADOR: La Lihertad: La Libertad, 19, October 1959, N.L.H. Krauss (AMNH); BELIZE: Toledo District: Id, 7 April 1974, Goodnight (AMNH); PANAMA: Ca- nal Zone: Barro Colorado Island, Id, 1 juv., 3 De- cember 1965, R.X. Schick (AMNH); 19, 2 juvs.. BRESCOVIT & RHEIMS— THE GENUS SCYTODES 315 Figures 7-12. — 7-8. Scytodes panamensis new species. 7. Male palp, prolateral view; 8. Distal area, retrolateral view. 9-10. Scytodes tegucigalpa new species. 9. Male palp, distal area (e = embolus opening); 10. Prolateral view. 11-12. Scytodes armatus new species. 11. Male palp, retrolateral view; 12. Male femur I, ventral spines. April 1953, A.M. Nadler (AMNH); 3 9, 2 juvs., 20 April 1953, A.M. Nadler (AMNH); 1$, 23 May 1952, T.C. Schneirla (AMNH); 2$, 1928 (AMNH); Id, 3-20 April 1953, A.M. Nadler (AMNH); Bar- bacoas Islands, Id, 14 December 1965, R.X. Schick & M. Moody (AMNH); BRAZIL: Rorai- ma: Ilha de Maraca, Id, 17 July 1987, A. A. Lise (INPA); Id 19, 18 July 1987, A. A. Lise (INPA); 19, 19 March 1987, A.A. Lise (INPA); Id, 29 March 1987, A.A. Lise (INPA); 4 9, 6 juvs., 31 January-14 February 1992, A.A. Lise (MCTP 1827); ld39, 2 juvs., 31 January- 14 February 1992, A.A. Lise (MCTP 1828); 19,1 juv., 31 Jan- uary-14 February 1992, M. Nascimento (MCTP 1966); (Estagao Ecologica de Maraca), 20 March 1987, A.A. Lise (MCTP 17623); Id, 17 March 316 THE JOURNAL OF ARACHNOLOGY 1987, A.A. Lise (MCN 17621); 16, 25 July 1987, A. A. Lise (MCN 17622); Amazonas: Manaus (Fa- zenda Esteio), 16, 15 October 1985, B.C. Klein (MCN 19876); Sao Gabriel da Cachoeira, Matura- cA 19, 13 October 1990, A.A. Lise (MCTP 1261); Para: Santarem, FMma de Urucurituba, 19, 24 January 1994, A.D. Brescovit (MCN 25354); 16, 24 January 1987, A.D, Brescovit (MCN 25030). Scytodes romitii Caporiacco (Figs. 3-6, 18-25) Scytodes romitii Caporiacco 1947: 22 (female ho- lotype from Diamont Point, East Demerara Dis- trict, Guyana, iO.V. 1936, deposited in MZS 519, examined); 1948: 626, figs. 17, 18; Brignoli 1983: 150. Scytodes cf. championi: Hofer 1990: 175. Scytodes championi: Brescovit & Hofer 1999: 105 (misidentification). Diagnosis. — The males of S, romitii differ from the other species, here included, by the retrolateral medially-positioned ventral serrat- ed sclerotized membrane (Figs. 4, 5, SM) and by slightly-narrowed median region in the male palpal bulb (Figs. 3, 4, 19, 20). The fe- males differ by the bulb-like shape of the an- terior pair of seminal receptacles very close to the posterior pair (Fig, 22). Male. — (Parque Nacional da Serra do Di- visor, IBSP 12305). Carapace yellow with light brown spotted pattern as shown in Fig, 18. Pedipalps light yellow with brownish stains. Labium and endites cream colored with brownish margins. Sternum as in S, championi but cream colored. Legs light yellow with scattered brown spots except on tarsi. Abdo- men grayish. Total length 4.50. Carapace slightly domed, 2.13 long, 1.76 wide. Eye di- ameters: PME 0.18, ALE 0.16, PLE 0.16. Lat- eral eyes and chelicerae as in S. championi. Labium 0,16 long, 0.22 wide. Sternum 0.88 long, 1.13 wide. Leg measurements: I - femur 10.00/ patella 0.63/ tibia 10.50/ metatarsus 15.00/ tarsus 0.88/ total 37.01/ II - 6.50/ 0.63/ 6.50/ 8.00/ 0.75/ 22.38/ III - 3.63/ 0.63/ 3.38/ 3.88/ 0.63/ 12.15/ IV - 6.00/ 0.63/ 5.75/ 6.88/ 0,75/ 20.01. Palpal femur presenting stridula- tory pick long and slender with oval and pro- jected socket (Fig. 6). Cymbium presenting single distal slender spine (Fig. 19). Bulb 0.70 long. Distal area ventrally with anterior ser- rated membrane followed by small triangular pocket (P; Figs. 4, 5, 20). Abdomen 2.38 long, 1,30 wide, rounded, covered with slender hairs. Female. — (Parque Nacional da Serra do Divisor, IBSP 12526). Coloration as in male. Total length 5.13. Carapace slightly domed, 2.38 long, 1.88 wide. Eye diameters: PME 0.14, ALE 0.14, PLE 0.16. Lateral eyes and chelicerae as in male. Labium 0.32 long, 0.26 wide. Sternum 1 .20 long, 0.94 wide. Leg mea- surements: I - femur 6.25/ patella 0.63/ tibia 7.13/ metatarsus 4.25/ tarsus 0.75/ total 19.01/ II - 4.38/ 0.63/ 4.13/ 4.50/ 0.75/ 19.01/ III - 2.75/ 0.50/ 2.63/ 2.88/ 0.63/ 9.39/ IV - 4.38/ 0.63/ 4.25/ 4.88/ 0.75/ 14.89. Epigynal fovea narrow, curved and parallel, widely separated. Positioning ridge semicircular (Fig. 21). Inter- nal genitalia with two pairs of seminal recep- tacles with short ducts. Central pair with strongly sclerotized ring at base (Fig. 22). Ab- domen 2.88 long, 1.75 wide, as in male. Variation. — Carapace pattern varies great- ly as shown in Fig. 18, 23-25. Males: Total length 3.50-4.88; carapace 1.63-2.38; femur I 5.63-11.38; bulb 0.38-0.64 {n = 10). Fe- males: Total length 3.50-6.00; carapace 1.50- 3.63; femur I 4.13-7.00 {n = 15). Genitalic morphology constant. Distribution. — North and northeastern Brazil. Material examined. — BRAZIL: Acre: Parque Nacional da Serra do Divisor, 19, 13 November 1996, R.S. Vieira (IBSP 9137); 3 9, 9 November 1996, R.S. Vieira (IBSP 8971); 29, 5-25 Novem- ber 1996, R.S. Vieira (IBSP 9268); (Tipologia 9, sitio 10), 19, 23 March 1997, L. Resende & R.S. Vieira (IBSP 12426); (Anil), 23, 10 November 1996, R.S. Vieira (IBSP 9494); (Tipologia 9, sitio 11), 29, 25 March 1997, L. Resende & R.S. Vieira (IBSP 12179); (Varzea Gibralta-Pedro), 13, 20 No- vember 1996, R.S. Vieira (IBSP 9355); (Tipologia 7, sitio 4), 29, 15 March 1997, L. Resende & R.S. Vieira (IBSP 12379); (Travessa Baixa), 1319, 16 November 1996, R.S. Vieira (IBSP 9407); (Tabo- cao), 1319, 17 November 1996, R.S. Vieira (IBSP 9188); (Tipologia 8, sitio 1), 33, 10 March 1997, L. Resende & R.S. Vieira (IBSP 12466); (Juazeiro), 19, 23 November 1996, R.S. Vieira (IBSP 9012); (Tipologia 8, sitio 4), 13, 14 March 1997, L. Re- sende & R.S. Vieira (IBSP 12305); (Tipologia 7, sitio 4), 3329, 18 March 1997, L. Resende & R.S. Vieira (IBSP 12526); Rio Branco (Reserva Extra- tivista de Humaita), 13, 12 April 1996, Eq. IBSP/ SMNK (IBSP 8748); Amazonas: Manaus (Igapo, Taruma-Mirim), 19, 5 February 1988, H. Hofer (SMNK 271); 13, 2 December 1987, H. Hofer BRESCOVIT & RHEIMS— THE GENUS SCYTODES 317 Figures 13-25. — 13-17. Scytodes championi EO.R-Cambridge. 13. Male body, dorsal view; 14. Male palp, retrolateral view; 15. Prolateral view; 16. Female epigynum, ventral view; 17. Dorsal view. 18-25. Scytodes romitii Caporiacco. 18. Male carapace, dorsal view; 19. Male palp, prolateral view; 20. Retro- lateral view; 21. Female epigynum, ventral view; 22. Dorsal view. 23-25. Male carapace, dorsal view, variation patterns: 23. Sao Mateus, Espirito Santo; 24. Serra do Teimoso, Jussari, Bahia; 25. Tefe, Ama- zonas. Scale lines = 0.25 mm. 318 THE JOURNAL OF ARACHNOLOGY (SMNK 272); 1 juv., 17 February 1988, H. Hofer (SMNK); 16, 8 October 1987, H. Hofer (SMNK); 161^,2 December 1987, H. Hofer (SMNK); Ic?, 3 October 1987, H. Hofer (SMNK); 19,1 February 1983, H. Hofer (SMNK); (Ilha da Marchantaria); 19 January 1988, H. Hofer (SMNK 946); (Fazenda Esteio ZF3-Km 23), 16, 25 February 1987, B.C. Klein (INPA); 1$, 7 November 1985, B.C. Klein (INPA); Id, 5 May 1985, B.C. Klein (INPA); Alto Solimoes, 19, 1 juv., December 1997, A. A. Lise (MCN 8894); Tefe (Esta^ao Ecologica do Mami- raua), Ic5'l9, 9-13 October 1992, S.H. Borges (MCN 22876); RondSnia: Porto Velho, 19, 15 April 1996, Eq. IBSP/SMNK (IBSP 8711); Bahia: 2632, Ceplac (MNRJ); Camacan (Fazenda Matia- pa), 29, 2 juvs., 16 October 1978, J.S. Santos (CE- PLAC); 19, 16 October 1978, J.S. Santos (CE- PLAC); 19, 16 October 1978, J.S. Santos (CEPLAC); Itamaraju, 19, Ceplac (MNRJ); (Fa- zenda Nossa Senhora das Neves); 3629, 14 Oc- tober 1978, J.S. Santos (CEPLAC); (Fazenda Pan Brasil), Id, 22 December 1969, Ceplac (MNRJ 13354); ld39, 20 June 1968, Ceplac (MNRJ 13388); Jugari 19, Ceplac (MNRJ); (Fazenda Ari- zona), Id, 4 March 1971, Ceplac (MNRJ); (Fazen- da Ribeirao do Antonio), 19,1 juv., 13 May 1970, Ceplac (MNRJ); (Fazenda Sao Francisco), 19, 26 November 1970, Ceplac (MNRJ); 3 9, 27 Novem- ber 1969, Ceplac (MNRJ); 26,27 November 1969, Ceplac (MNRJ 13345); Id, 24 September 1970, Ceplac (MNRJ 13062); 19, 8-9 April 1998, A.D. Brescovit et al. (IBSP 18576); (Fazenda Bethania), 29, 1 juv., 17 April 1971, Ceplac (MNRJ); Uru- giica. Id, 2 juvs., Ceplac (MNRJ 13381); (Fazenda Santa Tereza), 19, 21 October 1970, Ceplac (MNRJ); Jiigari, Reserva Natural da Serra do Tei- moso, 1 9, April 1998, A.D. Brescovit & R. Bertani (IBSP 18825); Ilheus (Ceplac), 19, 12 April 1998, A.D. Brescovit et al. (IBSP 18909); Lomanto Junior (Fazenda Mangabeira), 4d, 29 May 1968, Ceplac (MNRJ); Mascote (Fazenda Palestina), 2d, 11 May 1968, Ceplac (MNRJ); 7dl9, 11 May 1968, Ce- plac (MNRJ); Porto Seguro (Fazenda Sao Jorge), 19, 28 June 1970, Ceplac (MNRJ); Coaraci (Fa- zenda Boa Esperanga), 1 d, 24 March 1971, Ceplac (MNRJ); 19, 18 September 1970, Ceplac (MNRJ); 19,3 November 1970, Ceplac (MNRJ); 2 9, 17 Oc- tober 1970, Ceplac (MNRJ); 3d39, 28 January 1971, Ceplac (MNRJ); 19, 16 January 1971, Ce- plac (MNRJ); Mascote (Fazenda Palestina), lOd, 9 June 1968, Ceplac (MNRJ); Gandu (Fazenda Pedra Branca), Id, 5 February 1970, Ceplac (MNRJ); Prado (Fazenda Furado), Id, 26 September 1970, Ceplac (MNRJ); Espirito Santo: Sao Mateus (Re- serva Florestal Vale do Rio Doce), 1 9, 5-12 Jan- uary 1998, A.D. Brescovit et al, (IBSP 16955); 1 9, 5-12 January 1998, A.D. Brescovit et al. (IBSP 16758); Id, 4 juv., 5-12 January 1998, A.D. Bres- covit et al. (IBSP 21429); 19,7 juv., 5-12 January 1998, A.D. Brescovit et al. (IBSP 21436). Scytodes panamensis new species (Figs. 7, 8, 26-30) Types. — Male holotype from Fort Sher- man, Canal Zone, Panama, 15 August 1939, A.M. Chickering deposited in MCZ. Six male and 15 female paratypes deposited in MCZ and two male and three female paratypes de- posited in IBSP 24029, all with the same data as holotype. Etymology. — The specific name refers to the type locality. Diagnosis. — The males of S. panamensis differ from the other species, here included, by the dorsal rectangular hump on the male palpal bulb (Figs. 8, 27). The female differs from the other species by the transversal pair of oval seminal receptacles (Fig. 30). Male. — (Portobelo, Canal Zone, Panama). Carapace yellow with U-shaped dark brown pattern as shown in Fig, 26. Pedipalps yellow with brown longitudinal dorsal stripe and one or two scattered brown stains. Labium and en- dites yellow. Sternum cream colored with brown margins at base of legs and along slight grooves that extend towards center. Legs yel- low with pair of brown longitudinal ventral stripes along femur and single brown longi- tudinal dorsal stripe along tibia and metatar- sus. Abdomen grayish with two or three pos- terior longitudinal black stripes and few anterior scattered black stains (Fig. 30). Total length 4.50. Carapace slightly domed, 2.25 long, 1.88 wide. Eye diameters: PME 0.14, ALE 0.14, PEE 0.14. Lateral eyes and chelic- erae as in S. championi . Labium 0.24 long, 0.28 wide. Sternum 1.24 long, 0.92 wide. Leg measurements: I - femur 5.13/ patella 0.63/ tibia 5.63/ metatarsus 7.38/ tarsus 0.88/ total 19.65/ II - 3.63/ 0.63/ 4.00/ 4.38/ 0.75/ 13.39/ III - 2.50/ 0.50/ 2.13/ 2.63/ 0.63/ 8.39/ IV - 3.63/ 0.63/ 3.38/ 3.75/ 0.75/ 12.14. Palpal fe- mur as in S. championi. Cymbium with single distal spine (Fig. 27). Bulb 0.56 long, distal area with ventral spoon-shaped, slightly scler- otized pocket (P, Fig. 27). Abdomen 2.25 long, 1.38 wide, rounded, covered with slen- der hairs. Female* — (Portobelo, Canal Zone, Pana- ma). Coloration as in male. Total length 4.75. Carapace slightly domed, 2.38 long, 2.00 wide. Eye diameters: PME 0.14, ALE 0.14, BRESCOVIT & RHEIMS— THE GENUS SCYTODES 319 Figures 26-35. — 26-30. Scytodes panamensis new species. 26. Male body, dorsal view; 27. Male palp, retrolateral view; 28. Prolateral view; 29. Female epigynum, ventral view; 30. Dorsal view. 31-35, Scy- todes guttipes Simon. 31. Male body, dorsal view; 32. Male palp, retrolateral view; 33. Prolateral view; 34, Female epigynum, ventral view; 35. Dorsal view. Scale lines = 0.25 mm. PLE 0.16. Lateral eyes and chelicerae as in male. Labium 0.30 long, 0.36 wide. Sternum 1.32 long, 0.96 wide. Leg measurements: I - femur 3.00/ patella 0.50/ tibia 3.00/ metatarsus 3.75/ tarsus 0.50/ total 10.75/ II - 2.25/ 0.50/ 2.38/ 2.63/ 0.63/ 8.39/ III - 1.63/ 0.38/ 1.50/ 1.75/ 0.50/ 5.76/ IV - 2.25/ 0.63/ 2.38/ 2.38/ 0.50/ 8.14. Epigynal fovea semicircular, wide- 320 THE JOURNAL OF ARACHNOLOGY ly separated from each other. Positioning ridge semicircular (Fig. 29). Internal genitalia pre- senting posterior pair of seminal receptacles with long ducts (Fig. 30). Abdomen 2.38 long, 2.25 wide, as in male. Variation. — Males: Total length 3.63-5.75; carapace 2.00-2,75; femur I 4.13-6.50; bulb 0.48-0.68 {n = 15.) Females: Total length 4.25-5.75; carapace 2.38-2.88; femur I 2.50- 3.88 {n = 20). Some males with a single row of spines along the ventral face of the tibia 1. Genitalic morphology constant. Distribution. — Canal Zone, Panama. Material examined. — PANAMA; Canal Zone: Gatun, 1?, 11 jiivs., 15 February 1958, A.M. Chickering (MCZ); Fort Sherman, 73 1 1 9, August 1939, A.M. Chickering (MCZ); Fort Gulik, 2d, September 1979, H.J. Harlan (AMNH); Portobelo, 9d 189, lOjuvs., 12 August 1936, A.M. Chickering (MCZ). Scytodes guttipes Simon (Figs. 31-35) Scytodes guttipes Simon 1892: 438, pi. 9, fig. 13 (3d, 19, 5 immature syntypes from Venezuela, with no definite locality, deposited in MNHN AR1223, examined. Lectotype d and 2d, 9 and 5 immature paralectotypes hereby designated). Roewer, 1942: 329. Diagnosis. — Scytodes guttipes differs from the other species by the presence of a devel- oped dorsal membrane in the distal area of the male palpal bulb (Figs. 32-33) and by the slightly sclerotized projection between the fe- male seminal receptacles (Fig. 35), Male. — (Lectotype). Carapace yellow with dark pattern as shown in Fig. 31. Labium and endites yellow with brownish margins. Pedi- palps, sternum and legs yellow with black stains. Abdomen cream colored with few black transversal scattered stains. Total length 4.50. Carapace, 2.25 long, 1.75 wide. Eye di- ameters: PME 0.10, ALE 0.12, PLE 0.12. Lat- eral eyes and chelicerae as in S. championi. Labium 0.26 long, 0.20 wide. Sternum 1.14 long, 0.74 wide. Leg measurements: I - femur 6.88/ patella 0.63/ tibia 7.00/ metatarsus 9.00/ tarsus 0.75/ total 24.26/ 11 - 4.50/ 0.50/ 4.50/ 5.25/ 0.63/ 15.38/ III - 3.00/ 0.50/ 2.50/ 3.38/ 0.63/ 10.01/ IV - 4.25/ 0.50/ 4.13/ 4.63/ 0.75/ 14.26. Palpal femur as S. championi. Cym- bium with strong distal spine (Fig. 32). Bulb 0.54 long, medially narrowed (Fig. 33). Distal area with median ventral lance-shaped pocket (P; Fig. 32). Abdomen 2.25 long, 1.75 wide, rounded, covered with slender hairs. Female. — (Paralectotype). Coloration as in male. Total length 4.88. Carapace 2.50 long, 2.13 wide. Eye diameters PME 0.12, ALE 0.12, PLE 0.14. Lateral eyes and chelicerae as in male. Labium 0.32 long, 0.30 wide. Ster- num 1.38 long, 1.00 wide. Leg measurements: I - femur 4.38/ patella 0,63/ tibia 4.25/ meta- tarsus 5.63/ tarsus 0.75/ total 15.64/ II - 3.50/ 0.50/ 3.13/ 3.88/ 0.63/ 11.64/ III - 2.13/ 0.50/ 1.75/ 2.38/ 0,50/ 7,26/ IV - 2.75/ 0.50/ 2.50/ 3.25/ 0.75/ 9.75. Epigynal fovea narrow and semicircular. Positioning ridge semicircular (Fig. 34). Internal genitalia with two pairs of oval seminal receptacles (Fig. 35). Abdomen 2.38 long, 2.25 wide, as in male. Variation* — Males: Total length 3.88- 4.50; carapace 2.13-2.25; femur I 4.88-6.88 (/I = 2). Distribution. — Venezuela and Trinidad & Tobago. Other material examined. — TRINIDAD & TOBAGO: Mount St. Benedict (10°39'49"N, 61°23'56"W), Id, 27-30 June 1999, R. Pinto-da- Rocha (MZSP 18880). Scytodes cogu new species (Figs. 36-40) Scytodes championi: Valerio 1981: 86—87 (Mis- identification, only figs. 17 and 29). Types. — Male holotype, 5 9 and 5 imma- ture paratypes from San Jose, San Jose Prov- ince, Costa Rica, E. Schmidt, deposited in AMNH; and 2 9 paratypes, with the same data, deposited in IBSP 24026, Etymology. — Short for ''cogumelo.” Bra- zilian word for mushroom, due to the shape of one pair of seminal receptacles. Diagnosis* — The male of Scytodes cogu re- sembles S. vaiirieorum by the pronounced groove with lateral projections in the apex of the distal area of the male palpal bulb (Figs. 37, 50) but differs by the presence of short and narrowed median ventral projection and absence of slightly sclerotized membrane (Fig. 38). The females differ from the other species by the presence of a pair of anterior mushroom-like seminal receptacles and a pair of posterior curved truncated ones (Fig. 40). Male. — (Holotype). Carapace light brown with brown pattern as shown on Fig. 36. Ped- BRESCOVIT & RHEIMS— THE GENUS SCYTODES 321 Figures 36-45. — 36-40. Scytodes cogu new species. 36. Male body, dorsal view; 37. Male palp, pro- lateral view; 38. Retrolateral view; 39. Female epigynum, ventral view; 40. Dorsal view. 41-45. Scytodes vaurieorum new species. 41. Male carapace, dorsal view; 42. Male palp, prolateral view; 43. Retrolateral view; 44. Female epigynum, ventral view; 45. Dorsal view. Scale lines = 0.25 mm. ipalps yellow with dorsal longitudinal stripe. Labium and endites yellow. Sternum yellow with brown margins at base of legs and along slight grooves extending towards center. Legs yellow with scattered longitudinal stains. Ab- domen grayish with black pattern of transver- sal stripes with few scattered black spots be- tween them and lateral black stains (Fig. 36). Total length 4.25. Carapace 2.00 long, 1.63 wide. Eye diameters: PME 0.14, ALE 0.14, PLE 0.14. Lateral eyes and chelicerae as in S. championi. Labium 0.18 long, 0.28 wide. Sternum 1.20 long, 0.94 wide. Leg measure- ments: I - femur 3.00/ rest of leg absent/ II - 322 THE JOURNAL OF ARACHNOLOGY femur 2.75/ patella 0.63/ tibia 2.88/ metatarsus 3.63/ tarsus 0.63/ total 10.52/ III - 1.75/ 0.50/ 1 .63/ 2.00/ 0.63/ 6.5 1/ IV - 2.63/ 0.63/ absent/ absent. Palpal femur as in S. championi. Cym- bium with slender distal spine (Fig. 37). Bulb 0.54 long, distal area with dorsal, sclerotized membrane (Fig. 37). Abdomen 2.25 long, 1.38 wide, rounded, covered with slender hairs. Female. — (Paratype). Coloration as in male. Total length 4.38. Carapace 2.13 long, 1.75 wide. Eye diameters: PME 0.14, ALE 0.12, PEE 0.14. Eateral eyes and chelicerae as in male. Eabium 0.20 long, 0.24 wide. Ster- num 1.12 long, 0.92 wide. Eeg measurements: I - femur 2.13/ patella 0.50/ tibia 2.38/ meta- tarsus 2.88/ tarsus 0.63/ total 8.52/ II - 1.88/ 0.50/ 1.88/ 2.00/ 0.50/ 6.76/ III - 1.38/ 0.38/ 1.13/ 1.13/ 0.50/ 4.52/ IV - 2.00/ 0.50/ 1.75/ 1.88/ 0.63/ 6.73. Epigynal fovea semicircular, shalow. Positioning ridge semicircular (Pig. 39). Abdomen 2.25 long, 2.25 wide, as in male. Variation. — Females: Total length 4.25- 5.00; carapace 2.13-2.75; femur I 2.13-2.75 in = 7). Distribution. — Costa Rica. Material examined. — COSTA RICA: Three minutes south Liberia, Guanacaste Province, 1 $, 10 July 1966, S. Peck (AMNH). Scytodes vaurieorum new species (Pigs. 41-45) Types. — Male holotype from San Jeronimo Department, Guatemala, 24—26 July 1947, C. & P. Vaurie; and female paratype from the same locality, 26-27 July 1947, C. & P. Vaurie deposited in AMNH, Etymology. — The specific name is a pa- tronym in honor of the collectors of the types. Diagnosis. — Scytodes vaurieorum differs from S. cogii by the presence of a slightly sclerotized membrane located all around the distal area and by a finger-like dorsal projec- tion (Fig. 43). The female differs from the oth- er species by the presence of a sinuous posi- tioning ridge (Fig. 44) and by a pair of small seminal receptacles with long ducts (Fig. 45). Male. (Holotype). — Carapace light brown with brown pattern as shown on Fig. 41. Ped- ipalps light brown with few ventral spots. La- bium and endites yellow with brownish mar- gins. Sternum yellow with brown margins at base of legs and along slight groves extending towards center. Legs yellowish with many ventral scattered black spots, except on tarsi. Abdomen grayish. Total length 4.13. Carapace slightly domed, 2.13 long, 1.88 wide. Eye di- ameters: PME 0.12, ALE 0.10, PEE 0.12. Lat- eral eyes and chelicerae as in S. championi. Labium 0.14 long, 0.24 wide. Sternum 1.18 long, 0.88 wide. Leg measurements: I - femur 2.63/ patella 0.50/ tibia 3.13/ metatarsus 3.75/ tarsus 0.63/ total 10.64/ II - 2.13/ 0.50/ 2.13/ 2.38/ 0.63/ 5.88/ III - 1.50/ 0.50/ 1.25/ 1.50/ 0.50/ 5.25/ IV - 2.13/ 0.50/ 2.13/ 2.13/ 0.63/ 7.52. Palpal femur as in S. championi. Cym- bium with single apical slender spine (Pig. 42). Bulb 0.52 long, strongly curved inwards with distal area presenting prolateral concav- ity (Pig. 42). Abdomen 2.00 long, 1.75 wide, rounded, covered with slender hairs. Female. (Paratype). — Coloration as in male. Total length 4.63 Carapace slightly domed, 2.13 long, 1.88 wide. Eye diameters: PME 0.14, ALE 0.12, PLE 0.12. Lateral eyes and chelicerae as in male. Labium 0.24 long, 0.24 wide. Sternum 1.18 long, 0.90 wide. Leg measurements: I - femur 2.25/ patella 0.50/ tibia 2.38/ metatarsus 2.75/ tarsus 0.63/ total 8.51/ II - 1.75/ 0.50/ 1.75/ 2.13/ 0.50/ 6.63/ III - 1.25/ 0.50/ 1.13/ 1.00/ 0.38/ 4.26/ IV - 1.75/ 0.50/ 1.75/ 1.75/ 0.63/ 6.38. Epigynal fovea inconspicuous (Pig. 44). Internal geni- talia with pair of seminal receptacles on each side. Larger globose pair covering ducts of smaller pair and with lateral sclerotized area (Pig. 45). Abdomen 2.50 long, 2.25 wide, as in male. Distribution. — Known only from the type locality. Material examined. — Only the types. Scytodes tegucigalpa new species (Figs. 9, 10, 46-48) Types. — Male holotype from Tegucigalpa, Francisco Morazan Department, Honduras, November 1959, N.H.L. Krauss, deposited in AMNH. Id (IBSP 24027) and 1 d and 3 im- mature (AMNH) paratypes from same locality as holotype, 14 July 1948, Clark. Etymology. — The specific name is a noun in apposition taken from the type locality. Diagnosis. — The male of S. tegucigalpa re- sembles S. chiquimula by the presence of an apical projection on the bulb (Figs. 50, 51) but differs by the presence of bifid distal area and dorsal groove (Figs. 10, 47, 48). BRESCOVIT & RHEIMS— THE GENUS SCYTODES 323 i! Figures 46-53. — 46-48. Scytodes tegucigalpa new species. 46. Male carapace, dorsal view; 47. Male palp, prolateral view; 48. Retrolateral view. 49-53. Scytodes chiquimula new species. 49. Male carapace, dorsal view; 50. Male palp, prolateral view; 51. Retrolateral view; 52. Female epigynum, ventral view; 53. Dorsal view. Scale lines = 0.25 mm. Male, (Holotype). — Carapace yellow with brown pattern as shown in Fig. 46. Pedipalps yellow. Labium and endites yellow. Sternum yellow with brown margins at base of legs and along slight grooves extending toward center. Legs yellow with pair of ventral longitudinal stripes along femora and few scattered longi- tudinal stains along tibiae. Abdomen grayish. Total length 4.63. Carapace slightly domed, 2.25 long, 1.88 wide. Eye diameters: PME 0.12, ALE 0.12, PLE 0.14. Lateral eyes and chelicerae as in S. championi. Labium 0.28 long, 0.24 wide. Sternum 1.34 long, 0.98 wide. Leg measurements: I - femur 4.25/ pa- tella 0.50/ tibia 5.00/ metatarsus 6.13/ tarsus 0.75/ total 16.63/ II - 3.00/ 0.50/ 3.25/ 3.75/ 324 THE JOURNAL OF ARACHNOLOGY 0.75/ 11.25/ III - 2.13/ 0.50/ 1.88/ 2.13/ 0.75/ 7.39/ IV - 3.00/ 0.63/ 2.88/ 3.13/ 0.75/ 10.39. Palpal femur as in S, championi. Cymbium with single slender distal spine (Fig. 47). Bulb 0.56 long, distal area with slightly sclerotized dorsal membrane (Figs. 10, 47) and a short ventral triangular projection (Fig. 48). Abdo- men 2.38 long, 2.00 wide, rounded, covered with slender hairs. Female. — Unknown. Variation. — Males: Total length 4.63-5.00; carapace 2.25-2.75; femur I 4.25-5.88; bulb 0.56-0.64 {n = 3). Distribution. — Known only from the type locality. Material examined. — Only the types. Scytodes chiquimula new species (Figs. 49-53) Types, — Male holotype, 1 $ and 1 imma- ture paratype from Chiquimula (1250 ft.), Chiquimula Department, Guatemala, 21-23 July 1947, C. & P. Vaurie, deposited in AMNH, and female paratype, with the same data as holotype, deposited in IBSP 24028. Etymology, — The specific name is a noun in apposition taken from the type locality. Diagnosis. — The male of S. chiquimula dif- fers from S. tegucigalpa by the presence of a retrolateral curved lamina apically projected (Fig. 51) and by a distal triangular laminar projection (Fig. 50) on the distal area of palpal bulb. The female resembles S. cogu by a pair of anterior mushroom-like seminal receptacles but differs by the short ducts and the elliptical posterior pair (Fig. 53). Male. (Holotype). — Carapace light brown with brown pattern as shown on Fig. 49. Ped- ipalps yellow. Labium and endites yellow. Sternum yellow with brown margins at the base of each leg. Legs yellowish with many black small ventral longitudinal stains. Ab- domen grayish. Total length 4.88 Carapace slightly domed, 2.63 long, 2.00 wide. Eye di- ameters: PME 0.14, ALE 0.14, PLE 0.14. Lat- eral eyes and chelicerae as in S. championi. Labium 0.28 long, 0.30 wide. Sternum 1.40 long, 1.00 wide. Leg measurements: I - femur 4.00/ patella 0.63/ tibia 4.63/ metatarsus 5.63/ tarsus 0.75/ total 15.64/ II - 3.00/ 0.50/ 3.13/ 3.75/ 0.63/ 1 1.01/ III - 2.13/ 0.50/ 1.75/ 2.25/ 0.63/ 7.26/ IV - 3.00/ 0.63/ 3.13/ 2.88/ 0.63/ 10.27. Palpal femur with stridulatory pick short and strong with rounded and projected socket. Cymbium with single slender apical spine (Fig. 50). Bulb 0.60 long. Abdomen 2.25 long, 1.63 wide, rounded, covered with slender hairs. Female. (Paratype, AMNH). — Coloration as in male. Total length 3.88. Carapace domed, 2.38 long, 2.13 wide. Eye diameters: PME 0.12, ALE 0.12, PLE 0.12. Lateral eyes and chelicerae as in male. Labium 0.20 long, 0.22 wide. Sternum 1.26 long, 0.88 wide. Leg measurements: I - femur 3.13/ patella 0.63/ tibia 2.88/ metatarsus 4.00/ tarsus 0.75/ total 1 1.39/ II - 2.75/ 0.63/ 2.38/ 2.88/ 0.63/ 9.27/ III - 1.88/ 0.63/ 1.50/ 1.88/ 0.63/ 6.52/ IV - 2.63/ 0.63/ 2.63/ 2.75/ 0.75/ 9.39. Fovea in- conspicuous. Positioning ridge semicircular (Fig. 52). Abdomen 1.50 long, 1.75 wide, as in male. Variation, — Females: total length 3.88- 5.25; carapace 2.38-2.63; femur I 2.50-3.13 {n = 2). Distribution. — Known only from the type locality. Material examined. — Only the types. Scytodes zamorano new species (Figs. 54-56) Scytodes championi: Valerio 1981: 86 (misidenti- fication, only fig. 18). Types. — Female holotype and female par- atype from Zamorano, El Paraiso Department, Honduras, September 1953, N.H.L. Krauss deposited in AMNH. Etymology, — The specific name is a noun in apposition taken from the type locality. Diagnosis, — The female of S. zamorano differs from the other species herein included by the sac-like positioning ridge (Figs. 55, 56). Male. — Unknown. Female (paratype). — Carapace yellow with brown pattern (Fig, 54). Pedipalps yellow with longitudinal brown stains. Labium and endites yellow. Sternum yellow with brown margins at base of legs extending towards center along slight grooves. Legs yellow with pair of brown longitudinal stripes along ven- tral face of femora and few scattered longi- tudinal stains along tibiae. Abdomen uniform- ly gray. Total length 5,25. Carapace slightly domed, 2.75 long, 2.00 wide. Eye diameters: PME 0.14, ALE 0.14, PLE 0.14. Lateral eyes BRESCOVIT & RHEIMS— THE GENUS SCYTODES 325 and chelicerae as in S. championi. Labium 0.28 long, 032 wide. Sternum 1.42 long, 1,08 wide. Leg measurements: I “ femur 2,88/ pa- tella 0.63/ tibia 3,13/ metatarsus 3.75/ tarsus absent/ total 10,39/ II - 2.25/ 0.50/ 2.38/ 2.75/ 0.75/ 8,63/ III - 1.63/ 0.50/ 1.50/ 1.75/ 0.63/ 6.01/ IV - 2.25/ 0.63/ 2,38/ 2.38/ 0.75/ 8.39. Epigynal fovea very deep (Fig. 55). Internal genitalia with pair of anterior globose seminal receptacles and pair of posterior oval seminal receptacles with slender, strongly curved ducts (Fig. 56). Abdomen 2.50 long, 2,13 wide, rounded, covered with slender hairs. Variation. — Females: Total length 5,00= 5.25; carapace 2.50-2.75; femur I 2.88-3.13 (/I = 2). Distribution. — Known only for the type lo- cality. Material examined. — Only the types. Scytodes armata new species (Figs, 11, 12, 57-61) Scytodes championi: Valerio 1981: 87 (misidenti- fication, only fig. 28). Types*— Male holotype from La Selva, Puerto Viejo, Heredia, Costa Rica, December 1980, W. Eberhard; Id paratype from same locality, February 1981, W. Eberhard; Id par- atype from Cahuita, Limon, Costa Rica, 30 March 1979, J. Coddington; and 3 9 and 1 im- mature paratypes from Monteverde Commu- nity, Puntarenas, Costa Rica, July 1978, C.L. Kraig & P. Klass, all deposited in MCZ. Etymology.— The specific name refers to the strong ventral spines along male legs I and IL Diagnosis. — The males of Scytodes armata resemble those of 5, univittata Simon 1882 (see Brescovit & Rheims 2000, fig. 16) by the double row of spines along ventral face of the femur I but differ by a double row of spines also along femur II (Fig. 12). It differs from the other species, as well as S, univittata by the presence of a tubular retrolateral projec- tion on the distal area of the male palpal bulb (Fig. 58). The females resemble those of S. gertschi by the pair of anterior, rounded, mushroom-like seminal receptacles but differ by straight posterior area of epigynal plate (Fig. 60) and rounded shape of anterior pair of seminal receptacles (Fig. 61). Male (holotype), — Carapace yellow with brown pattern (Fig. 57). Pedipalps yellow with few prolateral brownish stains. Labium and endites yellow with brownish margins. Sternum yellow with brown margins at base of legs, extending towards center along slight grooves. Legs yellow with brown longitudinal stains along ventral face of the femora and very slightly along tibiae. Abdomen cream colored with dark brown pattern of posterior median transversal stripes (Fig. 57). Total length 438. Carapace slightly domed, 2.38 long, 2.00 wide. Eye diameters: PME 0.16, ALE 0.16, PLE 0.16. Lateral eyes and chelic- erae as in 5. championi. Labium 0.26 long, 0.28 wide. Sternum 1.30 long, 1.00 wide. Leg measurements: I - femur 5.75/ patella 0.63/ tibia 5.63/ metatarsus 8.25/ tarsus 0,88/ total 21.14/ II - 4,13/ 0.63/ 3.75/ 5.13/ 0.75/ 14.39/ III - 2,50/ 0,63/ 2.25/ 3.00/ 0.63/ 9.01/ IV - 3.50/ 0.63/ 338/ 4.50/ 0.75/ 12,76, Ventral faces of femora I-II with double row of spines, prolateral row strong and twice as long as less developed retrolateral row (Fig, 10). Palpal fe- mur with stridulatory pick as in S. championi. Cymbium with single distal spine (Fig. 58). Bulb 0.50 long. Distal area bifid (Figs. 58, 59). Abdomen 2.00 long, 1.63 wide, rounded, covered with slender hairs. Female (paratype). — Coloration as in male. Total length 6.25. Carapace slightly domed, 3.50 long, 2.75 wide. Eye diameters: PME 0.16, ALE 0.14, PLE 0.14. Lateral eyes and chelicerae as in male. Labium 0.38 long, 0,32 wide. Sternum 1.80 long, 130 wide. Leg mea- surements: I - femur 4.00/ patella 0.75/ tibia 4.38/ metatarsus 5.75/ tarsus 0.88/ total 15.76/ II - 4.75/ 0.75/ 3.25/ 4.13/ 0.88/ 13.76/ III - 2,38/ 0,75/ 2.13/ 2.63/ 0.75/ 8,64/ IV - 3.50/ 0.63/ 3.25/ 3.63/ 0.88/ 11.89. Epigynal fovea conspicuous and deep. Positioning ridge semi- circular (Fig. 60). Internal genitalia with pair of posterior seminal receptacles positioned close together with lateral sclerotized area (Fig. 61). Abdomen 2.75 long, 2.63 wide as in male. Variation* — Males: total length 4.38-5.50; carapace 2.38-2.75; femur I 5.75-6.88; bulb 0.50-0.54 {n = 4). Females: Total length 6.25-6.75 in - 3). Distribution* — Costa Rica. Material examined.-COSTA RICA: Puntar- enas Province, Monteverde Community (1480 m), 16, July 1978, C.L. Craig & P. Class (MCZ); He- 326 THE JOURNAL OF ARACHNOLOGY Figures 54-61. — 54-56. Scytodes zamorano new species. 54. Female carapace, dorsal view; 55. Female epigynum, ventral view; 56. Dorsal view. 57-61. Scytodes armata new species. 57. Male body, dorsal view; 58. Male palp retrolateral view; 59. Prolateral view; 60. Female epigynum, ventral view; 61. Dorsal view. Scale lines = 0.25 mm. redia, La Selva near Puerto Viejo, 16, 18 January 1979, J. Coddington (MCZ). Scytodes gertscM Valerio (Figs. 62-66) Scytodes gertschi Valerio 1981: 86, figs. 6, 15, 27, 30 (male holotype and female allotype from Bar- ro Colorado Island, Canal Zone, Panama, June 1950, A.M. Chickering deposited in MCZ, not examined) (examined by Valerio); Platnick 1989: 117. Diagnosis. — The male of S. gertscM differs from the other species by the presence of a dorsaLventrally elongated bulb (Fig. 64) and a dorsal triangular projection on the distal area of the male palpal bulb (Fig. 63). The female differs from the other species by the invagi- nated epigynal plate (Fig. 65) and subrectan- BRESCOVIT & RHEIMS— THE GENUS SCYTODES 327 Figures 62-66. — Scytodes gertschi Valerio. 62. Male carapace, dorsal view; 63. Male palp, prolateral view; 64. Retrolateral view; 65. Female epigynum, ventral view; 66. Dorsal view. Scale lines = 0.25 mm. gular pair of seminal receptacles (Fig. 66; VaF erio 1981: fig. 15). Variation, — Males: total length 3.00-3.75; carapace 1.63-1.88; bulb 0.36-0,40 {n = 2). Females: total length 3.63-4.50; carapace 1.75-2.00; femur I: 1.38-1.88 {n = 7). Material examined. — PANAMA^ Canal Zone: Barro Colorado Island, 19,11 February 1936, W.J. Gertsch (AMNH); Balboa, 237 9 & 17 juvs.. May 1964, A.M. Chickering (MCZ). Scytodes cubensis Alayon (Figs. 67, 68) Scytodes cubensis Alayon 1977: 2, figs, la-c (fe- male holotype from Loma Montecristi, Colorado, 328 THE JOURNAL OF ARACHNOLOGY Figures 67-68. — Scytodes cubensis Alayon. 67. Male palp, retrolateral view; 68. Prolateral view. Scale lines = 0.25 mm. Limonar, Matanzas, Cuba, March 1976, L.R. Her- nandez; and several male and female paratypes, deposited in Academia de Ciencias de Cuba, not examined); Brignoli 1983: 149. Male. — Described by Alayon 1977: 2, figs, la, b. Cymbium of palp with three slender dis- tal spines (Fig, 67). Bulb long, distal area with a small dorsal membrane on base of long, fi- liform embolus (Figs. 67, 68). New records.— TRINIDAD & TOBA^ GO: Mount St. Benedict (10°39'49"N, 6r23'56"W), 1629 (possible females, lack- ing abdomen), 27-30 June 1999, R.Pinto-da- Rocha (MZUSP 18860), ACKNOWLEDGMENTS We wish to thank the curators for loaning material for this study. Prof. Pedro Kiyohara and Miss Simone Perche de Toledo (IF/USP) for making the scanning electron micrographs. Martin J, Ramirez and Adalberto J. dos Santos for helpful comments on the manuscript. This work was supported by CNPq and ‘Tundagao de Amparo a Pesquisa do Estado de Sao Pau- lo” (FAPESP No. 99/05446-8; 00/00247-6). LITERATURE CITED Alayon, G. 1977. Nuevas especies de Scytodes La- treille, 1804 (Araneae, Scytodidae) de Cuba. Poeyana 177:1-20. Alayon, G. 1985. El genero Scytodes Latreille, 1804 (Araneae: Scytodidae), en Jamaica, com la descripcion de dos nuevas especies. Poeyana 293:1-10. Alayon, G. 1992. Nueva especie de Scytodes y des- cripcion del macho de S. noeli (Araneae: Scy- todidae). Poeyana 413:1-7. Banks, N. 1929. Spiders from Panama. Bulletin of the Museum of Comparative Zoology 69: 53-95. Brescovit, A.D. & H. Hofer. 1999. Four new spe- cies of litter inhabiting Scytodes spiders (Ara- neae, Scytodidae) from Amazonas. Studies on Neotropical Fauna & Environment, 34:105- 1 13. Brescovit, A.D. & C.A. Rheims. 2000. On the synanthropic species of the genus Scytodes La- treille (Scytodidae, Araneae) of Brazil, with synonymies and records of these species in other Neotropical countries. Bulletin of the British Arachnological Society 1 1 (8):320-330. Brignoli, P.M. 1976. Beitrage zur Kenntnis der Scytodidae (Araneae). Revue Suisse de Zoologie 83(1):125-19L Brignoli, P.M. 1983. A Catalogue of The Araneae Described Between 1940 and 1981. Manchester, Manchester Acad., 755 p. Caporiacco, L. di. 1947. Diagnosi preliminari di specie nuove di Aracnidi della Guiana Britanni- ca, reccolte dai professor! Beccari e Romiti. Monitore Zoologico Italiano 56:20-34. Caporiacco, L. di. 1948. Arachnida of British Gui- ana collected in 1931 and 1936 by Professors Beccari and Romiti. Proceedings of Zoology of the Society of London 1 18(3):607-747. Hofer, H. 1990. The spider community (Araneae) of a central Amazonian blackwater inundation forest (igapo). Acta Zoologica Fennica 190:173- 179. Nentwig, W. 1993. Spiders of Panama. Flora & Fauna Handbook n°l2. The Sandhill Crane Press, Inc., 274 p. Pickard-Cambridge, F.O. 1899. Arachnida - Ara- neida. Pp. 48-52. In Biologia Centrali-America- na. Godman & Salvin, London, vol. 2, Platnick, N.I. 1989. Advances in Spider Taxonomy 1981-1987. Manchester Univ. Press, Manches- ter., 673 p. Rheims, C.A. & A.D. Brescovit. 2000. Six new species of Scytodes Latreille 1 804 (Araneae, Scy- todidae) from Brazil. Zoosystema 22(4):721- 731. Roewer, C.F. 1942. Katalog der Araneae von 1758 bis 1940. Bruxellas, vol. 1, Pp. 1-1040. BRESCOVIT (fe RHEIMS— THE GENUS SCYTODES 329 Simon, E. 1892. Voyage de M. Simon en Vene- zuela, Arachnida. Annales de la Societe. Ento- mologique de France 61:423-462. Valerio, C.E. 1981. Spitting spiders (Araneae, Scy- todidae, Scytodes) from Central America. Bulle- tin of the American Museum of Natural History 170:80-89. Manuscript received 8 May 2000, revised 9 March 200L 2001. The Journal of Arachnology 29:330-344 A REVIEW OF THE CHINESE PSECHRIDAE (ARANEAE) Xin-Ping Wang: Department of Entomology, California Academy of Sciences, Golden Gate Park, San Francisco, California 94118 USA Chang-Min Yin: Hunan Biological Institute, Hunan Teachers University, Changsha, Hunan 410008 China ABSTRACT. The Chinese psechrid spiders of the genera Fecenia and Psechnis are reviewed. The species Fecenia haincmensis is newly synonymized with F. cylindnita. The species P. miniiis is considered a nomen dubium. The species P. senociilata is regarded as a valid species. The male is newly described for P. tingpingensis. Three new species are described: P. jinggangensis new species, P. rani new species, and P. taiwanensis new species. In all, nine psechrid species are recognized from China. The spinnerets, trichobothria, and tarsal organ morphology of P. tingpingensis are presented. A key to Chinese Psechnis species is also provided. Keywords: Psechridae, Psechnis, Fecenia, China Psechrid species of the genera Fecenia Si- mon 1887 and Psechnis Thorell 1878 are widespread from China (north to Qinling Mt., Shaanxi) and southeast Asia to New Guinea, with approximately 19 valid species (Platnick 2000). A revision of this family was pre- sented by Levi (1982), who gave detailed di- agnoses, illustrations, and descriptions of the family, genera, and species. Levi’s revision (1982) enabled further work on the species of this family possible (e.g., Murphy 1986; Yin, Wang & Zhang 1985). To date, seven psechrid species have been reported from China (Song, Zhu & Chen 1999): P. ghecu- anus Thorell 1897; P. kunmingensis Yin, Wang & Zhang 1985; P. minus Chamberlin 1924; P. sinensis Berland & Berland 1914; P, tingpingensis Yin, Wang & Zhang 1985; Fecenia cylindrata Thorell 1895; and F. haincmensis Wang 1990. The presence of P. torviis (O. R-Cambridge 1869) in Taiwan (Lee 1966; Hu 1984) was shown to be a mis- identification (Chen 1996; Song, Zhu & Chen 1999). Further collection and study of Chinese psechrids made this revision possible. In this paper, nine psechrid species are recognized from China. The species Fecenia hainanensis is newly synonymized with F. cylindrata. The species P. mimus, which was described based on an unidentifiable juvenile female (Chamberlin 1924), is considered a nomen dubium, and therefore the species P. seno- ciilata is removed from its synonymy. The male is newly described for P. tingpingensis. The female previously identified as P. sinen- sis by Levi (1982) is shown to be a new spe- cies. Three new species described in this study are: P. jinggangensis', P. rani; and P. taiwanensis. METHODS All measurements are in mm. All scales are 0.2 mm length. Leg measurements are shown as: total length (femur, patella + tib- ia, metatarsus, tarsus). The terms used in the genitalic descriptions follow Levi (1982). Because of the similar body color pattern, stable number of cheliceral teeth, and simi- lar leg spine distributional pattern at spe- cies-level, the species descriptions are fo- cused on the male and female genitalic structures. The material used in this study was based on collections made available through the courtesy of the following indi- viduals and institutions: N.L Platnick, American Museum of Natural History, New York, USA (AMNH); J. Margerison, The Natural History Museum, London, UK (BMNH); C.M. Yin, Hunan Biological In- stitute, Changsha, Hunan, China (HBI); 330 WANG & YIN— CHINESE PSECHRIDAE 331 M.S. Zhu, Hebei Teachers University, Shi= jiazhuang, China (HTU); J. Chen, Institute of Zoology, Beijing, China (IZB); P. Paetiei, Museo de Bergamo, Bergamo, Italy (MCB); C. Rollard, Museum National d'Histoire Na= turelle, Paris, France (MNHN). SPINNERETS, TRICHOBOTHRIA AND TARSAL ORGAN MORPHOLOGY A representative species, Psechrus tingpin- gensis, was chosen here for detailed spinner^ ets, trichobothria, and tarsal organ descrip- tions in order to form a basis for further comparison with other psechrids and also with other families in future study. This spe- cies was selected for the reason of well-pre- served spinnerets in the examined psechrid species and large numbers of available spec- imens. Cribellum large, divided, female with nu- merous spigots (Figs. 39, 40), male without spigots (Fig. 41). According to a study by Zhang et al. (1998) of the female juvenile cribellum of P. mimus (sensu Zhang et ah 1998), “there was still not any spigot visible on the seventh day of molting; there were few small spigots in the middle area of cri- bellum on the ninth day of molting, and many spigots appeared on the eleventh day juveniles but still no distinct segment.” Apex of anterior lateral spinneret (ALS) with two major ampullate spigots (MAP) at mesal margins, many short piriform spigots in both male and female; posterior mediae spinneret (PMS) strongly curved back an- teriorly (Fig. 36), with spigots situated on distal half of the segment, one minor am- pullate spigots (mAP) on distal end, 40=50 acieiform spigots in both male and female, and 11--12 cylindrical spigots (as shown in short arrows) in female arranged in two rows; posterior lateral spinneret (PLS) with approximately 30 aciniform spigots in both male and female, and at least 16 cylindrical spigots (as shown in short arrows) in female (Figs. 42=47). Trichobothrial base with hood transversely striated (Fig. 37). Tarsal organ oval to round (Fig. 38), situated dor- sally on distal tarsus, slightly anterior of most distal trichobothrium. KEYS TO CHINESE PSECHRUS SPECIES Males 1. Palpal femur modified with notch (Figs. 21, 26, 33) .................................. 2 Palpal femur without such modification ............................................ 4 2. Conductor base enlarged, with small tubercles (Fig. 19) .......................... senoculata Conductor base not enlarged, without tubercles ...................................... 3 3. Embolic base with 2 teeth (Figs. 31, 32) ................................... tingpingensis Embolic base with only 1 tooth (Figs. 24, 25) ................................... sinensis 4. Embolus short, much shorter than the bulb length (Figs, 5, 6) ..................... ghecuanus Embolus long, at least the bulb length (Figs. 13, 14) ....................... rani new species Females 1. Ventral abdomen with distinct white spot in front of cribellum ........................... 2 Ventral abdomen without distinct white spot in front of cribellum ........................ 7 2. Epigynum with slits more or less parallel (Fig. 29) .................. Jaiwanensis new species Epigynum otherwise (Figs. 9, 11, 22, 27, 34) ....................................... 3 3. Epigynai median sclerite lobed on sides, spermathecal heads situated laterad of spermathecae (Figs. 9, 10) ............................................................. kunmingensis Epigynai median sclerite not lobed, spermathecal heads situated mesad of spermathecae (Figs. 11, 12, 22, 23, 27, 28, 34, 35) ..................................................... . 4 4. Slits of epigynum wider apart anteriorly than posteriorly (Figs. 11, 22) .................... 5 Slits of epigynum wider apart posteriorly than anteriorly (Figs. 27, 34) .................... 6 5. Posterior copulatory ducts much larger than spermathecae (Fig. 23) ................. senoculata Posterior copulatory ducts much smaller than spermathecae (Fig. 12) ... . jinggangemis new species 6. Anterior epigynum strongly narrowed, width approximately % of posterior (Fig. 27) ...... sinensis Anterior epigynum moderately narrowed, width at least % of posterior (Fig. 34) ..... tingpingensis 1. Spermathecal heads situated mesad of spermathecae (Fig. 16) ................ rani new species Spermathecal heads situated anterad of spermathecae (Fig. 8) ..................... ghecuanus 332 THE JOURNAL OF ARACHNOLOGY Figures 1-4. — -Fecenia cylindrata. 1. Male palp, ventral view; 2. Male palp, retrolateral view; 3. Epi- gynum; 4. Vulva. TAXONOMY Fecenia cylindrata Thorell Figs. 1-4, Map 1 Fecenia cylindrata Thorell 1895: 64 (1 juv. syntype from Tharrawaddy, Myanmar, in Naturhistoriska Riksmuseet, Stockholm, examined by Levi Map 1. — Distribution of Fecenia cylindrata, Psechrus ghecuaniis, P. kunmingensis and P. jing- gangensis new species in China. 1982). Thorell 1897; 263; Pocock 1900: 212; Lehtinen 1967: 462, figs. 472, 473 (male); Levi 1982: 136, figs. 80-82 (male and female); Yang & Wang 1993; 29, figs. 1-4 (male and female); Song, Zhu & Chen 1999: 397, fig. 2310-Q (male and female). Fecenia hainanensis Wang 1990: 257, figs. 1-3 (fe- male holotype from Tongqian City, Hainan, Chi- na, in HBI, examined). Song, Zhu & Chen 1999: 397. NEW SYNONYMY Synonymy. — This species was erroneously described as F. hainanensis with one female specimen from Hainan, China. The only dif- ference between F. hainanensis and F. cylin- drata, according to Wang (1990), was the presence of a pair of long, oval, white spots on ventral abdomen. Apparently, such spots are present in F. cylindrata and other Fecenia species (Levi 1982). Later collection of F. cy- lindrata with both males and females from the same locality (Yang & Wang 1993) further showed that F. hainanensis is in fact a junior synonym of F. cylindrata. The species F. cy- lindrata was collected from Tongqian and Qionghai, Hainan, China (Wang 1990; Yang & Wang 1993). It is widespread and occurs in large numbers in Qionghai (Yang & Wang 1993). WANG & YIN— CHINESE PSECHRIDAE 333 Figures 5-8. — Psechrus ghecuanus. 5. Male palp, ventral view; 2. Male palp, retrolaterai view; 3. Epigynum; 4. Vulva. Diagnosis.— This species can be distin- guished from others by the presence of a me- dian depression on the epigynum, and by the shape and transverse direction of the median apophysis (Figs. 1-4). Description. — See Thorell (1895), Levi (1982) and Wang (1990). Material examined. — CHINA: Hainan: Jian- feng, 6 August 1990, 1 male and 1 female (M.B. Gu, HTU); Tongqian, 1 July 1984, female holotype of F. hainanensis Wang 1990 (M.Y. Liu, HBI). Distribution. — China (Hainan) (Map 1), Myanmar. Psechrus ghecuanus Thorell Figs. 5-8; Map 1 Psechrus ghecuanus Thorell 1897: 261 (female syntypes from Myanmar, in Naturhistoriska Riks- museet, Stockholm, examined by Levi 1982); Levi 1982: 123, figs. 29-33 (female); Yin, Wang & Zhang 1985:19, fig. 1 (A-I)(male and female); Song, Zhu & Chen 1999: 397, figs. 232A-B, M- N (male and female). Diagnosis. — This species is similar to P. torvus but can be distinguished by the short embolus, the simple embolic base (Figs. 5, 6), and the more or less parallel epigynal slits (Figs. 7, 8). Male.— See description of Yin, Wang & Zhang (1985). White spot in front of cribeL lum absent. Male palpal femur without mod- ification; palpal bulb duct more or less strong- ly curved, U-shaped; conductor long, lamella shaped; embolus short, slender; embolic base simple, not rectangular, but slightly triangular (Figs. 5, 6). Female. — See descriptions of Thorell (1897), Levi (1982), and Yin, Wang & Zhang (1985). White spot in front of cribellum ab- sent. Epigynal slits more or less parallel; epi- gyeal median sclerite wide, width about 1.25X length; copulatory ducts short, not dis- tinct; spermathecal heads apparent, situated anteriorly; spermathecae rounded, large, wide- ly separated (Figs. 7, 8). Material examined. — CHINA: Yunnan: Men- gla, 21 March 1978, 1 male and 1 female (J.F. Wang, HBI); Menglun, 31 July 1981, 2 females (J.F. Wang, HBI); Menghai, 23 March 1978, 1 male and 1 female (J.F. Wang, HBI). 334 THE JOURNAL OF ARACHNOLOGY Distribution. — China (Yunnan) (Map 1), India, Thailand, Myanmar. Psechrus kunmingensis Yin, Wang & Zhang Figs. 9, 10; Map 1 Psechrus kunmingensis Yin, Wang & Zhang 1985: 25, fig. 5(A-D) (female holotype and 3 female paratypes from Kunming, Yunnan, China, in HBI, examined). Song, Zhu & Chen 1999: 397, figs. 232C-D, O-P (male and female). Psechrus tingpingensis: Feng 1990: 34, fig. 9 (fe- male only) (misidentification). Diagnosis. — This species can be easily dis- tinguished by the laterally lobed epigynal me- dian sclerite, the lateral placement of the sper- mathecal heads, the shape of spermathecae (Figs. 9, 10) and the presence of strong apoph- yses at embolic base. Female. — Described by Yin, Wang & Zhang (1985). White spot in front of cribel- lum present. Epigynal slits not parallel; epi- gynal median sclerite elongated, with lateral margins lobed; copulatory ducts long, distinct, widely separated; spermathecal heads appar- ent, situated laterally, curved anteriorly; sper- mathecae transversely extended, large, widely separated (Figs. 9, 10). Male. — Illustrated by Song, Zhu & Chen (1999), but not described. The male speci- mens are not available in this study. Judging from the illustrations by Song, Zhu & Chen (1999), male palpal bulb duct only slightly U- shaped; conductor long, lamella shaped; em- bolus short, slender; embolic base with strong apophyses (figs. 2320-P in Song, Zhu & Chen 1999). Material examined. — CHINA: Yunnan: Kunm- ing, 5 April 1979, female holotype (J.F Wang, HBI); Kunming, July 1983, 2 female paratypes (M.Y. Liu, HBI); Kunming, 21 July 1981, 4 females (J.F. Wang, HBI); Kunming, 30 June 1999, 1 female (X. Xu, HBI). Distribution. — China (Yunnan) (Map 1). Psechrus jinggangensis new species Figs. 11, 12; Map 1 Types. — Female holotype from Jinggang- shan (N26.5E1 14.1), Jiangxi, China (4 Octo- ber 1996; C.M. Yin), deposited in HBI. Etymology. — The specific name refers to the type locality. Diagnosis. — This species is similar to P. kunmingensis but can be distinguished by the laterally concaved epigynal median sclerite, the rounded spermathecae, and the mesal placement of the spermathecal heads (Eigs. 11, 12). Female. — Total length 24.5. Carapace 9.0 long, 7.8 wide. Abdomen 15.5 long, 9.0 wide. Leg measurements: I: 63.2 (18.5, 22.2, 15.1, 7.4); II: 47.3 (13.0, 16.5, 12.0, 5.8); III: 33.5 (10.0, 10.5, 8.5, 4.5); IV: 46.0 (14.0, 15.5, WANG & YIN— CHINESE PSECHRIDAE 335 11.0, 5.5). White spot in front of cribellum present. Epigynal slits not parallel; epigynal mediae sclerite elongated, with lateral mar- gins concave; copulatory ducts widely sepa- rated anteriorly, approaching each other pos- teriorly; spermathecal heads apparent, situated mesally; spermathecae rounded, widely sepa- rated (Figs. 11, 12). Male.— Unknown. Other material examined. — None. Distribution. — China (Jiangxi) (Map 1). Psechrus rani new species Figs. 13-18; Map 2 Types, — Male holotype from Sanchahe, Maolan National Nature Reserve, Libo, Guizhou, China (6 October 1997; X.P. Wang); female paratype from Xiaoqikong, Libo, Guizhou, China (2 March 1995; J.C. Ran), de- posited in IZB. Etymology.- — The specific name is a pa- tronym in honor of Mr. Jing-Cheng Ran of the research department, Maolan National Natural Reserve, Guizhou, China, the collector of the paratype female. Notes. — The male and female are matched because their localities are close together and also the similar size. Diagnosis. — This new species seems clos- est to P. torvus but can be distinguished by the simple, small embolic base, the enlarged conductor base (Figs. 13, 14), and the more or less parallel lateral margins of epigynal me- dian sclerite, and the shape of spermathecae (Figs. 15, 16). Male. — Total length 18.0. Carapace 7.2 long, 5.6 wide. Abdomen 10.8 long, 4.8 wide. Leg measurements: I: 69.6 (18.4, 23.2, 19.2, 8.8); II: 53.4 (14.4, 18.0, 14.0, 7.0); III: 33.6 (9.6, 11.2, 9.2, 3.6); IV: 54.8 (15.2, 17.0, 15.0, 7.6). White spot in front of cribellum absent. Male palpal femur without modification; pal- pal bulb duct simply curved, slightly U-sha- ped; conductor long, lamella shaped, with en- larged base; embolus long, slender; embolic base simple, small, not rectangular (Figs. 13, 14). Female.— Total length 21.6. Carapace 8.0 long, 6.0 wide. Abdomen 13.6 long, 8.0 wide. Leg measurements: I: 54.8 (14,8, 18.4, 14.4, 7.2); II: 44.2 (12.4, 15.2, 11.0, 5.6); III: 31.2 (9.2, 9.6, 8.0, 4.4); IV: 46.2 (12.8, 14.4, 12.0, 7.0), White spot in front of cribellum absent. Epigynal slits more or less parallel; epigynal median sclerite with lateral margins wide apart medially, posteriorly, approaching each other anteriorly; width of epigynal median sclerite approximately 1.5 X length; copulato- ry ducts short but clearly visible; spermathe- cal heads apparent, short, situated mesally; spermathecae rounded, widely separated (Figs. 15, 16). Penultimate instar. — As indicated by Levi (1982), some sclerotized sculpturing occurs in the genital area in the penultimate instar. In 336 THE JOURNAL OF ARACHNOLOGY Figures 13-18. — Psechrus rani new species. 13. Male palp, ventral view; 14. Male palp, retrolateral view; 15. Epigynum; 16. Vulva; 17. Penultimate instar, epigynum; 18. Penultimate instar, vulva. the penultimate instar, the epigynum and vul- va (Figs. 17, 18) are clearly apparent and may be confused with adults stage (Figs. 15, 16), if no adults are collected and compared with it. Compared to the adult stage, the longitu- dinal grooves of the epigynum of the penul- timate instar are much shorter and not well developed, and the spermathecae and sper- mathecal heads are weaker, although the cop- ulatory ducts and fertilization ducts are as well developed as the adult stage. Perhaps this is one reason why the psechrid female geni- talia appear so variable. According to our col- lection of P. senoculata from various places in China, including Shaanxi, Hubei, Sichuan, Hubei, Hunan, and Guizhou Province, all adult female genitalia are stable, particularly the vulva. Other material examined. — CHINA: Guizhou: Libo, Maolan National Nature Reserve, Yaozai, 7 October 1997, 1 female penultimate instar (X.P Wang, IZB). Distribution. — China (Guizhou) (Map 2). Psechrus senoculata Yin, Wang & Zhang Figs. 19-23; Map. 2 Psechrus mimus: Xu & Wang 1983: 35, figs. 1-7 (male and female); Song 1987: 68, figs. 34A-D (male and female); Song 1988: 33; Chen & Zhang 1991: 40, fig. 31 (male and female); Zhang WANG & YIN— CHINESE PSECHRIDAE 337 Map 2. — Distribution of Psechrus rani new spe- cies and P, senoculata in China. et al. 1998: 11, figs. 2a-p (female); Song, Zhu & Chen 1999: 397, figs. 232E-F, Q-R, PI. 3C (fe- male) (misidentification). Psechrus sinensis: Hu 1984: 55, fig. 50 (male and female); Chen & Gao 1990: 25, figs. 27a-b (male and female) (misidentification). Psechrus senoculata Yin, Wang & Zhang 1985: 21. fig. 2(A-J) (female holotype from Sangzhi, Hu- nan, male allotype from Zhangjiajian, Daiyong, Hunan, and 1 male and 1 female paratypes from Huanglongdong, Hangzhou, Zhejiang, China, in HBI, examined. Feng 1990: 33, fig. 8 (male and female). Synonymy.— The species P. senoculata has been treated as a junior synonym of either P. mimus (Song 1988) or identified as P. si- nensis (see Hu 1984; Chen & Gao 1990). Chamberlin (1924) described P. mimus from an unidentifiable female juvenile from Su- Figures 19-23. — Psechrus senoculata. 19. Male palp, ventral view; 20. Male palp, retrolateral view; 21. Male palpal femur, showing femoral modification; 22. Epigynum; 23. Vulva. 338 THE JOURNAL OF ARACHNOLOGY Figures 24-28. — Psechriis sinensis. 24. Male palp, ventral view; 25. Male palp, retrolateral view; 26. Male palpal femur, showing femoral modification; 27. Epigynum; 28. Vulva. zhou, Jiangsu, China and should be consid- ered as nomen dubium. Further study of the types of R sinensis ( two male syntypes from Guiyang, Guizhou, China, in MNHN, exam- ined) showed that P. senoculata is a valid spe= cies rather than the synonym of P. sinensis. Diagnosis* — This species can be easily dis- tinguished from P. sinensis by the elongated, vase-shaped, anteriorly wider epigynal median sclerite (Fig. 22), the large, strongly expanded posterior part of copulatory ducts (Fig. 23), and the strongly enlarged, tuberculous con- ductor base (Fig. 19). Male* — Described by Yin, Wang & Zhang (1985) and Song (1987). White spot in front of cribellum present. Palpal femur modified with notch (Fig. 21); palpal bulb duct U- shaped; conductor short, lamella shaped; con- ductor base strongly enlarged, with numerous small tubercles; embolus short, slender, with rectangular base (Figs. 19-21). Female, — Described by Yin, Wang & Zhang (1985) and Song (1987). White spot in front of cribellum present. Epigynal slits wid- er apart anteriorly than posteriorly; epigynal median sclerite vase-shaped; copulatory ducts with posterior part enlarged, extending ante- riorly; spermathecal heads apparent; sperma- thecae rounded, relatively small, close to each other (Figs. 22, 23). Material examined. — CHINA: Hunan: Sang- zhi, 21 August 1984, female holotype (Y.J. Zhang, HBI); Daiyong, Zhangjiajian, 20 September 1984, male allotype (Y.J. Zhang, HBI); Chengbu, July 1982, 2 females (X.C. Ouyang, HBI); Liuyang, Mt. Dawei, 31 July 1994, 1 female (H.M. Yan, HBI); Changsha, Lukou, 30 June 1999, 1 female (Xu, HBI); Daoxian, 9 October 1991, 1 male (L.S. Gong, WANG & YIN— CHINESE PSECHRIDAE 339 Map 3. — Distribution of Psechrus sinensis, P. taiwanensis new species, and P. tingpingensis in China and Vietnam. HBI); Shimen, Mt. Huping, 25 June-7 July 1992, 1 female (XJ. Peng, HBI); Suining, 25 May 1995, 2 females (C.M. Yin & YJ. Zhang, HBI); Heng= yang, Mt Goulou (elev. 1500 m), 30 July 1997, 1 female (XJ. Peng, HBI); Shuangpai, 1 1 August 1993, 1 female (C.M. Yin, HBI). Hubei: Wudang- shan, from Zixiao to Naeya, 23 September 1997, 1 male and 4 females (X.P. Wang, AMNH); Xian- gyang, October 1990, 1 male and 3 females (J.F. Wang, HBI). Guizhou: Kaili, 3 October 1997, 1 male and 1 female (X.P. Wang, MCB); Zunyi, 22 September 1997, 1 female (X.P. Wang, AMNH). Sichuan: Chunqing, Jingyunshan, 26 September 1997, 1 female (X.P. Wang, AMNH). Zhejiang: Hangzhou, Huaeglongdong, 16 May 1982, 1 male and 1 female paratypes (Z.F. Chen, HBI). Shaanxi: Zhouzhi, Louguantai, June 1990, 1 male and 1 fe- male (X.P. Wang, AMNH). Distribution,— China (Hunan, Zhejiang, Hubei, Guizhou, Sichuan, Shaanxi) (Map 2). Psechrus sinensis Berland & Berland 1914 Figs. 24-28, Map 3 Psechrus sinensis Berland & Berland 1914: 131, figs. 1-3 (two male syntypes from Guiyang, Guizhou, China, in MNHN, examined). Lehtinen 1967: 261, fig. 474 (male) (incorrectly synony- mized with P. singaporensis); Levi 1982: 123, figs. 34, 35 (male only, female is P. taiwanensis sp. nov.); Song, Zhu & Chen 1999: 397, figs. 232G-H, S (male and female). Psechrus guiyangensis Yin, Wang, & Zhang, 1985: 24, fig. 4(A-D) (female holotype and paratypes from Guiyang, Guizhou, China, in HBI, exam- ined). First syeonymized by Song, Zhu & Chen (1999). Synonymy. — Study of P. sinensis male types and further collections of psechrids in China shows that P. guiyangensis is a junior synonym of P. sinensis (Song, Zhu & Chen 1999). As suspected by Levi (1982), the fe- male (from Taiwan) illustrated as P. sinensis in Levi's (1982) paper is a new species P. tai- wanensis, which will be described in this pa- per. Although Lehtinen (1967) listed P. sinen- sis as a junior synonym of P. singaporensis, this was not followed by later authors (Levi 1982; Plateick 1997; Platnick 2000). The spe- cies P. sinensis can be easily distinguished from P. singaporensis by the presence of white spot in front of cribellum, the strongly narrowed anterior part of epigynal median sclerite, the spermathecal shape, and the shape of conductor and embolic base. Diagnosis. — This species is similar to P. senoculata but can be recognized by the ab- Figures 29, 30. — Psechrus taiwanensis new species, female. 29. Epigynum; 30. Vulva. 340 THE JOURNAL OF ARACHNOLOGY Figures 31-35. — Psechnis tingpingensis. 31. Male palp, ventral view; 32. Male palp, retrolateral view; 33. Male palpal femur, showing femoral modification; 34. Epigynum; 35. Vulva. sence of tubercles on the conductor base, the different shape of the rectangular embolic base (Figs. 24, 25), and the anteriorly narrowed me- dian epigynal sclerite, and the narrowly sepa- rated copulatory ducts (Figs. 27, 28). Male. — Described by Berland & Berland (1914) and Levi’s (1982). White spot in front of cribellum present. Palpal femur modified with notch (Fig. 26); palpal bulb duct simply curved, slightly U-shaped; conductor short, la- mella shaped; conductor base normal, not en- larged; embolus short, slender; embolus with toothed rectangular base (Figs. 24-26). Female. — See Yin, Wang & Zhang’s (1985) description of P. guiyangensis. White spot in front of cribellum present. Epigynal slits approach each other anteriorly; epigynal median sclerite wider posteriorly than anteri- orly, with anterior part only about Va width of posterior part; copulatory ducts narrowly sep- arated medially, with anterior and posterior part moderately separated; spermathecal heads apparent, situated mesally on spermathecae; spermathecae rounded, widely separated (Figs. 27, 28). Material examined. — CHINA; Guizhou: Guiyang (Kouy-Tcheou, Env. De Kouy-Yang), 1909 and 1913, 2 male syntypes (Le R Cavalerie, MNHN); Guiyang, 30 September 1997, 1 female (X.P. Wang, AMNH); Guiyang, 4 July 1983, female holotype and 4 female paratypes of P. guiyangensis (Y.J. Zhang, HBI); Anshun, 2 July 1999, 2 females (X. Xu, HBI). Distribution. — China (Guizhou) (Map 3). Psechrus taiwanensis new species Figs. 29, 30, Map 3 Type. — Female holotype from Taiwan (1894; Holst), deposited in BMNH, examined. Etymology. — The specific name refers to the type locality. Diagnosis. — This species is similar to P. rani new species, but can be distinguished by WANG & YIN— CHINESE PSECHRIDAE 341 Eigures 36-41. — Psechrus tingpingensis . 36. Female spinnerets, ventral view, without left PLS; 37. Tri- chobothrium; 38. Tarsal organ; 39. Female cribellum; 40. Female cribellum, enlarged; 41. Male cribellum. the depressed epigynal median sclerite, the posteriorly enlarged copulatory ducts, the small spermathecae of female (Figs. 29, 30). Female. — For body measurements, see Levi’s (1982) description of female P. sinen- sis. White spot in front of cribellum present. Epigynal slits more or less parallel; epigynal median sclerite depressed, with width slightly longer than length; copulatory ducts apparent, widely separated, enlarged posteriorly; sper- mathecal heads apparent, situated mesally; spermathecae small, widely separated (Figs. 29, 30). Male. — Unknown. Other material examined. None. Distribution. — China (Taiwan) (Map 3). Psechrus tingpingensis Yin, Wang & Zhang Figs. 31-47, Map 3 Psechrus tingpingensis Yin, Wang & Zhang 1985: 23, fig. 3 (female holotype and 2 female para- 342 THE JOURNAL OF ARACHNOLOGY Figures 42-47. — Psechnis tingpiiigefisis, spinnerets, ventral view. 42. Female, ALS, left; 43. Female, PMS, both; 44. Female, PLS, left; 45. Male, ALS, left; 46. Male, PMS, left; 47. Male, PLS, left (short arrows refer to cylindrical spigots; MAP refers to major ampullate spigots and mAP refers to minor ampullate spigots). types from Tingping, Chenbu, Hunan, China, in HBI, examined). Song, Zhu & Chen 1999: 397, hgs. 232I-J (male and female). Diagnosis. — The male of this species is similar to P. sinensis and P. senocnlata in having a rectangular embolic base and modi- fied femur (Fig. 33), but can be recognized by the slightly bifid conductor apex, and the pres- ence of two apophyses on embolic base (Figs. 31-33). The female of this species is similar to P. sinensis but can be distinguished by the much wider anterior part of epigynal median sclerite, and the anteriorly spiral copulatory ducts (Figs. 34, 35). Male. — Total length 16.0-18.0. One medi- um-sized specimen measured: Total length WANG & YIN— CHINESE PSECHRIDAE 343 18.0. Carapace 8.0 long, 4.5 wide. Abdomen 10.0 long, 4.0 wide. Leg measurements: I: 62.1 (17.0, 20.0, 17.5, 7.6); II: 46.0 (14.0, 15.0, 12.0, 5.0); III: 28.0 (9.0, 8.5, 6.5, 4.0); IV: 45.2 (13.0, 14.0, 12.0, 6.2). White spot in front of cribellum present. Palpal femur mod^ ified; palpal bulb duct U-shaped; conductor long, bifid, with dorsal apophysis sharp, high- ly sclerotized, ventral one broad, membra- nous; conductor base not enlarged, but with numerous small tubercles; embolus short, broad with sharp apex; embolic base broad, with two strongly sclerotized apophyses (Figs. 31, 32). Female. — -Described by Yin, Wang & Zhang (1985). White spot in front of cribeL lum present. Epigynal slits approaching each other anteriorly; epigynal mediae sclerite wid- er posteriorly than anteriorly, with anterior part about Vi width of posterior part; copula- tory ducts spiral anteriorly, widely separated posteriorly; spermathecal heads apparent, sit- uated mesally; spermathecae rounded, widely separated (Figs. 34, 35). Material examiined. — CHINA: Hunan: Cheng- bu, Tingping, 31 July 1982, 1 female holotype and 2 female paratypes (J.F. Wang & YJ. Zhang, HBI); Shimen, 25 June to 5 July 1992, 1 male and 2 fe- males (XJ. Peng & L.P. Xie, HBI). Guangxi: Long- sheng, 7 August 1982, 12 females (J.F. Wang & YJ. Zhang, HBI); Ningming, May 1992, 1 male and 1 female (X. Pan, HBI); Ningming, 27 May 1997, 2 males and 2 females (YJ. Zhang, HBI). VIET- NAM: Hanoi: Tam Dao Mt. Forest Park, 2 May 1999, 2 males and 4 females (X.P. Wang, AMNH). Distribution. — China (Hunan, Guangxi), Vietnam (Hanoi) (Map 3). ACKNOWLEDGMENTS We thank J. Margerison of the BMNH, C. Rollard of the MNHN, J.C. Ran of the Re- search Department, Maolan National Nature Reserve, Guizhou, China, and M.S. Zhu of the HTU for the loan of specimens, J. Chen of the IZB for depository of some types. H.W. Levi of the Museum of Comparative Zoology at Harvard University and N.L Platnick (AMNH) kindly helped review the manu- script. R. Baptista of the National Museum of Natural History, Smithsonian Institution, Washington,- D.C. gave invaluable help. This research is in part based upon work supported by the National Science Foundation under Grant No. 9870232 to the Center for Biodi- versity and Conservation, AMNH, and the In- stitute of Ecology and Biological Resources, Hanoi, Vietnam. X.R Wang was supported by the Schlinger Foundation Postdoctoral Fel- lowship in Systematic Entomology of the Cal- ifornia Academy of Sciences. LITERATURE CITED Berland, J. & L. Berland. 1914. Description d’un Psechrus nouveau de Chine. Bulletin de la So- ciete entomologique de France 83:131-133. Chamberlin, R.V. 1924. Descriptions of new American and Chinese spiders. Proceedings of the United States National Museum 63(13): 1-38. Chen, S.H. 1996. A checklist of spiders in Taiwan. Annual of Taiwan Museum 39:123-155. Chen, X.E. & J.C. Gao. 1990. The Sichuan farm- land spiders in China. Sichuan Science and Tech- nology Publishing House, Chengdu, 226 pp. Chen, Z.F. & Z.H. Zhang. 1991. Fauna of Zhe- jiang: Araeeida. Zhejiang Science and Technol- ogy Publishing House, 356 pp. Feng, Z.Q. 1990. Spiders of China in colour. Hu- nan Science and Technology Publishing House, 256 pp. Hu, J.L. 1984. The Chinese spiders from the fields and the forests. Tianjin Press of Science and Techniques. 482 pp. Lee, C.L. 1966. Spiders of Formosa (Taiwan). Tai- chung Junior Teachers College Publisher, 84 pp. Lehtinen, P.T. 1967. Classification of the cribellate spiders and some allied families, with notes on the evolution of the suborder Araneomorpha. Annales Zoologici Fennici 4:199-468. Levi, H.W. 1982. The spider genera Psechrus and Fecenia (Araneae, Psechridae). Pacific Insects, 24:114-138. Murphy, J. 1986. Additional inform.ation concern- ing the spider family Psechridae. Bulletin of the British Arachnological Society 7:183-185. Platnick, N.L 1997. Advances In Spider Taxonomy 1992-1995, New York Entomological Society, 976 pp. Platnick, N.L 2000. The World Spider Catalog. American Museum of Natural History Web Doc- ument. Pocock, R.I. 1900. Arachnida. Fauna of British In- dia, Including Ceylon and Burma. London, 279 pp. Song, D.X. 1987. Spiders From Agricultural Re- gions Of China (Arachnida: Araneae). Agricul- ture Publishing House, Beijing, 376 pp. Song, D.X. 1988. A revision of the Chinese spiders described by Chamberlin. Sinozoologica 6:123- 136. Song, D.X., M.S. Zhu & J. Chen. 1999. The Spi- ders Of China. Hebei Science and Technology Publishing House, Shijiazhuang, 640 pp. 344 THE JOURNAL OF ARACHNOLOGY Thorell, T. 1881. Studi sur Ragni Males! e Papuan! III. Nel Museo Civico di Storia Naturale di Gen- ova 17:1-720. Thorell, T. 1895. Descriptive Catalogue Of The Spiders Of Burma. British Museum, London, 406 pp. Thorell, T. 1897. Viaggio di Leonard! Fea in Bir- mania e region! vicine. Nel Museo Civico di Sto- ria Naturale di Genova 37:161-267. Wang, J.F 1990. A new species of psechrid spider from China (Aeaneae: Psechridae). Natural Sci- ence Journal of Hunan Normal University 13: 257-258. Xu, Y.J. & L. Wang. 1983. A record of Psechriis mimus Chamberlin. Journal of Huizhou Teachers College 2:35-36. Yang, Z.L. & X.Z. Wang. 1993. A newly recorded of the spider genus Fecenia (Araneae: Psechri- dae) from China. Acta Arachnologica Sinica 2: 27-28. Yin, C.M., J.F. Wang & Y.J. Zhang. 1985. Study on the spider genera Psechrus from China. Jour- nal of Hunan Teachers College (Natural Science Edition) 1:19-27. Zhu, C.D, 1983. A list of Chinese spiders (revised in 1983). Journal of Bethune Medical University 9 (supplement): 1-130, Zhang, Z.H., P.P. Yuan, FP. Tong & Z.F Chen. 1998. Study on fine structure of the two spider (Metathelae: Cribellatae). Journal of Hangzhou University 25(l):74-79. Manuscript received 15 January 2000, revised 11 February 200 L 200 L The Journal of Arachnology 29:345-353 A COMPARATIVE STUDY OF THE BIOLOGY AND KARYOTYPES OF TWO CENTRAL EUROPEAN ZODARIID SPIDERS (ARANEAE, ZODARIIDAE) Stano Pekar: Department of Zoology and Ecology, Faculty of Science, Masaryk University, Kotlafska 2, 611 37 Brno, Czech Republic Jifi Krah Laboratory of Arachnid Cytogenetics, Department of Genetics and Microbiology, Faculty of Sciences, Charles University, Vinicna 5, Prague 2, 128 44, Czech Republic ABSTRACT. A comparison of the biology and karyotypes of Zodarion germanicum and Zodarion rubidum (Araneae, Zodariidae) which occur in central Europe was carried out. Surprisingly, these species were found to differ in a number of characters such as pattern of activity, reproduction and karyotypes. Zodarion germanicum was observed to be diurnal, whereas Z. rubidum is nocturnal. Courtship and mating were markedly longer and more complex in Z. germanicum than in Z rubidum. Females of Z. germanicum produced only one or two successive egg sacs including 17 eggs on average which they would guard, while females of Z. rubidum produced up to 5 egg sacs each having 4 eggs that they abandoned. The two species differ from each other also in number of chromosomes and the sex chromosome system. Results suggest these species belong to distant evolutionary lineages within the genus Zodarion. Keywords: Araneae, Zodariidae, activity, reproduction, karyotype About 500 species of Zodariidae have been described so far, most of which occur in the subtropical region (Jocque 1991). In Europe, only one genus, Zodarion, including 47 spe- cies (Bosmans 1993, 1997), occurs. These species are well-known for being myrmecoph- agous and for constructing remarkable retreats from soil grains (e.g., Nielsen 1932). This re- markable behavior was observed for the first time more than 100 years ago (Simon 1864; Santschi 1908). Since then there have been seven studies concerning the biology of these spiders (Wiehle 1928, 1953; Schneider 1971; Harkness 1976, 1977a, b, 1995; Jocque & Bil- len 1987; Couvreur 1990; Harkness & Hark- ness 1992). Nevertheless, a great majority of these investigations was centered upon the ant-eating behavior only. Thus, many other as- pects of biology of these fascinating spiders, such as pattern of activity or reproduction, are poorly understood or even unknown. In our study we focused on two species, Zodarion germanicum (C.L.Koch 1837) and Zodarion rubidum Simon 1914, the only rep- resentatives of the genus Zodarion in the Czech Republic and Slovakia. The former and larger species (body length 3. 5-6.0 mm) oc- curs abundantly in dry habitats associated with coniferous woodland only in central Eu- rope. The latter and slightly smaller species (3-4 mm), was known only from southwest- ern France (Denis 1935). But in the last de- cades it has spread into central Europe, for example, onto sand dunes in South Moravia, Czech Republic (Pekar 1994). Both species show invasive tendencies as they often occur in secondary habitats. Z. germanicum, for in- stance, was recorded from heather rimming forested peat-bog (Mahringova 1993), and Z rubidum on sandy substrates within the area of Berlin railway station (Germany) (Broen & Moritz 1987) or on mining dumps in Slovakia (Pekar 1994; Krajca 1996). At present, the ranges of the two zodariid species overlap in central Europe, but only in a few examples was sympatric occurrence proven at the eco- logical level (e.g., Jelmek 1999). Based on morphological characters of cop- ulatory organs, Bosmans (1997) classified 47 European species of the genus Zodarion into six groups. Zodarion germanicum was placed in “germanicum” while Z. rubidum was placed in “rubidum.” At the beginning of our observations, the two species appeared to be 345 346 THE JOURNAL OF ARACHNOLOGY very similar, both occurring in identical hab- itats and foraging on similar ant species (Pe- kar Linpubl. data). However, later investigation showed that these species mimic different ant species (Pekar & Krai unpubl.) and have dif- ferent activity. Thus our aim was to focus in detail on aspects of their biology which have been insufficiently studied in order to clarify differences between the study species which could have significance for further study of evolution within the genus Zodarion. STUDY AREA The study sites are situated in Slovakia which is in the center of the distribution of Z. germanicum and at the northeastern edge of the distribution of Z. rubidum. Zodarion rubidum was observed on a min- ing dump in Novaky town. The dump (about 25 years old) consists of Tertiary tuff and coal slate and is sparsely covered with vegetation, dominated by the grass Calamagrostis epi- geios (L.) Roth. Zodarion germanicum was observed on a steep outcropping in a nearby village Opatovce nad Nitrou, about 6 km from Novaky. This study area is a former sand pit adjacent to a pine forest {Pinus silvestris L.). It was abandoned some 15 years ago. The Neogene conglomerate sands of this site are mostly barren, with many stones and the cover is sparse vegetation dominated by the grass Dactyl is glome rata L. The elevation of both sites is 275-290 m. The average annual tem- perature of the area is 8.5 °C, and the average annual precipitation is 650 mm. Average bi- weekly temperatures for the sites are dis- played in Fig. 1. Soil surface temperature of the study areas was measured under a clear sky (on 7 and 8 June 1997) by means of a THERM 2246-2 thermometer at 0600, 1000, 1400 and 1800 h. Obtained data showed that the temperatures of study sites were very sim- ilar. METHODS The investigation took place both in the field and under laboratory conditions. From April to October 1997, weekly visits were made to the study areas to assess the propor- tion of adult spiders which were either run- ning on the ground or hidden in retreats. On one day in June 1997, the number of both spi- der species (seen during 5 min) and the num- ber of ants (seen during 30 sec) in a 1 circle drawn in the soil around three nest en- trances was assessed every hour (between 0600 and 2200 h). In June the sun rises at about 0445 h and sets at about 2045 h in the study area. The frequency of ant species hunt- ed by spiders was also recorded in June. A few egg sacs (3 in Z. rubidum and 2 in Z. germanicum) were collected, and the sizes of eggs were measured using a stereoscopic mi- croscope. Forty adult individuals (206 209) of each species were brought into the laboratory in June to investigate their behavior. The indi- viduals were kept in specimen containers (di- ameter 15 mm, 60 mm long) at 20 °C ± 2° and under natural ED regime (14:10). They were offered various substrates, such as soil, sand grains, paper, pine needles, leaves, and other plant material, all potentially useful for the construction of retreats. The relative fre- quency and size of retreats constructed was measured after three days. The substrate was moistened as it dried out, usually at three-day intervals. The reared specimens were fed in excess with ant workers of Tetramorium caes- pitum (L.). Then all 40 specimens were moved in pairs (male and female) to a Petri dish (diameter 60 mm) with a filter paper attached to the bottom, and kept separated by a paper barrier in order to study their reproductive behavior. As soon as the male began to ''search,” the barrier was removed. Style and duration of courtship and mating were observed under binocular stereo- microscope. After copulation males were re- moved. If the female was not receptive the males were immediately removed. After females laid eggs, the number of egg sacs produced and the incubation periods were recorded. Individual fecundity was assessed by summing the number of hatched offspring with the number of undeveloped eggs which were left in each egg sac. Data on the duration of mating, the number and size of eggs, and the incubation period for the two study species were compared using the permutation (exact) test since they did not meet the criteria required for parametric tests. A two-sided 2-sample randomization test after Manly (1997) was used. The simulation test procedure was constructed within RESAM- PLING STATS program (Simon 1993). Two different methods were used for chro- mosome preparations. The first method was PEKAR & KRAL— BIOLOGY OF TWO ZODARIID SPIDERS 347 Figure 1. — Relative frequency of mature specimens of Zodarion germanicum and Zodarion rubidum compared with the average temperatures (A) in the study region at biweekly intervals. used for preparation of chromosomes from subadiilt and adult individuals. The entire con- tents of the abdomen were dissected out in a hypotonic solution (0.075 M KCl). After 20- 25 min of hypotonic treatment, the tissues were placed into a small beaker with fresh fix- ative (a mixture of absolute methanol and gla- cial acetic acid, 3:1). The pieces of the tissues were incubated in a beaker in a refrigerator at 5 °C. During the first hour of incubation, the fixative was renewed twice (after 15 and 45 min of incubation). After 5-6 h the tissues were placed into a tube with new fixative, re- suspended, and centrifuged at 2000 G for 5 min. The supernatant was discarded and the sediment was diluted in fresh fixative to an optimal concentration of fixed cells. The sus- pension was thee dropped onto clean slides. The chromosomes from first instar speci- mens were obtained by a modification of the spreading technique used by Traut (1976) as follows. The entire contents of the abdomen were dissected out and treated in hypotonic solution as in the former case. Following a 15-30 min fixation in freshly prepared Carnoy fixative (ethanolichloroformiglacial acetic acid 6:3:1) the tissue was placed in a drop of 60% acetic acid on a clean slide. The tissue was quickly shredded as finely as possible with a pair of fine tungsten needles. The slide was then placed quickly on a warm histolog- ical plate (surface temperature of 40 °C) and the drop of dispersed tissue was allowed to evaporate while keeping it moving constantly using a fine tungsten needle. The slides obtained by both methods were air-dried at room temperature overnight, and stained with 5% Giemsa solution in Sorensen phosphate buffer (pH 6.8) for 5-6 min (Cokendolpher & Brown 1985). RESULTS Phenology, — The phenology diagrams for the study species are shown in Fig. 1. Season- al activity of spiders began in April when both juvenile and subadult specimens appeared on the ground and started hunting. Of the Z. ger- manicum specimens collected on 25 April 1997, 40% {n = 15) were adult, increasing to 80% (n — 25) within two weeks (11 May 1997). During that time, all individuals of Z rubidum (n = 17) were still subadult. On 25 May 1997 92% (n = 24) of specimens of Z. germanicum and 61% {n = 31) of Z. rubidum were adult. In 1997, mating began in April (Z. germanicum) or at the end of May (Z rubi- dum). The egg sacs were found on 17 June 1997, Examining cocoons the first free instar was found on 2 July for Z. germanicum and on 19 July for Z. rubidum. The last adults were recorded on 2 October for Z. germani- cum and on 30 October for Z. rubidum. Both species overwinter as juveniles hiding in re- treats. 348 THE JOURNAL OF ARACHNOLOGY Figures 2, 3.- — Igloo-shaped retreats. 2. Zodarion genmmicum on the lower surface of a stone, con- structed of soil and pine needles; 3. Zodarion rubidnm, constructed of sand grains and attached to the lower surface of a stone. Shelters. — The spiders rest and molt in re- treats (Fig. 2). The retreats are closed solitary “igloo-shaped” shelters, usually attached to a solid substrate, such as the lower surface of stones, usually near an ant nest entrance. Of- ten an aggregation of retreats attached to a stone was found (Fig. 3). In the field, the re- treats were constructed with a wide variety of materials (e.g., soil or sand grains, plant ma- terial, pine needles) held together by webbing. In the laboratory, the retreats were also con- structed of artificial material, e.g., paper. A new retreat was constructed when the old one was destroyed, or after molting. The material was collected by a spider from the vicinity of a retreat site. A spider brought a particle in its palps to a certain place, held it by a leg IV, attached it to the substrate with silk, and then continued this process until the retreat was complete. The construction of retreats lasted 0.5-2 h, depending on the availability of a suitable material. Nevertheless, in the labora- tory 45% (n = 40) of specimens of Z. ger- manicurn and 38% (n = 40) of Z. rubidum did not construct a retreat within 3 days. Such be- havior was observed also in the field, where many individuals did not construct a retreat and were found to rest in soil holes or rock crevices and in other similar shelters. The di- ameter of retreats constructed by Z. germani- cum was on average 5.6 ± 0.2 (S.D.) mm {n = 8) for adult males, 9.4 ± 0.2 mm {n = 10) for adult females and 4.1 ± 0.1 mm (r? = 12) for the first instar. In Z. rubidum it was on average 4.8 ± 0.2 mm {n = 6) for adult males, 5.3 ± 0.1 mm {n = 9) for adult females and 3.0 ± 0.1 mm {n — 10) for the first instar spiders. When no suitable material was of- fered, females usually spun a small sheet web (about 1 cm^) to hide under, whereas males did not construct such a web. Activity. — Zodarion germanicum was found to be active during the day. In June its activity began at about 0700 h and terminated approximately at 2100 h. The spiders were seen hunting and mating near the entrances of ant nests during all activity period, but before sunset they moved into an old retreat or con- structed a new one. There was a slight decline in activity between 1000 h and 1400 h when the temperature of soil surface reached 40 °C and the ants were most active (Fig. 4). During rainy or cool days (i.e., average day temper- ature about 15 °C), the number of active spi- ders was approximately halved. Individuals of Z. rubidum were active in the morning (0600- 0900 h) and in the evening (1830-2200 h) (Fig. 5). There was no spider active between 1000-1700 h. The nocturnal activity of this species was not investigated. The spiders were hunting and mating during both periods of ac- tivity; construction of retreats was recorded only in the evening. During these periods the surface temperature fell below 30 °C. Courtship and mating. — When a male of Z. germanicum approached a female, it began to move very slowly with the whole body vi- brating, with waving raised forelegs and PEKAr & KRAL— BIOLOGY OF TWO ZODARIID SPIDERS 349 E E 10 6 4 2 0 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 time (hours) B c ffl 6 7 8 9 10 11 12 13 14 16 16 17 18 19 20 21 22 time (hours) Figures 4, 5. — Activity (mean number, n = 3) of spiders compared to the activity of ant species which the spiders most frequently hunted. 4. Zo- darion germanicum and Formica cinerea in Opa- tovce nad Nitrou; 5. Zodarion rubidum and Tetra- morium caespitum in Novaky. dramming palps. When reaching a female, the male first lightly touched her with his fore- legs, then followed by a “sparring” with palps. If the female was receptive she became passive, stayed in a normal position; and the male, still vibrating, moved across her and in- serted palpal organs first from one side, then from the other side. This mating position is classified by Foelix (1996) as hype 3.’ The courtship of Z. rubidum was much shorter. The male quickly approached the female with twitching raised forelegs and drumming palps. After a short period of palpation, they copu- lated in the same way as in Z. germanicum. The copulation lasted on average 11.9 ± 0.3 (S.D.) min {n = 18) in Z. germanicum, and 1.3 ± 0.1 min {n = 15) in Z. rubidum. The difference in duration is highly significant {P < 0,001; 2-sample randomization test, 4999 simulations). Soon after this primary, “long” copulation, the female of both species could copulate again with another male but the sub- sequent mating (which involved attempts to insert a palp from each side) lasted less than 15 sec. Before and after such short mating the female was vibratieg/quivering. Fertility and brood care.—The eggs were laid in woolly silken sacs (Fig. 6). There were on average only 4.2 eggs in a cocoon of Z. rubidum but 16.5 eggs in a cocoon of Z. ger- manicum (Table 1). The difference in number of eggs between the species is highly signifi- cant {P < 0.001; 2-sample randomization test, 4999 simulations). The eggs of both species were cream-colored and did not stick to each other but rolled freely. The mean diameter of eggs was 0.79 in Z. rubidum and 0.9 mm in Z. germanicum. The eggs of Z. germanicum were significantly larger {P < 0.001; 2-sample randomization test, 4999 simulations) than those of Z. rubidum. Females of Z. rubidum produced an egg sac on average within 9 days after copulation whilst females of Z germanicum produced an egg sac on average within 14 days (Table 1). Females of Z. rubidum produced up to 5 egg sacs within a month while females of Z. ger- manicum produced one or (in one case) two successive egg sacs, after hatching offspring from the first one. The female of Z rubidum hid each egg sac in a separate retreat and kept on hunting ants without paying attention to Table 1 . — Comparison of differences (mean ± standard deviation) found in the reproduction of the two study species. Z. rubidum Z. germanicum Number of eggs/cocoon 4.2 ±0.1 {n = 13) 16.5 ± 0.3 {n = 15) Diameter of eggs (mm) 0.79 ± 0.01 {n = 12) 0.9 ± 0.01 {n = 24) Time to egg sac production (days) 5-13 {n - 13) 2-26 (« = 15) Incubation period (days) 57.5 ± 1.1 in = 9) 39.6 ± 1.9 {n = 5) 350 THE JOURNAL OF ARACHNOLOGY Figure 6. — Retreat of Zodarion rubidum includ- ing remnants of cocoon. Arrow points to the open- ing which spiderlings used to escape. The cocoon is the smooth-surfaced structure at the center. the cocoons except for placing an occasional (31%, n = 13) dead ant in the retreat contain- ing an egg sac. In contrast, the female of Z. gennanicum built a large retreat in which the egg sac was placed and stayed on guard in- side. She left the retreat approximately once during a four-day period to feed. The hrst instar emerged from the egg sac on average 57.5 days after laying eggs in Z. rub- idum, and after 39.6 days in Z. germaniciim (Table 1). Nevertheless, the difference in the incubation period is not significant {P = 0.17; 2-sample randomization test, 4999 simula- tions). In the laboratory, the female of Z. ger- mcmicum died at last, and was fed upon by some of the first instar spiders. Spiderlings stayed in the remains of the cocoon until the first molt, which occurred within a few days of emergence. They then dispersed from the co- coon through a tiny opening on the side with- out assistance of the female (Fig. 6), and each specimen constructed its own tiny retreat. Karyotype. — Both mitotic and meiotic phases were obtained from subadult and adult males. The first instar and females gave only mitotic phases. Acrocentric chromosomes pre- dominated in karyotypes of the both species but differences in the size of particular chro- mosomal pairs were apparent. The diploid chromosome number (2n) in Z. germaniciim was 29 for males (Figs. 8, 10), and 30 for females and in Z. rubidum 24 for males (Figs. 7, 9), and 26 for females. The comparison of meiotic male phases with mitotic metaphases of both sexes indicated an XO type of sex chromosome system in Z. germanicum, with male XO and female XX. The large X chro- mosome was acrocentric. Z. rubidum possess- es sex chromosome system X]X20 with male X,X2 and female X,XiX2X2. Both X chromo- somes were acrocentric but of different size. Similar to the majority of spiders analyzed so far, the X chromosome! s) in males of both species show greater condensation than auto- somes during prophase I (positive heteropyc- nosis) and lie on the periphery of meiotic fig- ures until metaphase II. In Z. rubidum, both X chromosomes are aligned closely to each other, not only during the first meiotic divi- sion, like the majority of species of spiders analysed so far, but even until metaphase II. DISCUSSION Because both species belong to the same genus, we expected that they might have sim- ilar life histories. Surprisingly, besides some similarities we also found several striking dif- ferences. Both species construct, like other species of the genus Zodarion studied so far, figloo-shaped’ retreats. Observed building procedure was in agreement with the one giv- en by Harkness (1977b) and Couvreur (1990). From an evolutionary point of view, we sup- pose the retreats are an apomorphy derived from simple burrowing behavior known from many other zodariid groups (Jocque 1991) used for protection in an unusual environment of ground-living ants. Surprisingly, the con- struction of the retreats does not seem to be an obligatory habit of all specimens, as sug- gested by both laboratory and field observa- tions. However, we suppose that even hiding in soil or in other similar places could be con- sidered as a remnant of burrowing behavior. Couvreur (1990) described two types of re- treats, open and closed, constructed by Z. rub- idum. According to him, the spiders hide in the open retreats when hunting but use closed retreats when resting. We have never observed open retreats either in a field or in a lab. In contrast to the laboratory findings of Harkness (1977b) on Z. frenatum Simon, we did not ob- serve marked differences between the number of males and females which constructed re- treats. PEKAr & KRAL— BIOLOGY OF TWO ZODARIID SPIDERS 351 Figures 7-10. — 7. Zodarion rubidum, mitotic metaphase of male; 8. Zodarion germanicum, two male daughter cells at metaphase 11. Note differences in the size of particular chromosomes, especially in Zodarion rubidum; 9. Zodarion rubidum, diplotene of male; 10. Zodarion germanicum, diplotene of male. Arrow points to X chromosome(s) displaying still weak positive heteropycnosis. Scale line = 0.01 mm. The onset of activity and maturity of Z. rub- idum was delayed in comparison with Z ger- manicum by about two weeks within which period the average temperature increased about 5 °C. This delay might be a conse- quence of the fact that the former species has spread to central Europe from southern Eu^ rope where the average day temperatures are considerably higher. The activity patterns of both species were clearly different. We ob- served Z rubidum to be active in the evening and in the morning. We did not investigate its activity during the night since Couvreur (1990) studied the nocturnal activity of this species in detail. Nocturnal activity was found also in other species of the genus Zodarion: Z. frenatum Simon 1884 (Harkness 1977a) and Zodarion sp. from Afghanistan (Schnei- der 1971). On the contrary, diurnal activity as observed in Z. germanicum has never been re- ported for any species of the genus Zodarion. Until now, information gathered about mat- ing and related behavior of the genus Zoda- rion has been very scarce and incomplete. Our observation revealed that although both spe- cies copulated in the same position, charac- teristic for “modern” wandering spiders (Foe- lix 1996), significant differences in courtship 352 THE JOURNAL OF ARACHNOLOGY and mating were observed. In Z. germanicum, courtship and mating took more time and were more complex than in Z. rubidum. Except for the “long” copulation that we consider a true one, we also recorded multiple short copulations in both species. Such cop- ulations were also noticed by Gerhardt (1928) who observed in Z elegans (Simon 1873) multiple copulations, each lasting only a few sec. We consider this a pseudo-copulation. Af- ter the true copulation, the females became Linreceptive and expressed this by a specific behavior (quivering) which threatened the male. A similar behavior was observed for the female of a thomisid spider, Xysticus cristatus (Clerck 1757) (Bristowe 1941). Females of both species hid their egg sacs in retreats as observed by Harkness (1995) in Z. frenatum but showed different brood-care strategies. While females of Z. germanicum produced only one or two egg sacs of approx- imately 16 eggs each and guarded the egg sacs, females of Z. rubidum produced on av- erage 5 egg sacs of approximately 4 eggs, and exhibited no further care. Wiehle (1953) found 25-50 eggs per egg sac in Z. germanicum. However, we found a maximum of 25 eggs per egg sac of this species. Regarding other Zodarion species, Harkness (1995) reported 9-12 eggs/egg sac of Z. frenatum. Unfortu- nately, he did not mention either how many egg sacs the female produced or whether the female guarded the egg sacs or not. With respect to the karyology, the family Zodariidae appears to be practically an un- known group. Only one short note on the number of chromosomes in '"Storena"" indica Tikader & Patel 1975 (2n, male = 22; X,X20) has been published (Datta & Chatterjee 1983). Males of the species examined possess a hap- loid number of chromosomes which is close to the mean male haploid number of chro- mosomes {n = 14.09) in spiders (Gowan 1985). Though both species studied are placed in the same genus, they differ considerably in the number of chromosomes as well as in the sex chromosome system. From an evolution- ary point of view, Z. rubidum exhibited a sex chromosome system that seems to be an an- cestral trait in spiders (Suzuki 1954; White 1973). This sex chromosome system was also found in the most primitive recent spider tax- on, i.e., in Mesothelae (Suzuki 1954). Thus we hypothesize that the system XO in Z. ger- manicum is derived from the X,X20. The large acrocentric X chromosome of this species might have originated by tandem fusion be- tween the ancestral acrocentric chromosomes Xj and X2 that are still conserved in the kar- yotype of “S'.” indica and Z. rubidum. Our comparative study revealed that al- though both species are placed in the same genus, they differ in a number of characters such as activity pattern, courtship, mating, brood care, karyotype and sex chromosome system. Our results support Bosman’s (1997) separation of these species into two different groups and suggest they might belong to dis- tant evolutionary branches of this genus. However, to understand further the evolution process within the genus Zodarion, we sug- gest additional research of these aspects of bi- ology in other representatives of this genus. We assume such investigation might also con- tribute a clarification of the evolution of mim- icry in the subfamily Zodariinae. ACKNOWLEDGMENTS We would like to thank J.M. Couvreur (WWF, Belgium) for providing us with his own papers as well as some unavailable papers from other authors. We are greatly indebted to Dr. R. Jocque (Koninklijk Museum voor Midden- Afrika, Belgium) who kindly commented on our manuscript and made some valuable sug- gestions. Our special thanks are extended to Dr. J. Hajer (University of J.E. Purkyne, Czech Re- public) for a help with rearing some specimens. The meteorological data were kindly provided by Meteorological station in Prievidza (Slovak- ia). SP was funded by the grant of the Grant Agency of the Czech Republic (no. 206/01/ P067) and the grant of the Masaryk University (no. 143100010). JK was funded by the grant of the Charles University (no. 111/1998/B BIO/PfF). LITERATURE CITED Bosnians, R. 1993. Revision of the genus Zodarion Walckenaer, 1833 in the Iberian peninsula and Balearic islands (Araneae, Zodariidae). Eos 69: 115-142. Bosnians, R. 1997. Revision of the genus Zodarion Walckenaer, 1833, part IT Western and central Europe, including Italy (Araneae: Zodariidae). Bulletin of the British Arachnological Society 10:265-294. Bristowe, WS. 1941. The Comity of Spiders. II. Ray Society, London. PEKAR & KRAL— BIOLOGY OF TWO ZODARIID SPIDERS 353 Broen von, B. & M. Moritz. 1987. Zum Vorkom- men von Zodarion rubidum (Simon, 1914) im Berliner Gebiet. Deutsche Entomologische Zeit- schrift 34:155-159. Cokendolpher, J.C. & J. Brown. 1985. Air-dry method for studying chromosomes of insects and arachnids. Entomological Newsletter 96:114- 118. Couvreur, J.M. 1990. Quelques aspects de la bioL ogie de Zodarion rubidum, Simon, 1914. Nieuwsbrief van de Belgische Arachnologische Vereniging 7:7-15. Datta, S.N. & K. Chatterjee. 1983. Chromosome number and sex determining mechanism in fifty- two species of spiders from north-east India. Chromosome Information Service 35:6-8. Denis, J. 1935. A propos de quelques Araignees du genre Zodarion Walck., appartenant a la faune fraegaise. Bulletin de la Societe d’Histoire Na- turelle de Toulouse 67:51-68. Foelix, R.F. 1996. Biology of Spiders. 2nd ed. Ox- ford University Press, New York. Gerhardt, U. 1928. Biologische Studien an grie- chischen, corsischen und deutschen Spinnen. Zeitschrift fur Morphologic und Okologie der Ti- ere 10:576-675. Gowan, T.D. 1985. The life history and reproduc- tion of the wolf spider Lycosa lentia Hentz. Dis- sertation, University of Florida, Gainesville. Harkness, R.D. 1976. The relation between an ant, Cataglyphis bicolor (F.) (Hymenoptera: Formi- cidae) and a spider, Zodarion frenatum (Simon) (Araneae: Zodariidae). Entomologist’s Monthly Magazine 111:141-146. Harkness, R.D. 1977a. Further observations on the relation between an ant, Cataglyphis bicolor (F.) (Hymenoptera: Formicidae) and a spider, Zoda- rion frenatum (Simon) (Araneae: Zodariidae). Entomologist’s Monthly Magazine 112:111- 123. Harkness, R.D. 1977b. The building and use of “shelters” by a hunting spider {Zodarion frena- tum Simon) in Greece. Zoologischer Anzeiger 199:161-163. Harkness, R.D. 1995. The use of “shelters” by an ant hunting spider, Zodarium frenatum Simon, as receptacles for eggs. Entomologist’s Monthly Magazine 131:46. Harkness, M.L.R. & R.D. Harkness. 1992. Preda- tion of an ant {Cataglyphis bicolor (F.) Hym., Formicidae) by a spider {Zodarium frenatum (Si- mon) Araneae, Zodariidae) in Greece. Entomol- ogist’s Monthly Magazine 128:147-156. Jelmek, A. 1999. Vyskyt teplomiinych druhu pa- vouku (Araneae) v udolf feky Jihlavy na Ces- komoravske vrchovine [Occurrence of thermoph- ilous spiders (Araneae) in the valley of the river Jihlava in Ceskomoravska vrchoviea]. M.Sc. the- sis, Charles University, Prague. Jocque, R. 1991. A generic revision of the spider family Zodariidae (Araneae). Bulletin of the American Museum of Natural History 201:1- 160. Jocque, R. & J. Billen. 1987. The femoral organ of the Zodariinae (Araneae, Zodariidae). Revue de Zoologie Africaine 101:165-170. Krajca, A. 1996. Spolocenstva pavukov (Araneae) na niklovych haldach v Seredi [Spider commu- nities of nickel dumps in Sered, Slovak Repub- lic]. M.Sc. thesis, Commenius University, Bra- tislava. Manly, B.F.J. 1997. Randomization, Bootstrap and Monte Carlo Methods in Biology. Chapmann & Hall, London. Mahringova, M. 1993. Stfevlikoviti brouci (CoL: Carabidae) a pavouci (Araneida) ruznych typu raselinisf CHKO Tfebonsko [Carabid beetles and spiders of various peat-bogs in protected area of Tfebonsko, Czech Republic]. M.Sc. the- sis, Charles University, Prague. Nielsen, E. 1932. The Biology of Spiders, With Special Reference to the Danish Fauna. Levin & Munskgaard, Copenhagen. Pekar, S. 1994. The first record of Zodarion rubi- dum and Ostearius melanopygius for Slovakia (Araneae: Zodariidae, Linyphiidae). Entomolog- ical Problems 25:97-100. Santschi, F. 1908. Quelques observations nouvelles et remarqees sur la variabilite de F instinct de ni- dification chez les fourmis. Journal fiir Psychol- ogie und Neurologic 13:136-149. Schneider, P. 1971. Ameisenjagende Spinnen (Zo- dariidae) an CataglypMs-H&si&m in Afghanistan. Zoologischer Anzeiger 187:199-201. Simon, E. 1864. Histoire naturelle des araignees (Araneides). Paris. Simon, J.L. 1993. Resampling: The New Statistics. 2ed ed. Resampling Stats, Wadsworth. Suzuki, S. 1954. Cytological studies in spiders III. Studies on the chromosomes of fifty-seven spe- cies of spiders belonging to seventeen families with general considerations on chromosomal evolution. Journal of Science of the Hiroshima University (B) 15:23-136. Traut, W. 1976. Pachytene mapping in the female silkworm Bombyx mori L. (Lepidoptera). Chro- mosoma 58:275-284. White, M.J.D. 1973. Animal Cytology and Evo- lution. Cambridge University Press, Cambridge. Wiehle, H. 1928. Beitrage zur Biologic der Ara- neen insbesondere zur Kenntnis des Radnet- baues. Zeitschrft fiir Morphologic und Okologie der Tiere 11:115-151. Wiehle, H. 1953. Zodariidae. Pp. 47-52. In Die Tierwelt Deutschland 42. (F. Dahl, ed.). Gustav Fischer Verlag, Jena. Manuscript received 15 June 2000, revised 9 Feb- ruary 2001. 2001. The Journal of Arachnology 29:354-366 UNDER THE INELUENCE: WEBS AND BUILDING BEHAVIOR OE PLESIOMETA ARGYRA (ARANEAE, TETRAGNATHIDAE) WHEN PARASITIZED BY HYMENOEPIMECIS ARGYRAPHAGA (HYMENOPTERA, ICHNEUMONIDAE) William G. Eberhard: Smithsonian Tropical Research Institute, and Escuela de Biologfa, Universidad de Costa Rica, Ciudad Universitaria, Costa Rica, email: archisepsis @ biologia. ucr. ac. cr ABSTRACT. On the evening that it will kill its host, the orb-weaving spider Plesiometa cirgyra, the larva of the ichneumonid wasp Hymenoepimecis argyraphaga induces the spider to perform highly ste- reotyped construction behavior and build an otherwise unique “cocoon web” that is particularly well- designed to support the wasp larva’s cocoon. Cocoon web construction behavior is nearly identical with the early steps in one subroutine of normal orb construction, and is repeated over and over. Usually all other normal orb construction behavior patterns are completely or nearly completely repressed. Experi- mental removal of the larva one or a few hours before cocoon construction would normally occur is sometimes followed by nearly normal cocoon web construction, and sometimes by construction of other highly altered web designs. The mechanism by which the larva induces these changes in the spider’s behavior is thus apparently a fast-acting chemical, with effects that are manifested gradually. Partial recovery of orb designs sometimes occurred several days later. Keywords: Parasite, manipulation of host behavior, orb construction behavior, Plesiometa, Hymenoe- pimecis Manipulation of host behavior by parasites is a widespread phenomenon (Holmes & Bethel 1972; Moore 1984; Barnard & Behnke 1990; Toft et al. 1991; Godfray 1994; Mc- Lachlin 1999; Poulin 2000), but most reports of behavioral modifications, especially those caused by insect parasitoids in other insects, involve only simple behavior patterns such as movement from one habitat to another, adop- tion of sleeping postures, or eating more or less (Wickler 1976; Godfray 1994; McLachlan 1999). Spider behavior is also influenced by insect parasitoids (Schlinger 1987). At least some of these changes may be due to rela- tively simple mechanisms, such as modifica- tion of particular receptors (Jenni et al. 1980). This report concerns an unusually selective behavioral modification by the larva of the parasitoid wasp Hymenoepimecis argyrapha- ga Gauld (Ichneumonidae), which apparently chemically induces expression of the early steps of one subroutine of orb web construc- tion in the spider Plesiometa argyra (Wal- ckenaer 1841) (Tetragnathidae), while sup- pressing all the rest of orb construction behavior (Eberhard 2000a). It may be the most finely directed alteration of host behav- ior ever attributed to an insect parasitoid. It has long been known that psychotropic substances can modify the forms of orb webs (Witt et al. 1968), but the details of how par- ticular steps of the spider’s construction be- havior are affected have never been deter- mined. Elucidation of which aspects of behavior are changed can have important con- sequences for the common use of details of building behavior as taxonomic characters (Eberhard 1982; Hormiga et al. 1995; Gris- wold et al. 1998), as well as how evolutionary transitions may have occurred. It has not al- ways been clear whether or not some variant behavior patterns should be recognized as sep- arate traits (Eberhard 1990). If particular be- havior patterns can be selectively induced, then the case for their independence from oth- er traits, and thus their potential usefulness as characters, is strengthened. Clarification of the behavioral effects of this wasp parasite on web construction behavior thus promises to improve understanding of the organization of behavior within the spider, and of the useful- ness of different behavioral characters in spi- der taxonomy. 354 EBERHARD— PARASITE MODIFICATION OF CONSTRUCTION BEHAVIOR 355 The life cycle of H. argyraphaga is the foF lowing (Eberhard 2000b). The female wasp attacks P, argyra as the spider rests at the hub of its orb, stings it into a temporary (10-15 min.) paralysis, and glues an egg to the spi- der’s abdomen. Subsequently the spider resu- mes normal activity, and builds apparently normal orbs to capture prey during the next approximately 7-14 days while the wasp’s egg hatches and the larva grows. The larva re- mains attached to the surface of the abdomen, and feeds by sucking hemoiymph through small holes it makes in the spider’s abdominal cuticle. The second instar larva, on the night that it will kill its host, induces the spider to construct an otherwise unique “cocoon web” of dragline silk, molts to the third instar, and then kills and consumes the spider. The next evening the larva spins a cocoon hanging by a line from the cocoon web. The larva (which is barely visible through the thin v/alls of the cocoon) pupates about 4 days later, and then emerges as an adult wasp after about 7 more days. METHODS Field observations were made near Parrita, Puntarenas Province, Costa Rica (elev. 10 m) in January and February of 1999 and 2000 in a mature plantation of African oil palm {Elaeis guineensis L.) where spider popula- tions were dense. Web measurements were performed in the morning, and thus did not include webs built later in the day (which may have different designs— Eberhard 1988). Con- struction of cocoon webs made by spiders car- rying wasp larvae was observed indoors near Parrita the night after the spiders were col- lected and transferred onto silk lines from P. argyra orbs that had been fastened to approx- imately horizontal 0.6 m dia. circular wire frames that were hung about 1 m above the floor. Larvae, which would kill their hosts that evening, could be reliably distinguished (15 of 15 cases) from others on the m.orning and af- ternoon of the same day, due to their larger size. Voucher specimens of wasps and spiders have been deposited in the U. S. National Mu- seum of Natural History, the Museum of Comparative Zoology at Harvard, and the Museo de Eetomologia of the Universidad de Costa Rica. The behavior of spiders from which the lar- va had been experimentally removed was ob- Figure 1. — Web of an ueparasitized adult Pie- siometa argyra. Scale bar = 3.0 cm. served after the spiders had been taken to San Antonio de Escazu (elev. 1300 m), where they were kept indoors at room temperature for up to two weeks. On the evening the larva was to be removed, the spider was kept in a small container (6 cm dia.) in which it could not spin a web, and then placed on a wire frame as soon as the larva was removed between 2100 and 0200 h. Because the spiders seemed to need air movement to induce web construc- tion, they were not kept in cages, but allowed to range freely in rooms. RESULTS FieM.^The orbs of spiders carrying wasp eggs and larvae were not distinguishable from the more or less horizontal, moderately open- meshed orbs of unparasitized spiders (Figs. 1, 2) (ANCOVA analyses showed no significant effects of parasitism by larvae, or by eggs and larvae {P = 0.91, 0.40). Even parasitized spi- ders found the morning of the day on which they would be killed by the wasp larva were on freshly made, apparently normal orbs. Oth- er than orbs, the only other webs on which unparasitized spiders occurred were small molting webs (Eberhard et al. 1993). These 356 THE JOURNAL OF ARACHNOLOGY 60= CO Cl O o ^ O 0 — ^ 5 D u • '■ CO 40 20 m © © O o o o □ o © o © a o © o o o © © ^ o o □ o o • g o omit • O O o o I© o © □ penult, no paras. 0 penult, with egg ■ penult, with larva ©adult no paras, ©adult with egg • adult with larva 20 30 Number of radii Figure 2. — Numbers of radii and sticky spiral loops (mean number of loops directly above, below, and to the sides of hub) in webs of spiders in the field that were parasitized (filled symbols) and unparasitized (open symbols). No differences between parasitized and unparasitized individuals were apparent. webs were rare, and several newly molted in- dividuals lacked such webs. Despite the dense spider populations, no egg sacs or webs as- sociated with egg sacs were seen; egg sacs may be hidden in leaf litter, as in the closely related Leucauge mariana (Keyserling) (Ibar- ra et al. 1991; V. Mendez pers. comm.). More than 100 cocoon webs were observed in the field. They almost always consisted of a few lines that radiated in a more or less hor- izontal plane from a “hub” or central area, where the cocoon’s suspension line was at- tached, and were each attached directly to a support (Eberhard 2000a, Figs. 3-5). Most ra- EBERHARD— PARASITE MODIFICATION OF CONSTRUCTION BEHAVIOR 357 Figures 3-5. — Typical cocoon web. 3. Dorsal view. Scale bar = 3.0 cm; 4. Detail of attachments to a leaf. Scale bar = 0.5 cm; 5. Lateral view. Scale bar = 3.0 cm. 358 THE JOURNAL OF ARACHNOLOGY dial lines had many branches near their tips and were thus attached at many adjacent points to the substrate (Fig. 4). They were also sometimes attached at multiple points in the central area. There were several other indica- tions, in addition to the planar arrangement of radial lines, that the webs from which cocoons were suspended represented modified orbs. Some cocoon webs had circular lines similar to those at the hubs of normal orbs (17% of 41 webs checked for this detail) (Figs. 6, 7), though in no case was the central portion of the hub empty, as in normal orbs. Some had one or more frame lines connecting the radial lines (29% of 42 webs checked for this detail) (Fig. 6). These frames were typically much shorter and nearer the hub than were the frame lines of normal orbs (Fig. 1). The most elab- orate cocoon web had a distinct hub, frame lines, and a mesh above and below the hub. At the opposite extreme, the two simplest co- coon webs consisted of a single strong line with the larva or the cocoon hanging from the central portion. Cocoon webs spanned smaller spaces than normal orbs of mature females. The distance between the two most distant points of attach- ment of anchor lines was smaller in cocoon webs (mean 36.6 ± 17.2 cm in a sample of 38) than in orbs (99.6 ± 47.5 cm in a sample of 31) {P < 0.001 with Mann Whitney U~ Test). These cocoon webs also had fewer an- chor lines (lines directed to the substrate) (mean 3.9 ±1.5 for cocoon webs, 5.3 ± 1.7 for the orbs; P < 0.001 with Mann- Whitney U-Test). Construction behavior. — Cocoon web construction behavior, observed in five spiders captured in the field the same day with larvae and a sixth three days after being collected, was very consistent. Early in the evening, the spider built several lines, repeatedly removing and shifting the points of attachment as typi- cally occurs during the preliminaries of orb construction of many species of orb weavers (Tilquin 1942; Eberhard 1990). It then re- mained more or less immobile until between 23:30 and 01:00, when construction activity occurred in bursts. Typically the spider added one to several radial lines in quick succession, and then spent a minute or more (up to 30 min) immobile at the hub before the next burst of activity. The spider's movements showed no signs of weakness or vacillation, and it moved directly from one attachment to the next as in normal frame and radius construc- tion. Radial lines were all in nearly the same plane and were added to the web using two similar, simple behavior patterns (Fig. 8, A and B). The spider began by attaching its dragline at the hub, then walked toward the substrate along a radial line, walked along the substrate a short distance and attached the line it had laid from the hub (Aj, Bj). Then it re- turned to the hub, laying a second dragline as it walked along this line or another radial line that it had laid previously and attached it at the hub (A2, B2). When the substrate was thin (a strand of wire, for instance) the spider usu- ally moved to the opposite side to make the attachment before returning to the hub, as is typical of frame construction in orbs (Tilquin 1942; Eberhard 1990). The two patterns differed in that either the lines were laid without attachments to previ- ously laid radial lines (A,, A2 in Fig. 8), or (more often) the spider attached its dragline one or more times to radial lines both on the way out and on the way back to the hub (B,, B2 in Fig. 8). Consecutive radial lines were always laid in different directions, as in orb construction by other araneoid spiders (Ti- lquin 1942; Dugdale 1969; Le Guelte 1966; Witt et al. 1968; Eberhard 1982). Each radial line was reinforced repeatedly, and the total amount of dragline silk in a cocoon web prob- ably represented a major fraction of that in an orb. The estimated total numbers of radial lines in two finished cocoon webs were 36 and 30. Thus the number of radial trips was on the same order as the typical number of radii (20- 35) in a normal orb (Fig. 1). The behavior of one further individual, col- lected four days previously and observed in San Antonio de Escazu, was very different. The spider descended to the floor about 1.5 below the wire hoop, formed a “hub” where several lines converged about 1 cm above the floor, and then made 5-10 very long radial excursions (up to 1.3 m each) walking on the surface of the floor. As it moved away from the hub it walked in a nearly straight line, at- taching its drag line periodically to the floor, but in some cases it gradually made an arc of up to more than 180° before it turned back and slowly retraced its path back to the hub. On at least four occasions the spider encoun- EBERHARD— PARASITE MODIFICATION OF CONSTRUCTION BEHAVIOR 359 Figures 6-7. — Dorsal view of an unusually elaborate cocoon web with hub loops and a frame line. Scale bars = 2.5 and 1,0 cm respectively. 360 THE JOURNAL OF ARACHNOLOGY Figure 8. — Diagrammatic representations of the sequences of behavior during construction of a cocoon web (A1-A2, and B1-B2) and a frame line in a typical orb (C1-C4). Stippling represents substrate, black spots represent points where the dragline was attached, and dashed lines represent lines laid earlier in the sequence (C,-C4 after Eberhard 1990). Cocoon web construction corresponds to the behavior in Cj and the first part of C2. EBERHARD— PARASITE MODIFICATION OF CONSTRUCTION BEHAVIOR 361 Figure 9. — Dorsal view of a web made by a ma- ture male on the night that it was killed by a wasp larva. Note multiple attachments to substrate of some radial lines (as in Figs. 4 and 10). Scale bar = 3.0 cm. tered an object that it could have climbed and thus have raised its drag line off the floor, but instead it struggled on across the floor. When I then removed the larva, and replaced the spi- der on the wire hoop after breaking the lines leading downward toward the floor, the spider again descended to the floor where it made another hub. Wasps generally avoided parasitizing ma- ture males (Eberhard 2000), but two larvae on mature males matured and made cocoons in captivity. One male spider did not make a co- coon v/eb (or indeed any supporting structure whatsoever); the wasp’s cocoon hung from a single short strand of spider silk. The second parasitized mature ro.ale spider, however, built an extensive web that resembled a cocoon web in being more or less planar, and having many attachments to the substrate on the eight that it was killed and consumed by the larva (Fig. 9). Experimental removal of larvae. — Larvae were removed from 22 spiders in captivity on the evening when the spider was to be killed. Four spiders built no webs that night. The 18 webs that were built were of three types. Three were more or less typical cocoon webs, with a low number of radial lines which were composed of multiple strands that were at- tached at many adjacent points on the sub- strate, and more or less converged at the hub (Figs. 10, 11). A fourth spider, which had al- ready begun cocoon web construction when I removed the larva, resumed cocoon web con- struction but did not return consistently to the hub, and made two additional “hubs”. Sim- plified, “vestigial” webs, that had only a few more or less radial lines converging at the point where the spider rested and large masses of silk lines loosely packed together near the central area, were built by 13 spiders (Figs. 12, 13). These radial lines were attached to the substrate at only one or at most a few points. Vestigial webs never had hub loops, temporary spirals or sticky lines, and only sel- dom had recognizable frame lines. One further web was a nearly normal orb, except that the center of the hub was not removed and some portions of the sticky spiral lacked sticky balls. The construction behavior of spiders from which larvae had been removed was observed for two cocoon webs and three vestigial webs. Cocoon web construction was very similar to that described above for spiders carrying lar- vae, including the frequent pauses between bursts of construction behavior, except that on some occasions the spider failed to attach its drag line at the hub when it returned after lay- ing a radial line. The drag line laid on the next trip away from the hub was thus not attached at the hub, but originated part way out the previous radial line (line 2~4 in B2 of Fig. 14). When this behavior was repeated over and over, the hub gradually expanded and became dispersed. The resulting web had large num- bers of more or less radial lines attached to the substrate close to each other, but a diffuse central area (Fig. 15). During vestigial web construction, the spi- der also made radial lines attached to the sub- strate just as above. On some return trips to the hub area, however, it broke and removed these lines, reeling them up and leaving them packed loosely together attached to the web. The final product of this process of repeatedly laying and then removing lines was a scanty array of more or less radial lines, and one or more large masses of fluff (Fig. 13). None of the 22 experimental spiders that built webs died on the evening the wasp larva was removed. In nine cases the spider built a second web on the following night, and the second web was of the same type built on the 362 THE JOURNAL OF ARACHNOLOGY Figures 10-13. — Webs made by spiders from which the wasp larva was removed on the night when the larva would have normally killed the spider. 10. Cocoon-web type, in which the few radial lines each had multiple attachments to the substrate. Scale bar = 3.0 cm. 1 1. Close-up of hub of web in Fig. 10. Scale bar = 1.0 cm. 12. “Vestigial” type web, in which a few radial lines were attached singly to the substrate (heavy white lines are from previous web of another spider). Scale bar = 3.0 cm. 13. Close-up of the hub of a vestigial web (different web from that in Fig. 12), showing several masses of fluff. Scale bar = 1.0 cm; all wire hoops were horizontal. first (two cocoon webs, seven vestigial webs). Five of the second vestigial webs had at least one hub loop. Due to deaths and emigrations, it was not possible to follow the spiders’ be- havior systematically on subsequent nights. Two spiders survived for a week, and gradu- ally built webs that were progressively more orb-like though still substantially altered (Fig. 16). DISCUSSION Comparison of cocoon web construction behavior with the early stages of normal orb construction (Eberhard 1990) indicates that it is probably homologous with the early steps of type “D” frame construction (Fig. 8 Cj- C4). Most anchor line construction in an orb involves removal of lines already in place, or shifting their attachments to each other (Ti- Iquin 1942; Eberhard 1990), but neither of these behavior patterns was ever seen during cocoon web construction. In type D anchor construction, however, which sometimes oc- curs as part of frame construction, the early stages do not involve removing or shifting lines (Eig. 8 C,, C2). Premature termination of this type of frame construction behavior when the spider returns to the hub after the first at- tachment to the substrate and followed by at- tachment of the spider’s drag line at the cen- tral area (x in Eig. 8 C2), would result in a sequence of operations identical to type A co- coon web construction (Fig. 8 A). Adding at- tachments to the line already in place on the way out would result in a sequence similar or identical to the second type of cocoon web construction behavior (Fig. 8 B). Similar at- tachments sometimes occur in the closely re- lated L. mariana during frame construction of types “A” and “C” of Eberhard (1990) but were not seen in conjunction with type D of Eberhard (1990) (the same individual often performed more than one type while building a given orb). A further resemblance to attach- EBERHARD— PARASITE MODIFICATION OF CONSTRUCTION BEHAVIOR 363 Figure 14. — Diagrammatic representations of co- coon web construction behavior of a spider with a wasp larva (Aj, A2) and a spider from which the wasp larva had been experimentally removed (Bi, B2). The experimental spider sometimes omitted the final attachment at the hub typical of cocoon web construction (attachment 3 in Ai and A2; see also Fig. 8 A2 and B2); when it moved away from the hub to make the next radial line, the dragline was thus displaced away from the hub (line 2-4 in Bj). Repeated omissions of this attachment resulted in a diffuse central area of the web (Fig. 15). ments of anchor lines built during orb web- construction by other orb weavers (Tilquin 1942: Eberhard 1990) was the attachmeet of radial lines to thin objects by moving to the opposite side of the object just before attach- ing. Thus, the spider built the cocoon web by apparently repeating the first portions of one type of frame construction over and over. Fur- Figure 15. — Cocoon-type web with dispersed hub built by a spider from which the larva was removed on the evening on which it would have normally killed its host. Scale bar = 3.0 cm; wire hoop was horizontal. Figure 16. — Orb-like web built by a partially re- covered spider. The wasp larva had been removed five days earlier, on the evening when it would have killed the spider, and the spider had spun a typical vestigial web on that night. Scale bar = 2.0 cm. ther evidence that cocoon webs were homol- ogous with orbs is the fact that when these webs had more than three radial lines, these were nearly always in approximately the same plane. In addition, some cocoon webs had frame lines, and a few had hub loops (Figs. 6, 7). The homology of cocoon and orb webs em- phasizes that perhaps the most extraordinary aspect of the wasp larva’s effect on the spider was not so much what the spider did, but what it did not do. Many aspects of normal orb con- 364 THE JOURNAL OF ARACHNOLOGY struction were completely absent, including both breaking, reeling up and replacing lines (e.g. Fig. 8 C4), and breaking and then re-at- taching lines. These two behavior patterns form integral parts of most types of both frame and radius construction in normal orbs of this and other species (Tilquin 1942; Eber- hard 1982 1990; Coddington 1986). A single failure to repress these behavior patterns could be disastrous for the wasp larva, as it would result in the removal of the many-stranded ca- ble of radial lines, and its replacement with a much weaker line. This indeed occurred in the vestigial webs built by spiders from which lar- vae were experimentally removed. Also com- pletely missing were production of the tem- porary spiral and sticky spiral, and removal of the central portion of the hub at the end of orb construction, which again would have resulted in considerable weakening of the support for the wasp’s cocoon. These differences between cocoon webs and normal orbs are appropriate to make the cocoon web stronger and less likely to be damaged by falling debris, and thus a more durable support for the wasp’s cocoon than an orb would be. Strong support for the cocoon may be important for the wasp’s survival, as in the related Hymenoe- pimecis robertsae some pupae died when heavy rains damaged cocoons (Fincke et al. 1990). The importance of the precision of the be- havior induced in the spider is also illustrated by the effect of occasional omission of one normal detail, the final attachment at the hub after a radial line was built (Fig. 8 A2, B2) that was seen in some spiders from which the lar- vae were experimentally removed. The result- ing lack of a clear central point of conver- gence produced webs that were much less appropriately designed to support the wasp’s cocoon (Fig. 15). It is not clear whether the aberrant behavior of one spider that laid radial lines on the surface of the floor instead of in the air was something that happens in nature (such webs would be missed in the field) or was an artifact of captivity. In some cases, claims that modification of host behavior associated with parasitism rep- resents an evolved adaptation by the parasite to promote its own reproduction have been controversial (Toft et ah 1991; Poulin 2000). There can be little doubt on this score with the species of this study, as the cocoon web design is both unprecedented in P. argyra or any closely related orb weaver, and seems es- pecially appropriately designed to increase the survival of the wasp. Induction of spinning behavior also occurs in several families of spi- ders parasitized by acrocerid flies; the spider spins a thin cell similar to that made just prior to moulting, and the larva clings to the web after emerging from the spider (Schlinger 1952, 1960, 1987). The changes in the behavior of P. argyra are induced chemically rather than by direct physical interference with the spider’s nervous system. The wasp larva contacts only the sur- face of the spider’s abdomen and limits itself to making small holes through which it im- bibes hemolymph (Eberhard 2000a,b). In ad- dition, some spiders built normal cocoon webs after the larva was removed. Some ichneu- monids modify host behavior and physiology via products injected by the female wasp when she oviposits (Gauld 1995). However, the lack of web modification in the days im- mediately following the attack by the wasp, the sudden abrupt shift in behavior that is co- ordinated with maturation of the larva, and the changes in webs produced by removing the larvae, all argue that the larva rather than the adult female wasp induced modified web con- struction behavior. Secretion of neuromodu- lators by parasitoid larvae has been implicated in behavioral changes produced in some insect hosts (Beckage 1997). The variety of web forms and construction behavior observed when the larva was removed prematurely sug- gest a complex, gradual effect rather than an abrupt, simple modification. The ability of Hymenoepimecis argyrapha- ga to induce specific behavior patterns in spi- ders indicates that even these fine behavioral details are independent units or modules at some level within the spider, and not just ar- tificial constructs. The additional web forms produced by experimentally removing larvae from spiders suggest even further subdivisions of building behavior. The problem of what constitutes a biologically realistic behavioral unit is crucial in the use of behavior patterns as taxonomic characters in orb-weavers Eber- hard 1982; Coddington 1986, 1990; Scharff & Coddington 1997; Griswold et al. 1998) as well as in other animals (Wenzel 1992). The results of this study suggest that it is reason- able to attempt to use even finer behavioral EBERHARD— PARASITE MODIFICATION OF CONSTRUCTION BEHAVIOR 365 details than those that have been used previ- ously in orb weaver taxonomy. The cocoon web of R argyra is similar to the secondarily reduced “asterisk” web found by Stowe (1978) in the distantly related araneid Wixia ectypa (Walckenaer), Whether or not this evo- lutionary transition involved chemical chang- es similar to those produced by H, argyra- phaga remains to be determined. ACKNOWLEDGEMENTS I thank LD. Gauid and H.W. Levi for iden- tifying the wasp and the spider respectively. This research was financed by the Smithson- ian Tropical Research Institute and the Vicer- rectoria de Investigacioe of the Universidad de Costa Rica. LITERATURE CITED Barnard, CJ. & J.M. Behnke. 1990. Parasitism and Host Behaviour. Taylor & Francis, London. Beckage, N.E. 1997. New insights: how parasites and pathogens alter the endocrine physiology and development of insect hosts. Pp. 3-36. In Para- sites and Pathogens, Effects on Host Hormones and Behavior (N.E. Beckage, ed). Chapman and Hail, New York, New York. Coddington, J.A. 1986. Monophyletic origin of orb webs. Pp. 319-363, In Spiders, Webs, Behavior and Evolution (W. Shear, ed). Stanford Univer- sity Press. Coddington, J.A. 1990. Cladistics and spider clas- sification: araneomorph phylogeny and the monophyly of orb weavers (Araneae: Araneo- morphae; Orbicular! ae). Acta Zoologica Fennica 190:75-87. Dugdale, B.E. 1969. The weaving of an engineering masterpiece, a spider’s orb, done at Frysoe Lakes, N. J., August 8, 1942, as observed and reported by B. E. Dugdale, structural engineer. Natural History 78(3):36-41. Eberhard, W.G. 1982. Behavioral characters for the higher classification of orb-weaving spiders. Evolution 36:1067-1095. Eberhard, W.G. 1988. Behavioral flexibility in orb web construction: effects of silk supply in dif- ferent glands. Journal of Arachnology 16:303- 320. Eberhard, W.G. 1990. Early stages of orb construc- tion by Philoponella vicina, Leucauge mariana, and Nephila clavipes (Araneae: Uloboridae and Tetragnathidae), and their phylogenetic implica- tions, Journal of Arachnology 18:205-234. Eberhard, W.G. 2000a. Manipulation of host spider web construction behavior by a wasp larva. Na- ture 406:255-256. Eberhard, W.G. 2000b. The natural history and be- havior of a parasitoid, Hymenoepimecis argyra- phaga (Hymeeoptera, Ichneumonidae) and its host, the spider Plesiometa argyra (Araneae, Te- tragnathidae). Journal of Hymenoptera Research 9:220-240. Eberhard, W.G,, S. Guzman-Gomez & K.M. Catley, 1993. Correlation between spermathecal mor- phology and mating systems in spiders. Biolog- ical Journal of the Linnean Society 50:197-209. Fiecke, O., L. Higgins & E. Rojas. 1990, Parasitism of Nephila clavipes (Araneae, Tetragnathidae) by an ichneumonid (Hymenoptera, Polysphinctini) in Panama. Journal of Arachnology 18:321-329. Gauid, LD. 2000. The re-definition of pimpline ge- nus Hymenoepimecis (Hymenoptera: Ichneu- monidae) with a description of putatively ple- siomorphic new Costa Rican species. Journal of Hymenoptera Research 9:213-219. Gauid, I.E., J. Ugalde & P.E. Hanson. 1998. Guia de los Pimplinae de Costa Rica. Revista de Biol- ogia Tropical. 46 (SupL 1): 1-189. Godfray, H.C. J. 1994. Parasitoids Behavioral and Evolutionary Ecology (Princeton University Press, Princeton, New Jersey, Griswold, C. E., J.A. Coddington, G. Hormiga & N. Scharff. 1998. Phylogeny of the orb-web building spiders (Araneae, Orbiculariae: Deino- poidea, Araneoidea). Zoological Journal of the Linnean Society 123:1-99. Hormiga, G., W.G. Eberhard & J.A. Coddington. 1995. Web construction behavior in Australian Phonognatha and the phylogeny of nephiline and tetragnathid spiders (Araneae, Tetragnathidae). Australian Journal of Zoology 43:313-364. Ibarra, G., O. Jaramillo M. & T. de Castro C. 1991. Observaciones sobre el comportamieeto reprod- uctor de las aranas del geeero Leucauge (Ara- neae: Tetragnathidae) del Soconusco, Chiapas, Mexico. Pp 65-66. Memorias del 26 Congreso Nacional de Entomologia, Veracruz, Mexico. Jenni, L., D.H. Molyneux, J.L. Livesley & R. Gal- un. 1980. Feeding behaviour of tsetse flies in- fected with salivarian trypanosomes. Nature 283: 383-385. Le Guelte, L. 1966. Structure de la toile de Zygiella x-notata Cl. (Araignees, Argiopidae) et facteurs qui regissent le comportemente de I’araignee pendant la construction de la toile. These Publi- cations de FUniversite Nancy Faculte des Sci- ences. McLachlan, A. 1999. Parasites promote mating suc- cess: the case of a midge and a mite. Animal Behaviour 57:1199-1205. Poulin, R, 1994. Meta- analysis of parasite-induced behavioural changes. Animal Behaviour 48:137- 146. Poulin, R. 2000. Manipulation of host behaviour by parasites: a weakening paradigm? Proceedings of the Royal Society of London, Series B 267:787- 792. 366 THE JOURNAL OF ARACHNOLOGY Scharff, N. & J.A. Coddington. 1997. A phyloge- netic analysis of the orb-weaving spider family Araneidae (Arachnida, Araneae). Zoological Journal of the Linnean Society 120:355-434. Schlinger, E.I. 1952. The emergence, feeding hab- its, and host of Opsebius diligens Osten Sacken. Pan-Pacific Entomology 28:7-12. Schlinger, E.I, 1960. A revision of the genus Ogco- des Latrielle with particular reference to species of the Western Hemisphere. Proceedings of the United States National Museum 1 1 1 :227-336. Schlinger, E.I. 1987. The biology of Acroceridae (Diptera): true endoparasitoids of spiders, Pp. 319-327, In Ecophysiology of Spiders (W. Nen- twig, ed.). Springer- Verlag, New York. Stowe, M.K. 1978. Observations of two nocturnal orbweavers that build specialized webs: Scolod- erus cordatus and Wixia ectypa (Araneae: Ara- neidae). Journal of Arachnology 6:141-146. Tilquin, A. 1942. La Toile Geometrique des Araig- nees. Presses Univ. France, Paris. Toft, C.A., A. Aeschlimann & L. Bolis. 1991. Par- asite-Host Associations. Oxford University Press, New York. Wenzel, J.W. 1992. Behavioral homology and phy- logeny. Annual Review of Ecology and System- atics 22:361-381. Wickler, W. 1976. Evolution-oriented ethology, kin selection, and altruistic parasites. Zeitschrift fiir Tierpsychologie 42:200-214. Witt, P.N., C. Reed & D.B. Peakall. 1968. A Spi- der’s Web Springer, New York. Manuscript received 15 December 2000, revised 1 May 200L 2001. The Journal of Arachnology 29:367-377 LIFE-CYCLES OF FOUR SPECIES OF PARDOSA (ARANEAE, LYCOSIDAE) FROM THE ISLAND OF NEWFOUNDLAND, CANADA J. Rs Pickavance: Biology Department, Memorial University of Newfoundland, St. John’s, Newfoundland, Canada A IB 3X9. email: rpickava@muii.ca ABSTRACT. Populations of four species of Pardosa, P. fuscula, P. groenlandica, P. hyperborea and P. moesta, were sampled during summer 1997 on the west coast of the Island of Newfoundland, Canada. Measurements of carapace width indicated that all four species fit a biennial life-cycle model where new individuals join the population in summer, live through the following winter, grow throughout the next year, live through the next winter, and thee mature, breed and die in the year following their second winter. All species showed only one defined recruitment of new spiderliegs during the sampling period, but at least two species may have extended periods of recruitment and some individuals may have an extended life-cycle. Keywords: Lycosidae, Pardosa, life-cycle At the beginning of the 20* century the conventional wisdom about araneomorph spi- der life-cycles was that most were annuals, ei- ther spring breeders or summer-autumn breed- ers (Emerton 1902). Palmgren (1939) provided one of the first exceptions when he described the two-year cycle of Dolomedes fimbriatus (Cierck 1757) where juveniles overwintered twice. Cloudsley-Thompson (1955) concluded that individuals of all three British species of Amaurobius C. L. Koch 1837 lived for about two years, overwintered twice, and spent their second winter as adults. Hackman (1957) described a similar two-year cycle for Trochosa ruricola (De Geer 1778). Dondale’s (1961) seminal work presented quantitative data for five species of spiders in Nova Scotia, Canada: Araniella displicata (Hentz 1847), Philodromus rufus Walckeeaer 1 826, P. cespitum (Walckenaer 1 802) and Eris militaris (Hentz 1845) were shown to be true biennials, while Pelegrina proterva (Walcken- aer 1837) was annual. He also concluded that nine other widespread and abundant spiders were biennials. In the last decades of the 20* century a number of studies have not only clearly demonstrated annual and biennial life histories for several different species, but also reported permutations of these two basic life- cycles. That is, within these two general cat- egories there are species that mature and breed at different times of year and species that are intermediates between strictly annual and strictly biennial (e.g. Eason & Whitcomb 1965; Toft 1976, 1979; Doedale 1977; Strat- ton & Lowrie 1984). In addition to such permutations of the an- nual/bieneial theme, the plasticity of spider life history has been demonstrated. For ex- ample, the same species may change from an- nual to biennial depending on geographical lo- cation, such as Philodromus cespitum that was annual on the warmer Niagara Peninsula, On- tario, Canada (Putman 1967), but biennial in colder Nova Scotia, Canada (Doedale 1961). Pardosa lugubris (Walckenaer 1802) was an- nual in the Netherlands (Vlijm et al. 1963) but biennial in Scotland, and this was attributed to differences in summer temperatures (Edgar 1971a, 1972). In addition, individuals of the same population may extend their life-cycle under particular circumstances. Edgar (1972) showed that P. lugubris in the Netherlands varied between annual and biennial depending on environmental conditions. Workman (1978) showed that in Norfolk, U.K., Trocho- sa terricola Thorell 1856, usually biennial with the second overwintering as adults, was occasionally triennial when juveniles hatched from late or second cocoons overwintered three times before breeding in their fourth year. Leech (1966) suggested that even more extended life-cycles may occur in particularly cold conditions. He surmised that two species 367 368 THE JOURNAL OF ARACHNOLOGY from Hazen (Ellesmere Island, NWT, Cana- da), Pardosa glacialis (Thorell 1872) and A/- opecosa exasperans (O. Pickard-Cambridge 1877), had a life-span of six or seven years, but that surmise was based on the unsupported assumption that each of the estimated six or seven instars lasted one year. These points raise a number of questions. Do spider species not yet examined have similar life histories to those already described? Do species in places not yet examined have similar life histories to those in known places? Do species at latitudes farther north than some of those previously examined show extended life histories, for ex- ample intermediate between those document- ed in Nova Scotia (Dondale 1961) and those hypothesised on Ellesmere Island (Leech 1966)? The Island of Newfoundland is an appro- priate location to examine these questions. The life histories of the species in this study have not been described before with the fol- lowing exceptions. Ricards (1967) reported the life history of what he called P. groenlcm- dica (Thorell 1872) in Montana, and Schmoll- er (1970) reported on what he called P. tristis Keyserling 1887 (identified as P. groenlandi- ca by Dondale 1999) in Colorado. But as Dondale (1999) pointed out, both authors were in fact dealing with complexes of two or more species, not monospecific populations of P. groenlandica, so their conclusions have limited significance. Buddie (2000) reported the life-cycle of Pardosa moesta Banks 1892 in Alberta, and his observations are directly relevant here. Life histories of spiders on the Island of Newfoundland have essentially nev- er been investigated. Hackman (1954) report- ed 220 species from the Island, but drew con- clusions for only one: Trochosa terricola was described as biennial, although the data pre- sented could support other interpretations. The present study sites on the Island of Newfoundland, at approximately 50°N, are farther north than previous life history work with the following exceptions. Buddie (2000) and Zimmerman & Spence (1998) reported life-cycles of lycosids and a pisaurid, respec- tively, from approximately 54°N in central Al- berta. However, both these studies were con- ducted at the George Lake area dominated by hardwoods such as aspen (indicative of higher summer temperatures) in the boreal transition region, whereas the present study was con- ducted in boreal forest dominated by fir and spruce (indicative of lower summer tempera- tures) (Ecological Stratification Working Group 1995). Leech (1966) reported from the Canadian arctic at approximately 82°N. Some European work has been conducted at lati- tudes farther north than Insular Newfound- land. Eor example. Toft (1976) reported from Denmark and Edgar (1971a) from Scotland, both at approximately 56°N. However, climate is not simply determined by latitude, and the generally more temperate European climate is indicated by the beech woods of the former study and the oak woods of the latter. METHODS Species and localities. — Lour species of Pardosa C. L. Koch 1847 (Lycosidae) were chosen for this study: P. fuscula (Thorell 1875), P. groenlandica, P. hyperborea (Tho- rell 1872) and P. moesta. Full descriptions of these species can be found in Dondale & Red- ner (1990). They were chosen both because the taxonomy of most Canadian lycosids is well established (Dondale & Redner 1990) so conclusions could be confidently assigned to individual species and preliminary investiga- tions in 1995 and 1996 found dense popula- tions of these species. Such dense populations lend themselves to sampling by hand as op- posed to using pitfall traps. Pitfall traps are useful measures of activity and have a long history of employment in ecological studies, but they are selective in trapping different spe- cies and different life-stages (Berghe 1992; Topping & Sunderland 1992). The Island of Newfoundland is in the boreal shield ecozone where the climate is heavily influenced by arctic currents, many areas are exposed to particularly harsh climatic condi- tions and the landscape is dominated by spruce-fir forest with extensive peatlands. Within that ecozone, the populations of this study were in, or immediately adjacent to, the northern peninsula ecoregion (South 1983; Ecological Stratification Working Group 1995). The populations chosen were all in Gros Mome National Park, Newfoundland, and were therefore largely protected from human inter- ference. The P. fuscula population was on an extensive peatland immediately below and around the highest land on top of Partridge- beiTy Hill behind the community of Woody PICKA VANCE— FANDOM LIFE-CYCLES 369 Point (49°30.2'N, 57°56,9'W). The P. groen- landica population was at the back of a pebble- cobble beach immediately north of the mouth of Baker’s Brook (49°39.5'N, 57°57.7'W). The P. hyperborea population was on the extensive treeless heath on the higher parts of Partridge- berry Hill behind Woody Point (49°30.0'N, 57°56.6'W). The P. moesta population was on the treeless coastal meadow immediately above and behind the beach at Lower Head, Shallow Bay (49°57.3'N, 57°46.2'W). Voucher speci- mens are deposited in the Newfoundland Mu- seum (catalogue numbers NFM ARA-01, -02, -03 and -04). Measurements.- — Doedale (1961) conclud- ed that carapace width (CW) was the most generally useful measurement to distinguish life history stages but the species he examined did not include lycosids. Hagstrum (1971) confirmed the essentials of that work with measurements of the lycosid Alopecosa kochi (Keyserling 1877). However, Toft (1976) claimed that linear measurement of tibia I gave the best discrimination between instars, but in support presented data for only one spe- cies, the linyphiid Helophora insignis (Black- wall 1841), His data supported the superiority of tibia I measurements for that particular spe- cies, but what is applicable to a linyphyiid may not be applicable to lycosids. To resolve this question, CW and tibia I of a number of samples of the lycosids of the present study were compared. Life-cycles.”-“Critical information for all four species was whether adults survived the winter and the general nature of the popula- tion immediately after the winter. Samples were therefore taken just after snow-melt at each site (May or early June). Preliminary ob- servations in 1995 and 1996 indicated that no adults were seen until July, except adult P. groenlandica which appeared in June. There- fore in 1997, sampling of P. groenlandica commenced in May and of the other species in June. At each sampling I tried to catch at least 50 specimens. I achieved this in all but the June samples of P. groenlandica and P. fuscula, when bad weather made these two larger and less numerous species harder to find. In one instance, P. groenlandica in June, it was necessary to sample the population on two consecutive days, June 1 and 2. Sampling dates and numbers caught for each species in 1997 are as follows. P. moesta: June 2, 98; July 7, 99; August 14, 143; September 14, 88. P. hyperborea: June 5, 74; July 3, 75; August 11, 55; September 15, 66. P. fuscula: June 5, 48; July 9, 135; August 11, 54; September 15, 104. F. groenlandica: May 15, 60; June 1 and 2 combined, 37; July 4, 68; August 12, 62; September 13, 74. Spiders were caught with an aspirator and transferred to snap-cap plastic vials. Only one spider was put in each vial to avoid intraspecific aggression and cannibal- ism. Spiders were taken to the laboratory, placed in a deep-freeze until comatose and then placed directly into 75% ethanol for stor- age and later examination. Although considered superior to pitfall traps for present purposes, hand collection nevertheless has two principal imperfections: lycosids are weather- sensitive (Vlijm & Kes- sler-Geschiere 1967) and may not be visible except under warm and windless conditions, and data from hand collections can be mis- leading because of conscious or unconscious size-selection by the collector. To offset weather problems, collections were made as far as possible on favorable days, when at least two individuals of the selected species were visible in a five-minute preliminary in- spection. To offset size-selection, a conscious effort was made to catch all individuals seen of the target species irrespective of size. Life-stages identified.- — -Three separate life-stages were identified: immature, subadult and mature. Mature contains a single instar and is clearly defined as adult males with fully developed functional palps and adult females with fully developed functional epigyna. The boundary between mature and subadult is clear cut. Subadults are close to becoming ma- ture, presumably within a molt or two of ma- turity (although total number of molts and number of molts within the subadult stage are unknown). Secondary sexual characters are pronounced but not complete: male palpal tar- si are significantly swollen with ventral sur- faces showing pronounced ogee curves; de- veloping female epigyna have obvious lateral sclerites. Immatures are either smaller speci- mens showing no differences that would in- dicate their future sex, or larger specimens with males showing at most a slight thicken- ing of the palpal tarsi and females showing no development of the lateral epigynal sclerites and distinguishable from potential males only by virtue of having no sign of any palpal 370 THE JOURNAL OF ARACHNOLOGY Table 1. — Males and females (raw data) caught in 1997. (%) = females with cocoons. P. fuscula d, ? P. groenlandica d, ? P. hyperborea d, ? P. moesta d, $ June 0, 0 (0%) 2, 4 (0%) 0, 0 (0%) 0, 0 (0%) July 19, 18 (89%) 2, 2 (50%) 21, 15 (47%) 35, 26 (4%) Aug. 1, 6 (83%) 0, 5 (20%) 0, 24 (54%) 17, 50 (92%) Sept. 0, 3 (0%) 0, 1 (100%) 0, 30 (80%) 0, 42 (43%) swelling. The immature life-stage contains several instars. The boundary between im- mature and subadult is not always clear cut, and conclusions drawn from the data must be in light of this imprecision. In addition to these three stages, very small newly or re- cently hatched spiderlings will be referred to occasionally. These were easily identified be- cause in previous years females of all four species caiTying spiderlings had been caught and so the size range of new spiders was well known. RESULTS Measurements. — One example will illus- trate the relative usefulness of measurements of the two different body parts. Figure 1 com- pares measurements of CW and tibia I length of the July 7 R. moesta sample and shows that both yield essentially similar information with the two different cohorts definitively separated and both larger and smaller cohorts spread over five or six units. Life-cycles. — Numbers of individuals with different CWs for all four species populations are displayed in Figs. 2-5. Numbers of males and females caught and the percentage of those females carrying a cocoon are shown in Table 1. All four species fit the generalized life-cycle illustrated in Fig. 6. Adults appear around the end of June and the sexes are pre- sent in approximately equal proportions in July. Males then either vanish by August or decline rapidly in numbers and have gone by September. Based on the synchrony of the sexes, the mating season is principally late June and July. Females persist to at least mid September (when sampling stopped) but do not survive the winter. The breeding season (here defined as females carrying cocoons) is July through to September. New spiders hatch in mid to late summer and join a population consisting partly of mid-size immatures (hatched the previous year) and partly of ma- tures that have just produced the new young. The two cohorts of immature individuals pre- sent at the end of the year survive the follow- ing winter and by the next spring have grown. The cohort of smaller immatures now be- comes the mid-size cohort that will grow throughout the year. The cohort of larger im- matures becomes sub-adult at or before the beginning of the year and then matures, breeds to produce a cohort of new spiderlings and in turn dies before the end of the year. In each species only one recruitment of new spi- derlings was seen within the sampling period and there was no direct evidence that females make more than one cocoon. Individual spe- cies are considered below. Pardosa moesta males and females ap- peared in July (Table 1). Males peaked in July, declined in August and were gone by Septem- ber. That was similar to P. fuscula, whereas males of the other two species had gone by August. Breeding was slightly later than in the other species because only 4% of females car- ried cocoons in July and it was not until Au- gust and September that a significant percent- age of females had cocoons. New spiderlings (modal CW 0.6 mm) appeared in September (Fig. 2), later than in R. hyperborea and P.groenlandica but the same time as R. fus= cula. The apparent shrinkage of immatures be- tween June and July is an artifact of sampling. Pardosa hyperborea males and females ap- peared in July (Table 1), and males were seen only in July. Significant percentages of fe- males carried cocoons in July, August and September. New spiderlings (modal CW 0.6 mm) appeared in August (Fig. 3). No new spi- derlings were caught in September despite 80% of females carrying cocoons in that month. The group of immatures in September with modal CW of 1.0 mm is seen as the new spiderlings of August grown to that size. The similar group of small immatures in August PICKAVANCE— PAi?D05A LIFE-CYCLES 371 Carapace width (mm) Figure 1. — Comparison of measurements of car- apace width and tibia I length for the same sample of Pardosa moesta July 1997. (CW 0.9-1. 1 mm) is not seen as new spiders hatched the previous month (July) because neither adults nor cocoons were seen until July, and no new spiderlings were seen in that month.Therefore the two groups of immatures in August, lying between CW 0.9 and 1.6 mm, are seen as one group with a wide size range. That same group grew to occupy the range CW 1. 1-1.8 mm in September. Therefore the breaks in the data at CW 1.2 mm in August and CW 1.6-1. 7 mm in September are arti- facts of sampling. Pardosa fuscula males and females ap- peared in July (Table 1). Males had declined significantly by August and were gone by September. Breeding was in July and August with over 80% of females carrying cocoons in each of those months. New spiderlings (modal CW 0.6 mm) (Fig. 4) appeared in September. Pardosa groenlandica males and females appeared in June (Table 1), the earliest ap- pearance of adults in this study. Males were present in June and July but were gone by August. Females were still seen in September. Females carried cocoons in July and August, Carapace width (mm) Figure 2. — Frequency distribution of carapace width measurements of the monthly 1997 samples of Pardosa moesta. and a single female with a cocoon was caught in September. New spiderlings (modal CW 0.8 mm) appeared in August (Fig. 5), and had grown to modal CW 1.0 mm by September. The single subadult taken in July (CW 3.4 mm) was the latest observation of this stage for any of the four species. Small numbers of small immatures (CW 0.8 mm) were also seen 372 THE JOURNAL OF ARACHNOLOGY Figure 3. — Frequency distribution of carapace width measurements of the monthly 1997 samples of Pardosa hyperborea. in June and July. In July these might have been a small number of early-hatching new spiderlings because both sexes had been pre- sent the previous month, but the same expla- nation is not applicable to June because there was no evidence of cocoon production in that month. This is the only species of this study where subadults were seen in September. The 0.6 1.0 1.4 1.8 2.2 2.6 3.0 Carapace width (mm) Figure 4. — Frequency distribution of carapace width measurements of the monthly 1997 samples of Pardosa fuscula. PICKAVANCE— FAi^DO^A LIFE-CYCLES 373 l.i 1.4 1.8 22 2.i 3.0 3.4 3.1 4.2 Carapace width (mm) Figure 5. — Frequency distribution of carapace width measurements of the monthly 1997 samples of Pardosa groenlandica. smaller immatures seen in September (modal CW LO mm) were new spiders hatched the previous month (August) with one month’s growth, and the smallest of this group may be a few new spiders hatched in September. The larger immatures (modal CW 1.9 mm) are seen as persistent juveniles from the previous year that were not large enough to become subadults, possibly in combination with some slightly older individuals hatched the previous month. DISCUSSION Measuremeets»-—Since measurements of CW and tibia I yielded essentially the same information either could have been used here. However, CW was adopted because it was easier to manipulate the carapace into position for measurement. This is contrary to the opin- ion of Toft (1976) who argued that tibia I was easier to measure and used a linyphiid as an example. No doubt this discrepancy is due to the morphology of the taxa under consider- ation: what is true for lycosids may not be true for linyphiids. The use of CW to establish life- cycle stages has frequently been reported, for example by Almquist (1969), Workman (1978) and Putman (1967). Life-cycles. — The four life-cycles demon- strated here are essentially similar to biennial species elsewhere, for example P. lugubris in Scotland (Edgar 1971a), Trochosa ruricola in Finland (Hackman 1957), P. moesta and P. mackenziana (Keyserling 1877) in Alberta 374 THE JOURNAL OF ARACHNOLOGY ISatyre [ g [ jyjature Subadult HIHH ImmMure Recruitment I Winter #1 Winter #2 Mating | ? ? Recruitment ' \ ' r“ i "■ I I I n^T r ”n i ! i n i “n ~r AUG SIP OCT I^OV-APR MAY JUN JUL AUG SIP OCT NOV- APR MAY JUN JUL AUG SEP OCT YEAR! YEAR II YEAR III Figure 6. — Generalized life-cycle of Pardosa moesta, Pardosa hyperborea, Pardosa fuscula and Par- dosa groenlandica on the Island of Newfoundland, Canada. (Buddie 2000), Dolomedes fimbriatus in Fin- land (Palmgren 1939) and D, triton (Walcken- aer 1837) in Alberta (Zimmermann & Spence 1998), Philodromiis rufus and P. cespitum in Nova Scotia (Dondale 1961), Araniella dis- plicata in Nova Scotia (Dondale 1961) and Eris militaris in Nova Scotia (Dondale 1961). Others, such as the three British species of Amaurobius, are biennials but the second overwintering is as adults not as older im- matures (Cloudsley-Thompson 1955). Pardosa moesta is the only one of the pre- sent four investigated elsewhere, and the bi- ennial life-cycle of this species in Alberta (Buddie 2000) differs from the present work only in detail. In Alberta mating was mid-May to early June (July in Newfoundland) and co- coon-carrying early June through August (principally August and September in New- foundland). These differences can be attribut- ed to climate. In Alberta the population lived in the boreal transition ecoregion dominated by hardwoods with a mean annual precipita- tion of ca. 500 mm and the following mean temperatures: annual 1°C, summer 14°C and winter — 13.5°C. The Newfoundland popula- tion lived in the northern peninsula ecoregion dominated by conifers with a mean annual precipitation of ca. 1 1 50 mm and the follow- ing mean temperatures: annual 3°C, summer 11°C and winter — 4.5°C (Ecological Stratifi- cation Working Group 1995). Apart from these generalizations, at least two of the present four species had a longer period of recruitment than immediately appar- ent from the data. Within the sampling period (which ended mid September) all species showed one recruitment of new spiderlings: P. fuscula and P. moesta in September and F. hyperborea and P. groenlandica in August. However, a significant number of P. hyper- borea and P. moesta in September carried co- coons that would have hatched after sampling stopped. The single P. groenlandica female (with a cocoon) caught in September is not an adequate basis for further discussion.Whether such an extended period was due to females carrying single cocoons over a longer period or to some females carrying more than one cocoon, whether this might be expressed bi- modally, and the implications of this for both synchrony of the sexes and the size range of the life-stages will be discussed below. Production of two or more cocoons by one female over an extended period has been re- ported for several lycosids. Some reports were based on the direct evidence of marking tech- niques (e.g, Vlijm et ak 1963 for P. amentata (Clerck 1757), F. monticola (Clerck 1757), and P. nigriceps (Thorell 1856) in the Neth- erlands) . Others were based on indirect evi- dence such as length of the cocoon-carrying season (e.g. Eason (1969) for P. lapidicina Emerton 1885 in Arkansas; Vlijm & Kessler- Geschiere (1967) for P. pul lata (Clerck 1757), P. nigriceps and P. monticola in the Nether- PICKAVANCE— LIFE-CYCLES 375 lands; Edgar (1971a) for P. lugubris in Scot- land; Toth et at (1997) for P. agrestis (Westr- ing 1862) in Hungary). These species all had an early start to mating followed by a mini- mum four-month cocoon-carrying season, but there are suggestions of two cocoons in a shorter period. For example, both Wolff (1981) for P. moesta and Buddie (2000) for P. moesta and F. mackenziana surmised that a second cocoon was likely because they were carried over 3 mo. In the present study there is only the indirect evidence of duration and late start of the cocoon-carrying season to in- dicate how many cocoons females carried. None of the four species here had cocoons before July, whereas they were typically seen in May or at latest June in reports of two co- coons elsewhere. The cocoon-carrying periods of all four present species seem shorter than reported elsewhere, but there was an unob- served, extended cocoon-carrying season for at least two species after sampling stopped in September as discussed above. Overall, the late start and shortness of the cocoon-carrying season suggest that second cocoons were not usual. But whether from one cocoon or two, F. moesta and F. hyperborea had extended re- cruitment periods. These might result in bi- modal recruitment and they have implications for both synchrony of the sexes and the range of sizes of life-stages. One type of bimodaiity was reported by Samu et ah (1998) in F. agrestis, where a long reproductive period had synchronous peaks of males and females in May and August, each preceded by a peak of subadults, with new spiderlings present from early June to Octo- ber. The present study is clearly distinguished by the shortness of the cocoon-carrying peri- od, the late start to mating and the lack of a double peak of subadults. However, there may be an unobserved, bimodal peak in recruit- ment produced by late hatching cocoons as predicted by Edgar (197 lb) for F. lugubris in Scotland. An extended recruitment period could af- fect synchrony because some late-hatched spi- derlings might not mature in concert with the majority of immatures that have just overwin- tered for the second time. Therefore some in- dividuals could be triennial, taking an extra year to mature, as reported for Trochosa ter- ricola (Workman 1978). This might explain the presence in July and early appearance in September of subadult F. groenlandica, the presence of small immature F. groenlandica in June and July, and the large immature F. hyperborea embedded among the adults in September. On the other hand, Edgar (1971b) showed that late hatched spiderlings rapidly catch up in size with their counterparts from an early hatch, thereby reducing or obliterat- ing the anticipated bimodal age distribution of these two cohorts. The wide variation in size-range of life-stag- es seen in the present study has been reported for other lycosids, for example by Eason & Whitcomb (1965), Almquist (1969) and Eason (1969). An extended period of recruitment would increase the number of instars occurring together, which would increase the size-range of life-stages, particularly the immatures that contain several instars. There may also be dif- ferences between early and late-hatched spi- derlings. Edgar (1971b) reported that later spi- derlings tended to be heavier than earlier (but whether heavier equals a larger CW is uncer- tain). On the other hand, Buddie (2000) said that later spiderlings were substantially smaller than earlier. Against this must be balanced the report that later spiderlings tend to catch up with earlier ones (Edgar 1971b). Either way, size difference of new spiderlings will to an extent increase the spread of the size-range of subsequent life-stages, particularly of the im- matures. Overall, such variations in size do not obscure the general conclusions of the present work, but may mask subtle attributes of the populations. The present work has increased the knowl- edge of spider life-cycles in northern locali- ties, compared these four species to other bi- ennials, and suggested that some individuals may extend their life-cycle beyond 2 yr. The question of v/hether and to what extent spiders can extend their life-cycles to accommodate increasingly difficult environmental condi- tions awaits further studies at more northern latitudes or perhaps higher elevations. ACKNOWLEDGEMENTS I am very grateful to Charles Dondale and James Redner for their encouragement and ex- pertise, to Parks Canada for permission to sample in Gros Morne National Park, to Carol Harding for her assistance in the field and to Chris Buddie for comments on the manu- script. I would like to thank Ian King for his 376 THE JOURNAL OF ARACHNOLOGY generosity with his time, and the Biology De- partment, Memorial University of Newfound- land, for materials and facilities. LITERATURE CITED Almquist, S. 1969. Seasonal growth of some dune- living spiders. Oikos 20: 392-408. Berghe, Van den E. 1992. On pitfall trapping in- vertebrates. Entomological News 103:149-156. Buddie, C.M. 2000. Life history of Pardosa moesta and Pardosa mackenziana (Araneae, Lycosidae) in central Alberta, Canada. Journal of Arachnol- ogy 28: 319-328. Cloudsley-Thompson, J.L. 1955. The life-histories of the British cribellate spiders of the genus Cin- iflo Bl. (Dictynidae). Annals and Magazine of Natural History 8:787-794. Dondale, C.D. 1961. Life histories of some com- mon spiders from trees and shrubs in Nova Sco- tia. Canadian Journal of Zoology 39:777-787. Dondale, C.D. 1977. Life histories and distribution patterns of hunting spiders (Araneae) in an On- tario meadow. Journal of Arachnology 4:73-93. Dondale, C.D. 1999. Revision of the groetdandica subgroup of the genus Pardosa (Araneae, Ly- cosidae). Journal of Arachnology 27:435-448. Dondale, C.D. & J.H. Redner. 1990. The Wolf Spi- ders, Nurseryweb Spiders, and Lynx Spiders of Canada and Alaska. Araneae: Lycosidae, Pisaur- idae, and Oxyopidae. The Insects and Arachnids of Canada Part 17. Agriculture Canada Publica- tions 1856. 383 pp. Eason, R.R. 1969. Life history and behavior of Par- dosa lapidicina Emerton (Araneae: Lycosidae). Journal of the Kansas Entomological Society 42: 339-360. Eason, R. & W.H. Whitcomb. 1965. Life history of the dotted wolf spider, Lycosa punctulata Hentz (Araneida: Lycosidae). Arkansas Academy of Science Proceedings 19: 11-20. Edgar, W.D. 1971a. The life-cycle, abundance and seasonal movement of the wolf spider, Lycosa {Pardosa) lugubris, in central Scotland. Journal of Animal Ecology 40: 303-322. Edgar, W.D. 1971b. Seasonal weight changes, age structure, natality and mortality in the wolf spi- der Pardosa lugubris Walck. in central Scotland. Oikos 22: 84-92. Edgar, W.D. 1972. The life-cycle of the wolf spider Pardosa lugubris in Holland. Journal of Zoology London 168: 1-7. Ecological Stratification Working Group. 1995. A National Ecological Framework for Canada. Ag- riculture and Agri-Food Canada, Research Branch, Centre for Land and Biological Resourc- es Research and Environment Canada, State of the Environment Directorate, Ecozone Analysis Branch, Ottawa/Hull. 125 pp. Emerton, J.H. 1902. The Common Spiders of the United States. Ginn and Company, Boston. 235 PP- Hackman, W. 1954. The Spiders of Newfoundland. Acta Zoologica Fennica 79:1-99. Hackman, W. 1957. Studies on the ecology of the wolf spider Trochosa ruricola Deg. Commenta- tiones Biologicae 16 (6): 1-34. Hagstrum, D.W. 1971. Carapace width as a tool for evaluating the rate of development of spiders in the laboratory and the field. Annals of the En- tomological Society of America 64:757-760. Leech, R.E. 1966. The spiders (Araneida) of Hazen Camp 81°49'N, 71°18'W. Quaestiones Entomo- logicae 2:153-212. Palmgren, P. 1939. Okologische und physiologische Untersuchungen uber die Spinne Dolomedes fim- briatus (CL). Acta Zoologica Fennica 24:1-42. Putman, W.L. 1967. Life histories and habits of two species of Philodromus (Araneida: Thomisidae) in Ontario. Canadian Entomologist 99:622-631. Ricards, D.R. 1967. An investigation of the ecology of the wolf spider Pardosa groenlandica Th. (Araneae, Lycosidae). Unpublished M.A.Thesis, University of Montana, Bozeman. 138pp. Samu, E, J. Nemeth, F. Toth, E. Szita, B. Kiss, & C. Szinetar. 1998. Are two cohorts responsible for the bimodal life-history pattern in the wolf spider Pardosa agrestis in Hungary? In Seldon, P. A. (ed.). Proceedings of the 17‘^ European Col- loquium of Arachnology, Edinburgh 1997. Pp. 215-221. Schmoller, R. 1970. Life histories of alpine tun- dra arachnida in Colorado. American Midland Naturalist 83:1 19-133. South, G.R. (Ed). 1983. Biogeography and Ecology of the Island of Newfoundland. W. Junk, Hague. 723 pp. Stratton, G.E. & D.C. Lowrie. 1984. Courtship be- havior and life cycle of the wolf spider Schizo- cosa mccooki (Araneae, Lycosidae). Journal of Arachnology 12:223-228. Toft, S. 1976. Life-histories of spiders in a Danish beech wood. Natura Jutlandica 19:5-40. Toft, S. 1979. Life histories of eight Danish wetland spiders. Entomologiske Meddelelser 47:22-32. Topping, C.J. & K.D. Sunderland. 1992. Limita- tions to the use of pitfall traps in ecological stud- ies exemplified by a study of spiders in a field of winter wheat. Journal of Applied Ecology 29: 485-491. Toth, E, J. Kiss, & B. Balint. 1997. Reproduction characteristics of Pardosa agrestis (Westring) (Araneae, Lycosidae) based on pitfall trappiong in winter wheat. Acta Phytopathologica et En- tomologica Hungarica 32: 313-317. Vlijm, L. & A.M. Kessler-Geschiere. 1967. The phenology and habitat of Pardosa monticola, P. nigriceps and P. pullata (Araneae, Lycosidae). Journal of Animal Ecology 36: 31-56. PICKAVANCE— FA^DOSA LIFE-CYCLES 377 Vlijm, L., A, Kessler & CJJ. Richter. 1963. The life history of Pardosa amentata (CL) (Araneae, Lycosidae). Entomologische Berichten Amster- dam 23: 75-80. Wolff, RJ. 1981. Wolf spiders of the genus Par- dosa (Araneae: Lycosidae) in Michigan, Great Lakes Entomologist 14: 63-68. Workman, C. 1978. Life cycle and population dy- namics of Trochosa terricola Thorell (Araneae: Lycosidae) in a Norfolk grass heath. Ecological Entomology 3:329-340. Zimmermann, M, & J. R. Spence. 1998. Phenology and life-cycle regulation of the fishing spider Do- lomedes triton Walckenaer (Araneae, Pisauridae) in central Alberta, Canadian Journal of Zoology 76:295-309. Manuscript received 1 November 2000, revised 4 June 2001. 2001. The Journal of Arachnology 29:378-387 SYNONYMS OF FRONTINELLA TIBIALIS (ARANEAE, LINYPHIIDAE) G. Ibarra-Nunez'^ J. A. Garcia*, M. L, Jimenez^ and A. Mazariegos*: 'El Colegio de la Frontera Sur. Carr. Antiguo Aeropuerto km 2.5, Apdo. Postal 36, Tapachula, Chiapas 30700, Mexico; ^Centro de Investigaciones Biologicas del Noroeste S.C. Apartado Postal 128, La Paz, Baja California Sur, 2300, Mexico, email: gibarra@ tap-ecosur.edu. mx ABSTRACT. Synonymy among three species of linyphid spiders of the genus FrontineUa F. O. Pickard- Cambridge 1902 is established based on field association between males and females, mating records, and morphological data from recently collected specimens. It is concluded that F. caudata Gertsch & Davis 1946 and F. lepidula Gertsch & Davis 1946 are both junior synonyms of F. tibialis F. O. Pickard- Cambridge 1902. A redescription of this species is included. RESUMEN. Se establece la sinonimia entre tres especies de aranas linffidas del genero FrontineUa F O. Pickard-Cambridge 1902, con base en datos de campo sobre la asociacion entre machos y hembras, registros de apareamientos, y datos de la morfologia de especimenes recientemente colectados. Se concluye qiie F. caudata Gertsch & Davis 1946 y F, lepidula Gertsch & Davis 1946 son sinonimos junior de F. tibialis F. O. Pickard-Cambridge 1902. Se incluye una redescripcion de esta especie. Keywords: Linyphiidae, FrontineUa, synonymy, Mexico, Chiapas. The spider genus FrontineUa was created by E O. Pickard-Cambridge in 1902 for eight spe- cies found in Central and North America. Pe- trunkevitch (1911) made FrontineUa a junior synonym of Linyphia Latreille 1 804 without ar- gument, but Blauvelt (1936) resurrected Frori- tinella from this synonymy when she made a revision of Linyphia and several related genera. Roewer (1942, 1954) and Bonnet (1956, 1957), in their respective catalogues, included Fronti- nella as a junior synonym of Linyphia, but they did not give any argument. Nevertheless, from that time on most authors considered Frontinel- la a valid genus name (Kaston 1938, 1948; Gertsch & Jellison 1939; Muma 1943; Brignoli 1983; Millidge 1984; Platnick 1989, 1993, 1997; Breene et al, 1993), and several new species were described (Gertsch & Davis 1946; Bryant 1948; Kraus 1955; Li & Song 1993). Millidge (1991) transferred F. uncata F. O. Pickard-Cam- bridge 1902 to the genus Novafrontina Millidge 1991. At present there are 15 species world- wide, two from China and 13 from the Ameri- cas, with nine of these species reported from Mexico: F. communis (Hentz 1850), F. laeta (O. Pickard-Cambridge 1898), F bicuspis F. O. Pickard-Cambridge 1902, F rustica F O, Pick- ard-Cambridge 1902, F. tibialis F O. Pickard- Cambridge 1902, F caudata Gertsch & Davis 1946, F huachuca benevola Gertsch & Davis 1946, F. lepidula Gertsch & Davis 1946 and F potosia Gertsch & Davis 1946. Only two of these have been recorded for the state of Chia- pas, Mexico: F. caudata and F. lepidula (Gertsch & Davis 1946; Hoffmann 1976). In 1995, we collected female and male specimens of FrontineUa in coffee plantations in southeast Chiapas. The female specimens were identified as F. caudata, some males as F. tibialis, and other males as F. lepidula. FrontineUa tibialis and F. lepidula were de- scribed from only one male specimen each, F. caudata was described from only female spec- imens. We found accompanying males in the webs of several F. caudata females. Most of these were identified as F. tibialis, one as F. lepidula. The finding of these pairs suggests synonymy. In their description of F. lepidula, Gertsch & Davis (1946) considered it near to F. tibi- alis but pointed out some differences: “This is a smaller species than tibialis Cambridge. The embolus of the male palpus is shorter and less strongly curved at the apex, and the pa- 378 IBARRA-NUNEZ ET AL.— SYNONYMS OE FRONTINELLA TIBIALIS 379 tella of the palpus is armed with a long curved spine instead of a short spun” Some of our male specimens could not be assigned to any of these two species because they have a “short spur” in one patella and a “long spine” in the other, showing another indica- tion of possible synonymy. Thus, we decided to collect more specimens of both sexes, to study and clarify the taxonomy of the three described species. METHODS The collecting site was the coffee plantation of the “Campo Agricola Experimental Rosa- rio Izapa” of the Institute Nacional de Inves- tigaciones Forestales, Agricolas y Pecuarias (INIFAP), Municipio of Tuxtla Chico, 20 km NNE of Tapachula, Chiapas, at 400 m eleva- tion. This place was selected due to the abun- dance of Frontinella. The hrst collecting pe- riod was from September-November 1995, and the second in January 1998. The spiders were collected by visual search of the webs of adult and subadult females (as it is difficult to differentiate these two age classes in the field); additionally, some solitary males were collected to have enough specimens for the taxonomic study. All individuals found on the same web were put in the same container and preserved in 70% ethanol. In the laboratory each specimen was tentatively identified and the age class and sex was determined. Speci- mens were deposited at the Coleccion de Ar- anas del Sureste de Mexico (ECOTA-AR, El Colegio de la Frontera Sur, Tapachula, Chia- pas, Mexico), the American Museum of Nat- ural History, New York (AMNH), and The Natural History Museum, London (NHM). To study the specimens, we considered the characters in the original description of each species. All adults (females and males) were measured to compare their variability in re- lation to the measurements of the correspond- ing type as noted in the original descriptions. The measured characters were total length, carapace length, carapace width, and length of each segment of the first leg from femur to tarsus. Specimens of both sexes were sent to Dr. N. 1. Platnick (AMNH), and to Mr. P. D. Hilly ard (NHM), for comparison with the cor- responding types in those museums. The par- atypes of F. caudata and of F. huachuca be- nevola, deposited in the Coleccion Nacional de Aracnidos (CNAN, Dr.Tila Maria Perez, Institute de Biologia, Universidad Nacional Autonoma de Mexico), were also examined for comparative analyses. We examined the patellar macroseta of both palpi for each col- lected male, and made some SEM photo- graphs of them. Additional information about the type specimens, not included in the orig- inal descriptions, was provided by Dr. N. L Platnick and by Mr. P D. Hillyard. The de- scriptions in the taxonomic section were based on the specimens collected in this work, with each measurement noted as average and range of variation (minimum and maximum); all measurements are in millimeters. Abbrevia- tions used in text: ALE: anterior lateral eyes; AME: anterior median eyes; PLE: posterior lateral eyes; PME: posterior median eyes. Fe- I (II, III, IV): first femora (second, third, fourth); Me-I (II, III, IV): first metatarsi (sec- ond, third, fourth); Pa-I (II, III, IV): first pa- tellae (second, third, fourth); Ta-I (II, III, IV): first tarsi (second, third, fourth); Ti-I (II, III, IV): first tibiae (second, third, fourth). Cy: cymbium; PCy: paracymbium. RESULTS We collected 76 individuals from 55 webs: 39 webs were occupied by only one individual (solitaries), the remaining 16 webs contained 37 spiders, from two to five individuals per web (Table 1). Most accompanied adult fe- males shared the web with only one male, but a few were found with several males. Most sub-adult females were alone on their webs, only one was found with a male. Inexplicably, one sub-adult female and one juvenile were found each on the web of an adult female. Most males were found on the webs of adult females, but two males were each on the web of a juvenile (Table 1). In one web, one adult female was found with four males (Table 2), and we observed the end of copulation between one of these males and the female. On a date subsequent to the sampling period, one of the authors (J. A. Garcia) observed another pair in a web, copulating at least two times during the 30 min of observation. The measurements from 30 collected fe- males showed that most mean values of these females are slightly smaller than those of the F. caudata holotype and paratype, but the var- iation of these types are well inside the vari- ation ranges of the collected females. Besides 380 THE JOURNAL OF ARACHNOLOGY Table 1. — Numbers of specimens collected on each web, with their corresponding sex/age. F = fe- male(s), M = male(s). Adult F Sub-adult F Adult M Juveniles No. in web totals Solitaries 18 14 7 39 with F 1 16 1 18 with 1 M 8 1 2 11 with >1 M 3 3 with sub-adult F 1 1 2 with juveniles 1 2 3 Age/sex totals 31 16 26 3 76 the size variations, the study of the adult fe- males showed no major differences among them, nor in respect to the characters noted in the description of F. caudata, or those ob- served in the paratype of this species. The epi- gyna were very similar for all females, para- type included, in respect to form and position of the openings, dorsal and ventral plates (as defined by Millidge 1984). Although some variability was recorded in the form and size of the opisthosoma and its caudal tubercle, this could be due to differences in nutritional condition, level of development of ovaries, or natural variability. Gertsch & Davis (1946) considered the length of the patellar macroseta, and length and curvature of the embolus as distinctive characteristics between F. tibialis and F. lep- idula. The collected male specimens showed only slight variability among them in embolus characteristics. On the contrary, for the length of the patellar macroseta we found two states, the “short spur” of Gertsch & Davis (as in the figs. 7a and 7b of F. tibialis by F. O. Pick- ard-Cambridge 1902, table XL), or the “long spine” (as in fig. 11 of F lepidula by Gertsch & Davis 1946). We found males with “long spines” on both palpi, males with “short spurs” on both palpi, but also males with a “short spur” on one palp and a “long spine” on the other (Ta- ble 2). This mixed condition seems to indicate that the macroseta is originally long, but it can break and lose its slender distal part, with only the thick base remaining. SEM images showed that the “short spur” is a patellar ma- croseta broken in the area where its diameter is reduced (Figs. 1, 2). Our data also show that both solitary and courting males (in the web of a female) show the three conditions. Likewise, one female specimen was accompanied with males show- ing these three conditions (Table 2). Also, the male observed copulating with a female on a date subsequent to the collect period had a broken macroseta in one palp and a complete macroseta in the other. As for the females, the measurements of the collected male specimens showed also over- lapping ranges of size among them, and with the holotype of F. lepidula. It was also found that the collected males have a mastidion on each chelicera (one small laterally directed tu- bercle at the anterior proximal surface of the Table 2. — Numbers of specimens collected alone or accompanied, and the pedipalp-patella macrosetae condition of the corresponding males. F = female, M == male(s), BM = broken macroseta, CM = complete macroseta. * The values with the same subscript letter corresponds to the same female specimen. M with 2 CM M with 2 BM M with 1 CM and 1 BM Solitary M 2 2 3 1 F with 1 M 2 4 2 1 F with > 1 M* 1 subadult F with 1 M 1 juvenile F with 1 M 2, + Ic 1 2a + lb + k 2 lb Totals 8 12 6 IBARRA-NUNEZ ET AL.— SYNONYMS OF FRONTINELLA TIBIALIS 381 cheliceral base, just below the clypeus)= The mastidioe size is variable, from sharply point- ed in some specimens to blunt and very re- duced in others. Some specimens have mas- tidia of different size, one pointed and one reduced. In some of the males, the chelicerae were so retracted that the mastidion was not directly visible because it was covered by the clypeus, but a careful examination showed its presence in all collected males. DISCUSSION In some linyphiid species, the males re- spond to a species-specific sex pheromone present in the silk of adult female webs by approaching the female and initiating court- ship behavior; in some cases the males can stay in the web for some time, and even cop- ulate several times with the female (Rovner 1968; Austad 1982; Suter & Reekes 1982; Watson 1986, 1995; Wiley-Robertson & Adler 1994). The presence of the males in the fe- males’ webs, and the copulations observed constitute sound evidence of conspecificity, especially since no other related species were found at the collecting site. The morphological similarity between the collected females and the paratype of F. cau- data (especially with regard to the epigynum), and the fact that the measurements of the ho- lotype and paratype of this species are inside the ranges of the collected females, indicate that F. caudata and the collected specimens are the same species. Platnick (pers. comm.) considered a female specimen sent to him as the same species as the F. caudata holotype, and the male speci- mens sent to him as the same species of the F. lepidula holotype. For the male specimens sent to be compared with F. tibialis, Hillyard (pers. comm.) considered that “the pedipalp macroseta is not significantly different”, but that “there is som,e doubt that these two spe- cies are coespecific”, because he noted some differences such as total length (“the type is slightly larger”), the size of the palp (“more robust in the type”), the shape of the embolus (broader and more curved at its tip), and the presence of mastidia (“the type does not have a single mastidion on the basal segment of each chelicera”). In the original description of F. tibialis, F. O. Pickard-Cambridge (1902) noted only the total length, which is not a reliable character because of the variability of the opisthosomal size in spiders (Blauvelt 1936; Hormiga 1994a). It was not possible to obtain other measurements from this type. Nevertheless, the maximum value of total length found from our specimens is only slightly below of that noted for the type of F. tibialis (4.4 vs. 4.5). Additionally, there is a high variability in size in the collected male specimens, where the smallest male is only one half of the largest (2.2 to 4.4). Furthermore, the range of varia- tion for each measured character overlapped among the three variants of the collected males (with 2 complete macroseta, with 2 bro- ken macroseta, and with one complete and one broken macroseta), and the measurements of the F. lepidula holotype were also inside these variation ranges. As F. tibialis was de- scribed from only one specimen, there were no records about its size variability. Thus, the slight difference in size does not contradict the conspecifity of our specim.ens with the type of F. tibialis. Concerning the differences between F. tib- ialis and F, lepidula in size and curvature of the embolus mentioned by Gertsch & Davis (1946), we consider these as minor differences in comparison with other Frontinella species (judging from the drawings of F. O. Pickard- Cambridge 1902; Blauvelt 1936; Gertsch & Davis 1946; and Song et al. 1999), where the pedipalpai bulbs (subtegulum, tegulum, em- bolic division, particularly the embolus and la- mella characteristica) are conspicuously dif- ferent among species. When the palpal bulb of the collected specimens is observed from different angles the embolus becomes more or less curved, and the lamella characteristica be- comes more or less wide. Additionally, the type of F. tibialis comes from a locality near the Gulf of Mexico coast, but the type of F. lepidula and the males we collected come from a locality near the Pacific coast of Mex- ico; therefore, it is possible that these are op- posite ends of a geographic variability spec- trum concerning the size and curvature of the embolus. F. O. Pickard-Cambridge (1902) noted ex- plicitly in his key to Frontinella species that F. tibialis lacks mastidia, but all our male specimens have one on each chelicera, al- though sometimes a very reduced one. The presence of mastidia was not mentioned by Gertsch & Davis (1946), but Platnick (pers. 382 THE JOURNAL OF ARACHNOLOGY comm.) confirmed its presence in the F. lep- idula holotype. Most species of Frontinella with known males have a mastidion on their chelicerae: F. laeta and F. bicuspis (E O. Pickard-Cambridge 1902), F. communis (Blauvelt 1936), F. huachuca and F. huachu- ca benevolo (Platnick pers. comm, and per- sonal observation of the male paratype of F. huachuca benevolo in the CNAN). Thus, the presence of mastidia seems not to be rare in this genus. The collected male specimens showed a high variability in the size of their mastidia. As F. tibialis was described from only one specimen, there were no records about mastidion variability. Also, it is possible that this difference corresponds to a spectrum of variability between the two populations (Gulf coast and Pacific coast). Gertsch & Davis (1946) said about Fron- tinella, “It is notable that the males of the three new species herein described all have the patella of the palpus set with a long dorsal spine. In all the other known species this spine is modified into a short spur.” As both spe- cies, F. tibialis and F. lepidula, were de- scribed each with only one specimen, the var- iation existing in this character was not observed. We do not know when and how the patellar macroseta breaks. As they are present only in the males, it is possible that these structures are related to reproductive activi- ties. It would be necessary to study the repro- ductive behavior of this species to know more about the function of this structure. Other similarities that support the hypoth- esis of conspecifity between the type of F. tib- ialis and the collected male specimens are the presence of macrosetae in the mesal border of the cymbium (Fig. 7), the relative size be- tween pedipalp’s patella and the tibia (tibia about twice as long as patella), the form of the pedipalp’s tibia (widening to its distal end. Figs. 6-8), and the form of the male sternum, narrowly produced between coxae IV (as in table XL, fig. 7 of F. O. Pickard-Cambridge 1902). From this evidence, we conclude that F. caudata and F. lepidula are junior synonyms of F. tibialis by the principle of priority (Ar- ticle 23 of the ICZN). As the original descrip- tion of F. tibialis is very short, we include here a redescription of this species based on the specimens collected in this work. TAXONOMY Figs. l-IO Frontinella tibialis F. O. Pickard-Cambridge 1902 Frontinella tibialis F. O. Pickard-Cambridge, 1902: 422, plate XL figs. 7a-b 6 ; Gertsch & Davis, 1946: 3. Linyphia tibialis, Petrunkevitch 1911: 255; Roewer 1942: 591; Bonnet 1957: 2531. Frontinella caudata Gertsch & Davis, 1946: 4, fig. 6 ?; Brignoli 1983: 294. NEW SYNONYMY Frontinella lepidula Gertsch & Davis, 1946: 4-5, figs. 10-11 3; Brignoli 1983: 294. NEW SYN- ONYMY. Types, — Male holotype of F. tibialis from Teapa, Tabasco, Mexico, in the collection of Goodman & Salvin, deposited in NHM (not examined). Female holotype of F. caudata from Chilpancingo, Guerrero, Mexico, depos- ited in the AMNH, female paratypes from Chilpancingo, Guerrero, Mapastepec and Ta- pachula, Chiapas, Mexico, deposited in the AMNH and in the CNAN (examined). Male holotype of F. lepidula from Tonala, Chiapas, Mexico, deposited in the AMNH (not exam- ined). Diagnosis. — This is a tentative diagnosis, because several species of this genus are known only from one sex, and we did not re- vise this genus. A possible autapomorphy of this species is the distinctive form of the la- mella characteristica in the male pedipalp (Figs. 4, 7), clearly different from that of all other known males. The tibia’s relative length, twice as long as the patella in the male pedi- palp (Fig. 3), separates this species from most others species with known males (with tibiae less than twice the patella length), except from F. potosia, but in F. potosia the patellar ma- croseta of the pedipalp does not have a wid- ened base as in F. tibialis. The position of the copulatory openings (on the lateral borders, about at the middle of the distance between the anterior and posterior borders of the dorsal plate. Fig. 9) distinguish this species from most other species with known females (hav- ing the copulatory openings on the anterior border of the dorsal plate), except from F. zhui Li & Song 1993, but in this species the pos- terior border of the dorsal plate is notoriously rounded and extended backwards. Description. — Male: {n = 26) Total length 3.25 (2.16-4.43), carapace length 1.49 (0.97- 1.86), carapace width 0.98 (0.70-1.20). Di- IBARRA-NUNEZ ET AL.~SYNONYMS OF FRONTINELLA TIBIALIS 383 Figures 1-6. — Frontinelia tibialis, scanning electron micrographs of male pedipalp structures: 1. Com- plete macroseta of patella; 2. Broken macroseta of patella showing fracture point; 3. Row of setigerous cusps on mesal face of right femur, v/ith enlargement of one cusp (inset) (scale bar for femur 50.0 pm); 4. Ectoventral viev/ of left pedipalp showing visible parts in unexpanded condition; 5. Ectodorsal view of right pedipalp in unexpanded condition showing paracymbium with setae (white aiTow) and membrane of cymbium (black arrow) (scale bar 46.5 pm); 6. Dorsal view of right tibia showing three retrolateral (white arrows) and two prolateral (black arrows) trichobothria (scale bar = 29.4 p.m). Abbreviations: Cy = cymbium; E = embolus; EM = embolic membrane; Fe = femur; LC = lamella characteristica; Pa = patella; Pcy = paracymbium; Sp = spine of lamella characteristica; SPT = suprategulum; ST = subte- gulum; T = tegulum; Ti = tibia. 384 THE JOURNAL OF ARACHNOLOGY Figures 7-10. — Frontinella tibialis, drawings of genitalia. 7-8. Left male pedipalp with expanded bulb. 7. Meso-ventral view; 8. Ecto-dorsal view; 9-10. Female epigynum. 9. Ventral view; 10. Dorsal view of cleared epigynum. Scale bars = 0. 1 mm. Abbreviations: Pedipalp: BH = basal hematodocha; Cl = column; Cy = cymbium; CyM = cymbium macrosetae; E = embolus; ED = ejaculatory duct; EM = embolic membrane; Fu: = fundus; EC = lamella characteristica; LP = lateral process of lamella characteristica; M = membrane; Pcy = paracymbium; Pe = petiole; SPT = suprategulum; ST = subtegulum; T = tegulum; Ti = tibia. Epigynum: A = atrium; CD = copulatory duct; CO = copulatory opening; DP = dorsal plate; FD = fertilization duct; S = spermatheca; VP = ventral plate. ameter of AME 0.08 (0.07-0.09), ALE 0.09 (0.07-0.11), PME 0.09 (0.08-0.09), PEE 0.08 (0.07-0.11). Separation between AME 0.06 (0.05-0.07), PME 0.08 (0.07-0.11), AME and ALE 0.11 (0.08-0.13), PME and PLE 0.13 (0.09-0.15). Clypeus height 0.25 (0.19-0.32). Length of Fe-I 2.38 (1.60-3.16), Pa-I 0.39 (0.30-0.50), Ti-I 2.03 (1.40-2.63), Me-I 2.33 (1.60-3.26), Ta-I 1.23 (0.90-1.53), Pa-II 0.35 (0.28-0.43), Ti-II 1.56 (1.08-2.03), Pa-III 0.28 (0.20-0.37), Ti-III 0.81 (0.54-1.07), Pa- IV 0.35 (0.24-0.47), Ti-IV 1.42 (0.94-2.00). Carapace with a distinct but shallow tho- racic groove. Eyes on low tubercles, anterior IBARRA-NUNEZ ET AL.— SYNONYMS OF FRONTINELLA TIBIALIS 385 eye row moderately recurved, posterior eye row slightly recurved, lateral eyes contiguous. Labium about 1.5 times wider than long, in close contact with sternum. Sternum scuti- form, produced behind between coxae IV. Cheliceral base with a narrow but distinct stridulatory band on the ectal side; with a mastidion on the anterior face near to its base, varying from a sharply pointed tubercle to a reduced bluet one; anterior face of chelicerae with scattered setigerous cusps. Chelicerae with 4“”5 promarginal teeth and 4-5 small re- tromarginal teeth. Endites with a diagonal ca- rina on the outer half of the distal border. Legs LII~IV-IIL Coxae IV separated by about one half of their width; Fe-I to IV with a few lon- gitudinal ventral and lateral series of setiger- ous cusps, more conspicuous on Fe-I and 11. Opisthosoma elongated-oval from above, about twice as long as wide; more or less rect- angular in lateral view, with a scarcely devel- oped rounded caudal tubercle. Pedipalp: Femur with small setigerous cusps on its dorsal, and ectal faces; mesal face smooth with only 5-6 setigerous cusps in a longitudinal line (Fig. 3). Patella with a dorso- distal protuberance that supports a macroseta. Tibia about twice as long as patella (Fig. 3), and widening to its distal end (Figs. 6-8). Cy elongated, about twice as long as maximum wide, narrowing toward its tip, with a trans- lucent, convex membrane on the proximal half of its ectal border; alveolus occupying almost all proximal face, leaving unoccupied about distal one teeth (Figs. 4-5, 7-8). PCy a small curved strip (intersegmental seosu Hormiga 1994b), similar in texture and coloration to Cy, about one sixth the Cy length, and touch- ing the Cy membrane (Figs. 5,8). Subtegulum transverse; tegulum trapezoidal, narrow on its distal border; suprategulum visible distal to te- gulum, between the embolic division and the alveolus (Figs. 4, 5). Embolic division con- nected to tegulum by a membranous column. Lamella characteristica elongated, proximally directed in the not expanded bulb, pointed on its proximal tip and reaching the base of Cy (Figs. 4); with one short lateral process on its mesal side (Figs. 7, 8), that reaches the middle of the mesal border of Cy, and with an incon- spicuous spine (visible only at high magnifi- cation, Fig. 4) on the opposite side (the spur of the lamella in Blauvelt’s 1936 description of F. communis). Embolus pointed and curved on its apex, more or less parallel to Cy, em- bolic membrane and membrane both parallel to embolus, ending as membranous strips that touch the embolus tip (Figs. 4, 5, 7 & 8). Carapace almost glabrous, sternum and en- dites with sparse setae. Chelicerae with sparse short setae on anterior face, apex of outer sides and a few setae bordering both cheliceral margins. Legs with scattered setae, more nu- merous and stiff on Me and Ta; stiff setae on setigerous cusps of Fe. Bristles on legs as fol- lows: 2 dorsal on Pa and Ti-I to IV; 1 ventral on Ti-I and II; 1 prolateral on Ti-I; I retrola- teral on Ti-I and II; 1 dorsal and 1 ventral on Me-III; 1 dorsal on Me-IV Opisthosoma with sparse, short setae and with two groups of long bristles on its anterior end, above each side of the pedicel. Pedipalp: with sparse setae form femur to Cy; patella with a long distal dorsal macroseta on dorso-distal protuberance, the macroseta proximal % is thick and curved to the outer side (forming an angle of about 90- from the femur axis), and the rest thin and more or less straight, tapering to a point, (Figs. 1, 2). Tibia with longer setae forming an incomplete ring near its distal border, with 2-3 retrolateral tri- chobothria and 1-2 prolateral trichobothria (Fig. 6). Cy with 3-4 short thick macrosetae on the distal half of its mesal border (Fig. 7). PCy with a few small setae on its distal half, visible only at high magnification (Fig. 5). Coloration: Variations observed on both fresh and older preserved specimens. Cara- pace, chelicerae and endites orange-brown, pars thoracica with faint dusky radiating lines, eye tubercles black. Endites becoming white- yellow towards their tip, with a black carina on the outer half of the distal border. Sternum and labium dusky orange-brown, with borders in front of endites and rear point infuscated. Pedipalpi dusky light-green to dark orange- brown. Legs with coxae to basal two thirds of femora light orange-brown, the rest dusky light-green, darker from tibiae to tarsi. Opis- thosoma creamy-gray with dorsal orange- brown tinge, and sides with a dark band and patches. Venter, caudal tubercle and spinnerets darker. Female: {n — 30) Total length 5.69 (4.46- 6.95), carapace length 1.96 (1.47-2.25), car- apace width 1.29 (0.87-1.67). Diameter of AME 0.09 (0.08-0.11), ALE 0.12 (0.11- 0.13), PME 0.11 (0.09-0.12), PLE 0.10 386 THE JOURNAL OF ARACHNOLOGY (0.09-0.11). Separation between AME 0.06 (0.05-0.08), PME 0.09 (0.08-0.1 1), AME and ALE 0.16 (0.13-0.17), PME and PEE 0.16 (0.11-0.17). Clypeus height 0.25 (0.20-0.29). Length of Fe-I 3.14 (2.56-3.56), Pa-I 0.63 (0.50-0.70), Ti-I 2.91 (2.60-3.30), Me-I 3.11 (2.40-3.55), Ta-I 1.53 (1.13-1.80), Pa-II 0.60 (0.53-0.63), Ti-II 2.25 (1.97-2.43), Pa-III 0.49 (0.43-0.50), Ti-III 1.26 (1.07-1.37), Pa- IV 0.55 (0.50-0.57), TLIV 2.20 (1.90-2.40). Female similar to male except in the follow- ing characters: labium not in close contact with sternum. Cheliceral base clearly thickened proximally, without mastidia nor setigerous cusps. Tarsi of pedipalpi with one simple claw. Legs I-IV-II-IIL Fe without series of setigerous cusps. Opisthosoma more or less trapezoidal in lateral view, with the rear side higher than the anterior side, and with a pronounced caudal tu- bercle projected beyond the spinnerets. Opis- thosoma with sparse short setae on the ventral plate of epigynum. Carapace and chelicerae brown to dark brown. Sternum, labium and en- dites dusky brown. Pedipalpi dusky light- green, darker to the tarsi. Legs with distal half of Fe light orange-brown. All Fe with a trans- verse dark gray band on the distal ventral bor- der. Opisthosoma dark brown to black, some specimens with a pair of small creamy white points on the middle of dorsum. Dorsum mar- gined with an irregular creamy white band in- cluding the caudal tubercle, incomplete in some specimens. Sides with another in'egular creamy white band at mid-height, and with four transversal (dorso-ventral) irregular dis- continuous creamy white bands on the poste- rior half. With a diffuse patch of creamy white just above the anal tubercle. Epigynum (Figs. 9, 10) wider than long. Ventral plate slightly convex, protruding very little from the abdominal wall. Dorsal plate about as wide as long, concave in its anterior half forming an epigynal atrium where are found the exposed rounded copulatory open- ings at each side, touching the border with the ventral plate (Fig. 9). In dorsal view (Fig. 10) copulatory ducts straight and short, pointing to the sides, and leading directly to the sper- mathecae which are curved, kidney shaped. Fertilization ducts thin, long, leaving sper- mathecae from the internal curvature to the midline, then making a loop around copula- tory ducts, very near to the copulatory open- ings, and then continuing more or less straight to the posterior border of the epigynum, in contact with the border between dorsal plate and ventral plate, and curving dorsally at the dorsal border of the genital opening (Fig. 10). Distribution. — MEXICO: Veracruz (Po- trero). Tabasco (Teapa), Guerrero (Chilpan- cingo), and Chiapas (Tonala, Mapastepec, Ta- pachula and Tuxtla Chico). ACKNOWLEDGMENTS We thank the following persons: the au- thorities of the Campo Agricola Experimental Rosario Izapa, (INIFAP) for permitting us to collect in their coffee plantation; T. M. Perez (CNAN, Institute de Biologla UNAM) for permitting us to examine the paratypes of F. caudata and F. huachuca benevola; G. Nieto (El Colegio de la Frontera Sur) for her assis- tance with the scanning micrographs; N.L Platnick (AMNH) and PD. Hillyard (NHM) kindly agreed to compare the specimens col- lected with the types deposited in their re- spective institutions; N. 1. Platnick and G. Hormiga (George Washington University) and L. Leibensperger (Museum of Comparative Zoology) kindly provided useful information and important literature. G. Hormiga, M. L. Draney, J. Miller and the editors of the Journal of Arachnology made many useful sugges- tions for the improvement of the manuscript. This work was supported in part by a grant from the Consejo Nacional de Ciencia y Tec- nologia, Mexico (CONACYT R28867-N). LITERATURE CITED Austad, S.N. 1982. First male sperm priority in the bowl and doily spider, Frontinella pyramitela (Walckenaer). Evolution 36 (4):777-785. Blauvelt, H.H. 1936. The comparative morphology of the secondary sexual organs of Linyphia and some related genera, including a revision of the group. Vol. 2, pp. 81-171, In Festschrift zum 60 Geburtstage von Professor Dr. Embrik Strand. Riga. Bonnet, P. 1956. Bibliographia Araneorum. Tome II, 2eme partie: C-E Toulouse, Les Artisans de LTmprimerie Douladoure. Pp. 919-1926. Bonnet, P. 1957. Bibliographia Araneorum. Tome II, 3eme partie: G-M. Toulouse, Les Artisans de LTmprimerie Douladoure. Pp. 1927-3026. Breene, R.G., D.A. Dean, M. Nyffeler & G.B. Ed- wards. 1993. Biology, predation ecology and sig- nificance of spiders in Texas cotton ecosystems with a key to the species. Texas Agriculture Ex- periment Station, College Station (B-1711). 115 pp. IBARRA-NUNEZ ET AL.— SYNONYMS OF FRONTINELLA TIBIALIS 387 Brignoli, RM. 1983. A catalogue of the Araneae described between 1940 and 1981. Manchester University Press. 755 pp. Bryant, E.B. 1948. The spiders of Hispaniola. Bul- letin of the Museum of Comparative Zoology 100 (4):33 1-447. Gertsch, W.J. & W.L. Jellison. 1939. Notes on a collection of spiders from Montana. American Museum Novitates 1032:1-13. Gertsch, W.J. & L.I. Davis. 1946. Report on a col- lection of spiders from Mexico. V. American Museum Novitates 1313:1-11. Hoffmann, A. 1976. Relacion bibliografica preli- minar de las aranas de Mexico (Arachnida: Ar- aneae). Institute de Biologia, Universidad Na- cional Autonoma de Mexico, Publicaciones Especiales No. 3. Mexico. 117 pp. Hormiga, G. 1994a. A revision and cladistic anal- ysis of the spider family Pimoidae (Araneoidea: Araneae). Smithsonian Contributions to Zoology 549:1-104. Hormiga, G. 1994b. Cladistics and the comparative morphology of linyphiid spiders and their rela- tives (Araneae, Araneoidea, Linyphiidae). Zoo- logical Journal of the Linnean Society 1 11:1-71. Kaston, B.J. 1938. New spiders from New England with notes on other species. Bulletin of the Brooklyn Entomological Society 33:173-191. Kaston, B.J. 1948. Spiders of Connecticut. Bulletin of the Connecticut State Geological and Natural History Survey 70:1-874. Kraus, O. 1955. Spinnen aus El Salvador (Arach- noidea, Araneae). Abhandlungen der Sencken- bergischen Naturforschenden Gesellschaft 493: 1-112. Li, S.Q. & D.X. Song. 1993. On the diagnostic characters of linyphiid spiders, with descriptions of some species (Araneae: Linyphiidae). Scien- tific Treatise on Systematic and Evolutionary Zo- ology 2:247-256. Millidge, A.E 1 984. The taxonomy of the Linyphi- idae, based chiefly on the epigynal and tracheal characters (Araneae: Linyphiidae). Bulletin of the British Arachnological Society 6(6):229- 267. Millidge, A.E 1991. Further linyphiid spiders (Ar- aneae) from South America. Bulletin of the American Museum of Natural History 205:1- 199. Muma, M.H. 1943. Common spiders of Maryland. Natural History Society of Maryland, Baltimore. 179 pp. Petrunkevitch, A. 1911. A synonymic index-cata- logue of spiders of North, Central and South America with all adjacent islands, Greenland, Bermuda, West Indies, Terra del Fuego, Gala- pagos, etc. Bulletin of the American Museum of Natural History 29:1-809. Pickard-Cambridge, EO. 1897-1905. Arachnida- Araneidea and Opiliones. Vol 2. In Biologia Cen- trali Americana. IX -f 610 pp. + 54 plates. Lon- don. Platnick, N.I. 1989. Advances in Spider Taxonomy 1981-1987: A supplement to Brignoli’s A cata- logue of the Araneae described between 1940 and 1981. Manchester Univ. Press. 673 pp. Platnick, N.I. 1993. Advances in Spider Taxonomy 1988-1991, With synonymies and transfers 1940-1980. New York Entomological Society. New York. 846 pp. Platnick, N.I. 1997. Advances in spider taxonomy 1992-1995, With redescriptions 1940-1980. New York Entomological Society. New York. 976 pp. Roewer, C.E 1942. Katalog der Araneae von 1758 bis 1940. 1 Band. Paul Budy, Bremen. 1040 pp. Roewer, C.E 1954. Katalog der Araneae von 1758 bis 1940. 2 Band, Abt. b. Institut Royal des Sci- ences Naturelles de Belgique, Bruxelles. Pp. 927-1751. Rovner, J.S. 1968. Territoriality in the sheet- web spider Linyphia triangularis (Clerck) (Araneae: Linyphiidae). Zeitschrift fuer Tierpsychologie 25:232-242. Song, D., M. Zhu & J. Chen. 1999. The Spiders of China. Hebei Science and Technology Publishing House, Shijiazhuang, China. Suter, R.B. & G. Renkes. 1982. Linyphiid spider courtship: releaser and attractant functions of a contact sex pheromone. Animal Behaviour 30: 714-718. Watson, P.J. 1986. Transmission of a female sex pheromone thwarted by males in the spider Lin- yphia litigiosa (Linyphiidae). Science 233 (4760):219-22L Watson, P.J. 1995. Dancing in the dome. Natural History 104(3):40-43. Wiley-Robertson, M. & P.H. Adler. 1994. Mating behavior of Florinda coccinea (Hentz) (Araneae: Linyphiidae). Journal of Insect Behavior 7(3): 313-326. Manuscript received 25 June 2000, revised 7 May 200 L 2001. The Journal of Arachnology 29:388-395 MONOAMINES IN THE BRAIN OF TARANTULAS (APHONOPELMA HENTZI) (ARANEAE, THERAPHOSIDAE): DIFFERENCES ASSOCIATED WITH MALE AGONISTIC INTERACTIONS Fred Punzo: Dept, of Biology, Box 5F, University of Tampa, 401 W. Kennedy Blvd., Tampa, Florida 33606 USA Thomas Punzo: 2782 Johnson Dr., Albuquerque, New Mexico 50052 USA ABSTRACT. Experiments were conducted to determine the effects of male-male agonistic encounters on changes in monoamine neurotransmitter concentrations in the supraesophageal ganglion (brain) of the tarantula, Aphonopelma hentzi. Serotonin levels were significantly reduced 30 min after fighting in both dominant (66.5 ±9.1 SE nmol/mg protein) and subordinate (42.8 ± 7.6) animals as compared to isolated controls (89,7 ± 13,2), and these differences persisted for up to 24 h. A similar decrease was found for octopamine concentrations in dominant (43.7 ± 7.7) and subordinate (31.2 ± 4.9) spiders when compared to controls (56.9 ± 5.8). In addition, serotonin and octopamine levels were significantly lower in subor- dinate vs. dominant spiders. Agonistic interactions had no effect on the concentrations of dopamine, norepinephrine, and epinephrine. In isolated control spiders, serotonin (89.7 ± 13.2 SE nmol/mg protein) was present in highest concentration in the brain, followed by octopamine (56.9 ±5.8 nmol/mg), dopamine (22.4 ± 3.8 pmol/mg), norepinephrine (15.3 ± 4.7 pmol/mg), and epinephrine (0.57 ± 0.2 pmol/mg). The results indicate that following agonistic encounters, monoamine concentrations in the brain decrease to different levels in winners and losers. This is the first demonstration that the establishment of social status causes changes in brain monoamines in spiders. Keywords: Agonistic interactions, Aphonopelma, CNS monoamines, males Activation of monoaminergic systems in the central nervous system (CNS) has been implicated in the mediation of short-term and chronic physiological stress responses as well as aggressive and social dominance relation- ships in numerous taxa (Eichelman 1987; Bicker & Menzel 1989; Haney et al. 1990; Summers et al. 1995). Most of this work has focused on social interactions in vertebrates exposed to staged encounters between con- specifics under laboratory conditions. For ex- ample, the social status of subordinate animals is associated with an increase in the utilization of the indolalkylamine neurotransmitter (NT) serotonin (5-HT, 5 -hydroxy tryptamine) in var- ious brain regions from fish to mammals (Ha- ney et al. 1990; Winberg et al. 1997; Punzo 2000a). In addition, activation of brain cate- cholaminergic systems including dopamine (DA), norepinephrine (NE), and epinephrine (Epi) in vertebrates is associated with in- creased levels of aggression and dominance (Eichelman 1987; Matter et al. 1998). In general, the brain or supraesophageal ganglion (SEG) of arthropods lies above the esophagus and consists of three major brain regions: the protocerebrum, which contains the ‘higher brain centers' including the optic lobes and nerves; the central body, corpora pedunculata and a variable number of ganglia and associated neuropils; the deuterocere- brum, which innervates the antennae (reduced in arachnids); and the tritocerebrum, which in- nervates the mouthparts (Horridge 1965; Gup- ta 1987). In spiders, the CNS is highly con- densed anteriorly, and the SEG consists primarily of the cheliceral and optic nerves, optic neuropils (masses of axonal fiber tracts) for primary and secondary eyes (‘corpora pe- dunculata’), and a loosely organized central body (Babu 1985; Wegerhoff & Breidbach 1995). There have been few studies on neuro- chemical parameters associated with behavior in general, and agonistic interactions specifi- cally, in arthropods and other invertebrates. For example, increases in protocerebral RNA and protein synthesis have been shown to ac- company learning in molluscs (Kerkut et al. 388 PUNZO & PUNZO— TARANTULA BRAIN MONOAMINES 389 1970; Adamo & Chase 1991), decapod crus- taceans (Punzo 1985), insects (Lin & Roelofs 1992; Punzo 1996), and spiders (Punzo 1988a). Cycloheximide-induced inhibition of brain protein synthesis impaired learning and memory in insects (Jaffe 1980) and spiders (Punzo 1988a), as well as innate phototactic behavior in tenebrionid and passalid beetles (Punzo & Jellies 1980). Changes in levels of brain monoamines have been implicated in a variety of ontogenetic shifts in behavior in honeybee workers including the onset of nest- guarding behavior (Moore et al. 1987) and discrimination between olfactory cues (Mac- millan & Mercer 1987). With respect to aggression and agonistic in- teractions, octopamine turnover rates in- creased significantly in crickets after fighting with conspecifics (Adamo et al. 1995). In- creased foraging activities and nest defense were correlated with higher concentrations of octopamine (OA), dopamine (DA), and sero- tonin (5-HT) in the SEG of worker honeybees (Harris & Woodring 1992). Indeed, it has been suggested that OA is part of a general arousal system which prepares insects for a variety of vigorous skeletal-muscular activities, territo- rial defense, and helps the animal deal with stressful conditions (Corbet 1991; Orchard et al. 1993). Increased 5-HT levels in the brain (SEG) have been implicated in the onset of flight behavior in weevils (Guerra et al. 1991). Lobsters exhibiting dominance over conspe- cifics exhibited higher levels of CNS 5-HT when compared with subordinate animals (Kravitz 1988). Changes in SEG amine concentrations as well as other NTs were shown to be associated with ontogenetic shifts in behavior in solifug- ids (Punzo 1993, 1994). First nymphal instars (Nl) typically have poorly developed chelic- erae, and are gregarious, do not hunt prey, and remain in the nest with their siblings and ma- ternal parent. However, after molting, second- instar nymphs (N2) possess functional chelic- erae and become aggressive (Punzo 1998a). They will cannabilize one another if they do not disperse from the nest. This pronounced increase in aggression is associated with sig- nificant changes in brain 5-HT and DA levels, although OA levels remained relatively con- stant throughout postembryonic development (Punzo 1994). In addition, later nymphal in- stars (N5 — N8) exhibited higher brain concen- trations of acetylcholine (ACh), norepineph- rine (NE), and acetylcholinesterase (AChE) as compared to younger instars (Punzo 1993). Theraphosid spiders exhibit a variety of ag- gressive behaviors. Some of these involve male-male agonistic interactions (Baerg 1958; Minch 1977; Punzo & Henderson 1999; Pun- zo 2000b), while others involve males and fe- males, especially during courtship and mating (Costa & Perez-Miles 1992; Shillington & Verrell 1997), and sometimes end in sexual cannibalism (Punzo & Henderson 1999). A previous study showed that agonistic interac- tions between paired conspecific males of Aphonopelma hentzi (Girard 1854) were ob- served in 24 out of 27 (88.9%) staged en- counters in the laboratory (Punzo & Hender- son 1999). These encounters were initiated by vigorous leg-fencing with each protagonist pushing forcefully against its opponent. These fencing bouts were interrupted from time to time with at least one of the males exhibiting a threat display (elevation of the anterior end of the body, first pair of legs and pedipalps, and opening of the fangs). At least one male exhibited a strike response toward his oppo- nent in 11 out of 24 cases (45.8%). In eight of these instances (33.3%), one of the males was killed. The purpose of this study was to investigate neurochemical parameters associated with ag- onistic interactions between males of the ther- aphosid spider, Aphonopelma hentzi. Neuro- transmitters and comodulators are important regulatory molecules required for the trans- mission of information (nerve impulses) along neural pathways involved in the control of motor movements as well as 'mood’ and mo- tivational states (Ansell & Bradley 1973). Specifically, we were interested in whether or not there were any differences in the concen- trations of monoamines (OA, DA, NE, Epi, 5-HT) in the brains (SEG) of dominant (‘win- ners’) vs. subordinate (‘losers’) males follow- ing agonistic encounters. To our knowledge, this is the first study to address neurochemical correlates of aggression in spiders. METHODS Animals. — Males were collected during July and August of 1997 at a site 3.5 km S of Elgin, Texas (30°32^N, 97°29'W; Bastrop County). This site consisted of a dry wash and surrounding flood plain consisting of sand. 390 THE JOURNAL OF ARACHNOLOGY gravel and adobe soils, with numerous rocks, rock crevices, and burrows. The dominant vegetation included prickly pear cactus (Opuntia), catclaw {Mimosa), mesquite {Pro- sopis), broom weed {Xauthocephalum), and mesquite grass {Bouteloua), Adult males rang- ing in size from 4. 2-6. 7 g were abundant dur- ing this period and were easily found moving about the surface between 2000-0300 h (Cen- tral Standard Time). Spiders were collected and weighed to the nearest 0.1 g using a Ohaus Model 87 portable electronic balance. Spiders were transported to the laboratory and housed individually in plastic cages (20 X 16 X 8 cm). They were provided with water ad libitum and fed three times per week to satiation on a mixed diet of crickets {Gryllus sp,), mealworms {Tenebrio molitor), and grasshoppers {Schistocerca sp.). They were maintained at 22 °C ± 1°, 65% RH, and a photoperid regime of 12L:12D in a Percival Model 805 environmental chamber (Boone, Iowa). Adult males were kept in these condi- tions for two weeks and then re-weighed on the day before the initiation of encounter tri- als. Since previous studies on arachnids have indicated that differences in body size can in- fluence the outcome of aggressive encounters (Faber & Bayliss 1993; Punzo 1998c, 2000b), only males of approximately similar size (6.2- 6.7 g) were used for encounter trials and sub- sequent neurochemical analyses. Voucher specimens have been deposited in the Inver- tebrate Collection at the University of Tampa. Encounter trials. — We used a rectilinear glass arena (26 X 16 X 12 cm) divided into halves by an opaque divider to stage conspe- cific male encounters as described by Punzo (1998c). To summarize, the floor of the arena was provided with a layer of loose sand to a depth of 2 cm. All observations were con- ducted under Black lighting (BioQuip One., Model 2804, Gardena, California). We used a Panasonic PS 150 tape recorder to record ver- bal descriptions of each encounter. Before each encounter trial, a male spider (chosen at random) was placed at each end of the arena, separated by the opaque divider. A trial was initiated by removing the divider and allowing the animals to interact. Within a pe- riod of time ranging from 0.5-8. 5 min over all trials, the contestants made contact with one another (usually with a front leg). In a few trials, one of the spiders would immediately attempt to flee after making initial contact with its opponent. These trials were not used in data analysis. In all other cases {n = 60 trials), following initial contact, one of the spi- ders would begin to push with its front pair of legs against its opponent. The other spider rapidly responded in a similar fashion (leg- fencing). In other cases, after initial contact, one or both spiders would exhibit the threat display, followed by another bout of leg-fenc- ing. In a few instances the fighting escalated until one or both spiders attempted to bite the other. An encounter trial was terminated if at any time during the encounter one of the spi- ders backed away and rapidly fled from the vicinity of the other spider and attempted to crawl out of the chamber. The spider that held its ground was recorded as the 'winner’ (dom- inant animal), and the spider that fled, the Toser’ (subordinate). Each pair of contestants were subjected to only one encounter trial as described by Summers & Greenberg (1995) in their study of male-male aggression in lizards. We conducted a total of 60 encounter trials comprising 60 pairs of contestants {n = 120). Neurochemical analyses of brain tis- sues*— Immediately following their designa- tion as dominant or subordinate (based on the outcome of encounter bouts), paired contes- tants were randomly assigned to one of three groups; each group consisted of 40 spiders (20 pairs). Spiders in group 1 (Gl) were anaes- thetized with CO2 thirty min after encounter trials, and their brains (SEG) removed in a cold room, weighed to the nearest 0.1 g on an electronic analytical balance, and frozen at -80°C as described by Punzo (1988b) for subsequent neurochemical analyses. Group 2 (G2) and group 3 (G3) spiders were anaesthe- tized and their brains frozen at 24 hr and 48 h, respectively, after encounter trials. In this way, we were not only able to determine what neurochemical changes, if any, followed male-male aggression, but also how rapid the response might be, and how long these chang- es might persist. The brains from another group of 20 spiders (G4) maintained in iso- lation and not exposed to encounter trials were used as controls. After thawing, all glandular and peripheral fatty tissue was carefully removed from the surface of the SEG (Murdock & Omar 1981). The SEG were then weighed to the nearest 0,01 g on a Sartorius Model 54C electronic PUNZO & PUNZO— TARANTULA BRAIN MONOAMINES 391 analytical balance. Brain protein determina- tions were conducted using the standard pro- cedure described by Lowry et aL (1951) and expressed as percent (brain protein/brain weight) (Meyer et al. 1984). The SEG from the dominant and subordinate spiders were an- alyzed to determine the concentrations of the monoamine eeurotransmitters, 5-HT, OA, DA, and NE, using high performance liquid chro- matography with electrical detection (HPLC- ED, Beckman Model 47 A) as described by Brandes et al. (1990). To summarize, each brain tissue sample was placed in a 750 pi glass vial and homogenized in 50 pi of a 200 mM perchloric acid (PA) solution. Following homogenization, an additional 50 pi of PA were added to each vial. Samples were then centrifuged at 10,000 g and 4 °C for 3 min in a Sorvall Model 100 A high speed refrigerated centrifuge. Twenty pi of supernatant were in- jected directly into the HPLC column (40 cm in length, with a 0.2 p pore diameter) packed with Hypersil and provided with a Hewlett- Packard 760E detector (0.40 V). The mobile phase (flow rate, 3000 psi) used to elute the monoamines consisted of 12% acetonitrile, 20 mM sodium acetate, 100 mM sodium dihy- drogen orthophosphate, 2.5 mM octane sul- fonic acid, and 0.3 mM EDTA disodium salt adjuested to pH 4.2 and filtered through a 0.45 pm filter. Each sample was compared to 5-HT and DA standards tested at the beginning of each assay run and retested at 30 min inter- vals. Monoamine concentrations were ex- pressed as nmol or pmol/mg protein as des- ribed by Meyer et al. (1984). All statistical procedures followed those de- scribed by Sokal & Rohlf (1995). Compari- sons between mean concentrations of mono- amine NTs for the various groups were conducted using an analysis of variance (AN- OVA), followed post-hoc by a Duncan’s mul- tiple range test at a significance level of 0.05, Significant differences between dominant and subordinate males following aggressive en- counters were determined using an indepen- dent-samples t test {P < 0.05), RESULTS Brain weights for all spiders ranged from 8.98-9.32 mg (mean: 9.11 ± 0.56 SE). Brain protein/brain weight (%) ranged from 7. 2-7. 6. An analysis of variance (ANOVA) indicated that there were no differences in mean brain weights and brain protein values between dominant, subordinate or isolated control spi- ders {P < 0.5), Serotonin was the monoamine found in the highest concentration in the brains of isolated control of A. hentzi (Table 1), This was followed in decreasing order by OA, DA, NE, and Epi. The effects of agonistic interactions be- tween coespecific males on SEG monoamine concentrations at various time intervals fol- lowing encounter trials are shown in Table 1. Serotonin (5-HT) levels were significantly re- duced in spiders losing aggressive encounters (subordinates) for up to 24 h as compared to dominant animals (t = 9.4, P < 0,01). This difference persisted for at least 24 h, with lev- els returning to normal after 48 h (F = 3.36, P < 0.05). In addition, the brains of dominant spiders contained significantly lower levels of 5-HT than those of the isolated controls (t — 6.8, P < 0.05) for up to 24 h. These changes in 5-HT and OA levels associated with fight- ing occurred quite rapidly since changes were detected after only 30 min following an en- counter. A similar pattern was found for OA levels which were also significantly reduced in sub- ordinate vs. dominant animals (t = 6,2, P < 0.02). However, the reduced levels of OA as- sociated with agoeisitic interactions returned to control levels within 24 hr. With respect to DA, NE, and Epi, no differences were found between spiders exposed to agonistic encoun- ters and controls at any time interval {P < 0.5). DISCUSSION Although the profile for monoamine con- centrations in the SEG of A. hentzi (5-HT > OA > DA > NE < Epi) is in general agree- ment with what little information is available on the neurochemistry of spiders, differences in the NT profiles for spiders from different families have been reported (Florey 1967; Meyer et al. 1984; Meyer 1991) Similar con- centrations were reported for NE and DA from the brain of the theraphosid, Aphonopeh ma eutylenum Chamberlin 1918, although no data were presented for 5-HT and OA (Meyer et al. 1984). In contrast, the brain of Pardosa amentata (Clerck 1932) contained much high- er concentrations of NE (174.9 pmol/mg ± 5.0 SE), a condition most likely associated with the noradrenergic system of the optical 392 THE JOURNAL OF ARACHNOLOGY Table 1 . — Concentrations of various monoamines (in nmol or pmol/mg protein) in the supraesophageal ganglia (SEG) of Aphonopelma hentzi following agonistic interactions between conspecific males. Brain analyses were conducted from tissue extracted from isolated control spiders, as well as from the brains of dominant and subordinant spiders removed 5 min (20 pairs; N = 40) 24 h {n = 20 pairs), and 48 h {n — 20 pairs) after an encounter trial. Data expressed as means; values in parentheses represent (± SE). Values followed by asterisks are significantly different than controls: ** (P < 0.01); * (P < 0.05). See text for details. Time after encounter Neurotransmitter Controls 5 min 24 h 48 h Serotonin (5-HT) (nmol/mg) Subordinate 89.7 (13.2) 42.8** (7.6) 51.3** (8.3) 92.2 (12.6) Dominant 66.5* (9.1) 73.9* (10.4) 86.3 (9.5) Octopamine (nmoi/mg) Subordinate 56.9 (5.8) 31.2** (4.9) 54.8 (8.1) 60.1 (10.6) Dominant 43.7* (7.7) 55.3 (5.2) 57.8 (8.2) Dopamine (pmol/mg) Subordinate 22.4 (3.8) 19.6 (5.1) 23.6 (7.1) 21.9 (5.5) Dominant 22.2 (7.8) 20.6 (3.5) 24.4 (6.2) Norepinephrine (pmol/mg) Subordinate 15.3 (4.7) 18.1 (5.8) 14.3 (2.9) 17.3 (6.6) Dominant 16.6 (4.4) 18.5 (3.1) 14.9 (4.1) Epinephrine (pmol/mg) Subordinate 0.57 (0.2) 0.61 (0.1) 0.55 (0.2) 0.58 (0.1) Dominant 0.52 (0.2) 0.63 (0.3) 0.56 (0.2) brain centers which are more highly devel- oped in salticids (Meyer & Jehnen 1980). Ly- cosids and agelenids also contained higher levels of DA and NE as compared to thera- phosids, although not as high as salticids (Meyer 1991). The most pronounced changes in mono- amine levels involved 5-HT. They occurred among subordinate males 30 min after fight- ing, although 5-HT levels were reduced in dominant males as well. Thus, fighting be- tween male spiders resulted in a decrease in SEG serotonin levels. To our knowledge, this is the first demonstration of an association be- tween monoaminergic activity and aggression in spiders. Serotonin has been identified with aggressive behavior in other arthropods as well. Injection of 5-HT into lobsters and cray- fish caused them to elevate and flex their tails, which represent behavioral acts associated with the expression of dominance (Yeh et al. 1996). These observations are interesting since a similar reduction in brain serotonin levels ac- companying fighting and territorial defense has been reported for a number of vertebrates. Indeed, it has been well established that a va- riety of stimuli, including social interactions, activate endocrine stress mechanisms in ver- tebrates, which are thought to be mediated by changes in CNS neurotransmitters brought about primarily via activation of monoamin- ergic systems (Ansell & Bradley 1973; Ei- chelman 1987). Eor example. Summers & Greenberg (1995) showed that 5-HT levels decreased significantly after one h and one day in the brains (diencephalon and non-optic lobe midbrain) of lizards (Anolis carolinensis) losing aggressive interactions. Similarly, no changes were detected for NE and DA levels over this time interval. However, subordinate males exhibited significantly lower DA levels after one week than did subordinates after one h. Changes in the serotonergic content and turnover between individuals of different so- cial status were found in the telencephalon and diencephalon of territorial vs. satellite males in the lizard, Sceloporus jarrovi (Matter et al. 1998). In addition, the levels of 5-HT in the telencephalon and diencephalon were found to decrease significantly following male-male aggression in rodents (Haney et al. 1990; White et al. 1991) and fish (Winberg et al. 1997). Immediately following aggressive defense of territories, territorial male lizards {S. jar- PUNZO & PUNZO— TARANTULA BRAIN MONOAMINES 393 rovi) exhibited higher Epi levels as compared to males that did not experience aggressive encounters (Matter et ah 1998). In contrast, no comparable changes in Epi levels in the SEG of A. hentzi were observed after fighting. Octopamine levels also decreased signifi- cantly in males of A. hentzi exposed to ag- gressive interactions. This suggests that an ac- tivation of the octopaminergic system, in addition to serotonergic activation, follows aggression in spiders. This is not surprising since OA appears to be central in eliciting the overall arousal response of arthropods (Krav- itz 1988; Corbet 1991), and elevated OA ac- tivity has been shown to accompany stress (Downer & Hiripi 1993; Harris & Woodring 1992), increased locomotor activity (Orchard et al. 1993; Adamo et al. 1995), courtship (Downer & Hiripi 1993), and a wide range of systemic physiological responses including respiration, gastrointestinal peristalsis, cardio- acceleration, Malpighian tubule filtration, gly- cogenolysis, and pheromone production (Cor- bet 1991) in insects. It has been further suggested that certain behavior patterns can be triggered by the activation of specific octo- paminergic pathways in arthropods, an idea known as the ‘orchestration hypothesis’ (Som- bati & Hoyle 1984). For example, administra- tion of exogenous OA has been shown to trig- ger diurnal hyperactivity in nocturnal moths (Shimizu & Fukamii (1981). Changes in OA levels in the CNS have been associated with ontogenetic shifts in specific behavioral acts in social insects (Bicker & Menzel 1989; Brandes et al. 1990), and an increase in OA activity was found in the brains of crickets following aggressive interactions between conspecifics (Adamo et al. 1995). Although most of the research on OA has focused on insects, some previous studies, in- cluding the present one, suggest that this monoamine plays an important role in regu- lating the behavior of other arthropods as well. For example, the tail flip response, an integral behavioral component of the escape response of crayfish, is enhanced by OA (Bicker & Menzel 1989). The injection of OA into freely moving lobsters elicited submissive body postures toward conspecifics (Kravitz 1988). The application of OA caused en- gorged ticks to detach from their hosts (Mason 1986). With respect to OA, direct comparison with vertebrates is not possible since OA has not been identified as a NT in this group. In conclusion, significant changes in brain concentrations of 5-HT and OA result from male-male agonistic encounters in tarantulas. It has been well established that 5-HT is a CNS monoamine involved in the expression of dominance and aggression in vertebrates, and the results of this study suggest that the establishment of social status in spiders causes changes in brain monoamine levels and may play a role in the elicitation of communicative displays as well. Future studies should focus on other species of spiders as well as other arachnids in order to determine if similar changes in monoamine profiles are associated with aggression in these groups. ACKNOWLEDGMENTS We are indebted to C. Bradford, R. Khatibi, M. Harvey, and R. B. Suter for commenting on an earlier draft of the manuscript, B. Gar- man for consultation on statistical analyses, and J. Bottrell for assistance in the collection of specimens. A University of Tampa Faculty Development Grant and a Delo Foundation Research Grant (DFRG- 10-204) to FP made much of this work possible. LITERATURE CITED Adamo, S.A., C.E. Linn & R.R. Hoy. 1995. The role of neurohormonal octopamine during ‘fight or flight’ behavior in the field cricket Gryllus bi- maculatus. Journal of Experimental Biology 198: 1691-1700. Adamo, S.A. & R. Chase. 1991. Central arousal and sexual responsiveness in the snail, Helix aspersa. Behavioral & Neural Biology 55:194-213. Ansell, G.B. & P. Bradley. 1973. Macromolecules and Behavior. Univ. Park Press, Maryland. Babu, K. S. 1985. Patterns of arrangement and con- nectivity in the central nervous system of arach- nids, Pp. 3-19, In Neurobiology of Arachnids. (F. G. Barth, ed.). Springer- Verlag, Berlin. Baerg, W.J. 1958. The Tarantula. University of Kansas Press, Lawrence, Kansas. Bicker, G. & R. Menzel. 1989. Chemical codes for the control of behavior in arthropods. Nature 337:33-39. Brandes, C., M. Sugawa & R. Menzel. 1990. High performance liquid chromatography (HPLC) measurement of monoamines in single honeybee brains reveals caste-specific differences between worker bees and queens in Apis melUfera. Com- parative Biochemestry & Physiology 97C:53-57. Corbet, S.A. 1991. A fresh look at the arousal syn- 394 THE JOURNAL OF ARACHNOLOGY drome of insects. Advances in Invertertebrate Physiology 23:81-116. Costa, EG. & F. Perez-Miles. 1992. Notes on mat- ing and reproductive success of Ceropelma lon- gisternalis (Araneae, Theraphosidae) in captivity. Journal of Arachnology 20:129-133. Downer, R.G.H. & L. Hiripi. 1993. Biogenic amines in insects. Pp. 23-38, In Insect Neuro- chemistry and Neurophysiology. (A.B. Borkovek & M.J. Loeb, eds.). CRC Press, Boca Raton, Florida. Eichelman, B. 1987. Neurochemical and psycho- pharmacological aspects of aggressive behavior. Pp. 697-704, In Psychopharmacology: Third Generation of Progress. (H. Y. Meltzer, ed.). Ra- ven Press, New York. Faber, D.B. & J.R. Bayliss. 1993. Effects of body size on agonistic encounters between male jump- ing spiders (Araneae: Salticidae). Animal Behav- iour 45:289-299. Florey, E. 1967. Neurotransmitters and modulators in the animal kingdom. American Society of Ex- perimental Biology 26:1 164-1 178. Guerra, A. A. & R.W. Steger & J.C. Guerra. 1991. The possible role of brain serotonin in seasonal dispersal behavior of the boll weevil. South- western Entomologist 16:92-98. Gupta, A.P. 1987. Arthropod Brain: Its Evolution, Development, Structure, and Functions. Wiley Interscience Pubis., New York. Haney, M., K. Noda, K. Kream & K.A. Miczek. 1990. Regional serotonin and dopamine activity: sensitivity to amphetamines and aggressive be- havior in mice. Aggessive Behavior 16:259-270. Harris, J.W. & J, Woodring. 1992. Effects of stress, age, season, and source colony on levels of oc- topamine, dopamine, and serotonin in the hon- eybee {Apis mellifera) brain. Journal of Insect Physiology 38:29-35. Honidge, G.A. 1965. Arthropoda: general anatomy, Pp. 801-964, In Structure and Function in the Nervous System of Invertebrates. Vol. 2 (T. H. Bullock & G. A. Horridge, eds.). W. H. Freeman, San Francisco, California. Jaffe, K. 1980. Effects of cycloheximide on protein synthesis and memory in preying mantids. Phys- iology & Behavior 25:367-371. Kerkut, G.A., M. French & R. Walker. 1970. The location of axonal pathways of identifiable neu- rons of Helix aspersa using procion yellow M- 4R. Comparative Biochemistry & Physiology 32: 681-690. Kravitz, E.A. 1988. Hormonal control of behavior: amines and the biasing of behavioral output in lobsters. Science 241:1775-1780. Lin, G. & C. Roelofs. 1992, The effects of biogenic amines on learning in arthropods. Archives of Insect Biochemistry & Physiology 20:359-366. Lowry, O.H., N.J. Rosebrough, A. Farr & R.J. Randall. 1951. Protein measurement with the fol- in reagent. Journal of Biological Chemistry 192: 265-275. Macmillan, C.S. & A.R. Mercer. 1987. An inves- tigation of the role of dopamine in the antennal lobe of the honeybee, Apis mellifera. Journal of Comparative Physiology A 160:359-366. Mason, S.T. 1986. Catecholamines and Behaviour. Cambridge Univ. Press, London. Matter, J.M., P.J. Ronan & C.H. Summers. 1998. Central monoamines in free-ranging lizards: dif- ferences associated with social roles and territo- riality. Brain Behavior & Evolution 51:23-32. Meyer, W. 1991. The biology of excitatory neuro- transmitters in the central nervous system of spi- ders. Histochemistry 59:177-189. Meyer, W. & R. Jehnen. 1980. The distribution of monoamine oxidase and biogenic monoamines in the central nervous system of spiders (Arachni- da: Araneida), Journal of Morphology 164:69- 81. Meyer, W., C. Schlesinger, & H.M. Poehling. 1984. Comparative quantitative aspects of putative neurotransmitters in the central nervous system of spiders (Arachnida: Araneida). Comparative Biochemistry & Physiology 78C:357-362. Minch, E.W. 1977. The Behavioral Biology of the Tarantula, Aphonopelma chalcodes. Ph.D. Dis- sert., Arizona State University, Tempe, Arizona. Moore, A.J.M., M.D. Breed & M.J. Moore, 1987. The guard honeybee: ontogeny and behavioral variability of workers performing a specialized task. Animal Behaviour 35:1159-1167. Murdock, L.L. & D. Oman. 1981. N-acetyldopa- mine in insect nervous tissue. Insect Biochem- istry 1 1 : 161-166. Orchard, L, J. Ramirez & A.B. Lange. 1993. A mul- tifunctional role for octopamine in locust flight. Annual Review of Entomology 38:227-249. Punzo, F. 1985. Recent advances in behavioral plas- ticity in insects and decapod crustaceans. Florida Entomologist 68:89-104. Punzo, F. 1988a. Learning and localization of brain function in the tarantula spider, Aphonopelma chalcodes (Orthognatha, Theraphosidae). Com- parative Biochemistry & Physiology 89A:465- 470. Punzo, F. 1988b. Physiological amino acids in the central nervous system of the tarantulas, Aphon- opelma chalcodes and Dugesiella echina (Or- thognatha: Theraphosidae). Comparative Bio- chemistry & Physiology 90C:38 1-383. Punzo, F. 1993. An analysis of free amino acids, neurotransmitters and enzymes in the nervous system of Solpugida (Arachnida). Comparative Biochemistry & Physiology 106C:699-703. Punzo, F. 1994. Changes in brain amine concentra- tions associated with postembryonic develop- ment in the solpugid, Eremobates palpisetidosus PUNZO & PUNZO— TARANTULA BRAIN MONOAMINES 395 Fichter (Solpugida, Eremobatidae). Journal of Arachnology 22:1-4, Punzo, F. 1996. Localization of brain function and neurochemical events associated with learning in insects. Recent Trends in Comparative Biochem- istry & Physiology 2:9-16. Punzo, F. 1998a. The Biology of Camel-Spiders (Arachnida, Solifugae). Kluwer Acad. PubL, Norwell, Massachusetts. Punzo, F 1998b. The effects of maternal nest guard- ing behavior by Eremobates marathoni Muma & Brookhart on the survivorship of offspring (So- lifugae, Eremobatidae). Bulletin British Arach- nological Society 11:54-56. Punzo, F. 1998c. Intraspecific male aggression in Arenotherus joshuaensis Brookhart & Muma, and Eremobates marathoni Muma (Solifugae, Eremobatidae). Southwestern Naturalist 43:291- 295. Punzo, E 2001. Neurochemical correlates of ago- nistic interactions and dominance between males of the brown anole, Anolis sagrei. Florida Sci- entist, 64:131-139. Punzo, E 2000b. Desert Arthropods: Life History Variations. Springer- Verlag, Heidelberg, Germa- ny. Punzo, E & L. Henderson. 1999. Aspects of the natural history and behavioral ecology of the ta- rantula spider Aphonopelma hentzi (Girard 1 854) (Orthognatha, Theraphosidae). Bulletin British Arachnological Society, 11:121-128. Punzo, F & J. Jellies. 1980. Effects of cyclohexi- mide-induced protein synthesis inhibition on the phototactic behavior of Tenebrio molitor (Cole- optera: Tenebrionidae) and Popilius disjunctus (Coleoptera: Passalidae). Journal of the Kansas Entomological Society 53:597-606. Shillington, C. & P. Verrell. 1997. Sexual strategies of a North American ‘tarantula’ (Araneae: Ther- aphosidae). Ethology 103:588-598. Shimizu, T. & J. Fukamii. 1981. Role of octopa- mine in noctuid activity patterns. International Pest Control 23:166-175. Sokal, B. & R.J. Rohlf. 1995. Biometry. 3rd ed. W. H. Freeman, New York. Sombati, S. & G. Hoyle. 1984. Generalizations of specific behaviours in a locust by local release into neuropil of the natural neuromodulator oc- topamine. Journal of Neurobiology 15:481-506. Summers, C.H. & N. Greenberg. 1995. Activation of central biogenic amines following aggressive interaction in male lizards, Anolis carolinensis. Brain Behavior & Evolution 45:339-349. Wegerhoff, R. & O. Breidbach. 1995. Comparative aspects of the chelicerate nervous system, Pp. 159-179, In The Nervous Systems of Inverte- brates: An Evolutionary and Comparative Ap- proach. (O. Breidbach & W. Kutsch, eds.). Birk- hauser, Basel, Switzerland. White, S.M., R.F. Kucharik & J.A. Moyer. 1991. Effects of serotonergic agents on isolation-in- duced aggression. Pharmacology Biochemistry & Behavior, 39:729-736. Winberg, S., Y. Winberg & R.D. Fernald. 1997. Ef- fect of social rank on brain monoaminergic ac- tivity in cichlid fish. Brain Behavior & Evolution 49:230-236. Yeh, S., R.A. Fricke & D.H. Edwards. 1996. The effect of social experience on serotonergic mod- ulation of the escape circuit of crayfish. Science 271:366-369. Manuscript received 6 August 2000, revised 2 May 200 L 2001. The Journal of Arachnology 29:396-412 HABITAT DISTRIBUTION AND LIFE HISTORY OF SPECIES IN THE SPIDER GENERA THERIDION, RUGATHODES, AND WAMBA IN THE GREAT SMOKY MOUNTAINS NATIONAL PARK (ARANEAE, THERIDHDAE) Grant Jeffrey Stiles and Frederick A. Coyle: Department of Biology, Western Carolina University, Cullowhee, NC 28723 USA. e-mail: coyle@wcu.edu ABSTRACT. Based largely on 668 one-hour samples collected during a survey of spiders in 16 major habitats of the Great Smoky Mountains National Park, habitat distributions, life history patterns, and other natural history traits are described for 14 species in the related theridiid genera Theridion, Rugathodes, and Wamba. Two to eight of these species were found in each of the 16 habitats. Among-habitat differences in the kinds and relative abundance of these species suggest that they may be good predictors of habitat. Richness, diversity, and evenness of this species assemblage are highest in middle to low elevation habitats. Rugathodes aurantiiis and R. sexpimctatiis, two boreal sister species, are abundant in the highest elevation habitats, but differ sharply in microhabitat and habitat preference. Theridion frondeiim is much more common in high elevation habitats than is its sister species, T. albidum, which is virtually limited to middle and low elevation habitats. Theridion lyricurn is most common in dry, pine-dominated forests. The three most common species (R. aurantiiis, R. sexpimctatiis, and T. frondeum) have a simple annual life cycle of five or six instars and similar phenologies: they mate during late spring {R. aurantiiis and R. sexpimctatiis) and early summer {T. frondeum) and over-winter in antepenultimate and/or penultimate instars. Female-biased sex ratios were observed in juvenile cohorts of these species. Rugathodes aurantiiis, its natural history previously unknown, places its webs on the undersides of broad-leafed herbs close to the ground and captures small flying insects. Adult females engineer partly folded leaf retreats, carry the egg sac when disturbed, help their instar II spiderlings exit the egg sac, and then share the retreat with these spiderlings for at least a few days. Rapid early development (about two weeks from oviposition to emergence from the egg sac), the presence of females with egg sacs throughout the summer, and smaller clutch sizes in late summer suggest that a typical R. aurantiiis female produces more than one clutch. Keywords: Spiders, habitat distribution, life history, Theridion, Rugathodes The spider genus Theridion Walckenaer 1805, as defined by Levi (1957, 1959, 1963) is cosmopolitan and large (over 90 species have been recorded from North America alone), and many of its species are abundant in favorable habitats. But published knowl- edge of these species consists of little more than taxonomic descriptions and brief, scat- tered natural history observations (e.g., Emer- ton 1902; Archer 1947; Comstock 1948; Levi 1957; Bristowe 1958; Toft 1976; Kaston 1981; Hanggi et al. 1995; Roberts 1995). Moreover, it is now generally accepted that Theridion is polyphyletic, and phylogenetic research is underway to determine the rela- tionships of these species (L Agnarsson & M. Arnedo pers. comm.). Wunderlich (1987) re- moved Rugathodes Archer 1950 from Theri- dion, and Platnick (1997), following Wunder- lich’s and Archer’s (1950) views, formally transferred from Theridion to Rugathodes two species included in our current study, Rugath- odes sexpunctatus (Emerton 1882) and Ru- gathodes aurantius (Emerton 1915). Wamba crispulus (Simon 1895), another species in our study and one considered by Levi (1963) to belong to Theridion, was transferred to Wam- ba by Wunderlich in 1995. The other eleven of the 14 species included in our study still remain in Theridion, but some of these may be removed following phylogenetic analysis (L Agnarsson & M. Arnedo pers. comm.). Both morphological (Levi 1957) and molec- ular (M, Arnedo pers. comm.) evidence indi- cate that Theridion frondeum Hentz 1850 and Theridion albidum Banks 1895 are sister spe- cies and that R. sexpunctatus and R, aurantius are also sister species. Morphological evi- dence indicates that Theridion cheimatos Gertsch and Archer 1942 will eventually be 396 STILES & COYLE— BIOLOGY OF THERIDION, RUGATHODES, AND WAMBA 397 transferred to Rugathodes (L Agnarsson pers. comm.). The need to learn more about habitat pref- erences, life histories, and other facets of spe- cies’ natural histories before being able to an- swer important questions about the structure and dynamics of communities has been em- phasized by many (e.g., Duffy 1978; Strong et ah 1984; Wilson 1992; Polis et al. 1996). Our goal is to begin providing these kinds of information for North American species of Theridion, Rugathodes, and Wamba, thereby making them more accessible to ecologists and other evolutionary biologists. The Great Smoky Mountains National Park Biosphere Reserve (GSMNP), due partly to its wide el- evation range (275-2013 m), large size (207,000 ha), and low temperate latitude (35°35'N, in the southern Appalachian Moun- tains), comprises a rich mosaic of biotic com- munities appropriate for pursuing this goal. By distributing a systematic sampling effort among the major habitats of this park, we have expanded our knowledge of the habitat distribution patterns of these 14 related theri- diid species, completed the first life history analyses for three of them, and added knowl- edge about their reproductive biology. We also hope that such knowledge will be- gin to help us evaluate this group’s potential as an indicator assemblage (a group of species that can be used to monitor and predict chang- es or species richness in biotic communities). The search for such indicator groups is an im- portant focus of some ecologists, conservation biologists, and environmental monitoring agencies (Thomas 1972; Kremen et al. 1993; Colwell & Coddington 1994; Russell et al. 1995; Norris 1999). An ideal indicator group should be easily sampled, abundant, diverse, geographically widespread, sensitive to envi- ronmental change, and important to commu- nity dynamics (Noss 1990; Kremen et al. 1993). Since spider taxa appear to meet these criteria (e. g., Uetz 1979; Coyle 1981; Bracher & Bider 1982; Christenson et al. 1990; Riech- ert & Bishop 1990; Coddington & Levi 1991; Carter & Rypstra 1995), they deserve to be included in the search for indicator taxa. METHODS Habitat distribution. — A team of four col- lectors used a modified Coddington sampling protocol (Coddington et al. 1996) to obtain the 668 one-hour ground (235), aerial (172), beat (206), and sweep (55) samples used in this project. Ground collection involved searching mostly on hands and knees, exploring leaf lit- ter, logs, rocks, and plant surfaces below knee level (ca. 0-49 cm above ground). Aerial sampling involved searching leaves, branches, tree trunks, and spaces in between, from knee height up to maximum overhead arm’s reach (ca. 50-220 cm above ground). Beating con- sisted of striking vegetation with a 1 m long stick and catching the falling spiders on a 0.5 m^ canvas sheet held horizontally below the vegetation. Hands and aspirators were used to collect the spiders into 80% ethanol. One sam- ple unit equaled 1 hour of uninterrupted effort with one of these three methods during which the collector attempted to collect every spider encountered. During each hour the team as a whole used all three methods in the same area. In the non-forest communities (grass bald, mountain wetland, and native grassland sites) 1-hour sweep sampling was substituted for ae- rial and/or beating methods (see Appendix); sturdy sweep nets with 38 cm diameter hoops were used and the number of sweeps per hour (175-400, mean and SD - 268 ± 47.7) de- pended primarily on vegetation structure and spider abundance. Two sets of samples (one in the spring and one in late summer) were collected from 17 focal sites, each site representing one of the 16 major habitat (community) types found in the GSMNR Habitat type, locality data, col- lecting dates, and sample sizes for each meth- od at each site are given in the Appendix. At each site (except for the high grass bald. Table Mountain pine, and Indian Creek wetland sites) equal or nearly equal numbers of sam- ples were collected with each of the methods employed. Two of the sites (low grass bald and heath bald) were sampled in 1995. The others were sampled in 1996. All adult and juvenile Theridion, Rugathodes, and Wamba specimens were sorted from each sample and identified to species. The pigment patterns are distinctive for each species in all but the first instar. The most similar species, T. frondeum and T. albidum, differ in the form of the lon- gitudinal median marks on the carapace; T. frondeum has two lines or a broad band and T. albidum has one line. Voucher specimens for each species have been deposited in the 398 THE JOURNAL OF ARACHNOLOGY National Museum of Natural History, Smith- sonian Institution. The relative abundance (mean number of individuals per 1-hour sample) of each species was computed for each of the 17 sites. This index of abundance does not show the often wide variation in number of individuals among 1-hr samples at each site. This varia- tion is due largely to the fact that each method samples only a subset of microhabitats, to spa- tial environmental variation within each site, and to seasonal changes in spider abundance correlated with species’ phenologies. Shannon diversity and Pielou’s evenness indices were used to measure the diversity of these theridiid species at each site (Magurran 1988). Life history. — For the three most common species (R. aurantius, R. sexpunctatus, and T. frondeum) tibia I length (ITL) was measured along its dorsal surface in every specimen col- lected at a site where the species was com- mon. Toft (1976) demonstrated that ITL dis- tinguishes spider instars more clearly than does either the length or width of the cara- pace. Measurements were performed with a Wild M-5 stereomicroscope at SOX magnifi- cation and are accurate to ±0.0185 mm. StatView 4.5 (Abacus Concepts) was used to generate ITL frequency distribution histo- grams for these samples. From these histo- grams it was possible to determine instar num- ber. Instars were also distinguished by the distinctive widths of the palpal tarsi of pen- ultimate and, for R. aurantius and R. sex- punctatus, antepenultimate males. The maxi- mum width of the palpal tarsus in dorsal view (PTW) was measured at lOOX magnification (accurate to ±0.00925 mm) for instai' III speci- mens of R. aurantius to confirm this. Phenol- ogy and generation time were determined by examining the relationship between instar dis- tribution and collecting date. These life his- tory analyses were based on ITL measure- ments of 375 R. aurantius, 139 R. sexpunctatus, and 843 T. frondeum individu- als (see Figs. 3, 5 and 6 for the sites and dates represented by these samples). The pattern of early postembryonic development was deter- mined by examining (at 24-lOOX magnifica- tion) the spiderlings and shed exuviae in elev- en R. aurantius and four T. frondeum egg sacs containing spiderlings. One field collected an- tepenultimate male and an antepenultimate fe- male of R. aurantius were reared to adulthood. Other observations. — The vertical micro- habitat distribution for each species was ana- lyzed by computing its relative abundance in aerial (above knee level) vs. ground (below knee level) samples. Beat and sweep samples were not used because each of these methods sampled spiders both above and below knee- level. The Mann Whitney U test was used to see if the relative abundance values for ground and aerial samples were significantly different (at P < 0.05). Field notes, sketches, measure- ments, and close-up photos of webs and spi- ders were used to characterize web structure and spider behavior. Prey items were collected from webs in the field. Egg diameters were measured with a Wild M-5 stereomicroscope at 50 X magnification with an accuracy of ±0.0185 mm. Clutch sizes were determined by counting the number of eggs and spider- lings in each field-collected egg sac. An un- paired r-test was used to determine if clutch size and body size differed significantly be- tween early and late summer samples of R. aurantius and if egg and body sizes of R. au- rantius differed significantly from those of T. frondeum. Several live specimens (predomi- nantly adult females, most with egg sacs) of R. aurantius and T. frondeum were kept in small transparent plastic terraria for a few weeks to observe rates of brood development and behavior. RESULTS Habitat distribution, — Fourteen Theri- dion, Rugathodes, and Wamba species were found in the GSMNP. At each of the seven highest elevation sites (over 1500 m) one of these species was much more abundant than any other (Table 1, Fig. 1). Rugathodes sex- punctatus was common (relative abundance = 0. 5-2.0) or abundant (relative abundance > 2.0) in the spruce-fir (1830 m) and spruce (1715 m) sites, R. aurantius was abundant in the high grass bald (1755 m) and beech gap (1645 m) sites, and T. frondeum was abundant in the northern hardwood (1615 m), red oak (1555 m), and low grass bald (1505 m) sites. Middle to low elevation sites (below 1400 m) tended to contain more species (2-8, mean and SD = 4.4 ± 1.6) than the high elevation sites (2-4, 2.4 ± 1.0) and to lack abundant species. Shannon diversity and Pielou eveness index values for the sets of these species at each site show this same pattern (Table 2). Table L— Relative abundance of Theridion, Rugathodes and Wamba species at 17 focal sites representing 16 major habitats in the Great Smoky Mountains National Park. Species arranged alphabetically by species name. The single specimen of T. alabamense was found in a leaf litter sample. STILES & COYLE— BIOLOGY OF THERIDION, RUGATHODES, AND WAMBA 399 o a 5-. Q « A a 5 I o a cd CO 'S s 0) i 3 II e e Ci, > S '5 6^ o e s ^ oo i I o ? § S 0 a 1 § e o w 4) ^ e m o m m cu CO p T o d- d- p d CO d d d d- in r- CO r-i CO 00 TO VO o CO o p CM CO d” o d d d d d d d d in On O d d m O' o in r- 00 T p p A CM d f— ( d c6 CO d d d CO 1—1 d d d d CO in »n o 5-^ 1— ( o d d d CO VO o o o o d d d d VO VO o o ro M ^ CNl ON p CO p O >r) O A o m CM TO ON CO o o o VO in n d d d d d d d ^2 QO ^ CO 1-1 CO (C g 6 m d ^ C m OT X .-N o m o ^ a,j= ^ bD g 0) ^ 4J O -G ..C S B ^ a « O w m X m o IT) ■5^ w O VN ^ On ^ in 2 CO in d ’-i ^ ^ W w C CO 2 ^ = “-s ■g 'd ^ d X « o o Ld ^ N K S 0 s ’Sm .s ‘b fi p o 0 in S o O ^ m c O I'i -a 0 2 ^ -o -d d e Cd Cd >0 E R d -S IT) 0 d A fo ^ X X > ^ ^ Native grassland (520) 0.04 0.67 0.04 Pine-oak (395) 0.02 0.92 400 THE JOURNAL OF ARACHNOLOGY 0 1 2 3 4 5 6 7 8 9 10 11 12 13 Spruce-fir (1830) 2 - r High grass bald (1755) 4 - 1 i Spruce (1715) 2 - 1 r Beech gap (1645) 2 - 1 Northern hardwood (1615) 3 “ yg.g.g.- 4- Red oak (1555) 2 - j Low grass bald (1505) 3 “ Heath bald (1390) 5 “ Mixed oak (1115) 3 - □ alabamense 1 3 frondeum Table Mountain pine (1005) 4 - S albidum B slaucescens Hemlock-hardwood cove (945) 5 - FI Hemlock (885) 5 - LJ aiirciiitius lA lyncuffi Hardwood cove (740) 8 - cheimatos H miirarium Wetland (Indian Cr.) (685) 5 - m crispuliis @ neshamini Wetland (Meadow Br.) (535) 5 - [1 differens S peniisylvanicum Native grassland (520) 3 - Pine-oak (395) 2 - Jlavoiiotatiun r= 1 • 'I =1 S sexpunctatus 1 1 1 ~r 1 0 1 2 3 4 5 6 7 8 9 10 11 12 13 Relative abundance (mean number of individuals per 1-hour sample) Figure 1 . — Stack-bar diagram showing relative abundance of Theridion, Riigathodes and Wamba species at 17 focal sites representing 16 major habitats in the Great Smoky Mountains National Park. Focal sites are listed by habitat in order from highest to lowest elevation (in m) and number of species found at that site are given after elevation. Table 2. — Species richness and diversity values for assemblages of Theridion, Riigathodes and Wamba species at 17 focal sites representing 16 major habitats in the Great Smoky Mountains National Park. Habitat and elevation of focal site (in m) No. of samples Observed richness Diversity index (Shannon) Evenness index (Pielou) Spruce-fir (1830) 24 2 0.25 0.36 High grass bald (1755) 24 4 0.28 0.20 Spruce (1715) 24 2 0.07 0.10 Beech gap (1645) 24 2 0.30 0.44 Northern hardwood (1615) 44 3 0.04 0.03 Red oak (1555) 48 2 0.06 0.09 Low grass bald (1505) 72 3 0.36 0.32 Heath bald (1390) 72 5 1.18 0.73 Mixed oak (1115) 45 3 0.70 0.64 Table Mountain pine (1005) 33 4 0.93 0.67 Hemlock-hardwood cove (945) 48 4 1.42 0.88 Hemlock (885) 48 5 1.40 0.87 Hardwood cove (740) 56 8 1.47 0.71 Wetland (Indian Cr.) (685) 17 5 1.28 0.80 Wetland (Meadow Br.) (535) 17 5 1.10 0.68 Native grassland (520) 24 3 0.64 0.58 Pine-oak (395) 48 2 0.11 0.16 STILES & COYLE— BIOLOGY OE THERIDION, RUGATHODES, AND WAMBA 401 The only middle to low elevation site with diversity and evenness values as low as those of the high elevation sites was the pine-oak site. The hardwood cove site appears to pos- sess the greatest number of species (8). Both wetland sites had the same five species, and at each of these two sites T. albidum and T. flavonotatum were more common than the other three species and about equally abun- dant. Theridion frondeum was found in more habitats (10 of 16) than any other species, and was especially abundant in the high elevation hardwood communities and the low grass bald, which is surrounded by high elevation hardwood (Table 1, Fig. 1). Theridion albi- dum, the sister species of T. frondeum, was found in almost as many habitats (8) but was virtually absent from high elevation habitats. Both of these species occur over a wider el- evation range (1220 m) than any other spe- cies. Rugathodes aurantius is restricted to, and its sister species, R. sexpunctatus, is most abundant in, high elevation communities, but wherever one of these species is abundant or common, the other is uncommon (relative abundance < 0.5) or absent. Theridion lyri- cum Walckenaer 1841 was found in nine hab- itats, including a wide range of middle to low elevation communities, but appears to be most common in dry, pine-dominated, forests. Three of the less common species appear to be associated primarily with a single com- munity type: T. flavonotatum Becker 1879 with wetland, T. differens Emerton 1882 with Table Mountain pine, and T. cheimatos Gertsch and Archer 1942 with native grass- land. Six species {T. alabamense Gertsch and Archer 1942, T. glaucescens Becker 1879, Wamba crispulus (Simon 1895), T. murarium Emerton 1882, T. neshamini Levi 1957, and r. pennsylvanicum Emerton 1913) were un- common wherever they occurred and, with one exception {Wamba crispulus), were found in only one or two habitats. Microhabitat distribution. — Eour species, T. albidum, T differens, T. frondeum, and R. sexpunctatus, were equally common below and above knee level (Fig. 2). Theridion chei- matos was found only below knee level. Ru- gathodes aurantius was more common below than above knee level, but the difference is not significant {P = 0.12). Theridion lyricum was significantly more common above than 0.5 1.5 albidum aurantius - cheimatos differens flavonotatum frondeum lyricum sexpunctatus El AERIAL ■ GROUND 0 0.5 1 1.5 2 Relative abundance (mean number of individuals per 1-hour sample) Figure 2. — Vertical microhabitat distribution of Theridion and Rugathodes species as indicated by relative abundance in aerial and ground samples from those sites where the species has been col- lected. Only those species found in ground and/or aerial samples at two or more sites are included. Aerial samples from 50 to 220 cm above ground; ground samples below 50 cm. Large asterisks in- dicate significant differences {P < 0.05), small as- terisks differences at 0.05 > P < 0.13. below knee level; T. flavonotatum exhibits the same pattern, but the difference is not signif- icant {P = 0.08). Theridion flavonotatum webs were common in the tops of relatively tall leafless stalks of dead herbaceous plants at both wetland sites. Nearly all R. aurantius webs were found on the undersides of the leaves of low herbs in clearings (grass balds, trailsides, and areas of sparse canopy in beech gap forest). Rugathodes sexpunctatus was col- lected primarily by beating low branches of young fir or ferns in spruce-fir and spruce for- est. All T. cheimatos specimens collected at the native grassland site were found on the ground near a drainage ditch. Theridion glau- cescens appeared only in beat samples. The single specimen of T. alabamense was col- lected in a leaf litter sample. Life history. — There are five instars in the life cycle of R. aurantius (Fig. 3). Instar I, which is confined to the egg sac, lacks eyes, pigment, spigots, and visible hairs (at lOOX magnification). Instar II has eyes, pigment around the eyes, functional spinnerets, and many fully developed hairs. It is also the ac- tive spiderling instar that emerges from the egg sac; this was confirmed by examining and measuring newly emerged spiderlings reared in the laboratory and by observing that instar II spiderlings inside the sac have the same ITL Number of individuals 402 THE JOURNAL OF ARACHNOLOGY 1 2 8 4 0 2 0 Beech gap forest - 14 June adult females penultimate female k.- adult males - p, . 1 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 Beech gap forest - 15 Aug 1 2 8 4 0 2 0 instar III females adult females instar III males 1 1 ^ 1 ^ 1 1 I I ^ 1 1 ' i 1 1 ^ 1 ' 1 1 1 i 1 ' 1 ■ 1 . i ' 1 > 1 r 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 High grass bald - 27 June 4 H 2 0 adult females 1 H 0 - 0 adult males 0 oV o'. 2 o'. 3 o'.4 o'. 5 o'.6 o'.7 0.8 0.9 I'.O I'.l l'.2 l'.3 1.4 1.5 1.6 1.7 1.8 1 6 8 0 2 0 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 Mt. Buckley - 14 Sept ’2 i I ^ 8 - instar II I mstar III females i 4 - h adult females 0 - ' I ' I W I ■ I ' I I ' I ' I I ' ^ 4 ^ ^^star III males ^ 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 Mt. Buckley - 14 Sept instar II 1 M 1 instar III females ■ adult females 1 1 ^ [— i — , — ^ ^ ^ ^ — m— 1 ' 1 ' 1 ' 1 ' 1 ' i ’ 1 ' 1 ^ 1 ■ instar III males 1 — ' — I — ' — ^ — ' — I — ' — I — ' — ^ ^ — I — ^ — I — ^ — i — ' — r High grass bald - 22 Sept : J 1 adult females _ jm n- _ ■ --^--1— -'-I j -L 1 1 1 1 [ 1 1 1 ^l^nstar III males 1 ' — 1 — , 'n " 1 — ^ — 1 — 1 — 1 — . — 1 — .— ] — , — -1 — 1 1 — 1 — 1 — ^ — 1 — ^ 1 — 1 — 1 — 1 — ^ ^ 1 — 1 — 1 — 1 — Instar summary - all data from above II III females penultimate females 40 30 20 1 0 0 adult females li penultimate female (reared) 1 instar III males adult males X penultimate male (reared) 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 Tibia I length (mm) Figure 3. — Size (ITL) frequency distribution histograms for five samples of Rugathodes aurantius from three sites. For each sample, females and individuals too young to be sexed are graphed separately from males. All five samples were pooled to generate the instai' summai'y histograms at the bottom. Arrows in instai* summary histogram mark ITL values of the penultimate instar exuvia of two specimens reaied to adulthood in captivity. STILES & COYLE— BIOLOGY OF THERIDION, RUGATHODES, AND WAMBA 403 Rugathodes aurantius months Figure 4. — Postulated phenologies for two Ru- gathodes and one Theridion species in the Great Smoky Mountains National Park. Based on size fre= quency distribution analyses and additional obser- vations. as the smallest solitary (post-dispersal) instar collected in the field. Instar III is the antepen- ultimate instar. This was confirmed by rearing an antepenultimate male and an antepenulti- mate female collected at Mt. Buckley to the adult instar and measuring the exuvia. The palpal tarsi of antepenultimate males were sig- nificantly (P < 0.0001) wider (PTW = 0.1 1 1- 0,130, mean and SD = 0.117 ± 0.008, n = 15) than those of same-aged females (0.065- 0.074, 0.067 ± 0.004, n = 15). Instar IV is the penultimate instar; this was confirmed by rearing the two aforementioned Mt. Buckley spiders and by finding one female of that size class at the beech gap site with an epigynum visible through her soon-to-be-molted cuticle. Instar V is the adult instar. The life cycle pat- tern most consistent with these data (Fig. 3) is that of one generation per year, with most individuals over-wintering in the antepenulti- mate instar and males and females maturing and mating in May and June (Fig. 4). This hypothesis was further supported by finding 15 individuals — all antepenultimate males and females — in leaf litter samples collected under the snow at the beech gap site on 12 February 1997. The absence of adult males from all summer and fall collections made after 27 June suggests that they die soon after mating. Many of the females collected at the beech gap site on 9 July and 15 August and at Mt. Buckley on 14 September were guarding sin- gle egg sacs. Sexual dimorphism in ITL first appears in instar III and is much greater in subsequent instars (Fig. 3). We found female- biased sex ratios (females/males) in every sample of instar III (6.0 at the beech gap on August 15 (« = 35), 1.5 at the same site on 12 February (n = 15), 6.0 at the high grass bald on 22 September (n = 57), and 2.1 at Mt. Buckley on 14 September (n = 40)), There are five instars in the life cycle of R. sexpunctatus (Fig. 5). Based on observations of its sister species, R, aurantius, we presume that instar II of R. sexpunctatus is the active instar that emerges from the egg sac. The males of instars III and IV can be distin- guished from each other and from females by distinctive widths of the palpal tarsi. lestar V is the adult instar. The life cycle pattern (Fig, 4) most consistent with the data is that of a single annual generation that over-winters chiefly in the penultimate iostar, as indicated by the relatively large number of penultimate spiders collected in September. A collection of six individuals (four penultimate females, one penultimate male, and one instar III female) at Mt. Buckley on 25 September also supports this phenology. Males and females apparently mature and mate in May and June, Adult males are absent from late summer collec- tions, suggesting that they die soon after mat- ing. Some adult females persist until at least mid-September, Sexual dimorphism in ITL first appears in instar III, is much greater in the following instars, and appears to be even more pronounced in R, sexpunctatus than in R. aurantius (Fig. 5). The sex ratios (females/ males) in instars III and IV of the total R. sexpunctatus sample are 1.6 {n ~ 29) and 1.1 (w = 30) respectively. There are five or six instars in the life cycle of T. frondeum (Figs. 6, 7). Development in the egg sac includes the same two iestars as in R. aurantius. Instar II spiderlings in the sac have the same ITL as the smallest solitary (post-dispersal) field-collected iestar; this shows that instar II emerges from the egg sac. Only in the penultimate instar do males have distinctively wider palpal tarsi than females. Number of individuals 404 THE JOURNAL OF ARACHNOLOGY Instar summary - all data from above instar III instar II females penultimate females adult females 4 d 0 instar III males penultimate males adult males 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5 1.6 Tibia I length (mm) Figure 5. — Size (ITL) frequency distribution histograms for six samples of Rugathodes sexpunctatus from three sites. For each sample, females and individuals too young to be sexed are graphed separately from males. All six samples were pooled to generate the instar summary histograms at the bottom. Number of individuals STILES & COYLE— BIOLOGY OE THERIDION, RUGATHODES, AND WAMBA 405 Low grass bald - 5 June 4- 2 - 0 4 2 0 instar III Penultimate females ■ u ■ penultimate males 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0 3.2 3.4 3.6 Low grass bald - 30 Sept instar III J instar II 4 penultimate female adult female 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0 3.2 3.4 3.6 Northern hardwood - 12 June 20 16 12 8 4 0 16 12 8- 41 0 instar III A penultimate females adult female -L I penull penultimate males L adult males *lpl-,L,Jal ,■ 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0 3.2 3.4 3.6 Northern hardwood - 14 Aug 2l 0 instar II adult females 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0 3.2 3.4 3.6 Hardwood cove - 18 June : penultimate females adult females. 1 - ■ 1 1 — , — 1 — ^ — 1 — , — 1 — , — 1 — ^ — 1 — ; — pj, — 1 — ^ — 1 — , — 1 — P — 1 — -T-~| , 1 1 [ adult males L ■ ■ ■ 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0 3.2 3.4 3.6 Hardwood cove - 24 Aug jU*. Instar III adult female 2 ^ instar JI 0 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0 3.2 3.4 3.6 Red oak - 24 June penultimate females J ^ ■■ .iiil ii adult females 1 ILl.i. j 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 J penult, males — , l| ^ adult males i “Mi 1 ' 1 ' 1 ' 1 ' iljifia, , ,1 : 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0 3.2 3.4 3.6 Red oak - 31 Aug ■ instar II 1 instar III adult females 1 - , , I, ' 1 1 ■. 1 ■ 1 . - 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0 3.2 3.4 3.6 Tibia I length (mm) Figure 6. — Size (ITL) frequency distribution histograms for eight samples of Theridion frondeum from four sites. For each sample, females and individuals too young to be sexed are graphed separately from males. 406 THE JOURNAL OF ARACHNOLOGY Instar Instar II III (& IV?) penultimate females adult females adult males 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0 3.2 3.4 3.6 Tibia I length (mm) Figure 7. — Instar summary histogram for Theridion froudeum. Generated by pooling all data in Fig. 6. An epigynum was visible through the soon- to-be-molted cuticle of several penultimate fe- males. There are two lines of evidence sug- gesting that the size class labeled instar III in Fig. 7 may actually include instar IV individ- uals as well. This size class is broad and skewed strongly to the left. And the seasonal histograms (Fig. 6) show that many of the in- dividuals that comprise the low right shoulder of this size class were collected in early June (Fig. 6), suggesting that a mid-May collection may have contained a large proportion of in- dividuals with ITL values between 0.5 and 0.7 mm. Such a class, if present, would have to be instar IV, and would mean that at least many adults of T. frondeum are in their sixth instar. The life cycle pattern most consistent with the data is that of one generation per year, with over-wintering primarily in the an- tepenultimate instar (Fig. 4). Males and fe- males apparently mature and mate in June and July. Since males at each site in June exhibit a higher ratio of adults to juveniles than do females, it is clear that males tend to mature before females. The absence of adult males from late summer collections suggests that they die off soon after mating. Sexual dimor- phism in ITL is apparent in penultimate and adult instars. The sex ratio (females/maies) in the total 7. frondeum penultimate instar sam- ple is 1.9 {n = 362) (Fig. 7). Other natural history observations. — Ru- gathodes aurantius .—Web placement and structure: Rugathodes aurantius webs are vir- tually restricted to the herb layer. Most adult female webs were found on the broad-leafed herbs Solidago glome rata Michaux, Angellica sp., Agertina altissima King & Robinson van roenesis, and Aster divaricatus L. All of these species were common at the beech gap forest site wherever tree gaps permitted the devel- opment of a rich herb layer. Most R. aurantius collected in the high grass bald were found in dense patches of S. glome rata. At Mt. Buck- ley (1998 m) webs were found almost exclu- sively in a small clearing dominated by S. glomerata (11 adult females and 40 juveniles were collected in a 1-hour ground sample) and were virtually absent from the adjacent spruce-fir forest (only one adult female was collected — from ferns — in 3 hours of ground collecting and beating). Adult female webs are primarily confined to the underside of a single leaf or leaflet. Roughly parallel strands of silk attached to the leaf edges bend or fold the leaf, and the bulk of the web is an asymmetrical cobweb sheltered within the resultant concav- ity. The spider typically rests near the center of the web on or within a few mm of the re- treat leaf’s ventral surface. Three types of leaf modification were observed in detail at the beech gap site on 9 July (n = 9 webs). The first type was characterized by strands of silk running from the tip to the base of the leaf and bending the leaf tip and adjacent edges downward and toward the leaf base, the sec- STILES & COYLE— BIOLOGY OF THERIDION, RUGATHODES, AND WAMBA 407 ond by lines pulling the lateral edges of the leaf to within 5 cm of each other to form a length-wise fold, and the third by lines con- necting a lateral edge fold near the base of the leaf to the tip fold, forming a cone-shaped re- treat. Juvenile spiders do not appear to modify leaf shape; those collected at Mt. Buckley on 14 September were found, like adult females, on the undersides of S. glomerata leaves, but their webs were smaller and positioned in the natural concavity of the leaf undersurface. Diet: Prey items found in the webs of two adult female and two instar II spiders at Mt. Buckley included three Homoptera (two green leaf hoppers [Cicadellidae] 3.5 and 3.4 mm long and one other homopteran 1,3 mm long), three small muscoid Diptera (2. 1-3.5 mm long), and two other small unidentified winged insects. Beside the trail to the high grass bald site, the web of one female con- tained her recently emerged instar II spider- lings and the exoskeletal remains of four midges (2. 2-2. 5 mm long) and one small (8.5 mm long) cranefly. Reproduction and brood care: Rugathodes aurantius females were found in nature guard- ing egg sacs close to the undersurface of their retreat leaf from early July to mid September. No female was observed with more than one egg sac. The spherical white egg sac is com- posed of a single fairly dense layer of kinky/ looped threads. The diameters of 5-10 R. au- rantius eggs in each of nine clutches (egg sacs) ranged from 0.48-0.59 mm (mean and SD = 0.54 ± 0.03 mm). Clutch sizes at two sites (beech gap forest and Mt. Buckley clear- ing) ranged from 3-47 (24.6 ± 13.7, n = 30). Clutch size was significantly larger {P < .0001) at the beech gap forest site on 9 July (11-47, 35.5 ± 9.7, n = 15) than on 15 Au- gust (3-23, 12.2 ± 6.2, n = 11). Clutch size of the 14 September Mt. Buckley sample was also significantly lower (14-24, 17.8 ± 4.3, n = A, P = .003) than that of the 9 July beech gap sample. Females collected with egg sacs at the beech gap site on 9 July were signifi- cantly larger {P = .0001) (ITL = 1.18-1.39, 1.30 ± 0.05, n = 14) than those collected on 15 August (1.15-1.30, 1.22 ± 0.04, n = 11). When disturbed, females with egg sacs typ- ically maintained contact with the sac and of- ten moved it. At least one of the fourth legs was used to position and move the sac, which, in a few cases, was clearly seen to be attached to the spinnerets. The time between oviposi- tion and emergence of instar II spiderlings from the sac was 13 days for the only brood oviposited in captivity. Spiderlings emerged from the eight field-collected egg sacs be- tween 2 and 9 days (mean and SD = 5.9 ± 2.3) after they were collected, also suggesting that the normal period of development in the egg sac is about 2 weeks or less. Several hours before spiderling emergence, the female re- peatedly and vigorously bit at the egg sac, pulling and stretching the silk to create a hole through which instar II spiderlings soon began to emerge. In the field, females were com- monly found with emerged instar II spider- lings, and in one case the remains of several Diptera were present in such a web, but we saw no evidence of communal feeding. One captive female captured Drosophila flies and placed them near her recently emerged instar II spiderlings, which appeared to increase in size (abdominal volume) over the course of three weeks; however we never actually ob- served the spiderlings feeding. Theridion frondeum. — Retreat placement and structure: Adult females were found in- side partly folded living leaf retreats from 30 to 220 cm above ground in a great variety of plants, both herbaceous (stinging nettle, ferns, blackberries, etc.) and woody (striped maple, sugar maple, etc.). Part of the leaf is folded downward longitudinally, transversely, or di- agonally. Some, for example, were folded downward sharply near the middle at one side with the opposing edges on that side fastened together with silk to make a roughly cone- shaped retreat. Diet: A wrapped and partly consumed 5.4 mm long adult female Pityohyphantes costa- tus (Hentz 1850) spider was found in a web occupied by a 3.3 mm long T. frondeum fe- male and her emerged instar II spiderlings. Reproduction and brood care: Theridion frondeum females were found in leaf retreats with egg sacs from early July to late August. No female was observed with more than one egg sac. The spherical white egg sac is com- posed of a single fairly dense layer of kinky/ looped threads. The diameters of ten eggs in each of two clutches ranged from 0.67 to 0.78 mm (mean and SD = 0.72 ± 0.03 mm). Clutch size ranged from 13-40 (26.8 ± 11.5, n = 5). The time from oviposition to the emergence of instar II spiderlings from the 408 THE JOURNAL OF ARACHNOLOGY egg sac took 1 3 days for the only viable brood oviposited in captivity. Two females were ob- served vigorously biting their egg sacs prior to spiderling emergence. In the summer spi- derlings are often found with the mother in her web, suggesting that they remain there for at least a few days before dispersing. Theridion albidum. — One female was found at the hardwood cove site on 24 August in a partly folded leaf retreat at knee level guarding an egg sac containing ten spiderlings that emerged from the sac as instar II spider- lings about 4 days later in captivity. Theridion differens, — One female was found at the Table Mountain pine site on 6 August 1997 guarding her pale grey-brown egg sac attached to her small conical silk re- treat in the junction of a leaf petiole and twig on a mountain laurel branch at about head height. She continued to cling tightly to her egg sac after being transported to the lab in a glass vial. Her entire clutch of 25 instar II spi- derlings emerged from the sac by the follow- ing morning. Theridion lyricum. — One female was col- lected at the hardwood cove site on 25 August with her egg sac containing 82 instar I spi- derlings. The pale grey-brown sac was com- posed of kinky/looped strands of silk, some of which were brown, not white. DISCUSSION Habitat and microhabitat distribution patterns. — The differences among the 16 sampled habitats (and the great similarity of the two widely separated wetland sites) in the kinds and relative abundances of Theridion and Rugathodes species (Fig. 1) suggest that these species may be good predictors of hab- itat. The presence of more of these species in the habitat (hardwood cove forest) celebrated for its high plant diversity (Whittaker 1956), than in some of the habitats (spruce fir, spruce, beech gap, and pine-oak forests) characterized by relatively low plant diversity (Whittaker 1956), suggests a positive correlation between the species richness of this set of spider spe- cies and plant species richness. However, the finding of only three of these spider species in the plant-rich (175 species (Stratton & White 1982)) low grass bald and five in the plant-poor (12 species (Cain 1930)) heath bald completely negates this correlation. Evidently, the number of Theridion and Rugathodes spe- cies present at a site is not a reliable predictor of plant species richness. It is of interest that the species found in the greatest number of habitats and over the great- est range of elevations {T. frondeum, T. albi- dum, R. sexpunctatus, T. lyricum, and T. dif- ferens) all have relatively wide geographic and latidudinal ranges (Levi 1957), while some of the species found in only one or two habitats {T. alabamense, T cheimatos, T. nes- hamini, and T. pennsylvanicum) have much smaller ranges (Levi 1957). This tendency for habitat specialists to occupy relatively small geographic ranges has been observed in many taxa (Stevens 1989), including species of Te- tragnatha, Neriene, and Araneus spiders liv- ing in the GSMNP (Aiken & Coyle 2000; Wright & Coyle 2000; Davis & Coyle 2001). Our observation that the sister species T. frondeum and T. albidum are often found at the same sites and that T. frondeum is gener- ally more common than T. albidum is consis- tent with Levi’s (1957) and Kaston’s (1981) observations. The much greater abundance of T. frondeum at high elevation sites is consis- tent with the geographic ranges of the two species; T. frondeum is more common and widespread at higher latitudes (Levi 1957), and is therefore probably better adapted to cold climates, than is T. albidum. Rugathodes aurantius and R. sexpunctatus, another pair of sister species (Levi 1957; 1. Agnarsson pers. comm.), are most common in high elevation communities — which is consistent with their basically boreal geographic ranges (Levi 1957) — but they exhibit striking habitat seg- regation. The distinctively different web placement substrates of the two species sug- gests that this segregation is based upon dif- ferent microhabitat requirements. Rugathodes aurantius builds its webs on the undersides of broad-leaved herbs, which are rare on the heavily shaded ground of spruce-fir and spruce forests, whereas R. sexpunctatus typi- cally lives on the foliage of young fir trees, which are rare or absent in the high grass bald and beech gap forest communities. The re- striction of R. aurantius to high elevation hab- itats may be the result of climatic require- ments, since an ample broad-leaved herb substratum is present at some of the other for- est sites (especially northern hardwood) where R. aurantius is extremely rare or absent. Since R. sexpunctatus is found at some of our mid- STILES & COYLE— BIOLOGY OF THERIDION, RUGATHODES, AND WAMBA 409 die elevation sites, it seems to be less narrow- ly restricted to boreal climates. Our data sug- gest that it may be restricted to fine-needle conifers, which is consistent with Levi’s (1957) observation that it is “usually found on coniferous trees.” Our observations of T. cheimatos and T. neshamini are consistent with Chamberlin and Ivie’s (1944) note that the former species was collected in moderate- ly damp places on the ground and with Levi’s (1957) observation that the latter species is associated with tall grass. Life history. — Except for reports that T. frondeum matures in late June or July and that adult females produce eggs in July and sur- vive through September (Emerton 1902; Com- stock 1948; Kaston 1981), there are no pub- lished descriptions of the life cycles of the three species we have analyzed {R. aurantius, R. sexpunctatus, and T. frondeum). Based on our observations and those of Toft (1976) on several Theridion species living in a Danish beech forest, we postulate that in most Ther- idion and Rugathodes species the life cycle contains five or six instars, with instar II emerging from the egg sac. Like the three spe- cies we studied, four of the six Theridion spe- cies whose life cycles Toft resolved have an- nual (one-year) life cycles and overwinter in the antepenultimate or penultimate instars. The three species we studied differ from one another in two aspects of their phenologies (Fig. 4): R. aurantius overwinters in the an- tepenultimate instar; R. sexpunctatus overwin- ters in the penultimate instar; and R. aurantius and R. sexpunctatus mature a few weeks and possibly one instar earlier than does T. fron- deum. The female-biased sex ratios we observed in antepenultimate and penultimate instars of these three species may be artifacts of sam- pling error or may be real. The strongly fe- male-biased ratios in the adult samples of all three species could be the result of the earlier maturation and/or shorter longevity of males, but the basis for bias in earlier instars is not so evident. Group living, which should favor selection of female sex-biasing mechanisms in social spiders (Aviles 1986, 1993), does not exist in these species. Web placement and stracture. — The di- versity of leaf retreat architectures engineered by adult females of R. aurantius and T. fron- deum is probably the result of a flexible re- sponse to variation in leaf form. We hypoth- esize that leaf retreat construction is an adaptation to protect the spider and her brood from rain, intense sunlight, and/or visual pred- ators. Diet. — The few prey items collected from R. aurantius webs suggest that small flying insects are a significant part of their diet. The data also show that both R. aurantius and T. frondeum, like other Theridion species (Bris- towe 1958), can capture prey considerably larger than themselves. Reproduction and brood care. — Given that T. frondeum females with egg sacs were significantly {P < .001) larger (mean carapace width = 1.12 mm ± 0.04; n = 5) than those of R. aurantius (0.85 ± 0.19; n = 5), it is not particularly surprising that the eggs in the two observed egg-stage clutches of T. frondeum were significantly larger than those of R. au- rantius {P < .001). In the light of Marshall & Gittleman’s (1994) findings that clutch size in- creases with body size across spider taxa, it is interesting that the observed clutch sizes of these two species did not differ. It is also note- worthy that the only clutch of T. lyricum we observed (carapace width of the mother = 0.96 mm) was nearly twice as large as the largest clutches of the other species. Toft (1976) found that females of some Theridion species produce two or more (up to five) egg sacs over the course of the summer. He postulated that this ability to produce mul- tiple clutches is a widespread fecundity-en- hancing trait of small spiders that compen- sates for a relatively small number of eggs per egg sac. Two of our observations strongly suggest that R. aurantius females also produce multiple clutches: 1) the single annual cohort of adult females is found in nature guarding egg sacs throughout a 2.5 month period, even though the interval from oviposition to spi- derling emergence from the egg sac is about 2 weeks; 2) there was a significant decrease in the beech gap forest site population’s mean clutch size between 9 July and 15 August, a pattern commonly observed in multi-clutch spiders (Toft 1976; Marshall & Gittleman 1994). Nevertheless, we cannot rule out the possibility that R. aurantius females produce only one clutch and that the observed decrease in clutch size is the result of delayed ovipo- sition by smaller females. This second hy- pothesis is consistent with the observation that 410 THE JOURNAL OF ARACHNOLOGY adult females in the 15 August beech gap sample had a significantly smaller mean ITL value than those of the 9 July sample. Our observations of R. aurantius and T. frondeum indicate that females devote consid- erable effort to brood care. Bent-leaf retreat construction and egg sac guarding have been observed in other Theridion species (for ex- ample Archer (1947) and Bristowe (1958)), and Emerton (in McCook 1890) described an unidentified Theridion female assisting her spiderlings in exiting the egg sac with the same behaviors we observed in R. aurantius and T. frondeum. Our observations that instar II spiderlings of these two species remain with the mother in her retreat for at least a few days after emerging from the egg sac suggests the interesting possibility of nutritional parental care like that observed in some European and Central American Theridion species (Nielsen 1932; Bristowe 1958; Kullman 1972; and au- thors’ personal observations in Costa Rica). We have observed no compelling evidence of such care, but the possibility deserves closer scrutiny, ACKNOWLEDGMENTS We thank Robert Edwards, Ricky Wright, Doug Toti, and Jeremy Miller for helping to sample and process the spiders used in this study. Keith Langdon provided logistic sup- port. Dan Pittillo identified the plants used by R. aurantius. Ingi Agnarsson and Miguel Ar- nedo kindly provided information from their cun'ent research on the systematics of Theri- dion and its relatives. Jim Costa, Roger Dumb, Keith Langdon, Marie Aiken, Melinda Davis, Ian Stocks, Rosemary Gillespie, and an anon- ymous reviewer made helpful comments on drafts of this paper. This research was sup- ported by National Science Foundation (DEB- 9626734) and National Park Service grants to FAC, and by a Western Carolina University Undergraduate Research Grant to GJS. LITERATURE CITED Aiken, M. & F.A. Coyle. 2000. Habitat distribution, life history and behavior of Tetragnatha spider species in the Great Smoky Mountains National Park. Journal of Arachnology, 28:97-106. Archer, A.E 1947. The Theridiidae or comb-footed spiders of Alabama. State Geological and Natural History Survey of Alabama. Museum Paper 22. Archer, A.E 1950. A study of theridiid and mimetid spiders with descriptions of new genera and spe- cies. Museum Papers of the Alabama Museum of Natural History, 30:1-40. Aviles, L. 1986. Sex-ratio bias and possible group selection in the social spider Anelosimus eximius. American Naturalist 128(1): 1-12. Aviles, L. 1993. Interdemic selection and the sex ratio:a social spider perspective. American Nat- uralist 142(2):320-245. Bracher, G.A. & J.R. Bider. 1982. Changes in ter- restrial animal activity of a forest community af- ter an application of aminocarb (Matacil). Ca- nadian Journal of Zoology 60:1981-1998. Bristowe, W.S. 1958. The World of Spiders. Col- lins, London. 304 pp. Cain, S.A. 1930. An ecological study of the heath balds of the Great Smoky Mountains. Butler Uni- versity Botanical Studies 1:177-208. Carter, RE. & A.L. Rypstra. 1995. Top-down effects in soybean agroecosystems: spider density af- fects herbivore damage. Oikos 72:433-439. Chamberlin, R.V. & W. Ivie. 1944. Spiders of the Georgia region of North America. Bulletin of the University of Utah 35(9): 1-267. Christenson, T.E., J. Cohn, & M.J. Pokora. 1990. Spiders as environmental monitors of heavy met- als. The Toxicologist 10:249. Coddington, J.A. & H.W. Levi. 1991. Systematics and evolution of spiders (Araneae). Annual Re- views of Ecology and Systematics 22:565-592. Coddington, J.A., L.H. Young, & F.A. Coyle. 1996. Estimating spider species richness in a southern Appalachian cove hardwood forest. Journal of Arachnology 24:1 1 1-128. Colwell, R.K. & J.A. Coddington. 1994. Estimating terrestrial biodiversity through extrapolation. Philosophical Transactions of the Royal Society of London B 345:101-118. Comstock, J.H. 1948. The Spider Book. Cornell University Press, Ithaca, New York. 729 pp. Coyle, F.A. 1981. Effects of clearcutting on the spi- der community of a southern Appalachian forest. Journal of Arachnology 9:285-298. Davis, M.J. & F.A. Coyle. 2001. Habitat distribu- tion and life history of Araneus spider species in the Great Smoky Mountains National Park. Bul- letin of the British Arachnological Society 12(2): 49-57. Duffy, E. 1978. Ecological strategies in spiders in- cluding some characteristics of species in pioneer and mature habitats. Symposium of the Zoolog- ical Society of London (42): 109-123. Emerton, J.H. 1902. The Common Spiders of the United States. Ginn and Company. 227 pp. Hanggi, A., E. Stockli, & W. Nentwig. 1995. Hab- itats of central European spiders. Miscellanea Faunistica Helvetiae, No. 4. 459 pp. Kaston, B.J. 1981. Spiders of Connecticut. Rev. ed. State Geological and Natural History Survey of Connecticut, Bulletin No. 70. STILES & COYLE— BIOLOGY OF THERIDION, RUGATHODES, AND WAMBA 411 Kremen, C., R.K. Colwell, TL. Erwin, D.D, Mur- phy, R.F. Noss, and M.A. Sanjayan. 1993. Ter- restrial arthropod assemblages: their use in con- servation planning. Conservation Biology 7(4): 796-808. Kullmann, E.J. 1972. Evolution of social behavior in spiders (Araneae; Eresidae and Theridiidae). American Zoologist 12:419-426. Levi, H.W. 1957. The spider genera Enoplognatha, Theridiofiy and Paidisca in America north of Mexico. (Araneae, Theridiidae). Bulletin of the American Museum of Natural History 112:1- 123. Levi, H.W. 1959. The spider genera Achaearanea, Theridion and Sphyrotinus from Mexico, Central America and the West Indies (Araneae, Theridi- idae). Bulletin of the Museum of Comparative Zoology, 121(3):57-163. Levi, H.W. 1963. American spiders of the genus Theridion (Araneae, Theridiidae). Bulletin of the Museum of Comparative Zoology 129(10):481- 589. Magurran, A.E. 1988. Ecological Diversity and Its Measurement. Princeton University Press, Princeton, New Jersey. 179 pp. Marshall, S.D. & J.L. Gittleman. 1994. Clutch size in spiders: is more better? Functional Ecology 8: 118-124. McCook, H.C. 1890. American Spiders and their Spinning Work. Vol. 2. Allen, Lane and Scott. 479 pp. Nielsen, E. 1932. The Biology of Spiders. Levin and Munksgaard, Copenhagen. 248 pp. Norris, K.C. 1999. Quantifying change through time in spider assemblages: sampling methods, indices, and sources of error. Journal of Insect Conservation 3:1-17. Noss, R.F. 1990. Indicators for monitoring biodi- versity: a hierarchical approach. Conservation Biology 4(4):355-364. Platnick, N.I. 1997. Advances in Spider Taxonomy 1992-1995. New York Entomological Society, New York, 976 pp. Polis, G.A., R.D. Holt, B.A. Menge, and K.O. Wi- nemiller. 1996. Time, space, and life history: in- fluences on food webs. Pp. 435-461, In Food Webs: Integration of Patterns and Dynamics. (G.A. Polis & K.O. Winemiller, eds.). Chapman & Hall. Riechert, S.E. & L. Bishop. 1990. Prey control by an assemblage of generalist predators: spiders in garden test systems. Ecology 71:1441-1450. Roberts, M.J. 1995. Spiders of Britain and Northern Europe. Harper Collies, London. 383 pp. Russell, G.J., J.M. Diamond, S.L. Pimm, & T.M. Reed. 1995. A century of turnover: community dynamics at three timescales. Journal of Animal Ecology 64:628-641. Stratton, D.A. & P.S, White. 1982. Grassy balds of Great Smoky Mountains National Park: Vascular plant floristics, rare plant distributions, and an assessment of the floristic data base. Research/ Resource Management Report SER-58, National Park Service, Natural Science Research Division, Atlanta, Georgia. Stevens, G.C. 1989. The latitudinal gradient in geo- graphical range: how so many species coexist in the tropics, American Naturalist 133(2):240-256. Strong, D.R., D. Simberloff, L.G. Abele & A.B. Thistle (eds.). 1984, Ecological Communities: Conceptual Issues and the Evidence, Princeton University Press, Princeton, New Jersey. Thomas, W.A. (ed.). 1972. Indicators of Environ- m.eetal Quality. Plenum Press. Toft, S. 1976. Life-histories of spiders in a Danish beech wood. Natura Jutlandica 19:5-40. Uetz, G.W. 1979. The influence in variation in litter habitats on spider communities, Oecologia 40: 20-42. Whittaker, R.H. 1956. Vegetation of the Great Smoky Mountains. Ecological Monographs 26(l):l-80. Wilson, E.O. 1992. The Diversity of Life. Harvard University Press, Cambridge, Massachusetts. 424 pp. Wright, R.L. & F.A. Coyle. 2000. Habitat distribu- tion, life history and behaviour of Neriene spe- cies in the Great smoky Mountains national Park (Araneae, Linyphiidae). Bulletin of the British Arachnological Society 1 1(7):293-304. Wunderlich, J. 1987. Die Spinnen der Kanarischen Inseln ued Madeiras: Adaptive Radiation, Bio- geographic, Revisionen und Neubeschreibungen. Triops Verlag, Langen, Germany. Wunderlich, J. 1995. Revision und Neubeschrei- bung einiger Gattungen der Familie Theridiidae aus der Nearktis und Neotropis (Aracheida: Ar- aneae). Beitrag Araneologie 4:609-615. Manuscript received 28 January 2000, revised 1 July 2001. APPENDIX Habitat type, locality data (with UTM coordi- nates), collecting dates, and sample sizes are given for each of the 17 focal sites (listed in order from highest to lowest elevation) and one accessory site. For focal sites, the number of ground, aerial, beat, and sweep samples are given in parentheses after total number of 1-hr samples. Whittaker (1956) pro- vides descriptions of the vegetation of most of these habitats. A bald is a natural treeless community on a well-drained high elevation site below the cli- matic tree-line. A beech gap forest is an orchard- like forest dominated by small gray beech trees (Fagus grandifolia) and is typically located on a 412 THE JOURNAL OF ARACHNOLOGY south-facing slope in a high mountain gap. Hard- wood cove forests are found in sheltered middle elevation sites and are characterized by a high spe- cies diversity of large trees and understory plants. Focal sites. — Spruce-fir forest: North Carolina: Swain County, 0.5 km SW Mt. Collins, N&S sides of Appalachian Trail, E2755, N39403, 1815-1845 m elev., 26 June and 14 Sept. 1996, 24 (8-8-8-0). High grass bald: North Carolina: Swain County, Andrews Bald, E2738, N39354, 1755 m elev., 27 June and 22 Sept. 1996, 24 (10-0-4-10). Spruce forest: North Carolina: Swain County, just SW junction of Noland Divide Trail and road to pumping station, E2755, N39382, 1715 m elev., 20 June and 7 Sept. 1996, 24 (8-8-8-0). Beech gap forest: North Carolina: Swain County, in hog exclosure below Appalachian Trail at 350 m E Road Prong Trailhead, E2786, N39433, 1645 m elev,, 14 June and 15 Aug. 1996, 24 (8-8-8-0). Northern hardwood forest: North Carolina: Hay- wood County, Cataloochee Divide just above Hem- phill Bald Trail at 200 m E Garrett’s Gap, E3055, N39359, 1615 m elev., 12&15 June and 14 Aug. 1996, 44 (15-15-14-0). Red oak forest: North Carolina: Swain County, Roundtop Knob, E of Noland Divide Trail about 2 mi SE Clingman’s Dome Road, E2770, N39364, 1555 m elev., 24 June and 31 Aug. 1996, 48 (16- 16-16-0), Low grass bald: North Carolina: Swain County, Gregory Bald, E240i, N39343, 1505 m elev., 3-5 June and 29-30 Sept. 1995, 72 (24-0-24-24-24). Heath bald: Tennessee: Sevier County, Inspira- tion Point on Alum Cave Trail, E2789, N39461, 1390 m elev., 25-26 May and 23-24 Sept. 1995, 72 (24-24-24-0). Mixed oak forest: Tennessee: Sevier County, E, S, & W slopes of Chinquapin Knob, E2639, N39512, 1083-1 144 m elev., 13 June and 13 Aug. 1996, 45 (15-14-16-0). Table Mountain pine forest: Tennessee: Sevier County, about 200 m N of route 441 loop NW of Chimneys picnic area, E2738, N39471, 976-1037 m elev., 6 June and 6 Aug. 1996, 33 (13-8-12-0), Hemlock-hardwood cove forest: Tennessee: Se- vier County, N&E Grotto Falls Trailhead at Roaring Fork Motor Trail, P. White veg. plot, E2772, N39512, 945 m elev., 22 May and 30 July and 1 Aug. 1996, 48 (16-16-16-0). Hemlock forest: North Carolina: Haywood Coun- ty, Cataloochee, 150 m S mouth of Palmer Branch at Caldwell Fork, E3107, N39436, 854-915 m elev., 4 June and 5 Aug. 1996, 48 (17-14-17-0). Hardwood cove forest: Tennessee: Sevier Coun- ty, along Porter’s Creek Trail at 200 paces above bridge over Porter’s Creek, E2830, N39508, 740 m elev., 18-19 June and 24-25 Aug. 1996, 56 (19- 18-19-0). Wetland (Indian Cr.): North Carolina: Swain County, marsh between Indian Creek Trail and In- dian Creek at 2 mi. NE of junction with Deep Creek Trail, E2817, N39296, 685 m elev., 27 May and 16 Aug. 1996, 17 (7-3-4-3). Wetland (Meadow Br.): Tennessee: Blount Coun- ty, marsh along Meadow Branch at 0.5 km ENE of Dosey Gap, E2527, N39470, 535 m elev., 23 May and 1 Aug. 1996, 17 (7-4-0-6). Native grassland: Tennessee: Blount County, Cades Cove, S side Abrams Creek about 0.3 mi. upstream from Cades Cove Loop Road bridge, E2426, N39423, 520 m elev., 5 June and 8 Aug. 1996, 24 (12-0-0-12). Pine-oak forest: Tennessee: Blount County, 300 m N of junction of Tabcat Creek and Maynard Creek, E2301, N39347, 395 m elev., 28-29 May and 2 Aug. 1996, 48 (16-16-16-0). Accessory site. — Tennessee: Sevier County, small clearing near top of Mt. Buckley and adjacent area of young fir on N slope near top of Mt. Buck- ley, E2720, N39380, 1980-2000 m elev., 14 Sept. 1996. 2001. The Journal of Arachnology 29:413-422 EVIDENCE FOR KIN-STRUCTURED GROUP FOUNDING AND LIMITED JUVENILE DISPERSAL IN THE SUB-SOCIAL SPIDER STEGODYPHUS LINEATUS (ARANEAE, ERESIDAE) Jes Johannesen: Institut ftir Zoologie. Abteilung fiir Okologie, Universitat Mainz, Saarstrasse 21, D-55099 Mainz, Germany, e-mail: Jesjo@oekologie.biologie.uni-mainz.de Yael Lubin: Mitrani Department of Desert Ecology, Blaustein Institute for Desert Research, Ben Gurion University of the Negev, Sede Boqer Campus, 84990 Israel. ABSTRACT. In sub-social spiders, restricted dispersal of young (i.e., natal philopatry) and the potential for inbreeding could contribute to within-population subdivision, thus resulting in a population structure similar to that found in social congeners. In this context, we analyzed the origin and mode of individual distribution patterns and their contribution to within-population structure in juveniles of the sub-social spider Stegodyphus lineatus. We investigated the distribution of juveniles for four months after leaving the maternal nest using allozyme genetic markers. We found that isolated groups of juveniles consisted predominantly of siblings, whereas larger aggregations of individuals showed mixing of different juvenile sibling groups. However, even within such aggregations, sibling groups could be identified. Within the population at large, a heterozygote deficit and an uneven distribution of alleles were found. This was caused by limited movement of juveniles and males away from the natal site. Thus, the within-population (intrademic) structure could be partitioned into two components, resulting from kin-groups and population subdivision into demes. We compare this type of population structure with that found in non-social and social species, and discuss whether it provides conditions that could favor the evolution of sociality. Keywords: Dispersal, sibling groups, allozymes, relatedness, group founding, intrademic structure. Species of the genus Stegodyphus Simon 1873 (Eresidae) experience extended maternal care, in which spiderlings continue to be fed beyond their first instars and remain together after the mother’s death (Schneider 1995). There are three social species (non-territorial permanently social, sensu Aviles 1997) in the genus. The widespread occurrence of extend- ed maternal care and juvenile cohabitation in the sub-social congeners suggests that per- manent sociality evolved in species that were preconditioned for a prolonged phase of tol- erance. This has been termed the sub-social route to sociality (Buskirk 1981). Other eres- ids (e.g., Eresus Walckenaer, 1805, Seothyra Purcell 1903) exhibit sub-social behavior (Kullmann & Zimmermann 1975; Y. Lubin personal observation). High population turnover, inbreeding in closed colonies and colony founding by one or a few related females are demographic characteristics typically associated with soci- ality in spiders (Riechert & Roeloffs 1993; Aviles 1997). Such populations may have low levels of genetic variation overall, and in par- ticular, individuals within groups will be ge- netically similar. The spatial distribution of genetic variation within a population is im- portant, because the evolution of social traits may depend on the degree of genetic similar- ity among interacting individuals. Sub-social species examined so far differ from social species in having greater genetic variability, no permanent group living, and less population structuring (Johannesen et al. 1998; Johannesen & Lubin 1999). The eresid Stegodyphus lineatus Latreille 1817, is con- sidered sub-social. After a phase of living communally in the maternal nest, young S. li- neatus disperse and settle singly. Spiders live one year, and their nests are often found in clusters separated by large distances of simi- lar, but uninhabited habitat (Lubin et al. 1998). Dispersing young initially settle in the vicinity of the maternal web (Lubin et al. 1998) but nest relocation is not uncommon (Ward & Lu- bin 1993). Lubin et al. (1998) suggested that limited movement of young during dispersal and their preference for certain species of shrubs results in clumped distributions. Using 413 414 THE JOURNAL OF ARACHNOLOGY genetic markers, Johannesen & Lubin (1999) showed that S. lineatus form relatively stable clusters in trees and that in such clusters, adult spiders experience limited dispersal. Further- more, these groups showed evidence of being established by single females, a pattern similar to that found in social species. However, such stable environments may be atypical sites for this species, and data on juvenile movements in more typical habitats following the initial natal dispersal phase are still lacking. The aim of the present study was to investi- gate juvenile dispersal of S. lineatus to explain the origin and amount of within-population ge- netic variance and the processes that produce this variance. We investigated the distribution of juveniles 4 mo after dispersal from the maternal nest. We examined whether clusters consisted of sibling groups, and if these remained separated or if they mixed with juveniles of other clusters. Finally, we tested for the occurrence of random mating at the population level. On the basis of a previous study (Johannesen & Lubin 1999) we predicted that clusters are predominately kin- groups. In contrast to the previous study, the present study analyzes a population in a wadi (dry riverbed), which is an unstable environ- ment that is subject to occasional flooding and frequent disturbance from grazing. We ask if a stable environment is required to generate struc- turing within the population or if intrademic structure is an intrinsic part of the life history of S. lineatus in both stable and unstable envi- ronments. METHODS Juvenile spiders were collected from sev- eral clusters in a small wadi near Sede Boker in the Negev Desert, Israel (Fig. 1) on 25-26 October, 1998, about 4 mo after they left the maternal nest. A juvenile cluster was subjec- tively defined: either according to a clustered location in a specific plant or if the cluster was near to an old maternal nest showing signs of reproduction (remains of the consumed moth- er, exuviae of spiderlings). For every cluster, the number of occupied and unoccupied ju- venile nests was noted. Two groups, 24A and 24B, were not discretely clustered but con- sisted of several single nests radiating from the bush-cluster of group 23 (Fig. 1). More than 200 juvenile nests were located: 79 webs were occupied by S. lineatus, four by another species, and 65 webs were empty. The re- maining webs were not checked. The 79 ju- venile spiders belonged to sixteen clusters. Thirteen of these clusters consisted of three or more individuals. The population and family structure was investigated by means of genetic similarity estimates using enzyme electrophoresis. Electrophoresis staining procedures followed Johannesen & Lubin (1999). The enzymes Aat-1, Pep-A and Adh were omitted from the analysis because they did not stain or stained too weakly to be interpreted in all juveniles. These three enzymes stained weakly in adult animals. To improve the resolution of ester- ase alleles, the enzyme was run in Tris-Ma- leate pH = 7.8 instead of Tris-Maleate pH = 7.0. To help investigate for an unequal distribution of alleles within the wadi, it was divided into an upper and lower half. Each wadi-half com- prised about half the individuals. Random mat- ing (Hardy-Weinberg proportions) within the to- tal wadi population and within the upper and lower wadi, respectively, was tested by the Lou- is & Dempster (1987) exact test using the pro- gram GENEPOP (Raymond & Rousset 1995). Genotypic linkage disequilibria were estimated according to Weir (1991). Allele frequency dif- ferences between upper and lower wadi were tested applying a RxC — test (GENEPOP) (Fig. 1). Estimates of genetic differentiation among spider groups were obtained by the F-statistic estimators of Weir & Cockerham (1984), using the program Biosys (Swofford & Selander 1989). Standard deviations were obtained by jackknifing over loci. F-statistics assume that populations (or groups) are defined as breeding units, i.e., the individuals in a group originated from random matings of the previous genera- tion. However, if a cluster consists primarily of siblings, the random mating assumption is vio- lated (Chesser 1991), because by having the same parents, genes of siblings are correlated. Thus, within-population structure may be con- founded owing to sampling both sibling groups and breeding units (multi-parental groups). This may lead to a population subdivision estimate based on variance between families, and not ac- tual population subdivision. To distinguish between genetic structure caused by breeding units and that caused by sibling groups, we used the approach outlined in Johannesen & Lubin (1999), combining a test for group relatedness and heterozygote ex- JOHANNESEN & LUBIN— DISPERSAL OF STEGODYPHUS LINEATUS YOUNG 415 Figure 1. — Sampling location and positions of juvenile Stegodyphiis lineatus in a wadi near Sede Boker, Israel. Filled circles represent sampled clusters. 416 THE JOURNAL OF ARACHNOLOGY cess. The test provides an indication of whether the relatedness of individuals in specific clus- ters is caused primarily by population subdi- vision (Wahlund effects) or by sibling relation- ships. In a two-allelic full-sibling group the average heterozygote excess due to the sibling effect is 7/6 pq (Rasmussen 1979). If individ- uals of a cluster are related, /?group > 0. and have an inbreeding index, /^group < 0. then this indi- cates a group of predominantly siblings. In a structured population, where the population in- breeding index Fy^ > 0, if > 0 and = 0, then this may indicate that a cluster con- sists of offspring from several parents; here the positive value of is a consequence of sub- division of the population (the Wahlund effect). It should be noted, however, that this test is a crude measure when dealing with only one or two polymorphic loci. If only one polymorphic locus is present and parents of a full-sibling group carried alternative alleles, then ~ 0. Fgroup values were tested for significance us- ing permutation tests (see Johannesen & Lubin 1999). The population-wide Wahlund effect (measured here as Fit) is little affected by the sibling effect if more than four sibling groups are sampled (Rasmussen 1979). In a random mating population consisting of family groups, Fn quickly approaches zero as the number of sibling groups increases. Relatedness of group individuals was esti- mated according to the method of Queller & Goodnight (1989). The population R estimate was obtained by jackknifing over clusters. Only clusters having three or more individuals were analyzed in the group comparisons. For esti- mates concerning the total wadi population, all sampled individuals were included in the anal- ysis. Kinship assignment of individuals to sib- ling groups was performed under the primary hypothesis of individuals being full siblings (paternal relatedness pF = 0.5, maternal relat- edness mF = 0.5) against the null hypothesis pF = 0 and mF = 0 using the program Kinship 1.0 (Queller & Goodnight 1989). The above hypothesis of full sibling is conservative, given that we lack information about the frequency of multiple mating in natural populations. RESULTS A deviation from random mating (Hardy- Weinberg proportions) was found within the wadi-population, P < 0.001. Pep-Bl, Est and Idh experienced a deficit of heterozygotes. whereas the two remaining high- polymor- phism loci Ak and Sorbdh and the two low- polymorphism loci Ldhl and Fum did not. All rare alleles were found only as heterozygotes. Furthermore, significant deviations from Har- dy-Weinberg proportions were found within both sections of the wadi (lower half P < 0.05, upper half P < 0.01). Significant differ- ences in allele distributions was found be- tween the upper and lower half of the wadi population (F < 0.001), e.g., the Est alleles 2 and 6 were found only in the lower half, whereas the Ldh allele 87 and Eum allele 124 were found only in the upper half. In the total wadi sample, genetic linkage disequilibrium was observed in 8 out of 21 pair-wise locus comparisons (38%). Genotype distributions are given in the Appendix. An average group relatedness of F = 0.25 ± 0.12 was observed. The relatedness esti- mate for single clusters ranged between —0.21 and 0.82 (Table 1). The relatively large stan- dard deviation indicated that some groups consisted of individuals differing in their ge- netic background. This was confirmed by the Fgroup estimators (Table 1 ), where F„roup ranged between —0.58 and 0.41. Out of the 13 ex- amined clusters, seven showed a heterozygote excess. Based on permutation tests, nine groups had a heterozygote excess, two of which were significant. The remaining groups exhibiting homozygote excess were part of larger clusters. The finding is corroborated by the F-statistics, Fit ^ 0.150 ± 0.067; Fjg = -0.070 ± 0.073; Fst = 0.209 ± 0.086, which showed that both kin (negative F,s) and Wah- lund effects (positive Fn) enhance Fst^ which is the differentiation among groups (Table 1). Kinship-analyses were performed within four groupings: clusters 3-8, cluster 12, ag- gregated clusters 24A, 24B, F21, F23, F25, and cluster 33, under the assumption that in- dividuals in these clusters were full siblings. Figure 2 (see also Appendix) illustrates the kinship assignment based on the rare non- overlapping AkIEstIPep-Bl allele combina- tions in clusters 3-8. Individuals carrying the allelic combinations tested as full-siblings be- longing to one of two sibling groups. Individ- uals from clusters 3-8 were regrouped into two groups consisting of the predicted full- sibling assignments based on Ak! Est! Pep-Bl allelic distributions, and the F and Fg^^p val- ues were estimated for these groups. The in- JOHANNESEN & LUBIN— DISPERSAL OF STEGODYPHUS LINEATUS YOUNG 417 Table 1. — Relatedness and inbreeding index estimates for clusters of juvenile S. Lineatus within a wadi population. Sibling group 1 and 2 estimates are based on rearrangement of individuals from groups 3-8 into two sibling groups based on allelic distributions (see text). Significant permutation /^group’s {P < 0.05) are presented in bold. Group estimates cluster Mean permutation N Relatedness F ^ group E group sd 1 2 2 4 -0.21 0.30 0.19 0.13 3 6 0.28 0.01 -0.12 0.28 4 7 0.11 0.13 0.05 0.12 5 3 0.00 0.41 0.29 0.12 6 5 0.45 0.24 0.09 0.33 7 1 — — — — 8 6 0.64 -0.13 -0.27 0.19 12 6 0.30 -0.05 -0.17 0.24 20 1 — — — — 24A 11 0.22 -0.13 -0.20 0.11 24B 5 -0.07 0.05 -0.10 0.22 F21 5 0.72 -0.58 -0.76 0.30 F23 5 0.09 -0.11 -0.25 0.22 F25 5 0.06 -0.11 -0.25 0.15 33 8 0.82 -0.52 -0.64 0.19 Sibling Group 1 6 0.52 -0.31 -0.24 0.08 Sibling Group 2 Population estimates Relatedness 14 0.55 -0.20 -0.21 Fst 0.08 0.25 ± 0.12 0.15 ± 0.05 -0.070 ± 0.073 0.209 ± 0.086 Figure 2. — Kinship assignment based on allele combinations of clusters 3-8. Ellipses indicate ju- venile clusters on separate plants. Individuals be- longing to sibling group 1 are depicted with squares (H), sibling group 2 members with circles (•), and non-assigned individuals by triangles (A). Two cen- ters of distribution can be recognized. dividuals that were not regrouped were omit- ted from the population analysis to avoid biasing population estimates. The two pre- dicted sibling groups showed a relatedness es- timate of i? ~ 0.50 and a significant excess of heterozygotes, F < 0 (Table 2). However, the genotype composition of sibling group 2 re- vealed that it does not consist of only full- siblings. At least one second mating must be assumed. Each group has a center of origin, and juveniles of the two groups mix (Fig. 2). Kinship analysis of cluster 12 showed two significant full-sibling groups for individuals 1, 3 and 5, and for 4 and 7, respectively. In the upper wadi, the kinship analysis for the group aggregation 24 A, 24B, F21, F23 and F25 did not give unequivocal results. The dis- tribution of rare alleles indicated, as previous- ly, mixing of juveniles but an individual as- signment to sibling groups was only possible for the F21 cluster, which indicated individu- als of a single sibling group. The analysis gave ambiguous results for the remaining in- dividuals due to the lack of rare-allele varia- 418 THE JOURNAL OF ARACHNOLOGY Table 2. — A comparison of within-locality genetic structure in sub-social and social spiders. Genetic variance components are divided into three categories and refer to the location at which an individual can be found relative to the population at large: (i) variance among kin-groups (after juvenile dispersal); (ii) demic effects at a locality (intra-locality Wahlund effect); and (iii) more than one genetically independent population unit within a locality (i.e., closed colonies). Source of variance, among: (i) Kin (iii) Closed Species groups (ii) Demes colonies Reference Sub-social species Eresus cinnaberinus Yes No No Johannesen et al. 1998 Stegodyphus. Uneatus Yes Yes No Johannesen & Lubin 1999 Social species Stegodyphus sarasinorum ? Yes Yes Smith & Engel 1994 Anelosimus eximius ? Yes Yes Smith 1986; Smith Agelena consociata ? Yes Yes & Hagen 1996 Roeloffs & Riechert 1988; Achaearanea wau ? Yes ? Riechert & Roeloffs 1993 Lubin & Crozier 1985 tion. All individuals of cluster 33 indicated full-siblings. However, individual 4 is more likely a half-sibling to the remaining individ- uals, as indicated by the Pep~BI genotype. Combining the kinship analysis and Fg^o^p es- timators suggests that most clusters consisted predominantly of siblings. DISCUSSION letrademic structure of S. lineatus. — We found that despite juvenile nest relocation within the first months of leaving the maternal nest (Ward & Lubin 1993; Lubin et al. 1998), most juveniles remained distinctly clustered and did not disperse over greater distances. Several months after leaving the maternal nest, juveniles in isolated clusters consisted mostly of sibling groups, whereas larger ag- gregations of juveniles originated from several parental pairs. In these larger groups, juve- niles of different parentage can mix. However, even within such aggregations, sibling group- ings could be identified. Furthermore, we found evidence for the occurrence of multiple mating; thus, some of the variation is likely to be due to multiple fathers. The suggestion that multiple mating is not uncommon derives also from field studies showing a high fre- quency of male infanticide and probable re- mating in a natural population (Schneider 1997; Schneider & Lubin 1996, 1997). In the present study, the lack of heterozy- gotes at the population level suggests that mating-dispersal did not take place throughout the population. If mating is random through- out a population, the population at large should obey Hardy- Weinberg proportions. In- deed, if a population is subdivided into family units, the population inbreeding coefficient is expected to be very slightly negative, i.e., showing an excess of heterozygotes (Cocker- ham 1969; Rasmussen 1979). A heterozygote deficit for an entire population that is divided into sibling groups is a strong indication for non-random mating within the population and, therefore, of population sub-structuring. Thus we conclude that S. Imeatus juveniles settle, and later mate, largely in the vicinity of the maternal nest. The existence of clusters of sib- lings suggest that new groups arise by single females colonizing new nesting sites. Despite male dispersal during the mating period (Schneider & Lubin 1996), female group founding must have a greater relative impact on population structure because the propagule-type of cluster founding (Slatkin 1977), combined with generally philopatric behavior, enhances genetic differences among groups more rapidly than male mating-dis- persal can break them down. Furthermore, if females disperse to previously empty sites, then the two phenomena of female philopatry and female dispersal will combine to create neighborhood-structured demes. Therefore, occasional dispersal by females to new local- JOHANNESEN & LUBIN— DISPERSAL OF STEGODYPHUS LINEATUS YOUNG 419 ities away from the cluster does not conflict with the observed structure. Female distribu- tions are the basis for kin-groups. This is sup- ported by a previous study of population structure, which showed that sibling groups could be found even among adult spiders and also suggested that male mating-dispersal was limited (Johannesen & Lubin 1999). The pre- vious results may have been biased because spiders were sampled from long-lived Acacia trees, which are relatively uncommon habitats for S. lineatus in the Negev region. The two studies combined show that a stable environ- ment is not necessary to create sub-structuring of a 5*. lineatus population. In addition, the present study shows that intrademic structure does not depend on the life stage sampled: juvenile and adult populations exhibited sim- ilar population structures. Rather, population sub-structuring into sibling groups is likely to be an inherent consequence of philopatric breeding and dispersal behavior. The type of population structure seen in S. lineatus, where spiders occur in family neigh- borhoods and further population subdivision results from restricted movement of males, may lead to the differential proliferation of both kin-groups and population subsets. One may think of a S. lineatus population as a dy- namic population where new patches arise con- stantly and old o nes disappear. Female group- founding and limited mating-dispersal within the population lead to the differential distribu- tion of genetic variation within populations This pattern can be seen in other populations in the Negev (Johannesen & Lubin 1999). Intrademic structure and social spider evolution. — The genetic structure observed among S. lineatus clusters complies with pre- dictions of increased intra-locality structure for social evolution. However, we need to know whether genetic similarity within groups per se can be extrapolated to provide a basis for the evolution of social behavior. If kin- groups in an open mating system are defined as population units then a significant among- group variance, i.e. similarity of group mem- bers, is inherent (e.g., Ingvarson & Giles 1999), but does not necessarily imply an ad- vantage to sociality. Genetic indices can be used to infer the origin and mode of individual distribution patterns (and structuring process- es) and may as such, be more informative ex- plaining social spider evolution than merely similarity of group members. Common for social spiders is the establish- ment of new colonies or populations by mated females or by several related individuals (Vollrath 1982; Lubin & Robinson 1982), and the presence of closed colony clusters. In oth- er words, genetic similarity is achieved by fe- male lineages (propagule migration model), not by migrants from different populations mating randomly in isolated populations (mi- grant-pool model). The former type of struc- turing process seems also true for the sub-so- cial S. lineatus, albeit in an open system. A comparison of S. lineatus and another sub-so- cial eresid, Eresus cinnaberinus (Olivier 1789) indicates that the genetic variance in S. lineatus is partitioned a step further than in E. cinnaberinus, where family groups are ob- served, but there is no intralocality subdivi- sion (Johannesen et al. 1998). However, in neither S. lineatus nor E. cinnaberinus were localities divided into more than one popula- tion unit, as has been found in social spiders (Table 2). Intralocality differentiation also seems limited or lacking in two non-social species that have been investigated in more detail at the within population level. In Phol- cus phalangioides (Fuesslin 1775) there is suggestive evidence that philopatry may cause some micro-structuring among cellar popula- tions within the same building, but also that frequent dispersal breaks it down repeatedly (Schafer et al. 2001). Vox Atypus ajfinis Eich- wald 1830, no within-population divergence was observed. Structuring processes are ac- tive, however, at distances of a few kilome- tres. Ballooning A. afftnis probably seldom drift beyond the bounds of the population (Pedersen & Loeschcke 2001). In contrast, virtually no structure was detected in the ex- cellent ballooner Argiope trifasciata (Forskal 1775) (Ramirez & Haakonsen 1999). The four social species that have been in- vestigated genetically, Stegodyphus sarasino- rum Karsch 1891, Anelosimus eximius (Key- serling IHS4), Age lena consociata Denis 1965 and Achaearanea wau Levi et al. 1982 have taken population subdivision one step further than S. lineatus by creating a closed genetic system of regularly inbreeding colonies. The general lack of allozyme allelic variation in social relative to sub-social species may also be evidence for a process of group closure 420 THE JOURNAL OF ARACHNOLOGY (Table 2). We emphasize that Table 2 at pre- sent is only suggestive and that three caveats should be noted: 1) The family component in the sub=social species can be estimated be- cause individuals could be assigned to specific grid-positions. This has not been possible in the social species where individuals were tak- en at large from a colony and within-colony family components have not been determined. 2) The term “closed colony” refers to at least two genetically independent colony clusters at one locality. Many localities probably contain clusters derived from a single founding event. Because of the general lack of genetic poly- morphism in social spiders, it could not be determined if these colonies have diverged into independent population units or not. 3) Because the localities of social spiders may consist of more than one population unit, the deme genetic variance is given a priori, A closed population structure will generate extreme variances between groups. If groups have different relative fitness, this may result in selection among groups. Groups with higher productivity (due to cooperation) should pro- duce more young, and therefore more dispers- ing propagules, than groups lacking this trait. To avoid invasion of selfish individuals, high population turnover (Aviles 1993) and dispers- al to establish new trait groups (Sober & Wil- son 1998) are essential. Two of the three social species of Stegodyphus were shown to experi- ence high population turnover (Seibt & Wick- ler 1988; Crouch & Lubin 2000). Further evi- dence for a change in population system, from an open system to a closed one, comes from the study of female-biased sex ratios in social spiders. High population turnover alone is not sufficient to produce these ratios. Two addi- tional components, group-level selection and enough population subdivision to create a ratio of the genetic variances favorable to the group level, are required (Aviles 1993). Intercolony selection can only take place once closed col- onies have been established (Aviles 1997). The genetic data presently compiled on social and sub-social spiders allows preliminary com- parisons of genetic patterns among species of different social levels relative to their breeding behavior (Table 2). One possible test to evaluate the significance of breeding structure in spider social evolution would be to compare the pop- ulation genetic structure of social and sub-social species with communal non-social species. The latter form coherent groups, and perhaps even kin-groups, but are unlikely to inbreed. A com- parison of genetic systems might identify pat- terns for the elucidation of underlying evolu- tionary processes. However, one should keep in mind that patterns do not create processes, rath- er patterns may be used only to infer processes (Templeton 1998). The central unsolved questions are thus how and why do open systems become closed, what ecological conditions make individuals refrain from dispersal altogether and remain clustered (see discussions in Aviles & Gelsey (1998) and Aviles (1999), and is a closed sys- tem required for the evolution of sociality in spiders? Thus, in a genetic context, one needs to distinguish between traits leading to and re- sulting from demic structure. ACKNOWLEDGMENTS We thank Ofer Eitan for help in the field and Jutta Schneider for commenting on the manuscript. Funding for this project came from the 'DFG-NCRD Agreement for the In- vitation of German Senior Scientists to Israel’ and the U.S. -Israel Binational Science Foun- dation (grant # 97-00418). This is publication number 313 of the Mitrani Department for Desert Ecology. LITERATURE CITED Aviles, L. 1993. Interdemic selection and sex ratio: a social spider perspective. American Naturalist 142:320-345. Aviles, L. 1997. Causes and consequences of co- operation and permanent-sociality in spiders. Pp. 476-498, In Social Behavior in Insects and Arachnids. (J. Choe & B. Crespi, eds.). Cam- bridge University Press, Cambridge, Mass. Aviles, L. 1999. Cooperation and non-linear dy- namics: ecological perspective on the evolution of sociality. Evolutionary Ecology Research 1: 459-477. Aviles, L. & G. Gelsey. 1998. Natal dispersal and demography of a subsocial Anelosimus species and its implications of sociality in spiders. Ca- nadian Journal of Zoology 76:2137-2147. Buskirk, R.E. 1981. Sociality in the Arachnida. Pp. 281-367, In Social Insects (H. R. Hermann, ed.) Academic Press, New York. Chesser, R.K. 1991. Gene diversity and female phil- opatry. Genetics 127:437-447. Crouch, T.E. & Y. Lubin. 2000. Effects of climate and prey availability on foraging in a social spi- der, Stegodyphus mimosarum (Araneae, Eresi- dae). Journal of Arachnology 28:158-168 JOHANNESEN & LUBIN— DISPERSAL OF STEGODYPHUS LINEATUS YOUNG 421 Cockerham, C.C. 1969. Variance of gene flow fre- quencies. Evolution 23:72-84. Ingvarson, P.K. & B.E. Giles. 1999. Kin -structured colonization and small-scale genetic differentia- tion in Silene dioicia. Evolution 53:605-611. Johannesen, J., T. Baumann, A. Seitz & M. Veith. 1998. The significance of relatedness and gene flow on population genetic structure in the sub- social spider Eresus cinnaberinus (Araneae: Er- esidae). Biological Journal of the Linnean Soci- ety 63:81-98. Johannesen, J. & Y. Lubin. 1999. Group founding and breeding structure in the subsocial spider Stegodyphus lineatus (Eresidae). Heredity 82: 677-686. Kullmann, E. & W. Zimmermann. 1975. Regurgi- tationsfiitterungen als Bestandteil der Brutsfur- sorge bei Haubennetz- und Rohrenspinnen (Ar- aneae, Theridiidae und Eresidae). Proceedings of the 6^^ International Arachnological Congress, Amsterdam. Louis, E.J. & E.R. Dempster. 1987. An exact test for Hardy -Weinberg and multiple alleles. Bio- metrics 43:805-811. Lubin, Y. D. & R. H. Crozier 1985. Electrophoretic evidence for population differentiation in a social spider Achaearanea wau (Theridiidae). Insectes Sociaux 32:297-304. Lubin, Y.D., J. Hennicke & J. Schneider. 1998. Set- tling decisions of dispersing Stegodyphus linea- tus (Eresidae) young. Israel Journal of Zoology 44:217-225. Lubin, Y.D. & M.H. Robinson. 1982. Dispersal by swarming in a social spider. Science 216:319-321. Pedersen, A. a. & V. Loeschcke. 2001. Conservation genetics of peripheral populations of the myga- lomorph spider Atypus affinis (Atypidae) in northern Europe. Molecular Ecology 10:1133- 1142. Queller, D.C. & K.F. Goodnight. 1989. Estimating relatedness using genetic markers. Evolution 43: 258-275. Ramirez, M.G. & K.E. Haakonsen. 1999. Gene flow among habitat patches on a fragmented land- scape in the spider Argiope trifasciata (Araneae: Araneidae). Heredity 83:580-585. Rasmussen, D.I. 1979. Sibling clusters and geno- typic frequencies. American Naturalist 113:948- 951. Raymond, M. & F. Rousset. 1995. GENEPOP (V.L2): population genetics software for exact tests and ecumenicism. Journal of Heredity 86: 248-249. Riechert, S.E. & R.M. Roeloffs. 1993. Evidence for and consequences of inbreeding in the coopera- tive spiders. Pp. 283-303, In The Natural History of Inbreeding and Outbreeding: Theoretical and Emperical Perspectives. (N. W. Thornhill, ed.). Chicago University Press, Chicago. Roeloffs, R.M. & S. E. Riechert. 1988. Dispersal and population genetic structure of the coopera- tive spider, Agelena consociata, in west African rainforest. Evolution 42:173-183. Schafer, M.A., A. Hille & G.B. Uhl. 2001. Geo- graphic patterns of genetic subdivision in the cel- lar spider Pholcus phalangioides (Araneae). He- redity 86:94-102. Schneider, J.M. 1995. Survival and growth in groups of a subsocial spider {Stegodyphus linea- tus). Insects Sociaux 42:237-248. Schneider, J.M. 1997. Timing of maturation and the mating system of the spider, Stegodyphus lineatus (Eresidae): how important is body size. Biological Journal of the Linnean Society 60:517-525. Schneider, J.M. & Y. Lubin. 1996. Infanticidal male eresid spiders. Nature 381:655-656. Schneider, J.M. & Y. Lubin. 1997. Infanticide by males in a spider with suicidal maternal care, Stegodyphus lineatus (Eresidae). Animal Behav- ior 54:305-312. Seibt, U. & W. Wickler. 1988. Bionomics and fam- ily structure of ’Family Spiders’ of the genus Ste- godyphus, with special reference to the African species S. dumicola and S. mimosarum. Verhan- dlungen Naturwissenschaftlichen Hamburg (NF) 30:255-303 Slatkin, M. 1977. Gene flow and genetic drift in a species subject to frequent local extinctions. The- oretical Population Biology 12:253-262 Smith, D.R.R. 1986. Population genetics of Anelo- simus eximius (Araneae, Theridiidae). Journal of Arachnology 14:201-217. Smith, D.R. & M.S. Engel. 1994. Population struc- ture in an Indian cooperative spider, Stegodyphus sarasinorum Karsch (Araneae, Eresidae) Journal of Arachnology 22:108-113. Smith, D.R. & R.H. Hagen. 1996. Population struc- ture and interdemic selection in the cooperative spider Anelosimus eximius. Journal of Evolution- ary Biology 9:589-608. Sober, E., & D.S. Wilson. 1998. Unto Others. The Evolution and Psychology of Unselfish Behavior. Harvard University Press, Cambridge. Swofford, D.L. & R.B. Selander. 1989. Biosys-1: a Fortran program for the comprehensive analysis of electrophoretic data in population genetics and systematics. Journal of Heredity 72:281-283. Templeton, A.R. 1998. Species and speciation: ge- ography, population structure, ecology, and gene trees. Pp. 32-43, In Endless Forms: Species and Speciation. (D. J. Howard & S. H. Berlocher, eds.). Oxford University Press, New York. Vollrath, F. 1982. Colony formation in a social spi- der. Zeitschrift fiir Tierpsychologie 60:313-324. Ward, D. & Y. Lubin. 1993. Habitat selection and the life history of a desert spider, Stegodyphus 422 THE JOURNAL OF ARACHNOLOGY lineatiis (Eresidae). Journal of Animal Ecology 62:353-363. Weir, B, S. 1991. Genetic Data Analysis. Sinauer, Sunderland. Weir, B. S. & C. C. Cockerham. 1984. Estimating F-statistics for the analysis of population struc- ture. Evolution 38:1358-1370. Manuscript received 20 August 2000, revised 15 May 2001. Appendix. Genotypes of seven polymorphic loci in S. lineatiis juveniles smapled at Sede Boker. Missing genotypes were too weak to score. Allele designations are relative distances, except for esterase which was run in two buffers and alleles were given numbers according to their mobility. Cluster 1 Id. AK EST FUM IDH LD2 PBl S02 Id. AK EST FUM IDH LD2 PBl S02 1 100100 44 100100 100100 100100 7 100100 44 100100 90100 100100 100106 10010 2 100113 45 100100 90100 100100 94106 77100 Cluster 20 Cluster 2 1 100100 45 100100 100100 100100 106106 77100 1 100100 44 100100 90100 100100 94100 100100 3 100113 44 100100 100100 100100 9494 7777 Cluster 24A 5 93100 55 100100 100100 100100 100106 7777 1 100100 45 100100 100100 100100 100106 7777 6 100113 44 100100 90100 100100 100106 7777 2 100113 55 100100 100100 100100 100100 77100 3 100113 45 100124 100100 100100 94106 77100 Cluster 3 4 100100 55 100100 100100 87100 100100 77100 1 100100 45 100100 100100 100100 100100 7777 5 100113 45 100124 100100 100100 100100 7777 2 100113 24 100100 90100 100100 100100 7777 6 100100 55 100100 90100 100100 100100 7777 3 100100 44 100100 100100 100100 100100 77100 7 100113 45 100124 90100 100100 100100 77100 4 100100 44 100100 100100 100100 100100 7777 8 100100 45 100124 100100 100100 94100 77100 5 100100 46 100100 100100 100100 100106 77100 9 100100 55 100124 100100 100100 100100 7777 6 100113 24 100100 9090 100100 100100 77100 10 100100 44 100100 100100 100100 100106 77100 11 100100 55 100100 100100 100100 100106 7777 Cluster 4 1 100100 44 100100 100100 100100 100100 7777 Cluster 24B 2 100100 46 100100 100100 100100 100100 77100 1 100100 44 100100 90100 87100 100100 77100 3 100100 46 100100 100100 100100 106106 7777 2 113113 55 100100 100100 100100 100100 10010 4 100100 46 100100 90100 100100 100106 7777 3 100100 44 100100 100100 87100 100106 77100 5 100100 44 100100 100100 100100 106106 77100 4 100100 44 100100 100100 100100 100106 77100 6 100113 24 100100 90100 100100 100100 77100 5 100113 44 100100 90100 100100 100106 77100 7 100113 24 100100 9090 100100 7777 Cluster F21 Cluster 5 1 100113 45 100100 90100 100100 100106 7777 1 113113 24 100100 90100 100100 100100 77100 2 100100 44 100100 90100 100100 100106 7777 3 100100 44 100100 100100 100100 100100 7777 3 100100 45 100100 90100 100100 100106 7777 4 100100 24 100100 9090 100100 100100 100100 4 100100 44 100100 90100 100100 100106 7777 5 100100 45 100100 90100 100100 100106 7777 Cluster 6 1 100100 44 100100 100100 100100 100100 7777 Cluster F23 2 100100 44 100100 100100 100100 100100 7777 1 100113 45 100124 100100 100100 100100 77100 3 100100 46 100100 100100 100100 100106 77100 2 100113 44 100100 100100 100100 100106 7777 4 100100 44 100100 9090 100100 100100 77100 3 100100 55 100100 100100 100100 100100 77100 5 100113 44 100100 100100 100100 100100 7777 4 100100 45 100100 90100 100100 100106 77100 5 100100 55 100100 90100 100100 100100 77100 Cluster 7 1 100113 44 100100 100100 100100 100100 77100 Cluster F25 1 100100 55 100100 100100 87100 100106 77100 Cluster 8 2 100113 44 100100 100100 100100 94106 77100 1 100113 24 100100 9090 100100 100100 7777 3 100100 55 100100 100100 100100 100106 77100 2 113113 24 100100 90100 100100 100100 100100 4 100113 44 100100 100100 100100 100106 7777 3 100113 44 100100 9090 100100 100100 100100 5 100100 45 100100 100100 87100 4 100113 44 100100 9090 100100 100100 100100 5 100113 44 100100 90100 100100 100100 77100 Cluster 33 6 100113 44 100100 9090 100100 100100 77100 1 100100 45 100124 100100 100100 9494 77100 2 100100 55 100100 100100 100100 9494 77100 Cluster 12 3 100100 45 100124 100100 100100 9494 77100 1 100113 45 100100 100100 100100 106106 7777 4 100100 45 100124 100100 100100 94106 77100 3 100113 45 100100 100100 100100 106106 7777 5 100100 45 100124 100100 100100 9494 77100 4 100113 44 100100 90100 100100 100106 77100 6 100100 55 100100 100100 100100 9494 77100 5 100100 44 100100 100100 100100 106106 7777 7 100100 55 100124 100100 100100 9494 77100 6 100100 44 100100 100100 100100 100106 77100 8 100100 55 100124 100100 100100 9494 77100 2001. The Journal of Arachnology 29:423-426 SHORT COMMUNICATION ON THE GENUS EILICA (ARANEAE, GNAPHOSIDAE) FROM ARGENTINA Violeta Medan: Division Aracnologia, Museo Argentine de Ciencias Naturales, Av. Angel Gallardo 470, C1405DJR Buenos Aires, Argentina ABSTRACT. Eittca pomposa new species, from Buenos Aires, Argentina, and the male of E. iiniformis (Schiapelli & Gerschman 1942) are described for the first time. New records from Argentina for E. Iiniformis, E. modesta and E. trilineata are included. Keywords: Eilica, Gnaphosidae, new species. Neotropical region The genus Eilica Keyserling 1891 can be easily distinguished from other gnaphosids by the presence of two or three translucent lam- inae on the cheliceral retromargin (Platnick 1975), similar to the lamina found in the closely related genus Callilepis Westring. These laminae are probably associated with their preference for preying on ants (Goloboff 2000). The genus is represented in Argentina by four species: E. modesta Keyserling 1891, E. trilineata (Mello-Leitao 1941), E. unifor- mis (Schiapelli & Gerschman 1942) (Platnick 1975, 1985; Platnick & Shadab 1981; Golo- boff 2000) and E. myrmecophila (Simon 1903) (Platnick 1985; Platnick & Shadab 1981). Recent collecting with pitfall traps in Cordoba and Buenos Aires provinces has re- sulted in large samples of previously poorly- known spiders, among which are a new spe- cies of Eilica and the previously unknown male of E. uniformis, which are described be- low. The distribution of the species of Eilica in Argentina is shown in Fig. 1. METHODS All specimens examined are deposited in Museo Argentino de Ciencias Naturales (MACN, Cristina Scioscia), Facultad de Cien- cias Exactas de Cordoba (FCEC, Luis Acosta) and Instituto Argentino de Investigaciones de Zonas Aridas (lADIZA, Sergio Roig-Junent). The format of descriptions follows Platnick (1975) and Brescovit & Hofer (1993). Mea- surements are in millimeters. Eilica uniformis (Schiapelli & Gerschman) Pigs. 1-4 Laronia uniformis Schiapelli & Gerschman 1942: 330, figs. 17, 19 (female holotype from Colonia Dora, Santiago del Estero, Argentina, in MACN, examined); Gerschman & Schiapelli 1967: 201, figs. 17-20. Eilica uniformis: Platnick 1975: 9, figs. 18-19. Diagnosis. — Males are very similar to those of E. rufithorax (Simon 1892) but can be distinguished by their smaller median apophysis and by the pattern of light and dark areas on the dorsum of the abdomen. Description. — Male: (Chancani). Total length 3.08. Carapace 1.24 long, 0.94 wide. Length femur/tibia: I 0.82/0.62; II 0.70/0.52; III 0.60/0.42; IV 0.90/0.72. Carapace light brown medially, darker, striped laterally. Ab- domen with dorsal distinctive pattern of light areas (Pig. 2). Palp (Pigs. 3, 4): Tibial apoph- ysis long, bent. Copulatory bulb: embolus with apical projecting lamella on base. Me- dian apophysis short, bent. Peg spination not provided because in both specimens most spines are lost and their insertions are not dis- tinctly visible. Female: Described by Schiapelli & Gersch- man (1942), Gerschman & Schiapelli (1967), and Platnick (1975). Other material examined. — ARGENTINA: Chubut: Puerto Lobos, Jan. 1975, (E.A. Maury, MACN), 1 $. Cordoba: Chancani, 19 Nov.-23 Dec. 1993, 1319 (C. Mattoni, MACN), 1329 (C. Mat- 423 424 THE JOURNAL OF ARACHNOLOGY toni, FCEC). Misiones: Santa Maria, 1944 (J.M. Vi- ana, MACN), 1?; Puerto Iguazu, 1954 (Schiapelli & Gerschman, MACN), 1 $; Puerto Libertad, 1953 (Schiapelli & Gerschman, MACN), 1 ? . Distribution. — Known only from Argenti- na (Fig. 1). Eilica pomposa new species Figs. 5-7 Type . — Male holotype from a community of Baccharis salicifolia (chilca) in Reserva Natural Otamendi, Buenos Aires province, Argentina, 18 March 1998 (Belen Fuentes and Osvaldo Di lorio, MACN 2780). Etymology. — The specific name is from the Latin word pompa, which means ostenta- tion. Diagnosis. — Eilica pomposa is closest to E. modesta, but it can be distinguished by having a larger median apophysis and a longer and pointed tibial apophysis. Description. — Male (holotype): Total length 3.10. Carapace 1.43 long, 1.08 wide. Length femur/tibia: I 0.90/0.73; II 0.70/ Figure 1. — Distribution of the species of Eilica in Argentina. A = E. modesta, ■ = F. imifonnis, • = E. trilineata, ^ — E. pomposa new species, ★ = E. myrmecophila. Symbols with a white square in- side indicate that the specific locality is not known. (Data from this paper and Platnick 1975, 1981, 1985.) 0.53; III 0.73/0.52; IV 1.00/0.88. Palp de- scription: protruding portion of embolar base twisted, median apophysis large, thick, curved. Tibial apophysis elongate, curved, pointed (Figs. 6, 7). Leg spination: Femora: 1 d O-l-l-O; II d O-l-l-O; III d O-l-l-O, p 1 ap; IV d O-l-l-O, r 1 ap. Tibiae: I v O-O-rl-2; II V O-rl-rl-2; III p l-l-O-O, v O-pl-2-2, r 0- 0-dl-dI; IV r O-O-dl-dl. Metatarsi: I v 0-2- rl-2 or V O-pl-pI-2; II v 0-2-0-2 or v 0-2-rl-2; III p 1 ap., v O-O-rl-2, r 1 ap.; IV p 1 ap., V O-pl-pl-2, r 1 ap. Tarsi: IV v 0- 0-2-2. Abdomen: Dorsal pattern of four pairs of white specks and several posterior dark lines (Fig. 5). Eemale. — Unknown. Other material examined. — A male from Ar- gentina, Mendoza, Reserva Nacunan, 22 Nov. 1997 (S. Lagos, lADIZA), seems to belong to the same species although it is from a very distant locality {ca. 1000 km apart, see Fig. 1). Further collections may help to elucidate the actual distribution of the species. ADDITIONAL RECORDS IN ARGENTINA Eilica modesta — Jujuy: No specific locali- ty, 17 Jan. 1966 (E.A. Maury, MACN), 1$. Cordoba: Tanti, 1950 (J.M. Viana, MACN), Id; Calamuchita, 1955 (J.M. Viana, MACN), 2 $ . Buenos Aires: Sierra de la Ventana, Dec. 1968 (E.A. Maury, MACN), 1 $ ; San Vicente, 1964 (Canto, MACN), 29. Santiago del Es- tero: No specific locality, March 1963 (Hep- per, MACN), 1 9 . Eilica trilineata. — Catamarca: Capillitas, 1 Eeb. 1981 (A. Roig, MACN), 2d2 9. San Juan: Las Tumanas, 14 April 1979 (A. Roig, MACN), 19. Rioja: Aimogasta, 1944 (J. Caceres Ereyre, MACN), 1 9 . ACKNOWLEDGMENTS I thank Susana Lagos for lending valuable collections of gnaphosids from the Reserva Nacunan, Mendoza province, to Belen Euen- tes and Osvaldo Di lorio for collecting the material in which E. pomposa was found, and Norman Platnick and Antonio Brescovit for critical reading of the manuscript. Einally, I wish to thank Martin Ramirez for guiding me throughout this work. MEDAN— NEW EILICA FROM ARGENTINA 425 Figures 2-7. — 2. Eilica uniformis, abdomen, dorsal view; 3. E. unifonnis, palp, ventral view; 4. E. uniformis, palp, retrolateral view. 5. Eilica pomposa new species, abdomen, dorsal view; 6. E. pomposa new species palp, ventral view; 7. E. pomposa new species, palp, retrolateral view. Scale = 0.5 mm. 426 THE JOURNAL OF ARACHNOLOGY LITERATURE CITED Brescovit, A.D. & H. Hofer. 1993. Aranhas dos generos Lygromma e Eilica, da Amazonia Cen- tral, Brasil (Araneae, Gnaphosidae). Iheringia, Serie Zoologia 74:103-107. Gerschman, B.S. & R.D. Schiapelli. 1967. El ge- nero Laronia Simon, 1892 (Araneae, Drassodi- dae). Acta Zoologica Lilloana 23:189-200. Goloboff, P.A. 2000. The family Gallieniellidae (Araneae, Gnaphosoidea) in the Americas. Jour- nal of Arachnology 28:1-6. Platnick, N.I. 1975. A revision of the spider genus Eilica (Araneae, Gnaphosidae). American Mu- seum Novitates 2579:1-19. Platnick, N.I. 1985. Notes on the spider genus Ei- lica (Araneae, Gnaphosidae). Journal of the New York Entomological Society 93(3): 1073-1081. Platnick, N.I. & M.U. Shadab. 1981. On the spider genus Eilica (Araneae, Gnaphosidae). Bulletin of the American Museum of Natural History 170: 183-188. Schiapelli, R.D. & B.S. Gerschman. 1942. Aranas argentinas (U parte). Anales. Museo Argentine de Ciencias Naturales “Bernardino Rivadavia,” 40:317-332. Manuscript received 1 October 2000, revised 6 Eebruary 2001. 200L The Journal of Aracheology 29:427-430 SHORT COMMUNICATION DISTINGUISHING THE FEMALES OF TROCHOSA TERRICOLA AND TROCHOSA RURICOLA (ARANEAE, LYCOSIDAE) FROM POPULATIONS IN ILLINOIS, USA Tlioiii.as R. Prentice* Department of Entomology, University of California, Riverside, California 92521 USA Keywords: Spider, Lycosidae, Trochosa The Palearctic lycosid species, Trochosa ruricola (De Geer 1778), is here recorded for the first time in the State of Illinois. It was previously known only from Massachusetts in the USA. The species is widely distributed in northern and middle Europe and Asia and was apparently introduced into Bermuda sometime prior to 1888 (Marks 1889; Banks 1902; Sier- wald 1988). Trochosa terricola Thorell 1856 is considered a Holarctic species, occurring in northern Europe as well as in North America, from Alaska to Newfoundland and south to northern California, Arizona, south^central Texas, and Alabama (Brady 1979; Roberts 1985; Dondale & Redner 1990). Both Tro- chosa C.L. Koch 1847 species are common throughout much of their respective ranges in northern Europe and the British Isles (Roberts 1995). They are the most prevalent of the four European congeners in the British Isles where they often occur in sympatry (Roberts 1985; Edwards 1993). Sometime prior to 1993, a thriving popu- lation of T. ruricola was discovered in Cape Cod, Massachusetts, which outnumbered the native T. terricola in pitfall trap samples by a ratio somewhat greater than 2:1 (Edwards 1993). In 1994, the first known specimens were collected in Canada (L’Acadie, Quebec) (Lalonge et al. 1997). During a 1999 biodi- versity study (funded by Chicago Wilderness) within three forest preserves in Lake County, Illinois, both T. ruricola and T. terricola were discovered in sympatry. In this study, how- ever, T. terricola outnumbered the non-native T. ruricola by a factor of approximately 2.5: 1 (males-L9:l; femaleS“-4.25:l). Edwards (1993) distinguished T. terricola from E rur- icola by the absence or presence (both sexes and all instars in either case), respectively, of a claw on the palpal tarsus. Trochosa ruricola males were also distinguished by a unique ridge or carina at the base of the fang on the outer (convex) face and by a slight, anteriorly directed bend or curl of the apical portion of the embolus (Edwards 1993). In T. terricola, the ridge on the fang is lacking and the apical portion of the embolus forms a circular loop (Brady 1979; Edwards 1993). The Canadian researchers found only T. ruricola in Quebec samples (Lalonge et al. 1997). In Lake County, Illinois pitfall trap samples from Grainger Woods, Elm Road, and Spring Bluff Forest Preserves, all Trochosa females and juvenile instars examined were equipped with a pectinate claw on the palpal tarsus. Grounded on Edwards’ presence or absence characters, I began to assign all Trochosa fe- males to T. ruricola, based on the presence of a palpal claw. However, in light of the domi- nance of T. terricola males in samples, I found it hard to believe that only T, ruricola females were subject to pitfall collection. Af- ter thorough reexamination of all Trochosa fe- males, I discovered that, in some specimens, there v/as relatively wide and usually unbro- ken, light-colored submarginal band on the carapace (Fig. 1). In others, the band was somewhat narrower and broken in several places (Fig. 2) or was barely discernible. Re- examination of the males revealed a similar difference in banding patterns. In T. ruricola males, this band was relatively wide and often unbroken (Fig. 3) or, if broken, only faintly 427 428 THE JOURNAL OF ARACHNOLOGY Figures 1-6. — Trochosa mricola and T. terricola from Lake County, Illinois. 1, 3, 5, T. ruricola. 1. Female carapace; 3. Male carapace; 5. Epigynum, ventral. 2, 4, 6, T. terricola. 2. Female carapace; 4. Male carapace; 6. Epigynum, ventral. Abbreviations: ir = internal longitudinal ridge; mer = medial portion so and in few places. In T. terricola males, the band was usually narrow and widely bro- ken in several places by the dark pubescence of the submedial regions of the carapace or was not discernable (Fig. 4). Females were tentatively separated on the basis of the sim- ilarity of their bands to the respective male pattern. Those with a wide and largely unbro- ken band were determined as T. ruricola and those without an apparent band or with a nar- PRENTICE— DISTINGUISHING FEMALES OF TWO TROCHOSA SPECIES 429 row, largely broken band determined as T. ter- ricola (the submarginal bands of the T. ruri- cola specimens illustrated in Figs. 1 & 3 were unusually narrow). Substantiation of the female submarginal band configurations appeared to be confirmed, not only by the corresponding male patterns, but also by the details of the ventral views of the respective epigyna (internal structural dif- ferences were found to be unreliable). Both Roberts’ illustrations and those presented here clearly show that, in T, ruricola, the orienta- tion of the spermathecae is directed lateroan- teriad (obliquely oriented) to the termini of the transverse portion of the median septum (Fig. 5; compare to Roberts 1985, fig. 62c). In T. terricola, the orientation is directed anteriad to the termini (Fig. 6; compare to Roberts 1985, fig. 62e). In T. ruricola, the internal lon- gitudinal ridges (darkened tube-like structures visible at the lateral edges of the longitudinal portion) are generally separated medially to- ward the posterior end of the longitudinal por- tion of the septum. They usually do not im- pinge on the posterior border of the transverse portion where they merge with the copulatory tubes (Fig. 5). In T. terricola, the internal ridges are usually contiguous (or nearly so) medially and are usually visible (ventral view) near the posterior border of the septum (Fig. 6). The transverse portion of the septum is relatively short (relative to the length of the longitudinal portion) and the anterior margins only slightly concave to moderately straight in T. ruricola (Fig. 5). By comparison, the transverse portion is relatively long and the anterior margins generally concave in T. ter- ricola (Fig. 6). The median ectal rim portions of the paired hood cavities are directed pos- teriad (median lateral edges parallel) or me- dioposteriad in T. ruricola (Fig. 5; Roberts 1985, fig. 62c) but are usually directed later- oposteriad (oblique) in T. terricola (Fig. 6; Roberts 1985, fig. 62e). Differentiation of the Lake County Trocho- sa females was generally conclusive by em- ploying only the submarginal band character (epigynal characters were also used to confirm placement). However, this banding pattern may appear to be somewhat subjective to fu- ture workers if only females of one of the two species occur in northern Illinois samples. But even in T. terricola females with well-devel- oped submarginal bands, the pattern is almost always widely broken in at least one region or more narrowly broken in several regions. In doubtful cases, especially in instances in which the epigyna of the respective females are very similar, a combination of the epigynal details (ventral view) and submarginal config- uration may have to be used to separate the species. In regions of North America where T. terricola females lack a palpal claw, Edwards’ characters should suffice to separate the two Trochosa females. I would like to thank F Catchpole for en- listing my services for the identification of Ar- aneae from Lake County, C. Dondale for ex- amining several Trochosa specimens and acknowledging the value of the submarginal band character, and M. Planoutene for pro- ducing the original artwork used in the illus- trations. All Lake County Trochosa speci- mens, except those retained by the author, are deposited in the Field Museum, Chicago, Il- linois. LITERATURE CITED Banks, N. 1902. Some spiders and mites from the Bermuda Islands. Transactions of the Connecti- cut Academy of Arts and Sciences 1 1:267-275. Brady, A.R. 1979. Nearctic species of the wolf spi- der genus Trochosa (Araneae: Lycosidae). Psy- che 86:167-212. De Geer, C. 1778. Memoires pour servir a I’histoire des insectes (Stockholm) 7(3-4): 176- 324. Dondale, C.D. & J.H. Redner. 1990. The insects and arachnids of Canada, Part 17. The wolf spi- ders, nurseryweb spiders, and lynx spiders of Canada and Alaska, Araneae: Lycosidae, Pisaur- idae, and Oxyopidae. Research Branch, Agricul- ture Canada, Publication #1856:1-383. Edwards, R.L. 1993. New records of spiders (Ar- aneae) from Cape Cod, Massachusetts, including two possible European immigrants. Entomolog- ical News 104:79-82. Koch, C.L. 1847. Die Arachniden, Niirnberg, Vier- zehnter Band, pp. 89-210, Funfzehnter Band, pp. 1-136, Sechszehnter Band, pp. 1-80. Lalonge, S., J.H. Redner, & D. Coderre. 1997. First Canadian records of Trochosa ruricola (De Geer), Ostearius melanopygius (O. Pickard- Cambridge), and Dictyna decaprini Kaston (Ar- aneae: Lycosidae, Linyphiidae, Dictynidae, re- spectively). Canadian Entomologist 129:371- 372. Marks, G. 1889. A contribution to the knowledge of the spider fauna of the Bermuda Islands. Pro- ceedings of the Academy of Natural Sciences, Philadelphia 1889:98-101. 430 THE JOURNAL OF ARACHNOLOGY Roberts, M.J. 1985. The Spiders of Great Britain and Ireland, Volume 1; Atypidae to Theridioso- matidae. Harley Books, Colchester, England. Roberts, M.J. 1995. Collins Field Guide: Spiders of Great Britain & Northern Europe. Harper Col- lins Publishers, London. Sierwald, P. 1988. Spiders of Bermuda. Nemouria: Occasional Papers of the Delaware Museum of Natural History 31:1-24. Thorell, T 1856. Recensio critica aranearum sue- cicarum quas descripserunt Clerckius, Linnaeus, de Geerus. Nova acta Regiae Societatis Scientar- um Upsaliensis (3) 2(1):61-176. Manuscript received I November 2000, revised 12 March 2001. 2001. The Journal of Arachnology 29:431-433 SHORT COMMUNICATION THE UNUSUAL EGG ROD OF THE SPIDER HOMALOMETA CHIRIQUI (ARANEAE, TETRAGNATHIDAE) AND OTHER BIOLOGICAL DATA Guillermo Ibarra-Nunez: El Colegio de la Frontera Sur, Carr. Antiguo Aeropuerto km 2.5, Apdo. Postal 36, Tapachula, Chiapas 30700, Mexico ABSTRACT. Field observations of Homalometa chiriqui (Araneae, Tetragnathidae), a common habitant in coffee plantations in Chiapas, Mexico, provide biological information on this poorly known species. Collection data revealed several generations per year. The web architecture and microhabitat selection of young juveniles differ from those of older juveniles and adult females. Females deposit their eggs inside the retreat forming a straight cylindrical egg-rod. Keywords* Homalometa, egg laying, web architecture, microhabitat, life cycle The American orb weaver genus Homalo- meta Simon 1897 (Tetragnathidae) is known only from three species: H. nigritarsis Simon 1897 from the Lesser Antilles (Island of Saint Vincent and Martinique), Panama and Mexico (Levi 1986); H, chiriqui Levi 1986 from Pan- ama, Costa Rica and Mexico (Ibarra & Garcfa 1998); and H. nossa Levi 1986 from Brazil. Very little is known about the biology of this genus: females of H. chiriqui have been “collected to the side of orb— both specimens with a set of eggs under a leaf; in the orb at night” (Levi 1986). No more biological in- formation on this genus occurs in the litera- ture. During studies on the diversity and ecology of spiders of coffee plantations in the Socon- usco region of Chiapas, Mexico (15°10'N, 92°20'W), we found H. chiriqui as a common resident in two coffee plantations. The spiders were collected by hand, after measuring the diameter of its web, the distance from it to the ground, and noting the placement and orien- tation of its web and retreat, and presence of eggs or spiderlings. Voucher specimens of these spiders were deposited in the Coleccion de Aranas del Sureste de Mexico (ECO-TA- AR) of El Colegio de la Frontera Sur, Tapa- chula, Chiapas, Mexico. In a two-year monthly sampling we found 481 specimens: 414 juveniles on webs, and 67 adults (54 females, 13 males). These speci- mens were found between 800--1000 m above sea level. All age classes and both adult sexes were found almost all year round, as well as females with eggs or young spiderlings inside the retreats (Table 1). This shows clearly that this species has several generations per year. There were differences in form and orien- tation of the web, as well as in the microhab- itat selection for the placement of the web and its associated retreat among younger and older juveniles and adult females. Young juveniles spin their small (2-4 cm in diameter) sym- metrical and almost horizontal orb webs on the upper side of a coffee leaf, supported mainly from its lateral borders, with the spi- der’s retreat below the web, on the leaf upper surface, where the spiderling is found most of the time. Larger juveniles and adult females spin a vertical asymmetrical orb web on the side of a coffee leaf (Fig. 1), also supported by other leaves and branches, with a mean vertical diameter of 1 1 cm (5-20 cm). In this case the retreat is built on the underside of a leaf in front of the web’s hub and slightly in- clined to the vertical. A signal line leads from the hub of the web to the entrance of the re- treat. The retreat is an elongated vertical vault made of a thin layer of silk, almost transpar- ent, with one exit hole for the spider on the upper side and the entrance on the lower side 431 432 THE JOURNAL OF ARACHNOLOGY Figures 1-4. — Homalometci chiriqui, female web, retreat with female, egg-rod and spiderlings. 1. Ver- tical asymmetrical web among coffee leaves showing signal line (arrow); 2. Retreat on underside of coffee leaf showing egg-rod with female on side (white arrow), entrance (lower black aiTow) and exit holes (upper black arrow); 3. Recently emerged spiderlings (arrow) around egg chorions and exuviae; 4. View of egg-rod in detail and female (arrow). IBARRA-NUNEZ— EGG-ROD OF HOMALOMETA CHIRIQUI 433 Table 1. — Accumulated monthly abundance of juveniles on webs and adults (by sex) of H. chiriqui, in a two-year sampling at two coffee plantations in the Soconusco region of Chiapas, Mexico. The number of females found with egg-rod or spiderlings is noted in parentheses. Month Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Total Juveniles 59 20 71 16 26 8 11 22 97 10 43 31 414 Females 5 + (2) I 4 (2) 1 2 + (4) 6 + (4) (1) 1 + (2) 0 9 + (2) 8 37 + (17) Males 0 1 0 1 3 1 1 0 2 0 4 0 13 Total 66 22 75 19 30 15 22 23 102 10 58 39 481 (Fig. 2). The spider (juvenile or adult) is nor- mally inside the retreat, but it drops very quickly when slightly disturbed if without off- spring. If with eggs or spiderlings, the female is more reluctant to abandon its retreat. The mean distance of the web to the ground was 135 cm for both small juveniles (35-204 cm) and larger juveniles and female adults (30- 350 cm). Adult females deposit their eggs inside the retreat, but they do not make a conventional egg-sac. Instead, they place the eggs in a line over a thin layer of silk, forming a straight cylindrical “rod,” three to four eggs in width (Figs. 2, 4). The eggs adhere to one another and are supported by the basal thin layer of silk. This is an unusual way to arrange the eggs, and no other species has been reported in the literature to put its eggs in a rod exactly like H. chiriqui. Females of the uloborid ge- nus Miagrammopes and some species in the genus Argyrodes make cylindrical egg-sacs, but in these cases the eggs are visibly wrapped with silk (Exline & Levi 1962; Opell 1984, 1989). In contrast, the egg-rod of H. chiriqui did not appear to be covered by silk. After hatching, the young spiderlings of H. chiriqui stay in the rod by the side of the egg chorions and first exuviae, with the female guarding them (Fig. 3) until they molt and become larg- er. Then, they began to move on the retreat until they abandon it. The change in the web architecture from horizontal and symmetrical webs (in young juveniles) to vertical and asymmetrical webs (in older juveniles and adult females) suggests that — in this taxonomic context — vertical and asymmetrical are the apomorphic conditions. It is possible that the vertical asymmetrical web and the egg-rod constitute generic syna- pomorphies that could be tested when more information on the other species of this genus be available. I thank Alvaro Garcia Ballinas and Manuel Alberto Moreno Prospero (EGOS UR) for their assistance in collecting and photograph tak- ing. Walter Peters generously give us facilities to work in the coffee plantations of Finca Ir- landa. Comments from G. Hormiga, the edi- tors of the Journal of Arachnology, and an anonymous reviewer improved an early draft of this paper. This work was supported in part by grant R28867-N of CONACYT-Mexico. LITERATURE CITED Exline, H. & H.W. Levi. 1962. American spiders of the genus Argyrodes (Araneae, Theridiidae). Bulletin of the Museum of Comparative Zoology 127:75-204. Ibarra-Nuiiez, G. & J.A. Garcia-Ballinas. 1998. Diversidad de tres familias de arahas tejedoras (Araneae: Araneidae, Tetragnathidae, Theridi- idae) en cafetales del Soconusco, Chiapas, Mex- ico. Folia Entomologica Mexicana 102:11-20. Levi, H.W. 1986. The Neotropical orb- weaver gen- era Chrysometa and Homalometa (Araneae: Te- tragnathidae). Bulletin of the Museum of Com- parative Zoology 151:91-215. Opell, B.D. 1984. Eggsac differences in the spider family Uloboridae (Arachnida: Araneae). Trans- actions of the American Microscopical Society 103:122-129. Opell, B.D. 1989. Do female Miagrammopes ani- motus (Araneae, Uloboridae) spin color-coordi- nated egg sacs? Journal of Arachology 17:108- 111. Simon, E. 1897. On the spiders of the Island of St. Vincent. Proceedings of the Zoological Society of London 1897:860-890. Manuscript received 1 October 2000, revised 19 March 2001. 2001. The Journal of Arachnology 29:434-437 BOOK REVIEWS: NEW VOLUMES ON THE “MINOR” ARACHNID ORDERS The Biology of Camel-Spiders (Arachnida, Solifugae). Fred Punzo. 1998. Kluwer Academic Publishers, Norwell, Massachusetts. 301 pp. ISBN 0-7923-8155-6. US$135.00 (hard cover). Whip Spiders (Chelicerata: Amblypygi). Their Biology, Morphology and Systematics. Peter Weygoldt. 2000. Apollo Books, Stenstrup, Denmark. 163 pp. ISBN 87-88757-6-3 US$43.00 (DKK 320.00) (hard cover). Hormiga (2000) recently attributed the dwindling number of publications on the “mi- nor” arachnid orders (in decreasing number of described species — Opiliones, Pseudoscorpi- ones, Scorpiones, Solifugae, Schizomida, Am- blypygi, Uropygi, Palpigradi, Ricinulei) — to two factors. First, these orders are less diverse, collectively comprising only 12.5% of arach- nid species, compared with the megadiverse Acari and Araneae, containing some 87.5% (M.S. Harvey unpubL data). Second, there has been an alarming decline in specialists on these taxa, despite the fragmentary knowledge about most aspects of their biology. At a time when major synthetic publications on the bi- ology of the minor orders are as scarce as the specialists working on them, it is gratifying to report on new volumes that will hopefully re- store interest and perhaps stimulate the devel- opment of new expertise in some of these in- triguing groups. As noted by Hormiga (2000), Tome VI of Grasse’s (1949) Traite de Zoologie remains the standard text for the anatomy and biology of most of the minor orders, only two of which have been dealt with in more recent volumes: Pseudoscorpiones (Weygoldt 1969); Scorpiones (Polis 1990). Until recently, pseu- doscorpions remained the most “accessible” of the minor orders, for which both a com- prehensive synthesis of the biology (Weygoldt 1969) and an up-to-date, fully referenced tax- onomic catalogue (Harvey 1990) were avail- able. Probably due to their medical importance, scorpions have enjoyed disproportionate at- tention, given that they represent only the third most speciose of the minor orders. A re- surgence in studies on scorpions has occurred in the decade since publication of the volume by Polis (1990), culminating in the recent publication of another two volumes, present- ing current research in scorpion biology (Brownell & Polis 2000) and cataloguing the taxonomic diversity of the order (Fet et al. 2000). A third volume, dealing with all as- pects of scorpion biology, from taxonomy, through ecology to neurophysiology, has just been published (Fet & Selden 2001). It was hoped that the publication of three volumes on Chelicerata in the series on Mi- croscopic Anatomy of Invertebrates (Harrison & Foelix 1999) would provide an update of the information provided in Grasse’s (1949) Tome VI for the remaining minor orders. Un- fortunately, these volumes dealt only with the Acari, Araneae and Scorpiones, while omit- ting eight of the minor orders altogether (Hor- miga 2000). However, two new volumes, which are the subject of this review, should redress this void for the Solifugae and Am- blypygi, two of the smaller and more enig- matic orders for which no synthetic treatments were previously available. The first of these volumes, The Biology of Camel-Spiders, has already received a thor- ough review by Hormiga (2000); what follows shall serve merely to supplement the latter. In 434 BOOK REVIEW 435 accordance with the previous reviewer, I am commending the author for filling a notorious gap in the arachnological literature by sum- marizing available knowledge on Solifugae into a single, clearly written volume. Regret- tably, poor production by the publishers, cou- pled with various inconstencies, errors and omissions in the content, detract from other- wise fascinating subject matter. Hormiga (2000) has already lamented the poor production of this volume, which is pep- pered with typographical errors, misaligned text, oddly positioned blank spaces, scientific binomens not set in different font style, etc. Treatment of the illustrations is particularly appalling. Some are unnecessarily large (p. 43) while others are poorly reproduced from adequate originals (p. 37). Still others are aw- ful regardless of size or reproduction (p. 181). Several are not centered on the page (p. 22) or leave large blank spaces between the illus- tration and the text (pp. 15, 220). In many cases, the text of the illustrations is too large (p. 52). Further criticisms concern aspects of the content. Eight chapters deal in somewhat hap- hazard fashion with the following topics: in- troduction to Solifugae, including mythology and folklore (10 pp.); functional anatomy and physiology (33 pp.); neurobiology (25 pp.); life history (35 pp.); ecology (43 pp.); behav- ior (45 pp.); phylogeey, biogeography and taxonomy, including identification keys (44 pp.); field and laboratory techniques (13 pp.). The apparent absence of a logical structure both among and mdthin the chapters inevitably leads to repetition, e.g., discussions of life his- tory in the introduction (p. 8) and of behavior in the chapter on life history (p. 72). Some sections of Chapters 2, 3 and Chapters 4 and 6 would have benefited from a merger. Valuable, if somewhat disproportionate coverage is devoted to the sections on natural history, habitat preference, diet composition, dispersion patterns, mating and hunting be- havior, burrowing biology and diel/seasonal activity patterns (Chapters 4-6), many of which draw on data from the author's own studies. The final chapter provides important information on collecting and rearing solifu- ges, again drawn from the author's experience. Solifuges are notoriously difficult to maintain alive, let alone rear under captive conditions and the author is the only person on record to have successfully reared a species of solifuge through an entire generation (Punzo 1998). Unfortunately, the sections presenting ma- terial outside of the author's expertise — anat- omy, physiology and neurobiology (Chapters 2, 3) and phylogeey, taxonomy and biogeog- raphy (Chapter 7) — provide an unfavorable contrast with the sections on ecological and life history aspects, both in breadth of cover- age and in accuracy of content. As previously noted by Hormiga (2000), the anatomical chapters are conspicuous for the paucity and poor quality of illustrations, including the complete absence of photographs (only 15 photographs appear in the entire book, all in Chapters 5, 6), while the chapter on system- atics contains several inconsistencies. Most notable is the presentation of Van der Ham- men's (1989) non-cladistic classification of the Chelicerata in Table 7-1 (p. 200), rather than the better justified system of Shultz (1990) — derived from a cladistic analysis of morphological data and recently supported by molecular data (Wheeler & Hayashi 1998) — which receives only secondary mention in the text (p. 202). The incomplete treatment of the chelicerate fossil record, with its implications for phylogeny have similarly been noted by Hormiga (2000). According to Fig. 7-1 (p. 198), Araneae are dated from the Carbonif- erous, despite Devonian spider fossils (Selden et al. 1991), while Solifugae are dated from the Tertiary, although the morphology of a Carboniferous solifuge is discussed a few pag- es later (pp. 211-214), To these errors and omissions can be added the inconsistent use of terminology, e.g., the old ordinal names Ar- aeeida, Scorpionida and Solpugida. On the positive side, this chapter provides a useful synthesis of identification keys, adapted from multiple sources, for the solifuge families of the world, as well as regional keys to the fam- ilies and genera of North America (including Mexico), Israel and South Africa, and to the South American genera of Ammotrechidae. The book concludes with an extensive refer- ence list (34 pp.) — more than 500 entries on all aspects of solifuge biology (33 by the au- thor, though admittedly only 16 of these deal with solifuges)' — itself a valuable introduction to the literature on the order. Indeed, the syn- thesis of such disparate information on soli- fuges from a large variety of scattered sources is certainly the most laudable aspect of a vol- 436 THE JOURNAL OF ARACHNOLOGY ume that will remain a landmark in the arach- nological literature, despite its many short- comings. Whip Spiders, the second volume under consideration in this review, follows the tra- dition of Weygoldt (1969) in presenting an- other immaculately illustrated compendium on a remarkable order of arachnids. Both thor- ough content and impeccable presentation make for an unfair comparison with The Bi- ology of Camel-Spiders. The publishers are similarly commended for the professional pro- duction of this volume. The book is organized into nine chapters, dealing with the following topics: introduction to the Amblypygi (1 p.); historical background to studies on Amblypygi (2 pp.); external morphology (8 pp.); genera of Amblypygi, in- cluding identification key (16 pp.); anatomy and general biology, including behavior (87 pp.); distribution and ecology (9 pp.); endan- gered species (1 p.); systematics (7 pp.); field and laboratory techniques (2 pp.). These chap- ters are followed by a bibliography of selected references (7 pp.), no fewer than 33 by the author, but the reader is refened to a more extensive bibliography on the website of the International Society of Arachnology. As with any such volume, there will always be differences of opinion as to how the subject matter could have been organized. For ex- ample, perhaps the large chapter on anatomy and general biology (Chapter 5) could have been split into smaller chapters and some of the smaller, but related chapters (e.g.. Chap- ters 1, 2, Chapters 4 and 8, Chapters 6, 7) merged into larger, more inclusive units. Sim- ilarly, some subsections of particular chapters (e.g., the three lines on fighting and territori- ality, p. 70) are arguably too short to warrant separate treatment and might have been better combined with others. But these are only mi- nor criticisms of an excellent book overall. As with his previous book on pseudoscor- pions, the author demonstrates proficiency in all aspects of the biology of his subjects — from morphology, through ecology and distri- bution, to systematics — thereby rendering fu- tile an attempt to find fault with the content. Aside from a distinct absence of errors and an unbiased treatment of conflicting opinions (e.g., the alternative hypotheses of chelicerate phylogeny presented in Chapter 8), the most obvious drawcard of this volume (subject mat- ter notwithstanding) are the many exceptional illustrations that accompany each discussion as supporting evidence. Just as the sections on external morphology and anatomy are liber- ally illustrated with aesthetically appealing line drawings and clear electron micrographs, so the sections on behavior are illustrated with neat photographic sequences documenting rit- ualistic postures assumed in encounters be- tween members of the same or opposite sex. Similarly, the sections on distribution, ecology and systematics are supported by maps, hab- itat photographs and cladograms, respectively. The author is further commended for pre- senting the subject matter in a manner that is accessible and informative to both general and specialist readers alike. For example, by draw- ing on his extensive experience in the com- parative morphology and systematics of Chel- icerata, the author presents a detailed scientific synthesis and commentary of the characters relevant for delimitation and diagnosis of Am- blypygi, as well as the phylogenetic relation- ships among them and their chelicerate rela- tives (Chapters 3, 4 & 8). However, the chapter on amblypygid genera should also be very useful to the general reader, for it in- cludes not only a synopsis of each genus, ac- companied by photographs of exemplar spe- cies, but a straightforward identification key. The sections on anatomy, behavior, distribu- tion and ecology are equally comprehensive and bear testament to the author’s many pub- lished scientific contributions. But the hard science is rendered accessible to the general reader by the regular inclusion of topics with broader interest, e.g., why are there so few species of Amblypygi, why is the courtship dance so prolonged and why do amblypygid females mate more than once. Of all the topics covered, it is without doubt the author’s meticulous studies on the com- plex mating behavior and related aspects of the reproductive biology of whip spiders (summarized in Chapter 5) that are most in- spiring. During the past 30 years, the author has personally collected, transported and reared more than 20 species of Amblypygi for these studies, some of which are still main- tained in laboratory colonies to this day. The difficulty in collecting many of these elusive animals in the wild, not to mention the pa- tience required to maintain them and observe their nocturnal activities, whether in the field BOOK REVIEW 437 or laboratory, demonstrate an uncommon ded- ication. Whip Spiders is yet another example of that dedication, and will certainly remain the standard text on Amblypygi for years to come, just as Weygoldt (1969) has remained the standard text on pseudoscorpions. I highly recommend this book as an essential addition to the libraries of all arachnologists. LITERATURE CITED Brownell, P.H. & G.A. Polis. (eds.) 2000. Scorpion Biology and Research. Oxford University Press, New York. Fet, V., W.D. Sissom, G. Lowe & M.E. Braunwald- er. 2000. Catalog of the Scorpions of the World (1758-1998). New York Entomological Society, New York. Fet, V.A. & P.A. Selden. (eds.) 2001. Scorpions 2001: In Memoriam Gary A. Polis. British Ar- achnological Society, Burnham Beeches, Buck- inghamshire, Grasse, P. (ed.) 1949. Traite de Zoologie. Tome VI. Masson et Cie, Paris. Harrison, F.W. & R.F Foelix. (eds.) 1999. Micro- scopic Anatomy of Invertebrates: Chelicerate Ar- thropoda. Volume 8, Parts A-C. Wiley & Sons, Inc., New York. Harvey, M.S. 1990. Catalog of the Pseudoscor- pionida. Manchester University Press, Manches- ter. Hormiga, G, 2000. Reviews: The Biology of Cam- el-Spiders (Arachnida, Solifugae). Systematic Biology 49:613-614. Polis, G.A. (ed.) 1990. The Biology of Scorpions. Stanford University Press, Stanford, California. Punzo, F. 1998, Natural history and life cycle of the solifuge Eremobates marathoni Muma, Brookhart (Solifugae, Eremobatidae). Bulletin of the British Arachnological Society 11:111-118. Selden, R, W.A. Shear & P.M. Bonamo. 1991. A spider and other arachnids from the Devonian of New York, and reinterpretations of Devonian Ar- aneae. Paleontology 34:241-281. Shultz, J.W. 1990. Evolutionary morphology and phylogeny of Arachnida. Cladistics 6:1-38. Van der Hammen, L. 1989. An Introduction To Comparative Aracheology. SPB Academic Pub- lishing, The Hague. Weygoldt, R 1969. The Biology of Pseudoscorpi- ons. Harvard University Press, Cambridge, Mas- sachusetts. Wheeler, W.C. & C.Y. Hayashi. 1998. The phylog- eny of extant chelicerate orders. Cladistics 14: 173-192. Lorenzo Prendiei: Percy FitzPatrick Insti- tute, University of Cape Town, Rondebosch 7700, South Africa t ' P S' ' ^ m •< it t>s \l^ # ifr » Hr (ttf ^u V^ *• i'/‘^•4► 4Wi .v’» .fc*;'iy^,»j, ,i. ’ , . i^|...-o’*, u.r ■•, A ' fi..'i:> ’>fK ■ '■ , ,■ ,>'V'ii^. ■ .'■■,* ' •' '•■ '• i^iW*t ' .. ■ 1'’^.' ■•’’ . « Jwi'tj ■: '•N'H ' ►- r.‘J''l **. •^', iH" ' *'v' 'M, <» • ^*'« ^*5r; “i ^ ♦ I ! i ^ »f*- ^ * ■ j. ( .'. jrf:i,ui>Ht y,n ■ •*'{!tt4V'»'‘' sr^ JM: ;<. ..■ J>7,;-i;.j •»/■ /r»»J, *i| nrt>|>.nxy' ( i.'MJftl^f ’V’ y , tmi.w . • ’t’*. . . ' 1.."' ' ''K ■'.. Hi- . 1-: •;,»?. *.H. .' • .-.. ‘ . ,M>‘’ ■ *' ■ ■/■ - = -u • -I , -., , .. ' ■ ' ■ ^..ii *%■ ■ ■ 1 ■- '. . .»»»' . 'i. -iJ )■#»< i • ♦.. .'i ■ -'. '* •' A; •■ '.tfiaol.lJ' . M ■•i 'vTattfljjl ,., .* i— i->f 1 ;U.» l‘ v* ' ' V • ■ ' • ■A*; • '"t h.% < »»t ^-V:AV ^ ^dl# ■ ■ V.*. is^i- »• . :'' ■ ..7' Mt; : '' 'l '■ V- ;* y-t\ ,, •J'.. lU il'j!.t"fijjpi f MArv -;'r| ».*'•, liltl - ' . '.v •' i.tlUiT'' * ii '|.V^>■•‘‘ •' ■ - omr (\ • _ . '^*1 . u'hit „ 1 ,4 ». t • - I t*. f *( ^7r'l V»|^< ^ : '.. ,'l •>J'rr; :ip^ r* .“'- J'M ■j^’ * ui . * ■iV'' 1 » . /i, d^’- ‘' ■i/ y-1. ' ‘ ! .ryTi iti ’ r w 1/ J ;■ aisii^. >v, r U'.-; ii^fif^ i.‘' ^„il- .,- .-.J^-. . p.- , 4f'*, v. • M 1 fj ' . ' ^ -■ r ■ *,rit- ‘ rr . »i4 . ilt.l fy -1 • : 1 .iUiA, .,: t^. '^hjpri ^'L ■ fM if-WV' u’ jri *’• f Aii'-n; .. V.. I ■ .^v nS’' |»iia « . : : : >u cii^Tr .^-£, -IMIy- «|^(. > r, , 46^.t *i QV-iyw i:, ^ ‘-^• *t INSTRUCTIONS TO AUTHORS (revised September 2001) General: Manuscripts are accepted in English only. Authors whose primary language is not English may consult the editors for assistance in obtaining help with manuscript preparation. All manuscripts should be prepared in general accordance with the current edition of the Council of Biological Editors Style Manual unless instructed otherwise below. Authors are advised to consult a recent issue of the Journal of Arachnology for additional points of style. Manuscripts longer than three printed journal pages should be prepared as Feature Articles, shorter papers as Short Communications. Submission: Send four identical copies of the typed mate- rial together with copies of illustrations to the Managing Editor of the Journal of Arachnology. Paula E. Cushing, Managing Editor; Denver Museum of Nature and Science, Zoology Department, 2001 Colorado Blvd., Denver, CO 80205-5798 USA [Telephone: (303) 370-6442; FAX: (303) 331-6492; E-mail: PCushing@dmns.org] The Managing Editor will forward your manuscript to one of the Subject Editors for the review process. You will receive correspondence acknowledging the receipt of your manuscript from the responsible Subject Editor or the Managing Editor, with the manuscript number of your man- uscript. Please use this number in all correspondence regarding your manuscript. Correspondence relating to manuscripts should be directed to the Managing Editor or the appropii. Subject Editor. After the manuscript has been accepted, the author will be asked to submit the man- uscript on a PC computer disc in a widely-used word pro- cessing program. The file also should be saved as a text file. Indicate clearly on the computer disc the word processing program used. Voucher Specimens: Voucher specimens of species used in scientific research should be deposited in a recognized sci- entific institution. All type material must be deposited in a recognized collection/institution. FEATUPIE ARTICLES Title page. — The title page will include the complete name, address, and telephone number of the author with whom proofs and correspondence should be exchanged, a FAX number and electronic mail address if available, the title in capital letters, and each author’s name and address, and the running head (see below). Abstract. — The heading in bold and capital letters should be placed at the the beginning of the first paragraph set off by a period. A second abstract, in a language perti- nent to the nationality of the author(s) or geographic region(s) emphasized, may be included. Keywords. — Give 3-5 appropriate keywords following the abstract. Text. — Double-space text, tables, legends, etc. through- out. Three levels of heads are used. • The first level (METHODS, RESULTS, etc.) is typed in capitals and on a separate line. • The second level is bold, begins a paragraph with an indent and is separated from the text by a period and a dash. • The third level may or may not begin a paragraph but is italicized and separated from the text by a colon. Use only the metric system unless quoting text or referenc- ing collection data. All decimal fractions are indicated by the period (e.g., -0.123). Citation of references in the text. Cite only papers already published or in press. Include within parentheses the sur- name of the author followed by the date of publication. A comma separates multiple citations by the same author(s) and a semicolon separates citations by different authors, e.g., (Smith 1970), (Jones 1988; Smith 1993), (Smith 1986, 1987; Smith & Jones 1989; Jones et al. 1990). Include a let- ter of permission from any person ./ho is cited as providing unpublished data in the form of a personal communication. Citation of taxa in text: Please include the complete taxo- nomic citation for each arachnid taxon when it appears first in the paper. For Araneae, this taxonomic information can be found on-line at http://research.amnh.org/entomology/spi- ders/catalog81-87/INTR02.html. For example, Araneus diadematus Clerck 1757. Literature cited section. — Use the following style and include the full unabbreviated journal title. Lombardi, S.J. & D.L. Kaplan. 1990. The amino acid com- position of major ampullate gland silk (dragline) of Nephila clavipes (Araneae, Tetragnathidae). Journal of Arachnology 18:297-306. Krafft, B. 1982. The significance and complexity of com- niunication in spiders. Pp. 15-66, In Spider Commu- nications: Mechanisms and Ecological Significance. (P.N. Witt & J.S. Rovner, eds.). Princeton University Press, Princeton, New Jersey. Footnotes. — Footnotes are permitted only on the first printed page to indicate current address or other information concerning the author. All footnotes are placed together on a separate manuscript page. Tables and figures may not have footnotes. Running head. — The author’s surname(s) and an abbre- viated title should be typed all in capital letters and must not exceed 60 characters and spaces. The running head should be placed near the top of the title page. Taxonomic articles.^ — Consult a recent taxonomic article in the Journal of Arachnology for style or contact the Subject Editor for Systematics. Papers containing the origi- nal taxonomic description of the focal arachnid taxon should be given in the Literature Cited section. Tables. — Each table, with the legend above, should be placed on a separate manuscript page. Only horizontal lines (usually three) should be included. Tables may not have footnotes; instead, include all information in the legend. Make notations in the text margins to indicate the preferred location of tables in the printed text. Illustrations. — Original illustrations should not be sent until the article is accepted for publication. Address all questions concerning illustrations to the Editor of the Journal of Arachnology: Dan Mott, Editor-In-Chief Lincoln Land Community College, 5250 Shepherd Road, PO Box 19256, Springfield, IL 62794-9256 USA [Telephone (217) 786-2384; FAX: (217) 786-2251; E-mail: dan.mott@llcc.cc.il.us] All art work must be camera-ready - i.e., mounted and labeled - for reproduction. Figures should be arranged so that they fit (vertically and horizon- tally) the printed journal page, either one column or two columns, with a minimum of wasted space. When reduc- tions are to be made by the printer, pay particular attention to width of lines and size of lettering in line drawings. Multiple photos assembled on a single plate should be mounted with only a minimum of space separating them. In the case of multiple illustrations mounted together, each illustration must be numbered sequentially rather than given an alphabetic sequence. Written on the back should be the name(s) of author(s) and an indication of top edge. Indicate whether the illustration should be one column or two columns in width. The overall dimensions should be no more than 1 1 inches (28 cm) x 14 inches (36 cml. Larger drawings present greater difficulty in shipping and greater risks of damage for which the Journal of Arachnology assumes no responsibility. In manuscripts for review, photo- copies should be included, and should be reduced to the exact measurements that the author wants to appear in the final publication. Make notations in the text margins to indi- cate the preferred position of illustrations in the printed text. Color plates can be printed, but the author must assume the frill cost, currently about $600 per color plate. Legends for illustrations should be placed together on the same page(s) and separate from the illustrations. Each plate must have only one legend, as indicated below: Figures 1-4. — A-us x-us, male from Timbuktu: 1. Left leg; 2. Right chelicera; 3. Dorsal aspect of genitalia; 4. Ventral aspect of abdomen. Figures 27-34. — Right chelicerae of species of A-us from Timbuktu: 27, 29, 31, 33. Dorsal views; 28, 30, 32, 34. Prolateral views of moveable finger; 27, 28. A-us x-us, holo- type male; 33, 34. A-usy-us, male. Scale = 1.0 mm. Assemble manuscript for mailing. — ^Assemble the sep- arate sections or pages in the following sequence; title page, abstract, text, footnotes, tables with legends, figure legends, figures. Page charges, proofs and reprints. — There are no page charges, but the author will be charged for changes made in the proof pages. Reprints are available only from the Allen Press and should be ordered when the author receives the proof pages. Allen Press will not accept reprint orders after the paper is published. The Journal of Arachnology also is published by BioOne. Therefore, if your institution is a member of BioOne, copies of each article can be down- loaded as PDF files. SHORT COMMUNICATIONS Short Communications are usually limited in length to three journal pages, including tables and figures. They will be printed in a smaller (10 point) typeface. The format for these is less constrained than for feature articles: the text must still have a logical flow, but formal headings are omit- ted and other deviations from standard article format can be permitted when warranted by the material being covered. CONTENTS The Journal of Arachnology Volume 29 Feature Articles Number 3 Gross muscular anatomy of Limulus polyphemus (Xiphosura, Chelicerata) and its bearing on evolution in the Arachnida by Jeffrey W. Shultz . . 283 A new species of Diplocentrus (Scorpiones, Diplocentridae) from Texas by Scott A. Stockwell & Andrew S. Baldwin 304 Notes on the genus Scytodes (Araneae, Scytodidae) in Central and South America by Antonio D. Brescovit & Cristina A. Rheims 312 A review of the Chinese Psechridae (Araneae) by Xin-Ping Wang & Chang-Min Yin 330 A comparative study of the biology and karyotypes of two central European zodariid spiders (Araneae, Zodariidae) by Stano Pekar & Jifi Krai 345 Under the influence: webs and building behavior of Plesiometa argyra (Araneae, Tetragnathidae) when parasitized by Hymenoepimecis argyraphaga (Hymenoptera, Ichneumonidae) by William G. Eberhard 354 Life-cycles of four species of Pardosa (Araneae, Lycosidae) from the island of Newfoundland, Canada by J.R. Pickavance 367 Synonyms of Frontinella tibialis (Araneae, Linyphiidae) by G. Ibarra-Nunez, J.A. Garcia, M.E'. Jimenez & A. Mazariegos 378 Monoamines in the brain of tarantulas {Aphonopelma hentzi) (Araneae, Theraphosidae): differences associated with male agonistic interactions by Fred Punzo & Thomas Punzo ‘ 388 Habitat distribution and life history of species in the spider genera Theridion, Rugathodes and Wamba in the Great Smoky Mountains National Park (Araneae, Theridiidae) by Grant Jeffrey Stiles & Frederick A. Coyle 396 Evidence for kin- structured group founding and limited juvenile dispersal in the sub-social spider Stegodyphus lineatus (Araneae, Eresidae) by Jes Johannesen & Yael Lubin 413 Short Communications On the genus Eilica (Araneae, Gnaphosidae) from Argentina by Violeta Medan 423 Distinguishing the females of Trochosa terricola and Trochosa ruricola (Araneae, Lycosidae) from populations in Illinois, USA by Thomas R. Prentice 427 The unusual egg-rod of the spider Homalometa chiriqui (Araneae, Tetragnathidae) and other biological data by Guillermo Ibarra-Nunez 431 Book Reviews New Volumes on the ‘‘Minor” Arachnid Orders The Biology of Camel-Spiders {Arachnida, Solifugae). written by Fred Punzo. & Whip Spiders {Chelicerata: Amblypygi). Their Biology, Morphology and Systematics. written by Peter Weygoldt. reviewed by Lorenzo Prendini 434